1  

Discourse Analysis and

the New Testament

Approaches and Results

edited by

Stanley E. Porter &

Jeffrey T. Reed

Journal for the Study of the New Testament Supplement Series 170

Studies in New Testament Greek 4 Copyright © 1999 Sheffield Academic Press

Published by Sheffield Academic Press Ltd

Mansion House 19 Kingfield Road Sheffield S11 9AS

England British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library ISBN 1-85075-996-0

2  

CONTENTS Preface Abbreviations List of Contributors STANLEY E. PORTER AND JEFFREY T. REED Discourse Analysis and the New Testament: An Introduction

Part I THEORY AND METHOD IN DISCOURSE ANALYSIS

EUGENE A. NIDA The Role of Context in the Understanding of Discourse JEFFREY T. REED The Cohesiveness of Discourse: Towards a Model of Linguistic Criteria for Analyzing New Testament Discourse STANLEY E. PORTER Is Critical Discourse Analysis Critical? An Evaluation Using Philemon as a Test Case MATTHEW BROOK O’DONNELL The Use of Annotated Corpora for New Testament Discourse Analysis: A Survey of Current Practice and Future Prospects

Part II DISCOURSE ANALYSIS AND THE GOSPELS AND ACTS

STEPHANIE L. BLACK The Historic Present in Matthew: Beyond Speech Margins ROBERT E. LONGACRE A Top-Down, Template-Driven Narrative Analysis, Illustrated by Application to Mark’s Gospel ROBERT E. LONGACRE Mark 5:1–43: Generating the Complexity of a Narrative from its Most Basic Elements WOLFGANG SCHENK The Testamental Disciple-Instruction of the Markan Jesus (Mark 13): Its Levels of Communication and its Rhetorical Structures JONATHAN M. WATT Pronouns of Shame and Disgrace in Luke 22:63–64

3  

GUSTAVO MARTÍN-ASENSIO Participant Reference and Foregrounded Syntax in the Stephen Episode TODD KLUTZ Naked and Wounded: Foregrounding, Relevance and Situation in Acts 19:13–20

Part III DISCOURSE ANALYSIS AND THE PAULINE CORPUS

RICHARD J. ERICKSON The Damned and the Justified in Romans 5:12–21: An Analysis of Semantic Structure JOHANNES P. LOUW A Discourse Reading of Ephesians 1:3–14 STEPHEN H. LEVINSOHN Some Constraints on Discourse Development in the Pastoral Epistles ERNST R. WENDLAND ‘Let No One Disregard You!’ (Titus 2:15): Church Discipline and the Construction of Discourse in a Personal, ‘Pastoral’ Epistle

Part IV DISCOURSE ANALYSIS AND THE GENERAL EPISTLES

ANDRIES H. SNYMAN Hebrews 6:4–6: From a Semiotic Discourse Perspective BIRGER OLSSON First John: Discourse Analyses and Interpretations JOHN CALLOW Where Does 1 John 1 End? Index of References Index of Authors

 

 

 

 

 

 

4  

PREFACE

The editors wish to thank the contributors to this volume for making their work available to a primary reading public that approaches discourse analysis through the study of the New Testament. The process of gathering work from the four corners of the globe has been one that has had a number of intriguing challenges to it. However, now that the essays have been compiled, we believe that they make a fitting contribution to an area of academic endeavour that we all believe is worth pursuing actively and enthusiastically.

The contributors, and especially the editors, wish to thank their respective academic and other institutions for the generous support that they have received, both materially and in other less tangible though no less real ways. This support has allowed and encouraged each of us in various ways to be able to carry on our research projects despite the constraints and urgent demands of numerous other tasks. Even though these tasks too are important ones, the advancement of research in the area of discourse analysis is now showing benefits from which several other subjects can profit, including biblical studies, translation theory and other areas of linguistics.  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5  

ABBREVIATIONS

AB Anchor Bible A1CS The Book of Acts in its First Century Setting AnBib Analecta biblica ANRW Hildegard Temporini and Wolfgang Haase (eds.), Aufstieg und Niedergang der

römischen Welt: Geschichte und Kultur Roms im Spiegel der neueren Forschung (Berlin: W. de Gruyter, 1972–)

ATANT Abhandlungen zur Theologie des Alten und Neuen Testaments AV Authorized Version BDF Friedrich Blass, A. Debrunner and Robert W. Funk, A Greek Grammar of the

New Testament and Other Early Christian Literature (Cambridge: Cambridge University Press, 1961)

BETL Bibliotheca ephemeridum theologicarum lovaniensium Bib Biblica BibTrans The Bible Translator BIS Biblical Interpretation Series BJRL Bulletin of the John Rylands University Library of Manchester BTB Biblical Theology Bulletin BWANT Beiträge zur Wissenschaft vom Alten und Neuen Testament BZ Biblische Zeitschrift CB Cultura bíblica CBQ Catholic Biblical Quarterly ConBNT Coniectanea biblica, New Testament ConBOT Coniectanea biblica, Old Testament CTL Cambridge Textbooks in Linguistics EKKNT Evangelisch-katholischer Kommentar zum Neuen Testament EstBíb Estudios bíblicos ETEL Edinburgh Textbooks in Empirical Linguistics ETL Ephemerides theologicae lovanienses EvT Evangelische Theologie FGrHist Die Fragmente der griechischen Historiker FN Filología neotestamentaria FRLANT Forschungen zur Religion und Literatur des Alten und Neuen Testaments FSÖTh Forschungen zur systematischen und ökumenischen Theologie FTL Forum theologiae Linguisticae FV Foi et Vie GTA Göttinger theologische Arbeiten HNT Handbuch zum Neuen Testament HTKNT Herders theologischen Kommentar zum Neuen Testament HTS Harvard Theological Studies ICC International Critical Commentary Int Interpretation JAAR Journal of the American Academy of Religion JBL Journal of Biblical Literature JETS Journal of the Evangelical Theological Society JOTT Journal of Translation and Textlinguistics JSHRZ Jüdische Schriften aus hellenistisch-römischer Zeit JSNT Journal for the Study of the New Testament JSNTSup Journal for the Study of the New Testament, Supplement Series JSOTSup Journal for the Study of the Old Testament, Supplement Series

6  

JTS Journal of Theological Studies KantSt Kant-Studien KBANT Kommentare und Beiträge zum Alten und Neuen Testament KEK H.A.W. Meyer, Kritisch-exegetischer Kommentar über das Neue Testament KNT Kommentar zum Neuen Testament LB Linguistica biblica LD Lectio divina LEC Library of Early Christianity NCB New Century Bible Neot Neotestamentica NIBC Nag Hammadi Studies NICNT New International Commentary on the New Testament NIGTC The New International Greek Testament Commentary NIV New International Version NovT Novum Testamentum NRSV New Revised Standard Version NTD Das Neue Testament Deutsch NTOA Novum Testamentum et orbis antiquus NTS New Testament Studies NTTS New Testament Tools and Studies ÖBS Österreichische biblische Studie RB Revue biblique RevQ Revue de Qumran RILP Roehampton Institute London Papers RR Religion and Reason RSV Revised Standard Version SBA Stuttgarter biblische Arbeiten SBG Studies in Biblical Greek SBL Sources for Biblical Study SBLDS SBL Dissertation Series SBLMS SBL Monograph Series SBLRBS SBL Resources for Biblical Study SBLSBS SBL Sources for Biblical Study SBLTT SBL Texts and Translations SBS Stuttgarter Bibelstudien SE Studia Evangelica I, II, III (= TU 73 [1959], 87 [1964], 88 [1964], etc.) SEÅ Svensk exegetisk årsbok SNTG Studies in New Testament Greek SNTS Society for New Testament Studies SNTSMS Society for New Testament Studies Monograph Series SO Symbolae Osloenses ST Studia theologica STK Svensk teologisk kvartalskrift START Selected Technical Articles Related to Translation StEv Studia Evangelica SPB Studia Post-Biblica

                                                            SBL Sources for Biblical Study 

TU Terminating Utterance 

7  

StudTheol Studia Theologica TBl Theologische Blätter TEV Today’s English Version THKNT Theologische Handkommentar zum Neuen Testament TLZ Theologische Literaturzeitung TRE Theologische Realenzyklopädie TSFBul Theological Students Fellowship Bulletin TTK Tidskrift for teologie og kirke TynBul Tyndale Bulletin TZ Theologische Zeitschrift UrbTb Urban Taschenbuch UTB Uni-Taschenbücher VF Verkündigung und Forschung VT Vetus Testamentum WBC Word Biblical Commentary WTJ Westminster Theological Journal WUNT Wissenschaftliche Untersuchungen zum Neuen Testament ZAW Zeitschrift für die alttestamentliche Wissenschaft ZKT Zeitschrift für katholische Theologie ZNW Zeitschrift für die neutestamentliche Wissenschaft ZTK Zeitschrift für Theologie und Kirche  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8  

LIST OF CONTRIBUTORS

Stephanie L. Black, Roehampton Institute London, England John Callow, Summer Institute of Linguistics, England Richard J. Erickson, Fuller Theological Seminary, Seattle, WA, USA Todd Klutz, University of Manchester, England Stephen H. Levinsohn, Summer Institute of Linguistics, England Robert E. Longacre, University of Texas at Arlington and the Summer Institute of Linguistics, Texas Johannes P. Louw, University of Pretoria, South Africa Gustavo Martín-Asensio, Roehampton Institute London, England Eugene A. Nida, American Bible Society, New York Matthew Brook O’Donnell, Roehampton Institute London, England Birger Olsson, Lund University, Sweden Stanley E. Porter, Roehampton Institute London, England Jeffrey T. Reed, Roehampton Institute London, England Wolfgang Schenk, Linguistic Institute for Theology, Saarbrücken, Germany Andries H. Snyman, University of the Orange Free State, Bloemfontein, South Africa Jonathan M. Watt, Geneva College and the Reformed Presbyterian Theological Seminary, Pittsburgh, Pennsylvania Ernst R. Wendland, Bible Society of Zambia, Lusaka  

 

 

 

 

 

 

 

9  

DISCOURSE ANALYSIS AND THE NEW TESTAMENT: AN INTRODUCTION Stanley E. Porter and Jeffrey T. Reed

We are increasingly living and working in a multi-disciplinary academic world. This has to a large extent always been true of academic biblical studies. The reason for this is that biblical studies has compelled its practitioners to combine a number of academic areas in the pursuit of analysis of a compact and defined corpus of texts. These academic areas have often involved history, theology and the study of languages. In more recent times, this arena has expanded to include methods such as literary criticism in its various forms, social-scientific criticism and canonical criticism. What has always remained as a viable area of study—although pursued in various ways and to varying degrees—has been the language of the New Testament. Study in this area has seen periods of increased activity and periods of decreased activity, almost to the point of morbundity. However, since the texts of the New Testament are written in a language, and an ancient one requiring specialized tools for study, pursuit of the study of the language of the New Testament has never been too far away from the central concerns of at least a few.

In recent years, in the field of linguistcs, discourse analysis (or textlinguistics) has been an actively developing field of study. It too is interdisciplinary in nature, in the sense that it brings various of the sub-disciplinary areas of modern linguistics to bear on the study of discourse. As a result there are a number of different models of discourse analysis that have established themselves as productive means of reading texts. From the start, discourse analysis by its very nature was concerned with seeing the results of its analysis. In a very real sense, discourse analysis cannot exist without attention to discourse, which is usually defined in terms of instantiations of real language use, whether it is in spoken or written form.

As a result of these two factors—the increasingly multi-disciplinary nature of New Testament studies and the interdisciplinary nature of discourse analysis—it is not surprising that discourse analysis has now made sizable inroads into the study of the New Testament. In recent years, there have been a number of volumes that have made contributions in this area. For example, one volume published a collection of essays that presented forms of discourse analysis, and then had responses by both supporters and skeptics of the method.1 Another volume had a collection of essays mostly representing one of the methods of discourse analysis.2 These have been valuable contributions, and have helped establish the place of a volume such as this.

This volume of essays in discourse analysis does not follow any given or particular method or model of discourse analysis. As a result, the four major methods of discourse analysis in New Testament studies—the South African, the Summer Institute of Linguistics, the Continental with its sub-areas, and the Hallidayan or functional approaches3—are all ably represented here by at least one example of each. However, they are not categorized in this way, and there are developments from these models represented as well. Instead, we have organized the essays with emphasis upon their practical results. In other words, we wish to illustrate how discourse analysis, no matter what model is being used, is often at its best when

                                                            1 S.E. Porter and D.A. Carson (eds.), Discourse Analysis and Other Topics in Biblical Greek (JSNTSup, 

113; SNTG, 2; Sheffield: Sheffield Academic Press, 1995). 

2 D.A. Black with K. Barnwell and S. Levinsohn (eds.), Linguistics and New Testament Interpretation: 

Essays in Discourse Analysis (Nashville: Broadman, 1992). 

3 See S.E. Porter, ‘Discourse Analysis and New Testament Studies: An Introductory Survey’, in Porter 

and Carson (eds.), Discourse Analysis, pp. 14–35, esp. 24–34. 

10  

it is reading texts rather than theorizing about its method. Nevertheless, method is important, so the volume begins with its first part devoted to issues of theory and method. Eugene Nida begins with a discussion of the role of context in understanding discourse. For most, if not all, methods of discourse analysis, context is an important factor, and one must have an understanding of it in order to understand discourse itself. This essay is followed by Jeffrey Reed’s attempt to define a model of cohesiveness. One of the almost axiomatic assumptions of discourse analysis has been the cohesiveness of discourse. In this essay, Reed attempts to come to grips with defining this sometimes elusive concept. Looking at recent developments in functional discourse analysis, Stanley Porter offers a critical analysis of Critical Discourse Analysis, applying it to Philemon. He finds that there are some positive contributions of this growing area of research, but that due caution must be exercised as well. The section concludes with discussion by Matthew O’Donnell of how annotated corpora can and should figure into future discourse analysis. Corpus linguistics is less a method than a means of amassing data, but how these data are amassed makes a significant difference in how the data can be applied.

In the next three parts of this volume, the various bodies of texts of the New Testament are examined. Part II is concerned with the Gospels and Acts. Stephanie Black examines the use of the historic present in Matthew’s Gospel where it is not used to introduce direct speech. Two passages, Jesus’ temptation and his prayer in Gethsemane, are seen to have significant concentrations of historic presents, which she analyzes. Robert Longacre follows with two essays on Mark’s Gospel. The first uses a top-down model of narrative analysis to examine the entire Gospel in terms of its sequence of episodes. The second focuses specifically upon a section in Mark 5. Wolfgang Schenk discusses the structure of Mark 13, usefully integrating discourse analysis with traditional discussion of the nature of this problematic discourse. In a slight turn from what is said to what is not said, Jonathan Watt examines how in Luke 22 reference is and is not made to Jesus, and its discourse implications. Gustavo Martín-Asensio takes the Stephen Episode in Acts 7 and, with the aid of Halliday’s transitivity network, examines the participant reference and its implications for foregrounding. Using a similar method, Todd Klutz takes a passage in Acts 19 to explore not only foregrounding, but relevance and the concept of situation.

Part III examines the Pauline corpus. Several of the most difficult passages in the Pauline corpus are tackled in this section. The section begins with Richard Erickson’s essay on Rom. 5:12–21, and its semantic structure according to the structure of its nuclear sentences. Johannes Louw presents a colon analysis of Eph. 1:3–14, attempting to outline the flow of its thought. Stephen Levinsohn discusses the Pastoral Epistles from the standpoint of relevance theory, paying particular attention to connective words. Ernst Wendland concludes with an essay on the neglected book of Titus. His approach is to examine its parenetic structure, with implications for a method that appreciates rhetorical structure as well. Part IV concludes with the General Epistles. From a semiotic perspective, Andries Snyman looks at the notoriously difficult Heb. 6:4–6, analyzing these verses in their immediate context, as well as in the context of the book. The two final essays are on 1 John. Birger Olsson provides a useful survey of past interpretation, showing how this letter has been more of a hotbed of debate than many would imagine. John Callow concludes with discussion from a discourse perspective of where 1 John 1 actually ends, comparing the results of and exploring the implications for translation.  

 

 

11  

Part I THEORY AND METHOD IN DISCOURSE ANALYSIS

THE ROLE OF CONTEXT IN THE UNDERSTANDING OF DISCOURSE Eugene A. Nida

In any symbolic system such as language, the role of the context is maximized and the role of any one focal element is minimized (Joos 1972). This fact about language applies to all levels, from sounds to discourse units, but it can be most easily recognized in understanding the meanings of words. Without a context, lexical units have only a potentiality to occur in various contexts, but in combination with contexts, words have meaning. For example, in order to understand the meaning of run, it is essential to put the word into some type of context, for example, the boy was running, the snake ran across the lawn, the water is running, his nose is running, his heart is running, the play ran for three weeks, he is running for election, he hit a run, he lives up the run. In each of these expressions it is the context that contributes the major semantic elements of the combined conceptual meaning, because in each instance the semantic role of run is the least that it contributes to the meaning of the combinations, for example, rapid movement in space, flow (of a mass), characteristic function, extension, competition, a unit activity of movement in space, and a narrow valley.

The verbal contexts for any focal lexical element are of two principal types: syntagmatic and paradigmatic. The syntagmatic contexts are usually those that occur in the same stretch of speech or writing, and the paradigmatic contexts are those involving possible substitutions of terms in the same contextual frames, for example, he ran to work and he walked to work. It is therefore possible to define certain distinctions of meaning of run or of walk on the basis of contrasting practical contexts.

The meanings of words are learned primarily in syntagmatic contexts, but the practical contexts of usage also contribute to meaning. For example, the meaning of the word stock differs appreciably as the result of situational contexts. When used by a grocery clerk, the term stock is likely to refer to merchandise, but when used by a broker the reference is likely to be to equities, and when used by a farmer talking about his farming enterprise, the term likely refers to cattle. The same word may also have different meanings depending on who is

speaking. For example, non-Christians used the Greek term tapeinos to refer to irresponsible, lower-class, worthless people, but Christians radically altered the meaning of this term when used among themselves and employed it in the sense of ‘humble’.

Fully 95 per cent of the meanings of words in one’s mother tongue are learned by means of syntagmatic and practical contexts. But the real existence of such meanings is not to be found in dictionaries but in people’s hands, as series of synapses in the networks of the brain that can be quickly activated. Lexicons do little more than record types of contexts in which such meanings are likely to occur.

The importance of context in language is being recognized in an increasing number of dictionaries, for example, the Longman Language Activator: The World’s First Production Dictionary (Summers 1993) and several dictionaries published primarily for the use of persons working in the European Community. In these latter dictionaries fully 85 per cent of the lexical entries consist of phrases rather than individual words, because it is at the level of the phrase that the conceptual meaning of lexemes is most relevantly defined. This means that lexicographical studies must move up from the level of atomic distinctions between isolated words (and even subatomic differences of componential features) to the molecular level of combinations of words.

12  

This same principle of contextual conditioning of meaning also applies to discourse, and the primary features of such contextual conditioning can perhaps be best understood in analyzing the story of The Father and Two Sons (Lk. 15:11–32).

1. The Father and Two Sons This remarkable parable, regarded by some literary critics as the best short story ever written, is a fascinating, open-ended account of a younger son who was lost and then found because he ‘came to himself, and of an older son who never realized he was lost.

The younger son asked for a division of the inheritance, and he soon turned his holdings into cash and went to a far country, where he wasted his money in reckless living. And after spending all he had, he was forced by a famine to seek work as a herder of pigs. He became so desperate that he even wanted to eat what the pigs were eating. After realizing that the servants at home had plenty to eat while he was starving to death, he decided to return to his father and to plead to become a hired worker. But his father threw his arms around his son and ordered a servant to bring the son a new garment, a ring for his finger and shoes for his feet. In addition, the prize calf was to be killed for a festival with music and dancing. But when the older brother returned from the field and learned what had happened, he accused his father of favoritism and refused to join in the celebration.

The attitude of the older son is evident in the way in which he speaks to his father about his younger brother as ‘your son’ rather than ‘my brother’, and although the storyline describes the younger son as living ‘recklessly’, the older son accuses his brother of ‘wasting’ his money on ‘prostitutes’. The older brother is quick to emphasize his own self-righteousness by claiming never to have disobeyed his father’s commands, and he shows his stinginess in complaining that he was never given a young goat to have a party with his friends, when in reality, not only all the goats but everything else belonged to the older son.

In the story of The Father and Two Sons there are a number of important cultural elements that would have been immediately recognized by any first-century Judean audience who heard or read this account. In the first place, they would have known that the younger son would have received only one-third of the value of the estate, because according to Jewish tradition the firstborn son received twice as much as any other sibling. Furthermore, a Jew could not have suffered more loss of self-esteem than having to associate with pigs, regarded as the most despicable of the unclean animals. The ring was of course a symbol of the younger son being recognized as a full member of the family. The clothes gave him social status, and the shoes indicated that he would not be a servant, since in New Testament times slaves and servants in the Mediterranean world normally went barefooted. Furthermore, the killing of a prize calf was very significant in New Testament times because meat was a rarity and therefore a more meaningful element in the celebration.

The setting of this parable is also very meaningful because there were two groups listening to Jesus: the outcasts who listened to him gladly and the Pharisees who were anxious to catch Jesus saying anything that might be used to bring accusations against him. Furthermore, the Pharisees not only opposed Jesus but they were completely contemptuous of the outcasts, whose violations of religious regulations had resulted in their being excluded from the synagogues.

The third type of context consists of the two immediately preceding parables concerning The Lost Sheep and The Lost Coin. Both of these parables have essentially the same theme of ‘lostness’ as in the parable of The Father and Two Sons. But the two preceding parables are unusual in that the meaning of each is explained in terms of joy and celebration in heaven over one repentant sinner. These verbal contexts certainly add a theological dimension to the third parable in ch. 15.

13  

In view of the various contextual features, the story of The Father and Two Sons can best be understood on three levels: first, as a triangular relation of the father at the apex of an equilateral triangle with the younger son to the left and the older son to the right, based on their order of relevance in the story. The relations between the three sets of persons may be described as basically antagonistic. This seems obvious in the relation between the younger brother and the older brother. And as the story develops, it seems clear that there must have been considerable tension between the older brother and his father. At first the younger son is alienated from his father and decides to depart, but later he returns.

The second interpretive level of this story represents the relation between the three sets of participants in the setting of the parable: Jesus, the outcasts, and the Pharisees. By analogy Jesus would also be the apex character, with the outcasts in the position of the younger son and the Pharisees in the position of the older son. The repentant outcasts were completely reconciled to Jesus, but they would undoubtedly have been angry at the Pharisees as the result of the Pharisees having despised the outcasts. The relation between Jesus and the Pharisees was also antagonistic in view of the repeated denunciations by Jesus of the hypocrisy of the Pharisees and their constant opposition to the teaching of Jesus.

In view of the verbal context of the immediately preceding two parables, there is evidently a third level of relevance to the parable of The Father and Two Sons, namely the theological dimension in which God assumes the same apical position of Jesus, with repentant sinners to the left and self-righteous persons to the right. The relation of God to repentant persons is the same as Jesus’ relation to the outcasts, namely full acceptance and in fact the violation of all kinds of socioreligious taboos in order to reach out to the outcasts.

In the same way that the Pharisees opposed Jesus, so self-righteous persons reject the grace of God and are never reconciled to his forgiveness of sinners, even as the older brother never joined in the celebration. In fact, the incompatibility between people’s self-righteousness and God’s grace is a focal element in this parable.

These three levels of relevance in the parable of The Father and Two Sons are roughly analogous to the three levels of firstness, secondness and thirdness in the semiotic system of Peirce (1934).

2. 1 and 2 Chronicles

The books of 1 and 2 Chronicles are very different from most other books of the Bible. First, there are nine chapters of genealogies and lists of persons returning to the promised land from Babylonia, and what follows seems to be only a retelling of certain events recorded in 1 and 2 Samuel and 1 and 2 Kings. An initial reading of 1 and 2 Chronicles almost inevitably raises a number of questions in the minds of many persons. Why the repetitions of the same events? And why the differences between the accounts in 1 and 2 Chronicles and the accounts in 1 Samuel through 2 Kings? The answers may be found in part by treating 1 Samuel through 2 Kings as the verbal context of 1 and 2 Chronicles, since much of what is in 1 and 2 Chronicles presupposes the earlier four books and clearly draws on them for crucial information.

In addition to these verbal contexts there are also historical and cultural contexts, but it is not always easy to separate the two, nor is it really necessary. Factors that are primarily historical involve (1) the time at which 1 and 2 Chronicles were written (they consist of only one discourse, but are divided into two books because of the normal size of scrolls), (2) the author (considered by many to be the same person who wrote the books of Ezra and Nehemiah), and (3) the need to define precisely those persons who were to participate in rebuilding the Temple and the city of Jerusalem as the capital of a reborn nation. The cultural contexts would include the revived sense of nationhood (as over against the Samaritans who had remained in the land) and the cultural identity as people called by God to be his people.

14  

The meaning of 1 and 2 Chronicles is best understood in terms of major and minor contrasts between these two books and their verbal contexts. The first nine chapters, consisting of genealogies (from the descendants of Adam to King Saul’s family) and lists of descendants of the various tribal patriarchs as well as important persons returning to the promised land, all point to the crucial importance of indicating precisely who are the true people of God, especially in view of the fact that the returning exiles refused to accept offers of help from those poorer people who had been left in the land and who had evidently married local Gentile persons.

In view of the emphasis on Jerusalem and the land of Judah, it is not strange that the descendants of Judah head the list of the various tribal groups. Furthermore, the descendants of King David receive special mention. This is fully in keeping with the emphasis on the Temple and the ritual ceremonies that occupy a seemingly disproportionate amount of space. Second Chronicles continues the history of Judah as though the kingdom of the north did not exist.

Three important episodes mentioned in 1 and 2 Samuel are not recorded in 1 and 2 Chronicles, namely the story of David and Goliath, David’s affair with Bathsheba, and the rebellion of Absalom. In 1 Chron. 20:5 Elhanan is credited with killing Lahmi, the brother of Goliath. This is evidently an attempt to explain a statement in 2 Sam. 21:19 in which Elhanan is listed as the one who killed Goliath. The omission of the story of David and Goliath is understandable since the account in 1 Samuel 17 seems seriously misplaced.

The omission of the story of David and Bathsheba and the subsequent killing of Uriah is also clearly understandable in the history of 1 and 2 Chronicles, in which David is portrayed as the true hero of the Jewish people, not only because of his prowess in war but also because of his organization of Temple worship.

Mention of the revolt of Absalom would have cast an unfavorable light on David, who was highly successful in war and in taking advantage of a period of weakness in the other two great powers in the region, namely Assyria and Egypt. But David was not a successful family man or father.

There are, however, some important differences between 1 and 2 Chronicles and their verbal contexts. For example, in 2 Sam. 24:1 it is the Lord who causes David to count the people of Israel, but in 1 Chron. 21:1 it is Satan who urges David to count the people. Such a change is not strange in view of the experience of the Jewish people in exile in Babylonia.

There are also some significant differences in the numbers of people involved in certain events, and in such instances the number of persons is usually greater in the books of 1 and 2 Chronicles, something that is typical of secondary sources. For example, in 2 Sam. 24:9 the total number of fighting men in Israel and Judah is given as 1,300,000, but in 1 Chron. 21:5 the total number is given as 1,570,000.

What is important about the accounts in 1 and 2 Chronicles is the matter of distinctiveness of emphasis between these two books and their remote verbal contexts. The rebirth of the nation, its center of worship and its renewed sense of identity and nationhood provide important insights into the turbulent subsequent history of the Jewish people during the times of the Maccabees, the destruction of Jerusalem, the mass suicide at Masada and the final rebellion under Bar Kokhba.

3. The Gospel of John

The significance of the Gospel of John can be better understood by regarding the context of this Gospel as the synoptic tradition in the Gospels of Matthew, Mark and Luke. The writing of the Gospel of John evidently took place some 30 to 40 years after the other three Gospels were written. This represents at least one full adult generation later. Accordingly, it is not strange that the Gospel of John seems to be a type of commentary on the synoptic accounts.

15  

As such it may be recognized as what cultural anthropologists would call a secondary source, because it interprets much of the meaning of the primary documents. For example, it considers the miracles as essentially symbols of the power and nature of Jesus.

The growing conflicts between Christians and Jews toward the latter part of the first century may explain the number of recorded controversies that Jesus had with the Jewish leaders in Jerusalem. Moreover, the emphasis in Mark about Jesus’ insistence that his disciples not reveal his true nature is radically altered by the Johannine accounts of Jesus arguing openly with Jewish leaders in Jerusalem about his relation to the Father. Even the placement of the cleansing of the Temple at the beginning of the Gospel would suggest that the Gospel of John places the conflict with the Jewish leaders as something integral to the entire ministry of Jesus and not merely an inevitable result of Jesus’ opposition to ‘the hypocrisy’ of the Pharisees. Furthermore, the numerous references to the leaders of the Jews as simply ‘the Jews’ has encouraged many persons to interpret the Gospel as being strongly anti-Semitic.

Perhaps the strangest omission in the Gospel of John is that there is nothing about the joint partaking of the bread and the wine as symbols of the atonement made possible by Jesus’ death. Instead of this ceremony there is the washing of the disciples’ feet in John 13, and the conversation during the Last Supper is extended for three chapters and is concluded in the prayer of ch. 17. Also, in contrast with the Lukan tradition, the Holy Spirit is ‘breathed on’ the disciples in Jn 20:22.

The Gospel of John clearly does not attempt to repeat the synoptic tradition but tries to interpret its significance in the light of other cultural contexts. Since, however, we lack information about the context of the historical setting, including the author or authors, the circumstances that prompted the writing, and the distinctive ideas of those concerned with this symbolic interpretation of Jesus’ life and death, we can only suggest possibilities. Without a reasonable amount of valid data, such speculations are more unsettling than meaningful. Perhaps one of the important aspects of any study of the Gospel of John is the recognition that without the historical context, much of the distinctive significance of the Gospel is lost.

Understanding the rich implications of the parable of The Father and Two Sons is possible in view of the immediate cultural, historical and verbal contexts, but in the case of 1 and 2 Chronicles the temporal distance of the verbal contexts and the uncertain relevance of historical and cultural contexts make the interpretation of 1 and 2 Chronicles more doubtful. But in the case of the Gospel of John, questions about so many of the contextual features weaken greatly any attempt to provide definitive interpretations or explanations of differences.

REFERENCES

Joos, Martin 1972 ‘Semantic Axiom Number One’, Language 48: 257–65. Peirce, Charles 1934 Collected Papers (Cambridge, MA: Harvard University Press). Summers, Della (ed.), 1993 Longman Language Activator: The World’s First Production Dictionary (Harlow: Longman).  

 

 

 

16  

THE COHESIVENESS OF DISCOURSE: TOWARDS A MODEL OF LINGUISTIC CRITERIA FOR

ANALYZING NEW TESTAMENT DISCOURSE Jeffrey T. Reed

Data without generalization is just gossip.

Robert Pirsig1 The analysis and understanding of ‘cohesiveness’ has been a central concern of discourse analysts ever since modern linguists turned their concerted attention to the study of complete discourse rather than isolated sentences. Despite new facets of discourse being investigated today, ‘cohesiveness’ still remains an area of vigorous study. How is it that speakers go about forming texts into a complete unit? How do they combine unrelated words and sentences into a meaningful whole? Why are some texts considered more coherent than others? Discourse analysts repeatedly seek answers to such questions, attempting to identify how a given language is used to create cohesive communication. W. Labov describes the task similarly: ‘The fundamental problem of discourse analysis is to show how one utterance follows another in a rational, rule-governed manner—in other words, how we understand coherent discourse.’2

Issues of cohesiveness, especially in the past century, have been a central concern of New Testament scholars as well. Source, form and redaction critics have investigated apparent seams in the Gospel accounts, demonstrating various types of subgenres which have been worked together into a complete document. For them, issues of cohesiveness are directly relevant to the composition of early Christian traditions. Other scholars have claimed to have found seams in the Pauline letters supposedly revealing the hands of later redactors who have patched together originally separate letters into their canonical form. For them, the apparent lack of cohesiveness has proved crucial for reconstructing original Pauline texts and their historical situations. Surprisingly, however, there has been little discussion of New Testament cohesiveness from a modern linguistic perspective nor are there any agreed upon linguistic criteria which can be appealed to when discussing matters of discourse cohesiveness in the New Testament. This is not due to a lack of study in secular linguistics. Several theories of discourse cohesiveness exist which are applicable to issues of New Testament cohesiveness. And the increasing interest by biblical scholars in discourse analysis should remedy this lack of theoretical discussion with regard to cohesiveness.

The present study seeks to sketch a model of cohesiveness based upon koine Greek which may prove beneficial in discussions of New Testament texts. It owes much to the systemic-functional theories of M.A.K. Halliday, as well as others who have played a key role in the development of discourse theories of cohesiveness. After briefly introducing the notion of linguistic cohesiveness, the study discusses two ways in which koine Greek (and other languages) is employed to create discourse cohesiveness.

1. The Notion of Cohesiveness

At a very basic level, linguistic cohesiveness3 refers to the means by which an immediate linguistic context meaningfully relates to a preceding context and/or a context of situation (i.e.

                                                            1 R. Pirsig, Lila: An Inquiry into Morals (New York: Bantam Books, 1991), p. 55. 

2 W. Labov, Sociolinguistic Patterns (Philadelphia: University of Pennsylvania Press, 1972), p. 252. 

3 I am primarily concerned with the way in which language is employed to create cohesiveness; thus, I 

prefer the phrase ‘linguistic cohesiveness’ in distinction to simply ‘cohesiveness’, since the later 

17  

meaningful relationships between text, co-text and context). Linguistic cohesiveness provides speakers with the means to produce a ‘message’ (i.e. theme) from individual and sometimes unrelated words and phrases. This is done by making explicit ‘the external relationship between one clause or clause complex and another, and … in a way which is not dependent on grammatical structure’.4 Halliday is not speaking only about how clauses are grammatically linked paratactically and hypotactically (i.e. grammatical structure); more than that, he is asserting that discourse gets its cohesive quality by means of semantic relations involving ‘elements of any extent, both smaller and larger than clauses, from single words to lengthy passages of text … [which] may hold across gaps of any extent’.5 His focus is clearly on the functional relationships between the linguistic forms of a discourse and how such relationships create coherence for the reader. He labels this characteristic of discourse COHESION.

In a more comprehensive discussion, M.A.K. Halliday and R. Hasan subdivide cohesion (what they call ‘textual meanings’) into two parts: (1) semantic and grammatical symmetry within a text and (2) thematic structure.6 These two aspects of textual meaning, it is argued, make a text a ‘text’. That there is a relationship both semantically and grammatically between the various parts of a given text, and that there is some thematic (prominent) element which flows through it, allow an audience to recognize it as a cohesive text rather than as a jumble of unrelated words and sentences.7 On the one hand, we expect discourse to be cohesive (recall Grice’s ‘cooperative’ principle), that is, its various linguistic elements should interrelate in a meaningful whole. Furthermore, certain elements must distinguish themselves as thematic (or prominent), that is, each discourse should be about something in particular, not everything in general.8 The present study focuses on the former type of cohesiveness, namely semantic and grammatical symmetry.

In any balanced theory of linguistic cohesiveness, the cohesiveness of a given text should be viewed as a continuum. At one pole of the continuum are texts with a high degree of unity and cohesiveness. At the opposite pole are texts which can be quickly recognized as a jumble of words and sentences with little ‘textuality’—a prime example of this that has received much study is schizophrenic discourse. Although a text might be elegantly unified or grossly fragmented, most texts lie somewhere between these two poles—neither altogether cohesive

                                                                                                                                                                                          could, strictly speaking, be a phenomenon created by the listener’s pragmatic inference (with 

perhaps very little of it based on the actual words of the text). We see this in studies where readers 

are presented with varying degrees of cohesive texts but told to interpret them with the premise that 

they make perfect sense; most readers will try to make sense of even the most incohesive texts, 

primarily because of the premise that communication is to be relevant (see D. Sperber and D. Wilson, 

Relevance: Communication and Cognition [Oxford: Basil Blackwell, 1986]). 

4 M.A.K. Halliday, An Introduction to Functional Grammar (London: Edward Arnold, 1985), p. 287. 

5 Halliday, Functional Grammar, p. 288. 

6 M.A.K. Halliday and R. Hasan, Cohesion in English (London: Longman, 1976). 

7 This is not to say that texts with a lesser degree of cohesion will not be interpreted by readers, but 

that cohesive texts reduce the interpretive choices and thus decision‐making labor of the reader. 

8 Of course, exceptions to such tendencies do exist; but they may be explained in terms of lack of 

cohesion and information flow—some schizophrenic discourse, for example, is an exception which 

proves the rule. 

 

nor altomaking incohesconglom

Tota

typicallythat do of the ‘about thsignificaarguing demonsgenre). that theliterary perfectltext in t

So, interprethe relatand the primarilexamplelinguistiform’s textual wamong INTERPR

PRESUPP

ogether incoclaims abo

ive on thismerate, mix

ally incohey, a discournot follow usual’ kindhe coherencant burden that the

strate this iThe fact th

ere is little integrity oy cohesive

the New Testhe questio

etation of cotionship betrelationshiply a mattere, knowledgic form on language. Iworld to ththe lingui

RETATION oPOSES the ot

ohesive. Wout textual s cline suged.

esive texts rse is cohesthe pattern

d (i.e. it follce of a par

of proof itext had a

in relation at cohesivetrue conse

of certain Pin the eyes

stament. on remains: ohesive or itween a texp between ar of historicge of Paul the page to

In this senshe extra-textstic items

of some elemther, in the

With this in cohesivene

ggest respe

are the exsive of an ‘us of other dlows shared

rticular Pauif arguing fan unusual to other coness is a sc

ensus amonauline lettes of another

What propincohesive

xt and its exa text and itscal inquiry and the soc

o an underse, the intertual world. of a text

ment in thesense that i

mind, the ess in the Nectively uni

 

xtreme exceunusual’ kindiscourse frd rules of l

uline text, fofor the tota

cohesive ontemporar

calar phenomng New Teers. What isr. We are h

perties of latexts? To e

xternal worls inner worwhen analy

cial world(sstanding of rpreter is crThe latter ritself. Thi

e discourse it cannot be

following New Testamity, singula

eptions in hnd (i.e. it corom the samanguage us

for exampleal incohesivquality, th

ry discoursemenon may

estament schs incohesivehard-pressed

anguage conexplain this ld (i.e. contld (i.e. co-teyzing the Ns) in which a concept i

reating a corelationshipis type of is dependeneffectively

scale mighment. The tar, undivide

human comontains noti

me languagee). Accordi, the interpveness of the interpretes (preferab

y in part accholars, for e in the eyd to locate a

ntribute to Halliday an

ext of situaext). The foNew Testam

he lived. Tin the culturohesive readp involves th

cohesion nt on that odecoded ex

ht prove usterms cohesed and dis

mmunicationiceable pecue users) or cingly, whenpreter wouldthat text. Ifter would bly from thcount for th

example, oyes of one ra totally inc

the producnd Hasan a

ation and oformer relatioment, requirThe reader lure of that liding by linthe interdep‘occurs wh

of another. xcept by rec

18 

seful for sive and

sordered,

n. More uliarities cohesive n talking d bear a f instead need to he same

he reason over the reader is cohesive

tion and appeal to f culture) onship is ring, for links the inguistic

nking the pendency here the The one

course to

19  

it.’9 These cohesive relationships may occur between words and phrases or even between sentences and pericopes. That such relationships occur in texts is not an overly sophisticated observation, but the question remains: How is language used to create these cohesive relationships? To answer this question Halliday and Hasan introduce the concept of cohesive ties.

In sum, cohesive ties refer to the use of a language system to form connections between linguistic items of a discourse. The nature of these connections is primarily semantic, that is, the linguistic ties are related in terms of meaning (not syntax). Cohesive ties consist of two types: organic and componential. These two concepts will now be developed in more detail, forming the major tiers of the present linguistic approach to New Testament cohesiveness. a. Organic Ties ORGANIC TIES primarily concern the conjunctive systems of language, such as particles which serve as markers of transition (e.g. γάρ, ἀλλά, δέ, καί). Organic ties are also signaled by prepositions, grammatical structure (e.g. genitive absolute using γίνομαι), and conventionalized lexical items (e.g. λοιπόν). Koine Greek had a replete system of lexical items which served the task of creating organic ties in discourse. Careful study of these can take the student of the New Testament a long way towards understanding the cohesiveness of a given text.

Organic ties make up the ‘logical’ system of natural language and consist of two functional systems: (1) interdependency or ‘taxis’, namely parataxis and hypotaxis (found at all levels of language) and (2) expansion and projection (limited to the levels of clause and paragraph).10 HYPOTAXIS is the logico-semantic relation between a dependent element and the element on which it is dependent (dominant element). It is a modifying relationship; one element is dependent on the other and the order of the elements varies. In hypotaxis, a secondary clause is dependent on a primary clause. PARATAXIS, on the other hand, is the logico-semantic relation between two linguistic elements of equal status and, thus, either could stand independently of the other. Parataxis is dependent on the order of linguistic elements (e.g. the first clause initiates and the second continues). In other words, the primary clause precedes the secondary clause. The logico-semantic relation between the primary and secondary clause may be one of PROJECTION or EXPANSION.

In PROJECTION, the secondary clause is ‘projected’ through the primary clause by means of

(1) a locution or (2) an idea. LOCUTION occurs with verbs of saying or hearing (direct or indirect discourse); in Greek the secondary clause is usually expressed with the infinitive (Acts 25:11 οὐ παραιτοῦμαι τὸ ἀποθανεῖν) or finite verb forms with particles such as ὅτι, εἰ or ὡς (Mk 1:37 λέγουσιν αὐτῷ ὅτι πάντες ζητοῦσίν σε). IDEA covers a broad range of projections, in which the secondary clause presents ‘an idea, a projection of meaning’.11 Another way of viewing such expressions is that a clause has shifted rank down into the slot of the complement; in this way, the idea is a way of ‘completing’ the process of the primary clause. These, like locutions, are commonly expressed with an infinitive (Jas 1:26 εἴ τις δοκεῖ θρησκὸς εἶναι) or ὅτι construction (Jn 11:27 ἐγὼ πεπίστευκα ὅτι σὺ εἶ ὁ Χριστὸς ὁ υἱὸς τοῦ θεοῦ) in the secondary clause.

                                                            9 Halliday and Hasan, Cohesion, p. 4; cf. P. Werth, Focus, Coherence and Emphasis (London: Croom 

Helm, 1984), p. 90. 

10 See Halliday, Functional Grammar, pp. 192–251, 302–309. 

11 Halliday, Functional Grammar, p. 197. 

20  

In the case of EXPANSION, the secondary clause ‘expands’ the primary clause in one of three ways: (1) elaboration, (2) extension or (3) enhancement. In ELABORATION, the secondary clause (or phrase) expands upon the primary by ‘elaborating’ on it (or some portion of it), that is, restating, specifying, commentating or exemplifying. In EXTENSION, the secondary clause ‘expands’ the primary clause by moving beyond it, that is, adding to it, giving an exception or offering an alternative. In ENHANCEMENT, the secondary clause ‘expands’ the primary clause by qualifying it with a circumstantial feature of time, place, cause or condition. For example, in Greek a preposition plus infinitive is used to expand the primary clause, with the preposition specifying the type of expansion. By way of simile, the three types of expansion may be likened to enriching a building: (1) elaborating the existing structure of a building; (2) extending it by addition or replacement; (3) enhancing its environment.12 The following diagram lists further distinctions of expanding, extending and enhancing textual relations and examples of how Greek (with English equivalents) is often used to express them.

ELABORATION (+)

Apposition (restate or re-present; epexegetical)

expository ὅτι, ἵνα, τοῦτο ἐστιν (in other words, that is, I mean, to put it another way)

exemplifying οὕτως, οὕτω, γέγραπται, ῥητῶς (for example, for instance, thus, to illustrate)

Clarification (summarize or make precise)

corrective μᾶλλον, μενοῦν, μενοῦνγε, ἀλλά, οὐχ ὅτι (or rather, at least, to be more precise, on the contrary, however)

particularizing μάλιστα (in particular, more especially)

summative λοιπόν, οὖν (in short, to sum up, in conclusion, briefly)

verifactive ὅλως, ὄντως (actually, as a matter of fact, in fact)

EXTENSION (=)

Addition

positive καί, δέ, τέ, πάλιν, εἶτα, ἐπί, καί … καί, τε … καί, τε … τε, μέν … δέ (and, also, moreover, in addition)

negative οὐδέ, μηδέ (nor)

Adversative ἀλλά, δέ, μενοῦν, μενοῦνγε, μέντοι, πλήν, παρά (but, yet, on the other hand, however)

Variation

replacive ἀντί, τοὐναντίον, μέν … δέ (on the contrary, instead)

subtractive ἐκτός, εἰ μή (apart from that, except for that)

                                                            12 Halliday, Functional Grammar, p. 203. 

21  

alternative ἤ, ἤ … ἤ, ἤτοι … ἤ (alternatively, or)

ENHANCEMENT (×)

Spatio-Temporal

following καί, δέ, κατά (then, next, afterwards)

simultaneous ὡς, ὅτε, ὅταν, πότε, ποτέ, καθώς, ἄμα, ἐφάπαξ (just then, at the same time)

preceding πρό, πρίν, πρῶτον, ἤδη, πάλαι (before that, hitherto, previously)

conclusive λοιπόν (in the end, finally)

immediate εὐθύς, εὐθέως (at once, immediately, straightaway)

interrupted ταχύ, ταχέως, αὔριον, μέλλω (soon, after a while)

repetitive ἄνωθεν, πάλιν, εἰς τὸ πάλιν (next time, on another occasion)

specific μεταξύ, σήμερον, αὔριον (next day, an hour later, that morning)

durative ἐν τῷ μεταξύ (meanwhile, all that time)

terminal ἕως, ἄχρι, μέχρι (until then, up to that point)

punctiliar νῦν, δεῦρο (at this moment)

Comparative

positive ὅμοιος, ὁμοίως, τοιοῦτος ὅμως, ὡς, ὡσεί, ὥσπερ, καθώς, καθά, καθό, ὡσαύτως (likewise, similarly)

negative ἤ, ἤπερ, negated ‘positive forms’ (in a different way)

Causal-Conditional

(1) causal

result διό, πρός, εἰς, ἵνα, οὖν, τοίνυν, τοιγαροῦν, ὡς, ὥστε (in consequence, as a result)

purpose ἵνα, ὅπως, ὥστε, μήποτε, μή πως (for that purpose, with this in view)

reason ὅτι, γάρ, διά, διότι, χάριν, ἕνεκεν, ἐπεί (on account of this, for that reason)

basis ἐπί, νή (on the basis of, in view of)

(2) conditional

positive εἰ, εἴπερ, ἐάν, ἐάνπερ, εἴτε … εἴτε, ἄν, πότερον (then, in that case, if, under the circumstances)

negative εἰ μή, ἐὰν μή (otherwise, if not)

22  

concessive καίπερ, καίτοι, καίτοιγε, κἄν [καί + ἐάν] (yet, still, though, despite this, however, even so, nevertheless)

Respective

positive ὧδε, ἐνθάδε (here, there, as to that, in that respect)

negative ἀλλαχοῦ (in other respects, elsewhere) Besides creating links in the discourse, another function of organic ties is to set

boundaries. They provide a means of organizing groups of componential ties (see below) into thematic sections/paragraphs. In other words, they may be used to set limits on how far the reader should look in the surrounding discourse to find cohesive relationships.

In sum, organic ties provide a vital means of creating and interpreting cohesiveness in Greek discourse. A thorough discourse analysis of any New Testament text should include a serious study of organic ties. Because of their sometimes inconspicuous presence, it is all too easy to forget that such language plays a vital role in communication and that it can contribute significantly to our understanding not only of the way an argument develops in a text but also of the way in which an author uses language to create that argument. b. Componential Ties Whereas organic ties generally concern various paratactic and hypotactic, logico-semantic relationships between clauses and paragraphs (and phrases), COMPONENTIAL TIES generally concern the meaningful relationships between individual linguistic components in the discourse (e.g. repetition of words). This generally amounts to semantic relationships between words or phrases. In order to account for the various semantic relationships between discourse components, Halliday and Hasan appeal to three types of componential ties: (1) co-reference, (2) co-classification and (3) co-extension.13 These are akin to the distinctions of reference, denotation and sense often discussed in linguistic semantics.

CO-REFERENCE (reference) refers to the cohesive ties between linguistic items of the same identity. In the sentence ‘John bought the suit which he gave to his brother’, the relative pronoun ‘which’ refers to the entity ‘suit’. Both lexical items—‘suit’ and ‘which’—share the same identity. The same is true of ὁ ἀστήρ and ὅν in Mt. 2:9. Co-referential ties may be either exophoric or endophoric. EXOPHORIC information is located in the context of situation and thus also in the context of culture—it may or may not be linked to the co-text as well. Often the interpretation of such elements is based on the cultural and situational presuppositions shared by the speaker and listener.14 Exophoric reference, because it points to information outside of the text, requires more interpretive effort on the part of the modern reader. Werth accordingly notes that

                                                            13 See Halliday and Hasan, ‘Text and Context: Aspects of Language in a Social‐Semiotic Perspective’, 

Sophia Linguistica 6 (1980), pp. 43–59. For an application of this model to a New Testament 

discourse, see J.T. Reed, ‘Cohesive Ties in 1 Timothy: In Defense of the Letter’s Unity’, Neot 26.1 

(1992), pp. 131–47. 

14 Cf. T. Givón, On Understanding Grammar (New York: Academic, 1979), p. 50. 

23  

the most easily acceptable discourse would be one relating to a completely standard context. The more effort that is required of the listener to find a suitable context (assuming normal intelligence and language-experience), the lower in acceptability is the discourse.15

Part of the art of New Testament interpretation is finding a suitable context which

explains all of the exophoric references in the text. The proliferation of New Testament interpretations testifies to the difficulty of finding and convincing others of this ‘suitable context’.

ENDOPHORIC information is located within the language of the discourse itself, that is, a linguistic item referring to other linguistic entities in the surrounding text—it may or may not refer to something in the context of situation as well (i.e. a word or phrase could invoke both an exophoric and an endophoric reference). Endophoric information may refer to elements in the preceding (ANAPHORIC reference) or following (CATAPHORIC reference) discourse.16 Anaphoric reference, one of the most frequently used devices of textual cohesiveness, is a semantic relationship between one entity A (antecedent), which may or may not be linguistic, and another entity B (anaphora such as pronouns, definite descriptions, repetitions and zero-anaphora [e.g. verbal suffix]), which must be linguistic, such that in some discourse, B corresponds to A.17 Consequently, anaphoric elements may also contribute to the incoherence of discourse. This occurs ‘when an anaphor (usually a pronoun) has more than one potential antecedent in preceding discourse, a situation which may easily arise in view of the limited intentional specifications of most anaphors’.18 Anaphors often have little semantic weight aside from their referential capabilities (e.g. relative pronouns).

Another area of linguistic research is directly related to co-referential ties and adds much to Halliday and Hasan’s discussion of reference, namely the notion of DEIXIS. The linguistic devices used to create co-referential ties are often referred to by linguists as deictic indicators. DEIXIS refers to the ability of language users to employ linguistic forms to ‘point to’ or ‘indicate’ elements of the co-text or context of situation. Levinson describes it as ‘the ways in which languages encode or grammaticalize features of the context of utterance or speech event’.19 Deictic indicators are treated below according to commonly recognized categories: person, time and place.

PERSON DEIXIS refers to the encoding of various participants (animate or inanimate) of a context of situation into a discourse.20 Languages often allow for naming conventions to                                                             15 Werth, Focus, p. 35. 

16 Studies in language typology suggest that ‘anaphoric indexicality is universally preferred over 

cataphoric indexicality’ (W. Dressier, ‘Marked and Unmarked Text Strategies within Semiotically 

Based Natural Textlinguistics’, in S.J.J. Hwang and W.R. Merrifield [eds.], Language in Context: Essays 

for Robert E. Longacre [Dallas: Summer Institute of Linguistics and the University of Texas at 

Arlington, 1992], pp. 5–18 [10]). This is a rule of language that is sensitive to the interpretive 

limitations of the reader: ‘Because anaphora refers back to what is already known and cataphora to a 

(potentially) uncertain future, the former establishes the more reliable sign relationship’ (p. 10). 

17 Cf. the important study of anaphor in Werth, Focus, pp. 61–65, 166–68. 

18 Werth, Focus, p. 20. 

19 S. Levinson, Pragmatics (Cambridge: Cambridge University Press, 1983), p. 54. 

20 See M.A.K. Halliday, Language as Social Semiotic: The Social Interpretation of Language and 

Meaning (London: Edward Arnold, 1978), p. 132, and his Functional Grammar, pp. 168–69. 

24  

identify participants. Generally, some nominal form plays this role. Pronouns then take over to make the discourse less redundant. Other means of signaling participants in Greek include the article and verbal suffixes (e.g. first, second, third person, singular and plural). Person deixis not only concerns the referents of a discourse, but also the roles played by those referents. Typical roles include the SPOKESPERSON(S) and RECIPIENT(S). The spokesperson produces the discourse, but can also be distinguished from its SOURCE. Indirect discourse is one way in which a spokesperson shifts to another person as the source of a message. For example, in 1 Cor. 12:16 the anatomical ‘ear’ is the source of a direct discourse (‘Because I am not an eye, I do not belong to the body’ [ὅτι οὐκ εἰμὶ ὀφθαλμός, οὐκ εἰμὶ ἐκ τοῦ σώματος]) which is recounted by the spokesperson, Paul (the actual source of the direct discourse). Paul at times appeals to this distinction between spokesperson and source. In 1 Cor. 7:10 he gives a command to married persons, adding that it is not his command but the Lord’s. Later in the letter he gives his own opinion regarding the pursuit of marriage by virgins (1 Cor. 7:25). In one case the source is the Lord; in another he is the source. Just as the spokesperson may not be the source of a message, the recipient may not be the TARGET. For example, in 1 Timothy a common argument has been that the intended target comprised a wider circle of believers, even though the letter is addressed solely to Timothy. Recipients may also be either RATIFIED or UNRATIFIED. To draw from 1 Timothy again, if the author’s real target was the larger church body (see 1 Tim. 4:13), he may have only ratified part of that group to hear his letter (e.g. elders and presbyters). These leaders (including Timothy) were then expected to see that the message eventually reached the entire church community.

TEMPORAL DEIXIS is another important way in which elements of a discourse may be tied to a context of situation.21 For example, in the dates of Greco-Roman epistles letter writers sometimes specify the anchorage point from which the date is to be understood using the word ἐνεστῶτος ‘present, current’: for example, P.Oxy. 8.1128.9–11 (173 CE) τοῦ ἐνεστῶτος τρεισκαιδεκάτου {τρισκαιδεκατου} ἔτους Αὐρηλίου Ἀντωνίνου Καίσαρος (‘in the present thirteenth year of Emperor Marcus Aurelius’). According to one grammatical model of the Greek language, verb tenses serve as temporal indicators in discourse. Recent research in Greek verbal aspect either abandons or significantly dilutes the idea of time in the verbal tense-forms. This debate is far from over, and presently there are exegetes from all three schools of thought (temporal, Aktionsart, aspect), and some with mixed categories. The point, however, is that the interpretation of tense-forms as temporal deixis is disputed. Some less-disputed deictic indicators of time in Greek include adverbs (e.g. τότε, νῦν, μέχρι), anaphora (e.g. demonstratives), and references to places (spatial and temporal deixis are closely related). All three classes are helpful for establishing temporal relationships with the context. The first class represents a type of direct temporal indication; the others are indirect. Discourse analysis of temporal deixis is complicated by the point of view of the speaker. For example, if a letter contains the sentence ‘Now I will tell you what I think about your previous correspondence’, does the word ‘now’ refer to the point in time at which the letter was written or the point in time when it was read? Similarly, does the aorist ἔγραψα in 1 Cor. 5:11 (νῦν δὲ ἔγραψα ὑμῖν …) refer to the time when the letter was written or the time when it was read?22 In deictic terms, the distinction is between the ‘moment of utterance (or inscription) or coding

                                                            21 Halliday, Functional Grammar, p. 176. 

22 The label ‘epistolary aorist’ does not solve this ambiguity; it only specifies one possible 

interpretation; there is no ‘temporal’ problem needing explanation if the Greek verb does not 

indicate absolute time. 

25  

time (or CT)’ and the ‘moment of reception or receiving time (or RT)’.23 Temporal deixis is even relative to the speaker/author, since at the precise point in time when the word ‘now’ or ἔγραψα was actually written, the author has not yet finished the clause. The point being emphasized here is that grammatical forms do not provide absolute indicators of time. They are relative to anchorage points which are determined by the speaker.

Levinson defines PLACE DEIXIS as ‘the specification of locations relative to anchorage points in the speech event’.24 The characteristic feature of place deixis is that it concerns the spatial locations of persons/objects relative to other persons/objects. An example of this may involve a speaker (anchorage point) pointing to an object across the room (relative location). This may be done explicitly by naming or describing locations, as in the sentence ‘The school is 20 miles away from the library’ and in Mt. 14:24 τὸ δὲ πλοῖον … σταδίους πολλοὺς ἀπὸ τῆς γῆς ἀπεῖχεν (‘but the boat … was many stadia from the shore’). Locations may also be specified deictically relative to the location of a speaker, as in the sentence ‘The dog is 50 feet away from here’ and in Jn 4:16 ὕπαγε φώνησον τὸν ἄνδρα σου καὶ ἐλθὲ ἐνθάδε (‘go call your husband and come here’). In both examples, ‘here’ and ἐνθάδε refer to something outside of the discourse. ‘Here’ and ἐνθάδε are the anchorage points around which other objects are relatively located. As in temporal deixis, place deixis is relative to the speaker, that is, the speaker determines the anchorage point around which other entities are located. Accordingly, because they are functionally related to other elements in the discourse, anchorage points contribute to textual cohesiveness.

CO-CLASSIFICATION (denotation)—the second type of componential tie—refers to cohesive ties between linguistic items of the same class or genus. One way to create this type of tie is by SUBSTITUTION, as in ‘I want the children to draw with crayons’ and ‘I want the teenagers to draw with pencils’. By substituting ‘teenagers’ for ‘children’ and ‘with pencils’ for ‘with crayons’ the two sentences form a cohesive tie of co-classification with respect to who should do the drawing and how it should be done. A co-classificational tie (of ‘sinning’) is created in Rom. 2:12 by substituting ἀνόμως with ἐν νόμῳ (ὅσοι γὰρ ἀνόμως ἥμαρτον … καὶ ὅσοι ἐν νόμῳ ἥμαρτον). Another way to convey co-classification is by ELLIPSIS (or zero-anaphora).25 For example, an individual might say to another, ‘I hit the ball so hard it went over the parking lot. How hard did you?’ A cohesive relationship exists between these sentences because of the elided element ‘hit the ball’. Both sentences do not refer to the same event; rather, they fall into the class of ‘ball-hitting’. Similarly, in Phil. 2:4 (μὴ τὰ ἑαυτῶν ἕκαστος σκοποῦντες ἀλλὰ τὰ ἑτέρων ἕκαστοι) the participle σκοποῦντες is elided after ἀλλά, creating a co-classificational tie of ‘considering’. The simple act of elision invokes a cohesive interpretation by the reader.

CO-EXTENSION (sense)—the third type of componential tie—refers to cohesive ties between linguistic items of the same semantic field, but not necessarily of the same class.26 In the sentences ‘John ate the pizza’ and ‘Susie gobbled down the cake’ the linguistic pairs ‘John’ / ‘Susie’, ‘ate’/ ‘gobbled down’ and ‘pizza’ / ‘cake’ do not refer to the same entities nor do they refer to the same class (e.g. pizza is not a kind of cake). Co-extensional ties are one of the most common ways of creating cohesiveness in texts. These ties are primarily lexical. By using words with similar senses speakers talk about similar things in similar ways.

                                                            23 Levinson, Pragmatics, p. 73. 

24 Levinson, Pragmatics, p. 79; cf. Halliday, Functional Grammar, p. 160. 

25 Cf. Werth, Focus, p. 176. 

26 Co‐classification is a subtype of co‐extension. 

26  

Halliday and Hasan distinguish between two types of co-extensional ties: (1) instantial and (2) general. INSTANTIAL LEXICAL RELATIONSHIPS arise from the particular demands of the text.27 For example, the author of 1 Timothy may be referring to the specific individual Τιμοθέος when he uses the vocative ὦ ἄνθρωπε θεοῦ in 1 Tim. 6:11. However, this understanding is based on knowledge derived from the co-text or context of situation (instances of the discourse) and not from the Greek language itself. That is, ὦ ἄνθρωπε θεοῦ, as a group of Greek words, is not a code-based substitute for Timothy. Instantial ties often prove difficult for the modern reader because their interpretation is based on knowledge of the immediate text or the context of situation; a study of other contemporary literature or of Greek semantics is of little or no help.

GENERAL LEXICAL RELATIONSHIPS originate from the language system itself; thus, they are shared by a group of language users. General co-extensions take five forms: reiteration, synonymy, antonymy, hyponymy and meronymy. REITERATION occurs when both members of the cohesive tie consist of the same lexical item. This is one of the more obvious forms of cohesive tie. However, the simple repetition of a lexical item does not imply total synonymy nor does it leave out the possibility that the same (spoken or written) lexical items have two quite different meanings with the same spelling (a monetary ‘bank’ or a dirt ‘bank’) and/or pronunciation (‘meet’ and ‘meat’). Furthermore, the repetition of some words, such as the article ὁ, does not necessarily indicate cohesiveness. That is, repetition is not a phenomenon of the code itself but of the code as it is used by a speaker/author in a particular discourse; hence its presence in discourse as a cohesive device must be argued for by the interpreter, not simply asserted (this is an important point for debates over literary integrity). SYNONYMY refers to cohesive ties created by lexical items sharing similar meanings (but not necessarily being totally synonymous), that is, words from the same semantic domain. ANTONYMY refers to cohesive ties created by lexical items opposite in meaning. It is not that antonyms are unrelated in meaning but that the antonyms differ in one or more semantic features but share others, that is, there is negativity and similarity.28 Thus ‘dog’ and ‘kite’ are not antonyms because they do not share anything in common that would allow the listener to recognize a semantic tie between the two. The cohesive tie, instead, is created by means of shared semantic features (e.g. the antonyms ‘husband’ and ‘wife’ share the notion of ‘marriage’ relationships). HYPONYMY refers to cohesive ties created by the inclusive relationships between lexical items. One lexical item is included in the total semantic range of another item (but not vice versa). This allows for a hierarchy of meanings in lexical systems, and is perhaps one of the most important attributes of human language which allows us to organize the world around us into meaningful categories. For example, ‘labrador’ may be considered a hyponym of ‘dog’; ‘dog’ is a hyponym of ‘animal’; ‘animal’ is a hyponym of ‘living beings’, and so on. Similarly, οὗς is a hyponym of μέλος. The one is included in the semantic range of another, which in turn is included in the semantic range of another, and so on. Hyponymy may be further distinguished according to contracting types (e.g. ‘People got on and off. At the news-stand businesspersons, returning to Paris, bought that day’s papers’) and expanding types (‘Tulips are cheap even in January. But then flowers seem to be necessary to the Dutch during the darkest season’). MERONYMY refers to part-whole relationships between lexical items. For example, the word ‘fur’ is a meronym of ‘dog’ or ‘cat’. Similarly, κόμη is a meronym of κεφαλή. The one is a part of the other. Because it is part of the other, a cohesive relationship may be created between the two in a discourse. A study of these five types of co-extension in discourse is not always straightforward, because individual words often have several

                                                            27 Halliday and Hasan, ‘Text and Context’, pp. 43–59. 

28 On this type of contrastive coherence, see Werth, Focus, pp. 87–89. 

27  

meanings (i.e. senses). The result is polysemic indeterminacy. This occurs ‘when certain senses of two items are potentially linkable’.29 Thus, this phenomenon of polysemic indeterminacy provides an important caveat for analyzing co-extensions in discourse. What one reader might take to be a cohesive tie because of a hyponymous relationship between two words, another reader might not take as cohesive because of a different meaning attributed to one of the words.

Through the analysis of co-referential ties (e.g. pronouns, demonstratives), co-classificational ties (e.g. substitution, ellipsis), and co-extensional ties of both instantial (i.e. those tied to the situational context) and general types (repetition, synonymy, antonymy, hyponymy and meronymy), the discourse analyst is able to demonstrate a major component of textual cohesiveness. As seen above, co-reference and co-classification are primarily expressed by grammatical networks in the language and co-extension is primarily expressed by lexical networks.

A nagging question remains: What makes one text seemingly more cohesive than another? Or, what criteria may an interpreter employ in a discussion of the cohesiveness of a New Testament text? To answer the former question Halliday and Hasan speak in terms of SEMANTIC CHAINS. A chain is formed by a set of discourse lexemes each of which is related to the others by the semantic relation of co-reference, co-classification and/or co-extension. If a text, for example, contains a participant who is identified using pronouns (e.g. ‘he’), demonstratives (e.g. ‘this’), or the person’s name (e.g. ‘Paul’ as the first-person participant of a letter), then these elements form a chain of co-reference. There are two types of chains: identity chains and similarity chains. IDENTITY CHAINS are expressed by co-referential ties and SIMILARITY CHAINS are expressed by co-classificational and co-extensional ties. Exposing the identity and similarity chains of a text, nevertheless, proves less than adequate when attempting to speak about the relative cohesiveness of a text. In order to determine relative textual cohesiveness, the discourse analyst should differentiate between peripheral, relevant and central tokens. PERIPHERAL TOKENS include those linguistic items which do not take part in a chain. This happens, for example, when a topic is brought up in a clause and then subsequently dropped from the discussion. It is isolated from other chains and, hence, is peripheral to the author’s larger argument. RELEVANT TOKENS include all linguistic items in the text which are part of one or more chains. It should not be concluded, however, that a high proportion of relevant tokens to peripheral tokens necessitates greater textual cohesiveness (although it may play some role). Textual cohesiveness is primarily occasioned by central tokens. CENTRAL TOKENS refer to linguistic items in chains which interact with linguistic items in other chains. For example, in the New Testament, a co-extensional chain of supernatural beings might interact with a co-extensional chain of miracles (e.g. God raised Jesus from the dead). If the two chains interact in more than one part of the text (esp. in close contexts), it is probable that the author is ‘on about’ a similar topic, thus creating cohesiveness and potential coherence in the text. He is establishing a thread in the discourse. She is using her language in an organizing manner. Central tokens, in essence, involve chain interaction. Halliday and Hasan claim that ‘the minimum requirement for chain interaction can be phrased as follows: for two chains x and y to interact, at least two members of x should stand in the same relation to two members of y’.30 In other words, two lexical items (the same or different) of the same chain must be used in conjunction with at least two other lexical items (the same

                                                            29 Werth, Focus, p. 21. 

30 Halliday and Hasan, ‘Text and Context’, p. 57. 

28  

or different) of another chain.31 Typically, chain interaction involves a chain of participants (e.g. ‘the Philippian Christians’) and a chain of events (e.g. ‘think’); however, chain interaction may occur when one chain of participants interacts repeatedly with another chain of participants (e.g. ‘Paul’ says, hopes, sends ‘the Philippians’). Halliday and Hasan’s principle is based on the view that, with respect to textual cohesiveness, the main basis for coherence lies in similarity. Chain interaction is a theory of similarity in texts—the view that cohesiveness is created by speakers saying similar kinds of things (e.g. chain 1) about similar kinds of phenomena (e.g. chain 2). In non-technical terms, chain interaction is the speaker’s being on about similar kinds of things. This understanding of language use is closely related to the principle of linguistic REDUNDANCY, that is, texts will typically transmit less information than the sum of their linguistic parts. Redundancy ‘serves to reduce the likelihood of an error in the reception of the message resulting from the loss of information during the transmission’.32 We repeat things, that is, we say similar kinds of things about similar kinds of phenomena, in part to ensure that our point is being understood.

2. Conclusion

It is a singularly unique feature of human language that we can combine long stretches of symbols to communicate meaning. How we combine individual parts into a complete whole has been a primary concern of discourse analysts. In this article, I have set forth some linguistic criteria for analyzing the cohesiveness of New Testament Greek discourse. My focus has been on (1) specialized linguistic symbols which developed in koine Greek to signal organic ties in discourse and (2) individual linguistic components (mostly words) which are related to one another in one of four meaningful relationships (hyponymy, antonymy, synonymy, meronymy). This is not, of course, a comprehensive model. To present one would mean entering into a complete discussion of discourse analysis. I have skipped over, for example, two important topics relevant to the study of discourse cohesiveness: information flow and word order.33 Instead I have dealt with linguistic matters which most would understand as directly related to critical issues surrounding the cohesiveness of New Testament discourse. For example, the literary integrity of Philippians has often been argued for by appealing to semantic parallels in the discourse, but the method of determining such parallels has typically been linguistically naive. Furthermore, this model might prove useful in the teaching of Greek, since it highlights to the student the means by which Greek authors could create cohesiveness. That is, it focuses on whole texts rather than simply on grammar isolated from texts.                                                             31 To limit chain interaction to ‘two’ may seem arbitrary, but it is the necessary lowest boundary since 

if only ‘one’ chain interaction were required then every clause of discourse would necessarily be a 

central token—a problematic conclusion. Admittedly, Halliday and Hasan are after a relative (scalar), 

not absolute, set of criteria for speaking about the cohesiveness of discourse. 

32 J. Caron, An Introduction to Psycholinguistics (trans. T. Pownall; New York: Harvester Wheatsheaf, 

1992), p. 5. 

33 On the former, see J.T. Reed, ‘Identifying Theme in New Testament Discourse: Insights from 

Discourse Analysis’, in S.E. Porter and D.A. Carson (eds.), Discourse Analysis and Other Topics in 

Biblical Greek (JSNTSup, 113; SNTG, 2; Sheffield: Sheffield Academic Press, 1995), pp. 75–106; and on 

the latter, see J.T. Reed, ‘Modern Linguistics and Biblical Studies: A Review Article’, in S.E. Porter and 

D.A. Carson (eds.), Linguistics and the New Testament: Critical Junctures (JSNTSup, 168; SNTG, 5; 

Sheffield: Sheffield Academic Press; forthcoming 1999). 

29  

In closing, it must be repeated that cohesiveness is both a product of the speaker’s use of the linguistic code and a result of the listener’s interpretation of the discourse. There is no guaranteed one-to-one correspondence between the authorial intent of cohesiveness and the reader’s response to it. In other words, the intended cohesiveness may not be the interpreted one. Discourse analysts have repeatedly made distinctions between these two sides of the communicative coin. Because of its complexity, discourse cohesiveness allows for multiple interpretations and this cannot be overlooked when arguing for the cohesiveness of a given New Testament text. Nonetheless, the fact that humans do in part communicate their intent in the form of strings of symbols suggests that there is some agreed upon convention of cohesiveness. To quote J. Gumperz:

It seems clear that knowledge of grammatical rules is an essential component of the interactive competence that speakers must have to interact and cooperate with others. Thus if we can show that individuals interacting through linguistic signs are effective in cooperating with others in the conduct of their affairs, we have prima facie evidence for the existence of shared grammatical structure.34

If we extend Gumperz’s argument to include linguistic signs linked together into complete

discourses, then the discourse analyst has a strong basis for making claims about the cohesiveness of a New Testament text when that analysis is based on a comprehensive understanding of koine Greek’s ability to be used in cohesive ways. The above study is an attempt to present some of the linguistic criteria necessary for such an analysis of New Testament cohesiveness.  

 

 

 

 

 

 

 

 

 

 

 

 

 

                                                            34 J. Gumperz, Discourse Strategies (Cambridge: Cambridge University Press, 1982), p. 19. 

30  

IS CRITICAL DISCOURSE ANALYSIS CRITICAL? AN EVALUATION USING PHILEMON AS A TEST

CASE Stanley E. Porter

1. Introduction

Critical Discourse Analysis (CDA) is a recent development in the area of discourse analysis, datable at least in English-speaking circles to 1979 (Fowler et al. 1979; Hodge and Kress 1979/1995; cf. Fowler 1996; Kress 1996). Since that time, it has developed in a number of different ways, so that a recent survey briefly describes eight important theoretical approaches to CDA, embracing work in the English-, French- and German-speaking worlds (Fairclough and Wodak 1997: 262–68). A recent volume brings together a collection of insightful theoretical and applied essays by a range of contributors, well illustrating recent developments in the field (Caldas-Coulthard and Coulthard 1996). In the light of the slow development of the incorporation of social-scientific methodology, including forms of modern linguistic analysis, into New Testament studies (see Reed 1995), it is not surprising to discover that CDA has not developed its own niche in terms of forms of discourse analysis used in study of the New Testament text (see Porter 1995). In fact, to my knowledge, reference to work in CDA is confined to a brief section in a recent article by Reed (1996: 237–38) and a few references in the recent volume and chapter by Reed, where he notes the contribution of Fairclough (though not CDA) to his thinking, especially that of two areas, the social use of language (Reed 1997a: 30, 35, 82, 103; Reed 1997b: 214) and the literary dimension, especially that of intertextuality (Reed 1997a: 55, 56). Thus, in this, to my knowledge the first article utilizing CDA for New Testament study, it may come as somewhat of a surprise to raise the question of whether CDA is critical. But critical in what sense? I wish to address three questions here. First, in what sense is CDA critical of itself? Secondly, in what sense is CDA critical of and to the discipline of linguistics? Thirdly and lastly, in what sense is CDA critical for the discipline of New Testament exegesis? For the sake of exposition, I will use Philemon as an example.

2. In What Sense Is Critical Discourse Analysis Critical of Itself?

This may seem like an odd question to ask at this stage, since the discipline of CDA is relatively young, and completely unproven in the area of New Testament studies. However, when one begins to examine some of the purported distinctives of CDA, one begins to note that self-criticism might well be one of the activities it should engage in at an early stage in order to try to ensure longevity. In section 3 below, I will examine its contribution to the field of linguistics. Here, I wish to assess two dimensions of this linguistic perspective. The first are the essential presuppositions of CDA, and the second is with regard to what Fairclough and Wodak (1997: 262–68) state are the eight important theoretical approaches to CDA.

Fairclough and Wodak (1997: 258) state that there are two distinctives of CDA that distinguish it from other forms of discourse analysis. The first is its view of ‘the relationship between language and society’, and the other is its ‘relationship between analysis and the practices analysed’. These statements in and of themselves say very little about the discipline, however, since sociolinguistics is by definition concerned with ‘language and society’, as they seem to admit obliquely (1997: 259) (the other statement is even more opaque, and will be explored below). In the course of discussion, however, it becomes clear what Fairclough and Wodak mean by ‘language and society’. After repeating a number of statements about the relationship between language, society and ideology, in which discourse is said to be socially constitutive and socially shaped (1997: 258), they assert that

31  

Critical discourse analysis applies to language types of critical analysis which have developed within ‘Western Marxism’. In broad terms, Western Marxism has given considerably more emphasis than other forms of Marxism to cultural dimensions of societies, emphasizing that capitalist social relations are established and maintained (reproduced) in large part in culture (and hence in ideology), not just (or mainly) in the economic ‘base’ (1997: 260).

Recent developments in central Europe might well make one wonder whether tying any intellectual theory too closely to Marxism, especially one concerned with practical results (as is CDA), is a very wise move. It could only be expected that the practical results would fail, when, as Karl Popper has rightly shown, the theory of Marxism itself is based on fundamental flaws (Popper 1974: 33–36).

Those invoked in the name of CDA are also not all of the same sort, including such figures as Louis Althusser, the Frankfurt School including Jürgen Habermas, and Michel Foucault. The problem becomes more acute when Fairclough and Wodak admit that ‘Foucault’s work on discourse was explicitly directed against Marxism and theories of ideology’ (Fairclough and Wodak 1997:261). From what will be seen below, there is nothing essential in CDA that is dependent upon a Marxist theory of history, ideology or power. In fact, their alignment with this ideology, besides smacking of overt political correctness, is undermined by their desire to insist upon their being social scientists. They continue that ‘CDA sees itself not as dispassionate and objective social science, but as engaged and committed’ (1997: 258), and ‘What is distinctive about CDA is both that it intervenes on the side of dominated and oppressed groups and against dominating groups, and that it openly declares the emancipatory interests that motivate it’ (1997: 259). Furthermore, ‘many analysts are politically active against racism, or as feminists, or within the peace movement, and so forth’ (1997: 258). A footnote clarifies that ‘We would like to mention the International Association for the Study of Racism (IASR), which many critical linguists belong to. This association gathers over 200 European scholars and—among other activities—reacts to racist discourse in the public sphere through resolutions, letters, expertise, etc.’ (1997: 281 n. 2). Surely someone is protesting too much! Note that CDA has taken on a life of its own—it is CDA that sees and acts, as if a discipline can do anything apart from those who are its practitioners.

Reed (1996: 237) recognizes that this language of CDA smacks of political correctness. But is life so simple that one can categorically posit that one should always be on the side of the dominated, and that emancipatory interests are always to be chosen? Can it really be thought that all linguists who use the methods of CDA endorse such an agenda? Must they? Fairclough and Wodak admit that ‘Critical discourse analysts do not always explicitly place themselves within this legacy, but it frames their work nevertheless’ (1997: 260). This seems to be, at best, one of those unfalsifiable theories that prove absolutely nothing—at worst, an attempt to legislate uncritical conformity to a (to my mind, suspect) agenda. Such statements make other statements by Fairclough and Wodak ring hollow: ‘But CDA is not an exception to the normal objectivity of social science … This certainly does not imply that CDA is less scholarly than other research: standards of careful, rigorous and systematic analysis apply with equal force to CDA as to other approaches’ (1997: 258, 259). It may be true that ‘social science is inherently tied into politics and formulations of policy’ (1997: 258–59), but surely it cannot be an objective endeavour (the word ‘objective’ is problematic, but it is theirs) if it pre-decides how these terms are defined and how one must insist on and respond to them. Dying the death of innumerable logical inconsistencies and qualifications, let us leave the Marxist agenda where it belongs.

The ideological agenda seems to account for the broadly inclusive scenario of what constitutes CDA. Eight approaches to CDA are noted by Fairclough and Wodak (expanded from the approaches noted by Fairclough 1992: 13–35). For example, it is noted that French

32  

Discourse analysis is highly dependent on forms of ideological theory, especially in the work of Pêcheux, who explores ‘the ideological effects of discursive formations in positioning people as social subjects’ (Fairclough and Wodak 1997: 263). Similarly, the Duisburg School has been ‘massively influenced’ (1977: 267) by Foucault in its treatment especially of far right movements in Germany, as has the Reading Analysis method, with its Foucaultian critique of social discourse practice, especially in fascist discourse. This kind of emphasis seems also to have influenced what is called Socio-Cognitive Studies, typified by the recent work of van Dijk, with his concerns for ‘the abuse of power and the reproduction of inequality through ideologies’ (Fairclough and Wodak 1997: 265; cf. van Dijk 1996), including confirmaton of the role that ‘elites’ play in racism, and the Discourse-Historical Method of Wodak and her study of antisemitism and racial prejudice in Vienna (cf. Wodak 1996), elaborating van Dijk’s socio-cognitive approach. What seems to distinguish these methods is that they are all top-down linguistic methods (Brown and Yule 1983: 234–36). That is, they begin with a large conception of the nature, purpose and function of discourse, and examine how this complex of ideas (ideologies?) is found to be present in sociolinguistic contexts and the texts that they produce.

Fairclough and Wodak also mention three other approaches. These include the Critical Linguists who came under the influence of Halliday and systemic-functional theory. This has been further developed in Social Semiotics by some of the same people who practised Critical Linguistics, and taken even further by one in particular (Fairclough himself) in what is called Sociocultural Change and Change in Discourse (cf. Fairclough 1992; 1996). In other words, these seem to be three very similar (or personal) variations on a Hallidayan discourse framework. Such a framework is characterized in contrast with the ‘abstract focus’ of French Discourse Analysis by ‘the attention it gives to grammar in its ideological analysis. Features of the grammatical form of a text are seen as meaningful choices from within the possibilities available in grammatical systems’ (Fairclough and Wodak 1997: 263). This bottom-up method seems to be reflected in the eight planks of practice outlined below.

More to the point, major approaches to CDA seem to be two in number, those governed by ideology imposed from the top down, consistent with the overt, and biased, presuppositions of some of its practitioners, and those based in the Hallidayan systemic-functional framework, with its attention to grammatical systems and working from the bottom up. The way in which these issues, as well as approaches, are presented in the theoretical literature, it appears that to date perspectives on CDA have yet to come to terms with the nature of CDA as a method. When it is shed of its ideological baggage, in many ways CDA appears to be a further development of the Hallidayan form of discourse analysis, perhaps with increased sensitivity to ideological issues as they are manifested in various situations of power (this is how Fairclough 1989 and 1992 appear). It will be treated on those terms below.

3. In What Sense Is Critical Discourse Analysis Critical for Linguistics?

Although those who practise CDA have not been as self-critical as they might have been, this does not mean that they have not had a healthy critical effect upon the rest of the discipline of linguistics. This is not the place for a thorough critique of all dimensions of CDA and what it has effected in relation to the wider field of linguistics. However, in several areas, CDA has brought to the fore changes in emphasis that have had a useful, modifying effect upon linguistic theory. Many of these have provided extensions of Hallidayan theory, but they are not less important for that.

One of these contributions is the emphasis upon the relationship between language and society being formative for discourse. What is meant by this is that language, as a form of ‘social practice’ (Fairclough and Wodak 1997: 258), maintains an interactive relationship between a discourse and its situation. So far, there is nothing new here, since many functional

33  

grammars, and especially systemic-functional grammar, have noted this relationship, and addressed the question of context. Context is an important dimension of linguistic discussion that has been underdeveloped, even in Hallidayan grammar. Reed has improved upon the usual Hallidayan framework, which emphasizes context of situation, by developing the context of culture (1997a: 41–58). CDA would wish to take this further by noting that there are situational, institutional and social structures that are part of this context of culture that require particular attention, especially in their ideological dimension (Fairclough 1989: 2). For the study of New Testament Greek texts, this would include not only the Hellenistic Greek language, whether or not seen as a standard language, varieties of language or dialects, and idiolects (as Reed notes; 1997a: 58), but also such factors as the institutional power of the Roman Empire, and its localized effect in terms of civil and miltary force, civic religion (including the incipient emperor cult) and even environmental engineering (such as architecture), and the place of various racial/ethnic/religious groups, of which Judaism was one, and Christianity soon became another. Part of the context of culture becomes Christianity’s own institutional structures, as churches were founded, organized, managed and developed, and drew into them converts from various social and economic strata, from the wealthy to slaves, and men and women.

A further implication of CDA is that, as Fairclough states (1989: 1), it helps to ‘correct a widespread underestimation of the significance of language in the production, maintenance, and change of social relations of power’. The importance of this topic has been appreciated in some social-scientific studies of the New Testament (see Petersen 1985), but merits much more focused study. Thus, a text is important for constituting its situational context as well as being shaped by it. The result is that CDA not only attempts to provide a description of sociolinguistic situations and conventions, but attempts to explain these relationships in terms of a variety of factors. One need not, as some practitioners of CDA have done, reductionistically attempt to explain all language practice as the manifestation of struggles for power, which seems to be less an explanation than a simplistic assumption. There is no denying that there were power relations in the early Church that certainly had a result in linguistic practice, such as the decree of the Apostolic Council in Acts 15. But there are other decisive factors for determining linguistic usage in the early Church, including theological belief systems (which in other contexts might well be characterized as ideological differentiations, although in the Christian Church it may not have been diametrically opposed ideologies so much as conflicting ideologies), social values and backgrounds, and even personal characteristics, manifested in idiolect. There is, of course, a tendency to extend the concept of explanation to one of prescription, especially if one’s explanatory agenda is ideologically committed. Although some CDA may well tend in this direction (as noted above), it is not a necessary conclusion. For example, rather than simply describing an idiolect, such as Mark’s paratactic style, one might well explain this in a number of ways. Confining explanations to those with at least a modicum of plausibility (i.e. rejecting the Holy Ghost dialect and related theories), one might posit Semitic influence reflecting the use of the waw in Mark’s first language, or one might posit an oral environment for linguistic production, in which the author was perhaps taking down the recollected account of another or was himself illiterate, or one might posit that his language reflects a particular register of written Greek, found also in a number of documentary papyri from Egypt. Whatever one decides on this issue, the temptation to say that Mark should have written his Gospel in another way must be resisted.

A third contribution of CDA is a necessary questioning of the Saussurian distinction between langue and parole. This may come as a surprise, since the distinction is one of the foundational concepts of much modern linguistic theory. Saussure has been interpreted to have made a distinction between langue as a system or code of language that stands outside of and prior to individual usage, which is found in parole. Parole consists of the individual

34  

choices of a particular language speaker. Hence the emphasis of much linguistics has been upon langue, rather than parole. CDA wishes to question the distinction and its definition. For example, the social factor in language usage does not enter into the Saussurian estimation of parole, and this sociolinguistic factor has led to needed re-evaluation of the Saussurian concept, and recognition that much language usage is determined by social factors, including such things as education, socio-economic background, racial and ethnic identification, besides personal factors. This social factor has not been widely appreciated in the study of the variety or varieties of Greek of the New Testament. Since the work of Moulton early in this century (1909: 60–97), there has been little attempt, no doubt due in large part because of the lack of knowledge regarding the biblical writers, to appreciate how their language reflects these factors. Furthermore, the concept of langue as a homogeneous, consistent and unified phenomenon is belied by linguistic variation, even by those who purport to speak the same language. The language of the books of Hebrews and Revelation, as well as the Pauline letters and the Gospels, shows a similar situation even within such a restricted corpus. The resolution of this difficulty, so far as CDA is concerned, is to assume a form of langue in terms of the underlying social conventions of language use, and actual use, with the concentration being upon actual usage. This is an extension of the functional distinction often made between code and text, with the text providing the focus of study in order to arrive at an estimation of code, which may vary according to a variety of factors. There is also a resultant move away from emphasis upon the individual to appreciation of language as in some sense group behaviour, that is, behaviour that reflects group conventions and has a reciprocal influence upon the group.

Despite what I have said above, a fourth and final dimension of CDA that has merit for the larger discipline is its activist role, summarized by Fairclough and Wodak when they say that a concern of CDA is ‘the relationship between analysis and the practices analysed’ (1997: 258). What that relationship is, however, needs to be examined. Reed (1996: 237) thinks that CDA has merit in forcing those in the humanities, including those in biblical studies, to demonstrate their value as disciplines through doing applied work that shows its practical value, including perhaps the analysis of power relations in missionary contexts. No doubt Reed is right that exegetical work must also ask larger hermeneutical questions regarding the values, truth claims and agendas of the texts analysed. Nevertheless, much systemic-functional discourse grammar, as well as discourse analysis as a whole, has long demonstrated its concern for practical work on real texts and in real contexts. In fact, discourse analysis began with treatment of actual texts that arose in oral contexts, and has now extended its scope to include written texts as well. But, in all of this, it was actual instances of linguistic practice that merited examination. What Fairclough and Wodak are speaking of, however, is something more, I think. It is not necessary to prejudge this relationship in terms of power relations to appreciate that discursive practice is, and has always been, engaged in for a variety of reasons, some noble and others not. Nevertheless, these discursive practices merit conscious analysis of their reasons, and an explicit method by which such analysis can take place.

4. In What Sense Is Critical Discourse Analysis Critical for New Testament Study?

Recent practitioners of CDA have arrived at eight principles of interpretative theory or method. As one might expect of such a model of discourse analysis, its engagement in practice, not just theorizing, is crucial to its conception of itself as a linguistic model. I have chosen the book of Philemon as a suitable text to demonstrate the model’s capacity. There are several reasons for its selection. One is its size. It is a short enough book that it is possible to discuss the entire book as a complete discourse, rather than needing to extract a particular portion for examination. Another is the fact that, at least on one interpretation, there are

35  

compatibilities between the ideological interests of CDA and the book of Philemon, especially in terms of social, political and power relations manifested and manipulated by linguistic practice. I rely upon the eight principles outlined by Fairclough and Wodak (1997: 271–80), recognizing that I and others might well wish to alter these, or select others for use, not constituting these as a definitive and comprehensive list for CDA. a. Critical Discourse Analysis Addresses Social Problems As Fairclough and Wodak (1997: 271) state, ‘CDA is the analysis of linguistic and semiotic aspects of social processes and problems. The focus is not upon language or the use of language in and for themselves, but upon the partially linguistic character of social and cultural processes and structures.’ The early Christian movement had to face a number of problematic social problems, some of them inherited from its primarily Jewish roots, and some of them simply reflecting the Graeco-Roman world in which Christianity developed as a religious cult with its own distinctive values. There are a number of passages in the Pauline letters alone that reflect these various social problems. Galatians 3:28 is, in effect, an encapsulation of many of these issues, when it asserts that being ‘in Christ’, or being a Christian, effects altered relations along racial and ethnic lines (Jew and Greek), socio-economically (slave and free), and in matters of gender (male and female). All of these issues and more are reflected in a more extended fashion elsewhere in the Pauline letters. For example, Galatians is concerned with the issue of Jewish legalism and Christian converts, with the conclusion that Gentile converts are not obligated to perform Jewish legalistic rituals. Romans 13 is concerned with the issue of the state, and allegiance to it, concluding that obedience to a just state is necessary to preserve social order, with the implication that the unjust state is not obliged obedience. The Haustafeln encapsulate a number of the important socio-economic issues involving males and females, children and parents, and slaves and masters.

Many scholars have noted that Christianity did not take an overt stance against slavery. In fact, three major bodies of data have been seen to be at best inconclusive, and quite possibly harmful in this regard. 1 Corinthians 7:21 has been questioned whether it endorses slaves taking the opportunity for freedom, or whether it endorses their not asserting this and thereby disrupting the social order. The Haustafeln, while providing reciprocal relationships not found in earlier Haustafel forms, in which masters are seen to be slaves of Christ, nevertheless endorse obedience of slaves, and say nothing to overthrow the socio-economic order, in fact, extending the concept of slavery as a general principle. Lastly, there is the book of Philemon.

The situation regarding Philemon has been variously appraised. The traditional view is that Onesimus was a runaway slave, and the major alternative is that, though a slave, he was perhaps not a runaway at all, but was nevertheless sent back with a letter to his master, whether that be Philemon or Archippus (see Pearson 1999; cf. Callahan 1997: 1–19, but who makes no attempt at a linguistic or CDA analysis of Philemon). According to either view, the letter is concerned with a complex social situation. The situation involves a number of figures (see Petersen 1985: 89–199; Wilson 1992). Mentioned in the letter are 11 characters. A preliminary analysis shows that there are at least five categories of participants. First are the writers Paul and Timothy, second the recipients Philemon, Apphia and Archippus (note also mention of the church in ‘your’ house), and third Onesimus, about whom there is discussion and who appears to figure as in some way an intermediary between the writers and the recipients. These appear to be the primary participants in the letter. There are two further groups, however. These include God our Father (Phlm. 3) and the Lord Jesus Christ (Phlm. 3, 25; cf. vv. 4, 5, 6, 8, 9, 16, 20, 23). These are secondary participants in that they are not directly responsible for addressing, receiving or responding to words or actions, but, as the references make clear, they are highly present figures nonetheless. The reference is to the risen Lord, linked with God the Father. Less significant secondary participants are Epaphras,

36  

Mark, Aristarchus, Demas and Luke, whom the authors call ‘my fellow-workers’ (Phlm. 24), and the saints or holy ones, mentioned in parallel with the Lord Jesus in Philemon 5.

How does one chart these relationships? Even among the primary participants there are several further distinctions. One is that although the letter prescript includes two people, only the first-person singular, not the first-person plural, is used from Philemon 4 on. It is clear that Paul is the primary speaker and writer of the letter, referring to himself by name (Phlm. 9, 19), and as an old man/elder (Phlm. 9). Noteworthy also is that, though three people are addressed in the adscript, reference is to the second-person singular from Philemon 2 on. This singular reference has led some to posit that Archippus was the recipient, according to Knox’s well-known theory, although most interpreters take it as Philemon on the basis of his name being listed first. If this is the case, then the social relations seem to be reducible to three major ones—Paul to Philemon, Paul to Onesimus, and Philemon to Onesimus.

Thus the social problems can be boiled down to two highly intertwined ones. One is the complex social convention of slavery, and the other is the personal dynamics of the three primary participants within a church context (which is a distinct type of social organization). One cannot speak of one without involving the other. Whether a runaway or not, Onesimus appears to be a slave of Philemon and has for some time been with Paul. Paul is sending Onesimus back to Philemon, but with the desire that Onesimus be restored to him, because he knows that Philemon will not be able to provide the same service. Along the way a third, economic relationship is introduced, but that one grows out of the other two and is almost incidental to it. b. Power Relations Are Discursive This principle of interpretation states that the human relations that exist are ones that involve power, and that this power can be, and often is, exercised through language. The letter form is, in fact, such a discursive form of power, or at least has the potential for it. The power resides literally in the hands of the writer of the letter, but it can extend much further than that. There are two major elements to the exercise of power through the letter form in Philemon. The first is the way in which people are invoked in the letter. It has been noted above that even though Paul and Timothy are the purported authors of the letter, it appears that Paul’s is the only voice heard. Why then is Timothy mentioned? This question has been asked many times in Pauline studies, and given various answers. One answer that has not been given very frequently is that it is a linguistic move of power on Paul’s part. That is, Paul speaks not simply for himself but with the support and endorsement of another leader of the Christian missionary movement. Even if the issue is one personal to Paul, he extends it beyond his personal sphere by involving Timothy. The invocation of others at the end of the letter conforms to a typical Pauline convention of including closing greetings from other Christians (see Rom. 16:21–23), but might well serve a similar purpose, that is, of letting the recipients know that this is not simply an issue between the two parties, but one known to others who take an interest as well. Likewise with those he addresses. Even though the issue Paul has is with Philemon (or Archippus about Philemon), he invokes three people, plus the church in ‘your’ house. None of them is invoked by name again, with the primary recipient only referred to as ‘you’ (second-person singular). One reason for this, cited by commentators, is that Paul wishes to involve several close to the recipient in his social relations, here members of the local congregation.

The second way that power is exercised is in terms of how the various participants are characterized. Of course, all of the characterizations are by the author of the letter, and so have a controlling function on how people are viewed and thought of, all in linguistic terms. Paul’s characterization of himself falls into four different categories. First, he is the writer of the letter, which he depicts in two ways, first in Philemon 1 at the beginning of the prescript and also in Philemon 19, where he refers to the fact that he is writing the letter with his own

37  

hand. Whereas professional scribes were used in the ancient world, often in contexts where the letter ‘writer’ was illiterate, in this context it appears legitimate to extrapolate that the personal involvement of the writer at the end of the letter—taking the pen from the scribe’s hand and writing several words on his own—is a move of power. The author is not dependent upon others, even upon a scribe, but is personally supervising the writing task, and the wording is therefore his, including the way in which he characterizes himself and others.

Most of Paul’s characterizations of himself and those he writes to and about, however, are more overtly linguistically framed. For example, Paul characterizes himself in terms of being a fellow-worker with Philemon (Phlm. 1), a brother of Apphia (Phlm. 2), and a fellow-soldier of Archippus (Phlm. 2), and as a brother of the recipient (Phlm. 7, 20), as well as a partner (Phlm. 17). This language places Paul on the same level as that of his recipients, as he does the secondary participants at the end of the letter (Phlm. 24). Paul also uses language that has a hierarchical element to it, however. At several places, he asserts his superiority to the letter’s recipient. For example, in Philemon 8, he says that he could order his recipient to do as he says, but that he chooses not to. He switches instead to a form of appeal (Phlm. 9, 10), placing himself in a position of dependence upon the favour of his recipient, to whom he appeals for a favour (Phlm. 14), characterizing himself as either an old man or elder and a prisoner (Phlm. 1, 9). Paul establishes a familial link with Onesimus the slave (Phlm. 16) by addressing him as son (Phlm. 10), but hoping that the recipient will have a relationship with him as brother (Phlm. 16). He also introduces an economic metaphor to characterize relationships, noting that, whereas there may be one of debtor and debtee between the recipient and Onesimus (Phlm. 18), the recipient himself is a debtor to Paul (Phlm. 19). Thus Paul introduces economic, personal, and more broadly social relations into his involvement with the recipients of the letter.

These linguistically constituted relations may at first appear to be random, but I think that further examination shows that they are far from that. The frame of these relations is that of the power structure implied by the letter form. Paul begins the letter from a position of strength. Even though he characterizes himself as a prisoner in Philemon 1, it is as a prisoner of Christ Jesus, who is invoked throughout the letter. For Paul, it would appear that this is a position of strength, as if the Lord Jesus were observing all that is said and hoped to be done in the letter. Within the context of his taking the linguistic initiative in writing the letter, Paul begins by characterizing his recipients as his equals, continued through the thanksgiving portion of the letter. However, once the thanksgiving is concluded, Paul begins the body of the letter, responsible for the content of his communication, with a very clear invocation of power. No matter how quickly he denies this position, note that it is grounded in his being ‘in Christ’, and sets the tone of his appeal. He appeals to his recipient to act on the basis of love, but it is clear that a proper response is meant to be seen as consonant with Paul’s desire as one in authority in Christ. Thus when Paul refers to himself as an old man or elder and as a prisoner in Christ Jesus, these references seem to function on two levels. On the literal level he may well have been an old man and in prison, but he is also the one who represents Christ’s authority to his recipient (notice that he is in chains for the gospel; Phlm. 13), and he is making a request for Onesimus, who is characterized first as a son. Paul then develops and makes clear how useful this ‘son’ has been and would be in the future. Paul’s language is consciously geared to place Onesimus and the letter’s recipient on the same level, one of inferiority to the apostle, implied from the outset by the way he uses the letter itself. Thus he speaks of Onesimus as a substitute for the letter’s recipient, both of them as brothers, and both of them as debtors. Part of the strategy, however, is to get the recipient to do what the apostle desires willingly, but without leaving the possible responses open to debate, and thus he mixes language of equality, such as brother, with highly allusive yet persuasive language about the recipient’s own debt to Paul for his very self (Phlm. 19). This is probably a reference to the recipient having become a convert through Paul’s ministry, indicating another

38  

commonalty between Onesimus and the recipient. Paul closes the body of the letter by an overt exercise of his authority, first in his expressing confidence in the recipient’s obedience, secondly in his reference to the recipient surpassing his request, and thirdly in his desire for preparation of a room (Phlm. 22). One might rightly raise the question of how it is that Paul can invoke confidence regarding obedience, when he has not issued any direct commands for the recipient to perform apart from welcoming Onesimus. Yet the message is clear: he wishes to have Onesimus returned to him, implying if not manumission at least an effective ending of his being the recipient’s slave. The request for the room might well be merely a formal close, yet the thought of Paul’s personal presence to check on whether the recipient has acted on Paul’s requests gives such requests the force of commands. c. Discourse Constitutes Society and Culture According to Fairclough and Wodak (1997: 273; citing Fairclough 1992), ‘It is useful to distinguish three broad domains of social life that may be discursively constituted, referred to as representations, relations and identities for short: representations of the world, social relations between people, and people’s social and personal identities.’ Each of these merits brief comment, although one might well subsume the first under the third.

Representations of the world are linguistically constituted in Philemon in terms of how the writer chooses to describe those about whom he writes. There is a sense in which this language is responsive to the world outside of the text, but another in which the text constitutes this world. For example, Paul uses the language of slave and prisoner. It appears to be the case that Onesimus is a literal slave, due to the play on words in Philemon 16. Philemon 16 characterizes Onesimus as a slave, depicting it as a low-level position from which his transformation as brother has in some way liberated him. His status is not simply linguistic, although it at least in part can be constituted that way. The same is true of Paul’s being a prisoner. On the one hand, he appears to be literally in chains and in need of help (Phlm. 10), but he is also a prisoner of Christ Jesus (Phlm. 1, 9), perhaps indicating literal imprisonment on the basis of his allegiance to Christ but also his own way of viewing the relation that exists.

Social relations between people are also constituted by discourse. There is much language of partnership, brotherhood and fellowship contained within this letter. Some of the relations may be literal, but most of them appear to be linguistically constituted, and thereby open to linguistic reconstitution on the basis of changed characterizations. Onesimus has now become a brother of the letter’s recipient, and both are in some sense debtors, which places them on a similar social level in relation to Paul.

People’s social and personal identities are also constituted linguistically. For example, Onesimus was a slave in so far as Roman law was concerned, but his having been in contact with Paul led to a change in his identity. Paul constituted him as a son, and as a brother of the letter’s recipient. This apparently not only constituted his changed status linguistically but effected a changed status within the Christian community and in relation to the letter’s recipient by means of such a characterization. The letter’s recipient also had his identities both socially and personally reconstituted linguistically. Whereas he had once been the slave’s owner, now he was depicted as his brother, and whereas he once probably thought of himself as Paul’s equal, perhaps even his superior because of economic status and security, he found himself his debtor. d. Discourse Does Ideological Work According to Fairclough and Wodak (1997: 275), ‘To determine whether a particular (type of) discursive event does ideological work, it is not enough to analyse texts; one also needs to consider how texts are interpreted and received and what social effects they have’. The book of Philemon seems to be predicated upon a conflict of social values, whether one takes the

39  

traditional or the revised estimation of its circumstances. According to the traditional view, the primary issue is slavery, in which Paul is calling for Philemon to, essentially, manumit Onesimus, so that Onesimus can return to Paul, his debts forgiven and his having been incorporated into the believing community as a brother, to serve Paul while he is in prison. According to the revised view, the primary issue is ‘the relationship between an imprisoned apostle and his free, wealthy, “partner”, Philemon’ (Pearson 1999), yet still revolving around the fortunes of the slave Onesimus. In either case, there is a conflict of ideology, revolving around differing conceptions of slavery and Church—social relations, the only real difference being which gives one priority over the other. Paul’s ideological basis seems to be that Church business, of which his own business is a clear and predominant part, takes priority over other personal interests, including the issue of slavery. One does not have to argue that Paul is calling for a slave revolt, since he clearly is not, or even that he is calling for an adjustment in slave-master relations within the Church (as I think he tacitly is), to note that he is instigating a small-scale and individual revolution. Whether Onesimus was a runaway slave or not, for Paul his priorities as apostle took precedence over the institution of slavery, of which this is an instantiation, and over the personal interests even of a potentially wealthy and influential member of the congregation. It is not hard to see that if such a set of ideological priorities were generalized, it would make it difficult for the institution of slavery to co-exist within a Christian context.

The history of interpretation of Philemon, however, is not one of the Christian church wholeheartedly endorsing such an ethic toward slavery. In that sense, the interpretation of the book was not consonant with its original discursive purpose. This is not the place to note the history of interpretation of Philemon, but there are a number of general points to note. One is the tradition that Onesimus was, in fact, manumitted and became a bishop. An interesting story, it is hard to prove that this Onesimus is the Onesimus of this letter. More to the point, however, is the fact that this letter was retained and eventually incorporated into the Pauline and then the New Testament canon. This indicates a plausible scenario that the scene as Paul has linguistically dictated it was fulfilled in ways at least approximate to his purpose, and that the letter was preserved as an example of the useful and productive correspondence of the apostle. e. Discourse Is Historical What is meant that discourse is historical is that ‘Discourse is not produced without context and cannot be understood without taking the context into consideration’ (Fairclough and Wodak 1997: 276). The CDA practitioner must rightly appreciate that

utterances are only meaningful if we consider their use in a specific situation, if we understand the underlying conventions and rules, if we recognize the embedding in a certain culture and ideology, and most importantly, if we know what the discourse relates to in the past. Discourses are always connected to other discourses which were produced earlier, as well as those which are produced synchronically and subsequently. In this respect, we include intertextuality as well as sociocultural knowledge within our concept of context (Fairclough and Wodak 1997: 276; cf. Reed 1997a: 50, 55).

Despite much use of the term intertextuality among biblical scholars, the term here is more expansively interpreted in terms of how texts are connected to each other, not being confined to such things as direct quotations of or allusions to other texts, such as the Old Testament.

As noted above, the notion of context has usefuly been expanded by CDA, and could benefit from further expansion in terms of New Testament studies, perhaps helping to overcome the impasse that has been reached in much biblical studies over the relatively limited specific background material known. A firm procedure has not been developed in terms of CDA, because of its virtually exclusive use in contemporary linguistic contexts, to

40  

reconstruct ancient historical backgrounds. For a book such as Philemon, one would need to have quantifiable historical knowledge of a number of important areas, including Roman slavery, the composition and structure of the early Church, including local house churches, Paul’s missionary strategy and those involved with him in it, and the organization of this early Church. Most of these have already been explored, even in terms of Philemon. The difficulty seems to be in judging what items of the larger context of culture are immediately relevant in discourse analysis. One of the benefits of discourse analysis has been the overt effort to quantify data, and thereby provide controls on an otherwise potentially unregulated use of background material. The context of culture, as broad as it is in theory, appears to be best incorporated into CDA in terms of those elements of the context of culture that are also relevant to the context of situation, including issues of genre and register, as well as field of discourse.

The intertextual dimension of study of Philemon might well be more productive at this stage, however. Apart from a few linguistically based plays on words, such as the meaning of Onesimus’s name (Phlm. 10), and other Christian conventions, there is not much that links the letter to the past, including not much that links it to the Old Testament. More profitable is examination of the Graeco-Roman side. For example, Paul’s use of the Hellenistic letter form might well benefit from further examination. Philemon has a four-part structure. This structure departs from the traditional three-part letter form and many other of Paul’s own letters, which were based around a five-part structure. Paul begins much like contemporary letters of the period, but he has expanded the word of greeting. Schnider and Stenger (1987: 1–41) characterize Paul’s letter openings as being oriental, but the parallels simply are not germane. It may be true that there was some influence of the Hebrew shalom on his conception of greeting, but more likely is that he has adapted the standard Hellenistic letter opening from ‘greetings’ to ‘grace and peace’, with a play on the Greek root. Paul’s thanksgiving is uniquely his, and only bears superficial resemblance to the health wish of other Hellenistic letters. This section has taken on its own proportions. In Philemon, Paul uses the thanksgiving to serve his linguistic discursive purposes, by separating his egalitarian words of greeting from the body of the letter, where he utilizes a set of variegated hierarchical words, beginning with bold words of authority and position. The body of the letter is very similar to the personal Hellenistic letter form, including discussion of personal relations and financial matters. The major difference is that these are all within the context of the church (Phlm. 2), with an appeal throughout to the authority of Jesus Christ. The letter closes with a wish for a visit, reflecting the desire for the apostolic presence. In placement and use of this convention, the letter is similar to others, although the use here is clearly designed with Paul’s linguistic purpose in mind. He then closes with greetings and a grace wish, a unique feature of the Pauline letter. f. The Link between Text and Society Is Mediated The linkage between text and society is mediated by orders of discourse. There is no direct linkage between the social values and a text except as it is mediated through the medium of discourse, and this discourse draws upon all of the available resources shaped by its underlying ideology (or in this case, theology) in its attempt to effect social change through discursive practice. I have already noted a number of the most readily apparent features used in doing this.

There are a number of other features of mediation worth noting, however. One is that the personal letter form, in which people write to each other about personal matters, often involving commendation or financial matters, is here used for this purpose but also for a larger theological purpose. This letter is not only about Onesimus, it is also about Philemon and his relation to Paul. Involved, therefore, is a complex social dynamic regarding the organization and function of the early Church. Paul is utilizing this letter form in an attempt to

41  

establish and maintain a certain position of authority, and the document has ended up being part of canonical Scripture for certain groups. Many CDA practitioners would, on the basis of their presupposition regarding the oppressed and the oppressors, want to resist the outcome of this scenario. This discursive situation, however, illustrates the need to evaluate every discourse on its own terms in so far as power and language are concerned. On the one hand, Paul is certainly attempting to utilize his linguistic capacity to effect certain changes upon his letter recipient. From this standpoint, he might well be seen as the one exercising power and authority, by means of language. On the other hand, however, Paul is a prisoner, a situation in the Roman world not to be taken lightly. From that standpoint, he is clearly the oppressed, one who does not have power and is in no position to exercise it, apart from his ability to communicate. He is clearly convinced that it is in his, and probably Onesimus’s, best interests for Onesimus to return to him to serve him, as well as being in the best interests of his letter’s recipient. One cannot be certain of the outcome, but retention of the letter seems to argue similarly.

Paul uses other forms of discourse to mediate the link between text and society. This includes his invocation of a number of different characterizations of people, drawing upon a vast repertoire of possible images. These include those of family relations (e.g. brother, sister, son, father), relations in the workplace (e.g. fellow-workers, servants), legal relations (e.g. prisoner), socio-economic relations (e.g. slave, debtor), and even military relations (e.g. fellow-soldier) to characterize the linguistic situation. g. Discourse Analysis Is Interpretative and Explanatory Having engaged in the CDA above, one is more clearly able to engage in interpretation and explanation of the text. The way the terms ‘interpretation’ and ‘explanation’ are used in CDA requires a brief comment. Interpretation involves placing the text within its meaningful structures, often construed in terms of power structures and ideological framework. Explanation involves the further step of reflecting upon why the text is construed the way it is, and what it means to be interpreted in this way. For much CDA, this involves going further and engaging in a self-reflective and self-critical proactive stance (Fairclough and Wodak 1997: 279). This might be thought desirable in certain biblical and theological contexts, but does not seem to be a necessary stage, without further development of a hermeneutical stance so far undeveloped in CDA. At this stage, I think that it is best to view CDA as a linguistic criticism focusing upon the text, no matter what other stages may develop with further thought about its hermeneutical implications.

With this framework in mind, Philemon can be observed as a letter written in terms of Paul seeing himself as one who has both a position of authority and a set of needs that he thinks can be best served by Onesimus. Realizing that Onesimus is a slave, and hence is the ‘property’ of another, he writes a letter, a potentially powerful form of discourse, to attempt to effect a change in the social relations. These social relations, as noted above, exist on several levels, including legal, economic and ecclesiastical. Paul is attempting to effect change in all three. As a result, he writes a letter in which he positions himself with regard to his own characterization of himself, and his characterization of others, including the recipient and a number of secondary figures. As a backdrop he invokes Jesus Christ as the ultimate witness to his linguistic actions, but he also feels empowered to create shifting discursive relations that draw in the various levels of social relations.

This is one interpretation of the letter. There may well be others as well. Explanations of Paul’s behaviour can and have varied also. Some have noted, and roundly condemned, such a manipulative stance by the apostle. He appears to have his goal of securing the services of Onesimus so much in mind that he is willing to engage in manipulation, perhaps verging on deception, in order to accomplish that goal. Perhaps there is an underlying thought that the apostle should not need to engage in such discursive practices, when a simple request should

42  

suffice to get Onesimus released to him. Perhaps it would have in another context. I think that there are larger considerations, however, that dictated Paul’s linguistic behaviour. Perhaps unconsciously, Paul seems to have realized that there was something incompatible with Onesimus continuing as a slave of Philemon while serving Paul. For Paul, at least in this letter, it appears that service to others should be voluntary, with one’s being a prisoner of Christ not a mandated but a voluntary state of affairs, one that cannot be imposed by another human. In effect, Paul is re-examining the institution of slavery within a Christian context, and finding that it is not compatible with at least his ethos. Therefore, he creates a linguistic environment in which it is proper for the letter’s recipient to release Onesimus from his legal bondage in order to submit himself voluntarily to service to Paul, all within the context of all of the participants functioning ‘in Christ’. h. Discourse Is a Form of Social Action One of the goals of CDA, in fact, for some the goal, as noted above, is ‘to uncover opaqueness and power relationships’ (Fairclough and Wodak 1997: 279). I hope that it can be used for other goals as well, such as uncovering the full range of linguistic means available in the creation of various types of discourse, especially as it is concerned with various social relations, including those of power, but not be confined to such discursive practice. There is the further instructive element of evaluation, as one draws on the wider context of culture in order to establish not only the fact of, but the reasons for, the kinds of discursive practice within a text. This should take the form of evaluation of the original author and recipients and their linguistically enfranchised relationships, but could also be extended to examination of how others have received and interpreted the discourse. In one sense, CDA of biblical texts is bound to be more limited than CDA of contemporary texts, since the kinds of texts analysed often have greater immediate impact on a particular contemporary audience. An analysis of a biblical text such as this may well, in another context with another text, have more to say to contemporary ecclesiastical practice. For example, an examination of passages on the role of women in the Church, or on the practice of certain spiritual gifts, may well be enlightened through CDA, and result in interpretative explanations that effect change in contemporary Church practice. I would not wish to fall victim to the criticism above, however, in saying that one ought to predecide these issues, since they are complex ones that require critical analysis before arriving at conclusions, rather than the other way around.

This interpretation of Philemon has, I hope, been able to lay bare some of the dynamics of a text that is full of interesting and potentially problematic linguistic elements. These include the nature and type of participants, their role relations in terms of social, political, legal and ecclesiastical structures, the nature and kind of request being made by Paul, and the means by which he goes about linguistically effecting such actions. At this stage of analysis, I am less concerned about proving this interpretation to be right, than I am about bringing to the fore the kinds of linguistic issues that need to be discussed in an analysis such as this.

5. Conclusion

In what sense is CDA critical? It is clear that it needs to be far more self-critical of its own methods and assumptions. Below or beyond or behind the rhetoric, there is a method that raises meaningful interpretative questions regarding texts. If these questions were brought more prominently to the fore, New Testament scholars might be inclined to give the method more serious attention, even if they do not share the same ideological commitments. I have tried to distinguish the ideological bias from what is essential in the method. In that sense CDA does offer significant insights that develop those already gained through other forms of linguistic analysis. If one is able to restrict CDA to its exegetical rather than its hermeneutical potential, a more promising way forward, at least for the time being, seems to be open to it. Its

43  

strengths seem to be those of bringing to the interpreter’s attention linguistic data that perhaps would not be noticed, or would even be downplayed in another context. These are especially concerned with describing the social relations that exist in discursive contexts. A further strength is its desire to go beyond description of linguistic practice and to engage in linguistic evaluation and explanation. It is not enough simply to say that this is the case, but one should evaluate why it is the case and what are the results of this case being as it is. The Pauline letters, even such a short one as Philemon, reflect complex linguistic dynamics that would appear to merit such critical discourse analysis.

BIBLIOGRAPHY

Brown, G., and G. Yule 1983 Discourse Analysis (CTL; Cambridge: Cambridge University Press). Caldas-Coulthard, C.R., and M. Coulthard (eds.) 1996 Texts and Practices: Readings in Critical Discourse Analysis (London: Routledge). Callahan, A.D. 1997 Embassy of Onesimus: The Letter of Paul to Philemon (Valley Forge, PA: Trinity Press International). Dijk, T.A. van 1996 ‘Discourse, Power and Access’, in Caldas-Coulthard and Coulthard 1996: 84–104. Fairclough, N. 1989 Language and Power (London: Longman). 1992 Discourse and Social Change (Cambridge: Polity). 1995 Critical Discourse Analysis: The Critical Study of Language (London: Longman). 1996 ‘Technologisation of Discourse’, in Caldas-Coulthard and Coulthard 1996: 71–83. Fairclough, N., and R. Wodak 1997 ‘Critical Discourse Analysis’, in T.A. van Dijk (ed.), Discourse Studies: A Multidisciplinary Introduction. II. Discourse as Social Interaction (London: Sage): 258–84. Fowler, R. 1996 ‘On Critical Linguistics’, in Caldas-Coulthard and Coulthard 1996: 3–14. Fowler et al. (eds.) 1979 Language and Control (London: Routledge). Hodge, R., and G. Kress 1979/1995 Language as Ideology (London: Routledge). Kress, G. 1996 ‘Representational Resources and the Production of Subjectivity: Questions for the Theoretical Development of Critical Discourse Analysis in a Multicultural Society’, in Caldas-Coulthard and Coulthard 1996: 15–31.

                                                            CTL Cambridge Textbooks in Linguistics 

44  

Moulton, J.H. 1909 ‘New Testament Greek in the Light of Modern Discovery’, in S.E. Porter (ed.), The Language of the New Testament: Classic Essays (JSNTSup, 60; Sheffield: JSOT Press): 60–97. Pearson, B.W.R. 1999 ‘Assumptions in the Criticism and Translation of Philemon’, in S.E. Porter and R.S. Hess (eds.), Translating the Bible: Problems and Prospects (JSNTSup, 173; Sheffield: Sheffield Academic Press): 253–80. Petersen, N.R. 1985 Rediscovering Paul: Philemon and the Sociology of Paul’s Narrative World (Philadelphia: Fortress Press). Popper, K. 1974 Unended Quest: An Intellectual Autobiography (London: Fontana). Porter, S.E. 1995 ‘Discourse Analysis and New Testament Studies: An Introductory Survey’, in S.E. Porter and D.A. Carson (eds.), Discourse Analysis and Other Topics in Biblical Greek (JSNTSup, 113; SNTG, 2; Sheffield: Sheffield Academic Press): 14–35. Reed, J.T. 1995 ‘Modern Linguistics and the New Testament: A Basic Guide to Theory, Terminology, and Literature’, in S.E. Porter and D. Tombs (eds.), Approaches to New Testament Study (JSNTSup, 120; Sheffield: Sheffield Academic Press): 222–65. 1996 ‘Discourse Analysis as New Testament Hermeneutic: A Retrospective and Prospective Appraisal, JETS 39: 223–40. 1997a A Discourse Analysis of Philippians: Method and Rhetoric in the Debate over Literary Integrity (JSNTSup, 136; SNTG, 3; Sheffield: Sheffield Academic Press). 1997b ‘Discourse Analysis’, in S.E. Porter (ed.), Handbook to Exegesis of the New Testament (NTTS, 25; Leiden: E.J. Brill): 189–217. Schnider, F., and W. Stenger 1987 Studien zum neutestamentlichen Briefformula (NTTS, 11; Leiden: E.J. Brill). Wilson, A. 1992 ‘The Pragmatics of Politeness and Pauline Epistolography: A Case Study of the Letter to Philemon’, JSNT 48: 107–19. Wodak, R. 1996 ‘The Genesis of Racist Discourse in Austria since 1989’, in Caldas-Coulthard and Coulthard 1996: 107–28.  

                                                            JSNTSup Journal for the Study of the New Testament, Supplement Series 

SNTG Studies in New Testament Greek 

JETS Journal of the Evangelical Theological Society 

NTTS New Testament Tools and Studies 

JSNT Journal for the Study of the New Testament 

45  

THE USE OF ANNOTATED CORPORA FOR NEW TESTAMENT DISCOURSE ANALYSIS: A SURVEY

OF CURRENT PRACTICE AND FUTURE PROSPECTS Matthew Brook O’Donnell

1. Introduction

Discourse analysis of the New Testament is, by necessity, primarily a text-based discipline. With only limited access to the context of situation and culture in which a particular discourse took place, it is difficult to make direct application of many of the studies that focus on the communicative and sociological aspects of discourse. This is not to deny the importance of attempting to reconstruct the context of culture and situation for the texts of the New Testament. However, it does mean that a discourse analysis of a New Testament book or section must begin with and focus primarily upon the textual component. At this level the discourse analyst is concerned with discovering patterns of language use that may reveal the structure and nature of the interpersonal and ideational functions in the text. This requires systematic and close study of the text.

Traditionally, texts have been ‘stored’ in printed form, and the text on the printed page has served as the object of analysis. Increasingly texts are being stored in machine-readable form, accessible by computer. There are a number of advantages of a machine-readable text over its hard-copy equivalent. For example, it is possible to create a concordance list of all the occurrences of a particular word form or phrase. The TLG (Thesaurus Linguae Graecae) database is perhaps the most well-known example of a large, machine-readable corpus of Greek texts. It can be used with various concordance programs to search for selected words or combinations of words. There are, however, considerable limitations to the use of such corpora, which simply contain the text without any lexical or morphological information (for general grammatical work as well as discourse analysis). This is particularly the case with an inflected language like Hellenistic Greek. For instance, to find all the occurrences of the word λόγος in a selected text or group of texts, it is necessary to perform individual searches for each inflected form (e.g. λόγος, λόγου, λόγῳ, λογόν, λόγοι, λόγων … etc.) and then to merge the results. In some cases it is possible to use a wild-card character and search for λόγο*, but this would not only find all the forms of λόγος (apart from dative singular and genitive plural), but also certain forms of other words such as: λογογράφος, λογομαχέω and λογοποιία. It is possible to overcome this problem by including the lexical form of each word in the text in a special tag following the word, e.g. λόγων <λόγος>. Then a search can be performed using the lexical tags.

There are many different kinds of information that can be added to the basic text to allow more detailed and specific analysis to be carried out. This study discusses the various levels of annotation that can take place and argues that it is important for these levels be kept distinct. The specific focus of this study is on the use of annotated corpora for the discourse analysis of the New Testament. It seeks to investigate (1) how annotated corpora have been used for discourse analysis, (2) what levels of annotation are necessary to provide the information required for discourse analysis and (3) methods for future development.

2. Corpus Annotation

a. What Is Corpus Annotation? In this study, I use the term corpus annotation instead of the perhaps more familiar term text tagging. It may appear as simply a terminological distinction, but there is an important difference. Tagging refers to the practice of appending a label or tag to individual words or

46  

word groups, such as a part-of-speech tag: word_N. Part-of-speech (POS) or grammatical tagging is perhaps the most common form of tagging now used. It is particularly useful when one wishes to disambiguate a word like run, which could be either a verb or a noun. An example of grammatical tagging of a Greek word would be: τήν_dafs. This follows the Friberg coding1 and signifies that τήν is an article (d), in the accusative case (a), with feminine gender (f), and singular number (s). Grammatical tagging is only one possible type of information that can be added to words or groups of words. It would also be possible to add semantic information about words, such as which semantic field they belong to, or to add morphological information relating to how the word has been formed. The addition of such tags to a machine-readable corpus is one example of corpus annotation. There are many other kinds of information that can be added to a text as well, such as section headings and dividers, information about date, authorship, provenance and genre, and so on. This kind of annotation requires more than a simple label or tag following a word or section. It is more accurate to speak of corpus annotation, which may include the use of tags, instead of using the term text tagging.

This discussion raises the question of definition—what exactly is corpus annotation? Perhaps first one must ask the question: what is a corpus and what is the distinction between a text and a corpus? Without getting too specific, a text can be considered a written discourse that, in its current form, is considered to be a unit. Thus, the Gospel of Matthew is a text, as is Romans and the first epistle of John. However, those three texts together do not necessarily constitute a corpus. Biber, Conrad and Reppen assert that ‘a corpus is not simply a collection of texts. Rather, a corpus seeks to represent a language or some part of a language.’2 That is, an intentional grouping of particular texts, according to specific criteria, forms a corpus.3 Grouping together the 13 letters in the New Testament that contain an attribution to Paul could be considered a corpus of Pauline letters. Extracting the purported sayings of Jesus from the Synoptic Gospels, and grouping them together, would constitute a corpus of the sayings of Jesus. Following this understanding of what constitutes a corpus, there is no reason why Matthew, Romans and 1 John could not form a corpus, if these three texts were representative of some particular language feature or variety. McEnery and Wilson provide the following definition of a corpus:

A corpus in modern linguistics, in contrast to being simply any body of text, might more accurately be described as a finite-sized body of machine-readable text, sampled in order to be maximally representative of the language variety under consideration.4

Corpus annotation is then the addition of linguistic information to a text or group of texts

that meets the definition of a corpus. Or, as defined by Leech, ‘[corpus annotation is] the

                                                            1 See B. Friberg and T. Friberg, The Analytical Lexicon of the New Testament (Grand Rapids: Baker 

Book House, 1994). 

2 D. Biber, S. Conrad and R. Reppen, Corpus Linguistics: Investigating Language Structure and Use 

(Cambridge: Cambridge University Press, 1998), p. 246. 

3 For a discussion of some of the criteria for the construction of a corpus, see J. Sinclair, Corpus, 

Concordance, Collocation (Oxford: Oxford University Press, 1991), pp. 13–26. Also, Biber, Conrad and 

Reppen, Corpus Linguistics, pp. 246–53; A. McEnery and A. Wilson, Corpus Linguistics (ETEL; 

Edinburgh: Edinburgh University Press, 1996), pp. 21–23, 149–51. 

4 McEnery and Wilson, Corpus Linguistics, p. 24. 

47  

practice of adding interpretative, linguistic information to an electronic corpus of spoken and/or written data’.5 He includes the word interpretative in his definition to indicate that when a corpus is annotated, the annotators must make certain choices, guided by their understanding of the meaning of the word forms and larger textual units in the corpus.6 As indicated above, there may be certain pieces of information, such as a text’s author and date, that might be added to a text. This type of information may not strictly be classified as linguistic information; however, it can be included as a type of annotation. b. Types of Annotation and their Discourse Levels It is necessary to distinguish between the different levels at which annotation can take place. These levels have often been equated with the traditional areas of linguistic study, that is, phonological annotation with phonology, grammatical or part-of-speech annotation with morphology, syntactical annotation relating to syntax, semantic annotation with lexical semantics and pragmatic annotation with pragmatics. The most recent collection of essays on corpus annotation, Corpus Annotation: Linguistic Information from Computer Text Corpora,7 edited by a number of the pioneers of the field, chooses to discuss the different kinds of annotation that can be added to a corpus in terms of these traditional linguistic levels. It also discusses the possibility of other levels of annotation, such as stylistic annotation to encode stylistic information and, for spoken corpora, prosodic annotation.

Given the purpose of this essay and the scope of this volume, it may also be helpful to tie the different varieties of annotation that can take place to the different levels (or units) of a text recognized in discourse analysis. Beginning with the phoneme (or grapheme for written texts) as the smallest unit, then the morpheme, as the smallest unit of meaning, one moves to the word, then the phrase or word group, then the clause or sentence and next the paragraph (or pericope) and finally the whole discourse.8 Reed refers to these different levels as co-textual levels of discourse.9 He points out the importance of recognizing and defining these different levels of discourse, because,

                                                            5 G. Leech, ‘Introducing Corpus Annotation’, in R. Garside, G. Leech and A. McEnery (eds.), Corpus 

Annotation: Linguistic Information from Computer Text Corpora (London: Longman, 1997), pp. 1–18 

(2) (emphasis his). 

6 ‘By calling annotation “interpretative”, we signal that annotation is, at least in some degree, the 

product of the human mind’s understanding of the text. There is no purely objective, mechanistic 

way of deciding what label or labels should be applied to a given linguistic phenomenon’ (Leech, 

‘Introducing Corpus Annotation’, p. 2). 

7 Garside, Leech and McEnery (eds.), Corpus Annotation. 

8 See S.E. Porter, Idioms of the Greek New Testament (Biblical Languages: Greek, 2; Sheffield: 

Sheffield Academic Press, 2nd edn, 1994), pp. 298–300, who uses a pyramid to display these 

different levels, beginning at the base with the word and culminating at the pinnacle, which 

represents the entire discourse. See also J.T. Reed, A Discourse Analysis of Philippians: Method and 

Rhetoric in the Debate over Literary Integrity (JSNTSup, 136; SNTG, 3; Sheffield: Sheffield Academic 

Press, 1997), pp. 42–47. 

9 See Reed, Discourse Analysis of Philippians, p. 42. 

48  

Ancient Greek writers generally did not orthographically separate discourse units as is frequently the case in English with the use of spaces between words, punctuation between sentences and indentation between paragraphs.10

This is demonstrated by examining any of the New Testament papyri, such as P66, one of

the Bodmer Papyri, which contains sections of the Gospel of John. The following two lines were selected at random from a picture of one of the papyrus sheets:11 ολογοσουτοσονειπενοτιζητη σεταιμεκαιουχευρησεταικ The text contains no indication of where a clause or sentence begins or ends, or even where the boundaries of words occur. In this particular case, the first line begins with a word, the article ὁ. A knowledge of Greek is required to divide the text extract into words. It reads, ο λογος ουτος ον ειπεν οτι ζητησεται με και ουχ ευρησεται κ … (‘this word which he spoke that you will seek me and you will not find …’). The section comes from Jn 7:36, and is missing the first two words of the verse, τίς ἐστιν (‘what is’). Knowledge of the use of vowels in the papyri is also required, to be aware that αι is often used for ε.12 So ζητησεται is not a future middle/passive indicative third singular, but actually ζητησετε, an active voice second-person plural.

The above example is given to illustrate that, when working with Hellenistic Greek texts (and specifically the New Testament), the process of annotation, according to the definition given by Leech, must begin not at the word level with the addition of part-of-speech information, but at the character (grapheme) level. The table below lists the different kinds of annotation that can be added to a corpus of Greek texts, and the discourse level to which each type of annotation applies.

Type of Annotation Discourse Level

Orthographic Grapheme (Character)

Morphological / Phonological Morpheme, Word

Grammatical Word

Syntactical Clause, Sentence

Semantic Word, Clause, Sentence, Paragraph

Discourse Paragraph, Discourse

                                                            10 Reed, Discourse Analysis of Philippians, p. 42. 

11 For the picture the transcription was taken from, see V. Matin, Papyrus Bodmer II, Evangile de Jean 

chap. 1–14 (Geneva: Bibliotheca Bodmeriana, 1956), p. 80. Note that the font used in the 

transcription is not an exact representation of the hand used by the scribe of P66. The size of letters 

and spacing is much more irregular on the papyrus, for example, the two lines are the same length 

on the manuscript. 

12 See E. Mayser, Grammatik der Griechschen Papyri aus der Ptolemäerzeit. I. Einleitung und Lautlehre 

(ed. H. Schmoll; Berlin: W. de Gruyter, 2nd edn, 1970), p. 86, for this instance of itacism. 

49  

Pragmatic Stylistic

Clause, Sentence, ParagraphClause Sentence, Paragraph

Orthographic Annotation. The most basic level of annotation of a text to be included in a corpus occurs at the character or grapheme level. At first it may seem strange to refer to the transcription of characters as textual annotation, that is, ‘the addition of interpretative and linguistic information’.13 However, as noted above, the earliest manuscripts of the New Testament do not contain word, sentence or paragraph breaks. Also, they are often fragmentary in nature or contain lacunae of various kinds, leaving only parts of letters visible. It is, therefore, both an interpretative and linguistic activity to produce a version of one of these texts. It is beyond the scope of this paper to discuss the details of papyrological work and issues, such as, whether one should produce a minimally reconstructive (or diplomatic) text or attempt to fill in the gaps.14 But, the importance of beginning the task of the construction and annotation of a corpus of New Testament texts should begin with the manuscripts at the grapheme level.

A related issue is the electronic representation or coding of Greek characters. Given that binary bits are the basic data unit for computers, all language information must be represented using codes. There are certain standard codes that have been devised for this purpose, the most common being ASCII code,15 utilized by many computer operating systems. The ASCII code for the character ‘A’ is binary 1000001 (decimal value 65). To represent Greek characters, each letter of the Greek alphabet, accents, breathing marks and other characters are assigned an ASCII equivalent. Though there is basic agreement over which characters represent the majority of the alphabet, different fonts and coding systems use different ASCII characters for accents and breathing marks. For example, the SuperGreek font uses ‘J’ for a rough breathing mark and ‘j’ for a smooth one, whereas beta-code, used in the TLG database, uses ‘(’ and ‘)’ for these two characters. It is important, therefore, when orthographic annotation is taking place to make a decision on how Greek characters will be represented. This is an explicit interpretative decision, and therefore illustrates further that it is appropriate to use the term ‘annotation’ for this activity. Morphological Annotation. Further interpretative decisions are presented during the annotation process by the lack of word boundaries and clause and sentence divisions. The morpheme is the smallest linguistic unit of meaning within the text, and can communicate lexical or grammatical information. For example, the morpheme -ου added to a noun or adjective stem indicates grammatical case and number (genitive singular). Reduplication of the first letter of a verb stem, with an intervening ε, indicates that the verb is in the perfect tense form, thus communicating stative verbal aspect. Other morphemes communicate particular lexical ideas, such as κοιν, which seems to contribute the idea of ‘commonalty,

                                                            13 Discussing orthographic annotation, Leech says ‘orthographic annotation might seem to be a 

contradiction in terms—since, as we have seen, orthography represents the text, while annotation 

interprets the text linguistically. However, up to a point orthographic information can be 

interpretative, in distinguishing the linguistic functions of various devices on paper’ (‘Introducing 

Corpus Annotation’, p. 14). 

14 For an introduction to Greek papyrology and a discussion of papyrological method, see E.G. Turner, 

Greek Papyri: An Introduction (Oxford: Clarendon Press, 1968). 

15 ASCII stands for American Standard Code for Information Interchange. 

50  

sharing’ to the words built upon it: κοινός (common), κοινόω (call common, defile), κοινωνία (fellowship, partnership).16

It would be possible, and instructive for discourse analysis, to mark morphemes within a corpus and thereby to make it possible for the smallest units of meaning to be searched for and identified. Perhaps a basic distinction could be made between lexical and grammatical morphemes. The example below, from Gal. 2:9, illustrates this kind of annotation:

Text of Galatians 2:9

Morphologically Annotated Text of Galatians 2:9

ἐγὼ <mor tp = l ἐγὼ>

γὰρ <mor tp = l γὰρ>

διὰ <mor tp = l διὰ>

νόμου <mor tp = l νόμ><mor tp = gc ου>

νόμῳ <mor tp = l νόμ><mor tp = gc ῳ>

ἀπέθανον, <mor tp = l ἀπο><mor tp = gm έ>

<mor tp = l θαν>

<mor tp = gpn ον>

ἵνα <mor tp = l ἵνα>

θεῷ <mor tp = l θε><mor tp = gc ῷ>

ζήσω. <mor tp = l ζή><mor tp=a σ><mor tp = gpn ω>

Χριστῷ <mor tp = l Χριστ><mor tp = gc ῷ>

συνεσταύρωμαι <mor tp = l συν><mor tp = ga ε>

<mor tp = l σταύρο><mor tp = gpn μαι> Here a morpheme (a string cccc) is indicated by a tag <mor tp = x1 [x2] [x3] “cccc”>.

Where the tp attribute indicates the type of morpheme, either lexical (x1 = l) or grammatical (x1 = g). If it is a grammatical morpheme, then one or two further values can be included to indicate what grammatical feature is communicated by the morpheme: aspect (x2 = a), person (x2 = p), number (x3 = n), mood (x2 = m) or case (x2 = c). This example is given merely as an illustration of what morphological annotation might look like at the morpheme discourse-                                                            16 See Reed, Discourse Analysis of Philippians, p. 43. On the morphology of the Greek of the New 

Testament, see J.H. Moulton and W.F. Howard, A Grammar of New Testament Greek. II. Accidence 

and Word‐Formation (Edinburgh: T. & T. Clark, 1929); W.D. Mounce, The Morphology of Biblical 

Greek (Grand Rapids: Zondervan, 1994); D.A. Black, Linguistics for Students of New Testament Greek: 

A Survey of Basic Concepts and Applications (Grand Rapids: Baker Book House, 2nd edn, 1995), pp. 

53–95. For a general introduction to morphology, see P. Matthews, Morphology: An Introduction to 

the Theory of Word‐Structure (CTL; Cambridge: Cambridge University Press, 1974). Also J. Lyons, An 

Introduction to Theoretical Linguistics (Cambridge: Cambridge University Press, 1968), pp. 180–94. 

51  

level. It will be evident that such annotation would be highly labour intensive, and it is not clear that the benefit gained from such work would justify this effort.

The above table suggests that morphological annotation could also take place at the word discourse-level. By this I am primarily referring to addition of word boundaries into the text.17 Producing a reading/regularized edition of an unical manuscript requires that the editor make interpretative decisions about where the text should be divided to form words. Also it is necessary to add accents to the words that are formed by this process. The addition of accents could be considered as a form of prosodic annotation, as they are thought originally to have been connected to pronunciation. However, due to the lack of either a native Hellenistic Greek speaker or recordings of one, accents cannot reliably be considered prosodic annotation, and should instead be treated as orthographic annotation at the word level.18 The example below illustrates this process, with a portion of Codex Alexandrinus.19 The text from the codex has been orthographically annotated, that is, each letter represented with a specific ASCII character (here the Super-Greek font), and then decisions made as to where the word divisions should take place.

Ⲉⲓⲡⲉⲛⲇⲉⲟⲕⲥ̅̅ⲥⲓⲙⲱⲛⲥⲓⲙⲱⲛ εἶπεν δὲ ὁ κύριος Σίμων Σίμων

ⲓⲇⲟⲩⲟⲥⲁⲧⲁⲛⲁⲥⲉⲝⲏⲥⲁⲧⲟⲩ̈ⲙⲁⲥ ἰδοῦ ὁ Σατανᾶς ἐξητησατο ὑμας

ⲧⲟⲩⲥⲓⲛⲓⲁⲥⲉⲱⲥⲧⲟⲛⲥⲓⲧⲟⲛ· τοῦ σινιάσαι ὡς τὸν σῖτον·

ⲉⲅⲱⲇⲉⲉⲇⲉⲏⲑⲏⲛⲡⲉⲣⲓⲥⲟⲩ̈ⲛⲁ ἐγω δὲ ἐδεήθην περὶ σοῦ ἵνα

ⲙⲏⲉⲕⲗⲉⲓⲡⲏⲏⲡⲓⲥⲧⲓⲥⲥⲟⲩ μὴ ἐκλείπη ἡ πίστις σου· This has also involved the interpretation of the characters ⲕⲥ̅ ̅as a nomen sacrum for the

word κύριος, and the ⲉ in ⲥⲓⲛⲓⲁⲥⲉ has been represented with αι, resulting in the aorist infinitive form σινιάσαι.20

Brief mention should also made here of how variant readings in different manscripts should be indicated in a corpus of New Testament texts. Most of the existing New Testament corpora follow one of the standard editions, either the UBSGNT or Nestle-Aland editions or perhaps the Textus Receptus. They do not include any text-critical apparatus for indicating

                                                            17 In the field of computational linguistics, this process is often referred to as tokenization; see G. 

Leech, ‘Grammatical Tagging’, in Garside, Leech and McEnery (eds.), Corpus Annotation, pp. 21–24. 

18 For discussions of accents in the Greek of the New Testament, see Mounce, The Morphology of 

Biblical Greek, pp. 47–63, and D.A. Carson, Greek Accents: A Student’s Manual (Grand Rapids: Baker 

Book House, 1985). Both Mounce and Carson present detailed classification rules for the use of 

accents. For a simpler and more elegant treatment, see S.E. Porter and J.T. Reed, Fundamentals of 

New Testament Greek (Biblical Languages: Greek, 1; Sheffield: Sheffield Academic Press, 

forthcoming). 

19 Transcribed from a photograph in F.G. Kenyon, The Codex Alexandrinus in Reduced Photographic 

Facsimile: New Testament and Clementine Epistles (London: British Musem, 1909). Codex 

Alexandrinus is an interesting example to consider in the present discussion, because it contains a 

system for marking paragraphs and the addition of diaresis (e.g. ⲓ̈ⲛⲁ) and the raised dot to indicate a 

clause or sentence break, though this is not that unusual among uncial manuscripts. 

20 See Mayser, Grammatik der Griechschen Papyri aus der Ptolemäerzeit, I, pp. 85–86. 

52  

variant readings. It is surprising that so little use has been made of corpus annotation techniques in the area of textual criticism.21 It is beyond the scope of this study to address that issue in detail;22 however, it should be noted that the inclusion of textual variants into a corpus of New Testament texts should be seen as an important part of the annotation process. Grammatical Annotation. The most common, and probably most familiar, form of corpus annotation is grammatical annotation, particularly part-of-speech tagging. Generally the process involves the insertion of a tag or label indicating some grammatical information about the word that immediately preceeds it. For example, the sentence The car turned the corner, could be grammatically annotated in the following way: The_AT car_NN turned_VB the_AT corner_NN. The tags used here indicate broad part-of-speech categories, article (AT), noun (NN) and verb (VB). There are different degrees of detail that can be specified with the tags, for example, an indication could be made whether nouns are singular or plural, or the tense and mood of verbs, and so on. Regardless of the degree of delicacy, it is important to maintain consistency in the tagset that is used.

Besides the lack of standardization in the way in which corpora are grammatically annotated, particularly concerning the kind and number of tags used, another drawback to previous work is the failure to maintain a clear separation between grammatical (primarily part-of-speech) information and syntactical, semantic and discourse information. Consider the function of the conjunction ‘and’ in the sentence It was very cold and the pipes had burst. The word ‘and’ serves to indicate some causal relationship between the two clauses. The pipes burst because it was very cold outside. Some tagging schemes include the functional/semantic categories such as cause in their codes for a conjunction. So ‘and’ might receive the tag CC (Conjunction Causal); other options might be a logical relation such as purpose, or the possibility that ‘and’ is functioning disjunctively.

The use of levels of discourse means that annotation at the level of the word should focus only on the word-level and the levels below. That is, analysis is restricted to the formal grammatical categories of the word, derived from its morphological properties. This means that less information can be given about the word, but reduces the level of interpretative information included at this level. As analysis progresses from the bottom-level, the grapheme to the morpheme, to the word, to the phrase and to the clause, and so on, it requires                                                             21 See R.A. Craft, ‘The Use of Computers in New Testament Textual Criticism’, in B.D. Ehrman and 

M.W. Holmes (eds.), The Text of the New Testament in Contemporary Research: Essays on the Status 

Quaestionis (Grand Rapids: Eerdmans, 1995), pp. 268–82, for a survey of the use of computers in 

textual criticism. Also K. Aland and B. Aland, The Text of the New Testament (trans. E.F. Rhodes; 

Grand Rapids: Eerdmans, 2nd edn, 1989), pp. 317–37. For an evaluation of New Testament textual 

criticism from a papyrological perspective, see S.E. Porter, ‘Why So Many Holes in the Papyrological 

Evidence for the Greek New Testament?’, in K. van Kampen and S. McKendrick (eds.), The Bible as 

Book: The Transmission of the Greek Text (Grand Haven, MI; London: British Library and Scriptorium, 

forthcoming). 

22 For an attempt to apply corpus annotation in the study of textual variants of the purported words 

of Jesus, following the TEI (Text Encoding Initiative) guidelines for textual criticism, see S.E. Porter 

and M.B. O’Donnell, ‘The Implications of Textual Variants for Authenticating the Words of Jesus’, in 

B. Chilton and C.A. Evans (eds.), Authenticating the Words of Jesus (NTTS, 28.1; Leiden: E.J. Brill, 

1998), pp. 97–133. This article presents the initial findings of a larger project that proposes a 

standardized methodology for the annotation of textual variants in machine‐readable corpora of 

New Testament texts, following international guidelines for text encoding. 

53  

increasingly more interpretative information and the use of less formal categories. This is an inevitable part of discourse analysis; however, the advantage of a bottom-up approach is that analysis begins with the formal categories and gradually becomes more subjective as it moves towards the higher levels (the paragraph and discourse levels).23 Whichever approach is adopted, either top-down or bottom-up,24 it is important to have a clear differentiation of which elements are to be analysed at each level, as well as what analytic categories are appropriate for that level.

The importance of keeping the annotation of different discourse-levels separate is particularly marked in the case of grammatical and syntactical annotation. Syntax is perhaps the most contentious area of linguistics. There is a large difference between transformational theories, which adopt a logical, sentence-based understanding, and functional theories, which focus on language in use. In the light of this, I strongly disagree with Leech, who suggests that The reason for considering grammatical tagging and syntactical annotation together is obvious: they both, in a general sense, specify the grammatical characteristics of a text.25

A discussion of what exactly constitutes grammar is beyond the scope of this paper. However, it seems clear that there is a world of difference between adding a tag to a word indicating its morphological features, such as whether it is a noun or a verb, and assigning the word a position in a constituent structure tree. Leech, on the other hand, continues his argument, stating:

In fact, there is a strong argument that these are not really distinct levels at all: grammatical tagging is merely a specification of the leaves (or preterminal nodes) of the phrase-structure (PS) tree which is a favoured model for syntactic annotation.26

It is clear from this quotation that he favours a phrase structure grammar for his linguistic analysis. Much of the work using corpora in computational linguistics has utilized similar grammars to produce parsed tree-banks of sentences, so Leech is correct in identifying it as the favoured model of analysis. Nevertheless, it is problematic to miss the distinction between the largely theory-neutral classification of words according to their part-of-speech and a theory-bound syntactical analysis, such as the use of phrase-structure trees.

In summary, I am suggesting that grammatical annotation should take place at the level of the word, and consist of simply assigning a standard tag to each word on the basis of its part-

                                                            23 On bottom‐up and top‐down approaches to discourse analysis, see Porter, Idioms of the Greek New 

Testament, p. 299. Reed (Discourse Analysis of Philippians, pp. 47–48) states: ‘To read from the 

bottom‐up is to begin by analyzing the smaller units of discourse and how they are combined into 

increasingly larger units … To read from the top‐down is to begin with an understanding of larger 

discourse functions (e.g. register/genre) and then to interpret the meaning of smaller units in terms 

of those functions.’ He compares bottom‐up analysis to inductive reasoning and top‐down to 

deductive reasoning. For a discussion of these two approaches to discourse analysis from a more 

psycholinguistic/processing perspective, see G. Brown and G. Yule, Discourse Analysis (Cambridge: 

Cambridge University Press, 1983), pp. 234–36. 

24 Porter (Idioms of the Greek New Testament, p. 299) suggests that ‘one must work through all the 

stages, from both directions, to provide a full analysis’. 

25 Leech, ‘Grammatical Tagging’, p. 19. 

26 Leech, ‘Grammatical Tagging’, p. 19. 

 

of-speecbased. Syntactiand senclause othe largMost ofprojectsmost sc(NP) analso becentral usefulneStructurpresentsstructur 

               27 For an

Annotat

Eyes and

Parsing’,

English: 

computa

Computa

28 See su

‘Syntacti

29 Two st

D.D. Sch

62; Chico

Testame

30 Palme

ch category

ical Annotantence discoor sentence er unit. Thef these corps, and there chemes is thnd verb phra an indicatto much Ness for the re in New Ts the syntacre could be i

                       overview of

ion: Treeban

d G. Leech, ‘S

, in E. Black, 

The IBM/Lan

ational lingui

ational Lingu

mmary table

ical Annotati

tudies that h

midt, Hellen

o, CA: Schola

ent Greek (SB

r, Levels of C

y. These cat

ation. The nourse-levels(e.g. noun p

ere are a varpora have b

are significhe marking ases (VP). Ttion of dep

NLP work, study of H

Testament Gctical analyindicated in

                       f existing syn

nks’, in Garsi

Syntactic Ann

R. Garside a

ncaster Appr

istics and a d

uistics: An Int

e of schemes

ion’, p. 49. 

ave attempt

nistic Greek G

ars Press, 19

BG, 4; New Y

Constituent S

egories sho

next type ofs. It usuallyphrase, verbriety of exiseen developcant differe(bracketing

Then, depenpendency rChomskyan

Hellenistic GGreek, pressis of Lk. 2

n a corpus in

ntactical ann

de, Leech an

notation: Lin

nd G. Leech 

roach (Amste

discussion of

troduction (C

s used in diff

ted to apply 

Grammar and

81), and M.W

ork: Peter La

Structure, p. 7

ould, for Gre

f annotationy involves b phrase), ansitent corpoped for specences in theg) and labelnding on therelations ben syntacticGreek.29 Pasents an app23:47b shown the follow

otation proje

nd McEnery (

nguistic Aspe

(eds.), Statis

erdam: Rodo

f computatio

Cambridge: C

ferent annot

Chomsky’s l

d Noam Cho

W. Palmer, L

ang, 1995).

70. 

eek at least

n that can tarecognizing

nd markingora annotatecific Natura annotationlling of cone syntacticaetween the al analysis almer, in hiplication ofwn in the t

wing form:

ects, see G. L

(eds.), Corpu

ects of Gramm

stically‐Drive

opi, 1993), pp

onal syntactic

Cambridge U

tation projec

inguistic the

msky: Nomin

evels of Cons

, be primari

ake place apg the constthem and thd with syntal Language

n schemes unstituents, thal theory bei

constituenthas failed

is study Lef Chomsky’ree diagram

 

Leech and E.

us Annotation

matical Tagg

en Computer 

p. 36–61. Fo

cal analysis, s

University Pre

ts, in Leech a

ories to Helle

nalizing Tran

stituent Stru

ily morphol

applies at thtituent unittheir positiotactical infoe Processin

used.27 But, hat is, nouning used, tht groups.28

to demonsevels of Con’s X-bar them below.30

. Eyes, ‘Synta

n, pp. 34–52

ging and Skel

r Grammars o

r an introduc

see R. Grishm

ess, 1986). 

and Eyes, 

enistic Gree

nsformations

ucture in New

54 

logically

he clause ts of the on within ormation. ng (NLP)

basic to n phrases here may

Though strate its nstituent eory. He The tree

actical 

2; also E. 

leton 

of 

ction to 

man, 

k are: 

s (SBLDS, 

w 

55  

[S [NP [NP [D1 ὁ D1] [N ἄνθρωπος N] NP] [D2

οὗτος D2] NP] [AP δίκαιος AP] [VP ἦν VP] ]

It is not clear, however, what benefit, at least for discourse analysis, would be gained by having a corpus of New Testament texts annotated in this manner. This kind of analysis, fundamentally sentence-based, that attempts to apply similar sentence-grammars to whole discourses, has not been entirely successful.31 Palmer’s initial discussion of studying language in use and the need for a suitable corpus seems promising;32 however, his study is somewhat disappointing in that it fails to demonstrate how X-bar theory can increase our understanding of the Greek of the New Testament.

There is, however, value in the study of Greek clause structure and word order. I am of the opinion that a functional approach to language, particularly the analysis of what has traditionally been labelled syntax, is of much more value for understanding Hellenistic Greek, as a number of recent studies have clearly demonstrated.33 Even within NLP projects utilizing PS trees, the usefulness of functional, as well as categorical, labels is beginning to be seen.34 There have also been a number of studies that have utilized Halliday’s functional grammar to annotate corpora.

In discussing the elements of the clause in Greek, Reed notes that Unmarked Greek sentences consist of at least a verbal element (predicator); in addition, a complement such as a direct object in the accusative case may also be present. To this basic structure may be added subjects, adjuncts … and conjunctions.35                                                             31 See S.E. Porter, ‘Discourse Analysis and New Testament Studies: An Introductory Survey’, in S.E. 

Porter and D.A. Carson (eds.), Discourse Analysis and Other Topics in Biblical Greek (JSNTSup, 113; 

SNTG, 2; Sheffield: Sheffield Academic Press, 1995), pp. 14–35. 

32 Palmer, Levels of Constituent Structure, pp. 19–29. 

33 See S.E. Porter, Verbal Aspect in the Greek of the New Testament, with Reference to Tense and 

Mood (SBG, 1; New York: Peter Lang, 1989); idem, Studies in the Greek New Testament: Theory and 

Practice (SBG, 6; New York: Peter Lang, 1996); idem, ‘Dialect and Register in the Greek of the New 

Testament: Theory’, in M.D. Carroll R. (ed.), Rethinking Context, Rereading Texts: Contributions from 

the Social Sciences to Biblical Interpretation (Sheffield: Sheffield Academic Press, forthcoming). Also 

Reed, Discourse Analysis of Philippians; idem, ‘Modern Linguistics and the New Testament: A Basic 

Guide to Theory, Terminology, and Literature’, in S.E. Porter and D. Tombs (eds.), Approaches to New 

Testament Study (JSNTSup, 120; Sheffield: Sheffield Academic Press, 1995), pp. 222–65; idem, 

‘Discourse Analysis’, in S.E. Porter (ed.), Handbook to Exegesis of the New Testament (NTTS, 25; 

Leiden: E.J. Brill, 1997), pp. 189–217; G. Martín‐Asensio, ‘Hallidayan Function Grammar as Heir to 

Rhetorical Criticism’, in S.E. Porter and D.L. Stamps (eds.), The Rhetorical Interpretation of Scripture: 

Essays from the 1996 Malibu Conference (JSNTSup, 180; Sheffield: Sheffield Academic Press, 1999), 

pp. 84–107. 

34 The PENN treebank corpus (under construction at the University of Pennsylvania) includes both 

syntactical and functional labels on its tree constitutents; see Leech and Eyes, ‘Syntactical 

Annotation’, pp. 42–47. 

35 Reed, Discourse Analysis of Philippians, p. 44. 

56  

From this he concludes that the four word groups ‘subject, verb, complement and adjunct … are an important part of the analysis of the microstructures of discourse’.36 I suggest that it would be more profitable, at least for the purposes of discourse analysis, to annotate a corpus with respect to these basic elements of the clause, than to attempt to ‘parse’ clauses and sentences and indicate dependency relations. One of the areas of discourse study where clause element or component information is useful is the study of word order. A corpus marked in this way would ease the process of determining marked and unmarked patterns from a large sample.37 Semantic Annotation. The next type of annotation that can be added to a corpus is semantic annotation. The table above indicates that it can be applied at all discourse-levels, beginning with the word. In fact, it would be possible to begin at the morpheme level, given that the morpheme is the smallest unit of meaning in a discourse. Much of the linguistic work on semantics has focused on lexical semantics, that is, the meaning of units at the word level,38 with some work at the clause and sentence level, utilizing formal logic.39 This emphasis on lexical semantics is reflected in the semantic annotation that has been carried out up to now.40 Utilizing semantic-field theory it is possible to place a tag following each word indicating the semantic field to which it belongs. For the New Testament, we are fortunate to have available a lexicographic resource that classifies all the words of the New Testament by semantic domain.41 It would be possible to annotate a section of text semantically by adding the domain number from the Louw-Nida lexicon after each word. For example, ὁ <92> ἄνθρωπος <9> ἦν <13> δίκαιος <88>. Each of the domains in the lexicon is divided into subdomains, and each word has both a major domain number and a subdomain classification. The article (ὁ), for instance, is given the number 92.24. In the above example, I have used only the major domain numbers for the sake of simplicity.

                                                            36 Reed, Discourse Analysis of Philippians, p. 45. See also Porter, Idioms of the Greek New Testament, 

pp. 286–87. 

37 See Reed, Discourse Analysis of Philippians, pp. 45, 116–18; Porter, Idioms of the Greek New 

Testament, pp. 286–97; idem, ‘Word Order and Clause Structure in New Testament Greek: An 

Unexplored Area of Greek Linguistics Using Philippians as a Test Case’, FN 6 (1993), pp. 177–206. 

38 See D. Cruse, Lexical Semantics (CTL; Cambridge: Cambridge University Press, 1986); For a study 

dealing specifically with lexical semantics of the Greek of the New Testament, see M. Silva, Biblical 

Words and their Meanings: An Introduction to Lexical Semantics (Grand Rapids: Zondervan, 2nd edn, 

1994). 

39 See R.M. Kempson, Semantic Theory (CTL; Cambridge: Cambridge University Press, 1981). 

40 For review of exisiting semantic annotation projects, see A. Wilson and J. Thomas, ‘Semantic 

Annotation’, in Garside, Leech and McEnery (eds.), Corpus Annotation, pp. 53–65. 

41 See J.P. Louw and E.A. Nida, Greek‐English Lexicon of the New Testament Based on Semantic 

Domains (2 vols.; New York: United Bible Societies, 1988). For a discussion of the methodology and 

classification scheme behind the lexicon, see Louw and Nida, Greek‐English Lexicon, I, pp. vi–xx; also 

E.A. Nida and J.P. Louw, Lexical Semantics of the Greek New Testament (SBLRBS, 25; Atlanta: Scholars 

Press, 1993). 

57  

There are certain words that have been classified as belonging to more than one domain, thus presenting a challenge to this sort of word-by-word annotation. There are two possible ways of handling this situation. The first, and most simple, is to include all the possible semantic tags for a word, leaving it to the user of the corpus to make a decision as to which domain to select. The previous example would then appear in the following form: ὁ <92> ἄνθρωπος <9, 10> ἦν <13, 85, 71, 58> δίκαιος <88, 34, 66>. The second method of handling multidomain words is for the annotator to attempt to disambiguate the different options. If the annotation is being done manually, then the annotator will make this decision on the basis of their understanding of the text and context. If the annotation process is automated, then a set of disambiguation rules will need to be used, perhaps using the concept of collocation.42 The example below shows the first three verses of Jude, which have been annotated with semantic information: each word has the major domain or domains assigned to it in the Louw-Nida lexicon. It is clear that many of the words belong to more than one domain, and thus require disambiguation. For the discourse analyst, however, the availability of a text annotated in this way will save vast amounts of time looking up words in the lexicon.

1.1 Ἰούδας <93> Ἰησοῦ <93> Χριστοῦ <93, 53> δοῦλος, <87, 37> ἀδελφὸς <10, 11> δὲ <89>

Ἰακώβου, <93> τοῖς <92> ἐν <83, 13, 84, 89, 90, 67> θεῷ <12> πατρὶ <10, 12, 87, 11, 58, 36> ἠγαπημένοις <25> καὶ <89, 91> Ἰησοῦ <93> Χριστῷ <93, 53> τετηρημένοις <13, 37, 36> κλητοῖς· <33>

1.2 ἔλεος <88> ὑμῖν <92> καὶ <89, 91> εἰρήνη <22, 25> καὶ <89, 91> ἀγάπη <25, 23> πληθυνθείη. <59>

1.3 Ἀγαπητοί, <25, 58> πᾶσαν <59, 78, 63, 58> σπουδὴν <68> ποιούμενος <90, 42, 13, 42, 41, 37, 57> γράφειν <33> ὑμῖν < 92> περὶ <83, 90, 67, 89> τῆς <92> κοινῆς <57, 53, 65, 89> ἡμῶν <92> σωτηρίας <21> ἀνάγκην <22, 71> ἔσχον <57, 18, 31, 49, 74, 90, 13, 90> γράψαι <33> ὑμῖν <92> παρακαλῶν <33, 25> ἐπαγωνίζεσθαι <39> τῇ <92> ἅπαξ <60> παραδοθείσῃ <57, 37, 33, 13> τοῖς <92> ἁγίοις <88, 53> πίστει. <31, 33> A corpus that has been annotated with semantic information could be utilized in a number

of ways by the discourse analyst. It could be used in the investigation of lexical cohesion, to help identify significant lexical chains within the discourse.43 It would also be useful not only for identifying the ideational component of the discourse, but also in the discovery of discourse structure.44 See the example below, in section 5b, for further discussion of the use of semantic annotation in discourse analysis.

                                                            42 See G. Hirst, Semantic Interpretation and the Resolution of Ambiguity (Cambridge: Cambridge 

University Press, 1987). 

43 On lexical cohesion, see M.A.K. Halliday and R. Hasan, Cohesion in English (London: Longman, 

1976), pp. 282–92. Also M. Hoey, Patterns of Lexis in Text (Oxford: Oxford University Press, 1991). For 

studies of cohesion in New Testament discourse, see G.H. Guthrie, ‘Cohesion Shifts and Stitches in 

Philippians’, in Porter and Carson (eds.), Discourse Analysis, pp. 36–59; and especially Reed’s work: 

J.T. Reed, ‘Cohesive Ties in 1 Timothy,’ Neot 26.1 (1992), pp. 131–47; idem, Discourse Analysis of 

Philippians, pp. 89–101. 

44 The thesis of Hoey’s work is that non‐narrative discourse (the focus of his study) does not exhibit 

external or generic structure, but that structure is created through lexical chains and patterns. That is 

to say, text is ‘organized rather than structured’ (Hoey, Patterns of Lexis, p. 29). 

58  

Moving beyond the discourse-level of the word, it is possible to discuss the semantics of the clause, sentence and paragraph. Higher levels of semantic annotation are, as far as I am aware, a virtually unexplored area of research. It would require the selection of a consistent annotation scheme. As with syntactical annotation, a functional approach would provide more useful information for discourse analysis than a formally based, logical approach to semantics. Discourse Annotation. The previous types of annotation discussed have applied primarily to the level of the sentence or below, though the possibility of semantic annotation at the paragraph and discourse level was mentioned. Once analysis moves above the clause/sentence unit, it becomes more difficult to quantify in a formal way. There are numerous features of discourse that could be added to a corpus of texts, such as information flow or structure; using notions such as theme and rheme; logical relationships between and within paragraph, sentence and clause units; the marking of items of prominence; the identification of discourse participants, actors and recipients. This variety leads Garside, Fligelstone and Botley to comment that ‘ “discourse annotation” is likely to appear a particularly ill-defined concept’,45 in comparison to the other forms of annotation already discussed. This situation is largely due to the numerous, different approaches to discourse analysis.46 Like much of modern linguistics, the field of discourse analysis is still in its early stages of development, and thus exhibits great terminological and theoretical diversity.47 The systemic-functional approach to discourse, pioneered in the work of Michael Halliday, seems particularly suited for application to New Testament discourse, most comprehensively demonstrated in the recent study of Philippians by Reed.48

There is not the space here either to describe Halliday’s system or to explore different possibilities for annotating a corpus with the components from this system. One example, drawing from the work of Garside, Fligelstone and Botley, shall suffice for the present.49 They select the feature of cohesion to demonstrate the addition of discourse information into a corpus. The standard linguistic work on cohesion (at least for English) is Cohesion in English by Halliday and Hasan. They deal with the different linguistic devices that are used to create cohesion within discourse: reference, substitution, ellipsis, conjunction and lexical cohesion. The final chapter presents a method for analysing cohesion in texts and an application to seven short texts.50 The basic concept for this analysis is the concept of a tie. ‘A tie is best                                                             45 R. Garside, S. Fligelstone and S. Botley, ‘Discourse Annotation: Anaphoric Relations in Corpora’, in 

Garside, Leech and McEnery (eds.), Corpus Annotation, pp. 66–84 (66). 

46 For an overview of some of the major approaches to discourse analysis, see D. Schiffrin, 

Approaches to Discourse (Oxford: Basil Blackwell, 1994). 

47 This is perhaps because analysis of discourse units relies on the analysis of smaller units, and 

particularly the analysis of the sentence. As discussed above, the analysis of the sentence falls under 

what has traditionally been referred to as syntax, and much of the diversity within linguistics involves 

different approaches to syntax. It is thus unlikely that anytime in the near future we will see 

homogenity in discourse theory. For a survey of the main approaches to discourse that have been 

applied to the New Testament, see Porter, ‘Discourse Analysis and New Testament Studies’. 

48 See Reed, Discourse Analysis of Philippians, esp. pp. 34–122. 

49 Garside, Fligelstone and Botley, ‘Discourse Annotation’, pp. 66–84. 

50 Halliday and Hasan, Cohesion in English, pp. 329–55. 

59  

interpreted as a RELATION between … two elements.’51 These two elements in the text are joined by one of the cohesive devices listed above. For example, consider the references to the character of Jesus in Mt. 11:1: καὶ ἐγένετο ὅτε ἐτέλεσεν ὁ Ἰησοῦς_G διατάσσων τοῖς δώδεκα μαθηταῖς αὐτοῦ_R, μετέβη_I ἐκεῖθεν τοῦ διδάσκειν καὶ κηρύσσειν ἐν ταῖς πόλεσιν αὐτῶν (‘And it was when Jesus finished commanding his twelve disciples, he left there in order to teach and to preach in their cities’). The first of the references is a full or grammaticalized reference (G), and is able to stand on its own without reference, either backwards (anaphoric) or forwards (cataphoric), to another element.52 This reference stands in relation with two other elements in the verse. There is a reduced reference (R) to Jesus in the third person personal pronoun αὐτοῦ. The second is the implied reference (1) contained in the verb μετέβη. These two references are anaphoric because they require the reader to ask the question Who is he? and to look back in the discourse for the answer. Halliday and Hasan present a coding system for marking in a text each of the cohesive devices they cover in their work. For example, in analysing an extract from Alice in Wonderland, they identify two cohesive devices in the clause ‘She looked at the Queen’.53 ‘She’ is given the code R12.6, meaning that it is a feminine singular non-possessive (HEAD term) pronominal reference item.54 To the word ‘Queen’, they attach the code L1.6, indicating that the word is a lexically cohesive device, the same as a previous item in the discourse, in an identical form.55 Their coding system is very detailed and requires a considerable amount of analysis to use it on a text.

Garside, Fligelstone and Botley describe the work carried at UCREL56 to annotate the cohesive devices in a corpus of several hundred thousand words. They comment that for large corpora of this kind the notation scheme used must be somewhat simplified in comparison to the one used by Halliday and Hasan on small, individual texts.57 The details of the scheme are

                                                            51 Halliday and Hasan, Cohesion in English, p. 329. 

52 This is somewhat simplified, because there is a sense in which a grammaticalized reference such as 

ὁ Ἰησοῦς may be acting anaphorically or cataphorically (endophoric relations). Also, it can be said to 

be acting exophorically, pointing to a referent outside the discourse. For a discussion of endophoric 

and exophoric relations, see M.A.K. Halliday and R. Hasan, Language, Context, and Text: Aspects of 

Language in a Social‐Semiotic Perspective (Victoria, Australia: Deakin University, 1985), pp. 75–77. 

53 Halliday and Hasan, Cohesion in English, p. 340. 

54 Halliday and Hasan, Cohesion in English, p. 333. 

55 Halliday and Hasan, Cohesion in English, p. 338. 

56 UCREL stands for University Centre for Computer Corpus Research on Language. It is situated at 

Lancaster University and has been engaged in numerous corpus annotation projects over the last 20 

years. 

57 Garside, Fligelstone and Botley, ‘Discourse Annotation’, p. 68. They state that when adding 

cohesive annotation to a hundred‐thousand‐word corpus, ‘speed of human analysis is important, but 

not at the expense of accuracy and consistency in marking. The result is a notation scheme which 

does not attempt to mark all possible theoretically justifiable distinctions, and in fact tends to be 

theoretically neutral.’ 

60  

not important here,58 but, basically, each antecedent in the discourse is marked with parentheses and a number, and every proform that refers back to it (in the case of anaphoric relations) is marked with a code indicating the type of relation and the number of its antecedent. They give the following example of the annotation scheme: (7 this week’s winner 7) said <REF = 7 he had rung (8 <REF = 7 his wife 8) and <REF = 7, 8 they had spoken to <REF = 7, 8:2 each other. The <REF indicates that the following word is an anaphoric reference back to the word or phrase enclosed in parentheses, and marked with the same number. They explain that ‘here the symbol <REF = 7, 8:2 indicates an anaphoric referential link from a proform two orthographic words long (each other); the link is to a pair of antecedents (this week’s winner and his wife)’.59 Following this scheme, the previous example from Mt. 11:1 could be annotated as καὶ ἐγένετο ὅτε ἐτέλεσεν (1 ὁ Ἰησοῦς 1) διατάσσων τοῖς δώδεκα μαθηταῖς <REF = 1 αὐτοῦ, <REF = 1 μετέβη ἐκεῖθεν τοῦ διδάσκειν καὶ κηρύσσειν ἐν ταῖς πόλεσιν αὐτῶν. It might also be useful to indicate the type of reference, either grammaticalized, reduced or implied.

Discourse annotation in corpora is still at a very early stage of development, and there are many other kinds of annotation that could take place, the marking of cohesive devices being just one of them. However, small-scale test markups will be required to investigate both the value of each scheme, for the discourse analyst, and the degree of effort required to apply the annotation. c. Standardization of Annotation Even this brief overview of the different types of corpus annotation, and the different levels at which they can take place, has shown the variation that exists in different annotation schemes. Due to the relative newness of the field of corpus annotation, there have been few standards or guidelines on how annotation should be applied. Individual researchers and teams of researchers have designed their own tagsets and markup schemes, guided by the purpose of each particular project. This is primarily because the annotation has been seen merely as a means to an end and not as a goal in its own right. However, with the emergence of corpus annotation as a field of its own, and the availability of larger corpora,60 some sort of agreed standard is required as corpora are used by different users for multiple purposes.61

                                                            58 See Garside, Fligelstone and Botley, ‘Discourse Annotation’, pp. 68–78. For more detailed 

descriptions of the UCREL anaphoric corpus, see S. Fligelstone, A Description of the Conventions Used 

in the Lancaster Anaphoric Treebank Scheme (UCREL Technical Report; Lancaster: Department of 

Linguistics, Lancaster University, 1991). Also idem, ‘Developing a Scheme for Annotating Text to Show 

Anaphoric Relations’, in G. Leitner (ed.), New Directions in English Language Corpora: Methodology, 

Results and Software (Berlin: Mouton de Gruyter, 1992), pp. 153–70. 

59 Garside, Fligelstone and Botley, ‘Discourse Annotation’, p. 69. 

60 The British Nation Corpus (BNC) currently contains 100 million words, in comparison to the 1 

million in the Brown Corpus, which was one of the first electronic corpora in wide usage. 

61 ‘The production of an annotated corpus is without doubt an expensive task, in terms of both time 

and effort, and therefore the reusability and shareability of such a resource is of great importance. 

Standardization of annotation practices can ensure than an annotated corpus can be used to its 

greatest potential.’ So say P. Kahrel, R. Barnett and G. Leech, ‘Towards Cross‐Linguistic Standards or 

61  

A standard for all types of annotation, at all levels, is still a long way off. However, the development of the Text Encoding Initiative (TEI) Guidelines for Electronic Encoding and Interchange marks a considerable step in this direction. There is not room to explain the nature or background of these guidelines here. The TEI guidelines have become the international standard for text annotation in a broad range of disciplines.62 Kennedy suggests that It is hoped that the Text Encoding Initiative will help establish norms for those compiling electronic corpora for a wide variety of research purposes as to the levels of markup and the features which should be marked.63 The guidelines include standards for indicating the information structure and format of documents, through the use of SGML64 tags, such as <p> … </p> to mark off a paragraph. They also include standards on the encoding of different languages and the recording of textual variants when working from manuscripts. As yet, however, the guidelines do not present a standard for higher level information, such as grammatical and syntactical annotation.65 The UCREL team has been in collaborative work with other European researchers to develop a set of guidelines for grammatical and syntactical annotation, called the EAGELS (Expert Advisory Group for Language Engineering Standards), but these are currently only in the initial stages of development.66

3. Existing Annotated Corpora of the New Testament and their Use

The recent development of software for studying Greek and Hebrew has opened many new avenues of research into the biblical texts and languages. Perhaps most widely recognized is the work of the GRAMCORD Institute, headed by Paul Miller.67 They have developed powerful search software that allows words or combinations of words to be searched for, either in their inflectional or lexical forms. There is not space to describe such software or its

                                                                                                                                                                                          Guidelines for the Annotation of Corpora’, in Garside, Leech and McEnery (eds.), Corpus Annotation, 

pp. 231–42 (242). 

62 See discussion in G. Kennedy, An Introduction to Corpus Linguistics (London: Longman, 1998), pp. 

82–85; also S. Hockey, ‘Textual Databases’, in J. Lawler and H.A. Dry (eds.), Using Computers in 

Linguistics: A Practical Guide (London: Routledge, 1998), pp. 111–15. For summary of the TEI 

guidelines, see C.M. Sperberg‐McQueen and L. Burnard (eds.), Guidelines for Electronic Text Encoding 

and Interchange (Chicago: Text Encoding Initiative, 1994). 

63 Kennedy, Introduction to Corpus Linguistics, pp. 84–85. 

64 SGML = Standard Generalized Markup Language. A popular subset of SGML is HTML (Hypertext 

Markup Language) used for producing pages for the worldwide web. See C.F. Goldfarb, The SGML 

Handbook (Oxford: Clarendon Press, 1990). 

65 Kennedy (Introduction to Corpus Linguistics, p. 85) notes, ‘When we move beyond the 

representation of formal organizational features of the text, however, to tagging of word classes, 

there are few signs yet that there can be compatibility of tagsets between different corpus projects’. 

66 See Kahrel, Barnett and Leech, ‘Towards Cross‐Linguistic Standards’, pp. 235–42. 

67 For information contact GRAMCORD Institute, Paul A. Miller, 2218 NE Brookview Drive, Vancouver, 

WA 98686, USA. 

62  

use for grammatical study here.68 However, it should be noted that what allows GRAMCORD and similar programs (BibleWorks and BibleWindows) to perform such powerful search operations is the underlying text that has been annotated with grammatical information. Hahne has provided a useful comparison of both the different search programs and the grammatically tagged texts they utilize.69 He identifies the three main corpora used by these programs: the GRAMCORD text, the Friberg text and the CCAT text. These texts all contain grammatical annotation at the word level, that is, each word has been given at least one part-of-speech tag. However, these tags are not always based solely on the morphology of the word, as suggested in section 2 above. For instance, the GRAMCORD tagset provides two levels of classification for conjunctions. A conjunction is indicated by the code C, followed by its class, coordinating (C) or subordinating (S), followed by its subclass. A coordinating conjuction, for example, can be continuative, correlative, disjunctive, interrogative, copulative, inferential, explanatory or adversative.70 Distinctions such as these are clearly operative only at the clause level and above, and are highly subjective and interpretative. It is not that such distinctions should not be included in a corpus, but they do not belong at the word level of annotation. The majority of the tags in the GRAMCORD text are restricted to the formal, morphological categories of the word.

The original Friberg text (it has recently been revised), on the other hand, includes many more functional categories in its tagset. The creators of the text, Timothy and Barbara Friberg, have long been involved in both translation and discourse analysis of the New Testament. The text reflects this background, and has included numerous discourse features, but they have included many of them in the grammatical tags attached to each word. An example of this is their treatment of participles. They use a distinctive tag (R) to indicate the, so-called, imperatival participle. This functional categorization can only be decided with reference to the co-text, at the level of the clause and above. Also, many of the participles in the text, particularly those which are functioning adverbially, are given a person and number tag. In Jude 3, two of the participles, ποιούμενος and παρακαλῶν, are marked up in this way: ἀγαπητοί, πᾶσαν σπουδὴν ποιούμενος (vppmnm1s) γράφειν ὑμῖν περὶ τῆς κοινῆς ἡμῶν σωτηρίας ἀνάγκην ἔσχον γράψαι ὑμῖν παρακαλῶν (vppanm1s) ἐπαγωνίζεσθαι τῇ ἅπαξ παραδοθείσῃ τοῖς ἁγίοις πίστει.71 The participles are modifying the finite verb ἔσχον, which does grammaticalize person and number (first person singular). This information has been added to the participles to indicate

                                                            68 See Porter and Reed, Fundamentals of New Testament Greek, the section ‘Where to go from here’, 

in the final chapter, for a description of the GRAMCORD software and its use for grammatical study. 

69 See H. Hahne, ‘Avoiding the Pitfalls of Computer‐Assisted New Testament Grammatical Analysis’, in 

F.R. Poswick (ed.), Bible and Computers: Desk and Discipline. The Impact of Computers on Biblical 

Studies: Proceedings of Association Internationale Bible et lnformatique (Paris: Champion‐Slatkine, 

1995), pp. 223–36. 

70 See H. Brown, Accordance: Software for Biblical Studies Version 2.0 User Guide (Altamonte Springs, 

FL: OakTree Software, 1996), Appendix B, pp. 4–5. 

71 ‘Beloved, while eagerly preparing to write to you about the salvation we share, I find it necessary 

to write, appealing to you to contend for the faith that was once for all entrusted to the saints’ 

(NRSV, modified to reflect the grammatical point made above). 

63  

that their ‘subject’ is the first person singular implied referent of ἔσχον.72 Though understandable, from a discourse perspective, this kind of annotation is potentially very misleading, especially when searches for all first person singular references in a discourse are carried out (see section 5 below).73

4. The Use of Annotated Corpora for Discourse Analysis

In a volume such as this, the reader would be forgiven at this point for wondering exactly what all this discussion of corpus annotation has to do with discourse analysis. The field of corpus annotation is such a recent development, that it seemed necessary to outline its theory and practice in some detail. As yet, there have been only a handful of discourse studies that have utilized annotated corpora. Most often, studies in discourse analysis focus either on one particular grammatical or discourse feature of a language or select a small text for illustrative purposes. McEnery and Wilson note that ‘the amount of corpus-based research in pragmatics and discourse analysis has to date been relatively small’.74 This is partly due to the limitations of the standard tools for accessing corpora, such as concordances and simple search software. Discourse analysis requires attention to wider co-textual information than that available from a concordance listing. However, the sizable quantity of data that needs to be examined and analysed to carry out a discourse analysis of even a short section of text, seems ideally suited to computerized storage and manipulation.75

Building on his earlier corpus work, which focused upon variation in language (register) according to genre,76 Biber has carried out a number of studies of both individual discourse features across different types of text (conversation, speeches, news reportage, academic writing) and combinations of grammatical features across a text in order to investigate information flow and discourse structure. He suggests that ‘the major concern in discourse analysis is the way that meaning unfolds over the course of a text’.77 Focusing on a corpus of scientific texts (‘research articles from experimental science’78), he collects frequency counts

                                                            72 (vppmnm1s) stands for verb, participle, present, middle, nominative, masculine, first person 

singular. 

73 See Hahne, ‘Avoiding the Pitfalls’, for a more detailed discussion of the interpretative implications 

of using biblical search software for exegesis. 

74 McEnery and Wilson, Corpus Linguistics, p. 98. 

75 For examples of studies utilizing a corpus to investigate specific features of discourse, see P. Collins, 

‘The Modals of Obligation and Necessity in Australian English’, in K. Aijmer and B. Altenberg (eds.), 

English Corpus Linguistics: Studies in Honour of Jan Svartvik (London: Longman, 1991), pp. 145–65; A. 

Stenström, ‘Expletives in the London‐Lund Corpus’, in Aijmer and Altenberg (eds.), English Corpus 

Linguistics, pp. 239–53. M. Stubbs, Text and Corpus Analysis (London: Blackwell, 1996); J. Thomas 

and A. Wilson, ‘Methodologies for Studying a Corpus of Doctor‐Patient Interaction’, in J. Thomas and 

M. Short (eds.), Corpora for Language Research: Studies in Honour of Geoffrey Leech (London: 

Longman, 1996), pp. 92–109. 

76 See D. Biber, Variation across Speech and Writing (Cambridge: Cambridge University Press, 1988). 

77 Biber, Conrad and Reppen, Corpus Linguistics, p. 122. 

78 Biber, Conrad and Reppen, Corpus Linguistics, p. 123. 

64  

of the tense and voice forms of verbs from these texts. The figures are broken down according to the different sections of the articles, for example, introduction, methods, results and discussion. For example, the following mean scores (per 1000 words) were found for the present tense-form: Introduction 47.9, Methods 21.1, Results 35.9 and Discussion 60.6.79 He notes that frequency counts of this type ‘provide useful average characterizations of each section [of the text]’.80 However, they do not provide information on how a discourse progresses and develops. To obtain this information, Biber presents a map of the progression of verbs throughout a text. Adopting a tabular display, he uses four columns to display this data. The first two columns plot verb tense and the others voice. For each clause of the discourse containing a verbal form, a mark is placed in the relevant column. This way it is possible to identify sections where, for example, the present tense predominates.81 These charts confirm the findings of earlier frequency counts, but also reveal an interesting pattern of information flow in the scientific texts examined.82 They note that at the transition points within the structure, for example, moving from the introduction to the method section, there is not an immediate switch from the present tense to the past, as the simple frequency figures would suggest. Instead, the transition takes place over a number of sentences before the formal start of the section, and a number of sentences into the new section.

Biber’s study presents only limited and initial findings, but it clearly illustrates the ways in which annotated corpora could aid in the process of discourse analysis, as well as offering new insights not easily identified through manual methods. He concludes that ‘corpus-based techniques can act like a pointer, locating those portions of a text that are particularly interesting from a discourse-analytic perspective. They do this by providing a comprehensive analysis of the text.’83

5. Examples of the Use of Annotated Corpora for New Testament Discourse Analysis

To conclude, I present three examples of the use of annotated sections of the New Testament in the course of discourse analysis. Each of the examples uses the book of Jude as a sample corpus, due primarily to the book’s manageable size. The examples are largely undeveloped and meant simply to be suggestive of what types of analysis would be possible, given the availability of the kind of annotated corpora discussed above (see section 2). The examples demonstrate the use of a corpus annotated with grammatical, semantic and discourse information respectively.

                                                            79 Biber, Conrad and Reppen, Corpus Linguistics, p. 124. 

80 Biber, Conrad and Reppen, Corpus Linguistics, p. 126. 

81 See Biber, Conrad and Reppen, Corpus Linguistics, pp. 126–30, for a description of the charts and 

the program developed to produce them. They state that ‘the output of the program is set up in 

columns to show the movement between past and non‐past tense, and between active and passive 

voice’. 

82 They suggest that ‘the maps can be used … to identify systematic departures from the expected 

patterns, leading to the identification of rhetorically salient shifts in discourse’ (Biber, Conrad and 

Reppen, Corpus Linguistics, p. 127). 

83 Biber, Conrad and Reppen, Corpus Linguistics, p. 130. 

65  

a. The Use of a Grammatically Annotated Corpus In section 4, brief reference was made to a study by Biber that made use of a grammatically annotated corpus to study the grammatical characteristics of scientific discourses. Following a similar method for Jude, and selecting a number of grammatical features that are marked in a grammatically annotated corpus, produces the chart below. The features selected were aspect, voice, mood, person-number and initial conjunctions. The aspect categories follow Porter’s system, thus aorist forms grammaticalize perfective aspect, present and imperfect imperfective, and perfect and pluperfect stative aspect.84 The revised Friberg morphological text was used to produce the above chart. One of the problems with this text is the tagset used for indicating voice—there are a number of different tags to deal with middle, passive and, so-called, deponents.85 The first three columns under the voice section represent the basic active, middle and passive tags, and the fourth column captures everything else. The mood section of the chart requires no explanation. The optative has been omitted due to its low occurrence in the New Testament. Person-number includes verb and pronoun counts and, unfortunately, the person-number value given many participles in the Friberg text (see section 3). The final columns of the chart are based on what I have referred to as initial conjunctions. The first two columns indicate whether a καί or δέ respectively, occurs within the first three words of the verse. The third column indicates that neither of those words occurs in this span. This could indicate asyndeton, or the presence of another conjunctive word. There are obvious problems with this classification, such as the use of the beginning of the verse, instead of considering clauses. However, it serves as a rough indication of the logical structure of the discourse, which might reveal some interesting patterns for closer investigation.  

 

 

 

 

 

                                                            84 See Porter, Idioms of the Greek New Testament, pp. 20–28. 

85 There are seven different voice tags used by the text. A = active, M = middle, P = passive, E = middle 

or passive, D = deponent middle, O = deponent passive, N = middle or deponent passive. See M. 

Bushnell, BibleWorks for Windows: Installation Guide and Pocket Reference (Montana: Hermeneutika 

Software, 1996), p. 74. Without this unnecessary complication of the Greek voice system, it still 

poses a considerable problem. Since only the future and aorist tense‐forms possess distinct forms for 

the middle and passive voices, it is not clear how the ambiguity in the other tense‐forms should be 

handled when adding grammatical annotation to a corpus. See D.B. Wallace, Greek Grammar beyond 

the Basics: An Exegetical Syntax of the New Testament (Grand Rapids: Zondervan, 1996), p. 410 n. 8. 

For a discussion of some of the difficulties of understanding and translating voice in the Greek of the 

New Testament, and some suggestions for a reassessment of the voice system, see M.B. O’Donnell, 

‘Some New Testament Words for Resurrection and the Company They Keep’, in S.E. Porter, M. Hayes 

and D. Tombs (eds.), Resurrection (RILP, 5; JSNTSup, 186; Sheffield: Sheffield Academic Press, 

forthcoming 1999). 

 

Simof the Bdisplay The plodiscoursnumericsignifica               86 This is 

Testame

merely c

without 

example

authorsh

sample o

and eval

Use of St

D.A. Cars

Sheffield

milar informaBible searchthis inform

ot merely prse analyst cal patternsant.86 It is b                      one of the m

ent discourse

counting feat

addressing t

e, the use or 

hip of certain

of text fails t

luation of su

tatistics in th

son (eds.), Li

d: Sheffield A

ation to thath programs

mation in a tresents a sumust take

s found inbeyond the                       main difficult

e studies (an

tures and th

the meaning

absence of t

n New Testa

o take into c

ch studies, s

he Discussion

inguistics an

Academic Pre

t displayed , such as Atabular manmmary of tthis inform

n such a e scope of t

ties with the 

d discourse s

en drawing i

 within the l

the article ha

ment docum

consideration

see M.B. O’D

n of Authors

d the New T

ess, forthcom

in the abovAccordance nner so thatthe featuresmation andplot, but t

this study to

 use of both 

studies in ge

interpretativ

inguistic fram

as been used

ments. Howev

n the discou

Donnell, ‘Ling

hip of the Ne

Testament: C

ming 1999).

ve profile caor BibleW

t grammatic of the text—

d interpret these are o carry out

descriptive a

eneral). Too o

ve conclusion

mework of th

d as one of th

ver, simply c

rse function 

guistic Finger

ew Testame

ritical Junctu

an easily be orks. It is ucal patterns —it is a deit. There mnot necessa systemat

and inferenti

often studies

ns regarding 

he quantified

he features t

counting artic

of the article

rprints or Sty

nt Documen

ures (JSNTSup

collected uuseful to becan be ind

escriptive chmay be intsarily lingutic treatmen

tial statistics 

s have stopp

language us

d feature. Fo

to determine

cles in a spe

e. For a sum

yle by Numbe

ts’, in S.E. Po

p, 168; SNTG

66 

 

using one e able to dentified. hart. The teresting uistically nt of the

in New 

ped at 

se, 

or 

e 

cific 

mary 

ers? The 

orter and 

G, 5; 

67  

grammar of the book of Jude. It is only possible to make a few observations from the above chart.

Examining the distribution of person and number within a discourse reveals the interpersonal functions within the discourse, that is the participants in the text.87 This is particulary the case with the first and second person because, as Porter notes, they ‘imply that the participants be present, whereas third person does not’.88 Tracking grammatical person through a discourse may help to discover or re-enforce the understanding of the structure of the text. For instance, it seems intuitively reasonable to expect a high occurrence of second-person plural references in the paranetic section of a New Testament letter. A comparison of a plot of second-person reference in a letter with a verse outline from a commentary could be informative for this task.89 The person system also contributes to the creation of prominence in discourse.90 The third-person singular, especially in verbs, is often used to provide the background or setting material. It is used in narrative to describe the actions of characters. The pattern of the third-person singular in Jude is noteworthy partly because of the relative infrequency of third-person singular references. Considering the totals for person use in Jude, it is surprising to see that the third-person singular is only the third most common,91 with concentration in the narrative sections (vv. 9, 14–16). As significant as when the third singular does occur, is its absence. There are seven verses (17–23) without one occurrence:

                                                            87 In his description of language as social‐semiotic, Halliday identifies three elements of context: field, 

tenor and mode, realized in the text, respectively, by ideational, interpersonal and textual meanings. 

See, for example, Halliday and Hasan, Language, Context, and Text, pp. 24–28. For a detailed and 

clear adaption of Halliday’s system to New Testament discourse, see Reed, Discourse Analysis of 

Philippians, chap. 2, esp. pp. 60–122. 

88 Porter, Idioms of the Greek New Testament, p. 76. He makes reference to Lyons, Introduction to 

Theoretical Linguistics, pp. 276–78. On person deixis, see Reed, Discourse Analysis of Philippians, pp. 

95–96. 

89 Porter (Idioms of the Greek New Testament, p. 301) suggests that ‘shifts in grammatical person … 

are often useful indicators of the closing of one discourse unit and the beginning of another’. He 

gives the example of Rom. 1:18–32 and 2:1–29, with a shift from third to second person. 

90 See Porter, Idioms of the Greek New Testament, pp. 302–304; Reed, Discourse Analysis of 

Philippians, pp. 105–119. For a discussion of prominence in Jude, with specific reference to verbal 

aspect, see J.T. Reed and R.A. Reese, ‘Verbal Aspect, Discourse Prominence, and the Letter of Jude’, 

FN 9 (1996), pp. 181–99. 

91 First singular 4×, first plural 7×, second singular 1×, second plural 24×, third singular 14× and third 

plural 28×. 

 

 

This becor omiswell as occurrenimperfeimperfeHowevethe perf

The aspect, the thirdrecognitframewthat pro

               92 The fig

aspect a

the Lette

comes moression coinci

a section nces of the

ective aspecective aspecer, vv. 18–2fective tense diagram abimperative d-person sintion of sucork. Reed h

ominence in

                      gures are: pe

nd the verb‐

er of Jude’. 

e significanides with thwhere the imperative t in this secct in the w24 contains e-forms. bove summmood and

ngular, the pch patterns has presente

n discourse ‘

                       erfective 30×

‐forms in Jud

nt from a dishe main clu

use of themood are f

ction (vv. 17whole book,

the highest

marizes thesethe second-perfective ais of little

ed such a f‘is rarely sig

×, imperfectiv

de, see Reed 

scourse persuster of sece indicativefound. In ad7–24). Ther,92 which at concentrat

e observatio-person plu

aspect and thuse if it c

framework gnalled by

ve 40× and s

and Reese, 

spective, whcond-persone decreasesddition, one e is a notabare distribution of the i

ons: markedural accomphe indicativcannot be iin his treatmone device,

tative 9×. Fo

‘Verbal Aspe

hen it is notn plural refe

and then notes the ply large num

uted throughimperfectiv

d occurrencpanied by nove mood. Asinterpreted ment of pro but more o

or a more in‐

ect, Discourse

 

ted that this ferences in

ends and tpredominancmber of verhout the di

ve and the lo

ces of the imotable omiss noted abovwithin a d

ominence. Hoften is the

‐depth analy

e Prominenc

68 

absence Jude, as the only ce of the

rbs in the iscourse. owest of

mperfect ssions of ve, mere

discourse He notes result of

sis of 

ce, and 

69  

a combination of grammatical and semantic features’.93 This lends support to the patterns which have been identified in Jude 17–24. Reed does not specifically discuss person reference in his treatment of prominence devices, but he does examine both aspect and mood. Concerning verbal aspect, he states:

One of the discourse functions of verbal aspect is to indicate the prominence of clauses in relationship to the larger paragraph or discourse … Thematic prominence may be signalled by the present and imperfect tense-forms (imperfective aspect), as well as the future tense-form.94

Verses 17–24 show the highest and most sustained concentration of the imperfective aspect, and a lack of the perfective. This warrants closer investigation, to see whether there might be a thematic peak in the discourse. Turning to mood, Reed says:

Modality may be used to distinguish between background and thematic prominence. In non-narrative, the imperative mood is … used in thematic material, due to its semantic attribute of ‘direction’ (i.e. the speaker directs or commands others to do something).95

A closer reading of Jude 17–23 confirms the findings from the grammatical plot. In terms of the letter form, this section is clearly the paranetic section. It follows the body, which has presented a number of historical and narrative examples in preparation for the directive commands beginning in v. 17. In terms of information flow, the writer has clearly reached a peak at this point in the discourse. In addition to the high frequency of the imperfective aspect in Jude, the numbers of participles are also notable and warrant closer investigation.96 The kind of grammatical discourse plot presented above can only serve as a descriptive summary of selected grammatical features of the text. There are many more features that could be included in such a numerical analysis. It serves as a starting point for more detailed investigations, such as the study of aspect in Jude by Reed and Reese.

The length of Jude makes it relatively easy to keep track of the grammatical and discourse features of the text, and thus this kind of analysis might be seen as unnecessary. However, this example serves as an indicator of what could be done with longer discourse sections. Longer texts present a number of difficulties not encountered in this study of Jude. For example, if a grammatical discourse plot of the book of Acts were produced and person usage were examined, a sporadic smattering of the first person plural would be seen, with significant clusters around 16:11–21; 20:5–15; 21:1–25; 27:1–29. These sections have been referred as the ‘we’ passages, sections where the narrator uses the first-person plural in presenting the narrative.97 The discourse plot does bring these sections to light, but they are not shown to be                                                             93 Reed, Discourse Analysis of Philippians, p. 111. 

94 Reed, Discourse Analysis of Philippians, pp. 113–14. 

95 Reed, Discourse Analysis of Philippians, p. 115; cf. Porter, Verbal Aspect, pp. 335–36. 

96 Though there may not be anything all that unusual or significant about the number or distribution 

of participles. Similar plots of a significant number (if not all) of the books of the New Testament 

(particularly the letters) would be required to see whether the distribution in Jude is marked in any 

way. 

97 For a recent and provocative treatment of the ‘we’ passages in Acts, see S.E. Porter, The Paul of 

Acts: Essays in Literary Criticism, Rhetoric, and Theology (WUNT, 115; Tübingen: Mohr Siebeck, 1999), 

pp. 10–46. 

70  

as marked as perhaps they should be, because of the sparse, but persistent occurrence of the first-person plural throughout the rest of the book. A closer look at the co-text of these occurrences, however, reveals that the vast majority of the first-plural occurrences outside of the ‘we’ passages are found either in quotations or reported speech. This is just one example of the problems with simply counting and plotting the frequency of selected grammatical features without a sufficient awareness of the co-text in which they are found. Perhaps a useful and further level of annotation would be the marking off of different sections within a text, so that a grammatical search could be refined to consider just speech or just narrative sections. b. The Use of a Semantically Annotated Corpus Section 2 above introduced the concept of semantic annotation, and presented examples of the use of Louw and Nida’s lexicon to give each word a semantic tag, indicating the domain (or domains) in which it is found in the lexicon. Once this has been carried out for a large enough portion of a discourse it is possible to plot a semantic map of the discourse. The semantic plot in Appendix A shows such a semantic map for the book of Jude. The y-axis of the graph is marked with the number of each of the 93 domains in the Louw-Nida lexicon, and the x-axis has one mark per verse. The number in a particular (x, y) position indicates the number of words from domain y occurring in verse x. So tracing a row of the graph (moving horizontally) enables a selected semantic field to be traced across the text. And following a column upwards (moving vertically) shows which domains are represented in a particular verse.

The problem of words occurring in more than one domain was discussed above. For this example, the first domain of each semantic tag was taken and used for that word. So, for example, in v. 2, εἰρήνη is listed in domains 22 and 25, but in the plot it is represented by the ‘1’ at (2, 22).98 There are obvious difficulties with this policy, particularly for words that occur in a large number of domains. But as an initial indicator of lexical cohesion in a discourse, this method serves well as the starting point for more detailed and exact investigation. An examination of semantic fields within a discourse reveals information about both cohesion within the text,99 and thematic structure and prominence.100

I have added some labels to the semantic plot of the book of Jude in order to begin the process of investigation and interpretation. As with the grammatical plot in the previous section, not every cluster of occurrences from a specific domain is significant from a discourse perspective. Certain words, such as conjunctions and prepositions, occur so

                                                            98 (2, 22) represents x–y coordinates for v. 2 and domain 22. 

99 See Guthrie, ‘Cohesion Shifts’, p. 41. He says ‘it is useful to identify lexical items that play key roles 

in two or more units [paragraphs]. Semantic threads in a discourse are woven most often with the 

same, or related lexical items.’ He goes on to list relations between words by collocation, reiteration, 

synonymy, antonymy, hyponymy or metonymy. Apart from collocation, these are the sense relations 

that are central to the Louw‐Nida lexicon; see Nida and Louw, Lexical Semantics, pp. 35–114. 

100 Reed (‘Identifying Theme in the New Testament: Insights from Discourse Analysis’, in Porter and 

Carson [eds.], Discourse Analysis, pp. 75–101) states that ‘a semantic analysis of the discourse is an 

important part of determining prominence’ (pp. 83–84). He goes on to identify certain semantic 

fields that are more prominent than others, drawing on work by S. Wallace, ‘Figure and Ground: The 

Interrelationships of Linguistic Categories’, in P.J. Hopper (ed.), Tense‐Aspect: Between Semantics and 

Pragmatics (Amsterdam: Benjamins, 1982), pp. 201–23. 

71  

frequently in discourse, and are assigned to so many domains in the Louw-Nida lexicon, that it is probably unwise to make too much of the plot of domains 83 to 92. Domains 12, 25, 33, 67 and 93 have been highlighted as domains that show a high and consistent frequency of occurrence across all 25 verses. This too does not mean that these domains necessarily contain highly thematic words for the book of Jude. For instance, domain 33, one of the largest in the Louw-Nida lexicon, contains words of communication.101 It seems intuitively obvious that just about every section of New Testament discourse will contain a high number of domain 33 words. But observing where they cluster in a passage may still be informative. Likewise, domain 12 is made up of words about and related to Supernatural Beings and Powers,102 and, given the nature of much of the writing in the New Testament, references to θεός and κύριος are not surprising. But again, the distribution of these words throughout a passage may be of interest, and this is graphically illustrated by the semantic plot. In Jude, alongside references to θεός and κύριος are other domain 12 words: ἀρχάγγελλος (v. 8), διαβόλος (v. 8) and πνεῦμα (vv. 19, 20).

A closer examination of the words in domain 25, Attitudes and Emotions,103 might also provide some insight into the semantic texture and ideational function of Jude. The table below lists all of the words plotted from this domain. Both positive and negative emotions and attitudes are displayed in this list, for example, love and fear. To make any conclusions of a discourse semantics kind would require a closer look at the actors and sensers of these emotions. This would be best carried out in coordination with examination of the major participants in the discourse, which is discussed in the next section.  

verse Word verse Word

1 τοῖς ἠγαπημένοις to the beloved ones 16 ἐπιθυμίας lust/desire

2 ἀγάπη love 16 θαυμάζοντες marvelling

3 ἀγαπητοί beloved 17 ἀγαπητοί beloved

5 βούλομαι I desire 18 ἐπιθυμίας lust/desire

9 οὐκ ἐτόλμησεν he did not dare 20 ἀγαπητοί beloved

12 ταῖς ἀγάπαις love 21 ἀγάπῃ in love

12 ἀφόβως without fear 23 φοβῷ in fear

13 αἰσχύνας shame 24 ἀγαλλιάσει in rejoicing  

                                                            101 See Louw and Nida, Greek‐English Lexicon, I, pp. 388–445. 

102 See Louw and Nida, Greek‐English Lexicon, I, pp. 136–49. 

103 See Louw and Nida, Greek‐English Lexicon, I, pp. 288–320. 

72  

The previous observations, focusing on specific domains across the whole discourse, represent one use of the semantic plot. The aim is to identify significant semantic chains104 throughout the whole book, through which lexical cohesion is achieved.105 Another use of the domain plot would be to take a series of verses and move vertically up the chart noting the domains that occur within that section. On the chart in Appendix A, I have marked vv. 11–14 as a section that could warrant closer study. This span was decided on the basis of the cluster of domains 1, 2 and 3106 coinciding with 5 occurrences of domain 93 words (Names of Persons and Places). In these verses, reference to biblical characters (Cain, Baalam, Enoch, Adam, Korah) and natural substances and plants (clouds, wind, trees, waves, sea) are prominent. There is nothing spectactular about such an observation—it would be evident to any careful reader of the text. However, charting the semantic domains in this way helps summarize the complex semantic relations within the discourse. Within these four verses it can be seen that v. 12 contains three words from domain 23, Physiological Processes and States:107

These ones are the spots on your love feasts, eating together (συνευωχούμενοι 23.14) without fear, guiding themselves, clouds without water carried along by the winds, autumn trees without fruits (ἄκαρπα 23.202) having died (ἀποθανόντα 23.99) twice and been uprooted.

It must be granted that these words come from different subdomains within domain 23 and are quite separate from each other. Perhaps it is simply coincidence that three words from the same domain occur within v. 12 (that is for the analyst and interpreter to decide), being beyond the scope of this study. Also within the span of vv. 11–14 are two of the five domain 20 (Violence, Harm, Destroy, Kill) words in Jude. In fact, another two of these five are found one verse either side of the span. After v. 15, there are no further occurrences of words from this domain (see chart in Appendix A).

Finally, I have highlighted certain clusters of words that could prove to be significant. It is instructive to study the order and arrangement of the domains in the Louw-Nida lexicon.108 There is not always a semantic progression when moving from one domain to the next, but often there is some kind of relationship. For example, moving from domain 28 to domain 31, there is a definite focus on cognitive processes: 28: Know; 29: Memory and Recall; 30: Think; and 31: Hold a View. This group of domains is found as a significant cluster between v. 3 and v. 11.

                                                            104 See Reed, Discourse Analysis of Philippians, pp. 296–99. He uses the term semantic chains instead 

of domains because his analysis is limited just to the book of Philippians. Thus semantic chains are 

significant lexical clusters and chains in an individual discourse, while semantic domains describe 

related words in the language as a whole. This is probably a useful distinction to make. 

105 See Hoey, Patterns of Lexis, pp. 100–124. 

106 1: Geographical Objects and Features; 2: Natural Substances; and 3: Plants, See Louw and Nida, 

Greek‐English Lexicon, I, pp. 1–37. 

107 See Louw and Nida, Greek‐English Lexicon, I, pp. 248–77. 

108 The ‘Table of Domains’ in I, pp. xxiv–xxv, provides a useful and concise summary. 

73  

c. The Use of a Corpus Annotated with Discourse Information The final application is the use of discourse annotation to study the major participants in Jude. The text of Jude was marked up using the UCREL scheme for anaphoric annotation described in section 2 above. In addition, the level of reference has been added. Thus each reference is either reduced, where a pronoun is used to refer back to the grammaticalized referent, or implied, where the reference is implicit within the verb form. These levels are indicated by a lower case r or i placed before the referent number. So, <REF=i2 indicates that the following word is an implied anaphoric reference to the referent marked with the number 2. The first five verses of Jude, annotated in this way, are shown below:

1 (1 Ἰούδας (2 Ἰησοῦ Χριστοῦ 2) δοῦλος 1), (1 ἀδελφὸς δὲ (3 Ἰακώβου 3) 1), (4 τοῖς ἐν (5 θεῷ πατρὶ 5) ἠγαπημένοις καὶ (2 Ἰησοῦ Χριστῷ 2) τετηρημένοις κλητοῖς· 4)

2 ἔλεος <REF=r4 ὑμῖν καὶ εἰρήνη καὶ ἀγάπη πληθυνθείη. 3 (4 Ἀγαπητοί 4), πᾶσαν σπουδὴν ποιούμενος γράφειν <REF=r4 ὑμῖν περὶ τῆς κοινῆς

<REF=r1, r4 ἡμῶν σωτηρίας ἀνάγκην <REF=i1 ἔσχον γράψαι <REF=r4 ὑμῖν παρακαλῶν ἐπαγωνίζεσθαι τῇ ἅπαξ παραδοθείσῃ τοῖς ἁγίοις πίστει.

4 παρεισέδυσαν γάρ (6 τινες ἄνθρωποι‚ οἱ πάλαι προγεγραμμένοι εἰς τοῦτο τὸ κρίμα‚ ἀσεβεῖς 6), τὴν (5 τοῦ θεοῦ 5) <REF=r1, r4 ἡμῶν χάριτα μεταιθέντες εἰς ἀσέλγειαν καὶ τὸν μόνον δεσπότην καὶ (2 κύριον <REF=r1, r4 ἡμῶν Ἰησοῦν Χριστὸν 2) ἀρνούμενοι.

5 Ὑπομνῆσαι δὲ <REF=r4 ὑμᾶς <REF=i1 βούλομαι, εἰδότας <REF=r4 [ὑμᾶς] πάντα ὅτι (5 [ὁ] κύριος 5) ἅπαξ λαὸν ἐκ γῆς Αἰγύπτου σώσας τὸ δεύτερον τοὺς μὴ πιστεύσαντας <REF=i5 ἀπώλεσεν.

The following six referents from the book of Jude where marked: Jude, Jesus Christ,

James, the recipients of the letter, God and the opponents (‘they’).109 Once the corpus has been annotated in this manner, it is possible to produce tables of the references to these six discourse participants. These tables are shown in Appendix B. The table below shows an extract from the table for the opponents of the recipients of the letter, often referred to as ‘they’. The first column shows the grammaticalized reference to these participants, ‘certain people’ (v. 4), ‘the ungodly ones written about in ancient times for judgment’ (v. 5). Grammaticalized references introduce either new participants or new information about existing participants in the discourse. The next column contains reduced references (all anaphoric within Jude), which do not present any new information about the participants themselves. Instead, they point the reader back to a previous grammaticalized reference. Most frequently, reduced reference is realized in the discourse through a pronoun. It is instructive to note what kind of pronoun is used; for example, personal, demonstrative, relative or reflective.110 The tables in Appendix B show that of the nine reduced references to Referent 6

                                                            109 I have made the assumption that when the writer of Jude uses the first‐person plural that the 

recipients of the letter are included in this group. This is not necessarily the case, and requires more 

detailed interpretative work before it can be accepted as valid. But for the initial analysis, it seems a 

fair assumption to make. Certain commentators may also dispute the grouping of all the ‘they’ 

references into one participant group, but again, for initial analysis, it seems the simplest option. 

110 See Porter, Idioms of the Greek New Testament, pp. 128–38. 

74  

(‘they’), two are realized through personal pronouns,111 one with a reflexive, one relative and five through the use of a demonstrative pronoun.  

REFERENT 6 ‘They’ (Opponents of Recipients) (16–9–10)

verseGrammaticalized

Reference verse

Reduced Reference

verse Implied Reference

4 τινες ἄνθρωποι 8 οὗτοι 8 μιαίνουσιν 88/53

4

οἱ πάλαι προγεγραμμένοι εἰς τοῦτο τὸ κρίμα, ἀσεβεῖς

10 οὗτοι 8 ἀθετοῦσιν 31/76

8 ἐνυπνιαζόμενοι σάρκα 11 αὑτοῖς, 8 βλασφημοῦσιν. 33

Referent 4, the recipients of the letter, exhibits 20 reduced references: 2 are realized by reflexive pronouns and 18 through personal pronouns. It is not possible to make a direct comparison of these figures, as there is not a first- or second-person demonstrative or relative pronoun. However, where there are a number of third-person referents within a discourse, a comparison of the realization of reduced references could be revealing.

It is clear from the participant reference charts in Appendix B that referents 4 and 6 are central to the discourse. In terms of description, the major focus is upon the opponents of the letter of Jude, referred to as ‘they’. The letter recipients receive only four grammaticalized references in comparison to the 16 for the opponents. Reduced references, on the other hand, show a reversed distribution: recipients 20, opponents 9. Implied references are linked with a finite verb form, so an examination of the semantic domains of the processes of which each of the participants is the grammatical subject would also be useful. This is included in the final column of the charts. There is a correlation between the verses in which the recipients are actors (vv. 17, 21, 22, 23) and the thematic section of the discourse identified from the grammatical discourse plot above (see section 5). Also, implied references to the opponents occur between vv. 8 and 11. d. Summary It is hoped that these three examples have demonstrated the potential of adding annotation to a corpus for doing New Testament discourse analysis. They clearly illustrate the fact that adding annotation and using computerized tools does not produce instant or miraculous answers to interpretative and analytic issues within a discourse. On the other hand, corpus annotation of this type does aid the analyst in processing and interpreting the vast amount of data required for even a partial discourse analysis of any text.

                                                            111 Strictly, the third‐person personal pronoun is not found in the New Testament. However, the 

intensive pronoun, αὐτός fills its place. See Porter, Idioms of the Greek New Testament, pp. 129–30. 

75  

6. Future Developments in Text Annotation

There is really no limit to the amount and kinds of annotation that can be added to a corpus of New Testament texts. The focus of this essay has been upon the co-textual features of discourse. It is not difficult to see how broader contextual features could also be added to a corpus to allow more comprehensive analysis. Information such as theorized authorship, date and provenance, as well as more general cultural and sociolinguistic data, could also be codified.112 Due to the relative infancy of the field of corpus annotation,113 there is the opportunity to develop standards and guidelines that will avoid the duplication of effort. Given the growing importance of the Text Encoding Initiative both in the humantities and in other disciplines, it would seem wise for those working specifically with New Testament texts to conform their corpora to the TEI guidelines. This will allow for re-usability of texts and minimize incompatibility.114 In addition, this paper has argued that it is vital to keep the different types of annotation, tied to levels of discourse, separate and not to allow them to become confused.

The purpose of this study has been to introduce readers, particularly those involved in New Testament discourse analysis, to the field of corpus annotation. The sheer quantity of data produced from analysing even a short discourse, such as the book of Jude, will become obvious to such researchers. The addition of linguistic information to a computer corpus serves not only to help organize and evaluate this data, but also to reveal patterns of language that would be difficult to detect, due to the amount of data, in an unaided manual analysis. However, data without theory are next to useless, and even misleading. Corpus annotation should only ever be used as a tool to test and refine a particular theory of discourse.

 

 

 

 

 

 

 

 

 

                                                            112 See R. Hudson, Sociolinguistics (CTL; Cambridge: Cambridge University Press, 1980); Porter, 

‘Dialect and Register in the Greek of the New Testament’. 

113 The collection edited by Garside, Leech and McEnery (Corpus Annotation) claims to be the first 

book to provide a survey of the field. 

114 The use of the TEI guidelines, and specifically the SGML markup language, is discussed in more 

detail in reference to a corpus of the words of Jesus, in porter and O’Donnell, ‘The Implications of 

Textual Variants’. 

 

 

 

 

Appeendix A—Semmantic Plot off the Book off Jude 

76 

 

 

77  

Appendix B: Main Referents in Jude

REFERENT 1 Jude (2–7–2)

verse

Grammaticalized Reference verse Reduced

Referenceverse Implied Reference

1 Ιούδας Ἰησοῦ Χριστοῦ δοῦλος 3 ἡμῶν 21 ἡμῶν 3 ἔσχον 57/31/18/49/74/90/13/90

1 ἀδελφὸς δὲ Ἰακώβου 4 ἡμῶν 25 ἡμῶν 5 βούλομαι 25/30

4 ἡμῶν 25 ἡμῶν17 ἡμῶν

Pronouns (7): Personal 7 

REFERENT 2 Jesus Christ (7–0–0)

verse Grammaticalized Reference verse Reduced Reference verse Implied Reference

1 Ἰησοῦ Χριστοῦ 1 Ἰησοῦ Χριστῷ 4 κύριον ἡμῶν Ἰησοῦν Χριστὸν

17 τοῦ κυρίου ἡμῶν Ἰησοῦ Χριστοῦ

21 τοῦ κυρίου ἡμῶν Ἰησοῦ Χριστοῦ

25 Ἰησοῦ Χριστοῦ 25 τοῦ κυρίου ἡμῶν

 

REFERENT 3 James (1–0–0)

verse Grammaticalized Reference verse Reduced Reference verse Implied Reference1 Ἰακώβου

 

 

 

 

 

 

78  

REFERENT 4 Recipients of Jude (4–20–5)

verse Grammaticalized Reference verse Reduced Reference verse Implied Reference

1 τοῖς ἐν θεῷ πατρὶ ἠγαπημένοις καὶ Ἰησοῦ Χριστῷ τετηρημένοις κλητοῖς

2 ὑμῖν 17 ἡμῶν 17 μνήσθητε 29

3 Ἀγαπητοί 3 ὑμῖν 18 ὑμῖν· 21 τηρήσατε 13/37/36

17 ἀγαπητοί 3 ἡμῶν 20 ὑμεῖς 22 ἐλεᾶτε 88 20 ἀγαπητοί 3 ὑμῖν 20 ἑαυτοὺς 23 σῴζετε 21/23

4 ἡμῶν 20 ὑμῶν 23 ἐλεᾶτε 88 4 ἡμῶν 21 ἑαυτοὺς

5 ὑμᾶς 21 ἡμῶν

5 [ὑμᾶς] 24 ὑμᾶς

12 ὑμῶν 25 ἡμῶν

17 ὑμεῖς 25 ἡμῶν

Pronouns (20): Personal 18 (1P 7–2P 11)—Reflexive 2

REFERENT 5 God (8–3–3)

verse Grammaticalized Referenceverse

Reduced Reference

verse Implied Reference

1 θεῷ πατρὶ 14 αὐτοῦ 5 ἀπώλεσεν, 20/57/27

4 τοῦ θεοῦ 15 αὐτοῦ 6 τετήρηκεν, 13/37/36

5 [ὁ] κύριος 24 αὐτοῦ 14 ἦλθεν 15/1314 κύριος 21 θεοῦ

24 τῷ δυναμένῳ φυλάξαι ὑμᾶς ἀπταίστους καὶ στῆσαικατενώπιον τῆς δόξης αὐτοῦ ἀμώμους ἐν ἀγαλλιάσει

25 μόνῳ θεῷ 25 σωτῆρι ἡμῶν

Pronouns (3): Personal 3—Demonstrative 0—Reflexive 0—Relative 0

79  

REFERENT 6 ‘They’ (Opponents of Recipents) (16–9–10)

verse Grammaticalized Reference verse

Reduced Reference

verseImplied Reference

4 τινες ἄνθρωποι 8 οὗτοι 8 μιαίνουσιν 88/53

4 οἱ πάλαι προγεγραμμένοι εἰς τοῦτο τὸ κρίμα, ἀσεβεῖς 10 οὗτοι 8 ἀθετοῦσιν 31/76

8 ἐνυπνιαζόμενοι σάρκα 11 αὐτοῖς, 8 βλασφημοῦσιν. 33

12 οἱ ἐν ταῖς ἀγάπαις ὑμῶν σπιλάδες συνευωχούμενοι ἀφόβως 12 οὗτοί 10 οἴδασιν 28/32/29/87

12 ἑαυτοὺς ποιμαίνοντες 13 οἷς 10 βλασφημοῦσιν, 33

12 νεφέλαι ἄνυδροι ὑπὸ ἀνέμων παραφερόμεναι 16 οὗτοί 10 ἐπίστανται, 28/32

12 δένδρα φθινοπωρινὰ ἄκαρπα δὶς ἀποθανόντα ἐκριζωθέντα 16 ἑαυτῶν 10 φθείρονται. 20/88

13 κύματα ἄγρια θαλάσσης ἐπαφρίζοντα τὰς ἑαυτῶν αἰσχύνας 16 αὐτῶν 11 ἐπορεύθησαν 15/41/23

13 ἀστέρες πλανῆται 19 οὗτοί 11 ἐξεχύθησαν 47/15/59/90

16 γογγυσταὶ 11 ἀπώλοντο. 20/57/27

16 μεμψίμοιροι

16 κατὰ τὰς ἐπιθυμίας ἑαυτῶν πορευόμενοι

16 θαυμάζοντες πρόσωπα ὠφελείας χάριν

19 οἱ ἀποδιορίζοντες19 ψυχικοί 19 πνεῦμα μὴ ἔχοντες

Pronouns (9): Personal 2—Demonstrative 5—Reflexive 1—Relative 1

80  

Part II DISCOURSE ANALYSIS AND THE GOSPELS AND ACTS

THE HISTORIC PRESENT IN MATTHEW: BEYOND SPEECH MARGINS Stephanie L. Black

1. Introduction

Scholars have long observed how frequently the historic present appears in Matthew, in fact in all the Gospels except Luke,1 but the use Matthew makes of the historic present in discourse structure has yet to be adequately explored.2 As he constructs his Gospel, why does the evangelist choose at some points to employ the historic present, that is, a present tense-form in past-referring narrative passages where the reader might expect an aorist or imperfect finite verb? In this essay I will offer some initial steps toward a discourse approach to the historic present in Matthew by examining his use of these forms in the two narrative passages that have the highest concentration of historic present finite verbs: the temptation of Jesus (4:1–11) and Jesus’ prayer in Gethsemane (26:36–46). I will show that within these passages Matthew intentionally juxtaposes present tense-forms with aorist tense-forms in a structured manner for dramatic effect.

This study looks closely at sentences in narrative that are not speech margins. The high incidence of the historic present in speech margins, that is, in sentences in which λέγω or φημί introduces quoted speech, has understandably received a good deal of attention.3 By my count, present tense-forms occur 79 times as the main verbs in past-referring narrative sentences in Matthew, 66 of these (84%) in sentences which introduce quoted speech.4 Fifty-nine of the 79 (75%) are forms of λέγω, and there is one occurrence of φημί (14:8). In six                                                             1 See A.T. Robertson, A Grammar of the Greek New Testament in the Light of Historical Research 

(New York: Hodder & Stoughton, 4th edn, 1934), pp. 866–69; BDF, p. 167; N. Turner, A Grammar of 

New Testament Greek. III. Syntax (ed. J.H. Moulton; Edinburgh: T. & T. Clark, 1963), pp. 60–62. 

2 A pioneering attempt in this area is S.H. Levinsohn, Discourse Features of New Testament Greek 

(Dallas: Summer Institute of Linguistics, 1992), pp. 141–60. Levinsohn focuses primarily on speech 

margins, but does suggest that one function of the historic present in Matthew’s narrative is to 

‘detach’ (highlight or downgrade) either the preliminary or concluding event or speech of an 

incident. ‘Most commonly, detachment downplays preliminary events but highlights concluding 

ones, in relation to the event from which they are detached’ (Levinsohn, Discourse Features, p. 147.) 

Levinsohn also argues that the historic present is used to introduce a new paragraph when that 

paragraph is a new section in a larger incident. Levinsohn seeks to explain a number of individual 

occurrences of the historic present in Matthew, particularly those in speech margins, but does not 

address Matthew’s use of the historic present in the overall narrative structure of the passages 

presented here. 

3 See Turner, Syntax, p. 61; Levinsohn, Discourse Features, pp. 141–60; B.M. Fanning, Verbal Aspect in 

New Testament Greek (Oxford: Clarendon Press, 1990), pp. 231–32; K.L. McKay, A New Syntax of the 

Verb in New Testament Greek (New York: Peter Lang, 1994), p. 42. 

4 This includes only those uses of the historic present that occur in Matthew’s narrative framework, 

and not those that occur in parables. 

81  

more instances, a verb which is not itself a verb of speaking appears in a present tense-form in combination with a present participle of λέγω to introduce quoted speech (2:13; 2:19; 3:1; 9:14; 15:1 and 22:16). However, for reasons to be discussed below, a primary focus of this essay will be Matthew’s use of the historic present in clauses that are not speech margins, but that function in the narrative framework of Matthew’s Gospel in combination with present-tense speech margins.5

My interest here is primarily language-in-use. I will not try to define what the historic present ‘means’. Rather, I will examine, first, where Matthew chooses to use historic present forms, and secondly, what contribution those choices make to the structure of the passages in question. a. Methodology These observations on the historic present in Matthew’s narrative have grown out of a larger project on connective words in Matthew. In order to analyse the use of intersentential conjunctions (that is, conjunctions introducing and linking sentences) to study their use in discourse cohesion, a computer database of sentences in Matthew was created. Each of these sentences, about 2300 in all, was analysed for a number of factors, including its intersentential conjunction (or asyndeton, i.e. no conjunction); any constituents of the sentence fronted before the main verb; the tense-form of the main verb; the discourse type in which the sentence occurs (narrative, exposition, speech or Old Testament quotation), and other characteristics such as whether the sentence functions as a speech margin or whether it occurs in a parable.

To compare variations among discourse types in the use of connective words, each sentence was designated as narrative (events recounted from the evangelist’s viewpoint in roughly temporal order, forming the framework for the discourse), exposition (longer discourse sections attributed to Jesus, including parables), speech (conversation; short statements attributed to Jesus, especially as a reply to another’s question or statement), or Old Testament quotation (longer citations as indicated in Nestle-Aland, Novum Testamentum Graece, 27th edn, which forms the text for this research). These distinctions proved to be fruitful in analysing the framework of the evangelist’s discourse. It became possible to isolate the narrative sentences, dropping out the intervening speech and expository segments, and, as a result, to follow more closely the underlying structure of Matthew’s storytelling.

An examination of Matthew’s use of τότε in narrative led to the observation of its association with the present tense, and prompted an interest in exploring further how the present tense is used in the discourse structure of Matthew.6 Occurrences first in speech margins in narrative and then in other narrative clauses were selected using the database. In this way, Matthew’s use of the present tense in narrative outside of speech margins became more readily observable, and the passages of interest here were identified. b. Traditional Views of the Historic Present in Matthew Grammarians agree that the use of the historic present is characteristic of Matthew, though it is somewhat less common in Matthew than in Mark or John. Hawkins lists 78 occurrences of                                                             5 By ‘narrative framework’ I mean the series of clauses that recount and evaluate events in the 

Gospel from the evangelist’s viewpoint in roughly temporal order, omitting those clauses that 

contain speech or exposition by characters within the narrative. See genre definitions under 

‘Methodology’, below. 

6 By my count, τότε occurs as the intersentential conjunction 55 times in narrative in Matthew, 20 of 

these with present tense verbs. 

82  

the historic present in narrative in Matthew, plus 15 more in parables.7 Matthew sometimes agrees with Mark in the use of the historic present, but is not dependent on Mark’s choices. Hawkins observes that Matthew parallels only 21 of Mark’s 150 occurrences of the historic present,8 and Fanning notes that he uses the historic present an additional 72 times independently of Mark.9

Two observations about the use of the historic present in the New Testament repeatedly appear in both traditional and more recent grammars. The first is that the historic present is used where the author wants to make the narrative more vivid.10 This usage is said to be common in classical Greek, the papyri, the LXX, Josephus and other writings, as well as in the New Testament.11 The use of the historic present to make narrative more lively is found in a number of languages, both ancient and modern,12 although several grammarians suggest that Luke’s reluctance to use it is due to his disdaining it as too vulgar.13 Robertson recognizes the potential narrative impact of the historic present when he suggests, ‘Modern literary English abhors this idiom, but it ought to be preserved in translating the Gospels in order to give the same element of vividness to the narrative’.14

While modern grammarians generally agree that the historic present is used for dramatic effect, how that effect is created is a point of debate. Fanning writes The commonly accepted explanation of the Greek historical present … is that the present is used to bring a past occurrence into immediate view, portraying the event as though it occurs before the reader’s eyes … It is the argument of this book that … the key feature which prompts the use of the present is the temporal transfer, not some sort of aspectual effect.15 However, in each of the two passages to be considered below the historic present is intermingled with other tense-forms in a single series of closely related events. It is difficult to understand how Matthew can be repeatedly thrusting one action within the pericope into the present time (rhetorically) and then bringing his audience back to a subsequent action which is narrated as occurring in past time. In both of these passages, there are too many changes of tense-form in an uninterrupted series of events to make this an adequate explanation. Porter argues that ‘several recent grammarians rightly indicate that different tense-forms can obviously be used in similar temporal contexts, and that the imperfective verbal aspect

                                                            7 J.C. Hawkins, Horae Synopticae (Oxford: Clarendon Press, 2nd edn, 1909), pp. 148–49. My count 

includes the same 78, plus an additional occurrence of λέγει, in 26:25. 

8 Hawkins, Horae, p. 148. 

9 Fanning, Verbal Aspect, p. 238. 

10 See BDF, p. 167; Turner, Syntax, p. 60. See also S.E. Porter, Verbal Aspect in the Greek of the New 

Testament, with Reference to Tense and Mood (New York: Peter Lang, 1989), p. 196; Fanning, Verbal 

Aspect, pp. 226–39; McKay, Syntax, p. 42. 

11 See BDF, p. 167; Turner, Syntax, p. 60. 

12 See Robertson, Grammar, p. 866; Turner, Syntax, p. 60. 

13 See Robertson, Grammar, p. 867; BDF, p. 167; Turner, Syntax, p. 61. 

14 Robertson, Grammar, p. 868. 

15 Fanning, Verbal Aspect, pp. 226–27. 

83  

(present tense-form) is used whenever one wishes to draw added attention to a given event’.16 As I will show, in 4:1–11 and 26:36–46 Matthew uses the historic present in a manner best explained by an aspectual approach to verbal tense-forms. He alternates the historic present with other tense-forms to help create the narrative’s structure and make his storytelling more engaging to his readers. In 4:1–11, he uses this juxtaposition of tenses to signal the development to a climax, and in 26:3–46, to distinguish between two interwoven storylines.

The second point of consensus is that the historic present is commonly associated with verbs of speaking, especially in Matthew’s Gospel. English speakers will recognize this type of use in the ‘he says … then she goes …’ forms found in informal English today. Turner notes that in post-classical Greek there is an increasing tendency to find the historic present with λέγει and other verbs of speaking, as well as with verbs of seeing and verbs of motion.17 Hawkins observes that in Matthew λέγει, λέγουσιν and φησίν ‘constitute slightly more than three-fourths of the whole number, being 59 out of the 78; while in Mark they constitute less than half, being 72 out of the 151’.18 Recent commentaries on Matthew continue this emphasis on his use of the historic present in speech margins, as when Luz writes: With λέγω, it [the present tense] usually is left standing or is even newly written by the evangelist. Since Matthew in his narratives likes to tighten the narrative and thus lets the dialogue become prominent, perhaps the historical present also is a means of directing the attention of the readers to the most important element in the narratives, namely, the sayings of Jesus.19 c. Rethinking Speech Margins It is not surprising that so much attention has been paid to the historic present in speech margins in Matthew—there simply are so many of them. An analysis by discourse type shows that about half of the sentences functioning as speech margins in exposition and speech have main verbs with present tense-forms, while about one-quarter of the sentences functioning as speech margins in the narrative framework have present tense-form main verbs. In each case, sentences functioning as speech margins are more likely to occur in the present tense than are non-speech margins in the same discourse type: • In exposition, 51% of the sentences functioning as speech margins have main verbs

with present tense-forms (53 of 103 sentences). Only 29% of sentences in exposition that are not speech margins have present tense-forms (193 of 665 sentences).

• In speech, 49% of the sentences functioning as speech margins for inset speeches or quotations have main verbs with present tense-forms (27 of 55 sentences). Only 35% of sentences that are not speech margins have present tense-forms (234 of 678 clauses).

                                                            16 S.E. Porter, Idioms of the Greek New Testament (Sheffield: Sheffield Academic Press, 2nd edn, 

1994), p. 31. Porter continues, ‘Even though the historic present and other tense‐forms may be used 

in past contexts, this does not mean that the historic present and, for example, the aorist mean the 

same thing—their sphere of temporal reference may overlap, but their verbal aspect is different. 

Whereas the aorist is merely used in its common narrative function, the present form draws added 

attention to the action to which it refers.’ 

17 Turner, Syntax, p. 61. 

18 Hawkins, Horae, p. 148. 

19 U. Luz, Matthew 1–7: A Commentary (trans. W.C. Linss; Minneapolis: Augsburg, 1989), p. 52. 

84  

• In narrative, the contrast is more striking: 22% of the sentences functioning as speech margins have main verbs with present tense-forms (66 of 299 clauses), while less than 4% of sentences in narrative that are not speech margins have present tense-forms (15 of 417 sentences20).

In other words, while speech margins in exposition and speech are about one and a half

times as likely to be in the present tense as other sentences, in narrative speech margins are almost six times more likely to be in the present tense than other clauses. As stated above, the prevalence of present tense-form speech margins in Matthew has inevitably led to scholarly attention to them in treatments of Matthew’s Gospel. However, it is my contention that a linguistic approach suggests a different understanding of the data. Given that it is by no means unusual for present tense-forms to be used with speech margins in biblical Greek (and in other languages), their interpretative significance may often be overstated. To the extent that their use approaches that of a stereotyped idiom, the communicative force of present tense-form speech margins diminishes.

Lyons’s dictum, ‘The “meaningfulness” of utterances (and parts thereof) varies in inverse proportion to their degree of “expectancy” in context,’ is relevant here.21 Lyons states that meaning ‘is in principle quantifiable with respect to “expectancy” (or probability of occurrence) in context’.22 The more likely a form is to appear in a given context, the less meaning it tends to convey. Conversely, the less likely a form is to appear in a given context, the more meaning it tends to convey in that context. Put more simply, when an author settles for a stereotyped word or phrase or form—whatever one normally says in such circumstances—less freshness of thought and real meaning is communicated. When an author reaches for a word or phrase, or in this case a tense-form, that might be considered unusual or in some way unexpected at a particular point in a narrative, more thought and intended impact can be assumed to lie behind that choice.

I suggest that the use of the historic present in speech margins in Matthew, especially with λέγω, approaches that of a stereotyped idiom.23 Its use is routine enough that it often conveys little new information. It is not that there is never interpretative significance in the choice of a present tense-form in a speech margin when another tense-form could have been used. In fact, I will show below how Matthew uses the historic present in both speech margins and non-speech margins to structure two narrative passages. However, the exegete should be cautious about reading too much meaning into a verb of speaking simply on the basis of its occurring in the present tense.

While this approach tends to decrease the importance one places on present tense-form speech margins, it should heighten our interest in the less predictable uses of the historic

                                                            20 This includes the 13 historic present forms dealt with here, plus two uses of εἰμί in what might be 

described as editorial comment: the list of the names of Jesus’ disciples (10:2), and the translation of 

his cry from the cross (27:46). 

21 J. Lyons, Introduction to Theoretical Linguistics (Cambridge: Cambridge University Press, 1968), p. 

415. 

22 Lyons, Introduction, p. 415. 

23 Fanning comes to a similar conclusion regarding the historic present with verbs of speaking in 

Mark. He then builds on Buth’s work in developing a list of discourse functions for the historic 

present in Mark (R. Buth, ‘Mark’s Use of the Historical Present’, Notes on Translation 65 [June 1977], 

pp. 7–13). Fanning, Verbal Aspect, pp. 231–32. 

85  

present, specifically the 13 occurrences found in Matthew’s narrative framework which do not introduce quoted speech. These are found in 3:13, 15; 4:5, 8 (two occurrences), 11; 17:1 (two occurrences); 26:36, 40 (two occurrences), 45; and 27:38. As the analysis above shows, while speech margins in exposition and speech are about one and a half times as likely to be in the present tense as are other sentences in the same genre, speech margins in narrative are almost six times more likely to be in the present tense than are other sentences. Or stated the other way around, by far the least expected use of the historic present in Matthew is in narrative clauses which are not speech margins. Since these 13 occurrences are much less predicable, when we do find Matthew choosing to use the present tense in his narrative framework, it is worth asking what the choice of such a form at this point in the narrative contributes to the discourse.24

2. The Historic Present in Discourse Structure

To restate the question, how does the use of the historic present, especially in those sentences which are not speech margins, contribute to discourse structure in Matthew?

The first thing to observe when considering these 13 occurrences of the present tense in Matthew’s narrative framework is that 12 of them are found in clusters at important points in the narrative: Jesus’ baptism (3:13, 15); Jesus’ temptation (4:5, 8 [two occurrences], 11); Jesus’ transfiguration (17:1 [two occurrences]), and Jesus’ prayer in Gethsemane (26:36, 40 [two occurrences], 45).25 There is one other occurrence in narrative, when two thieves are crucified alongside Jesus (27:38).

The unexpectedness of the historic present in narrative sentences outside speech margins suggests that where they do occur the evangelist has made a conscious choice to use them. The clustering of 12 of these in close proximity to the narrative ‘seams’ of 4:17 and 16:21, ἀπὸ τότε ἤρξατο ὁ Ἰησοῦς … 26 and at the beginning of the passion narrative increases the likelihood that they have discourse-structuring functions. The first occurrences of the historic present noted above appear in chs. 3 and 4, chapters that are often understood to work together in Matthew’s introduction of Jesus’ unique identity as the Son of God. The two                                                             24 Several scholars associated with the Summer Institute of Linguistics have taken a similar approach 

in studies of Mark and John. Each makes a distinction between the historic present in speech margins 

and in clauses which do not introduce speech, and each finds a prominence function for historic 

present forms in the discourse structure of the Gospel. See D. Boos, ‘The Historical Present in John’s 

Gospel’, START 11 (April 1984), pp. 17–24; Buth, ‘Mark’s Use of the Historical Present’, pp. 7–13; J. 

Callow, ‘The Function of the Historic Present in Mark 1:16–3:6; 4:1–41; 7:1–23; 12:13–34’, START 11 

(April 1984), pp. 9–17; S.H. Levinsohn, ‘Preliminary Observations on the Use of the Historic Present in 

Mark’, Notes on Translation 65 (June 1977), pp. 13–28. 

25 As mentioned above, Turner notes the increasing tendency to find the historic present with verbs 

of motion, ‘especially coming and going (also frequ. in LXX later historical books)’ in post‐classical 

Greek (Turner, Syntax, p. 61). Of the 12 occurrences listed here, at least 4 are verbs of motion: 

ἔρχομαι (26:36, 40, 45), παραγίνομαι (3:13), and perhaps ἀφίημι (4:11; cf. 3:15, which also has ἀφίημι but in the sense of ‘permit’). παραλαμβάνω appears 3 times (4:5, 8; 17:1) and ἀναφέρω once (17:1). The other verbs are δείκνυμι (4:8) and εὑρίσκω (26:40). Characteristic of Matthew, 6 of the 12 

occur with the conjunction τότε: 3:13, 15; 4:5, 11; 26:36, 45. 

26 See J.D. Kingsbury, Matthew: Structure, Christology, Kingdom (Philadelphia: Fortress Press, 1975), 

pp. 7–8. 

86  

appearances of the historic present in 17:1, where Jesus takes Peter, James and John up on a mountain for his transfiguration, closely follow the ‘seam’ in 16:21 in the narrative framework. Chapter 26, the beginning of the passion narrative, has the highest concentration of historic present forms (both speech margins and non-speech margins), with 14 of the 79 occurrences in Matthew being found here.

The smaller set of instances where a verb which is not itself a verb of speaking appears in a present tense-form in combination with a present participle of λέγω to introduce quoted speech (e.g. ‘Then John’s disciples come to him, saying …’ 9:14) are more difficult to categorize. There are six of these occurrences in Matthew (2:13; 2:19; 3:1; 9:14; 15:1; 22:16), which thus far I have treated as sentences functioning as speech margins. However, further observation of ‘compound’ speech margins (in which a verb that is not a verb of speaking appears in any tense-form in combination with a present participle of λέγω) suggests that in some discourse features they are more similar to other narrative sentences than to other speech margins. Their distribution of intersentential conjunctions follows the overall pattern for narrative sentences, rather than showing the increased incidence of δέ and asyndeton, and decreased use of καί that is characteristic of sentences with verbs of speaking. And the frequency of present-tense forms in such compound speech margins appears to be somewhat higher than in all narrative sentences (6 of 97, or 6%), but not as high as in other speech margins. This raises the question of whether present tense-forms appear in these sentences because of their function as speech margins, or whether they, too, have discourse structuring functions—although of course these need not be mutually exclusive functions—or whether other factors influence their use.27 The occurrence in 3:1 certainly fits into the concentration of present tense-forms in chs. 3 and 4, and indeed 3:1 begins the section which ends in 4:16. The uses in 9:14, 15:1 and 22:16 serve as margins for questions challenging Jesus’ attitudes toward tradition or law.

The use of all these historic present tense-forms in the macrostructure of Matthew merits further investigation. However, in the rest of this study I will primarily address microstructures: Jesus’ temptation in the wilderness (4:1–11), and his prayer in Gethsemane (26:36–46). In fact, as shall be seen when examining these two passages with the highest concentration of historic present forms (each having a mix of speech margins and other narrative sentences), the evangelist juxtaposes present and aorist or imperfect tense-forms for dramatic impact in the structure of these two important pericopes. a. Jesus’ Temptation (Matthew 4:1–11) The temptation of Jesus stands at the end of Matthew’s introductory section (1:1–4:11), immediately before Jesus begins his public ministry. Jesus’ 40 days of fasting and subsequent temptation in the wilderness parallel Israel’s 40 years of wandering in the desert. Jesus repeats Israel’s experience of hunger and temptation to idolatry, but this time there is no failure. Lest we overlook the parallel, Matthew includes three quotations by Jesus from Deuteronomy 6–8, drawing attention to the Deuteronomy narrative. Davies and Allison write, ‘All-important for a right understanding of our pericope is Deut 8:2–3: “And you shall remember all the way which the Lord your God has led you these forty years in the wilderness, that he might humble you, testing you to know what was in your heart, whether you would keep his commandments, or not. And he humbled you and let you hunger …” ’28 Hagner adds that

                                                            27 Again, lexical choice may play a role here in the use of the historic present with verbs of motion: in 

both 9:14 and 15:1, the verb is προσέρχομαι. 

28 W.D. Davies and D.C. Allison, A Critical and Exegetical Commentary on the Gospel According to 

Saint Matthew (3 vols.; Edinburgh: T. & T. Clark, 1988), I, p. 352. 

87  

‘whereas Israel failed its test in the wilderness, Jesus succeeds, demonstrating the perfection of his own sonship. This account is placed here deliberately because it serves as an important prolegomenon to the ministry of Jesus.’29 At a significant point in his development of Jesus’ identity as the faithful Son of God, Matthew engages his readers’ attention by incorporating a cluster of historic present forms into his storytelling.

In order to make the framework of 4:1–11 more explicit, I have arranged the passage by narrative sentences below, showing for each its intersentential conjunction and any fronted constituents (that is, sentence components that occur before the main verb), giving an English translation for the main constituents of the sentence (following the RSV as much as possible), and noting the tense-form of the main verb. ‘Aor ptc’ refers to an aorist participle placed before the main verb, and the symbol ‘Ø’ indicates the lack of an intersentential conjunction. Sentences with present tense-forms are shown in italics.

Speech sentences have been omitted, shown by ‘[speech]’ in the places where they occur in context. It is not that the speech sections are unimportant in the narrative. Quite the opposite, the speech sections convey the content of the exchanges between Jesus and the devil, and carry the weight of the parallels with Deuteronomy. However, my interest here is in the methods by which Matthew engages the reader in his story, some of which become more apparent when the narrative framework is considered alone.

Based on structural observations as well as content, the passage has been divided into four sections, one section for each of the three exchanges between Jesus and the tempter, and a fourth concluding section:

4:1 τότε + fronted subject + ‘Jesus was led up by the Spirit into the wilderness’ (aor)

4:2 καί + aor ptc + adverb + ‘he was hungry’ (aor)

4:3 καί + aor ptc + fronted subject + ‘the tempter said to him’ (aor)

[speech]

4:4 δέ + aor ptc + ‘he [Jesus] said’ (aor)

[speech]

4:5a τότε + ‘the devil takes him to the holy city’ (pres)

4:5b καί + ‘he set him on the pinnacle of the temple’ (aor)

4:6 καί + ‘he says to him’ (pres)

[speech]

4:7 Ø + ‘Jesus said to him’ (aor)

[speech]

4:8a πάλιν + ‘the devil takes him to a very high mountain’ (pres)

4:8b καί + ‘he shows him all the kingdoms of the world …’ (pres)

4:9 καί + ‘he said to him’ (aor)

                                                            29 D.A. Hagner, Matthew 1–13 (WBC, 33A; Dallas: Word Books, 1993), p. 62. 

RSV Revised Standard Version 

88  

[speech]

4:10 τότε + ‘Jesus says to him’ (pres)

[speech]

4:11 τότε + ‘the devil leaves him’ (pres)

4:11b καί + ἰδού + fronted subject + ‘angels came’ (aor)

4:11c καί + ‘they ministered to him’ (imperfect; also imperfect in Mark)

Although Matthew agrees with Mark’s order in placing Jesus’ temptation immediately

after his baptism and his attestation of Jesus as the beloved Son of God, Matthew’s account is quite distinct from Mark’s. Each begins with a short introductory statement (Mt. 4:1; Mk 1:12), which differ both in the lexical form used to describe Jesus’ progress into the desert and in the tense-form used. Matthew has an aorist form where Mark has a historic present. Matthew 4:2–10 is not paralleled in Mark. Instead, Mark has a one-verse summary (1:13). Mark’s last phrase ‘the angels ministered [imperfect] to him’ then appears in almost identical form at the end of Matthew’s account (4:11), with the same imperfect tense.

Although there are a number of differences, Mt. 4:1–11 parallels Lk. 4:1–13, if Luke’s order is rearranged as Lk. 4:1–4, 9–12, 5–8, 13 (in other words, the second and third interchanges between Jesus and the devil are reversed). Luke places his version of Jesus’ genealogy between the baptism and temptation pericopes. Characteristically, Luke uses no present tense-forms in the main narrative clauses in his account of Jesus’ temptation.

Matthew, then, stands alone against Mark in including an extended account of Jesus’ temptation in the wilderness, and against Luke in Matthew’s reworking of the material, particularly in his use of the historic present to draw his readers into the narrative.

Observing the narrative framework as organized above reveals that Matthew increasingly uses the historic present, intermixed with other tense-forms, to convey the growing tension as the narrative builds to a climax in the third temptation. The pericope’s framework consists of four sections, each with four sentences, except for the final section, which has three sentences, lacking a concluding statement by Jesus. In the first exchange, no present tense-forms are used. In the second exchange two appear, as the devil presents a more enticing temptation. Then in the climactic third exchange, three present tense-forms appear, most notably as Jesus gives the final rebuttal, ‘You shall worship the Lord your God and him only shall you serve’.30 Only one present tense-form appears in the final section as the devil, defeated, departs.

It is possible that the imperfect tense-form, διηκόνουν, in 4:11 serves as a less ‘marked’ instance of the imperfective aspect in the final section after the climax of the pericope, that is, contributing to the structure as the present tense-form does, but with less foregrounding emphasis. This is the approach to the imperfect Porter takes when he writes, ‘The imperfect is similar in function to the historic use of the present. Although they share the same verbal

                                                            30 There is a textual variant of λέγει for εἶπεν in v. 6, but if this were the preferred reading, it would only strengthen the argument for a greater concentration of historic present forms at the climax of 

the pericope. See also R.H. Gundry, Matthew: A Commentary on his Literary and Theological Art 

(Grand Rapids: Eerdmans, 1982), pp. 57–58. 

89  

aspect the present is used to draw even more attention to an action.’31 It is certainly difficult to see how a traditional understanding of the imperfect as continuous, progressive or iterative makes sense in this context (although Fanning labels διηκόνουν in 4:11 an ‘inceptive imperfect’).32 Robertson observes, ‘Where the aorist and the imperfect occur side by side, it is to be assumed that the change is made on purpose and the difference in idea to be sought. In juxtaposition the aorist lifts the curtain and the imperfect continues the play’, citing 4:11 as an example.33 A fuller study of the interaction between aorist, imperfect and present tense-forms in Matthew’s narrative from an aspectual approach which takes into consideration planes of discourse may reveal more.34

Other syntactical choices contribute to the structure of the pericope. The subject is fronted to mark the first appearance of each of the active participants in this passage: Jesus (4:1), the tempter (4:3) and the angels (4:11). The intersentential conjunctions add to the development of the structure as well. Each of the four sections begins with τότε, except the climactic third section, where τότε accompanies Jesus’ final reply. Here the reversed order (with τότε in the last sentence of the section) helps mark the dramatic peak of the passage. In each section there is a first sentence, usually with τότε, setting the scene in motion, followed by two καί sentences carrying the scene forward, followed by a statement by Jesus introduced by a differing conjunction. The exception is the final section, which has no fourth sentence, as there is no need for Jesus to speak again once he has given his final reply and the temptations have ceased. The clauses introducing Jesus’ statements move from less to more ‘marked’ as speech margins: from δέ to asyndeton to τότε, and from aorist to present tense,35 although the use of φημί in 4:7 also represents a marked lexical choice.36 Levinsohn cites Jesus’ final reply in 4:10 as an example of the use of the historic present in speech margins in Matthew ‘to detach the concluding event or speech of an incident. The rhetorical effect is often, but not

                                                            31 Porter, Idioms, p. 34. Fanning also notes that the present and imperfect share the same ‘basic 

aspect‐value’, which he describes as portraying ‘a specific situation (action or state) viewed as it is 

going on, with the imperfect necessarily used with respect to past time’ (Verbal Aspect, pp. 240–41). 

32 Fanning, Verbal Aspect, pp. 252–53. 

33 Robertson, Grammar, p. 838. 

34 There is a similar sequence in 8:14–15, closely paralleled in Mk 1:31 and Lk. 4:39, where a short 

unit, the healing of Peter’s mother‐in‐law, also ends with διακονέω in an imperfect tense‐form: καὶ ἠγέρθη καὶ διηκόνει αὐτῷ. 

35 δέ is the most common conjunction introducing sentences functioning as speech margins in 

narrative in Matthew, with 42% (125 of 299) following this pattern. Asyndeton occurs with 13% of 

speech margins in narrative (40 of 299) and τότε in 9% (27 of 299). Aorist main verbs appear in 69% 

of sentences functioning as speech margins in narrative (205 of 299), and present tense‐forms in 22% 

(66 of 299). 

36 φημί occurs 11 times in Matthew’s narrative framework (4:7; 8:8; 14:8; 17:26; 19:21; 21:27; 22:37; 

26:34; 27:11, 23, 65). These 11 speech margins all introduce significant statements or 

pronouncements. In seven of them Jesus is the subject of the verb (4:7; 17:26; 19:21; 21:27; 22:37; 

26:34; 27:11); in two Pilate is the subject (27:23; 27:65); in one it is the daughter of Herodias 

demanding John the Baptist’s head on a platter (14:8); and in one it is the centurion confessing his 

understanding of Jesus’ authority (8:8). 

90  

always, that of highlighting the act concerned.’37 He observes elsewhere that one use of τότε in Matthew is to introduce a concluding speech ‘to which an incident has been building up’.38

This analysis of the function of the historic present in Mt. 4:1–11 both confirms established exegetical insights about the climactic structure of the pericope, and answers questions about the seemingly random use of tenses by Matthew in this passage. Concerning the pericope’s climactic structure, Davies and Allison note that

the three temptations exhibit a spatial progression, from a low place to a high place. The first takes place in the desert, the second on a pinnacle in the temple, the third on a mountain from which all the kingdoms of the world can be seen. This progression corresponds to the dramatic tension which comes to a climax with the third temptation.39

In identifying the third temptation as the climax of dramatic tension, Davies and Allison argue only from the sense of the passage, in spite of the fact that they offer detailed analyses of phrases and clauses in the Greek text. In fact, the progression that they identify from the first temptation to the third directly parallels the progression from less to more use of the historic present in the account. Matthew is using both lexical choice (desert to temple to mountain) and syntactical structures (increasing use of the historic present) to build to a climax in his narrative.

Without taking into consideration the way he uses the historic present to structure discourse above the level of individual sentences, Matthew’s mix of tenses seems arbitrary. Hagner demonstrates the difficulty faced by those few scholars who attempt to explain the changes in tense in this passage when he writes:

Here, in παραλαμβάνει, ‘take,’ [4:5] and in λέγει, ‘says,’ of v 6, he shifts to the historical present, but not consistently (ἔστησεν, ‘placed,’ is again aorist, like the verbs in vv 1–4). If Q had the narrative in the historical present, Luke has been more consistent in changing the verbs to aorist; more probably, however, Matthew inconsistently alters some of the verbs to the historical present (by influence of oral tradition?).40

And again, ‘The aorist εἶπεν, “he said,” [4:9] is surprisingly used after two successive historical present tenses in the preceding verse. Further historical present tenses are used in vv. 10 (λέγει, “he says”) and 11 (ἀφίημι, “leaves”).’41

If Matthew has been inconsistent about altering verbs from some other source, it appears that he has done so intentionally, with an eye to the structure of his discourse. Of course, historic present forms may have existed in whatever sources Matthew used and he may have taken over any or all of them. My interest lies in Matthew’s final product and the results of the decisions he made to introduce or retain historic present forms at significant points in his discourse. As has been shown, these choices do not appear to be arbitrary, but contribute to a highly structured narrative.

                                                            37 Levinsohn, Discourse Features, p. 144. Levinsohn uses the term ‘detachment’ to refer to the 

highlighting or downgrading of information. 

38 Levinsohn, Discourse Features, p. 52. 

39 Davies and Allison, Matthew, I, p. 352. 

40 Hagner, Matthew 1–13, p. 66. 

41 Hagner, Matthew 1–13, p. 68. 

91  

In Jesus’ temptation in the wilderness, Matthew moves from fewer to more instances of the historic present in subsequent sections of the passage to convey increasing drama as the narrative builds to a climax. In the first exchange between Jesus and the devil, no present tense-forms are used; in the second exchange two appear; in the climactic third exchange, three present tense-forms appear. Intermixed with other tense-forms in a balanced structure, these make his storytelling more vivid. b. Jesus’ Prayer in Gethsemane (Matthew 26:36–46) The Gethsemane pericope offers the most concentrated use of the historic present in narrative in Matthew, in terms of numbers of occurrences in close proximity, as well as showing examples of its association with τότε. Again, in order to see the narrative framework, I have arranged the passage by sentences, showing for each its intersentential conjunction, giving an English translation for the main constituents of the sentence (following the RSV as much as possible), and noting the tense-form of the main verb. As before, ‘aor ptc’ refers to an aorist participle placed before the main verb; speech segments have been omitted, shown by ‘[speech]’ in the places where they occur; and sentences with present tense-forms are shown in italics:

26:36a τότε + ‘Jesus goes with them’ (pres)

26:36b καί + ‘he says to the disciples’ (pres)

26:36c [speech]

26:37 καί + aor ptc + ‘he began to be sorrowful and troubled’ (aor)

26:38a τότε + ‘he says to them’ (pres)

26:38b [speech]

26:39a καί + aor ptc + ‘he fell on his face’ (aor)

26:39b [speech]

26:40a καί + ‘he comes to the disciples’ (pres)

26:40b καί + ‘he finds them sleeping’ (pres)

26:40c καί + ‘he says to Peter’ (pres)

26:40b–41 [speech]

26:42a πάλιν ἐκ δευτέρο + aor ptc + ‘he prayed’ (aor)

26:42b [speech]

26:43a καί + aor ptc + ‘again he found them sleeping’ (aor)

26:43b γάρ + ‘their eyes were heavy’ (impf)

26:44 καί + aor ptc + aor ptc + ‘he prayed for the third time’ (aor)

26:45a τότε + ‘he comes to the disciples’ (pres)

26:45b καί + ‘he says to them’ (pres)

26:45c–46 [speech]

92  

Matthew 26:36–46 closely parallels Mk 14:32–42. Each of the present tense-forms in Matthew’s version occurs in Mark as well, but one of Mark’s historic present forms (Mk 14:33) appears as an aorist participle in Mt. 26:37.42 The two versions also differ in the presence of the characteristic Matthean conjunction τότε. The intersentential conjunction in Mark’s version is always καί, except for the γάρ in v. 40 (also γάρ in Mt. 26:43b). Gundry suggests that by using τότε in v. 36, ‘Matthew relates the disciples’ prayerlessness, especially Peter’s, to the foregoing prediction of falling into sin. The one fulfills the other.’43 Luke’s version, which shows some divergence from the other two Synoptics, characteristically does not use the present tense.

Thus the presence of the historical present in this pericope cannot be said to be unique to Matthew. However, the contrast with Luke, and the fact that Matthew elsewhere uses or omits the historic present independently of Mark, indicates that Matthew makes conscious choices to introduce (if one accepts the possibility of Matthean priority) or retain (accepting the stronger hypothesis of Markan priority) historic present forms in this passage. Whatever the origin of these forms, Matthew again makes use of the storytelling power of the juxtaposition of different tense-forms to draw his readers into his narrative.

The analysis of the narrative framework shown here suggests that this passage is most naturally seen as one unit with two intertwining narrative threads. The first storyline, containing historic present verb forms, follows Jesus’ interactions with his disciples: Then Jesus goes with them … and he says to the disciples … Then he says to them … And he comes to the disciples and he finds them sleeping, and he says to Peter … Then he comes to the disciples and he says to them … The other storyline, containing aorist verbs, along with aorist participles to give more depth of detail, focuses on Jesus in isolation—in his grief, in prayer to his Father, and as he gazes on his sleeping disciples and walks away without wakening them:44 And … he began to be sorrowful and troubled … And … he fell on his face … Again a second time … he prayed … And … again he found them sleeping, for their eyes were heavy … And … he prayed for the third time …

                                                            42 It is not unusual for Matthew to have participles where Mark has finite forms. See, for example, v. 

39, which has προσευχόμενος (προσηύχετο in Mk 14:35), and λέγων (ἔλεγεν in Mk 14:36). 

43 Gundry, Matthew, p. 531. 

44 Some might suggest that the imperfect form in v. 43 echoes the imperfective aspect of the present 

tense‐forms used in Jesus’ other interactions with the disciples. However, the fact that seven of the 

ten γάρ clauses in Matthew’s narrative framework (4:18; 7:29; 9:21; 14:3, 4, 24; 19:22; 26:43; 27:18; 

28:2) occur with imperfect tense verbs (five of them with εἰμί, as here) precludes the argument that 

Matthew has marked this clause by his choice of tense‐form. Porter warns that because εἰμί and some other μι verbs are ‘aspectually vague’, lacking forms which distinguish between imperfect and 

aorist tenses, ‘attempts to press the formal classification of, for example, ἤμην, etc., in terms of 

Aorist and/or Imperfect are misguided, and efforts to decipher the aspectual value of any form are 

also doomed to frustration’ (Verbal Aspect, p. 447). 

93  

The disciples’ behaviour, imperceptive and inept in the light of the unfolding events leading to Jesus’ crucifixion, is contrasted with Jesus’ lonely agony in anticipation of the ‘cup’ of God’s wrath he knows he will soon drink. The tension between the two is conveyed by the use of present tense-forms in Jesus’ interactions with his disciples, set against the series of aorist verbs describing Jesus’ movements alone in the garden. The choice of present tense-forms, the more heavily ‘marked’ tense in past-referring narrative, to describe Jesus’ interactions with his disciples suggests that it is the conflict between Jesus and the disciples that is being brought forward at this point in the discourse.45

Narrative critics such as Kingsbury note that conflict is a central element in Matthew’s plot, including conflict between Jesus and his disciples.46 During the passion events,

the disciples, although they want to stand by him (26:41), are totally unable to cope with events and end up by failing him … [The] disciples boldly express false confidence in their ability to remain loyal to Jesus: they all insist that if need be, they will die with him rather than deny him (26:30–35) … Peter and the two sons of Zebedee, though confident that they possess the fortitude to face death with Jesus, cannot even muster the strength to watch with him but instead fall asleep (26:36–46).47

Isolating the narrative framework from the rest of the passage allows us to see how Matthew displays this conflict in the structure of his narrative. As in the temptation pericope, Matthew integrates semantic contrast (Jesus’ interactions with the disciples set against his more isolated moments in the garden) with syntactic contrast (historic present forms alternated with aorist forms) to create a more compelling account of the dramatic tension inherent in these events.

And again, without observing how Matthew is utilizing the historic present at the discourse level, or at least above the level of individual sentences in this section of the discourse, it is difficult to explain why Matthew incorporates present tense-forms at particular points. For example, Allen merely comments on v. 36, ‘Mt., against his custom, retains the historic present’ (and similarly regarding v. 40), and on v. 37, ‘Mt. avoids, as usual, the historic present’.48 A discourse approach reveals the rationale behind Matthew’s choice of tense-forms in this passage.

Analysis of the narrative framework of Jesus’ prayer in Gethsemane (Mt. 26:36–46) suggests that the passage is one unit with two intertwining narrative threads. The first storyline, containing historic present verb forms, follows Jesus’ interactions with his disciples. The other storyline, recounted in aorist verb forms, focuses on Jesus in isolation—in his grief, in prayer to his Father, and as he walks away from his sleeping disciples. The two contrasting storylines reveal the dramatic conflict between Jesus and his unperceiving disciples.

3. Conclusion

I have presented two examples of Matthew’s use of the historic present in narrative discourse in his Gospel: the temptation of Jesus (4:1–11), and Jesus’ prayer in Gethsemane (26:36–46).

                                                            45 See Porter’s discussion of verbal opposition and planes of discourse, Idioms, pp. 22–23. 

46 J.D. Kingsbury, Matthew as Story (Philadelphia: Fortress Press, 2nd edn, 1988), p. 3. 

47 Kingsbury, Matthew as Story, p. 16. 

48 W.C. Allen, A Critical and Exegetical Commentary on the Gospel According to S. Matthew 

(Edinburgh: T. & T. Clark, 3rd edn, 1912), pp. 278–79. 

94  

The rarity of the historic present in narrative outside of speech margins indicates that where it occurs Matthew has made a conscious choice to use it. The high concentration of the historic present in both speech margins and non-speech margins at these two dramatic points in the Gospel suggests that it has discourse-structuring functions. An analysis of Matthew’s use of the historic present in units above the level of individual sentences uncovers the rationale behind his seemingly arbitrary selection of tense-forms in these two passages.

In 4:1–11 and 26:36–46 Matthew uses the historic present in a manner best explained by an aspectual approach to verbal tense-forms. He intentionally juxtaposes present and aorist or imperfect tense-forms within these passages for dramatic effect. This helps create the narrative’s structure and makes his storytelling more engaging to his readers: in one passage, to indicate development to a climax (4:1–11), and in the other, to distinguish between two interwoven storylines (26:36–46).

In each of these passages Matthew integrates semantic content with syntactical structures to develop his narrative. The syntactical forms he includes are not in themselves remarkable. The historic present in narrative and especially in speech margins is well attested in classical Greek and increasingly in the post-classical period. But the thoughtful use of the historic present in imaginatively structured episodes and in dramatic juxtapositions with other tense-forms creates a powerful narrative that engages his audience’s attention, drawing them into its unfolding plot. For those who undertake the study of Matthew as story, a fuller understanding of such discourse-structuring techniques can lead to a greater appreciation of Matthew as storyteller.

95  

A TOP-DOWN, TEMPLATE-DRIVEN NARRATIVE ANALYSIS, ILLUSTRATED BY APPLICATION TO

MARK’S GOSPEL Robert E. Longacre

1. Theoretical Foundations

Narrative analysis is necessarily dependent on the basic characteristics of narrative. What makes a story a story? A story is not an essay or a sermon, or a food recipe, or a set of procedures. It has a storyline, that is, a succession of happenings which are recounted. It involves a certain departure from the routine and expected. But nevertheless, it takes the stuff of life, sometimes with great detail and apparent arbitrariness, and weaves it into what Ricoeur calls an emplotment, or plot1 for short. It necessarily has participants involved in some sort of struggle, however refined or crude. It has to bring such struggle to a head and resolve it someway, even if the resolution is not a happy one.

Obviously, there is some sort of narrative template according to which stories are made. Since classical times (beginning with Aristotle’s writing on drama) such a template has been recognized, although various writers have expressed it differently. The schema I have held to for some time now2 (but cf. Labov and Waletzky,3 Rumelhart,4 and others) has the following elements: (1) Stage, (2) Inciting Incident, (3) Mounting Tension, (4) Climax, (5) Denouement, (6) Closure.

The purpose of (1) Stage is to lay the foundation for creating the storyworld, time, place, circumstances and participants (not necessarily the ones dominating the ensuing story). Element (2) brings in that which is unexpected and routine-breaking, so that ‘thereby hangs a t ale’. Element (3) typically involves a series of episodes which complicate the plot. Elements (4) Climax and (5) Denouement are further episodes which are somewhat correlative. The French terms nouement ‘tying it up’ and denouement ‘untying it’ capture this reciprocity quite well. The last element, Closure, brings the curtain down. It may, however, be preceded by one or more episodes of Lessening Tension or even Final Suspense which are consequent on the preceding denouement. The latter reports an event which makes resolution possible; it may leave many details to be worked out.

But while this is the underlying template, the actual story which is produced is like a theme with variations which it is the privilege of narrators to develop to their liking. Narrators elaborate the episodic surface structure of the story. They cannot tell all there is to tell at every stage of their narrative or the story would be both infinitely long and infinitely tiresome.                                                             1 P. Ricoeur, Time and Narrative (3 vols.; trans. Kathleen McLaughlin and David Pellauer; Chicago: 

University of Chicago Press, 1984–88). The material of special interest to us here is vol. 1, Chapter 3. 

2 Cf. R.E. Longacre, An Anatomy of Speech Notions (Lisse: Peter de Ridder, 1976), and idem, The 

Grammar of Discourse (New York: Plenum, 2nd edn, 1996). The particular form in which I have 

expressed the narrative schema is taken from W. Thrall, A. Hibbard and H. Holman, A Handbook to 

Literature (New York: The Odyssey Press, 1961). 

3 W. Labov and J. Waletzky, ‘Narrative Analysis: Oral Versions of Personal Experience’, in J. Helm (ed.), 

Essays on the Verbal and Visual Arts (Seattle: University of Washington Press, 1967), pp. 12–44. 

4 D.E. Rumelhart, ‘Notes on a Schema for Stories’, in D.G. Bobrow and A.M. Collins (eds.), 

Representation and Understanding (New York: Academic Press, 1975), pp. 211–36; idem, 

‘Understanding and Summarizing Brief Stories’, in D. LaBerge and S.J. Samuels (eds.), Basic Processes 

in Reading: Perception and Comprehension (Hillsdale, NJ: Erlbaum, 1977), pp. 265–304. 

96  

They must be highly selective. They will, in fact, have portions of their story that they present summarily and other parts concerning which they give considerable detail. The latter are the great moments of the story, which I will call action peaks5 in this essay. These great moments typically involve such sections of the underlying template as the Inciting Incident, the Climax and the Denouement. Narrators may also prepose a Title and/or Aperture to their story and may postpose a Finis. Aperture and Finis are characteristically formulaic and culture-determined.

A story of any length will typically have plots and subplots, that is, it may have story-within-story. Or narrators may find it necessary, if they are to guide the hearer/reader skillfully through the tale, to group the happenings in wavelike successive units. In either case the template is applied recursively.6 It is quite improbable to find a story of any length that is a simple linear string of happenings reported in successive episodes. Furthermore, narrators may group certain episodes of their story so as to suggest a special parallelism of one account with another account which consists of similar incidents. Narrators may even (especially if from a Semitic background) use a chiastic arrangement which suggests that there is a central episode which is pivotal. At all events the narrator is likely to echo at beginning and end—and maybe also in the center—certain themes.

The above brings us to the consideration that a story has not simply happenings and participants but themes. The latter can surface in various ways in a story: in background material which is not on the storyline, in the simple recurrence of certain ‘key’ words, and in reported speech. The latter is so important that a story may contain a Didactic Peak,7 that is, an episode in which action ceases and themes are developed via monologue and/or dialogue. Thus every novel or story of Ayn Rand8 contains a sermon, by its chief participant, on the virtue of self-reliance and the perniciousness of all forms of altruism and collectivism. In the book of Genesis the Flood Narrative9 contains besides the action peak in ch. 7, a didactic peak in ch. 9 whose themes are covenant and promise. In the Joseph story10 with which the book of Genesis terminates, Joseph’s call to prepare himself for an audience with Pharaoh and his installation as grand vizier after the interview are elements of an action peak that bracket a didactic peak in which Joseph interprets Pharaoh’s dreams and develops as theme the providence of God (total passage, Gen. 41:14–45). Thus, a didactic peak may occur in a distinct section of narrative from the action peak or it may occur in conjunction with the latter.

Analytically all this implies that a narrative analyst can proceed somewhat as follows in the preliminary approach to a text.11

                                                            5 Longacre, Grammar, Chapter 1; idem, ‘Discourse Peak as Zone of Turblence’, in J.R. Wirthe (ed.), 

Beyond the Sentence (Ann Arbor: Karoma Publishers, 1985), pp. 81–98. 

6 Longacre, Grammar, Chapter 7. 

7 Longacre, Grammar, Chapter 1. 

8 For example, Atlas Shrugged (New York: The New American Library, 1943) and Fountainhead (New 

York: The New American Library, 1957). 

9 R.E. Longacre, ‘The Discourse Structure of the Flood Narrative’, JAAR 47.1 (1979), pp. 89–133. 

10 R.E. Longacre, Joseph, a Story of Divine Providence (Winona Lake, IN: Eisenbrauns, 1989). 

11 This top‐down, template‐driven analysis amounts to a beginning sketch; to be more adequate the 

analysis needs to be extended downward to include relations within the paragraph in which 

sentences and groups of sentences are related according to what I term ‘interclausal relations’ 

97  

1. Search for natural fissures, joints or seams, in a narrative. This search requires both an intuitive following of ‘hunches’ and a sensitivity to formal marking by occurrence of conjunctions and other sequence signals, by cyclic recurrence of staging (time and locational expressions, change of participant slate, or radically changed circumstances), and by markers that an action sequence is slowing down or terminating.

2. Try to match underlying template segments with surface segments. In general the latter

will be episodes of the story. A valid match between a template segment and a surface segment establishes the function of the surface segment.

3. Apply the template recursively as much as necessary so as to obtain a coherent picture

of what is going on semantically and structurally. There is no point in trying to establish the moons of Jupiter in solar orbits, nor in viewing the hand as a direct appendage of the body instead of as an appendage of the arm which is attached to the body. Hierarchical structuring12 must be recognized in astronomy, biology and in textlinguistics. We can expect, therefore, to find within a story embedded discourses up to several layers—as many layers as seem necessary to obtain a plausible grouping.

4. Watch for peak-marking whether in respect to actions or themes. Action peaks are

marked by a variety of means13 which I will summarize here as (a) augmentation of the storyline (verb forms, sentence lengthening or shortening, crowding the storyline with a rapid sequence of happenings, with minute components of actions, or even with paraphrase of actions as pseudo-happenings); (b) immediacy (detail and dialogue); and (c) maximum interlacing of participants, that is, the ‘crowded stage’ effect.14 As mentioned above we can expect to find peak-marking applied to surface units which correspond to the Inciting Incident, the Climax or the Denouement. Didactic peaks are large segments of reported speech in which thematic material is developed.

5. Watch for parallelism and chiasm in the development of the story. The former can

determine a compound discourse, where successive events in two sections of a story are developed in a parallel fashion not generally characteristic of the rest of the story. While detailed use of chiasm is often best analyzed as an overlay over the successively episodic

                                                                                                                                                                                          (Grammar, Chapters 3 and 4); or according to the system of ‘rhetorical relations’ formulated by W. 

Mann and S. Thompson, Relational Propositions in Discourse (Technical Report ISI/RR‐83‐115; Marina 

del Rey, CA: Information Sciences Institute, 1983); and by the same two authors, Rhetorical Structure 

Theory: A Theory of Text Organization (Technical Report ISI/RS‐87‐190; Marina del Rey, CA: 

Information Sciences Institute, 1987); ‘Rhetorical Structure Theory: A Framework for Analysis of 

Texts’ (International Pragmatics Association Papers in Pragmatics 1.79–105). For a narrative analysis 

that is both top‐down and template‐driven and also carried down through interclausal relations 

within component paragraphs, see Longacre, Joseph. 

12 Longacre, Grammar, Chapter 9. 

13 Longacre, Grammar, Chapter 2. 

14 R.E. Longacre, Storyline Concerns and Word‐Order Typology in East and West Africa (Studies in 

African Linguistics Supplements, 10; Los Angeles: University of California, 1990), pp. 8–9. 

98  

structure,15 a looser chiasm can be employed so that one episode which cross-references both to the beginning and also to the end is developed as a Pivot in the episodic structure of the story. Such a pivot may take peak-marking.

2. Narrative Analysis of the Gospel According to Mark

It is assumed here that Mark’s Gospel can be analyzed as a narrative. The broad category ‘narrative’ embraces both fact and fiction. But even a story rooted and grounded in fact—as I believe this one to be—must be shaped according to narrative conventions if it is to be successful with its readers.16 We could, of course, argue that the Gospel is a special genre developed in order to present the mighty words and works of Jesus. But even the recognition of a specifically Gospel genre—with, for example, approximately a quarter of its bulk devoted to the last week in Jesus’ life—can hardly escape classification as narrative. At any rate, common to all four canonical Gospels is the plot that turns on the struggle between Jesus and the establishment of his day—a struggle that culminates in trial, crucifixion, death and resurrection. And the presence of plot is a diagnostic trait of narrative.

I have set about to develop a top-down, template-driven and functional analysis of Mark’s Gospel with application of the assumptions of the previous section to the analysis of the text. The editors (Aland et al.) present us with a Greek text of Mark neatly divided into 91 pericopes but it may be regarded as a foregone conclusion that the Gospel is not a simple linear sequence of these pericopes. Most assuredly there is a higher organization into bigger blocks. Most of the pericopes have, however, an introductory element that may be temporal, locative, circumstantial or participant-presentative. Very frequently motion verbs with given participants as subject introduce a pericope. While in my organization of the text of the

                                                            15 Thus, my template‐driven, linear‐recursive analysis of the Genesis Flood Narrative (‘Discourse 

Structure’) can be fruitfully compared with Wenham’s chiastic analysis of the same (G.J. Wenham, 

‘The Conference of the Flood Narrative’, VT 28 [1978], pp. 336–48). Likewise my analysis of Mark in 

this article may be compared with M.P. Scott’s masterful chiastic analysis in ‘Chiastic Structure: A Key 

to the Interpretation of Mark’s Gospel’, BTB, pp. 17–26. I do not feel that a linear‐recursive analysis 

and a chiastic analysis are ultimately incompatible; rather, they reflect different modes of linguistic 

structuring. The linear‐recursive analysis is an extrapolation from the grammar of the clause and the 

sentence. Chiastic analysis addresses itself to the content and colligational characteristics of a text, 

what the Pikes call ‘referential structure’. Cf. K. Pike and E. Pike, Grammatical Analysis (Summer 

Institute of Linguistics Publications in Linguistics, 53; Dallas: Summer Institute of Linguistics and the 

University of Texas at Arlington, 1977). I believe, therefore, that the type of analysis that is presented 

here and chiastic analyses such as those proposed by Wenham and Scott are complementary rather 

than contradictory. Ethel Wallis of the Summer Institute of Linguistics also has published a chiastic 

analysis of Mark: ‘Mark’s Goal‐Oriented Plot Structure’, JOTT 10 (1995), pp. 30–45. One matter of 

considerable interest is that even my linear‐recursive analysis presented here by setting up EPISODE 

3 as pivotal makes the Transfiguration account central and in this respects agrees with Scott’s taking 

Mk 9:7 as central to his chiastic structuring. We have, in effect, a two‐map problem with a central 

piece of topography common to the two maps and thus facilitating the relating of the two. 

16 Ricoeur’s discussion (Time, III) of the mutual relationships of fact and fiction within the overall 

narrative genre is a good contribution to this contemporary issue. Cf. also R.E. Longacre, ‘Paul 

Ricoeur’s Philosophy and Textlinguistic Analysis’, in The Nineteenth LACUS Forum 1992 (Lake Bluff, IL: 

The Linguistic Association of Canada and the United States, 1993), pp. 47–55. 

99  

Gospel into a hierarchy of units I do not always mention the introductory elements of a pericope in the discussion below, they are carefully catalogued in the display that constitutes an appendix to this paper. I do not make much use of the term ‘pericope’ in this analysis; rather I recognize them as episodes, which are typically on the lowest level of discourse embedding.

In that the Gospels present both the works and words of Jesus, I frequently posit didactic peaks on various levels of organization in the text. Outstanding works (miracles) are frequently found marked as peaks in various ways; such a marked account of a miracle I posit as climactic on various levels of structure. While the Gospel of Mark as a whole has a climax with a matching denouement, most of the embedded discourses end in a climax without a corresponding denouement. Typically an embedded discourse ends with an outstanding work which is reported in vividness and detail; each such climax further complicates the overall plot and makes increasingly urgent the thematic question ‘Who is he?’

The use of the historical present in this Gospel has long been considered to be a prime analytical concern.17 No amount of local contextual explanation in terms, for example, of continuity and discontinuity, can answer the simple question as to why it is used at all and especially why clusters of historical presents characterize the text at certain points. In keeping with my assumptions regarding peak-marking, according to which tampering with the tense of the verb is a specially sensitive area in many languages, I take the historical present, when occurring other than pericope-initial, when not limited to speech verbs, and when clustering within a passage, to be a peak marker. This has sometimes turned my analysis in unexpected directions. For example, I tried at first to handle the episode of Christ’s prayer in the garden of Gethsemane as simply another episode in the progress towards the cross, but the proliferation of historical presents in the passage, Mk 14:32–42, forced me to change my analysis at this point. Ralph Enos’s suggestive article is of special relevance in this regard; he, however, discusses isolated uses of the historical present which I do not consider here.

A suggested gross segmentation of the Gospel follows immediately below with further discussion following and with a more detailed analysis displayed in the appendix to this essay. In both analyses, capital letters signify primary constituents, that is, constituents of the Gospel as a whole. Capitalization of only the first letter of a word indicates constituents on the first level of discourse embedding. Italics symbolize constituents on a lower level of discourse embedding.                                                             17 A considerable bibliography could be cited at this point. Stephen Levinsohn’s work of recent years 

attempts to go at the matter of the historical present in the Gospels and Acts largely in terms of local 

cohesion, in what I would characterize as a ‘bottom‐up’ approach in contrast to my ‘top‐down’ 

analysis as illustrated in this article. Cf. S.H. Levinsohn, ‘The Historic Present and Speech Margins in 

Matthew’, in S. Hwang and W. Merrifield (eds.), Language in Context: Essays for Robert Longacre 

(Dallas: Summer Institute of Linguistics and the University of Texas at Arlington, 1992), pp. 451–74, 

and the same material presented as Chapter 10 in S.H. Levinsohn, Discourse Features of New 

Testament Greek: A Coursebook (Dallas: Summer Institute of Linguistics, 1992). I also want to cite 

here an excellent but little‐known work by R. Enos, ‘The Use of the Historical Present in the Gospel 

According to Saint Mark’, The Journal of the Linguistic Association of the Southwest 3.2 (1981), pp. 

281–98. Enos’s article is especially noteworthy for its extensive bibliography of work done on the 

historical present prior to the writing of his article, including some earlier work of Levinsohn. 

100  

TITLE/APERTURE of whole 1:1 STAGE: The ministry of John the Baptist 1:2–8 EPISODE 1 (INCITING INCIDENT): 1:9–13 (either a brief embedded discourse or a

compound paragraph) The Spirit comes on Jesus and confirms his Sonship; the heavens are ‘split’

1:9–11 The Spirit ‘drives’ him out to be tempted by the Devil 1:12–13 EPISODE 2 The rise to prominence 1:14–5:43 EPISODE 3 At full tide; Jesus a power figure and nurturer 6:1–8:26 EPISODE 4 (PIVOTAL): ‘Who is he?’ 8:27–9:50 EPISODE 5 The last journey to Jerusalem 10:1–52 EPISODE 6 (PEAK): 11:1–16:8 Episode 1 (Inciting Incident): The Triumphal Entry 11:1–11 Episode 2 (The DIDACTIC PEAK of the whole book): Teaching amid controversy

11:12–13:37 Episode 3 (The ACTION PEAK of the whole book): Events which culminate in the

crucifixion (CLIMAX) and resurrection (DENOUEMENT) 14:1–16:8 I will not comment on here nor attempt to justify the above segmentation but will reserve

such comment for the sections below which discuss EPISODES 1–6 in order. I do, however, mention briefly here the way in which I have analytically disposed of the first 13 verses of Mark’s Gospel. The ambiguity of v. 1 as to TITLE/APERTURE is discussed extensively by Cranfield.18 It structures plausibily either way. John the Baptist’s ministry as forerunner is cited by Mark as fulfillment of Mal. 3:1 and Isa. 40:3. This resort to Old Testament quotation is noteworthy in Mark since he is not given to as frequent a use of such quotations as is Matthew. The Baptist’s own words in vv. 7 and 8 underline the preparatory nature of his own ministry: ‘After me will come one more powerful than I …’ In terms of narrative structure the presentation of the Baptist and his ministry qualify as STAGE for all that follows. Nothing of the main story happens in its stage but the groundwork is laid for what follows. Later in his Gospel Mark felt obliged to recount the imprisonment and death of John the Baptist in the flashback account which is found in Mk 6:14–29 even though the account does not integrate too well into the ongoing story. a. EPISODE 1 (INCITING INCIDENT): 1:9–13 The passage 1:9–13 qualifies well as EPISODE 1, the INCITING INCIDENT of the whole Gospel. While the Greek text of Aland et al. makes vv. 12–13 a separate pericope, probably because of the parallelism with Mt. 4:1–11 and Lk. 4:1–12, Mark here abbreviates and makes vv. 12–13 one compound unit with vv. 9–11. The only transition particle is Mark’s omnipresent εὐθύς. There is, however, a new locale, the wilderness. The Holy Spirit does not enter the scene as a new participant but is carried over from v. 10. Satan is the new participant in the new locale. It seems simpler neither to compromise the ongoing unity of vv. 9–11 with vv. 12–13 nor to disregard their differences. Whether to consider the whole one compound paragraph or a short embedded discourse is somewhat irrelevant; the two seem to go together as the INCITING INCIDENT of the Gospel. As the Spirit descends on him the heavenly voice declares him to be the Son of God, thus preparing us as readers for the works and words                                                             18 C.E.B. Cranfield, The Gospel According to St Mark (Cambridge: Cambridge University Press, 1959), 

pp. 34–35. I take occasion here to commend Cranfield’s sensitivity to the flow and structure of 

Mark’s Gospel; he is not as slavishly verse‐by‐verse in his approach as are many traditional 

commentaries. 

101  

of power that follow. Furthermore, such an explicit word as to the identity of the man Jesus does not occur again until we hear it in the voice from the cloud Mk 9:7, and finally from Jesus’ own mouth at the time of the trial in 14:61–62—although this is implied in Peter’s confession in 8:29 as well. Indeed the passage before us sheds light on the whole central and theologically pivotal portion EPISODE 4 (8:27–9:50), which I have entitled ‘Who is he?’

Similarly, the temptation of Jesus prepares us for the heavy emphasis on exorcism which characterizes Mark’s Gospel as well as for the conflict with the establishment of his day. We are, as it were, taken behind the ensuing scenes along the lines indicated later by Paul: ‘For our struggle is not against flesh and blood, but against the rulers, against the authorities, against the powers of this dark world, and against the spiritual forces of evil in the heavenly realms’ (Eph. 6:12).

Therefore, in a fundamental sense 1:9–13 sets us up for all that follows and may plausibly be considered to have the characteristics of an inciting incident. b. EPISODE 2: The Rise to Prominence 1:14–5:43 It is plausible to consider that the balance of ch. 1 and the entire four chapters that follow constitute an embedded discourse with 1:14–15 patterning as Stage, ch. 4 as Didactic Peak and ch. 5 as Action Peak. The latter recounts in great detail three miracles of Jesus, the exorcism of the demons in the Gadarene demoniac and the raising of Jairus’s daughter, interrupted by the healing of the woman with a long-term disturbance of her menstrual cycle. The last is itself unique, no other recorded miracle has such a story interrupting and bracketed by another story. Moreover, the argument that ch. 5 is to be regarded as the Action Peak of the first part of the book is reinforced by the consideration that special peak-marking features are present as well. The second half of the chapter, the miracle-within-miracle, is well marked by special verb forms. To begin with, the case history of the afflicted woman is given in a long string of participles culminating in the verb ἥψατο, ‘she touched’ (v. 27). This is an unusually long chain of participles preceding the finite verb and is reminiscent of medial-final chaining in languages of Papua New Guinea.19 Then there is a clustering of historical presents in the bracketing story, the raising of Jairus’s daughter. It is noteworthy that Luke Johnson in his recent commentary on Luke20 also considers that Luke’s recounting of these miracles constitutes the end of a major section of Luke. With the performance of this exorcism, the healing, and a raising of one from the dead this part of Mark’s Gospel comes to a climax and is given appropriate surface development as such.

Chapter 4, concerning Jesus’ teaching in parables and giving a sample ensemble of the same, can be considered to be an embedded discourse which constitutes the didactic peak of this second main section of the Gospel. The internal structure of this one-chapter embedded discourse ends with an action episode, the calming of the storm (4:35–41). It is closely spliced onto the teaching that precedes by the connective expression in v. 36 ‘And sending away the crowds they took him as he was in the boat [where he had been seated teaching]’.

Episodes one to three of the embedded discourse which manifests main EPISODE 2 likewise pattern as embedded discourses whose constituents are on a still lower level of embedding (represented with italic letters). Thus, within EPISODE 2, Episode 1 (1:16–45) is the story of Jesus’ initial ministry in and around Capernaum; it is an embedded narrative with

                                                            19 Among the various works that could be cited here I take my 1972 monograph as an introduction to 

languages of this type: Robert E. Longacre, Hierarchy and Universality of Discourse Constitutents in 

New Guinea Languages (Washington, DC: Georgetown University Press, 1972). 

20 L.T. Johnson, The Gospel According to Luke (Collegeville, MN: Liturgical Press, 1993). Johnson 

makes tracing the discourse‐structure and flow one of the main concerns in his commentary. 

102  

five episodes, the last of which may be its action peak. These episodes recount, presumably in chronological order, the calling of the four fishermen (1:16–20), an exorcism in the synagogue (1:21–28), healings and exorcisms at the house of Peter (1:29–34), a preaching tour which begins the next morning (1:35–39) and, finally, his healing of a leper which provoked such a large following of people that he could no longer openly enter a town (1:40–45). The latter is certainly climactic but may perhaps lack the specific marking which we expect to find in a peak. It is not implausible to suggest that this embedded narrative is of Petrine origin.21

Episode 2 of the embedded discourse which constitutes EPISODE 2 of the main story is considered by some to reflect a topical arrangement of stories22 which are concerned with the beginnings of Jesus’ conflict with the authorities of his day. This, of course, does not deny the possibility that whatever their principle of selection they could have occurred substantially in the order in which they are introduced. The Episode is set off locationally by indicating a return to Capernaum and temporally by διʼ ἡμερῶν, ‘after some days’. The lower level episodes of this embedded discourse which manifests Episode 2 are five of which the first may be the inciting incident and the last is at least the climax and may have some peak marking: Episode 1: Jesus heals a paralytic and presumes (?) to forgive his sins (2:1–12 with several historical presents serving as peak marking in an inciting incident); Episode 2: Jesus encounters criticism by calling Levi and attending a banquet with him and his friends (2:13–17); Episode 3: (unfriendly) questions about fasting (2:18–22); Episode 4: the disciples are criticized for plucking and eating grain on the Sabbath (2:23–27); Episode 5: Jesus heals a man with a withered hand on the Sabbath and there is first mention of a plot to kill him (3:1–6, plainly climactic; the fact that all speech verbs are in the historical present may be peak marking in this instance). c. EPISODE 3: Jesus a Power Figure and Nurturer 6:1–8:26 This major episode, which pictures Jesus at full tide, has a compound structure with two parts each expounded by an embedded discourse. In each of the parallel discourses which constitute the compound structure, there is a feeding miracle. Furthermore, each embedded discourse ends with miracles of healing which are told in typically Markan (Petrine?) detail accompanied by scintillating dialogue and the use of the historical present for non-speech verbs. These similarly peak-marked endings of the parallel discourses make it awkward to attempt to account for all of 6:1–8:26 as one linear string of episodes on the same level. The first of the two embedded discourses also embeds some lower-level discourses in two of its episodes.

Part 1 (6:1–7:37) has four episodes, a didactic peak and an action peak. The first three Episodes are somewhat disparate but probably belong here to a new main EPISODE after the clearly marked and brilliant climax that constitutes ch. 5. Episodes 1 and 2 are the rejection at Nazareth (‘his own country’) 6:1–6 and the sending out of the Twelve 6:7–13. Episode 3 (6:14–29) reports Herod’s conjecture that Jesus was John the Baptist come back from the dead; then gives in a flashback the death of the latter. It thus for the first time raises the question ‘Who is he?’ that is thematic in EPISODE 4. Episode 4 (6:30–56) of this embedded discourse appears to be an embedded discourse with three episodes; the whole could be entitled ‘Miracles performed back and forth across the lake’. Two non-healing miracles are

                                                            21 So Cranfield, Mark, p. 61. In respect to the first four episodes he remarks that they are ‘a closely 

articulated group of four narratives of Petrine origin’. What I consider the climax of this section he 

considers simply ‘a link to what follows’ (p. 90). 

22 Cranfield, Mark, p. 61. 

103  

reported here; both emphasize Jesus’ power over nature. In Episode 1 (6:30–44), the disciples return from their mission and Jesus feeds the 5000. In Episode 2 (6:45–52), Jesus walks on the water. These supremacy-over-nature miracles are told with great clarity and vividness, but do not otherwise have peak marking nor are they positioned in their discourses where we might expect a peak to be positioned. Episode 3 (6:53–56), compared to what precedes, is more of a summary statement of healings in and around Gennesaret. Mark’s Gospel is not characterized by large teaching blocks as is Matthew’s, but 7:1–23 is a large section of teaching by Markan standards. I consider this passage to be Episode 5, didactic peak of the whole embedded discourse which constitutes Part 1 of EPISODE 3. This stretch of teaching deals with Jesus’ impatience with criticism of his disciples having eaten with unwashed hands and asserts that what comes out of a person’s inner being defiles him, not what descends into one’s stomach. Perhaps this is not inappropriately located here by Mark in a section where Jesus is pictured as one providing food.

I find it plausible in regard to 7:24–37 to believe that the two pericopes which are found here are meant to be taken together in that they have cross-referencing settings in 6:24, which refers to Jesus’ going away to the environs of Tyre, and in 6:31, which refers to his leaving the environs of Tyre, coming through the environs of Sidon, and eventually proceeding towards the Sea of Galilee through the Decapolis. On this ballistic movement out of the land of Israel and returning around its northern fringe, two miracles of healing are performed: Episode 1, the healing of the Syrophoenician woman’s daughter, and Episode 2, the healing of the deaf and dumb man. Both miracles are recounted in vivid detail with reported dialogue; in addition the account of the second miracle features the verbs ‘bring’ and ‘beseech’ in the historical present (v. 32) and reports the popular evaluation ‘He’s done all things well; he makes the deaf to hear and the dumb to speak’. I therefore label this whole passage 7:24–37 as action peak of the discourse which constitutes Part 1 of EPISODE 3. As we see from here and the conclusion of Part 2 of this compound discourse—as well as from the placement of ch. 5, at the end of EPISODE 2—a typically Markan way of bringing a discourse to a close is by giving its last pericope special peak marking, reserving the historical present of non-speech verbs for the second of two such accounts if two are found in the overall unit.

Part 2 of the compound unit that manifests EPISODE 3 has a somewhat simpler structure but is in certain ways parallel to the discourse found in Part 1. Its Episode 1 (8:1–9) recounts the feeding of the 4000; Episode 2 (8:10–12) depicts the Pharisees demanding a sign and getting Jesus’ enigmatic answer; Episode 3 (8:13–21) records the disciples’ confusion on receiving Jesus’ warning against the ‘leaven’ of the Pharisees and the Herodians; again the food motif occurs reinforced with Jesus’ explicit back reference to the two feeding miracles (vv. 19–20). Finally, Episode 4 (8:22–26) is an action peak in many ways marked like the action peak of the preceding part of this compound discourse: both in respect to the detail, the dialogue, and the use of the historical presents of the verbs ‘bring’ and ‘beseech’. Only in this account, among all of Mark’s miracle accounts, do we find a man reported as being healed in two stages! d. EPISODE 4 (PIVOTAL): 8:27–9:50 While there is no reason to believe that this block as a whole is not consecutive upon what precedes and anterior to what follows, it has certain unique properties which lead me to characterize it as in some sense central and pivotal in the narrative. The theme of the whole major EPISODE 4 seems to be ‘Who is he?’ first raised in effect by Herod in EPISODE 3, Part 1, Episode 1 (6:14). But here the matter of Jesus’ identity is peculiarly foregrounded. The answer given is twofold: (1) he is the Messiah, the Christ; and (2) he is to be a suffering Messiah. In Episode 1 (8:27–9:1) of the embedded discourse found here, Peter understood the first point but resisted the second and was summarily rebuked by his Lord, who not only identified himself with suffering but extended ‘cross-bearing’ to be the lot of all his true

104  

disciples. Christ’s rebuke of Peter is very severe. In addressing Peter as ‘Satan’ it is as if Jesus is recalling the temptation in the wilderness—although Mark gives none of the cross-avoiding nature of the temptations as do the parallels in Matthew and Luke.

But if we are in doubt as to whether there is an allusion here to the INCITING INCIDENT in respect to both the Baptism and the Temptation, the central episode of this embedded discourse, that is, Episode 2 (9:2–13), the transfiguration and the immediately following incident, removes this doubt. In the Transfiguration the voice from the cloud ‘This is my beloved Son’ echoes the baptismal voice ‘You are my beloved Son’—and anticipates the trial scene where Jesus is forced to confess his identity as ‘The Christ, the Son of the Blessed One’ (14:61). Furthermore, just as the Baptism is followed by the Temptation in the INCITING INCIDENT, so the transfiguration is followed by a conflict with Satan in the exorcism of the evil spirit from the child in 9:14–29—which is represented as a difficult exorcism which the disciples could not perform in spite of their having been empowered to exorcise demons in 3:15.

Two further Episodes conclude this major EPISODE 4. In Episode 3 (9:30–32) of this embedded discourse, Jesus foretells his death and resurrection a second time, and Episode 4 (9:33–50) records further teaching arising out of the quarrel as to who would be the greatest. Episode 3 further reinforces the point that Jesus as Messiah will prove to be a suffering Messiah. Episode 4, probably a hortatory discourse with three points, may well qualify as a didactic peak. Only the motion verb and the new locality mentioned in 9:33 keep us from joining the two Episodes. But Episode 3 is relatively unelaborated and could perhaps be grouped with what follows. If we were to make such a grouping then the whole embedded discourse which constitutes EPISODE 4 could be reduced to three episodes with the transfiguration and its aftermath as central, that is, pivotal in the pivotal section of the Gospel. The didactic material at the end of ch. 9, brings in a new motif, speaking or acting ‘in my name’ (vv. 37–41) which is at this juncture quite congruent with the identification of who Jesus is as developed in the previous passages. Granted who he is, the Messiah, the suffering Messiah, and the beloved Son of God, then words and deeds performed ‘in his name’ are appropriate.

Taking EPISODE 4 as pivotal to the whole Gospel, we see that Mark has constructed it well with Peter’s confession and the Transfiguration as the two main pieces but with congruent material combined with them. There is a backward look at the INCITING INCIDENT and an insistent forward look towards the momentous events, suffering, death and resurrection, the recounting of which constitutes the ACTION PEAK of the Gospel. Verse counting is revealing—in spite of the lateness and occasional arbitrariness of verse division. I count 316 verses preceding EPISODE 3, PIVOTAL, and 312 verses following it! e. EPISODE 5: The Last Journey to Jerusalem 10:1–52 I take 10:1a to signify the beginning of the last journey: ‘Jesus then again left that place and went into the region of Judea and across the Jordan’. Cranfield23 chooses to consider that the last journey begins in 8:27 in the neighborhood of Caesarea Philippi but I do not see how the detour over to the Mount of the Transfiguration fits this too well. At all events, making 8:27–9:50 a pivotal portion distinct from what precedes and what follows makes it more convenient to believe that the fateful journey begins with 10:1. Note also 10:17, ‘As Jesus started on his way’, and the somewhat more explicit reference to their being on their way to Jerusalem in 10:32, and, of course, the arrival in Jericho in 10:46. The travelogue discourse which constitutes ch. 10 has a clear Stage in 10:1, three Episodes, a further Episode 4 which is a complex Didactic Peak, and Episode 5, which is an Action Peak. The latter, exploiting a

                                                            23 Cranfield, Mark, p. 266. 

105  

device previously noted in Mark, consists of a healing miracle told with vivid detail and dialogue and with a historical present in its interior.

In this embedded discourse, Episode 1 (10:2–12) recounts a question concerning divorce and Jesus’ answer; Episode 2 (10:13–16) recounts his blessing the children; and Episode 3 (10:17–31) gives the story of the rich man who inquired concerning eternal life but found the price too high—along with ensuing teaching. Episode 4 patterns as Didactic Peak, 10:32–45; it is an embedded discourse with two subepisodes. Episode 1 (10:32–34) pictures Jesus resolutely leading the way towards Jerusalem while the disciples follow in fear. The first part of v. 32 may be the stage for what follows. Jesus teaches them in vv. 32b–34 even more explicitly concerning his betrayal, rejection and death at the hand of the Gentiles, and resurrection. Episode 2 (10:35–45) recounts the request of the sons of Zebedee that they be given pre-eminence in the coming glory along with Jesus’ solemn answer that they will indeed drink from his cup and be baptized with his baptism. All this could possibly be regarded as a distinct episode from what follows in vv. 41–45. Notice however, that in trying to calm the indignation of the other disciples against James and John, Jesus utters a saying fraught with deep meaning concerning his coming death: ‘For even the Son of Man did not come to be served, but to serve, and to give his life as a ransom for many’ (10:45). It seems plausible that this final reference by Jesus to his death forms with vv. 33–34 an inclusio which unites all 10:32–45 into one unit, namely Episode 4 of the embedded discouse which constitutes EPISODE 5 in the larger context. The contrast is exquisite: Jesus is preoccupied with his coming death, while the two disciples are preoccupied with their possible coming prerogatives and pre-eminence; even Jesus’ answer to them is couched in symbolic figures of grief and suffering!

Finally all the above is followed by a typically Markan action peak (10:46–52), the healing of blind Bartimaeus. Note the wealth of detail and the dialogue. Even the giving of the blind man’s name is of itself noteworthy. His yelling for help in an attempt to gain Jesus’ attention, his calling him ‘Son of David’, the attempts of the crowd to quiet him, his making all the more of a scene, and finally Jesus’ stopping in his tracks and asking for the man to be brought to him—all this is high drama. But more is to come: they call to the blind man (historical present); he throws aside his rags and springs to his feet to have his royal audience; Jesus makes him state his need and then heals him on the spot. So Bartimaeus becomes one of the many in the crowd following Jesus. All this is first-rate storytelling and is delivered by Mark in a style reserved for great moments in his Gospel. If the many vivid details here and in other such passages are from Mark’s having heard such incidents first hand from Peter—as many suggest—then we are bound to commend Mark’s placement of the incidents which are told in this vivid storytelling style; his placement of them so as to close out structural sections of his story, reflects the skill of a major craftsman.24 As already mentioned, each such vividly described miracle properly functions as a climax in terms of increasing the confusion and embarrassment of Jesus’ adversaries—and thus making inevitable the final CLIMAX and DENOUEMENT.

                                                            24 Cranfield (Mark, pp. 11–12) believes that ‘there are four different kinds of narrative material’: (1) 

narratives with vividness of detail which may be of Petrine origin, (2) narratives ‘which give the 

impression of being units of oral tradition which have been worn smooth by frequent repetition’, (3) 

narratives which although based on tradition were possibly constructed by Mark himself, and (4) 

‘brief summary statements indicating in general terms what was happening during a certain period’. I 

rather suggest that, whatever the variation in source, the narratives have been subtly shaped and 

adapted by Mark himself to fit the varying needs of particular contexts in which they are found. 

106  

f. EPISODE 6 (DIDACTIC PEAK and ACTION PEAK): 11:1–16:8 I assume that this extensive embedded discourse has three major Episodes, the first of which is Episode 1, the Triumphal Entry (11:1–11); the second of which is Episode 2, the DIDACTIC PEAK (11:12–13:37); and the third of which is Episode 3, the ACTION PEAK (14:1–16:8). An alternative analysis in which the two peaks would each be separate major structural EPISODES of the entire Gospel is precluded by the fact that the triumphal entry seems to be the Inciting Incident for all that follows while the embedded discourses which encode the two peaks each have their own inciting incidents. Nevertheless, while structural concerns of this sort must be given their proper weight the two peaks which are so introduced must be considered to be high points of the entire Gospel. Here structural and semantic concerns are somewhat askew, and it is best not to sacrifice one set of concerns to the other. I therefore accept the anomaly that the twin peaks of the entire Gospel are encoded as episodes of an embedded discourse. This necessitates a further layer of embedding than I have used earlier in the analysis. Thus, while I will continue to use EPISODE for a major section of the entire Gospel, Episode for the next lower level of embedding, and Episode for the layer below that, I will need to refer to a still deeper level of embedding as Episode. Episode 1, The Triumphal Entry into Jerusalem (11:1–11). This is clearly the inaugural event of Passion week. Furthermore, the account is marked as a great moment of the story by the use of the historical present not only at the opening in v. 1: ‘And when they draw near to Jerusalem … he sends two of his disciples’ (v. 2 ‘and says to them’); but also in the interior of the account in vv. 4 and 7: ‘And they loose it [the colt] … And they bring the colt to Jesus, and they throw their garments up onto it.’ It is plausible to take 11:1–11 as the inciting incident of all that follows in the Gospel from 11:12 to the conclusion. The triumphal entry provoked the debates of the last week and its salvific events. Episode 2, DIDACTIC PEAK (11:12–13:37). I have combined three pericopes into Episode 1, the Inciting Incident of the embedded discourse which constitutes Episode 2, in that they form an ABA chiastic sequence: the cursing of the fig tree Episode 1, the cleansing of the temple Episode 2, the withering of the fig tree Episode 3. The three pericopes are closely tied chronologically, with the cursing of the fig tree and its withering taking place on successive mornings, and the cleansing of the temple on arrival at Jerusalem the first of the two days. As a chiasm, presumably the center section dominates and may shed some light on the interpretation of the surrounding segments. Is the fig tree meant to symbolize the nation of Israel here? But, although Jesus ends this whole Episode 2 DIDACTIC PEAK with a prophecy of the desolation of Jerusalem and the nation, he never in any place curses the city and its people. Eliminating the curse from the metaphor as possibly not germane to the comparison, there may still be symbolism in the apparently flourishing but fruitless tree and its withering.25 In terms of the narrative template I take this to be an inciting incident on the grounds that the chiastic structure, even in absence of historical presents except in v. 15, amounts to a kind of peak-marking, that is, the Inciting Incident of the discourse which constitutes Episode 2. A similar chiastic unit (14:1–11) patterns as the Inciting Incident of Episode 3, the ACTION PEAK. It is the parallel use of these two chiastic structures to mark inciting incidents, as well as the strongly marked account: of the Triumphal Entry as also an inciting incident, that provides the clue that a complicated situation of discourse embedding is

                                                            25 For discussion concerning this incident and its possible symbolic value, see Cranfield, Mark, pp. 

254–57, and W.L. Lane, The Gospel According to Mark (Grand Rapids: Eerdmans, 1974), pp. 398–402. 

107  

present here. We have various intertwined structures that need to be untangled so as to permit the assignment of the proper inciting incident to its appropriate structure.

Episode 2, the DIDACTIC PEAK, once past Episode 1, its inciting incident, unrolls as a series of controversial exchanges between Jesus and his opponents and culminates in the Olivet Discourse which is an answer to an inquiry on the part of the disciples. In some of the controversial exchanges the opponents of Jesus take the initiative in propounding questions, in other exchanges Jesus himself takes the initiative. In the latter case, here and in the accompanying display in the appendix, I mark with an asterisk* Episodes in which Jesus takes the initiative. Perhaps the shifting inititatives indicate a further dialogue-like structure of proposal and response between certain of the episodes, but I have not at present recognized any such groupings. I proceed now to present what follows in Part 1 as Episodes 2–11 below.

Episode 2, 11:27–33: Jesus silences a question regarding his authority by countering with a question regarding what authority lay behind the ministry of John the Baptist.

*Episode 3, 12:1–12: Jesus, taking the initiative, gives the parable of the Tenant Farmers which his opponents recognize as having been spoken against them and incites their desire to arrest him (v. 12).

Episode 4, 12:13–17: Jesus answers the question about paying tribute to Caesar with, ‘Render unto Caesar the things that are Caesar’s and unto God the things that are God’s’.

Episode 5, 12:18–27:

Jesus answers a reductio ad absurdum argument of the Sadducees regarding the resurrection by teaching that there is no sex in the afterlife while nevertheless stoutly affirming the resurrection and the life to come.

Episode 6, 12:28–34: On being questioned regarding the greatest commandment Jesus reaffirms love to God and other humans and commends his interrogator.

*Episode 7, 12:35–37: Jesus, seizing the initiative, questions them as to how the Messiah can simply be called David’s son when the latter addresses him as Lord in Ps. 110:1.

*Episode 8, 12:38–40: Jesus denounces the scribes.

*Episode 9, 12:41–44: Jesus commends the widow’s offering.

Episode 10, 13:1–2: In response to a remark of the disciples, Jesus foretells the destruction of the Temple.

Episode 11, 13:3–36:

Didactic peak of the whole discourse which is DIDACTIC PEAK. This is a hortatory discourse (the Olivet Discourse) given in response to further inquiry from the disciples. It has at least three or four points and is not analyzed here.

                                                            * Jesus takes the initiative in verbal dueling. 

108  

Episode 3, ACTION PEAK, chs. 14, 15 and 16:1–8:

a narrative discourse pivoting around Episode 3, 14:32–42 (see below).

Episode 1, Inciting Incident 14:1–11:

This is a chiastic structure ABA with the plot to kill Jesus as Episode 1, 14:1–2; the anointing at Bethany as Episode 2, 14:3–9; and Judas’s agreeing to betray Jesus as Episode 3, 14:10–11. In this little embedded discourse Mary’s anointing of Jesus, told with great detail and pictured as provoking considerable discussion and criticism, is clearly the central piece. It stands as an example of reckless, abandoned worship, even though surrounded by plotting and betrayal. In commending her Jesus said, ‘She came beforehand to anoint my body for burial’—thus focusing even Mary’s act of devotion on his coming suffering and death. Not once does the narrator break step in the remorseless march towards the finale.

Episode 2, 14:12–31: The Last Passover. Here there are subepisodes marked by temporal expressions and verbs of motion; with the first two subepisodes, the introductory verbs of motion are historical presents.

Episode 1, 14:12–16: Preparations.

Episode 2, 7–21: Prophecy of betrayal.

Episode 3, 14:22–26: Institution of the Lord’s Supper.

Episode 4, 14:27–31: Peter’s denial foretold.

Episode 3, Pivotal 14:32–42: Jesus prays in Gethsemane. This episode cannot simply be treated as another in the ongoing string of successive episodes. It contains nine verbs in the historical present, three of which are speech verbs (and hence not very evidential of special marking) but six of which are motion verbs. These verbs occur scattered through the passage not merely at its head. We do not find a similar spate of historical presents until the portrayal of the mocking and crucifixion in Episode 8 below—where again nine examples of this tense occur. The final spate of historical presents is in Episode 11 below, the resurrection, where three such verbs occur. Taking Episode 8 below as climax marked for Action Peak, and Episode 11 as denouement marked for Peak, we raise the question as to the status of this episode, Episode 3 (the ordeal in Gethsemane).

What is happening narrative-wise is not hard to explain: the narrator is, as it were, gathering his feet under him for the final sprint. But it occurs late in the total stretch of the discourse expounding the ACTION PEAK for an inciting incident, and besides we have already assigned this function to 14:1–11. We simply have to view Mark’s narrative as marking, in addition to the Inciting Incident, three great moments in the Passion Narrative: Gethsemane, Calvary and the Empty Tomb. I label the first of the three Pivotal along with the Climax and Denouement, represented in the latter two. Mark pictures the sufferings of Christ as beginning in Gethsemane and properly underscores that point. The concept of pivotal has also been used in reference to the total structure of Mark’s Gospel in our setting up of main EPISODE 3 in this function. Episode 4, 14:43–52: The betrayal and arrest of Jesus. The verb tenses again merit comment here. Curiously enough, not only is there a verb in the historical present in the first verse of

109  

this passage, ‘And immediately while he was still talking, up comes Judas’, but also in its last vv. 51–52: ‘And a young man, wearing nothing but a linen garment was following along, and they seize him. And he fled away naked.’ While it is not unusual to have historical presents of verbs of motion and speech which initiate a section or which occur both initially and in the interior of a section, it is somewhat rare to find them closing out a passage. Two explanations are possible: (1) the verses in question are in reality not an addendum to what preceded but a separate section (as punctuated by Aland et al.), or (2) Mark writes of himself in these verses and the historical present is resultant on his own vivid personal recollection of the incident.26 It is also noteworthy that a perfect occurs in pluperfect sense in 14:44: ‘The one who betrayed him had given a sign (δεδώκει)’. Episode 5, 14:53–65: Jesus before the Council. Episode 6, 14:66–72: Peter’s denials. Episode 7, 15:1–15: Jesus is sentenced by Pontius Pilate. Episode 8 Peak, 15:16–32 (maybe two subepisodes): Jesus is mocked and crucified. Nine historical presents are found in this passage. No longer is the historical present simply limited to motion, transportation, and speech verbs. The verbs in the historical present here are συγκαλοῦσιν ‘they call together’ v. 16; ἐνδιδύσκουσιν ‘they clothe [him] in’ v. 17; περιτιθέασιν ‘they place upon’ v. 17; ἐξάγουσιν ‘they lead [him] out’ v. 20; ἀγγαρεύουσιν ‘they commandeer’ v. 21; φέρουσιν ‘they take away’ v. 22; σταυροῦσιν ‘they crucify’ v. 24; διαμερίζονται ‘they divide’ (in an Old Testament quotation) v. 24; σταυροῦσιν (two thieves) v. 27.

Thus, not only do we have an impressive number of historical presents but a more unrestricted domain of their occurrence. The crucial verb ‘crucify’ occurs twice in the historical present in the passage. Historical presents dominate the action part of the episode until the action merges into reporting of speech acts (mockeries and taunts) in vv. 29–32, where no more historical presents are found. Episode 9, 15:33–41: The death of Jesus. Episode 10, 15:42–47: The burial. Episode 11, 16:1–8: The resurrection. There are three historical presents in this passage: ἔρχονται ‘they come’ v. 2; θεωροῦσιν ‘they see’ v. 4; λέγει ‘he says’ 6. Again these verbs characterize three out of the eight verses in this brief account.

In regard to the almost universally felt problem regarding the abrupt ending of the Gospel according to Mark—a problem felt from the earliest centuries when at least two attempts were made to ‘finish’ the Gospel—I adopt the suggestion that Mark himself terminated the Gospel in this abrupt manner. Just as the Gospel starts abruptly so it by his intention ends abruptly with denouement as peak—without further wind-down or closure.27

                                                            26 Cranfield, Mark, pp. 438–39; and Lane, Mark, pp. 526–28. 

27 For a brief bibliography of older authorities (prior to 1955) that accepted the view here suggested, 

see Cranfield, Mark, p. 471—although he himself rejects this explanation. 

110  

3. Evaluation of the Exegetical Worth of Such an Analysis

What then? Is the analysis offered here simply another tedious example of overstructuring on the part of the analyst? In sidestepping this accusation I point out here some theological implications of the analysis. To begin with, however, let us recognize that our bracketing and labeling is no more a part of the authorial intent of the author Mark than phonological and grammatical analytic conventions figure with us in our immediate intuitive use of language. All such analytical devices—including those invoked in this essay—are attempts to make the message of a text explicit. But in invoking the narrative template as the starting point we appeal to something rooted in our cognitve structure. We know, for example, from how people react, that pointless stories, plots without resolution and ceaseless rounds of events without prominence or progress are not tolerated by listeners or readers. Furthermore, in invoking recursion, story-within-story or simply pause-for-station-identification breaks we tread again on ground provided by basic facts of language structure. From these cognitive and linguistic concerns the present analysis has come about.28

I point out here as of special interest several analytical results of considerable exegetical import: (1) The setting up of a compound discourse in EPISODE 2 elevates the two feeding miracles of Jesus above the nagging doubt that we have here simply a source-motivated doublet. Each feeding miracle is in a separate but parallel embedded discourse with startlingly similar closures in 7:31–37 and 8:22–26. By thus twice presenting Jesus as nourisher/provider a strong theological point is implicitly made (cf. Gen. 41:32) which another Gospel writer, John, explicitly develops in ch. 6 of his Gospel: Christ the bread of life. (2) Miracles of healing told in detail with sparkling dialogue mark climaxes in several parts of the ongoing work, namely ch. 5, 7:31–37 along with 8:22–26, and finally 10:46–52. Each such climactic display of Jesus’ supernatural power makes more acute and agonizing the challenge to the establishment of his day and makes more crucial the question ‘Who is he?’ (3) The putting together of 8:27–9:50 as central and pivotal to the whole work, connecting plausibly with the baptism and temptation in the Inciting Incident and with the trial scene at the end, foregrounds the question of the identity of Jesus: the claim is made that the historical Jesus is a supernatural figure, the Christ. All this is vindicated by the denouement of the Gospel, the resurrection. Even a linear/recursive analysis such as that here presented yields at this point to the presence of chiastic elements in Mark’s composition. (4) The recognition of didactic peaks in Mark is a way of showing how larger blocks of teaching are deployed characteristically before action peaks, as in ch. 4, the parables; 7:1–23, clean and unclean foods; 10:32–45, servanthood in the light of Jesus’ sufferings; and especially 11:12–13:37, the controversy-enveloped last teachings. (5) The manner in which Mark underscores the scene of Jesus’ praying in the garden constrains us to recognize that the grand action finale has three great events: Gethsemane, Calvary and the empty tomb—with profound theological implications. (6) The recognition of local chiastic structures in 11:12–26 and in 14:1–11 has further import. With the former, the cleansing of the Temple is surely inciting and provocative of much of the following controversy, but it is wrapped up in the enacted parable of the cursed and withering figtree. In the latter passage, while conspiracy and betrayal lead to all that follows, Mark gives the story a wonderful twist by putting in, as the key of the chiasmus, the anointing of Jesus at Bethany—implying that devotion to the wonderful figure revealed in his Gospel will in the end outweigh conspiracy and dark betrayal.                                                             28 I also note here the appearance of Mark Wegener’s book Cruciformed: The Literary Impact of 

Mark’s Story of Jesus and his Disciples (Lanham, MD: University Press of America, 1995). Wegener’s 

study and mine have in common a concern for wholistic analysis and discourse movement. His book 

is especially to be commended for its further concern with reader impact of the Gospel. 

111  

Appendix DISPLAY OF THE DISCOURSE STRUCTURE OF THE GOSPEL ACCORDING TO MARK

1. Basic Assumptions

A. Very probably Mark is not a simple linear sequence of pericopes—91 according to Aland et al.—such as we might obtain by attention to breaks based on all possible transition markers, e.g. time expressions (T), locative expressions (L), or introduction of new participants. Very probably a series of more inclusive blocks is present and these should have some relation to the typical narrative template. B. Inciting incidents and peaks, both didactic and action, are very probably involved—as well as certain episodes which may be called pivotal. C. Multiple occurrences of the historical present (HP), especially when not limited to verbs of speech or to verbs of speech and motion, may contribute to such marking as suggested in B above. Ralph Enos’s suggestion (1981), that pericope-initial HP’s may mark material of special theological importance, is of possible relevance. For this reason a pericope-initial HP is labelled below with the sign #. Typically these are motion verbs or verbs of speech. APERTURE of whole book 1:1 STAGE of whole book (ministry of John) 1:2–8

EPISODE 1 (INCITING INCIDENT) 1:9–13. A brief embedded discourse or a compound paragraph: The Holy Spirit comes on Jesus and confirms his sonship; the heavens are ‘split’ The Holy Spirit drives# him out to be tempted by the Devil EPISODE 2 The rise to prominence 1:14–5:43 Stage: Time, place, circumstances, Jesus and his message 1:14–15 Episode 1 An embedded discourse with five episodes, the last of which may constitute an action peak:

1:16–45. Episode 1 Calls four fishermen 1:16–20 Episode 2 (motion verb#, place name) Exorcism in the synagogue 1:21–28 Episode 3 Healings and exorcisms at the house of Peter 1:29–34 Episode 4 (T, motion verb) Preaching tour begins next morning 1:35–39 Episode 5 (action peak?) (motion verb#, participant) Heals a leper and great crowds follow so that he can

no longer openly enter a city 1:40–45 Episode 2 (Capernaum, T ‘after some days’) Tangling with critics (embedded discourse with five

episodes, the first of which is Inciting Incident and the last of which is Peak) 2:1–3:6: Episode 1 (Inciting Incident) (several HP’s) Jesus heals a paralytic and presumes (?) to forgive his sins.

(The latter is the big point according to Matthew who omits all the interesting details.) 2:1–12 Episode 2 (motion verbs, L, and πάλιν) Jesus calls Levi and eats with sinners 2:13–17 Episode 3 (motion verb#, speech verb#, participant switch) Questions about fasting 2:18–22 Episode 4 (motion verb, L) Plucking and eating grain on the Sabbath 2:23–27

                                                            HP Historical present 

# A pericope‐initial HP that may be a mark of special theological importance 

T Time expression 

L Locative expression 

112  

Episode 5 Action Peak (πάλιν, all speech verbs are HP) Jesus heals a man with a withered hand on the Sabbath; first mentioning of a plot to kill him 3:1–6

Episode 3 Increasing confrontation and misunderstanding (embedded discourse with four episodes): Episode 1 (many place names, Jesus, motion verbs) Attempted withdrawal, great crowds, teaches from a

boat) 3:7–12 Episode 2 (motion verb#, verb ‘call’#, L) Choosing of the Twelve 3:13–19 Episode 3 (2 motion verb#, L, and πάλιν) Great crowds; his family sets out to fetch him home; those who

ascribe his power to Beelzebub receive a solemn warning 3:20–30 Episode 4 (motion verb#, participant switch; all verbs of speech are HP) Arrival of his family; Jesus says

that his disciples are his true family—is this a peak? 3:31–35 Episode 4 didactic peak (motion verb#, L, and πάλιν) Parables and aftermath 4:1–41. Internal structure?

Two episodes, the second of which is an action peak which concludes a hortatory discourse: Episode 1 One long compound paragraph with clear aperture and closure—reinforced by a mid-paragraph

reference in vv. 11–12; the theme is teaching in parables 4:1–34 Episode 2 (action peak) This section is tightly connected to what precedes by ‘evening of same day’ and

‘they take# him as he was in the boat’ (L, internal HP’s). Jesus stills a storm on the lake; for the first time the question is raised ‘Who is he?’ 4:35–40

Episode 5 action peak of whole embedded discourse (i.e. of 1:14–5:43) Great miracles in great detail. All of ch. 5 is an embedded discourse with two episodes the second of which has more explicit peak-marking:

Episode 1 (L, motion verb) Healing of Gadarene demoniac 5:1–20 Episode 2 (L, motion verb, proper names Jesus and Jairus; story within a story with woman’s case history

given in string of participles, HP’s in the bracketing story) 5:21–43 EPISODE 3 compound narrative discourse with two somewhat parallel parts, Jesus at full tide as power

figure and nurturer 6:1–8:26 Part 1 embedded discourse with four episodes, didactic peak and action peak. Episode 1 (2 motion verbs#, L) Rejection at Nazareth 6:1–6 Episode 2 (‘calls’#, new participants) Jesus sends out the Twelve 6:7–13 Episode 3 (new participant, Herod raises question of Jesus’ identity) flashback records death of John the

Baptist 6:14–29 Episode 4 embedded discourse with three episodes 6:30–56 (back and forth across the lake): Episode 1 (motion verb#, participants, name Jesus) Disciples return and Jesus feeds the 5000 6:30–44 Episode 2 (motion verb, L) Jesus walks on water 6:45–52 (motion verb# in v. 48) Episode 3 (motion verb, L) Healing in and around Gennesaret 6:53–56 Episode 5 didactic peak (motion verb#, new participants) Disputation about the ‘clean’ and the ‘unclean’

7:1–23 Episode 6 action peak, Miracles performed on excursus through territory of Tyre and return; the two

subepisodes have cross-referencing settings, dialogue and detail: Episode 1 (motion verb, L, participant) The faith of Syrophoenician woman 7:24–30 Episode 2 (motion verb, L, participant, πάλιν) Healing of the deaf and dumb man 7:31–37 (‘bring’# and

‘beseech’# in v. 32; cf. 8:22–26 below) Part 2 embedded discourse with three episodes and action peak: Episode 1 (time, participants, πάλιν) Feeding of the 4000 8:1–9 Episode 2 (motion verb, L, participants) Pharisees demand a sign 8:10–12 Episode 3 (motion verb, πάλιν) The ‘leaven’ of the Pharisees and Herod 8:13–21 Episode 4 action peak (motion verb#, ‘bring’#, ‘beseech’# L, participant) Blind man healed in two stages!

(unique in miracles of Jesus) 8:22–26 EPISODE 4 (CENTRAL AND PIVOTAL) ‘Who is He?’ 8:27–9:50 (316 verses precede this passage; 312

verses follow), embedded discourse with four episodes: Episode 1 (L—on the road to Caesarea Philippi, name Jesus) Peter’s confession, Jesus predicts his death,

rebukes Peter, and teaches on cross-bearing 8:27–9:1 Episode 2 (time, L, ‘takes’#, motion verb#, names of Jesus and three disciples) Transfiguration and

aftermath, embedded discourse with two episodes:

113  

Episode 1 Transfiguration 9:2–13 (the voice from the cloud, cf. 1:11) Episode 2 (motion verb, L, participants) Jesus heals the boy with the evil spirit 9:14–29 (motion verb# in

v. 25) Episode 3 (motion verb, L) Jesus foretells his death a second time 9:30–32 Episode 4 (motion verb, L) Teaching arising out of the quarrel as to who would be the greatest 9:33–50

(new theme ‘in my name’ vv. 36, 38, and 39–41) EPISODE 5 The last journey to Jerusalem, embedded discourse with six episodes ch. 10: Stage (2 motion verb#, πάλιν [2×]) 10:1 Episode 1 (participant?—textual variation) Questions on divorce 10:2–12 Episode 2 (motion verb, participants) Jesus blesses the children 10:13–16 Episode 3 (motion verb, participant) The rich young ruler; resultant teaching 10:17–31 Episode 4 didactic peak, compound discourse with two subepisodes 10:32–45: Episode 1 (motion verb, L—in the road going up to Jerusalem, name Jesus) Jesus leads the way up to

Jerusalem and for third time predicts his death 10:32–34 Episode 2 (motion verb#, proper names) Request of James and John provokes further teaching on

servanthood, culminating with an insightful saying on the meaning of his death 10:35–45 Episode 5 action peak (motion verb#, participant) Healing of blind Bartimaeus. Story told with dramatic

detail and dialogue (‘call’# in v. 49) 10:46–52 EPISODE 6 DIDACTIC PEAK and ACTION PEAK 11:1–16:8: Episode 1 Inciting Incident (motion verb#, ‘send’#, ‘say’# L, T, three more HP’s occur internal) The

triumphal entry 11:1–11 Episode 2 DIDACTIC PEAK 11:12–13:37, In the verbal dueling below * marks where Jesus takes the

initiative: Episode 1 Inciting Incident, a chiastic structure ABA Episode 1 (motion verb, T, L) The cursing of the fig tree 11:12–14 Episode 2 (motion verb#, L) The cleansing of the Temple 11:15–19 Episode 3 (motion verb, T) The withering of the fig tree; lesson on faith 11:20–26 Episode 2 (2 motion verbs#, T, participants) The authority of Jesus is questioned 11:27–33 *Episode 3 Parable of the Tenants 12:1–12 Episode 4 (motion verb#, T, L, participants) Question regarding paying tribute to Caesar 12:13–17 Episode 5 (motion verb#, participants) Question regarding the resurrection 12:18–27 Episode 6 (motion verb, participants) Question regarding the greatest commandment 12:28–34 *Episode 7 (L) Jesus questions them about David’s son 12:35–37 *Episode 8 Jesus denounces the scribes 12:38–40 *Episode 9 (L) Jesus commends the widow’s offering 12:41–44 Episode 10 (motion verb, L, response to a remark of his disciples) Jesus foretells the destruction of the

temple 13:1–2 Episode 11, didactic peak of this embedded discourse (motion verb, L, answer to a question put by three

disciples) ‘Olivet discourse’ (hortatory, with at least three or four points; analysis not given here) Episode 3 ACTION PEAK, chs. 14, 15 and 16:1–8: Episode 1 (T) Inciting Incident in a chiastic structure ABA Episode 1 (T, participants) The plot to kill Jesus 14:1–2 Episode 2 (motion verb, L, participants) The anointing at Bethany 14:3–9 Episode 3 (motion verb, participants) Judas agrees to betray Jesus 14:10–11 Episode 2 The Last Passover 14:12–31 Episode 1 (‘send’#, long T) Preparations 14:12–16 Episode 2 (motion verb#, T) Prophecy of betrayal 14:17–21 Episode 3 Institution of the Lord’s Supper 14:22–26 Episode 4 (Proper name Jesus) Peter’s denial foretold 14:27–31 Episode 3 pivotal (motion verb#, L, eight HP’s) Jesus prays in Gethsemane 14:32–42

                                                            * Jesus takes the initiative in verbal dueling. 

114  

Episode 4 (motion verb#, participants-Judas) Betrayal and seizure 14:43–52 (HP in the isolated incident mentioned at the close in v. 52)

Episode 5 (motion verb#, proper name Jesus, participants) Jesus before the council 14:53–65 Episode 6 (motion verb#, participants) Peter’s denials 14:66–72 Episode 7 (motion verb, T, L, participants) Jesus is sentenced by Pontius Pilate 15:1–15 Episode 8 Action Peak (motion verbs, both initial and internal, new participants, nine HP’s) (maybe two

subepisodes) Jesus is mocked and crucified 15:16–32 Episode 9 (T marked twice, darkness, proper name Jesus) Jesus dies 15:33–41 Episode 10 (motion verb, T, participant) The burial 15:42–47 Episode 11 Peak (denouement) (motion verb, T marked twice, L, participants, three HP’s) The

resurrection 16:1–8

115  

MARK 5:1–43: GENERATING THE COMPLEXITY OF A NARRATIVE FROM ITS MOST BASIC

ELEMENTS Robert E. Longacre

1. Introduction

This essay is intended as a sequel to my overview of the Gospel of Mark.1 As I say in n. 11 to that essay, ‘This top-down, template-driven analysis amounts to a beginning sketch; to be more adequate the analysis needs to be extended downward to include relations within the paragraph in which sentences and groups of sentences are related according to what I term “interclausal relations” ’.2 But the analysis of such lower-level groupings turns on more than simply logical and rhetorical relations, as important as they are. The analysis necessarily must also take into account a hierarchy of information types, that is, the storyline and different sorts of relatively foregrounded and relatively backgrounded pieces of information, as reflected in the use of tense/aspect in the verbs of the component clauses.3 In this regard, the priority of the Greek aorist has long been recognized as crucial to Greek narrative. But other forms in addition to the aorist have their place in building a Greek narrative and it is the intent of this essay to attempt a preliminary systematization of these functions and their semantic reflexes. I will try to demonstrate here the intimate tie-in between constituency structure and the functions of various forms of the Greek verb in reference to Mark 5.

In section 2 of this essay I consider Mk 5:1–20, the account of the healing of the Gadarene demoniac, present a stepped tree diagram of this passage, and posit a saliency scheme for the various forms of the Greek verb that are used within it. In the course of doing this it is necessary to recognize a distinction between the function of the participle when preposed to the main clause and when postposed to the main clause, and to relate this distinction to the grammar of ‘chaining’ languages found around the world.

In section 3 of this essay I consider Mk 5:21–43, the account of the raising of Jairus’s daughter along with the interrupting account of the healing of the woman with the issue of

                                                            1 See the previous essay in this volume, ‘A Top‐Down, Template‐Driven Narrative Analysis, Illustrated 

by Application to Mark’s Gospel’. 

11 This top‐down, template‐driven analysis amounts to a beginning sketch; to be more adequate the 

analysis needs to be extended downward to include relations within the paragraph in which 

sentences and groups of sentences are related according to what I term ‘interclausal relations’ 

(Grammar, Chapters 3 and 4); or according to the system of ‘rhetorical relations’ formulated by W. 

Mann and S. Thompson, Relational Propositions in Discourse (Technical Report ISI/RR‐83‐115; Marina 

del Rey, CA: Information Sciences Institute, 1983); and by the same two authors, Rhetorical Structure 

Theory: A Theory of Text Organization (Technical Report ISI/RS‐87‐190; Marina del Rey, CA: 

Information Sciences Institute, 1987); ‘Rhetorical Structure Theory: A Framework for Analysis of 

Texts’ (International Pragmatics Association Papers in Pragmatics 1.79–105). For a narrative analysis 

that is both top‐down and template‐driven and also carried down through interclausal relations 

within component paragraphs, see Longacre, Joseph. 

2 See R.E. Longacre, The Grammar of Discourse (New York: Plenum Press, 2nd edn, 1996), Chapters 3 

and 4. 

3 R.E. Longacre, ‘Two Hypotheses Regarding Text Generation and Analysis’, Discourse Processes 12 

(1989), pp. 416–60. 

116  

blood. I give stepped diagrams of these accounts with prose commentary on the same. In analyzing this part of the chapter the understanding of the role of the historical present becomes especially crucial.4

In section 4 I demonstrate the feasibility of generating Mk 5:21–43 by obtaining an abstract of the story as given in the clauses that have aorist verbs, and progressively fleshing out the abstract of the story by adding in order the following: clauses with the historical present, then clauses with preposed participles, then finally clauses with verbs in the imperfect—thus building up the text in four stages by a kind of lamination. This leads to a recognition of the peculiar elevated style used in the second half of Mark 5, a consideration of the thrust and import of this style, and the suggestion of a special salience scheme which could be posited to accommodate this passage and others.

Finally, in section 5 I compare Matthew’s highly abbreviated, but parallel, account of this same passage with the Markan account. Without committing myself to any theory of literary dependency, I attempt to show that what is omitted in Matthew relates quite systematically to material found in certain information bands in Mark.

Mark 5 has been chosen as the testing ground for the theory of the interdependence of verb saliency and constituency structure because it occurs at a crucial spot in the structure of the Gospel, and for that reason has a certain fulness of structure and style. It concludes the first main section of the Gospel—which I entitle ‘The Rise to Prominence’ (1:14–5:43)—following the Stage (1:2–8), the Ministry of John the Baptist; and Episode 1, Inciting Incident (1:9–13), the Holy Spirit comes on Jesus at his baptism and confirms his Sonship, then drives him out to be tempted of the Devil. Chapter 5, with its three miracles—the healing of the Gadarene demoniac, the raising of Jairus’s daughter, and the healing of the woman with the disorder in her menstrual cycle—marks the peak of the whole embedded discourse that treats of Jesus’ rise to prominence.5 Mark typically winds down major and some minor sections of his Gospel with miracle stories told very dramatically with great detail and with reported dialogue. Mark 5 illustrates these peak-marking features. The occurrence of one story interrupting another story occurs nowhere else in either Mark or the other synoptics than in this pericope. Therefore this feature is highly unusual. In addition, the accounts in this chapter are given with consummate narrative skill. While all the three miracles that occur in ch. 5 are marked with typically Markan detail and dialogue, the two interwoven stories of the second half of the chapter are especially marked by ringing the changes on the various possible forms of the Greek verb.

2. Mark 5:1–20, the Healing of the Gadarene Demoniac

In terms of the categories of the narrative template6 we can distinguish sections of this story: Stage, 5:1–5; Inciting Incident, 5:6–8; Climax, 5:9–10; Denouement, 5:11–14, Closure, 5:15–20. Each of these sections has the structure of a recognizable paragraph type.

The Stage is expounded by a narrative comment paragraph. The Setting of this paragraph is a simple sentence with a motion verb in the aorist. Motion verbs are primarily verbs of

                                                            4 For a little‐known but careful piece of work I cite R. Enos, ‘The Use of the Historical Present in the 

Gospel According to Saint Mark’, The Journal of the Linguistic Association of the Southwest 3.2 

(1981), pp. 281–98. See also S.H. Levinsohn’s Discourse Features of New Testament Greek: A 

Coursebook (Dallas: Summer Institute of Linguistics, 1992), Chapter 10. 

5 See Longacre, ‘Top‐Down, Template‐Driven’. 

6 See Longacre, Grammar, Chapter 2. 

117  

orientation; here the verb ἦλθον, ‘came’, fulfills the narrative strategy of getting Jesus to the other side of the lake where he can interact with the Gadarene demoniac: ‘And he came to the other side of the lake to the country of the Gerasenes’. The thesis of this sentence is a simple sentence with a preparatio and a relative clause. The preparatio (see next section), a genitive absolute, links the coming event with the previous sentence and signals an impending change of actor: ‘And as he came up out of the boat …’ The sentence base reveals the new actor: ‘immediately there met him [aor.] out of the tombs a man with an unclean spirit who had [impf.] his dwelling in the tombs’. After this initial identification of the man, the Comment part of the paragraph follows. The Comment is expounded by a chiastic observation paragraph consisting of evidence 1, conclusion, and evidence 2. The conclusion, the key of the chiasm, states briefly with a verb in the historical present (conclusion of v. 4): ‘And no one is able to tame him’. The preposed evidence 1 is found in the second half of v. 3 and most of v. 4; the latter is an involved infinitival construction which does not translate smoothly into English: ‘No one could bind him with a chain—though often having been bound with chains and foot-fetters, he burst the chains and pulled apart the foot-fetters’. The postposed evidence 2 in the second half of v. 5 further emphasizes the man’s desperate condition with skillful parallel phrases: ‘And all night and all day, in the tombs and on the mountains, he was crying out and cutting [periphrastic impfs.] himself with stones’.

The Inciting Incident can be taken to be vv. 6–8 which constitute a narrative sequence paragraph. Its first Sequential Thesis is expounded by a coordinate sentence with preparatio: ‘And seeing Jesus from afar, he ran [aor.] and knelt [aor.] before him’. Sequential Thesis 2 is somewhat more involved in that it is expounded by a narrative reason paragraph. The Thesis of this embedded paragraph is a Quotation sentence with a preparatio and a speech verb in the historical present. ‘And crying out with a loud voice, he says “What do we have in common, Jesus Son of the Most High God? I put you on oath before God that you don’t torture me”.’ The Reason unit of the embedded paragraph gives the reason why the demoniac was so disturbed: ‘For Jesus had been commanding [impf.] him, “Come out of the man, you unclean spirit” ’.

I take the Climax to be vv. 9–10—although there is certain flow of the Inciting Incident into the Climax. The structure of the Climax is a resolved dialogue paragraph7 with a terminus at its end. The semantic structure is that of question and answer. Jesus asks the man his name, and the demons reply, ‘Legion is my name, for we are many’. As to speech verbs in the quotation formulas, the quotation formula of the question employs an imperfect verb and the quotation formula of the answer, a historical present. The terminus adds (with another impf. verb), ‘And they kept beseeching him not to send them out of the country’. The Climax pictures the struggle as squarely joined, with Jesus intent on exorcising the demons and the demons unwilling to come out and be gone.

The Denouement is swift and dramatic (vv. 11–14) and is reported in an Execution paragraph. The Setting of this paragraph is a simple sentence whose ‘be’ verb is in the imperfect: ‘And there was near there a great herd of swine grazing on the mountainside’. The Plan element of the paragraph is given in a simple sentence with a consecutive (see next section), in this case the verb ‘say’ which introduces a quotation: ‘And they beseeched him [aor.], saying, “Send us into the swine that we may enter into them”.’ The Execution is expounded by a narrative result paragraph whose Thesis is very terse: ‘And he gave permission [aor.] unto them’. The Result is more complex: it consists of a narrative sequence paragraph with two Sequential Theses, each expounded by a coordinate sentence. The Sequential Thesis 1 is a coordinate sentence with a preparatio and three coordinated bases: ‘And the unclean spirits having come out of the man, entered [aor.] into the swine, and the

                                                            7 Cf. Longacre, Grammar, Chapter 5: ‘Repartee: Dialogue Paragraphs’. 

118  

herd rushed down [aor.] a cliff into the lake (around two thousand), and were drowned [aor.] in the lake’. Sequential Thesis 2 is expounded by a coordinate sentence which at first glance seems to have three coordinated bases but on closer look it appears that the first two coordinated bases are associated into a coordinate sentence against the third. Consequently we consider that the first base of the coordinate sentence is itself expounded by a coordinate sentence: ‘And the ones who cared for them fled [aor.] and announced [aor.] to the town and countryside’. These two clauses group with the second base of the higher layer of sentence structure: ‘And they came to see what had happened’.

The Closure, which treats of ensuing events (vv. 15–20), is one long narrative sequence paragraph with three sequential theses, the third of which is expounded by a simple dialogue paragraph which consists of a proposal, an answer and a correlate. The Setting of this narrative Sequence paragraph is expounded by a coordinate paragraph with three bases. The Setting is depictive and graphic, with the verbs of the first two bases historical presents and the verb of the last base an aorist: ‘And they come, and they see the one who had been possessed sitting, clothed, and in his right mind, the one who had had the legion, and they feared [aor.]’. Sequential Theses 1 and 2: ‘And those who had seen how it happened to the possessed one gave their account [aor.], and concerning the swine. And they began [aor.] to ask him to leave their country.’ Sequential Thesis 3, as mentioned above, is expounded by a simple resolved dialogue paragraph. The Initiating Utterance (Proposal) of this dialogue is reported in a simple sentence with preparatio; it functions as an indirect quotation ‘And as he was getting into the boat [gen. abs.], the one who had been possessed was begging [impf.] him that he might be with him’. The Resolving Utterance (Response) is expounded by a contrast sentence, the second base of which is adversative and embeds a quotation sentence: ‘And Jesus did not permit [aor.] him, but says [historical present] to him, “Go back to your house and your family, and tell them how much the Lord has done for you and has had mercy on you”.’ The last sentence (v. 20) I label a Correlate.8 A correlate reports activity which correlates with the Response of a Dialogue or Execution paragraph. Here the correlate is the coordinate sentence that reports the actions of the man in response to Jesus’ injunction: ‘And he went away [aor.], and he began [aor.] to proclaim in the Decapolis how much Jesus had done for him and had had mercy upon him’.

This is summarized and graphically illustrated in Diagram 1, which is a stepped (tree) diagram; abbreviations within the diagram correspond to the units set up discursively above. In constructing this diagram I have used my apparatus for describing paragraph types.9

Diagram 1. Stepped (Tree) Diagram of Mark 5:1–20

Stage: (N) Comment paragraph (5:1–5)

Setting: and he came (aor.) to the other side of the lake to the country of the Gerasenes Thesis: Simple S with preparatio

Preparatio: And as he came up out of the boat (gen. abs.) Base: immediately there met (aor.) him a man out of the tombs with an unclean spirit who had (impf.) his dwelling in the tombs.

Comment: chiastic Observation paragraph:

                                                            8 Cf. R. Mansen and K. Mansen, ‘The Structure of Sentence and Paragraph in Guajiro Narrative 

Discourse’, in R.E. Longacre and F. Woods (eds.), Discourse Grammar: Studies in Indigenous 

Languages of Colombia, Panama, and Ecuador, Part 1 (Dallas: Summer Institute of 

Linguistics/University of Texas at Arlington, 1976), pp. 147–258. 

9 Longacre, Grammar, Chapter 4. 

119  

Evidence 1: And no one could (impf.) bind him with a chain—though often having been bound with chains and foot-fetters he burst the chains and pulled apart the foot-fetters (infinitival constructions) Conclusion: No one is able (HP) to tame him Evidence 2: By night and day among the tombs and on the mountainsides he was (impf.) crying out and cutting himself with rocks

Inciting Incident: (N) Sequence paragraph (5:6–8) Seq. Thesis 1: Coord S with preparatio

Preparatio: And seeing Jesus from afar Base 1: he ran (aor.) Base 2: and knelt before him (aor.)

Seq. Thesis 2: (N) Reason paragraph Thesis: Quotation S with preparatio

Preparatio: And crying out with a loud voice QF: he says (HP) Q: ‘What do we have in common, Jesus, Son of the Most High God; I put you on oath before God that you don’t torture me.’

Reason: For Jesus had been (impf.) commanding him, ‘Come out of the man you unclean spirit’.

Climax: Simple dialogue paragraph (5:9–10) IU (Q): and Jesus asked (impf.) him, ‘What is your name?’ RU (A): And he says (HP) to him, ‘Legion is my name, for we are many’. Terminus: And they kept beseeching (impf.) him not to send them out of the country.

Denouement: Execution paragraph (5:11–14)

Setting: And there was (impf.) near there a great herd of swine grazing on the mountainside Plan: Simple S with consecutive

Base: And they beseeched (aor.) him, Consecutive: Quotation S

QF saying Q: ‘Send us into the swine that we may enter into them’.

Execution: (N) Result paragraph Thesis: And he gave permission (aor.) to them Result: (N) Sequence paragraph

Seq. Thesis 1: Coord S with preparatio Preparatio: and the unclean spirits having come out of the man Base 1: entered (aor.) into the swine,

                                                            HP Historical present 

Coord Coordinate 

QF Quotation Formula 

Q Quotation 

IU Initiating Utterance 

(Q) Question 

RU Resolving Utterance 

(A) Answer 

120  

Base 2: And the herd rushed down (aor.) a cliff into the lake (around 2000), Base 3: and were drowned (aor.) in the lake.

Seq. Thesis 2. Coord S Base 1. Coord S

Base 1: And the one who cared for them fled (aor) Base 2: and announced (aor.) to the town and counryside;

Base 2: and they came to see what had happened. Closure: (N) Sequence paragraph (5:15–20)

Setting: Coord S Base 1: And they come (HP) Base 2: And they see (HP) the one who had been possessed sitting, clothed, and in his right

mind, the one who had had the Legion. Base 3: and they feared (aor.).

Seq. Thesis 1: And those who had seen how it happened to the possessed one gave their account (aor.), and concerning the swine. Seq. Thesis 2: And they began (aor.) to beg him to leave their country Seq. Thesis 3: Simple Dialogue paragraph

IU (Prop): Simple S with preparatio (participle in gen. abs.) Preparatio: And as he was getting into the boat Base: the one who had been possessed was begging (impf.) him that he might be with him.

RU (Response): Contrast S Base 1: And Jesus did not permit (aor.) him, Adversative Base: Quotation S

QF: but he says (HP) to him Q: Coord S

Base 1: ‘Go back to your house and your family, Base 2: and tell them how much the Lord has done for you and has had

mercy upon you.’ Correlate: Coord S

Base 1: And he went away (aor.) Base 2: and he began (aor.) to proclaim in the Decapolis how much Jesus had done for

him and had mercy upon him, Base 3: and everyone marveled (aor.).10

In the above passage, the role of the aorist is crucial to the storyline and the imperfect fills

in details. Note the depictive role of the imperfect in vv. 3–5 above, and in v. 11 where the grazing herd is introduced. In v. 8 the imperfect occurs in a reason slot which is ancillary to the preceding thesis and may indicate protracted activity rather than simple utterance: ‘For Jesus had been commanding him …’ Note also in v. 18 where the imperfect depicts the healed man as beseeching Jesus to let him go with him. Here too an ongoing activity, rather than a simple event, is depicted. Since Jesus’ response to him uses a verb with the aorist this presumably makes the response outrank the original initiating utterance. The historical present occurs a few times where it possibly dominates surrounding imperfects (vv. 3–5) but is outranked by the aorist.

For the most part, then, the storyline of koine Greek narrative is carried by clauses with verbs in the aorist. Topics which have not been discussed as yet are the function of the

                                                            Prop Proposal 

10 In this diagram, the following abbreviations are used: Coord = Coordinate; IU = Initiating Utterance; 

RU = Resolving Utterance; A = Answer; Prop = Proposal; (Q) = Question; Q = Quotation; QF = 

Quotation Formula; Inc Inc = Inciting Incident. 

121  

participle11 and of the historical present. Preposed participles (whether nominative or genitive absolute) which are dependent on a clause with an aorist supply the immediate backup to the storyline by adding preliminary detail. When clauses with the historical present occur—especially when clustered and not simply limited to verbs of speech and motion—they constitute a kind of secondary storyline, which can be thought of as either a demotion from the primary storyline or a promotion of privileged background material. This is most clearly seen in the next section where the structure of the second half of the chapter is presented. Preposed participles which are dependent on a clause with a verb in the historical present provide back-up for the clause that they accompany. Clauses with the imperfect function as background and often encode something conceived of as an ongoing activity rather than as a punctiliar event. As with the aorist and the historical present, clauses with the imperfect can have as backup preposed participial clauses. In addition to these concerns, clauses with ‘be’ verbs in the imperfect and verbless clauses supply setting.

In the above paragraph, the postposed participle is not mentioned. It appears that the postposed participle is of the same semantic rank as the verb that it follows; that is, it is consecutive on the preceding main verb and continues its function. This applies not only to the indicative forms which are found in narrative but to the imperative forms which are found in hortatory discourse.

In terms of universal grammar koine Greek represents a curious union of two distinct chaining structures which are found in many parts of the world:12 (1) Medial-final chaining structures where final verbs in final clauses are preceded by clauses with verbs less fully inflected and variously called gerunds, participles and co-verbs, or simply ‘medial’ verbs; and (2) initial-consecutive chaining structures where an initial clause with a fully inflected verb is followed by clauses with less fully inflected verbs, variously called sequential or consecutive. In respect to medial-final chaining structures the (otherwise unmarked) medial verb may be of the same salience rank as the final, or of lesser salience, or of greater salience. Greek preposed participles apparently are like medial verbs in such a language as Koreete (Ethiopia) in respect to being of lesser salience than the main verb which they precede. On the other hand, the Greek postposed participle, which continues the sense and function of the main verb, is like many African initial-consecutive systems. This conflux of systems can be seen in such a passage as Mt. 28:19–20 where the preposed aorist participle πορευθέντες ‘going’ precedes the main aorist imperative μαθητεύσατε ‘make disciples’ and is followed by two consecutive participles, βαπτίζοντες ‘baptizing’ and διδάσκοντες ‘teaching’. Here the preposed participle ‘going’ represents a preparatory action to what follows while the two postposed participles continue the action of the main verb and are semantically coordinate with it, issuing in a threefold command ‘Make disciples … baptize … teach’.

I summarize these ranking concerns in Diagram 2. In this diagram the three ranks—aorist, historical present, and imperfect—are grouped along with indication of consecutive forms and preparatory forms. In the stepped diagrams of this paper I refer to preposed participles as ‘preparatio’ and to postposed participles as ‘consecutives’. For the broader concerns reflected

                                                            11 For a very careful and almost encyclopedic treatment of the preposed Greek participle, see A. 

Healey and P. Healey, ‘Greek Circumstantial Participles: Tracking Participants with Participles in the 

Greek New Testament’, Occasional Papers in Translation and Textlinguistics 4.3 (1990), pp. 177–259. 

12 Longacre, Grammar, pp. 285–87. For more detail regarding chaining structures, see R.E. Longacre, 

Hierarchy and Universality of Discourse Constituents in New Guinea Languages (Washington, DC: 

Georgetown University Press, 1972); idem, Storyline Concerns and Word‐Order Typology in East and 

West Africa (Studies in African Linguistics, 10; Los Angeles: University of California, 1990). 

122  

in the diagram, note the priority of the aorist over the imperfect with the occasional historical present as a construction of intermediate status. This is, in a sense, the primary dimension of the diagram. The participles in two broad functions are like an intersecting second dimension: the preposed ancillary function and the postposed co-ranking function.

The following salience scheme recognizes the primacy of the aorist over the imperfect and in addition assigns a place in the scheme of things to the historical present—even though it is present in force only in a few passages such as those being examined here (but cf. Mark’s accounts of Gethsemane, the crucifixion and the resurrection). Isolated occurrences of the historical present with speech verbs and verbs of motion is a related problem not considered in this paper.

Diagram 2. Cline of Dynamicity for the Greek Verbs Found in Mark 5

1:1. Aorist and its consecutives (postposed participles) 1:2. Preposed participles dependent on an aorist

2:1. Historical present and its consecutives 2:2. Preposed participles dependent on the historical present 3:1. The imperfect and its consecutives

3:2. Preposed participles dependent on the imperfect 4. ‘Setting’: be verbs and verbless clauses.

Notes: 1:1, 2:1 and 3:1 are reminiscent of African chaining structures of the sort where the initial and the consecutive following it are of the same salience rank.13 1:2, 2:2 and 3:2 are reminiscent of Papua New Guinea, South America and Ethiopian Highlands in languages where the medial verb is one peg lower in salience than the final on which it depends.14

3. Mark 5:21–43, Jairus’s Daughter and the Woman with the Issue of Blood

I now present in this section stepped (tree) diagrams of Mk 5:21–43. This presentation is somewhat more detailed than that found in regard to the account which is found in the first part of this chapter from Mark’s Gospel. The increased detail of presentation is due to increased intricacy of the text itself at this point. I will also explain here in more detail the analytical apparatus assumed in the previous section.

Two interwoven structures are involved here: the bracketing story concerning the raising of Jairus’s daughter, and the bracketed story of the woman with the issue of blood (disorder of the menstrual cycle). In presenting the constituent structure of these accounts, our first recourse is again to the narrative template, that is, the natural characteristics of a story which embody its semantic structure in Stage, Inciting Incident, Rising Tension, Climax, Denouement and Closure—even as we did above in the story of the demoniac. But this does not access the constituent structure as such, since almost any size level unit may manifest one of these semantic functions, from an embedded discourse to a sentence. For this reason we again must also invoke models from paragraph and sentence structures. Here we have bundlings of clauses into looser and tighter structures which correspond respectively to the

                                                            13 Longacre, Storyline Concerns, pp. 111–43, and 173–77. 

14 Actually, there are three possible relationships of medial to final verbs: (1) the final verb outranks 

the medial, (2) the medial outranks the final, and (3) the medial and final are of equal rank. These 

concerns were not raised at the time of my New Guinea studies (Longacre, Hierarchy) but are 

presented as alternatives in Longacre, Storyline Concerns, where medial‐final chaining languages of 

the Ethiopian highlands are seen to display all three possibilities. 

123  

paragraph and sentence levels. A further complication is that a very short embedded discourse, such as a pericope which consists of a parable, realizes the whole semantic discourse structure within the domain of a single paragraph, that is, the narrative discourse structure collapses into the paragraph.

Specifically, I analyze the bracketing story, the raising of Jairus’s daughter, into discourse-level constitutents where elements of the narrative template correspond to paragraphs which can be isolated in the story. The bracketed discourse, the story of the woman with the flow of blood, seems to be best analyzed as a single paragraph in which the semantic elements of the narrative template are not distinguished in the surface constituent structure. Within the displays, when a sentence is posited, the constituents are labelled as appropriate to that level. Thus one or more sentence bases are posited as constitutents of the sentence. Within paragraphs, paragraph level units such as theses and sequential theses are posited as appropriate for that level. Our sentence boundaries correspond as a whole to those indicated in the Greek New Testament as punctuated by Aland et al., except that I take the colon (raised dot) as often indicating a new structural sentence. a. The Discontinuous Story of the Raising of Jairus’s Daughter This narrative has four parts: the Stage (5:21; the Inciting Incident (5:22–24); the Climax (5:35–36); and the Denouement (5:37–43), which correspond respectively to the physical position and circumstances; the request by Jairus on behalf of his sick daughter, the persistence of Jesus in the face of being told that the little girl had died, and the raising of the girl from the dead. Between the Inciting Incident and the Climax occurs the intervening story of the woman with the issue of blood. If a story is to be discontinuous, a somewhat natural position for the intervening element is to follow the Inciting Incident—as in the stories of Joseph and Jacob in the Hebrew Bible.15

The Stage (v. 21) is expounded by a compound sentence with a preparatio which is a participle in the genitive absolute (cf. 5:2 and 18). The sentence is coordinate; its first base has a verb in the aorist and its second base has the verb ‘be’ in the imperfect. The motion verb of the preparatio gets Jesus to the other side of Lake Tiberias (cf. 5:1). The sentence bases report the gathering of a great crowd and Jesus’ being with them by the lakeside. So far, business as usual!

The Inciting Incident (vv. 22–24a) is given in two sentences: a long coordinate sentence with three bases which reports the coming of Jairus with the request that Jesus come and heal his daughter; and a simple sentence reporting that Jesus went off with him as requested. The bases of the long coordinate sentence have verbs in the historical present: ‘comes’ in the first base, ‘falls at his feet’ in the second base, and ‘greatly beseeches him’ in the third base. The verb form ‘falls’ in the second base is preceded by a preparatio ‘seeing him’. The verb ‘beseeches’ in the third base is accompanied by a consecutive ‘saying’ which introduces an embedded quotation sentence. Thus, the quotation formula has a consecutive verb which is equal in salience with the preceding verb ‘beseeches’. The Quotation itself has a double purpose margin ‘that you may put your hands on her that she might live’, but the whole sentence is elliptical, in that the clause ‘My daughter is dying’ is not followed by some such element as ‘I ask you’—which we might have expected. Perhaps the roughness of the grammar expresses the desperation of a father who is distracted and somewhat incoherent. The second sentence ‘And he went away with him’ has an aorist verb which reports the crucial fact that Jesus went off with Jairus towards the house where the sick child lay.

The Climax (vv. 35–36) is expounded by a Complex Dialogue paragraph which reports an initiating utterance, a proposal on the part of emissaries from Jairus’s house that, since the

                                                            15 R.E. Longacre, ‘Genesis as Soap Opera’, JOTT 7.1 (1995), pp. 1–8. 

124  

daughter has died, he shouldn’t trouble the teacher any further about coming to heal her; and a continuing utterance, a counter-proposal on the part of Jesus to Jairus that the latter should not despair but believe.

The proposal of the initiating utterance is given in a sentence that has both a preparatio and a consecutive. The preparatio (gen. abs.) ‘and while he was yet talking’ refers back to the embedded story which ends with Jesus telling the woman to go in peace; it thus serves to connect the bracketed material smoothly to the bracketing material. The sentence base reports the arrival of the emissaries: ‘they come from the synagogue ruler’s house’. The consecutive, with the participle ‘saying’, introduces another embedded quotation sentence ‘Your daughter has died; why trouble the Teacher further?’ Jesus’ counter-proposal is given in a sentence with a preparatio and a base: ‘And Jesus, having heard the word which was spoken, says to the synagogue ruler “Don’t fear; only believe” ’. All the main verbs found in the Climax are historical presents.

The Denouement (vv. 37–43) is the longest and most complex part of the narrative. The Climax left us with a tightly knotted situation: The daughter is dead but Jesus, far from turning back, persists in his purposes. The resolution of the dilemma is expressed in a long narrative sequence paragraph. The first Sequential Thesis of the sequence paragraph is expounded by another complex dialogue paragraph which after some preliminary moves ends up with Jesus insisting to the mourners that the child was not dead but merely sleeping, followed by the scornful response of the mourners. The preliminary moves of this complex dialogue paragraph include a setting reported in a negated aorist: ‘And he didn’t let anyone go with him except Peter, James and John, the brother of James’. Jesus’ arrival at the house and what he saw on arrival are reported in a coordinate sentence whose verbs are, again, historical presents: ‘He comes to the synagogue ruler’s house and sees much commotion, crying and wailing’. The Initiating Utterance is given in a quotation sentence with a preparatio: ‘And coming in [present participle], he says to them, “Why are you making this commotion and crying? The child is not dead but sleeping”.’ This remark of Jesus is followed by what is, in effect, the counter-remark of the mourners but given in a clause with an imperfect verb, probably implying ‘They persisted in laughing at him’.

The second Sequential Thesis of this narrative sequence paragraph should probably be regarded as expounded by a reason paragraph. The verbs of the Reason (vv. 40–41) are in the historical present and report in sequence Jesus’ actions in approaching the girl. The thesis, which expounds the consequences of this chain of actions, reports the resurrection of the child in a crucial aorist form followed by some imperfects: The clause with the aorist is further marked by the occurrence of εὐθύς, ‘immediately’. Base 2 of the same sentence reports the child walking about a bit and a reason is appended ‘for she was [impf.] twelve years old’. The structure of the Reason unit—the cause and explanation of the resurrection of the child—is that of an embedded narrative sequence paragraph (vv. 40b–41). The first Sequential Thesis of this embedded sequence paragraph is a coordinate sentence with a preparatio: ‘And putting them all out [part.], he takes with him the father of the child and the mother and those with him and he enters where the child was’. Here again the main verbs are historical presents. The second Sequential Thesis is a quotation sentence with a preparatio. The structure of this quotation sentence is complicated by the narrator’s parenthetical remark as to the meaning of the Aramaic phrase which Christ used. ‘And taking the child by the hand [part.], he says to

her [historical present] “Talitha koum”, which is, being interpreted, “Little Girl, I say unto you, Get up”.’ Finally, as already commented on above, the girl got up.

The last Sequential Thesis of the main sequence paragraph is expounded by another embedded narrative sequence paragraph; but now all the verbs are aorists—after the long                                                             part. participle. 

125  

spate of historical presents in the interior of the story: ‘And they were astonished with great astonishment’. Then, in another coordinate sentence: ‘And he commanded them not to tell anyone, and he told them to give her something to eat’. The structure of this narrative, as discussed above, is given in the stepped diagram which constitutes Diagram 3.

Diagram 3. Stepped (Tree) Diagram of the Discontinuous Story of the Raising of Jairus’s Daughter

Mark 5:21–24, 35–43 Stage (5:21): Coord S

Preparatio (gen. abs. with aor. part.): And Jesus having passed over again to the other side, Base 1: a great crowd gathered (aor.) together unto him, Base 2: and he was by the sea (impf.).

Inc Inc (5:21–24a): (N) Sequence paragraph Thesis 1: Coord S

Base 1: There comes (HP) one of the synagogue rulers named Jairus Base 2: Simple S

Preparatio: and seeing him (part.) Base: falls (HP) at his feet

Base 3: Simple S Base: greatly beseeches (HP) him Consecutive: (Q Sentence)

QF: saying (postposed part.) Q: (Sentence with purpose margins and ellipsis): ‘My daughter is dying; [I ask you] to come (prep.) put your hands on her that she may be healed and live.’

Thesis 2: Simple S: And he went away (aor.) with him. (This story is broken here by the insertion of the story of the healing of the woman with the issue of blood)

Climax (5:35–36): Complex Dialogue paragraph

IU (Prop.): Simple S Preparatio ([gen. abs. with preposed part.] and backreference to the end of the interposed story): And while he was yet talking, Base: they come (HP) from the synagogue ruler’s house Consecutive:

QF: saying (postposed part.): Q: ‘Your daughter has died; why trouble the Teacher further?’

CU (C-Prop.): Simple S Preparatio (preposed part.): And Jesus having heard the word spoken Base: Quotation S.

QF says (HP) to the synagogue ruler: Q: Coord S: ‘Don’t fear; only believe’.

Denouement (5:37–43): (N) Sequence paragraph Seq. Thesis 1: Complex Dialogue paragraph

Setting: Simple S: And he didn’t let (neg. aor.) anyone follow along with him except Peter, James and John the brother of James. Lead-in: Coord S

Base 1: And he comes (HP) to the synagogue ruler’s house,

                                                            Inc Inc Inciting Incident 

CU Continuing Utterance 

C‐Prop counter Proposal 

126  

Base 2: and he sees (HP) commotion, crying and much wailing. IU (Rem.): Quotation S

Preparatio: And coming in (pres. part.) QF: he says (HP) to them Q Coord. S

Base 1: ‘Why are you making this commotion and crying? Base 2: Contrast S

Base 1: The child has not died, Base 2: but she is sleeping.’

CU (non-speech verb; counts as c-Rem): Simple S: And they laughed (impf.) at him. Seq. Thesis 2: Reason paragraph

Reason: (N) Sequence paragraph Seq. Thesis 1: Coord S

Preparatio: And putting them all out (pres. part.) Base 1: he takes with him (HP) the father of the child and the mother, and

those with him Base 2: And he enters (HP) where the child was (impf.)

Seq. Thesis 2: Quotation S Preparatio: And taking the child by the hand, QF he says (HP) to her Q: ‘Talitha koum’ (which is interpreted)

Q ‘Little girl’ QF I say to you, Q ‘get up’.

Thesis: Reason sentence Base: Coord S

Base 1: And immediately the child stood up (aor.) Base 2: and she was walking about (impf.)

Reason: for she was twelve years. Seq. Thesis n (N) Sequence paragraph

Seq. Thesis 1: And they were astonished (aor.) with great astonishment Seq. Thesis 2: Coord S

Base 1: And he commanded (aor.) them not to tell anyone Base 2: And he told (aor.) them to give her something to eat.16

In commenting on the structure thus conceptualized and diagrammed, I note: (1) It

structures plausibly as a low level (embedded) narrative built on the narrative template even as is the entire Gospel of which it is a part. Breaking off after the Inciting Incident, it resumes after the interruption with a Climax and Denouement. In the course of the pericope, a lot of detail and liveliness is added by the use of the historical present while aorists occur at the most decisive points in the story. b. The Bracketed Story: The Healing of the Woman with the Issue of Blood (Mark 5:24b–34) This account probably can be considered to consist of one long narrative result paragraph. Semantically the functions of Stage, Inciting Incident, Climax and Denouement can be                                                             Rem Rem = Remark 

c‐Rem counter Remark; Rem = Remark; part. = participle. 

16 Besides the abbreviations used in Diagram 1, the following further abbreviations are used in this 

diagram: CU = Continuing Utterance; c‐Prop = counter Proposal; c‐Rem = counter Remark; Rem = 

Remark; part. = participle. 

127  

distinguished within the paragraph, but it hardly seems possible or necessary to break the paragraph into four subparagraphs which correspond to these functions.

The paragraph Setting is expounded by a coordinate sentence with two bases; both verbs are imperfects: ‘A large crowd was following him and jostling him’.

The Thesis of the Result paragraph is expounded by a reason sentence which has a long and complex structure. This sentence has an involved preparatio, followed by the sentence base, which is dominated by a verb in the aorist, and which is in turn followed by a reason margin. The preparatio which accompanies the base is a striking construction: it consists of a long chain of preposed participles organized as if it were an embedded coordinate sentence with all its verbs reduced to participles. The first five participles go together and give the woman’s medical case history: ‘A certain woman being in an issue of blood for twelve years, and having suffered many things from many physicians, and having spent all that she had, and having got none the better but rather coming out worse …’ (Notice how the last two participles are equivalent to a contrast sentence with its main verbs reduced to participles.) The embedded construction which gives the case history consists of participles that are connected with καί, ‘and’. Two further participles occur without introducing conjunctions: ‘having heard concerning Jesus, coming up behind him in the crowd’. The sentence base follows with its electrifying aorist verb: ‘she touched his garment’. The appended reason is fittingly given in a clause with the imperfect; she must have been repeatedly reassuring herself as she pushed her way through the tightly packed and jostling crowd, ‘If I but touch his clothing, I shall be healed’.

The Result part of the paragraph contains two theses which tell us, in effect, that she immediately realized that something had happened, and so did Jesus. Thesis 1, which reports to us the woman’s point of view, is a coordinate sentence both clauses of which have aorist verbs: ‘And immediately her bleeding ceased and she knew/sensed in her body that she was freed from her plague’. Thesis 2 reports the same event from Jesus’ perspective. A simple sentence with a complex preparatio and a quotation verb in the imperfect reports Jesus’ words ‘Who touched my clothing?’ The preparatio is an embedded coordinate sentence with its verbs reduced to participles: ‘And Jesus immediately realizing that power had gone out of him, and having turned about in the crowd …’ The question of Jesus leads to a puzzled, if not impatient, counterquestion from the disciples: ‘You see the people jostling you and you ask “Who touched me?” ’ The next sentence, a sort of indirect further question on Jesus’ part is given with a verb in the imperfect; the sense might be ‘But he persisted in looking around to see who had done it’. Here the interplay with the disciples ends and the very next sentence gives from a different quarter the resolving utterance which finally answers Jesus’ question. This sentence, itself coordinate, has a preparatio which consists again of a coordinate sentence reduced to participles: ‘And the woman fearing and shaking, and knowing what had happened to her …’ Then follow the three coordinated sentence bases all with aorist verbs: ‘came, and fell at his feet, and told him the whole truth’. Finally, Jesus terminates the interview by acquiescing in what has happened and blessing the happy but frightened woman: ‘Daughter your faith has saved you; go in peace, and be healed of your plague’. See Diagram 4 for a summary presentation of this structure.

Diagram 4. Stepped (Tree) Diagram of Mark 5:24b–34

(N) Result paragraph

Setting 5:24b Coord S Base 1: A large crowd was following (impf.) Base 2: and jostling him (impf.)

Thesis: 5:25–28 Reason S Preparatio: (embedded) Coord S (preposed parts.)

Base 1: (embedded) Coord S (preposed parts.)

128  

Base 1: A certain woman being in an issue of blood for 12 years: Base 2: and having suffered many things from many physicians Base 3: and having spent all she had Base 4: (embedded) Contrast S (preposed parts.)

Thesis: and having got none the better Opp. Thesis: but coming out worse

Base 2: having heard concerning Jesus Base 3: coming up in the crowd from behind

Base: touched (aor.) his garment; Reason: for she was saying (impf.), ‘If I but touch his clothing I shall be saved’.

Result: (N) Coord paragraph 5:29–34 Thesis 1: Coord S 29:

Base 1: And immediately her (εὐθύς) bleeding ceased (aor.), Base 2: And she knew (aor.) in her body that she was freed from her plague.

Thesis 2: Complex Resolved Dialogue paragraph 5:30–34 IU (Q): Simple S v. 31

Preparatio: (embedded) Coord S (preposed parts.) Base 1: And Jesus immediately (εὐθύς) having known that power had

gone out of him, Base 2: having turned about in the crowd

Base: was saying (impf.), ‘Who touched my clothing?’ CU (c-Q): And the disciples were saying (impf.) to him ‘You see the people jostling you and you ask ‘Who touched me?’. CU (indirect; c-Q): And he was looking about (impf.) to see who had done it. RU (A): Coord S

Preparatio: (embedded) Coord S (preposed parts.) Base 1: And the woman fearing and shaking (parts.) Base 2: and knowing what had happened to her (part.)

Base 1: came (aor.) Base 2: and fell at his feet (aor.) Base 3: and told (aor.) to him the whole truth

TU (Acq): And Jesus said (aor.) to her, ‘Daughter, your faith has saved you; go in peace, and be healed of your plague’.17

Note especially in this paragraph the string of participles which encodes the woman’s case

history and her preparatory actions in approaching Jesus. Strings of participles of this length are uncommon in the Greek New Testament (cf. Mt. 14:19); where they do occur, they are reminiscent of chains of medial-final clauses in languages of Papua New Guinea. In the passage at hand the preposed participles serve to summarize neatly the woman’s case history and preparatory actions while at the same time building up suspense relative to the aorist verb ‘she touched his clothing’ which follows. Imperfects also find their place in this narrative in representing the information that a large crowd was following and jostling Jesus, in reporting the woman’s encouraging herself to take the action that she did, and in reporting Jesus’ sideplay with the disciples regarding who touched him. While some compression is evident in

                                                            c‐Q counter Question; TU = Terminating Utterance; Acq = Acquiescence. 

TU Terminating Utterance 

Acq Acquiescence 

17 Besides the abbreviations used in the preceeding diagram, the following abbreviations are used 

here: c‐Q = counter Question; TU = Terminating Utterance; Acq = Acquiescence. 

129  

the use of the preposed participles in the bracketed discourse, there is a certain expansion evident in the use of the imperfects.

In the two sections that follow I will demonstrate the generation of a story from the storyline downwards by progressively adding elements of less salience,18 and will then compare Matthew’s stripped down account of the same passage to show what elements he omits.

4. Generating the Stories from their Abstracts

In performing this demonstration I handle the two stories together—the bracketing and the bracketed story as presented by Mark. I will refer to this composite account here simply as ‘the story’.

A quite bare, unadorned abstract of the story can be obtained by putting down in order the clauses whose verbs are aorist; I note some of the deficiences of this abstract—especially in terms of participant reference in parenthesis at certain places. Sometimes, however, a noun phrase early in the sentence identifies the subject of the main clause before the string of preposed participles in the preparatio, so this is included in the abstract whenever the main verb is aorist. Jesus is the default participant (as central) who is often referred simply by third-person masculine pronoun or the corresponding verb affix.

Display 1 (clauses with verbs in the aorist)

A great crowd gathered together unto him. And he went off with him (with whom?) And she (a certain woman, as preposed topic of entire sentence) touched him. And immediately her bleeding ceased. And she knew in her body that she was free from her plague. And she came, and fell before him, and told him the whole truth. And he said to her, ‘Daughter, your faith has saved you; go in peace, and be healed of your

plague’. And he didn’t permit anyone except Peter, James and John to follow along with him. And immediately the little girl got up. And they were astonished with great astonishment. And he commanded them strictly that they should tell no one about this. And he said that they should give her something to eat. The bare abstract of the story is fleshed out considerably by adding clauses which contain

verbs in the historical present, thus making the latter appear somewhat as a surrogate for the aorist. I repeat all the clauses given under the above diagram and add the clauses which contain a historical present. More quotations occur here in this partially rounded out abstract of the story, because in this text there is more reported speech introduced by quotation formulas in the historical present than by speech verbs in the aorist. Also I count as equivalent to a historical present any consecutive participle that follows a historical present; just as above in one instance we counted as equivalent to an aorist a consecutive participle that followed an aorist.

                                                            18 Longacre, ‘Two Hypotheses’. 

130  

Display 2 (abstract plus clauses with historical present)

A great crowd gathered together unto him. There comes to him a synagogue ruler named Jairus. He falls at his feet. And he greatly beseeches him, saying, ‘My daughter is at the point of death. [I beseech you that] coming you put your hands upon her and she will be healed and live.’

And he went off with him. And she touched him. And immediately the flow of her blood stopped. And she realized in her body that she had been healed of her plague. And she came, and fell before him, and told him the whole truth.

And they come from the rulers of the synagogue’s house saying, ‘Your daughter has died. Why trouble the Teacher further?’ And Jesus says to the ruler of the synagogue, ‘Don’t fear; only believe’.

And he permitted no one to go with him except Peter, James and John, the brother of James. And he comes to the house. And he sees the confusion, crying and wailing. And he says to them, ‘Why are you mourning and weeping? The child is not dead; she is only

sleeping.’ And he takes the father of the child and the mother and those with him and he goes in to where

the child was. And he says to her ‘Talitha koum, which is being interpreted, Little girl, I say to you, Get up!’

And immediately the little girl got up. And they were astonished with great astonishment. And he commanded them strictly that they should tell no one about it. And he told them to give her something to eat.

In the above, note that it is the bracketing story that is especially rounded out by the

addition of clauses whose verbs are historical presents. This affords contrast with the bracketed story where the narrative is especially rounded out by participles and clauses with verbs in the imperfect.

I now enrich the above by adding preposed participles which I have called preparatio above.

Display 3

And Jesus having crossed over again to the other side, a great crowd gathered together unto him;

And there comes a ruler of the synagogue named Jairus, And he greatly beseeches him, saying, ‘My daughter is at the point of death. [I beseech you that] coming you will put your hands on her that she may be healed and live.’

And he went off with him. And a woman being in an issue of blood for twelve years, And having suffered many things from many physicians, And having spent all that she had, And having got none the better, but rather coming out worse, having heard about Jesus, coming up in the crowd from behind,

she touched his clothing. And immediately the flow of her blood ceased,

131  

And she realized in her body that she was healed of her plague. And the woman frightened, and trembling, knowing what had happened to her,

came, and fell at his feet, and told him the whole truth. While he was yet talking (part. in gen. abs.)

they come from the ruler of the synagogue’s house, saying, ‘Your daughter has died; why do you trouble the teacher further?’

And Jesus, having heard the word that was spoken, says to the ruler of the synagogue, ‘Don’t fear; only believe.’

And he permitted no one to go with him except Peter, James and John the brother of James. And they come to the house of the ruler of the synagogue, And he sees confusion, crying and wailing.

And coming in he says to them, ‘Why are you mourning and crying? The child hasn’t died; she’s only sleeping.’

But he, putting them all out, takes with him the father of the child and the mother and those with him, and goes into where the child was.

And taking the child’s hand,

he says to her, ‘Talitha koum’, which is being interpreted, ‘Little girl, I say to you, Get up’.

And immediately the little girl got up. And they were astonished with great astonishment. And he commanded them strictly that they should tell no one about it. And he said that they should give her something to eat.

To all the above we need to add clauses with verbs in the imperfect in order to generate

the actual text of the story. Of course, a further parameter, that of participant identification and tracking, weaves itself in along with the concerns centering around the varying forms of the verbs.

As we have seen above, the subject is typically identified at sentence onset before the preparatio. Also we have taken note of Jesus’ status as central participant and by default receiving minimal overt reference. The imperfect is especially used in rounding out the bracketing story, the raising of Jairus’s daughter, but it is also relevant to the bracketed story. I now display the full text of the story with the clauses with the imperfect indented the furthest to the right.

Display 4

And Jesus having crossed over again to the other side, A great crowd gathered together unto him.

And he was beside the lake. And there comes a ruler of the synagogue named Jairus, And he greatly beseeches him, saying, ‘My daughter is at the point of death, [I beseech you that] coming you will put your hands upon her that she may be healed and live’.

And he went off with him. And there followed him a great crowd, And they were jostling him.

132  

And a certain woman being in an issue of blood for twelve years, And having suffered many things from many physicians, And having spent all that she had, And having got none the better, but rather coming out worse, having heard concerning Jesus, coming up behind in the crowd,

she touched his clothing for she kept saying to herself, ‘If I but touch his clothing, I will be healed’.

And immediately the flow of her blood ceased, And she sensed in her body that she was healed of her plague.

And immediately Jesus realizing that power had gone out of him having turned about in the crowd,

said, ‘Who touched my clothing?’ But the disciples said, ‘You see this crowd jostling you and you say, “Who touched me?” But he kept looking around to see who had done this.

And the woman frightened, and trembling, and knowing what had happened to her,

came, and fell at his feet, and told him the whole truth. And while he was yet talking,

they come from the ruler of the synagogue’s house saying, ‘Your daughter has died, why trouble the teacher further?’

And Jesus having heard the word that was spoken, says to the ruler of the synagogue, ‘Don’t fear, only believe’.

And he didn’t let anyone follow along with him except Peter, James and John the brother of James.

And he comes to the house, And he sees confusion, crying and wailing,

And coming in, he says to them, ‘Why are you mourning and crying? The child hasn’t died, she’s only sleeping.’

And they laughed at him. But he, putting them all out,

takes along with him the father of the child and the mother, and those that accompanied him,

he comes into the place where the child was. And taking the hand of the child,

he says to her, ‘Talitha koum’, which is being interpreted, ‘Little girl, I say to you, Get up’.

And immediately the little girl got up. And she walked about. For she was twelve years.

And they were all astonished with great astonishment. And he commanded them strictly that they should tell no one about it. And he told them to give her something to eat.

133  

In considering the progressively laminated text displayed above, several comments come to mind: (1) This is not a run-of-the-mill koine Greek New Testament narrative. The role of the historical present is such that in this passage it patterns as a secondary storyline in a manner reminiscent of certain languages of East Africa, for example, Avocaya.19 The aorist clauses preserve the position of prominence but the historical presents add much to flesh out the story. The normal expectation would be that a koine Greek New Testament narrative would primarily be analyzable as an interplay between the aorist and the imperfect, but the intervention of the historical present is very striking in this account. (2) The role of the preposed participle is also especially heavy in this story with seven participles preceding the aorist ‘she touched his clothing’ and three participles preceding the coordinated aorist clauses ‘She came, fell before him, and told the whole truth’. Clearly, if the majority of koine Greek New Testament stories were structured like this one, our views of both Greek grammar and of discourse structure would be quite different from those that we hold at present.

What then? I see no way of accounting for the distinct structure of this story but by positing that its peculiar elaboration correlates with its position in the context of the whole Gospel. Why does Mark pull out all the stops on his narrative organ at this point? I refer again here to the essay on which this one is built and to the analysis propounded in that essay.20 All of Mark 5—including the story of the Gadarene demoniac in Mk 5:1–20—is the climax of the first grand wave of development in the Gospel. After the Aperture (1:1), Stage (1:2–8) and Inciting Incident (1:9–13), there occurs a section of the Gospel, beginning at 1:14 and terminating at the end of ch. 5, that describes the rise to prominence of Jesus as a teacher and miracle worker, and the beginnings of his controversy with the establishment of his day. This section, which is an embedded narrative discourse, contains a didactic peak in ch. 4 (the parables) followed by an action peak in ch. 5. Mark presents Jesus as one mighty in word and deed by this dual peak structuring in the first grand section. The action peaks of Mark are customarily miracle stories told with recourse to detail and dialogue. In Mark 5 the account of the healing of the demoniac exemplifies these features. But in the second part of the chapter, Mark outdoes himself. Not only do the features of detail and dialogue persist but the grammar and discourse structure are, as it were, reshaped by the unusual prevalence of the historical present and the preposed participle. Even the overall effect of the story-within-story contributes to this effect. For anything at all similar to Mark’s elevated style here we can look at the Gethsemane account in Mk 14:27–31 and the mocking and crucifixion in Mk 15:16–32. These portions of Mark occur in the action peak of the whole Gospel, where the Gethsemane episode is pivotal (the culmination of what precedes and the beginning of the culmination), the mocking and cruxifixion are Climax, and the resurrection account (16:1–8) patterns as Denouement. In the latter, however, there are only three historical presents compared to eight uses of this tense in the Gethsemane episode and nine such uses in the mocking and crucifixion. I believe that any attempt to analyze the role of these clusters of historical presents in Mark in terms of purely local constraints flies in the face of Gödel’s theorem that the consistency of a system cannot be explained wholly within the system itself. Therefore, I ultimately have to explain the clusters of historical presents in Mark in terms of total context in the Gospel.

                                                            19 In an unpublished data paper of Lynne Callinan and Eileen Kilpatrick, whose main points are, with 

the permission of the authors, incorporated into Longacre, Storyline Concerns, pp. 91–99. 

20 Longacre, ‘Top‐Down, Template‐Driven’. 

134  

5. Comparison with Matthew’s Account in Matthew 9:18–26

Matthew’s account in comparison with Mark’s seems to omit practically all the details which make Mark’s account so vivid and appealing. Matthew collapses into one the two episodes in Mark where the ruler first comes and asks Jesus to heal his daughter and then receives word on the way to his house that his daughter is dead. The three historical presents which Mark uses to depict Jairus’s coming to Jesus, falling before him, and beseeching him to heal his daughter are collapsed into a preparatio ‘coming’ and an imperfect ‘he beseeched him’. The six historical presents found in Mk 5:38–40—‘comes in’, ‘looks at/beholds’, ‘says’, ‘takes with him’, ‘goes in’ and ‘says’ (to the little girl)—are variously reduced; the first two to preposed participles, the third to an imperfect, the fourth and sixth to zero and the fifth again to a participle. Matthew is much more summary: ‘he took the girl by the hand and she got up’.

Mark’s mention of the following and jostling crowd—depicted in imperfects—is omitted entirely with only the routine observation that the disciples accompanied him. The case history of the woman with the issue of blood, given in five participial clauses in Mark, is also omitted along with the mention that the woman had heard about Jesus (still another participle in Mark). Matthew settles for one participle ‘having come up behind him’ and the main aorist verb ‘she touched [the edge of his cloak]’. The reason clause with the imperfect of the woman’s inner speaking occurs, however, as in Mark, ‘for she said to herself, “If I but touch his clothing I shall be healed” ’. Along with the omission of reference to the jostling crowd is omitted the reference to Jesus’ turning around, questioning the disciples, and making visual search to see who touched him—all Markan imperfects.

In all these ways the dramatic detail and dialogue of Mark are reduced to something approaching the macrostructure, that is, a bare summary, of the story.21 It is interesting that the omitted material in Matthew’s account corresponds to Mark’s development via historical presents, participles and imperfects. This, in turn, emphasizes that Mark’s extensive use of these three constructions is essentially embellishment. This throws us back again on the centrality of the aorist in koine Greek storytelling.

                                                            21 T. van Dijk, Text and Context (London: Longmans, 1977); idem, Macro‐structures: An 

Interdisciplinary Study of Global Structures, Interaction, and Cognition (Hillsdale, NJ: Lawrence 

Erlbaum, 1980). 

135  

THE TESTAMENTAL DISCIPLE-INSTRUCTION OF THE MARKAN JESUS (MARK 13): ITS LEVELS OF

COMMUNICATION AND ITS RHETORICAL STRUCTURES Wolfgang Schenk

1. Introduction

In this essay and in my other works on discourse analysis, I follow the methodological insights of German Textlinguistics.1 In particular, my model of textlinguistics works on the basis of the Peirce-Morris Triad: Syntax, Semantics, Pragmatics.2 My use of this and other

                                                            1 Cf. A. Eschbach and W. Rader (eds.), Literatursemiotik, I–II (Tübingen: Narr, 1980), which does not, 

as some do, uncritically join together heterogenous text‐units without testing the different starting 

points, methods and results of their scientific‐theoretical and methodological premisses (cf. E. 

Ströker, Einführung in die Wissenschaftstheorie [Darmstadt: Wissenschaftliche Buchgesellschaft, 2nd 

edn, 1977]; A. Menne, Einführung in die Methodologie [Darmstadt: Wissenschaftliche 

Buchgesellschaft, 2nd edn, 1984]; K. Eberhard, Einführung in die Erkenntnis‐und Wissenschaftstheorie 

[UrbTb, 386; Stuttgart: W. Kohlhammer, 1987]; and my review in LB 64 [1990], pp. 92–95). A good 

example of a semiotic approach is M. Tietzmann, Strukturale Textanalyse: Theorie und Praxis der 

Interpretation (UTB, 582; Munich: Fink, 1977). Too many articles uncritically adopt one model, 

following the slogan of F. Schiller: ‘Wie er räuspert und wie er spuckt, das habt ihr ihm glücklich 

abgeguckt’ (Wallensteins Lager [1798], scene 6; cf. J.W. Goethe, Faust [1790], act 1, scene 2, 

concerning the dog: ‘Ich finde nicht die Spur von einem Geist, und alles ist Dressur’; and ironically 

scene 3: ‘Am besten ist’s auch hier, wenn ihr nur einen hört und auf des Meisters Worte schwört’) as 

demonstrated alarmingly by D.C. Greenwood, Structuralism and Biblical Text (RR, 32; Berlin: Mouton, 

1985); cf. my consenting review in TLZ 112 (1987), cols. 21–22. 

2 Cf. E. Holenstein, Linguistik, Semiotik, Hermeneutik: Plädoyers für eine strukturale Phänomenologie 

(Frankfurt: Suhrkamp, 1976); idem, Von der Hintergehbarkeit der Sprache: Kognitive Unterlagen der 

Sprache (Frankfurt: Suhrkamp, 1980); C.S. Peirce, How to Make our Ideas Clear (ed. K. Oehler; 

Frankfurt: Klostermann, 3rd edn, 1985), and my review in LB 60 (1988), pp. 115–18, in concordance 

with the communicative ‘Organon‐Modell’ of K. Bühler, Die Axiomatik der Sprachwissenschaft 

(Frankfurt: Klostermann, 2nd edn, 1976 [1969]) (= KantSt 38 [1933], pp. 19–90); idem, Sprachtheorie: 

Die Darstellungsfunktion der Sprache (Stuttgart: Fischer, 2nd edn, 1965; 3rd edn, 1978 [1934]); cf. A. 

Eschbach (ed.), Bühler‐Studien I–II (Frankfurt: Suhrkamp, 1984); idem, ‘K. Bühlers Zeichenbegriff und 

seine Beziehung zu Wittgensteins Spätphilosophie’, Zeitschrift für Semiotik 6 (1984), pp. 397–420; 

R.E. Innis, ‘Die Überwindung der Assoziationspsychologie durch zeichentheoretische Analyse: James, 

Peirce, Husserl und Bühler’, Zeitschrift für Semiotik 10 (1988), pp. 331–55; S.E. Larsen, 

‘Phänomenologie und Interdisziplinarität: Brøndal versus Bühler—ein historisches Lehrstück’, 

Zeitschrift für Semiotik 11 (1989), pp. 353–62; E.‐O. Gerke, ‘In Memoriam K. Bühler [1897–1963]: 

Zeichenbedeutung beim Tier und beim Menschen‐Kritische Anmerkungen zu K. Bühlers 

Bedeutungstheorie’, Zeitschrift für Semiotik 15 (1993), pp. 413–30. All ancient insights into rhetoric 

belong to the primary level of communication, text‐pragmatics (relation: addressor/addressee; cf. my 

article ‘Sprachwissenschaft’, in TRE, forthcoming). 

136  

textlinguistic models has been detailed elsewhere.3 In this essay I apply the model to an analysis of Mark with respect to its communicative and rhetorical structures and functions.                                                             3 See my Naumburg Principal Address 1972, ‘Die Aufgaben der Exegese und die Mittel der Linguistik’, 

TLZ 98 (1973), cols. 881–94, and my ‘Wort Gottes zwischen Semantik und Pragmatik: Eine Anfrage 

der empirisch‐linguistischen Exegese an die existenzial‐hermeneutische Interpretation’, TLZ 100 

(1975), cols. 481–94; ‘Linguistik (Bibelwissenschaft)’, in Theologisches Lexikon (Berlin: Union, 1979), 

pp. 279–81 (2nd edn, 1981, pp. 345–47). It is treated more fully in Die Philipperbriefe des Paulus: 

Kommentar (Stuttgart: W. Kohlhammer, 1984), pp. 13–28; cf. my response to the reviews presented 

at the Seminar ‘The Role of the Reader’, SNTS Meeting, Trondheim 1985: ‘Die Rollen der Leser—

oder: Der Mythos des Lesers’, LB 60 (1988), pp. 61–84 (= ‘The Roles of the Reader or the Myth of The 

Reader’, Semeia 48 [1989], pp. 55–80), and taken up also in ‘Der Philipperbrief oder die 

Philipperbriefes des Paulus? Eine Entwort an V. Koperski’, ETL 70 (1994), pp. 122–31; my review of P. 

Wick, Der Philipperbrief (BWANT, 135; Stuttgart: W. Kohlhammer, 1994), in TLZ 120 (1995), cols. 

344–46; see also my ‘Der Brief des Paulus an Philemon in der neueren Forschung’, ANRW, II, 25.4 

(1987), pp. 3439–95; further suggestions for ways out of the ‘signifier‐trap’ appear in my Die Sprache 

des Matthäus: Die Text‐Konstituenten in ihren makro‐und mikrostrukturellen Relationen (Göttingen: 

Vandenhoeck & Ruprecht, 1987); cf. my applications of the theory in ‘Das Präsens historicum als 

makrosyntaktisches Gliederungsmerkmal im Matthäusevangelium’, NTS 22 (1976), pp. 464–75; ‘Das 

“Matthäusevangelium” als Petrusevangelium’, BZ 27 (1983), pp. 58–80; Kommentiertes Lexikon zum 

vierten Evangelium: Seine Textkonstituenten in ihren Syntagmen und Wortfeldern (Lewiston, NY: 

Edwin Mellen Press, 1993); ‘Die Umcodierungen der matthäischen Unser‐Vater‐Redaktion in Joh 17’, 

in A. Denaux (ed.), John and the Synoptics (BETL, 101; Leuven: Peeters, 1992), pp. 589–607; and 

‘Interne Strukturierungen im Schlußsegment Joh 21: Συγγραφή, Ἐπίλογος, Σατυρικόν’, NTS 38 

(1992), pp. 507–30. I also discussed it programmatically in ‘Was ist ein Kommentar?’, BZ 24 (1980), 

pp. 1–20; cf. special applications: ‘Text‐linguistische Aspekte der Strukturanalyse, dargestellt am 

Beispiel 1Kor 15, 1–11’, NTS 23 (1977), pp. 469–77 (French version in Semiotique et Bible 6 [1977], 

pp. 44–47); ‘Die makrosyntaktische Signalfunktion des lukanischen Textems ὑποστρεφειν’, StEv 7 (1982), pp. 444–50; ‘Textverarbeitung in Frühjudentum, Frühkirche und Gnosis’, in K.‐W. Tröger (ed.), 

AT Frühjudentum‐Gnosis (Berlin: Evangelische Verlagsanstalt, 1980), pp. 299–313; ‘Wird Mk auf der 

Couch materialistisch? Oder: Wie idealistisch ist die “materialistische Exegese”?’, LB 57 (1985), pp. 

95–106; ‘Die Paränese Hebr 13, 16 im Kontext des Hebräerbriefes: Eine Fallstudie semiotisch 

orientierter Textinterpretation und Sachkritik’, StudTheol 39 (1985), pp. 73–106 (cf. ‘Textlinguistik als 

Kommentierungsprinzip: Heb 4, 14–16’, NTS 26 [1979], pp. 242–52); ‘Code‐Wandel und christliche 

Identität: Der Kanon des NTs als semiotisches Problem’, LB 61 (1988), pp. 87–114; ‘Interpretatio 

Romana‐Interpretatio Graeca: Der hellenistische Synkretismus als semiotisches Problem’, in P. 

Schmitter and W.H. Schmitz (eds.), ‘Innovationen in Zeichentheorien’ (Münster: Nodus, 1989), pp. 

83–121; ‘Heiliger Geist—eine Simulations‐Kategorie?’, LB 66 (1992), pp. 5–38; ‘Offenbarung—ein 

Simulationskategorie? Offenbarung, die nichts offenbart’, in E. Güttgemanns (ed.), Das Phänomen 

der Simulation (FTL, 17; Bonn: Linguistica Biblica, 1991), pp. 107–42. A textlinguistic criticism of the 

synoptic ‘minor agreements’ is in ‘Zur Frage einer vierten Version der Seesturm‐Erzählung in einer 

Mt/Lk Agreement‐Redaktionsschicht (“Dt‐Mk”): Versuch einer textsemiotischen Geltungsprüfung von 

A. Fuchs’, in G. Strecker (ed.), Minor Agreements (GTA, 50; Göttingen: Vandenhoeck & Ruprecht, 

1993), pp. 93–118. The ecclesiological aspect of Mk 13 is explained in part II.3, ‘Markus der Lehrer, 

die Leser als seine Schüler und Jesus als sein Sprachrohr’, in ‘Die ältesten Selbstverständnisse 

christlicher Gruppen im ersten Jahrhundert’, ANRW, II, 26.2 (1995), pp. 1357–1467, esp. 1433–40. 

137  

As it is well known from the very beginning of this second biography of Jesus (if we count Q as being the first),4 Mark’s discourse clearly has four levels of communication, going step by step from the highest to the lowest stair or from the outermost room (or box) through a mediating room (argumentation) to the innermost (narration, including reported speech): ‘The text of the narrator dominates the dialogue of the characters’.5 In contrast to a one-dimensional ‘composition-analysis’ (even many narrative analyses remain on the syntagmatic level of text-sequence),6 we must establish not only a second dimension (text-semantics in their word-fields), but most of all a four-dimensional point of view, including text-pragmatics. From this perspective the text appears not only as a two-dimensional plain but a four-dimensional relief: ‘Narratology as a Semiotic Theory’.7 The embedded text-segments are thus treated as ‘hypo-texts’.8 For example,

• Mk 1:1 First and highest level (intratextual para-communicative9 opening): implied author as

sender of information to intended reader as receiver of the following information.10

                                                            4 F.G. Downing, ‘A Genre for Q and a Socio‐Cultural Context for Q’, JSNT 55 (1994), pp. 3–26. If Q can 

be defined as a biography of a philosophical teacher (like Lucian, Demonax; Downing, ‘Genre’, p. 24) 

then Mark can be further categorized as a biography of the philosophical teacher and his pupils (cf. 

R.A. Burridge, What Are the Gospels? [SNTSMS, 70; Cambridge: Cambridge University Press, 1992]; 

V.K. Robbins, Jesus the Teacher: A Socio‐Rhetorical Interpretation of Mark [Philadelphia: Fortress 

Press, 1984; Minneapolis: Fortress Press, 2nd edn, 1989]). 

5 M. Bal, On Meaning‐Making: Essays in Semiotics (Sonoma, CA: Polebridge Press, 1994); see my 

review in TLZ 120 (1995), p. 247; cf. her Narratologie (Paris: Klincksieck, 1977; ET Narratology: 

Introduction to the Theory of Narrative [Toronto: Toronto University Press, 1985]), and On Story‐

Telling: Essays in Narratology (Sonoma, CA: Polebridge Press, 1991). 

6 Cf. the overview and history of research: G.R. Beasley‐Murray, Jesus and the Future: An Examination 

of the Criticism of the Eschatological Discourse, Mk 13, with Special Reference to the Little Apocalypse 

Theory (London: Macmillan, 1954); R. Pesch, Naherwartungen: Tradition und Redaktion in Mk 13 

(KBANT; Düsseldorf: Patmos, 1968), pp. 74–82; K. Grayston, ‘The Study of Mk 13’, BJRL 56 (1974), pp. 

371–87; D. Wenham, ‘Recent Studies of Mk 13’, TSFBul 71 (1975), pp. 6–15; 72 (1975), pp. 1–9; E. 

Brandenburger, Mk 13 und die Apokalyptik (FRLANT, 134; Göttingen: Vandenhoeck & Ruprecht, 

1984), pp. 22–42; W.R. Telford, ‘The Pre‐Markan Tradition in Recent Research’, in F. van Segbroeck et 

al. (eds.), The Four Gospels (Festschrift F. Neirynck; BETL, 100; Leuven: Peeters, 1992), pp. 693–723, 

esp. 701–702. 

7 Bal, Meaning, pp. 21–76, cf. pp. 130–31, 211; cf. also methodologically, N.R. Petersen, Literary 

Criticism for New Testament Critics (Philadelphia: Fortress Press, 1978); idem, ‘When Is the End not 

the End? Literary Reflections on the Ending of Mk’s Narrative’, Int 34 (1980), pp. 141–66; and 

‘Literarkritik, The New Literary Criticism, and the Gospel According to Mk’, in van Segbroeck et al. 

(eds.), The Four Gospels, pp. 935–48. It is preferable to speak of ‘Textlinguistics’ instead of ‘discourse 

analysis’ (or ‘literary criticism’), which seems too generously—and that means too extensively—

defined. 

8 Bal, Meaning, pp. 254–55. 

9 The well‐established difference (by A. Tarski, R. Carnap and others) between the semantical stairs of 

‘object language’ and its ‘meta‐language’ in order to avoid semantic antinomies (T. Lewandowski, 

•

•

•

 

• Mk 1:2sender o

• Mk 1:3reader a

• Mk 1:7b

As iprimary

It is notwith theshall caanalysis               Linguisti

(analogo

(Lewand

Kompos

Essays in

hard to u

prefer to

eventua

semiotic

10 G. Arn

68 (1977

(1988), p

pp. 24–3

11 Cf. the

Sackgass

Testame

review in

What Is 

12  

2 Second eof argument3–7a Third as the spectab–8 Fourth

is well knowy level in wh

teworthy thae mere sepall it, ‘recepts’ with W.                       isches Wörte

ously on the 

dowski, Wört

ition des Ma

n Honor of L.

use when be

o use the pre

lly create the

c meta‐langu

old, ‘Mk 1.1 

7), pp. 123–2

pp. 3–20; cf. 

31, 58–64. 

e distinctions

se?’, in E.S. M

ent (JSNTSup

n TLZ 120 [19

Redaction Cr

embedded ltation to the(further) em

ator. (further) em

wn at the vehich the aut

2.

at post-Secoaration of ‘trtion-analysi Marxsen                       erbuch, I–III [

pragmatic le

terbuch, p. 4

atthäusevang

 Hartman [O

eing knowled

eposition παρe indispensa

uage, or, as it

und die Eröf

27; F. Matera

‘Substitution

s by J.R. Dona

Malbon and E

p, 109; Sheffi

995], cols. 62

riticism? (Ph

evel (interte intended rmbedded le

mbedded lev

ery end of thor speaks

The Structu

ond World Wradition’ anis’),11 but (win 195612 w                     Heidelberg: 

evel) betwee

47; D. Hellho

gelium’, in T.

Oslo: Scandin

dgeable of th

ρά: ‘The use ble if relativ

t is more app

ffnungswend

a, ‘The Prolog

n of Meta‐Eb

ahue, ‘Redac

E.V. McKnigh

eld: JSOT Pre

25–28), follo

iladelphia: F

textual metreader as theevel: implie

vel: Reporte

this second directly to t

ure of Mark

War researcnd ‘redactionwith regard who claim                      Quelle, 2nd 

en ‘communi

olm, ‘Substit

. Fornberg an

navian Unive

he Greek mea

of language 

e distance b

propriately c

dungen in gr

gue as Interp

bene’ (first d

ction Criticis

ht (eds.), The

ess, 1994), p

owing the dis

Fortress Pres

ta-narrative e receiver oed author a

ed speech o

biography,the addresse

k 12:41–13:

ch, or ‘Redan’ (or, betteto Mk 13) aed that Ma                      edn, 1976], 

cation’ and i

utionelle Gli

nd D. Hellho

rsity Press, 1

aning of the 

of another h

etween the 

alled, para‐la

riechischen u

pretative Key

degree) in He

m: Has the H

e New Litera

pp. 27–57, es

stinction out

s, 1969). 

start): impf conviction

as the narra

of a narrated

Mk 16:8b,ee.

:37

aktions-gescer, ‘editorialas a synchroark’s comp                      pp. 449, 482

ts ‘meta‐com

ederungsme

lm [eds.], Tex

1995], pp. 11

preposition 

hermeneutic

object‐langu

anguage’ (Ba

und lateinisch

y to Mark’s G

ellholm, ‘Glie

Hauptstrasse

ry Criticism a

pecially 29–3

lined by his t

plied authon. ator to the i

d figure.

, d finishes

chichte’, stal criticism’—onic ‘comp

position cam                      2–83) and 

mmunication

erkmale und 

exts and Cont

1–76, esp. 15

μετά. Insteac theory can 

uage and the

al, Meaning, 

hen Schrifte

Gospel’, JSNT

ederungsmer

e Become a 

and the New

–31, 31–34 (c

teacher N. P

138 

or as the

intended

with the

arted not —or as I ositional me from                   

n’ 

die 

texts: 

5–18) is 

ad I 

e 

p. 87). 

n’, ZNW 

T 34 

rkmale’, 

w 

cf. my 

errin, 

139  

                                                                                                                                                                                          W. Marxsen, Der Evangelist Markus: Studien zur Redaktionsgeschichte des Evangeliums (FRLANT, 67; 

Göttingen: Vandenhoeck & Ruprecht, 2nd edn, 1959 [1956]), pp. 101–28. The composition of Mk 13 

is treated in essentially the same synchronic way by E. Haenchen, Der Weg Jesu: Eine Erklärung des 

Markus‐Evangeliums und der kanonischen Parallelen (Berlin: W. de Gruyter; 2nd edn, 1968 [1966]), 

pp. 435–60; J. Lambrecht, Die Redaktion der Markus‐Apokalypse (AnBib, 28; Rome: Biblical Institute 

Press, 1967); J. Schreiber, Theologie des Vertrauens: Eine redaktionsgeschichtliche Untersuchung des 

Markusevangeliums (Hamburg: Furche, 1967), pp. 126–52; K. Tagawa, ‘Mk 13: La tâtonnement d’un 

homme réaliste éveillé face à la tradition apocalyptique’, FV 76 (1977), pp. 11–44; L. Schottroff, ‘Die 

Gegenwart in der Apokalyptik der synoptischen Evangelien’, in D. Hellholm (ed.), Apocalypticism in 

the Mediterranean World and the Near East: Proceedings of the International Colloquium on 

Apocalypticism Uppsala, August 12–17, 1979 (Tübingen: J.C.B. Mohr, 2nd edn, 1989 [1983]), pp. 707–

28; G.R. Beasley‐Murray, ‘Second Thoughts on the Composition of Mk 13’, NTS 29 (1983), pp. 414–20 

(p. 418: ‘it is difficult not to acknowledge the evangelist’s hand throughout Mk 13’). This line is picked 

up by D.E. Aune, Prophecy in Early Christianity and the Ancient Mediterranean World (Grand Rapids: 

Eerdmans, 1983), pp. 184–87, 399–400; idem, The New Testament in its Literary Environment (LEC, 8; 

Philadelphia: Westminster Press, 2nd edn, 1989 [1987]), pp. 238–40; Robbins, Teacher, pp. 43–45, 

171–79; J. Dupont, Les trois Apocalypses synoptiques (LD, 121; Paris: Cerf, 1985) (see the consenting 

review by T.F. Glasson, JTS 37 [1986], pp. 532–53); W.S. Vorster, ‘Literary Reflections on Mk 13, 5–37: 

A Narrated Speech of Jesus’, Neot 21 (1987), pp. 203–24; T.J. Geddert, Watchwords: Mk 13 in 

Markan Eschatology (JSNTSup, 26; Sheffield: JSOT Press, 1989) (cf. p. 18: the most assumed 

ambiguities are deliberately created by the ‘writer with a carefully considered and subtly 

communicated theological perspective’); C.C. Black, ‘An Oration at Olivet: Some Rhetorical 

Dimensions of Mk 13’, in D.F. Watson (ed.), Persuasive Artistry: Studies in New Testament Rhetoric in 

Honor of G.A. Kennedy (JSNTSup, 50; Sheffield: JSOT Press, 1991), pp. 66–92 (cf. idem, The Disciples 

According to Mark: Markan Redaction in Current Debate [JSNTSup, 27; Sheffield: JSOT Press, 1989]); 

Petersen, ‘Literarkritik’, pp. 943–47; A. Yarbro‐Collins, ‘The Eschatological Discourse of Mk 13’, in van 

Segbroeck et al., The Four Gospels, pp. 1125–40; J. Verheyden, ‘Persecution and Eschatology: Mk 

13.9–13’, in van Segbroeck et al. (eds.), The Four Gospels, pp. 1141–59. 

Concerning the materials used, J. Lambrecht (‘Die Logia‐Quelle von Mk 13’, Bib 47 [1966], pp. 321–

60) was the first to point to a reconstructed Q‐text here instead of only postulated oral ‘traditions’ 

(an initial stage which I welcomed: W. Schenk, ‘Der Einfluß der Logienquelle auf das 

Markusevangelium’, ZNW 70 [1979], pp. 141–65; Synopse zur Redenquelle der Evangelien 

[Düsseldorf: Patmos, 1981] and defended in TLZ 117 [1992], cols. 842–43, review of J. Schüling, 

Studien zum Verhältnis von Logienquelle und Markusevangelium [Würzburg: Echter Verlag, 1991]). I 

am even more convinced that Mark (together with Colossians) knew and used Paul in a way of ‘active 

reception’: ‘Sekundäre Jesuanisierungen von primären Paulus‐Aussagen bei Mk’, in van Segbroeck et 

al. (eds.), The Four Gospels, pp. 877–904. This view is exemplified in my Das biographische Ich‐Idiom 

Menschensohn in den frühen Jesus‐Biographien: Der Ausdruck, seine Codes und seine Rezeptionen in 

ihren Kontexten (FRLANT, 177; Göttingen: Vandenhoeck & Ruprecht, 1997) and ‘Die rhetorische 

Funktion der Fastenwarnung Mk 2, 20’, in W.L. Petersen et al. (eds.), Sayings of Jesus: Essays in 

Honour of T. Baarda (Leiden: E.J. Brill, 1997), pp. 251–76. 

In contrast, the ‘diachronical‐evolutionary’ approaches assume closed speech‐groups (E. Lohmeyer, 

Das Evangelium des Mk [KEK, 1.2; Göttingen: Vandenhoeck & Ruprecht, 8th edn, 1967 (1937)], pp. 

140  

distinctive materials. He stressed the possibility of pre-Markan texts that must be treated as secondary texts in their own right. The second major step forward was accomplished by C.C. Black in 1991 who focused on the ‘rhetorical dimensions of Mk 13’.13 In this paper I will go a step further and perform a semiotic analysis of this ‘hypo-text’. It is intended as a proposal for analysing the structure of other New Testament texts. In the following analysis the following symbols are used = Retrospective, = Markan connection, and ^ = Markan linking

                                                                                                                                                                                          285–86; V. Taylor, The Gospel According to St Mark [London: Macmillan, 2nd edn, 1966 (1952)], pp. 

636–44; W.G. Kümmel, Verheißung und Erfüllung [ATANT, 6; Zürich: Zwingli, 3rd edn, 1956], pp. 88–

97) or related complex traditions (L. Hartman, Prophecy Interpreted: The Formation of Some Jewish 

Apocalyptic Texts and of the Eschatological Discourse Mk 13 [ConBNT, 1; Lund: C.W.K. Gleerup, 

1966], a Christian Daniel‐Midrash, originated by Jesus). See D. Wenham, The Rediscovery of Jesus’ 

Eschatological Discourse (Sheffield: JSOT Press, 1984), who returned to an old ‘Ur‐Gospel’, 

independently used by the synoptics in order to re‐establish ipsissima‐vox‐Jesuanicity (critically 

reviewed by J.D.G. Dunn, JTS 38 [1987], pp. 163–66). Others assume, since the last century, a Jewish 

submission, the so‐called Apocalyptic Pamphlet: G. Hölscher, ‘Der Ursprung der Apokalypse Mk 13’, 

TBL 12 (1933), cols. 193–202; Pesch, Naherwartungen, pp. 207–23; E. Schweizer, Das Evangelium 

nach Markus (NTD, 1; Göttingen: Vandenhoeck & Ruprecht, 1967), pp. 149–63 (reduced at vv. 14–20 

with Kümmel, while vv. 5–13, 24–27 form another unit of tradition). This ‘Flugblatt’ is claimed to be 

Jewish‐Christian by H.J. Schoeps, ‘Ebionitische Apokalyptik im NT’, ZNW 51 (1960), pp. 101–11, 

followed by W. Grundmann, Das Evangelium nach Markus (THKNT, 2; Berlin: Evangelische 

Verlagsanstalt, 1959), p. 266 concerning vv. 14–27; F. Hahn, ‘Die Rede von der Parusie des 

Menschensohns Mk 13’, in R. Pesch and R. Schnackenburg (eds.), Jesus und der Menschensohn 

(Festschrift A. Vögtle; Freiburg: Herder, 1975), pp. 240–66, followed agreeingly by J. Gnilka, Das 

Evangelium nach Mk, II (EKKNT, 2.2; Neukirchen‐Vluyn: Neukirchener Verlag, 2nd edn, 1986 [1979]), 

pp. 211–12, and R. Pesch, Das Markusevangelium (HTKNT, 2; 2 vols.; Freiburg: Herder, 4th edn, 1991 

[1977]), II, pp. 266, 281–82, 287–88, 295–96, 300, 304–305, 311, and idem, ‘Mk 13’, in J. Lambrecht 

(ed.), L’Apocalypse johannique et l’Apocalyptique dans le NT (BETL, 53; Leuven: Peeters, 1980), pp. 

355–68 (revising his thesis of 1968 of a post‐Mark addition), which is criticized strongly and 

substantially by F. Neirynck, ‘Mc 13: Examen critique de l’interpretation de R. Pesch’, in Lambrecht 

(ed.), L’Apocalypse, pp. 369–401 (= Evangelica, I [BETL, 60; Leuven: Peeters, 1982], pp. 565–97), and 

underlined in his response to Pesch’s Naherwartungen in ‘Le discours anti‐apocalyptique de Mk 13’, 

in Lambrecht (ed.), L’Apocalypse, pp. 598–608 (= ETL 45 [1968], pp. 154–64); Brandenburger, Mk 13, 

pp. 43–73; C. Breytenbach, Nachfolge und Zukunftserwartung nach Mk (ATANT, 71; Zürich: Zwingli, 

1984), pp. 280–330; G. Theissen, Lokalkolorit und Zeitgeschichte in den Evangelien (NTOA, 8; 

Göttingen: Vandenhoeck & Ruprecht, 1989), pp. 249–50, 295 (renewing an early location of the 

assumed ‘apocalyptic broadsheet’ in the Caligula‐crisis). The ‘editorial approach’ 

(‘Traditionsgeschichte’) with the ‘broadsheet’‐hypothesis seems to be a rather germano‐centric 

phenomen. But ‘the theory of a “little apocalypse” underlying Mk 13 can no longer be regarded, 

without qualification, as a sententia recepta of synoptic criticism’ (Black, ‘Persecution’, p. 66 n. 3; 

Yarbro‐Collins, ‘Mk 13’, pp. 1129–32; Verheyden, ‘Mark 13’, pp. 1141–50). 

13 The most important insights for the arrangement are given by Black, ‘Mark 13’, p. 75, but cf. also 

the detailed descriptions of ‘Invention’ (pp. 78–82), and ‘Style’ (pp. 82–90). 

141  

Mk 12:41–13:37: The tripartite instruction (using three chreia patterns)14 of his pupils at the end of the third day in the Jewish Temple as the beginning of the separation-phase of the teacher-disciple relationship: Preparation of the students for independence from the teacher.

1. Mk 12:41–44: Short Teacher-Disciple-Chreia (indicated by change of place): Jesus Gives Warning/Example of Self-Sacrifice and the Obsolete Jewish Temple System (after 11:15 and 7:11–12 there is no ‘praise’ because it is not on behalf of the gospel as it is in 8:35; 9:41; 10:21).15

1.1. Expositio (A: change of place):

41 καὶ—καθίσας ^ (36; 11:2, 7) ^ *κατέναντι (11:2) τοῦ γαζοφυλακίου (2 Esdras 20:38; 22:44; Josephus, War 5.200)

ἐθεώρει (5:38) πῶς ^ ὁ ^ *ὄχλος (37) (βάλλει ^ *χαλκὸν (6:8) ^ εἰς (11:23; 2:22; 9:42–47) τὸ γαζοφυλάκιον (41a).

1.2. Propositio (B: interaction): The poor felt compelled to be like the wealthy:

καὶ πολλοὶ ^ (37) πλούσιοι (10:25; 2 Cor. 8:2, 9) ἔβαλλον (41b) ^ πολλά (41d) 42 καὶ—ἐλθοῦσα (18) μία χήρα (40!) πωχὴ (18, 21; 2 Cor. 8:4, 9) ἔβαλεν (41c) λεπτὰ δύο,

1.2.1. Meta-narrative commentary for the author’s Roman readers:

—ὅ ^ ἐστιν κοδράντης (Latin-Roman!)— 1.3. Probatio (C: new instructions for Jesus’ pupils):16

43 καὶ—*προσκαλεσάμενος (10:42) ^ τοὺς ^ μαθητὰς αὐτοῦ (11:14 Inclusio!)— εἶπεν αὐτοῖς,

                                                            14 K. Berger, Formgeschichte des NT (Heidelberg: Quelle, 1984), pp. 80–93 (p. 83: ‘Chrien sind keine 

alttestamentlich‐jüdische Gattung, sondern hellenistisch‐griechischen Ursprungs’): cf. idem, 

‘Hellenistische Gattungen und NT’, ANRW, II, 23.2 (1984), pp. 1031–432, esp. pp. 1092–110; R. Hock 

and E.N. O’Neil (eds.), The Chreia in Ancient Rhetoric, I (SBLTT, 27; Atlanta: Scholars Press, 1986); V.K. 

Robbins, ‘The Chreia’, in D.E. Aune (ed.), Greco‐Roman Literature and the New Testament (SBLSBS, 

21; Atlanta: Scholars Press, 1988), pp. 1–23; B.L. Mack and V.K. Robbins, Patterns of Persuasion in the 

Gospels (Sonoma, CA: Polebridge Press, 1989), pp. 1–29, 31–67. 

15 This text is mostly treated as a positive example which is unlikely in view of the preceding context 

Mk 12:38–40; contra E.S. Malbon, ‘The Poor Widow in Mk and her Poor Readers’, CBQ 53 (1991), pp. 

589–604, and idem, ‘The Major Importance of the Minor Characters in Mk’, in Malbon and McKnight 

(eds.), Criticism, pp. 58–86, 67, 76–81. 

^ Markan linking 

Retrospective 

16 This kind of three‐step progression (A, B, C, indicated by προσκαλεσάμενος τοὺς μαθητὰς αὐτοῦ) as a structuring pattern in Mark was detected by V.K. Robbins, ‘Summons and Outline in Mark: The 

Three‐Step Progression’, NovT 23 (1981), pp. 97–114; idem, ‘Mk 1.14–20: An Interpretation of 

Intersection of Jewish and Greco‐Roman Traditions’, NTS 28 (1982), pp. 220–36; idem, Teacher, pp. 

22–53. In contrast to his outline (Teacher, pp. 43–45) I wish to propose that Mk 13:1 is not a new 

start of a section, but continues the προσκαλεσάμενος τοὺς μαθητὰς αὐτοῦ heading of Mk 12:43. 

142  

Ἀμὴν (11:23; Q 12.37) λέγω (43a) ὑμῖν (11:33) ὅτι (11:23; 9:41) 1.3.1 Conclusio: Ironical evaluation:

ἡ χήρα (42a) αὕτη ἡ πτωχὴ (42a) πλεῖον (Q 12.23) πάντων (33) ἔβαλεν (42b) τῶν βαλλόντων (43d) εἰς τὸ γαζοφυλάκιον (42b)

1.3.2 Ratio (γὰρ):

44 πάντες ^ (43d) γὰρ ἐκ τοῦ περισσεύοντος (2 Cor. 8:14) ἔβαλον (43d), αὕτη (43d) δὲ ἐκ τῆς ὑστερήσεως (2 Cor. 8:14; Phil. 4:12) αὐτῆς ^ πάντα (44a) ὅσα εἶχεν (10:21; 2 Cor. 8:12) ἔβαλεν (44a),

ὅλον (30, 33) τὸν βίον (AnthGrJulÄgypt 6.25.6) αὐτῆς 2. Mk 13:1–2: Short Teacher-Disciple-Chreia (indicated by change of place): Present

brilliance and coming destitution of the Jewish National Sanctuary: 2.1. Foolishness of disciples during the final departure from the Jewish Temple:

1 καὶ—*ἐκπορευομένου (11:19) αὐτοῦ ἐκ τοῦ ἱεροῦ (35)— *λέγει αὐτῷ εἷς τῶν μαθητῶν αὐτοῦ (43),

a *∆ιδάσκαλε (14, 19, 32), *ἴδε ποταποὶ λίθοι (10; Q 13.34–35) b καὶ ποταπαὶ οἰκοδομαί (10; Q 13.34–35).17

2.2. Teacher’s Retort (oracle of the speedy destruction):

c 2 καὶ ὁ Ἰησοῦς (35—the first renominalization of the name) εἶπεν (1b) αὐτῷ b′ Βλέπεις (38) ταύτας τὰς μεγάλας (10:42–43) οἰκο δομάς; (Q 13:34–37) a′ οὐ ^ μὴ18 ἀφεθῇ (22) ὧδε (11:3) λίθος ἐπὶ λὶθον (Q 13.34–35) ὃς οὐ ^ μὴ

(2c) *καταλυθῇ 14:58; 15:29; 2 Cor. 5:1; Gal. 2:18). 3. Mk 13:3–37: Amplified Teacher-Disciple-Chreia (indicated by change of place:19

                                                            Markan connection 

17 This qualis‐argument in relation to destruction seems to be an allusion to the last words of Nero: 

qualis artifex pereo (Suetonius, Nero 49; cf. M. Grant, Roms Cäsarer [Munich: Beck, 1978], p. 214). 

18 ‘The use of the subjunctive or future for a prohibition or strong denials’ (Mk 9:1, 41; 10:15; 13:2, 

19, 30; 14:25, 31) is a classical, non‐Hellenistic feature of speech used to highlight the expression of 

Mark’s rhetorical Jesus, which is even adopted in LXX. ‘It conveyed not only emphasis but more 

important a solemn, biblical tone, especially suited to prophetic utterances and the speech of Jesus’ 

(J.A.L. Lee, ‘Some Features of the Speech of Jesus in Mk’s Gospel’, NovT 27 [1985], pp. 11–25, esp. 

18–19, putting classic expression into his mouth only parallel to God in LXX‐Pentateuch or in Virgil’s 

Aeneid, where ‘the speech of God is coloured by archaisms and other features adding 

impressiveness’ [p. 25; cf. J.A.L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch 

[Chico, CA: Scholars Press, 1983], pp. 123–24). 

19 There remains the issue of whether to treat this section as ‘apocalyptic’ or an ‘Apocalypse’. This 

artificial classification is a nineteenth‐century invention (cf. J.M. Schmidt, Die jüdische Apokalyptik: 

Die Geschichte ihrer Erforschung von den Anfängen bis zu den Textfunden von Qumran [Neukirchen‐

Vluyn: Neukirchener Verlag, 2nd edn, 1976]; K. Müller, Studien zur frühjüdischen Apokalyptik [SBA, 

11; Stuttgart: Bibelwerk, 1991], pp. 35–52, 195–200) and bears the landmarks of all categories which 

are, by definition, ‘extensional’ (e.g. ‘miracle’ or ‘piety’; cf. U. Forell, Wunderbegriffe und logische 

Analyse [FSÖTh, 17; Göttingen: Vandenhoeck & Ruprecht, 1967], pp. 25–48, on 

143  

                                                                                                                                                                                          ‘Adäquatheitsbedingungen für Definitionen’, and pp. 96–118 on extensional ‘unbestimmte 

Startpunkte der Begriffsbildung’ and ‘Sätze über Begriffe bei extensional unbestimmten Gedanken‐

subjekten’). If one says ‘apocalypse’ (significant) does one mean ‘apocalypse’ (signified), and if one 

means ‘apocalypse’ (semantically) does this adequately describe the literary empirical facts 

(sigmatically)? ‘Daß die Dinge viel differenzierter bedacht werden müssen als die existential 

verdünnten Wertungen von W. Schmithals, Die Apokalyptik: Einführung und Deutung (Göttingen: 

Vandenhoeck & Ruprecht, 1973) versteht sich von selbst’ (Müller, Apokalyptik, p. 213 n. 62; cf. also E. 

Schüssler Fiorenza, ‘The Phenomenon of Early Christian Apocalyptic: Some Reflections on Method’, in 

Hellholm [ed.], Apocalypticism, pp. 295–316, 303–305). The definition proposed by J.J. Collins 

(‘Apocalypse: The Morphology of a Genre’, Semeia 14 [1979], pp. 1–20 [9]) has been influential in the 

Apocalypse group of the SBL Genre Project: ‘Apocalypse is a genre of revelatory literature with a 

narrative framework, in which a revleation is mediated by an otherworldly being to a human 

recipient disclosing a transcendent reality which is both temporal, insofar as it envisages 

eschatological salvation, and spacial, insofar as it involves another, supernatural world’ (adopted by 

Schüssler Fiorenza, ‘Apocalyptic’, p. 296; Robbins, Teacher, pp. 167–68, though both concede that 

there is no otherworldly mediator in Mk 13 at all). This seems to me to be in part a regression from 

J.J. Collins’s better insights in his ‘Apocalyptic Eschatology as the Transcendence of Death’, CBQ 36 

(1974), pp. 21–43. First, it seems more appropriate to ‘avoid the use of the noun “Apocalypse” as 

though this represents a recognizable body of teaching’ (and even more of the derivated adjective 

‘apocalyptic’); ‘an “Apocalypse” is a literary type, a genre, and can embody teaching of varied kinds, 

just as parables and allegories are literary forms’ (T.F. Glasson, ‘The Temporary Messianic Kingdom 

and the Kingdom of God’, JTS 41 [1990], pp. 517–25, esp. p. 518 n. 2, in agreement with M.A. Knipp 

and J. Carmignac, ‘Les dangers de l’eschatologie’, NTS 17 [1971], pp. 365–90; idem, ‘Qu’est‐ce que 

l’Apocalyptique? Son emploi à Qumran’, RevQ 10 [1979/81], pp. 3–33; cf. already T.F. Glasson, 

‘Schweitzer’s Influence: Blessing or Bane?’, JTS 28 [1977], pp. 289–302; idem, ‘What is Apocalyptic?’, 

NTS 27 [1981], pp. 98–105; idem; ‘Theophany and Parousia’, NTS 34 [1988], pp. 269–70). Using the 

SBL‐Collins‐definition, we are left with the smallest common denominator. Only ‘aufgrund dieses 

sehr weiten Rahmens kam J.J. Collins zu der Definition … Die gemeinsame Implikation aller dieser 

Elemente sei “Transzendenz”: Die Art der Offenbarung erfordert einen transzendenten Mittler wie 

auch transzendente Objekte.—Diese Definition ist trotz ihrer (scheinbaren) Weite eine moderne 

Entscheidung und nicht ohne einen Hauch von Willkür. Wir gehen demgegenüber den umgekehrten 

Weg, indem wir nicht nach “der” Gattung Apokalypse fragen und demnach nicht versuchen, die 

Klassifikation seit der Mitte des 2. nachchristlichen Jahrhunderts nachzuvollziehen oder zu kritisieren, 

sondern indem wir fragen, welche literarischen Gattungen sich hinter den mit Apokalypse benannten 

Texten wirklich verbergen. Wir gehen also induktiv vor und setzen nicht voraus, daß Mk 13 parr., Lk 

17, 20ff und die ApkJoh zu einer Gattung gehören müssen oder auch nur je für sich als einheitliche 

Gattung konzipiert sein müssen. Dieser Weg ist weniger an allgemeinsten Strukturen (Collins), 

sondern stärker an der literarischen Gestalt orientiert’ (Berger, Formgeschichte, p. 296). The Uppsala‐

Colloquium 1979 was a symptom of methodological stagnation, so that ‘seit geraumer Zeit die 

Forschung auf der Stelle tritt. Dieser Still‐stand ist unlängst aufwendig dokumentiert worden’ 

(Hellholm, Apocalypticism). Weiterführend ist der Beitrag von H. Stegemann, ‘Die Bedeutung der 

Qumranfunde für die Erforschung der Apokalyptik’, in Hellholm, Apocalypticism, pp. 495–530’ 

(Müller, Apokalyptik, p. 15). Following the line of Carmignac he claims that ‘Bei “Apokalyptischer 

Literatur” muß es sich schon um ein regelrechtes Buch handeln, das speziell zu dem Zweck abgefaßt 

worden ist, “himmlisches Geheimwissen” bewußt als solches zu traktieren und es dennoch einem 

144  

                                                                                                                                                                                          bestimmten Leserkreis zu “offenbaren” ’ (p. 499). It is totally misleading to define ‘Apocalypticism’ as 

a daughter of ‘Eschatology’ or to speak of ‘apocalyptic Eschatology’. ‘Eschatology’ means ‘eine 

spezifische Ausprägung “heilsgeschichtlicher” Orientierung, die von der Zukunft eine “Wende zum 

Besseren” erhofft. In der Regel wird diese “Wende” bewirkt durch eine göttliches Eingreifen in den 

Verlauf der Dinge … Konstitutiv für “Eschatologie” ist allein, daß man eine solche Wende zum 

Besseren von einem künftigen Eingreifen Gottes erhofft’ (p. 500). Everybody has to bear in mind the 

categorical ‘Eigenständigkeit der Apokalyptik gegenüber der Eschatologie’ because: ‘Der für die 

“Apokalyptik” zentrale Gedanke, der literarisch die Apokalypsen konstituiert, ist nicht das—nur 

tatsächlich häufig benutzte—Geschichtskonzept der “Eschatologie”, sondern die Mitteilung von 

“himmlischen Geheimwissen”, das durch diese Schriften “offenbart” wird und gegebenenfalls eben 

auch den künftigen Verlauf der Geschichte klären mag’ (p. 501). Therefore, the incorrectly called 

‘Ten‐Weeks‐Apocalypse’ (1 En. 93.3–10 + 91.11–17) is a ‘Ten‐Weeks‐Instruction’ but not at all an 

‘apocalypse’. The definition of ‘Apocalypticism’ recently proposed by J. Maier, ‘Apokalyptik im 

Judentum’, in H. Althaus (ed.), Apokalyptik und Eschatologie: Sinn und Ziel der Geschichte (Freiburg: 

Herder), pp. 43–72, esp. p. 46, must be restricted to ‘Eschatology’ only: ‘Apokalyptik ist ein Symptom, 

das auftritt, sobald auf der Basis der deuteronomistisch‐eschatologischen Geschichtsbildes die 

Überzeugung Platz greift, die eigene Gegenwart sei jene entscheidende Phase, in der im 

menschlichen wie übermenschlichen Bereich eine Wende zur Periode des endgültigen Heils einsetzt, 

so daß alles darauf ankommt, wie Israel als die für den Geschichtsprozeß verantwortliche 

Erwählungsgemeinschaft sich verhält.’ Thus, probably the oldest known Apocalypse, the 

‘Astronomical Book’ recorded within the Enoch‐Pentateuch (1 En. 72–82; 4Q208–211) is not at all 

concerned with eschatology (Stegemann, ‘Bedeutung’, pp. 504–507; cf. idem, Die Essener, Qumran, 

Johannes der Täufer und Jesus [Freiburg: Herder, 3rd edn, 1994 (1993)], pp. 119, 133). The same 

must be said of the only Hebrew written Apocalypse, Jubilees (= ‘Books of Dividing up Times’, 2Q19–

20; 3Q5; 4Q176b, 216–24; 11Q12; cf. J.C. VanderKam, The Books of Jubilees: A Critical Text [Leuven: 

Peeters, 1989]), where God himself revealed the mysteries of the heavenly tablets to Moses for the 

rewritten Pentateuch: ‘Die Autorität Gottes selbst “himmlisches Wissen” und dessen “Offenbarung” 

in Form eines Buches sind die Konstituenda von “Apokalyptik”, was auch immer Gegenstand und 

Inhalt der “neuen Offenbarung” sein mag, in unserem Fall ein neuer Kalender mit alien detailliert 

dargelegten Konsequenzen und Implikationen’ (pp. 506–507, 509; cf. idem, Qumran, pp. 131–32, 

168, 172). The key‐term is ‘the secret’ (Aramaic raz, Greek μυστήριον) of dreams, scriptures, riddles, 

parables, and so on, which became now revealed (I. Willi‐Plein, ‘Das Geheimnis der Apokalyptik’, VT 

27 [1977], pp. 61–81). This term is totally absent in Mk 13, which is not at all an ‘Apocalypse’ (N. 

Walter, ‘Tempelzerstörung und synoptische Apokalypse’, ZNW 57 [1966], pp. 38–49). So it is 

meaningless if Mk 13 is treated as ‘anti‐apocalyptic’ (Pesch, Naherwartungen, p. 164) or ‘apocalyptic’ 

(so Brandenburger, Mk 13; his criteria are not at all decisive: [1] the teaching is neither ‘esoteric’ nor 

directed to an ‘esoterischen Zirkel’ [pp. 15, 164–65, 168, ‘Sondergruppe der Auserwählten’]; [2] the 

two initial questions in 13:4 are not typical ‘apokalyptische Schulfragen’; [3] the communicative 

relation is not between ‘der in die endzeitlichen Geheimnisse eingeweihte Lehrer’ and his 

‘einzuweihenden Schüler’ [pp. 19–20]; [4] the typical Markan phrase ἐν ^ ἑκείναις ^ ἡμέραις 13:17, 

[19–20], 24 is not as such an ‘apokalyptischer Terminus’ [pp. 5, 17–18, 23] as is revealed by Mk 1:9 

and 8:1; [5] the subordinating function of the oracles contradict the description of 13:5b–27 as 

‘Lehrmodus des eschatologisierten Geschichtsüberblicks’ [pp. 16–19, 164–65]). I regard this chapter 

to be eschatologically oriented but not an ‘apocalypse as such’ (Black, ‘Mk 13’, p. 67; C. Rowland, The 

145  

3.1 Mk 13:3–4: Propositio: A special learning-group20 which is alternative to the Jewish Temple (outermost inclusion A corresponds to A′ 13:23):

3 Καὶ—καθημένου (41) αὐτοῦ εἰς *τὸ ^ Ὄρος ^ τῶν ^ Ἐλαιῶν (11:1) *κατέναντι ^ τοῦ (11:2; 12:41) ἱεροῦ (1a)— *ἐπηρώτα (34) αὐτὸν *κατʼ ^ ἰδίαν (9:28)

*Πέτρος (11:21) ^ καὶ ^ Ἰάκωβος ^ καὶ ^ Ἰωάννης (10:41) καὶ Ἀνδρέας (3:18), 4 Εἶπὸν (2a) ἡμῖν πότε (9:19) ταῦτα ἔσται , καὶ τί τὸ σημεῖον

ὅταν μέλλῃ ταῦτα (4b) συντελεῖσθαι πάντα (44; LXX Dan. 12:7) 3.2. Mk 13:5–37 Argumentatio: Elaborated Response from Teacher (concerning the

Markan double-question of vv. 4c–d):21 5 ὁ δὲ Ἰησοῦς (2a—renominalization) *ἤρξατο ^ λέγειν (4a; 10:28, 32, 47) αὐτοῖς,

3.2.1. Mk 13:5b–23: Help for different dangers during the present time of the speaker (and

listeners)—disaster oracles: 3.2.1.1. Mk 13:5b–6 (outer-inclusion B corresponding to Section B′ 13:21–22): Danger of

deliberate attempts to mislead: (a) Exordium/subpropositio: expected reaction (imperative): βλέπετε ^ (2b)

^ μή (2; Gal. 5:15; 1 Cor. 8:9; 10:12) τις ὑμᾶς (43) πλανήσῃ (24, 27; 1 Cor. 15:33) b) Ratio (‘Asyndeton-causal’ reason): First I-oracle:

6 πολλοὶ (41)22 ἐλεύσονται (42) ἐπὶ τῷ ὀνόματί μου (9:37, 39, 38, 41)                                                                                                                                                                                           Open Heaven: A Study of Apocalypticism in Judaism and Early Christianity [New York: Crossroad, 

1982], pp. 9–72, 351–57]. 

20 This group of four disciples forms an inclusion with the beginning of Mk 1:16–20. Their mention is 

misunderstood if taken in the categories of ‘private’ vs. ‘public’ (e.g. Yarbro‐Collins, ‘Mk 13’, pp. 

1127–28) or ‘esoteric’ vs. ‘exoteric’ (e.g. Brandenburger, Mk 13, p. 15: ‘Schülerfrage eines 

esoterischen Kreises’, ‘esoterisches Schulgespräch’). It is most of all a special ‘learning‐group’ 

(advanced students) concerned thoroughly with the themes of ‘rejection’ and ‘vindication’, indicating 

that they will be known by the readers as martyrs (W. Stegemann, ‘Zur Rolle von Petrus, Jakobus und 

Johannes im Markusevangelium’, TZ 42 [1986], pp. 366–74). They are distinguished from the 

‘undergraduate learning‐group’ of the students (disciples) and also from the mere ‘beginners’ 

(ὄχλος). 

21 ‘Mk 13:5b–37 appears to exhibit general conformity with, as well as creative adaption of, this 

typical’ classical taxis/dispositio (Black, ‘Mk 13’, p. 71); but I find Black’s claim unconvincing and, 

furthermore, in contradiction to this is his misleading thesis: ‘There is a chiastic relationship between 

the implied objects of the proemium (the questions in V. 4) and the kinds of topics treated in the 

narratio (VV. 6–36). The disciples’ second query, τί τὸ σημεῖον, is addressed by roughly the first three‐quarters of the narratio (VV. 6–27); the remainder, VV. 28–36, takes up their first question, 

πότε ταῦτα ἔσται (pp. 74–5).’ The (1) stasis/conquaestio‐character (an sit, quid sit, quale sit) of Mark’s 

two‐part questions of students for more facts, (2) their definition, (3) their interpretation (p. 73), and 

the understanding of the nature of the chreia contradict the proposed question‐answer 

machrochiasm. 

146  

λέγοντες ^ ὅτι Ἐγώ εἰμι, καὶ πολλοὺς (6a) πλανήσουσιν (5c).

3.2.1.2. Mk 13:7–8 (inner inclusion C corresponding to section C′ (13:14–20): Dangers of

ongoing historical disasters: (a) Exordium/subpropositio: Expected reaction (imperative) of those who are well informed (ὅταν

as protasis of the main sentence): 7 ὅταν (4d) δὲ ἀκούσητε ^23 πολέμους καὶ ^ ἀκοὰς πολέμων, μὴ (5) θροεῖσθε24

(b1) First meta-narrative commentary: —δεῖ (9:11) γενέσθαι (10–11; LXX Dan. 2:28–29, 45), *ἀλλʼ (27) *οὔπω (8:17, 21) τὸ τέλος (4d)— (c) Ratio (reason): Second I-oracle:

ἐγερθήσεται (25; Phil. 1:17) γὰρ (44) ἔθνος ἐπʼ ἔθνος (10:33, 42; 11:17; 2 Chron. 15:6) καὶ βασιλεία (3:24; 6:23; Q 11.17–18) ἐπὶ βασιλείαν, ἔσονται (4b) σεισμοὶ κατὰ τόπους ἔσονται (8b) λιμοί

                                                                                                                                                                                          22 This may be a rhetorical hyperbaton. Petersen (‘Literarkritik’, pp. 944–45 n. 13) treats ‘the “many” 

messianic claimants as a function of Mk’s rhetorical strategy for dealing with a single claimant 

supported by one or more prophets … Mk was dealing with a single would‐be messiah, a “Christian” 

(“in my name”) among “Christians” (a group of the “elect”), and that Mk’s use of the plural a) 

relativized his opponent’s claim while at the same time b) enabling his readers to say: Ah, we have 

one of those whom Jesus talked about.’ 

23 This implies that the addressees are distant (cf. Phil. 1:27, 30; 2:26; 4:9) from the locale of these 

wars—contrary to Mk 13:14. The ὅταν‐protasis with subjunctive points to an iterative use, which is 

indicated by the double plural‐object (F. Blass, A. Debrunner and F. Rehkopf, Grammatik des 

neutestamentliche Griechisch [Göttingen: Vandenhoeck & Ruprecht, 14th edn, 1976], §373.3). ὅταν with subjunctive is indefinite, but because of Mark’s style of using ὅταν with indicative (3:11; 11:19, 

25), the eventualis must be expressed in a translation (‘would …’). So in 13:11, 28–29 the plurals also 

point to iterative events, while 13:14 is a singular event and the ὅταν‐protasis is followed by an active reaction‐apodosis, which shows that the ὅταν‐protasis in this case is conditional. This cannot be forced into a so‐called ὅταν‐‘Gliederungsstruktur’ (so Hahn, ‘Mk 13’, pp. 245–47; countered by 

Brandenburger, Mk 13, pp. 36–39, who recalls rightly on p. 38 n. 69 the quotation by Lambrecht, 

Redaktion, p. 198: ‘hatte mit Reecht herausgestellt, daß sich das ὅταν ἴδητε in V. 14 and 29 jeweils 

durch verschiendenen Nachsätze und Objekte unterscheide ein bewußter Bezug also 

unwahrscheinlich sei’). 

24 This first ‘call to attention’ (13:5, 9, 23, 33) also implies the notion of ‘do not panic’: θροεῖσθαι als 

“Anfregung” angesichts des nahen Endes wie 2 Thess 2:2’ (Schottroff, ‘Gegenwart’, pp. 710–11 n. 8, 

p. 720). Its antonym is the word‐field of positive reaction: ὑπομένειν 13:13 = ἀγρυπνεῖν 13:13 = γρηγορεῖν 13:34–35, 37 (cf. Geddert, Watchwords). 

147  

(b2) Second meta-narrative commentary: —*ἀρχὴ ^ (10:6) ὠδίνων (Gal. 4:19, 27) ταῦτα (4d)—

3.2.1.3. Mk 13:9–13 (central section D): Dangers of ongoing persecution:25

(a) Exordium/subpropositio: Expected reaction (imperative): 9* βλέπετε (5b, 12:38) δὲ ὑμεῖς (5c)

ἑαυτούς (Phil. 3:2a, b, c: acc.) b) Ratio (‘Asyndeton-causal’ reason): Third I-Oracle:

a παραδώσουσιν (10:33) ὑμᾶς (9a; 2 Cor. 4:11) b εἰς συνέδρια b′ καὶ εἰς συναγωγὰς (39; Q 12.11) a′ δαρήσεσθε (12:3, 5; 2 Cor. 11:20) b″ καὶ ἐπὶ ἡγεμόνων καὶ βασιλέων (6:14–27) a″ σταθήσεσθε (11:5)

ἕνεκεν ^ (8:35; 10:29) ἐμοῦ (6; Qmt 6.22) εἰς ^ μαρτύριον ^ αὑτοῖς (1:44; 6:11) (c1) Third meta-narrative commentary:

—10 καὶ εἰς πάντα (43f) τὰ ἔθνη (8; 11:17; Gal. 3:8, 14; Rom. 1:5) πρῶτον δεῖ (7c) *κηρυχθῆναι *τὸ ^ εὐαγγέλιον (1:14; 10:29; Gal. 2:2)— (a′) Expected reaction (double-imperative) of those well informed (ὅταν as protasis of the main sentence):

11 καὶ ὅταν (7) ἄγωσιν (1:38) ὑμᾶς (9b) παραδιδόντες (9b), μὴ ^(7) προμεριμνᾶτε (Q 12.11)

τί λαλήσητε (11:23 + acc. pupils; 12:1; 2 Cor. 4:13) ^*ἀλλʼ (7d) ὃ ^ ἐὰν δοθῇ (14) ὑμῖν (9a) ἐν ἐκείνῃ τῇ ὥρᾳ (Q 7.21; 12.12) ^ τοῦτο (Markan inversio) λαλεῖτε (11b), (c′) Fourth meta-narrative commentary:

—οὐ ^ (5) γάρ (8a) ἐστε ὑμεῖς (11d; as 9a) οἱ λαλοῦντες (11d) ^ *ἀλλὰ (11d) τὸ πνεῦμα (36; 2 Cor. 4:13) τὸ ἅγιον (Q 12:12)— (b′) Ratio (same function as the ὅταν protasis vv. 7, 11, 14): Fourth I-Oracle:

a 12 καὶ παραδώσει (11a) ἀδελφὸς (19–20) ἀδελφὸν εἰς θάνατον ^ (10:33; 2 Cor. 4:11)

b καὶ πατὴρ (10:29; Q 12.53) τέκνον (19; 10:29–30 , b′ καὶ ἐπαναστήσονται ^ τέκνα (12a) ^ ἐπὶ (LXX Mic. 7:6) γονεῖς (Rom. 1:30) a′ καὶ ^ θανατώσουσιν (Rom. 8:36) αὐτούς 13 καὶ ἔσεσθε (8b, c) μισούμενοι ὑπὸ πάντων (10) διὰ (2 Cor. 4:11) τὸ ὄνομά (6) μου (6,

9).

                                                            25 This section is well treated: J. Dupont, ‘Le persécution comme situation missionare’, in J. Wanke et 

al. (eds.), Die Kirche des Anfangs (Festschrift H. Schürmann; Leipzig: Benno, 1977), pp. 97–114; 

Schottroff, ‘Gegenwart’; R. Kühschelm, Jüngerverfolgung und Geschick Jesu: Untersuchung der 

synoptischen Verfolgungsankündigungen Mk 13.9–13 par. und Mt 23.29–36 par. (ÖBS, 5; 

Klosterneuburg: Bibelwerk, 1983); M. Hooker, ‘Trial and Tribulation in Mk 13’, BJRL 65 (1982–83), pp. 

78–99; G. Dautzenberg, ‘Das Wort von der weltweiten Verkündigung des Evangeliums (Mk 13.10) 

und seine Vorgeschichte’, in K. Kertelge (ed.), Christus bezeugen (Festschrift W. Trilling; Leipzig: 

Benno, 1989), pp. 150–64; Verheyden, ‘Persecution’. 

148  

(a″) Conclusio: Expected reaction (conditional instead of imperative) of those well informed:

ὁ ^ δὲ ὑπομείνας (Rom. 12:12) εἰς τέλος (7d) ^ οὗτος (Markan inversio) σωθήσεται (8:35; 10:26, 52; LXX Joel 3:5; Rom. 10:13, 1 Cor. 5:5).

3.2.1.4. Mk 13:14–20 (inner inclusion C′ corresponding to C 13:7–8): Dangers of historical disasters (final eradication of the Jewish national sanctuary as the new ‘Sodom’) during the present time of the author/reader: (a) Exordium/subpropositio (ὅταν as protasis of the main sentence):

14 ^ Οταν (11a) δὲ ἴδητε (1b)26 τὸ βδέλυγμα τῆς ἐρημώσεως (LXX Dan. 12:11) ἑστηκότα (9; pf. masc.!)

* ὅπου (9:48) οὐ (11) δεῖ (10), (b) Imperative of the fifth meta-narrative commentary directly addressed to the readers (parenthesis within a speech to the hearers):

—ὁ ἀναγινώσκων (12:10, 26; 2 Cor. 1:13; 3:2; 1 Thess. 5:27) νοείτω (7:18; 8:17; Rom. 1:20)— (c) Expected reaction of a group, not addressed directly27 to readers (‘women’) and spoken from afar:

τότε (3:27) οἱ ἐν τῇ Ἰουδαίᾳ (10:1; Rom. 15:31) φευγέτωσαν (5:14; → 16:8) εἰς τὰ ὄρη (LXX Gen. 19:17) (c′) Exemplificatio: 15 ὁ [δὲ] ἐπὶ τοῦ δώματος μὴ (11) καταβάτω (9:9) μηδὲ εἰσελθάτω (11:11, 15) *ἆραί (11:23) τι ἐκ τῆς οἰκίας (40) αὐτοῦ 16καὶ ὁ εἰς τὸν ἀγρὸν (11:8) μὴ (15) ἐπιστρεψάτω (8:33) εἰς τὰ ὀπίσω (LXX Gen. 19:17) *αἶραι (15b) τὸ *ἱμάτιον (11:7–8) αὐτοῦ. (d) Sixth meta-narrative commentary:

—17 οὐαὶ (Q 10.13; cf. 4 Ezra 13:16 as signal of commentary28 ^ δὲ ταῖς ἐν γαστρὶ *ἐχούσαις (1 Thess. 5:3)

                                                            26 This means the addressees are near, as in 13:1, 29 (cf. Phil. 1:27, 30; 2:26, 28; 4:9), the place of this 

war—contrary to Mk 13:7: ‘Der Unterschied zwischen den hörbaren und den sichtbaren Ereignissen 

in V. 7f und v. 14f ist sehr genau zu beachten, ebenso die Tatsache, daß V.28f nur auf die sichtbaren 

Ereignisse Bezug nimmt’ (Hahn, ‘Mk 13’, p. 26 n. 29). 

27 This is noted by Brandenburger, Mk 13, p. 69 (‘unpersönliche Formulierung’), but not treated in its 

synchronic function. He too rashly sees here an apocalyptic pre‐Markan text. 

28 The ‘parallels’ are often treated only as ‘morphemes’, that is ‘morphologically/syntagmatically’ as 

‘significants’ (in cases of ‘intertextuality’ as direct ‘reception’), sometimes as ‘signifieds’ (‘monemes’), 

but seldom in terms of their communicative (text‐pragmatic) functions in the original texts. We must 

bear in mind, ‘Auch die Funktion der Kontexte wird von verschiedenen Sprach‐, Kunst‐und 

Interpretationsverständnissen bestimmt. Die traditionelle Hermeneutik der Altphilologie und der 

Theologie versuchte mittels Parallelstellen‐Methode Text‐gleichungen herzustellen. Peter Szondi 

(1929–1971) hat in seinem Traktat “Über philologische Erkenntnis” ’ (1962; cf. idem, Schriften, II 

[Frankfurt: Suhrkamp, 1978]) ‘mit guten Gründen betont, daß Parallelstellen nicht Beweise, sondern 

149  

καὶ ταῖς θηλαζούσαις *ἐν ^ ἐκείναις ^ ταῖς ἡμέραις (8:1)— (b′) Expected reaction (imperative) of the informed addressees: Intercession for the women refugees from Judaea:29 18 προσεύχεσθε (40) ^ δὲ ^ *ἵνα (19)

μὴ γένηται (7) χειμῶνος (d) Ratio: Seventh meta-narrative commentary (of an author who knows the notion of running out of time in the present days from 1 Cor. 7:29, 31):

—19 ἔσονται (8b, c, 13) γὰρ (11) αἱ ἡμέραι ἐκεῖναι (17) θλῖψις (4:17) a *οἵα (9:3) οὐ (14) γέγονεν (18b) *τοιαύτη (10:14)b ἀπʼ ^ *ἀρχῆς (8) ^ κτίσεως ^ (10:6) c ἣν ^ ἔκτισεν ὁ θεὸς (34; here in centre; synonyms: 10:32) b′ ἕως (LXX Dan. 12:1) τοῦ νῦν (10:30; Phil. 1:5: time of author!)a′ καί οὐ ^ μὴ (530) γένηται (19b).a 20 καὶ εἰ μὴ *ἐκολόβωσεν31 ^ κύριος (36) ^ τὰς ^ ἡμέρας (19), b οὐκ ^ (19) ἂν32 ἐσώθη (13) πᾶσα (10, 13) σάρξ (Gal. 2:16; Rom. 3:20).b′ ^ *ἀλλὰ (11b) διά τοὺς ἑκλεκτοὺς (Rom. 8:33)

οὓς ἐξελέξατο (1 Cor. 1:27–28)a′ *ἐκολόβωσεν ^ (20b) ^ τὰς ^ ἡμέρας (20a)—

3.2.1.5. Mk 13:21–22 (outer-inclusion B′ corresponding to section B 13:5c–6): Dangers of

deliberate attempts to mislead:

                                                                                                                                                                                          bestenfalls Hinweise sein können’ (P. Rusterholz, ‘Über die [Un‐] Interpretierbarkeit literarischer 

Texte’, Zeitschrift für Semiotik 13 [1993], pp. 303–17, esp. 313). 

29 The break between v. 17 and v. 18 in E. Nestle and K. Aland, Novum Testamentum Graece 

(Stuttgart: Bibelstiftung, 26th edn, 1979), p. 134, is misleading. The motif of intercession for a third 

group in Mk 13:18 is overlooked, because the Markan text is interpreted ‘silently’ from the Matthean 

editorial addition ‘your flight’. 

30 Classical feature in speech to highlight the expression of Mark’s rhetorical Jesus (Lee, ‘Features’, 

pp. 18–23). 

31 ‘Der Indicative‐Aorist ἐκολόβωσεν blickt zurück, V. 20 ist ein vaticinium ex eventu, aus der Realität 

genommen und nicht aus apokalyptischer Tradition … Gott hat wegen der Auserwählten die Tage der 

Fluchtkatastrophe in Judäa verkürzt, sonst wäre die Bevölkerung restlos umgekommen’ (Schottroff, 

‘Gegenwart’, p. 708). This verb is used for defective stones (W. Bauer, K. Aland and B. Aland, 

Griechisch‐Deutsches Wörterbuch zu den Schriften des NT [Berlin: W. de Gruyter, 6th edn, 1988], col. 

898) but is used as an expression of time even in 3 Bar. 9.7 (Apocalypse of Baruch [Greek], second 

century in view of its dependence on 2 Baruch) but in a non‐eschatological sense concerning the 

waxing and waning moon: ‘Bei der Übertretung des ersten Adam war er bei Samael als er die 

Schlange als sein Kleid anzog. Er (der Mond) verbarg sich nicht, sondern vergrößerte sich, und es 

erzürnte Gott über ihn, drückte ihn zusammen und verkürzte sene Tage’ (W. Hage, ‘Die griechische 

Baruch‐Apokalypse’, JSHRZ 5 [1979], pp. 15–44, esp. 30). 

32 Classical expression in speech used to highlight the statement of Mark’s rhetorical Jesus: with 

indicative in the apodosis of an ‘unreal’ condition (Lee, ‘Features’, p. 23). 

150  

(a) Expected reaction (disbelief) of informed persons (its protasis is functionally equivalent to ὅταν):

21 καὶ τότε (14b) ἐάν ^ τις ^ ὑμῖν ^ (11) εἴπῃ (11:3; 1 Cor. 10:28), *Ἴδε (1c) ὧδε (2c) ὁ Χριστός ^ (35, *Ἴδε ( 21b) ἐκεῖ (11:5)

μὴ (16) πιστεύετε (11:23–24, 31; Q 17.23) (b) Ratio: Fifth oracle (reason):

22 ἐγερθήσονται (8a) γὰρ (19a) ψευδό (10:19: LXX Deut. 5:20)—^ χριστοι (21b) καὶ ψευδο (22a)—προ (11b)—φῆται33

καὶ δώσουσιν (11) σημεῖα καὶ τέρατα πρὸς τὸ ἀπο πλανᾶν (5f Inclusio; LXX Deut. 13:2, 6),

—εἰ ^ *δυνατόν (Gal. 4:18; Rom. 12:18; parenthetic authorial commentary),—τοὺς ἐκλεκτούς (20c Inclusio).

3.2.1.6. Mk 13:23: I-Conclusio: the double-question is answered (outermost inclusion A′

corresponding to A 13:4a, 5a–b—with asyndeton-causal as apodosis, formulated in the perfect tense from the author’s perspective):

23 ὑμεῖς (21; as in 9, 11 redundancy increased) δὲ *βλέπετε (9a, Inclusio to 5b) προ (22a)—είρηκα (22a; Inclusio to 4a, 5a) ὑμῖν πάντα (4a Inclusio).

3.2.2. Mk 13:24–27: Excursus (‘salvific’ verb of 13:13 as the ‘providence’ verb of 13:20)

eighth authorial speech: oracle of salvation for the future of the author/readers: the imminent gathering and formation of the community from total darkness to absolute light:34 a 24 ἀλλὰ (20) *ἐν ^ ἐκείναις ^ ταῖς ^ ἡμέραις (19–20, 17)

μετὰ ^ τὴν θλῖψιν (19) ἐκείνηνb(-) ὁ ἥλιος (1 Cor. 15:41) σκοτισθήσεται (Rom. 1:21; 11:10),

καὶ ἡ σελήνη (1 Cor. 15:41) οὐ (20) δώσει (22) τὸ φέγγος35 αὐτῆς, c(-) 25 καὶ οἱ ἀστέρες (1 Cor. 15:41; LXX Isa. 13:10) ἔσονται (8b, c, 19)36

                                                            33 The first root has a passive stem (to be anointed by God) and the second an active as its 

complement. As in prior traditions, the prophets are God’s anointed ones who have his Spirit (1QM 

11.7–8; CD 2.12–13; 6.1; 4Q521 frg. 2 col. 2.1; frg. 8.9; fr. 9.3); in subsequent traditions, Eph. 3:5 

contains a comparable expression (with only one article for both terms; cf. 2:20); ‘apostles’ with the 

passive voice (an apostle has been sent) complemented by the active voice: ‘prophets’ as the 

revealer of God’s plan. 

34 ‘Wenn die natürlichen Lichtquellen erlöschen und es finster wird auf Erden, wird die Ankunft des 

Menschensohns mit um so wirkungsvollerem Glanz umgeben sein’ (G. Wohlenberg, Das Evangelium 

des Mk [KNT, 2; Leipzig: Deichert, 3rd edn, 1930], p. 335). This view is treated as ‘abwegige 

Überlegungen’ by Pesch (Naherwartungen, p. 161 n. 653), who misses the ring‐composition, 

assuming that in all cases here we have to assume ‘kosmische Gerichts‐Methaphern’ (he falls into the 

significant‐trap). To the contrary, the function of this excursus is evident: as 13:26–27 expounds the 

foregoing salvific verb σωθήσεται in 13:14, 20, so also Mk 13:24–25 expounds the phrase 

ἐκολόβωσεν τὰς ἡμέρας in 13:20a, d linked with the salvific act: God’s plan is to cut short the days 

(and nights). 

35 Used for the ‘moon’: Ps.‐Xenophon, Cyneget 5.4; Philo, Som. 1.23 (in others for the ‘sun’, LXX Joel 

2:10 = 4:15 for the ‘stars’; Bauer, Aland and Aland, Wörterbuch, col. 1704). 

151  

ἐκ τοῦ οὐρανοῦ (12:25) πίπτοντες (9:20; LXX Isa. 34:4

καὶ αἱ δυνάμεις (22: 12:24) αἱ ἐν τοῖς οὐρανοῖς (25a; plur. 12:25) σαλευθήσονται (Q 7:24).37

                                                                                                                                                                                          36 ‘Non‐Jewish readers in the Roman empire Mk 13:24–25 would not refer to a cosmic catastrophe 

and the end of the world, but to the end of pseudo‐gods of the Greco‐Roman pantheon’ (B.M.F. van 

Iersel, ‘The Sun, Moon, and Stars of Mk 13:24–25 in a Graeco‐Roman Reading’, paper at the SNTS 

Congress, Prague, 1995). In Rome ‘Sol’ was associated with Jupiter as ‘Luna’ was with ‘Iuno’ or 

‘Diana’ (G. Radke, Zur Entwicklung der Gottesvorstellung und der Gottesverehrung in Rom 

[Darmstadt: Buchgesellschaft, 1987], pp. 129–30). ‘Auf einem Giebelfries sprengt die Himmelsgöttin 

“Urania/Caelestis” auf einem Löwen (dem Sternbild des Hochsommes) dahin, wahrend ihr der 

Sonnengott “Sol/Apollo” auf seiner Quadriga folgt und im Rücken der Göttin ein Stern leuchtet’ (E. 

Simon, Die Götter der Römer [Darmstadt: Buchgesellschaft, 1990], pp. 103–104). ‘Die Antefixe 

bringen ja keine fortlaufende mythische Erzählung, sondern weisen auf verschiedene Aspekte der 

Hauptgöttin hin’ (pp. 156–57). During the Roman ‘triumphus’ of commanders and emperors, the 

triumphal robe (ornatus Iovis) was taken from the statue of the god in the temple of Jupiter 

Capitolinus and the one who had triumphed was dressed with it (H.P. L’Orange, ‘Sol Invictus 

Imperator: Ein Beitrag zur Apotheose’, SO 14 [1935], pp. 86–114) as was the case with Vespasian and 

Titus after the end of the Jewish War (Josephus, War 7.123–57 at the climax of this book!). It is very 

likely that Mk 15 designates the cross‐road of Jesus (maybe as a direct response to Josephus) as an 

exaltation and ‘anti‐triumph’ against the background of this Roman triumphal procession of 

Vespasian and Titus (T.E. Schmidt, ‘Mk 15:16–32: The Crucifixion Narrative and the Roman Triumphal 

Procession’, NTS 41 [1995], pp. 1–18); so it is also likely that Mk 13:24–25 uses such language in this 

manner. There may be a direct connection in Mk 13:9–13 regarding the persecutions caused by 

conflicts with the Roman ‘state religion’ (Schottroff, ‘Gegenwart’, pp. 715–20, quotation p. 719) and 

the quotation of Pliny, Nat. Hist. 2:18–19: ‘Gott zu sein, bedeutet für den Sterblichen, dem 

Sterblichen zu helfen, und dies ist der Weg zum ewigen Ruhm. Ihn gingen die vornehmsten Römer, 

auf ihm wandelt jetzt göttlichen Schrittes zusammen mit seinen Kindern der größte Herrscher aller 

Zeiten, Vespasianus Augustus … Dies ist die älteste Sitte … daß man solche Helfer unter die Götter 

versetzt.’ Already the third triumph of Demetrios Poliorketes in Athens 291/290 BCE is recorded: ‘Er 

weilt heiter, wie es dem Gotte ziemt, und schön und lachend unter uns. Ein erhabener Anblick: die 

Freunde alle im Kreise, in ihrer Mitte er selbst, den Sternen gleich die Freunde, doch der Sonne er 

selbst. Du Sohn des mächtigsten Gottes …’ (FGrHist, 76 frg. 13; translated in Umwelt des 

Urchristentums, II [Berlin: Evangelische Verlagsanstalt, 1967], p. 103; cf. C. Habicht, Athen: Die 

Geschichte der Stadt in hellenistischer Zeit [Munich: Beck, 1995], pp. 98–99). 

37 Both plurals in this fourth sentence may indicate a summary of the first three sentences. It may be 

a concluding consecutive (or, less likely, ‘causal’) construction (if it does not refer to powers like 

lightning, hail, etc., or alternatively to fixed stars, if the preceding context is concerned with 

wandering planets, which is well known and documented since Sumerian culture in 2000 BCE; see H. 

Schmökel, Kulturgeschichte des Alten Orient [Stuttgart: Kröner, 1971], pp. 195–98). The use of 

δυνάμε[ι/εω]ς in a ring‐composition in the elements c: c′ signals this correlation: ‘Daß sie erschüttert 

werden, zeigt an, daß eine ihnen überlegene Kraft im Himmel zu wirken beginnt’ (F. Hauck, Das 

Evangelium des Mk [Leipzig: Deichert, 1931], p. 159), which is not misleading as is Pesch’s 

(Naherwartungen, p. 162 n. 653) view of cosmic‐metaphorical expressions of judgment). The Markan 

concept is unlike the ‘Ten‐Epochs‐Instruction’, 1 En. 93:1–10; 91:11–17 (which is not an ‘apocalypse’; 

152  

d(+) 26 καὶ τότε (21a) ὄψονται (9:4; Q 13.38)38

τὸν υἱὸν ἀνθρώπου (10:45) ἐρχόμενον (contra 6; Q 12.40) ἐν νεφέλαις (9:9; LXX Dan. 7:13)

c′ μετὰ δυνάμεως (25b) πολλῆς (6) b′(+) καὶ δόξης (8:38; 10:37; 1 Cor. 15:41; LXX Dan. 7:14) a′(+) 27 καὶ τότε (26a) ἀποστελεῖ (12:13; Subj. 3:14 + 6, 7 = Q 10:3)

τοὺς ἀγγέλους (12:24; cf. 1:13; 8:38; Q 7.27) καὶ ἐπισυνάξει τοὺς (Q 3.17; 13:34) ἐκλεκτοὺς (20c, 22d)

ἐκ τῶν τεσσάρων ἀνέμων (6:51; LXX Zech. 2:10) ἀπʼ ἄκρου γῆς (9:20) ἕως ἄκρου οὐρανοῦ (25a, b; Q 10:21)

3.2.3 Mk 13:28–31: The continuing programme of learning as the conclusion of the

answer to 13:3–4 (continuing the outermost inclusion A′ to A 13:3–4): The predictable imminence of the coming and gathering: (a) Ratio (exemplum/illustratio as amplificatio of the argumentation by analogous reasoning):

28 Ἀπὸ δὲ τῆς *συκῆς (11:20–21; LXX Isa. 34:4) μάθετε (Rom. 16:17; 1 Cor. 4:6; 14:31, 35) τὴν παραβολὴν (12:1, 12)

ὅταν (14) *ἤδη (11:11; Q 3.9) ὁ κλάδος (4:32) αὐτῆς ἁπαλὸς γένηται (19) καὶ ἐκφύῃ τὰ *φύλλα (11:13; contra to LXX Jer. 34:4),

γινώσκετε (12:12) ὅτι ἐγγὺς τὸ θέρος (4:29; Q 10:2) ἐστίν.39

                                                                                                                                                                                          cf. K. Koch, ‘Sabbatstruktur der Geschichte’, ZAW 95 [1983], pp. 403–30, 429; there is no ‘heavenly 

revelation’ at all and the name of God is never used: ‘In der Zehn‐Epochen‐Lehre meldet sich der 

G.F.W. Hegel des Altertums! Wollen wir eine solche Schau von Geschichte und Eschaton noch 

apokalyptisch nennen?’ [p. 422]); 1 En. 91:16 reads for the tenth Seven‐Epoch: ‘Und ein neuer 

Himmel wird erscheinen, und alle “Kräfte de Himmels” strahlen auf und leuchten sieben‐fach. 

Danach gibt es viele “Wochen”, deren Zahl kein Ende hat’ (F. Dexinger, Henochs 

Zehnwochenapokalypse und offene Probleme der Apokalyptikforschung [SPB, 29; Leiden: E.J. Brill, 

1977], pp. 179–80; Koch, ‘Sabbatstruktur’, pp. 410–11; S. Uhlig, ‘Das äthiopische Henochbuch’, JSHRZ 

5 [1984], pp. 469–780, 714). 

38 The subjects of this verb of perception do not refer to the previously mentioned sun, moon, stars 

and powers, since in this word‐field of ‘seeing’ they became unable to see (in the ancient sense, 

where perception is seen as a productive act: S.R. Garrett, ‘Lest the Light in You be Darkness’, JBL 110 

[1991], pp. 93–105; J.A. Darr, ‘Watch How you Listen [Lk 8.18]: Jesus and the Rhetoric of Perception 

in Luke–Acts’, in Malbon and McKnight [eds.], Criticism, pp. 87–107; D.O. Via, ‘Mt’s Dark Light and 

the Human Condition’, in Malbon and McKnight [eds.], Criticism, pp. 348–61). The subjects must be 

the last‐mentioned persons, viz. ἐκλεκτοί of 13:20, 22 (as this first description and Pauline adaption 

emphasized), which is not incidentally renominalized in 13:27. It is not unusual in Mark for a subject 

not to be mentioned after sentences of silent continuation, as has just been seen in 13:20d, e, in 

which God is the subject named in 13:20a (cf. the important role of the Markan Jesus as eye‐opener; 

E.S. Johnson, ‘Mk 8.22–26: The Blind Man from Bethsaida’, NTS 25 [1979], pp. 370–83). We must 

reject the conjecture, however, that the subjects must be enemies and persecutors. There is an 

‘absence of any indication that the’ Markan ‘Son of Man functions as judge’ at all (Yarbro‐Collins, ‘Mk 

13’, pp. 1136–37). 

153  

29 οὕτως (10:43; Q 11.30; 17.24, 26) καὶ ὑμεῖς (23; 7:18), ὅταν (28b) ἴδητε (14) ταῦτα (4b, c = 5–22) γινόμενα (28b),

γινώσκετε (28c) ὅτι ἐγγύς (28e) ἐστιν ἐπὶ *θύραις (11:4).40

b) Confirmatio: Seventh I-oracle: a 30 ἀμὴν (43; Q 12:37) λέγω (23) ὑμῖν (29a) ὅτι

b οὐ ^ μὴ παρέλοῃ (6:48; Q 16:17) ἡ γενεὰ αὕτη (8:38; Q 11:30–32, 51) c μέχρις οὗ (Gal. 4:19) ταῦτα πάντα (23, inclusio to 4) γένηται (29b).

c) Conclusio: Eighth I-oracle:

b′ 31 ὁ οὐρανὸς καὶ ἡ γῆ (27b) παρελεύσονται (30b; Q 16:17), a′ οἱ δὲ λόγοι (2a, 4a, 5a, 23b, 30a) μου οὐ ^ μὴ (30b) παρελεύσονται (31a).

3.2.4. Mk 13:32–37: Peroratio: The constant attention towards the unpredictable suddenness of the coming and gathering: (a) Subpropositio: Ninth meta-narrative commentary:41

32 Περὶ δὲ τῆς ἡμέρας (24a) ἐκείνης (24a) ἢ τῆς ὥρας (11; Q 12.39–40) *οὐδεὶς ^ (Q 10:22) οἶδεν (12:24; Q 12.39),

— οὐδὲ (12:10) οἱ ἄγγελοι (27a) ἐν οὐρανῷ (31a) οὐδὲ ὁ υἱός (26; contra Q 10:22)— ^ εἰ ^ μὴ ὁ πατήρ (8:38; 11:25; Q 10.21)

(b) Expected reaction of those being instructed (imperative): 33 *βλέπετε (23 inclusion to 5b) ἀγρυπνεῖτε (2 Cor. 6:5; 11:27) (c) Ratio (deductive conclusion from the propositio): a οὐκ ^ (24) οἴδατε ^ (32a) ^ γὰρ

^ πότε (4 inclusio) ὁ καιρός (12:2; 1 Cor. 4:5; Rom. 13:11) ἐστιν. (d) Confirmatio: Inductive conclusio from exemplum/illustratio:

b 34 ^ *ὡς (4:26, 31) ἄνθρωπος (26; 4:26; 12:1; Q 19.12.21) ἀπόδημος —ἀφεὶς (2) τὴν οἰκίαν (15; 10:29) αὐτοῦ καὶ δοὺς (24; Q 19.12, 24) τοῖς δούλοις (12:2, 4; Q 19.13, 15) αὐτοῦ τὴν ἐξουσίαν (11:33; 3:15;

6:7; 2 Cor. 13:10) ἑκάστῳ τὸ ἔργον (1 Cor. 3:5, 13; Gal. 6:5 from 2:6) αὐτοῦ—

καὶ τῷ *θυρωρῷ (29) ἐνετείλατο (10:3; Q 4.10) ^ *ἵνα (18) γρηγορῇ (35 contra 8 and 22). (e) Expected reaction of the instructed students (imperative):

b′ 35 γρηγορεῖτε (34d; 1 Cor. 16:13)42 οὖν (10:9; 11:31)43

                                                                                                                                                                                          39 Note the Latin syntax, as in Mk 4:28 and 12:31. 

40 Note the double temporal expression in Mark’s style and also the use of a classical plural (different 

from his previous narrative usage in 1:33; 2:2; 11:4). Here again Mark highlights the expressions of 

his rhetorical Jesus (Lee, ‘Features’, p. 23). 

41 In view of this function, the break between 13:32 and 13:33 in Nestle and Aland (Novum 

Testamentum Graece, p. 136) is misleading. 

154  

(f) Ratio (reasoning by applicatio of the exemplum/illustratio): a′ οὐκ (33b) ^ οἴδατε (33b) ^ γὰρ ^ πότε (33b) ὁ κύριος (37; Q 12.42–43, 45–46) τῆς οἰκίας ἔρχεται (26a)

(f1) Authorial commentary: a —ἢ *ὀψὲ (11:9)

b ἢ μεσονύκτιον (6:48) b′ ἢ *ἀλεκτορο*φωνίας (14:30, 72) a′ ἢ *πρωΐ (11:20)—

(f2) Conclusio negativa:

36 μὴ (21)—ἐλθὼν (35c) ἐξαίφνης— εὕρῃ (11:13; 1 Cor. 4:2; 2 Cor. 5:3) ^ ὑμᾶς (29) ^ *καθεύδοντας (5:39; 1 Thess. 5:6–7).

(g) Conclusio positiva universalized: Expected reaction of all instructed readers:

37 ὃ δὲ ὑμῖν (36) λέγω (30a, 31b) ^ πᾶσιν (13, 20) λέγω (37a),

γρηγορεῖτε (35a).

3. Summary

1. ‘Mark’, the author as teacher, looks on his implied readers as his pupils (from a synchronic perspective), to whom he narrates and explains, via his rhetorical Jesus (his mouthpiece), their common contemporary situation in order to strengthen their hope for final vindication in the near future.

2. The primary level of communication (author/readers) is not restricted to the well-known

command to the reader in Mk 13:14b, but includes many more aspects within the narrated speeches of Jesus: Mk 13:7c, 8c, 10, 14b, 17, 19–20, 24–27, 32, 35c. These eight postpositive elements placed after the arguments are direct forms of address to the readers: This is what you can/should see/conclude/learn from the preceding argument.

3. The imperatives in the second person addressed to the narrated addressees (Group of

Four Students: 13:5b, 7b, 9a, 11b, d, [13b], 18a, 21c, 28a, c, 29b, 33a, 35a, 37c) function as propositiones, which are substantiated (syndetically or asyndetically) by eight oracles (13:6, 8a–c, 9b–d, 12–13a, 22, 24–27, 30, 31) as their rationes et conclusiones.

4. Within the oracles of disaster, Mk 13:14d–20 has in view a third group of persons

(women in Judaea) who are not identical to the addressees, since in Mk 13:18 the addresees are called upon to intercede on behalf of this third group (of women).

                                                                                                                                                                                          42 This semantic field is not only a synonym to 13:33 and an antonym to 13:7b but can be counted as 

a rhetorical antonym to 13:8, 22 ἐγερθήσ[ε/ον]ται as well. As is well known, this verb is a Hellenistic formation from the perfect ἐγρήγορα, ‘to be in a state of wakefulness’ (Blass, Debrunner and 

Rehkopf, Grammatik, §73; Bauer, Aland and Aland, Wörterbuch, col. 334). 

43 A classical feature used in speech to highlight the expression of the Mark’s rhetorical Jesus (Lee, 

‘Features’, p. 23). 

155  

5. The segment Mk 13:28–31 intends the whole testamental speech (while following the oracle of salvation in Mk 13:24–27) to be a program of learning.

6. The last segment Mk 13:32–37 functions as the peroratio of the whole testamental

speech.

156  

PRONOUNS OF SHAME AND DISGRACE IN LUKE 22:63–64* Jonathan M. Watt

1. Introduction

The observation of Saussure that ‘Language is a social fact’ is more specifically drawn out by Halliday (1978: 3) when he states that ‘language actively symbolizes the social system, representing metaphorically in its patterns of variation the variation that characterizes human culture’. When a speaker selects from his repertoire one form of speech over another, his sociocultural environment is being incorporated, in some way or another, into his operational code.

In 1960 Brown and Gilman presented one instance of this language/sociology connection, citing a topic which has become popular fare for sociolinguistic observation. In ‘The Pronouns of Power and Solidarity’ (1960), they discussed formal versus informal second-person singular pronouns in French, English, Italian, Spanish and German. They concluded (253) that one’s pronominal usage ‘gives away … class status and … political views’, and added that some optional variations in usage may even ‘express transient moods and attitudes’. The authors also argued that a ‘nonreciprocal power semantic’ hinges upon the person who wields the most power during the course of a dialogue, be it physical, familial, political, material, military or ecclesiastic in derivation. They also showed that what has come to be called the ‘dispensation right’ permits superiors to address each other mutually with formal designations, while superior-inferior relationships remain subject to nonreciprocity in terms of address.

Though neither of the primary languages of the Bible offers this type of pronominal option, unequal relationships can certainly be flagged using other methods. Garr (1994), for example, showed how familial terms (e.g. ‘my brother’, Gen. 33:5–10 passim) could be invoked in attempts at mutuality and equality, while the omnipresent biblical phrases ‘my lord’ and ‘your servant’ attest to social asymmetries in many situations portrayed in the Old Testament.

The present essay considers how this power semantic features into one New Testament situation, the Arrest Narratives of Jesus as recorded in the Gospel of Luke, and how it finds a rather poignant application. Proceeding under the assumption that language is indicative of social status, I suggest that in the Arrest Narratives a ‘null set’ option is brought into play, that is, that it is the absence of a specific term of address that flags social asymmetry, and that this absence is applied by those who wield power in order to designate contempt for a prisoner who has offended social mores.

Moreover, I will demonstrate that this nonreciprocity of address is in keeping with similar conventions that appear in a variety of Old Testament texts, and that it is consonant with acknowledged Jewish practices postdating the New Testament. It will be suggested that the appearance of this idiosyncrasy in the Gospel of Luke is consistent with a self-conscious

                                                            * English Bible quotations are from the New American Standard Bible. Numerical counts are based 

upon the United Bible Societies’ Greek New Testament (3rd edn). I wish to thank Stanley Porter and 

Christina Paulston for assistance in identifying certain materials that proved useful to the 

development of this essay. Appreciation is also given to participants in the Sociolinguistics and 

Biblical Studies Consultation Group of the Society of Biblical Literature, who made helpful comments 

following its presentation on 26 November 1996. 

157  

language style that various writers suggest is typical of the author/compiler of the Third Gospel.

2. Background

This inquiry begins by noting a third-person singular pronominal ambiguity that appears in Lk. 22:63–64, a curiosity which has already received attention from various writers. Lying within Jesus’ arrest and interrogation narratives, these verses report: ‘And the men who were holding him in custody were mocking and beating him, and they blindfolded him …’—the word in each case being the pronoun αὐτός with accusative inflection. Although the antecedent closest to the three explicit pronouns is ‘Peter’ (v. 61), all the references are obviously to Jesus.

Cadbury commented on this incohesiveness in the narrative amid his discussion (1920: 101–102) of ‘a few cases where Luke’s omission of details given in Mark makes the situation obscure or abrupt’. Cadbury suggested that Luke had ‘fail[ed] to make plain that the chief

actor has changed, using in 22:63ff. the simple auton of Jesus although the last antecedent is Peter’. More recently, Marshall (1978: 845) has commented on the ambiguity, stating that ‘it seems likely that two different sources have been joined, although there is no agreement as to the cause of the difficulty’.

Yet a preferable solution to this problem may be found in a text-level feature of the Lukan narrative, one which may have originated in a social convention that appears in the very literary traditions that formed the backdrop for the Gospels. It has nothing to do with hypothetical textual seams or an unfortunate ambiguity of language, but instead reflects a technique by which numerous societies—including some of those depicted in the Bible—have identified a dishonored person: the omitting or erasure of his name. This solution draws from what Venneman has called the ‘presupposition pool’ (Brown and Yule 1983: 79) which, in this case, appears to have permitted contempt and shame to be indicated by certain language variants.

The convention of treating a disgraced person as if he were nameless appears in various Old Testament texts and also in later Jewish literature. All things being equal, it may tentatively be considered as current in first-century Jewish Palestinian society as well, and, if there, it could certainly find a place in the context- and dialect-sensitive Gospel of Luke. The substitution of a pronoun for Jesus’ name in Lk. 22:63–64 was not accidental but rather played a role in the author’s colorful and realistic accounting style of relational asymmetry between a religious-political power alliance and its condemned prisoner.

3. Namelessness in the Old Testament

The privilege of naming someone or something is an experience shared in all societies, so it is hardly surprising that the activity is frequently reported in the Old Testament narratives and, as is well known, sometimes appears with whimsical etymologies. ‘Essential to the being, existence, and character of God and man[humankind] are their names’ (Kaiser 1975: 361). But the inverse of biblical onomatology—the removal of someone’s name in cases of contempt and dishonor—is also worthy of consideration.

One catalyst for the removal of a name is infamy and disgrace, as in the case of the eighth-century BCE King Pekah, who usurped the Israelite throne by the assassination of Pekahiah (741–740) and then opted for an ill-fated alliance with the Judean Ahaz and the Syrian Rezin. Cited for electing pragmatic diplomacy over divine aid, he is reported to have done ‘evil in the sight of the Lord … not depart[ing] from the sins of Jeroboam’ (2 Kgs 15:28). Consequently, when Isaiah (7:9) makes reference to the king, he employs the periphrastic

158  

phrase ‘the son of Remaliah’ (7:9), thus omitting Pekah’s first name while preserving the identity for his audience.

A similar treatment was applied to Pekah’s Judean counterpart, J(eh)oram, whose regnal account (2 Kgs 3:6–8:16) fills some half-dozen chapters—yet he remains nameless throughout. In fact, dual namelessness appears in 6:32 where, in an offhand reference to his infamous father, Jehoram is designated ‘the son of a murderer’ (though it must be acknowledged that the Semitic ‘son of’ phrase may, at times, be nothing more than a circumlocution for the person himself).

Another way to ‘lose’ one’s name is through premature death. Names in many societies are passed through patrilineage (whatever the formal convention), so a man without a son faces discontinuity of name. Suffering death prior to the begetting of children terminates the family name, as Isa. 53:8, 10 show poignantly:

As for his generation, who considered That he was cut off out of the land of the living. … If he would render himself as a guilt offering, He will see his offspring, He will prolong his days.

This manner of name loss is mentioned in Ps. 41:5 as well, in which the writer laments: ‘My enemies speak evil against me, “When will he die, and his name perish?” ’

With a slight twist, the Old Testament also shows that one could lose his name if the offspring who would have inherited it were to perish prematurely. The import of the curse found in Josh. 6:26 (the subsequent result being duly reported in 1 Kgs 16:34) is as much aimed at the perpetrator as at the offspring:

Cursed before the Lord is the man who rises up and builds this city Jericho; with the loss of his first-born he shall lay its foundation, and with the loss of his youngest son he shall set up its gates.

An imprecation that specifically connects premature death with loss of name appears in Isa. 14:22:

‘And I will rise up against them’, declares the Lord of hosts, ‘and will cut off from Babylon and survivors, offspring and posterity’, declares the Lord.

This kind of name loss appears in the imprecatory Ps. 137:9, which is not merely a curse upon the author’s Babylonian captors but upon their posterity, since the destruction of the offspring relegates the parents to obscurity: ‘How blessed will be the one who seizes and dashes your little ones/Against the rock’. The idea is most explicit in Ps. 109:13 and Jer. 11:19:

Let his posterity be cut off; In a following generation let their name be blotted out. But I was like a gentle lamb led to the slaughter; And I did not know that they had devised plots against me, saying: ‘Let us destroy the tree with its fruit, And let us cut him off from the land of the living, That his name be remembered no more’.

The idea was expressed conversely (1 Sam. 24:21) when Saul begs David to spare his life, referring to no fewer than three successive generations: ‘So now swear to me by the Lord that you will not cut off my descendants after me, and that you will not destroy my name from my father’s household’. And when God promises to maintain the existence of his own people, 2 Kgs 14:27 reports that he ‘did not say that He would blot out the name of Israel from under heaven’, whereas Israel’s national enemies inclined to the opposite direction, intending to ‘wipe them out as a nation, that the name of Israel be remembered no more’ (Ps. 83:4).

159  

Still another occasion for the obscuring of one’s name in the Old Testament texts is religious infidelity. The following examples from Ps. 44:20 and Jer. 23:27 illustrate this ironically, because they refer to the name of God rather than men:

If we had forgotten the name of our God, Or extended our hands to a strange god … [Deceptive prophets] intend to make my people forget my name by their dreams … just as their fathers forgot my name because of Baal …

To ‘forget’ someone’s name is to absolve oneself of personal cognizance. Erasure of another’s name means complete disavowal, as illustrated once again in Ps. 9:5–6 and Zeph. 1:4:

You have rebuked the nations, you have destroyed the wicked; You have blotted out their name forever and ever. The enemy has come to an end in perpetual ruin, And you have uprooted their cities. The very memory of them has perished. And I will cut off the remnant of Baal from this place, And the names of the idolatrous priests along with the priests. Not only may God be portrayed as participating in this activity when he ‘blots out the

name’ of his enemies (Ps. 109:13), but a writer/editor may apply the same technique throughout a text, as Lillian Klein (1988: 142) suggests as been done in the latter chapters of the book of Judges:

The gradual disappearance of naming correlates roughly with an increase in worldly values and clan identification, so that the characters become less individual and more emblematic even as their values become more worldly and less abstract. Cynthia Miller (in conversation) suggests a similar technique was being used by the writer

of Ruth, who identified the main character only pronominally during the third chapter of the book, that is, during the process of her admittance to Israelite society (the mention of her name in v. 9 is from her own lips). However, Ruth will be reintroduced by name again in the fourth chapter, apparently as an indication of her successful assimilation.

Thus it appears that, in the Old Testament, namelessness can be a means of showering contempt upon moral and ethical corruption. Those whose behavior was considered disgraceful could be marked by means of this excision.

4. The Namelessness of Jesus

Various New Testament passages also witness, if subtly, to the connection between namelessness and disgrace. Vague paternity, for example, has been a cause for ridicule in many societies, especially when it is the father who supplies the child’s surname. Such probably lies behind the following comments (Jn 8:19, 41; 9:29) cast at Jesus in the course of various altercations:

And so they were saying to him, ‘Where is your father?’ They said to him, ‘We were not born of fornication …’ ‘But as for this man, we do not know where he is from.’

If one’s father could not be identifed (or did not wish to be), one would lack a credible reputation in any mononymous culture which employed the convention ‘the son of X [his father]’.

160  

Whatever the insinuations during the peripatetic ministry, there was no doubt about Jesus’ reputation when he stood before his captors. Malina and Neyrey (1991: 57), in an article on ‘Honor and Shame in Luke–Acts’, recount the litany of dishonoring actions done to him, including Judas’s sardonic kiss, the binding of his hands, the blindfolding, pinioning and mutilating of the prisoner. Yet actions are usually accompanied by words, and a text that records disgracing behavior might also retain the condemning language as well. For as Halliday (1978: 3) states, there is

a twofold function of the linguistic system … both as expression of and as metaphor for social processes … in the microencounters of everyday life where meanings are exchanged, language not only serves to facilitate and support other modes of social action that consititute its environment, but also actively creates an environment of its own, so making possible all the imaginative modes of meaning, from backyard gossip to narrative fiction and epic poetry. The context plays a part in determining what we say; and what we say plays a part in determining the context [emphases mine]. Consequently, it is suggested that the ambiguous pronouns of Lk. 22:63–64, along with

similar forms in the matrix text, functioned in this dual capacity. They were considered appropriate to the event and recalled for the reader the vividness of that event. A perusal of the context in which these verses occur reveals that, in fact, Jesus’ name is absent throughout most of the Lukan arrest narratives, and the few instances where it does appear evidence a restricted usage.

The events arising from the evening seizure of Jesus on the Mount of Olives, up to (but not including) the morning approach to Golgotha, are recorded in Mt. 26:57–27:31; Mk 14:53–15:20; Lk. 22:54–23:25; and Jn 18:12–19:16. Luke stands out from the other three Gospels in its expansion of the morning council proceedings, in which Jesus stands in the center of a hostile Jewish crowd.

Not even once during the course of the arrest and Peter’s denial (Lk. 22:54–62), the mockery (22:63–65), or the morning proceedings (22:66–67), is Jesus identified by name. Instead, the preponderance of the masculine pronoun αὐτός ‘he/him’, the demonstrative pronoun/adjective οὗτος ‘this/that one’, or just ‘the man’ is striking, as this listing from the Lukan narrative surrounding vv. 63–64 will show:

22:54—‘they arrested him’ 22:56—‘[Peter] was with him too’ 22:57—‘I [Peter] do not know him’ 22:59—‘this man was with him too’ 22:66—‘they led him to the chamber’ 23:1—‘they led him to Pilate’ 23:2a—‘they began to accuse him’ 23:2b—‘we found this one misleading our nation’ 23:3—‘Pilate asked him’ 23:4—‘I find no guilt in this one’ 23:6—‘Pilate asked whether the man was a Galilean 23:7—‘Pilate sent him to Herod’ 23:8—‘Herod wished to see him … having heard about him’ 23:9—‘Herod questioned him at length, but he answered nothing’ 23:11—‘they treated him with contempt … and sent him to Pilate’ 23:14—‘this man’ (appears twice) 23:15—‘[Herod] sent him back to me … yet nothing worthy of death has been done by him’ 23:16—‘I will punish and release him’ 23:18—‘Away with this one!’ 23:21—‘Crucify him!’

161  

23:22—‘what evil has this one done?’ ‘I found nothing worthy of death in him.’ ‘I will … release him.’

23:23—‘But they were insisting he be crucified.’ Now it must be admitted that some of these examples are nothing more than routine pronominal anaphora, such as would appear routinely in any text. An isolated case proves nothing. But the narrative, seen as a whole, contains a noteworthy repetition of pronouns. The pronominal ‘cohesiveness’ (Halliday and Hasan 1976) of the Arrest Narratives works endophorically, allowing the reader either to gaze anaphorically (backward) to the explicitly named Jesus who was arrested ‘with clubs and spears’ (22:52), or cataphorically (forward) to his accompaniment by the cross (23:26) on the way to Calvary. The vacuum of identity created by the pronouns is filled with tokens of shame.

The following comparison of the Gospel accounts supports this apparent feature of namelessness. The chart indicates the number of appearances of ‘Jesus’ in the arrest narrative of each Gospel: the frequency column to the right is a percentage (based on the ratio between the number of appearances of his name per quantity of verses in that Gospel’s Arrest Narrative), indicating how often he is specifically named:

Gospel Number of

Verses Number of

Appearances Frequency

Matthew 50 15 0.30

Mark 40 9 0.23

Luke 43 3 0.07

John 45 19 0.42 The frequency of Jesus in Luke’s Gospel is considerably lower than that of the other Gospels; his name never appears on Jewish lips (contra the synoptic parallels, Mt. 26:69, 71; Mk 14:67). His name appears only in 23:8 in connection with Herod, and again in 23:20, 25 during his examination by Pilate, with neither examiner being, properly, a Jew. Conversely, the Lukan text may indirectly support Brown and Gilman’s observation (1960: 256) that social equals employ equivalent terminology to enhance fraternity, when it notes, in connection with this incident, that ‘Herod and Pilate became friends’.

Yet how do these figures compare with Luke’s own usage of the name of Jesus elsewhere? In Luke 4–20, Jesus appears 71 times across 763 verses, yielding a frequency of only 0.09 (i.e. not much more than the arrest narratives). However, these chapters incorporate blocks of Jesus’ own teachings, materials which would have little need of self-reference. Two other sections may provide a more reliable comparison: Luke 4–5 contains 14 appearances of Jesus in 83 verses, while Luke 8–9 has 19 appearances across 118 verses, yielding frequencies of 0.17 and 0.16, respectively. These chapters, from Luke’s own hand, hold more than double the frequency of Jesus than the arrest narratives. It would appear, therefore, that while Luke in general uses pronouns in greater proportion than the other Gospels, he becomes even more restrictive in those parts of his text which cover the arrest.

It has long been observed that the Greek text of Luke–Acts shows considerable variation in dialect (or register or style), containing heavy Semitic influence on the infancy narratives while holding to a standard koine in the latter chapters of Paul’s ministry (Guthrie 1979: 115; Plummer 1960: xlix; Foakes-Jackson and Lake 1922: 36, 39). These and other writers have been impressed by Luke’s ‘adaptability’ and ‘versatility’, stating that he could be ‘as Hebraistic as the LXX, and as free from Hebraisms as Plutarch’, finding the ‘uneven character of the Greek’ to be unusual and vivid.

162  

Luke’s sensitivity to language register, whatever the motive or mechanism, makes him a good candidate for retaining cultural markers such as namelessness and shame. The disparity between the limited use of Jesus’ name in the Lukan arrest narratives, and its more frequent appearance in the parallel sections of the other Gospels, as well as in Luke’s own work elsewhere, suggests an intentional effort on his part to retain and reflect the custom. The frequent pronominal anaphora of the Arrest Narratives in Luke would comprise additional evidence for what Porter (1989: 141) has labeled ‘Semitic enhancement’ in the third Gospel.

Not surprisingly, it appears that the same namelessness technique has been reflected elsewhere in Luke’s Gospel as well. In Lk. 10:37, for example, a lawyer answers Jesus’ question about who had shown the requisite mercy to the hypothetical man on the road to Jericho. The lawyer identifies the good deed doer periphrastically as ‘the one who showed mercy’ rather than utter the simpler designation ‘Samaritan’. The technique seems to appear yet again in Acts 5:28, in which Jesus is referred to by the high priest as ‘this name’ and his death is mentioned indirectly using the phrase ‘this man’s blood’. One wonders if there is not a touch of irony in Peter’s statements (Acts 4:10, 12) to the high priestly family that healing and salvation are ‘by the name of Jesus Christ’ and ‘by no other name under heaven’.

The namelessness phenomenon reappears in the centuries following the New Testament era, being found in the ‘reluctance to pronounce the name Jesus which has become so

common in Jewish orthodoxy’; Jesus was referred to simply as taluy ‘the hanged one’ (Bruce

1981: 121). Specifically, a medieval Jewish antigospel known as the Tol’doth Yeshu alleged that Jesus was of uncertain lineage, thus appearing to reflect the previously cited examples from the Gospel of John. Most of its versions, according to Howard (1988: 61), claim that

Mary originally named her son Yehoshua. After the circumstances of his birth came to light and the problems created for Israel by his later actions, his name was changed to Yeshu (Yod-Sin-

Waw). The explanation is given that these letters stand for Yimmah shemo vezrichro, ‘may his name and memory be blotted out’. Evidence from the Old Testament thus suggests that those who occupy the seat of legal or

prophetic judgment can distance themselves from, and show contempt for, a disgraced kinsman by relegating him to obscurity. The custom appears to find an analogue in the Lukan narratives, and fits with subsequent Jewish custom. The technique of namelessness provides a desirable explanation for the pronominal ambiguity in Lk. 22:63–64. This intentional omission of Jesus’ name is a fitting sociolinguistic correlate to Bruce Shelley’s observation (1982: 15) that ‘Christianity is the only major religion to have as its central event the humiliation of its God’.

REFERENCES

Bammel, Ernst

1967 ‘Christian Origins in Jewish Tradition’, NTS 13: 317–35. Brown, Gillian, and George Yule

1983 Discourse Analysis (CTL; Cambridge: Cambridge University Press).

                                                            NTS New Testament Studies 

CTL Cambridge Textbooks in Linguistics 

163  

Brown, R., and A. Gilman 1960 ‘The Pronouns of Power and Solidarity’, repr. in Pier Paolo Giglioli (ed.), Language in Social

Context (Harmondsworth: Penguin Books, 1982): 252–82. Bruce, F.F.

1981 The Book of Acts (NICNT; Grand Rapids: Eerdmans). Bultmann, Rudolf

1957 Die Geschichte der Synoptischen Tradition (Göttingen: Vandenhoeck & Ruprecht). Cadbury, H.J.

1920 The Style and Literary Method of Luke (Cambridge, MA: Harvard University Press). Catchpole, David R.

1971 The Trial of Jesus (Leiden: E.J. Brill). Daube, David

1956 The New Testament and Rabbinic Judaism (University of London: Athlone Press). Foakes-Jackson, F.J., and Kirsopp Lake (eds.)

1922 The Beginnings of Christianity, II (London: Macmillan). Garr, W.

1994 ‘Politeness Strategies in Biblical Hebrew’; paper presented to the Sociolinguistics and Biblical Studies Consultation Group of the Society of Biblical Literature in Chicago, 20 November. Guthrie, Donald.

1979 New Testament Introduction (Downers Grove: Inter-Varsity Press). Halliday, M.A.K.

1978 Language as Social Semiotic: The Social Interpretation of Language and Meaning (Baltimore: Edward Arnold).

1985 An Introduction to Functional Grammar (London: Edward Arnold). Halliday, M.A.K., and R. Hasan

1976 Cohesion in English (London: Longman). Howard, George

1988 ‘A Primitive Hebrew Gospel of Matthew and the Tol’doth Yeshu’, NTS 34: 60–70. Kaiser, W.C., Jr

1975 ‘Name’, in Merrill C. Tenney (ed.), Pictorial Encyclopedia of the Bible, IV (Grand Rapids: Zondervan): 360–66. Klein, Lilian

1988 The Triumph of Irony in the Book of Judges (Sheffield: Almond Press). Krauss, Samuel

1977 Das Leben Jesu nach jüdischen Quellen (Hildesheim: Georg Olms).

                                                            NICNT New International Commentary on the New Testament 

164  

Malina, Bruce J., and Jerome H. Neyrey 1991 ‘Honor and Shame in Luke–Acts’, in Jerome H. Neyrey (ed.), The Social World of Luke–Acts

(Peabody: Hendrickson): 25–65. Marshall, I. Howard

1978 Commentary on Luke (NIGTC; Grand Rapids: Eerdmans). Plummer, Alfred

1960 The Gospel According to St Luke (ICC; New York: Charles Scribner’s Sons). Porter, Stanley E.

1989 Verbal Aspect in the Greek of the New Testament, with Reference to Tense and Mood (SBG, 1; New York: Peter Lang). Schlichting, Günter

1982 Ein jüdisches Leben Jesu (Tübingen: J.C.B. Mohr). Shelley, Bruce L.

1982 Church History in Plain Language (Dallas: Word Books). Taylor, Vincent

1972 The Passion Narrative of St Luke (Cambridge: Cambridge University Press).

                                                            NIGTC The New International Greek Testament Commentary 

ICC International Critical Commentary 

SBG Studies in Biblical Greek 

165  

PARTICIPANT REFERENCE AND FOREGROUNDED SYNTAX IN THE STEPHEN EPISODE Gustavo Martín-Asensio

Ever since Dibelius’s influential work on Acts, it has been customary to assume a radical disjunction between compositional artistry and cohesiveness on the one hand, and historicity on the other. It was Dibelius’s argument that the author of Acts could not possibly have had sources for his ‘second treatise’, because the first-century church was wholly uninterested in recording events.1 Consequently, he argued, the vast majority of the material in Acts is the result of the author’s ‘new literary freedom’.2 A decade before Dibelius inaugurated the age of Form-Criticism, the last of the great source-critics, Adolf Harnack, had already affirmed that the possibility of the writer having rewritten a ‘we’ source to conform it to his own style was to him ‘absolutely unimaginable’.3 His conclusion was a foretaste of Dibelius’s: what is stylistically polished and consistent with the rest of the work cannot have been grafted in from a source, and must, therefore, be considered to be the author’s own creation. The same assumption has since then been held by scholars who, unlike Dibelius and his many modern followers, argue for the existence of sources behind most, if not all, of Acts. Max Wilcox is a case in point. His ‘Foreword to the Study of the Speeches in Acts’4 offers convincing arguments for the existence of sources behind the speeches. Yet, Wilcox relies on the familiar criteria for identifying the source-based material, namely the alleged awkwardness of style, lack of adequate links and unusual vocabulary or turns of phrase.5

The literary analysis of the speeches has been further complicated by the debate over the meaning and relevance for Acts of the famous programmatic statement found in Thucydides’ History of the Peloponnesian War.6 In his classic essay mentioned above, Cadbury affirmed

                                                            1 Martin Dibelius, ‘Style Criticism of the Book of Acts’, in his Studies in the Acts of the Apostles (trans. 

Mary Ling; London: SCM Press, 1956), pp. 1–25, esp. p. 4. But see Martin Hengel’s chapter ‘The 

Earliest Christian Histories as Sources for a History of Earliest Christianity and the Unity of Kerygma 

and Historical Narrative’, in idem, Acts and the History of Earliest Christianity (trans. John Bowden; 

London: SCM Press, 1979), pp. 40–49, esp. pp. 43–44. 

2 Dibelius, Studies, p. 196; see also the classic essay by H.J. Cadbury, ‘The Speeches in Acts’, in F.J. 

Foakes‐Jackson and Kirsopp Lake (eds.), The Beginnings of Christianity, V (London: Macmillan, 1933), 

pp. 402–27, esp. pp. 406–407. 

3 Adolf Harnack, Luke the Physician (trans. J.R. Wilkinson; London: Williams and Norgate, 2nd edn, 

1909), p. 53. In regard to the ‘we’ sections, Harnack denies adamantly that Luke had ‘short notes 

which refreshed his memory’, concluding that the material is all Luke’s (Luke, p. 53). E. Haenchen 

feels that Harnack was rash in reaching this conclusion (E. Haenchen, The Acts of the Apostles [trans. 

B. Noble and G. Shinn; Oxford: Basil Blackwell, 1971], p. 32). 

4 Studies in Judaism in Late Antiquity 12.1 (1975), pp. 207–24. 

5 Wilcox, ‘Foreword’, p. 213; see also Javier Colmenero‐Atienza, ‘Hechos 7, 17–43 y las corrientes 

cristológicas dentro de la primitiva comunidad cristiana’, EstBíb 33 (1974), pp. 31–62, esp. pp. 43, 50. 

6 For an excellent and up‐to‐date discussion of this issue see W.F. McCoy, ‘In the Shadow of 

Thucydides’, in Ben Witherington III (ed.), History, Literature and the Book of Acts (Cambridge: 

Cambridge University Press, 1996), pp. 3–23; see also S.E. Porter, ‘Thucydides 1.22.1 and Speeches in 

Acts: Is There a Thucydean View?’, NovT 32 (1990), pp. 121–42. 

166  

that the speeches in Acts amounted to ‘purely literary artifices’, a feature he discerned in most Hellenic and Graeco-Roman historians, including Thucydides.7 But Cadbury seems to have ignored the fact that, in the classical historiographical tradition of Thucydides and Polybius, two essential elements were given similar consideration, namely narratio or fidelity to real history, and exornatio, or perfection of style, the two being considered by no means mutually exclusive.8

It seems to me that the mentioned dichotomy between that which is ‘literary’ and that which is ‘historical’, between compositional dexterity and reliance upon sources, makes the author of Luke–Acts look far less skilled than he actually was, and is ultimately both unwarranted and unnecessary. Several recent works have demonstrated that Luke–Acts is the product of a writer who was both a faithful editor and skilled literateur.9 In the study that follows I shall focus on the Stephen story as an episode within the larger work of Acts. My method of analysis is linguistic and literary rather than historical, not because I oppose the two, but because I believe any study of a New Testament document must begin by allowing the author himself to tell his story, focusing not only on his literary style as an indicator of the presence or absence of sources, but on the various functions that his language and style is made to serve in that text. The notion of linguistic foregrounding, defined as that prominence

                                                            7 Cadbury, ‘The Speeches’, pp. 402, 405. According to Cadbury, Thucydides frankly admitted that in 

composing speeches, ancient historians ‘probably rarely relied on any real knowledge of what was 

actually said’ (p. 405). Yet, this is in fact the opposite of what Thucydides himself asserts: … ἐχομένῳ ὅτι ἑγγύτατα τῆς ξυμπάσης γνώμης τῶν ἀγηθῶς λεχθέντων, οὕτως εἴρηται (History of the 

Peloponnesian War 1.22.1). Furthermore, Cadbury may have contradicted himself. On p. 402 he 

affirmed the conformity of the book of Acts to the ancient practice of ‘adorning historical works with 

imaginative speeches’, while on p. 425 he stated that although for classical historians speeches were 

the ‘most prized parts’ of their works and composed them most carefully, Luke did not follow their 

model. In an earlier essay (H.J. Cadbury, The Style and Literary Method of Luke. II. The Treatment of 

Sources in the Gospel of Luke [HTS, 6; Cambridge, MA: Harvard University Press, 1920], p. 73) 

Cadbury had expressed great confidence in the accuracy with which Luke follows his Markan source. 

In light of this, Cadbury’s radically different assessment of Acts seems at best difficult to accept. For a 

convincing study of Luke’s editorial activity in both the Gospel and Acts, see Ben Witherington III, 

‘Editing the Good News: Some Synoptic Lessons for the Study of Acts’, in idem (ed.), History, pp. 324–

47. 

8 See Helen. F. North, ‘Rhetoric and Historiography’, Quarterly Journal of Speech 42 (1956), pp. 234–

42, esp. 237, 242. 

9 Marion L. Soards, The Speeches in Acts (Louisville, KY: Westminster/John Knox Press, 1994), p. 12; 

S.E. Porter, ‘The “we” Passages’, in D. Gill and C. Gempf (eds.), The Book of Acts in its Graeco‐Roman 

Setting (A1CS, 2; Grand Rapids: Eerdmans; Carlisle: Paternoster Press, 1994), pp. 545–74, esp. pp. 

567–70; See also Johannes Munck, The Book of Acts (AB, 31; Garden City, NY: Double‐day, 10th edn, 

1981), pp. xxv, xxxix–xli. This is, however, by no means a novel view on the subject. Seventy years 

ago, E. Jacquier affirmed: ‘L’examen linguistique de Actes tend à prouver que toutes les parties du 

livre ont été écrites par le même auteur, quoiqu’il ne soit pas impossible qu’un éditeur habile ait 

utilisé des sources diverses, soit dans la premiè partie, soit dans la seconde, et n’ait retravailléreces 

diverses parties en leur imprimant une certaine conformité de style …’ (Les Actes des Apôtres [Paris: 

Librairie Victor Lecoffre, 2nd edn, 1926], p. clviii, emphasis mine). 

167  

that can be shown to be consistent and motivated in light of the overall theme(s) of a text,10 provides a link between stylistic and functional analysis and is capable of revealing important insights into the writer’s agenda in producing this ancient text.11

1. The Language of the Stephen Episode in Recent Study

Of the considerable amount of scholarship dedicated to the Stephen episode in recent years,12 the question of the relation of Stephen’s speech to the charges brought against him has often stimulated the most heated debate.13 In arguing for or against the speech as response to the charges of 6:13–14, or for or against the uniformity of its style with previous and later material, scholars often make reference to items of the language used that are seen by them as indicators of emphasis or of major turning points in the narrative or speech. Thus, for example, Kilgallen argues that direct speech is used by Luke when he wishes to mark climactic elements in his story;14 Soards affirms that the phrase καὶ νῦν indicates ‘major moments’ in speeches,15 and, following Conzelmann, asserts that relative clauses are indicative of ‘kerygmatic style’ and are, therefore, of central importance;16 Richard writes that by means of the repetition of a term, phrase or theme, the writer ‘draws particular attention to the components central to his views’;17 Colmenero-Atienza offers a stylistically based proposal as to who the figuras centrales may be in Stephen’s speech;18 lastly, Barrett affirms that οὗτος is ‘emphatic’ in 7:35–38 and that τοῦτον τὸν Μωϋσῆν in 7:35 is ‘brought to the

                                                            10 See M.A.K. Halliday, ‘Linguistic Function and Literary Style’, in idem, Explorations in the Functions of 

Language (London: Edward Arnold, 3rd edn, 1976), pp. 103–38. 

11 An agenda he has, at least partially, revealed in his prologues of Lk. 1:1–4 and Acts 1:1. E. Richard, 

however, has stated that only by examining the author’s sources ‘will it be possible to assess the 

redactive process and therefore the writer’s intention’ (‘The Polemical Character of the Joseph 

Episode in Acts 7’, JBL 98.2 [1979], pp. 255–67 [257]). 

12 A.F.J. Klijn, ‘Stephen’s Speech—Acts VII. 2–53’, NTS 4 (1957–58), pp. 25–31; Colmenero‐Atienza, 

‘Hechos’; John Kilgallen, The Stephen Speech (AnBib, 67; Rome: Biblical Institute Press, 1976); E. 

Richard, Acts 6:1–8:4, the Author’s Method of Composition (SBLDS, 41; Missoula, MT: Scholars Press, 

1978); Richard, ‘The Polemical Character’; J. Dupont, ‘La structure oratoire du discours d’Etienne 

[Actes 7]’, Bib 66 (1985), pp. 153–67; Dennis D. Sylva, ‘The Meaning and Function of Acts 7.46–50’, 

JBL 106.2 (1987), pp. 261–75; Craig C. Hill, ‘Acts 6.1–8.4: Division or Diversity?’, in Witherington (ed.), 

History, pp. 129–53. 

13 For a summary of the discussion see Kilgallen, The Stephen Speech, pp. 6–10. 

14 Kilgallen, The Stephen Speech, pp. 37, 39. One wonders, however, whether direct speech is 

climactic, especially when we consider the several ways in which the writers introduce direct speech 

when they deem it necessary: ‘and he said …’ (7:2); ‘Paul stood up … and said …’ (17:22), etc. 

15 Soards, The Speeches, pp. 92, 107. 

16 Soards, The Speeches, pp. 34, 68. 

17 Richard, Acts 6:1–8:4, p. 103. 

18 Colmenero‐Atienza, ‘Hechos’, p. 42. 

168  

beginning of the sentence for emphasis’.19 Though these statements are all valid as working hypotheses, the unsystematic and haphazard way in which they are presented raises important questions that remain unanswered. Of these, none is more crucial than the question of criteria: how exactly is a ‘major moment’ or a ‘central figure’ to be determined? In relation to what are these items major or central? Without an answer to these questions potentially fruitful insights such as the ones just mentioned will always suffer from a condition we could term ‘untestable impressionism’, that is, they are statements of opinion, rather than systematic and linguistically principled analyses that are open to evaluation by all.20

2. Major Participants in the Stephen Episode

Of the various theories that have been put forward regarding the structure of the speech, the biographical criterion as presented by Colmenero-Atienza21 seems to fit the data best. In the following pages, I hope to show that the entire Stephen episode (Acts 6–7) is designed by the author to underline one single feature that unites the careers of Stephen, Joseph and Moses. Far from a slightly revised ‘neutral’ account of Israelite history,22 the narrative and speech form a cohesive and masterfully crafted chronicle that successfully advances the writer’s overall theses in Acts.

In his essay entitled ‘Participant Reference in Koine Greek Narrative’,23 Stephen Levinsohn explores the various means available to Koine Greek writers for the introduction of and further reference to the characters in a narrative. More specifically, Levinsohn suggests the classification of ‘default’ and ‘marked’ encoding of participants, a notion based on the observation that the more significant a character is in a narrative, the more coding material is normally assigned to it to mark that significance. Thus, Levinsohn notes that in the Gospels any explicit reference to Jesus as subject of a clause when he has already been introduced is marked encoding, while the absence of explicit subject is generally the norm and is considered the ‘default’ encoding.24 While I am generally in agreement with Levinsohn’s theory, I would add that, at least until an exhaustive study of participant reference in the New Testament is carried out, statements of ‘norm’ or ‘default’, of which ‘marked’ encoding is

                                                            19 C.K. Barrett, The Acts of the Apostles (2 vols.; ICC; Edinburgh: T. & T. Clark, 1994–98), I, pp. 363–65. 

20 See Ronald Carter, ‘Introduction’, in idem (ed.), Language and Literature (London: George Allen & 

Unwin, 1982), pp. 1–17, esp. 4–6, for a well‐argued call to a linguistically based literary criticism. 

21 Colmenero‐Atienza rejects both the chronologically based structure proposed by Martini, and the 

geographically based structure suggested by Scharlemann, arguing instead for a biographical 

arrangement as follows: (A) vv. 2–8, Abraham; (B) vv. 9–16, Joseph; (C) vv. 17–43, Moses; (D) vv. 44–

50, the house of God; (E) vv. 51–53, the final invective (‘Hechos’, pp. 41–43). Though I am in 

agreement with the choice of a biographical structure, I do not concur with Colmenero‐Alienza’s 

inclusion of Abraham as a figure of equal importance to Joseph and Moses in the structure of the 

speech. See below on this point. 

22 See Haenchen, Acts, p. 288. Here, too, Haenchen follows Dibelius. 

23 In D.A. Black et al. (eds.), Linguistics and New Testament Interpretation: Essays in Discourse 

Analysis (Nashville: Broadman Press, 1992), pp. 31–44. 

24 ‘Participant Reference’; see also Stanley E. Porter, ‘Word Order and Clause Structure in New 

Testament Greek’, FN 6 (1993), pp. 177–206, esp. p. 194. 

169  

held to be a departure, should only be made relative to specific texts. I will later discuss a particular means of participant reference that seems unique to Luke, and can be properly understood only in light of his agenda in the Acts of the Apostles. First, however, I wish to suggest the following scale as a means of summarizing the ways in which the various participants are referred to in the Stephen episode:

1. Full explicit subject (proper noun used) e.g. Στέφανος δὲ … ἐποίει τέρατα καὶ σημεῖα 2. Abbreviated explicit subject (pronoun or article used) πρὶν ἢ κατοικῆσαι αὐτὸν ἐν Χαρράν 3. Non-explicit subject ἦν ἀστεῖος τῷ θεῷ 4. Non-subject participant25 (e.g. direct or indirect object) ὁ θεὸς τῆς δόξης ὤφθη τῷ πατρὶ ἡμῶν Ἀβραάμ

In reading a narrative such as the Stephen episode, even the literarily untrained reader will acquire impressions concerning what participants strike him or her as ‘more important’. The above scale is a means of grounding such impressions on the language system and the author’s use of it. Thus, in a narrative such as the Stephen episode, the importance or centrality of a character may be gauged by the frequency of his or her appearing as full or abbreviated explicit subject (categories 1 and 2). Conversely, the centrality of the characters in the story will decrease proportionately to their appearance as non-explicit subjects or non-subject participants. Though this method of participant reference analysis is somewhat more complex than Levinsohn’s described above, it takes full account of the fact that a participant may appear often as other than the subject of a clause or sentence, a fact that may yield important hermeneutical insights, as we shall see below. Further, the above scale is, unlike Levinsohn’s ‘default’ and ‘marked’ notion, capable of accounting for the transitivity choices26 that are often made by writers to differentiate a participant or participants from the rest.

In our story, human participants are many and varied, including ‘the disciples’ (6:1–2), Jewish and Greek Christians (6:1), the ‘seven’ (6:3–5), Stephen (6:5–7:2; 7:54–60), priests (6:7), certain synagogue members (6:9–10), ‘men’ (6:11), false witnesses (6:13), Jesus (6:14), Moses (6:11, 14), the high priest (7:1), God (7:2–8), Abraham (7:2–8), the patriarchs (7:8–9), Joseph (7:9–18), ‘another king’ (7:18–19), Moses (7:20–44), the Pharaoh’s daughter (7:21),

                                                            25 In Hallidayan functional grammar, the semantic framework for the representation of processes 

consists of three essential elements: the process itself (realized in texts by a verbal form), the 

participants (realized by actors or subjects and goals or direct objects) and circumstances (realized by 

adjuncts or circumstancial complements). Therefore, a particular character such as Abraham may 

appear several times as a participant in the narrative, though not as an actor or subject. See M.A.K. 

Halliday, An Introduction to Functional Grammar (London: Edward Arnold, 1985), pp. 101–44. 

26 The term ‘transitivity’ is here used in the rather wide functional grammar sense of ‘the different 

types of processes that are recognized in the language, and the structures by which they are 

expressed’. The key types of processes (and corresponding clause‐types) that Halliday recognizes are 

material processes (processes of doing), mental processes (processes of sensing) and relational 

processes (processes of being). See Halliday, Functional Grammar, pp. 101–57. See also Paul Hopper 

and Sandra Thompson, ‘Transitivity in Grammar and Discourse’, Language 56.2 (1980), pp. 251–99; 

and, for a critique of the latter, José María García‐Miguel, Transitividad y complementación 

prepositional en español (Verba, 40; Santiago: University of Santiago de Compostela Press, 1995). 

170  

the sons of Israel (7:23, 37), an Egyptian (7:24, 28), a prophet (7:37), our fathers (7:39), Aaron (7:40), David (7:46), the prophet (7:48), your fathers (7:52) and the witnesses (7:58). Of these participants, four stand out from the rest because of the sheer bulk of material dedicated to them, namely Stephen himself, Abraham, Joseph and Moses. Using the above scale, I shall examine each in turn. a. Stephen In the 16 verses that deal directly with him, Stephen appears 19 times. The distribution of these references is as follows:

Participant reference to Stephen: Acts 6:5–7:1; 7:55–60

1. Full explicit subject27 1×

2. Abbreviated explicit subject28 2×

3. Non-explicit subject29 10×

4. Non-subject participant30 6× The analysis of these references yields some interesting observations. First, though in these 16 verses Stephen’s presence ‘on stage’ is strong from the outset (of the seven ‘deacons’, it is he alone who is set aside for specific description in 6:5, and is referred to four times by name), the type of clauses in which these references appear encode processes that lack an external agent in all but two cases, that is, they make no reference to causation. In other words, though the spotlight is clearly on Stephen, he is not described as a dynamic protagonist who is directly involved in bringing about events and doing things. Of the 13 clauses where he is the subject participant, only two encode material processes, processes of ‘doing’ where Stephen is an agent acting upon a medium.31 The rest of these clauses encode relational processes (processes of being), or mental and verbal processes, where Stephen is speaking in his defense, crying out, looking to the heavens, seeing the glory of God, being full of the Holy Spirit, etc. The six references under my fourth category above (non-subject participant) appear in clauses where Stephen is on the receiving end of largely hostile actions (five out of six). It is interesting to note that the first and last times that Stephen is mentioned by name fall

                                                            27 Once in an independent clause: Στέφανος δὲ πλήρης χάριτος καὶ δυνάμεως ἐποίει τέρατα καὶ σημεῖα μεγάλα, 6:8. 

28 Twice in independent clauses: ὁ ἄνθρωπος οὗτος οὐ παύεται λαλῶν, 6:13; and ὁ δὲ ἔφη, 7:2 (in this 

case we have the article functioning as pronoun, a usage not at all infrequent in Luke–Acts. See BDF 

§251). 

29 ἐλάλει, 6:10; ὑπάρχων … ἀτενίσας εἰς τὸν οὐρανὸν εἶδεν … 7:55; εἶπεν, ἰδοὺ θεωρῶ … 7:56; θεὶς δὲ τὰ γόνατα ἔκραξεν … καὶ τοῦτο εἰπὼν ἐκοιμήθη, 7:60. 

30 ἐξελέξαντο Στέφανον, 6:5; συζητοῦντες τῷ Στεφάνῳ, 6:9; ἀκηκόαμεν αὐτοῦ, 6:11; συνήρπασαν αὐτόν, 6:12; ἀκηκόαμεν γὰρ αὐτοῦ, 6:14; ἐλιθοβόλουν τὸν Στέφανον, 7:59. 

31 These two clauses are Στέφανος δὲ … ἐποίει τέρατα καὶ σημεῖα (Stephen … used to do signs and wonders, 6:8), and θεὶς δὲ τὰ γόνατα (he fell on his knees, 7:60), which Stephen does as he is being 

stoned! 

171  

under my ‘non-subject participant’ category: ‘they chose Stephen’ (6:6) and ‘they stoned Stephen’ (7:59).

The cumulative effect of these choices made by the writer in the transitivity system of Greek amounts to what we might call the incapacitation of Stephen through linguistic means. In the 16 verses discussed above, Stephen is consistently found in situations where he appears to be carried along by events, rather than effecting them in any way; he is a patient victim, who in the face of opposition, arrest and deadly force responds only with words. Lastly, Stephen’s role as a ‘persecuted one’, described in terms reminiscent of Christ’s passion,32 is further underlined by his relationship to the Holy Spirit in this story. As Sylva has noted,33 the strong invective of 7:51–53 includes the only instance in which the Spirit is mentioned in reference to someone other than Stephen: σκληροτράχηλοι καὶ ἀπερίτμητοι καρδίαις καὶ τοῖς ὠσίν, ὑμεῖς ἀεὶ τῷ πνεύματι τῷ ἁγίῳ ἀντιπίπτετε ὡς οἱ πατέρες ὑμῶν καὶ ὑμεῖς (7:51). The other four references to the Holy Spirit, both before and after the invective, consistently describe Stephen as being full of the Holy Spirit (6:3; 6:5; 7:55) or speaking in or by the Spirit (6:10). The message is clear: those who have consistently opposed the Holy Spirit in Moses and the prophets will naturally continue opposing Him in Stephen, for as he is true heir of the saints of old, Stephen’s accusers are true children of their murderous ancestors.34 b. Abraham Given that Stephen begins his speech with Abraham, and roughly seven verses are taken up with his career, the exegete is immediately tempted to consider the patriarch a central character in the structure of the speech, on a par with Joseph and Moses.35 Upon closer examination of participant reference and clause structure, however, a rather different picture emerges:

Participant reference to Abraham: Acts 7:2–8

1. Full explicit subject 0×

2. Abbreviated explicit subject 0×

3. Non-explicit subject36 5×                                                            32 For the parallels between Stephen and Christ in this story, see Robert F. O’Toole, The Unity of 

Luke’s Theology (Wilmington, DE: Michael Glazier, 1984), pp. 63–67. 

33 Sylva, ‘Acts 7:46–50’, pp. 273–74. 

34 Sylva seems wholly justified in seeing vv. 51–53 as a direct reply to the charges leveled against 

Stephen: ‘the function of Acts 7:51–53 is to claim that the false accusations leveled against Stephen 

are but another example of the Jewish resistance to God’s message because the Holy Spirit is linked 

to Stephen in Acts 6:1–8:1, and because Stephen’s claim that the Jews, like their fathers, always 

resist the Holy Spirit immediately follows Stephen’s response to the false accusations leveled against 

him’ (‘Acts 7:46–50’, p. 274). 

35 See, for example, Colmenero‐Atienza, ‘Hechos’, p. 43; Soards, The Speeches, p. 59; Kilgallen, The 

Stephen Speech, p. 35, though Kilgallen notes (p. 43) that ‘God is the one who brings about all that 

happens to Abraham …’ 

36 Twice in dependent clauses: πρὶν ἢ κατοικῆσαι αὐτὸν ἐν Χαρράν (7:2), and ἐξελθὼν ἐκ γῆς Χαλδαίων (7:4); three times in independent clauses: κατῴκησεν ἐν χαρράν (7:4), ἐγέννησεν τὸν Ἰσαὰκ καὶ περιέτεμεν αὐτόν (7:8). 

172  

4. Non-subject participant37 7× Two features of the above scale are particularly striking. First, Abraham is never referred to by means of an explicit subject, one of the most common ways of highlighting a character in Greek.38 Secondly, in the syntax of these six verses Abraham occupies primarily the slot of complement (direct or indirect object) in clauses where the subject is for the most part God, who appears three times as full explicit subject,39 seven times as non-explicit subject40 and once as a non-subject participant.41 It seems clear, then, that in this brief overture to Stephen’s particular history of Israel it is the ‘God of glory’,42 rather than Abraham, who occupies center stage. The writer thus emphasizes the role of Abraham’s God as the great initiator of Israel’s life as a nation, a fact he conveys with great force by means of two ergative clauses in 7:4 and 7:8.43 Once this fundamental point is made, the author is ready to focus in on the individual careers of Joseph and Moses: ‘And [God] gave him the covenant, and it was thus that [Abraham] begat Isaac … and Isaac Jacob, and Jacob the twelve patriachs.’ c. Joseph

Participant reference to Joseph: Acts 7:9–14, 18

1. Full explicit subject44 2×

2. Abbreviated explicit subject 0×

                                                            37 ὁ θεὸς τῆς δόξης ὤφθη τῷ πατρὶ ἡμῶν Ἀβραάμ (7:2), εἶπεν πρὸς αὐτόν (7:3), σοι δείξω (7:3), μετῴκισεν αὐτὸν εἰς τὴν γῆν ταύτην (7:4), οὐκ ἔδωκεν αὐτῷ κληρονομίαν (7:5), ἐπηγγείλατο δοῦναι αὐτῷ (7:5), ἔδωκεν αὐτῷ διαθήκην (7:8). 

38 See Porter, ‘Word Order’, p. 200. 

39 ὁ θεὸς τῆς δόξης ὤφθη τῷ πατρὶ ἡμῶν (7:2); ἐλάλησεν δὲ οὕτως ὁ θεός (7:6); and ὁ θεὸς εἶπεν 

(7:7). 

40 εἶπεν πρὸς αὐτόν (7:3a); εἰς τὴν γῆν ἣν ἄν σοι δείξω (7:3b); μετῴκισεν αὐτόν (7:4); οὐκ ἔδωκεν αὐτῷ κληρονομίαν (7:5a); ἐπηγγείλατο δοῦναι αὐτῷ (7:5b); κρινῶ ἐγώ (7:7); ἔδωκεν αὐτῷ διαθήκην περιτομῆς (7:8). 

41 λατρεύσουσίν μοι (7:7). 

42 This expression occurs one other time only in the Old Testament: Ps. 28:3 in the LXX or Ps. 29:3 in 

the MT (אל־הכבוד). 

43 μετῴκισεν αὐτὸν εἰς τὴν γῆν ταύτην εἰς ἣν ὑμεῖς νῦν κατοικεῖτε (7:4); ἔδωκεν αὐτῷ διαθήκην περιτομῆς (7:8). See also Nils A. Dahl, ‘The Story of Abraham in Luke–Acts’, in Leander E. Keck and J. 

Louis Martyn (eds.), Studies in Luke–Acts (London: SPCK, 1976), pp. 139–58, esp. pp. 143–44, where 

Dahl argues that the remainder of the speech is fundamentally an account of the fulfillment of God’s 

word to Abraham. 

44 ἀνεγνωρίσθη Ἰωσὴφ τοῖς ἀδελφοῖς αὐτοῦ (7:13); Ἰωσὴφ μετεκαλέσατο Ἰακὼβ τὸν πατέρα αὐτοῦ 

(7:14). 

173  

3. Non-explicit subject45 1×

4. Non-subject participant46 5× Although in the above clauses Joseph appears for the most part in the complement slot (i.e. as a non-subject participant), he is referred to four times by name, two of which are explicit subjects. The writer begins his account of Joseph by detaching him from the rest of the patriarchs in the following manner: ‘and Isaac [begat] Jacob, and Jacob the twelve patriarchs’ (7:8). ‘And the patriachs were jealous of Joseph and sold him into Egypt. But God was with him’ (7:9). I wish to suggest that the verses that follow are primarily an elaboration of this theme (a point argued in a slightly different manner by Richard47) and may be summarized aptly as follows:

(i) Adversity (7:9a) (ii) Blessing (7:9b–10) (iii) Adversity (7:11–12) (iv) Blessing (7:13–16)

The betrayal of Joseph by his brothers, expressed in the strongest terms, is followed by an equally strong affirmation of God’s presence and blessing in his life, a blessing that results in salvation for Joseph’s family at the time of their greatest need. The christological allusions of this passage cannot be missed.48 d. Moses

Participant reference to Moses: Acts 6:14; 7:20–40, 44

1. Full explicit subject49 7×

2. Abbreviated explicit subject50 8×                                                            45 In a dependent clause: ἀποστείλας (7:14). 

46 οἱ πατριάρχαι ζηλώσαντες τὸν Ἰωσήφ (7:9); ἐξείλατο αὐτὸν ἐκ πασῶν τῶν θλίψεων αὐτοῦ (7:10); ἔδωκεν αὐτῷ χάριν καὶ σοφίαν (7:10); κατέστησεν αὐτὸν ἡγούμενον (7:10); ἀνέστη βασιλεὺς ἕτερος [ἐπʼ Αἴγυπτον] ὃς οὐκ ᾔδει τὸν Ἰωσήφ (7:18). 

47 See Richard, ‘The Polemical Character’, pp. 258–67. Richard’s essay has shown beyond reasonable 

doubt the intentionality of the writer’s choice of words in 7:9: both ζηλώσαντες and ἀπέδοντο, 

predicated of Joseph’s brothers and without further comment on the events that preceded this 

development in the Genesis account, serve to emphasize the victimization of Joseph at the hands of 

his brothers. 

48 See Kilgallen, The Stephen Speech, p. 52. 

49 ἃ παρέδωκεν ἡμῖν Μωϋσῆς (6:14); ἐγεννήθη Μωϋσῆς (7:20); ἐπαιδεύθη Μωϋσῆς (7:22); ἔφυγεν δὲ Μωϋσῆς (7:29); ὁ δὲ Μωϋσῆς ἰδὼν ἐθαύμαζεν (7:31); ἔντρομος δὲ γενόμενος Μωϋσῆς (7:32); ὁ γὰρ Μωϋσῆς οὗτος … οὐκ οἴδαμεν τί ἐγένετο αὐτῷ (7:40). 

50 ὃς ἀνετράφη μῆνας τρεῖς (7:20); ἐκτεθέντος δὲ αὐτοῦ (7:21); μὴ ἀνελεῖν με σὺ θέλεις (7:28); προσερχομένου δὲ αὐτοῦ κατανοῆσαι (7:31); οὗτος ἐξήγαγεν αὐτούς (7:36); οὗτός ἐστιν ὁ Μωϋσῆς (7:37); οὗτός ἐστιν ὁ γενόμενος (7:38); ὃς ἐξήγαγεν ἡμᾶς ἐκ γῆς Αἰγύπτου (7:40). 

174  

3. Non-explicit subject51 16×

4. Non-subject participant52 15× The size of the Moses section of the speech (it takes up roughly 35% of the total, vs. 11.6% for the Joseph section), together with its careful chronological structure (three ‘forty year’ periods) and the ‘more than human dignity’53 with which its protagonist is invested, reveal something of the centrality of the figure of Moses in Stephen’s oration. Far from being a ‘neutral’ account of the great Old Testament hero’s career, however, this is the climactic point of the writer’s unique history of Israel, carefully crafted to cohere with and consummate all that has preceeded it. Like Joseph in the preceding section, Moses is described in terms reminiscent of Christ.54 As was the case with both Stephen and Joseph above, the reader is left with the impression that Moses is hardly in control of the events of his life, but rather seems often overwhelmed by them. This impression is created by the syntax of this section: of the 31 clauses in which Moses appears as a subject (explicit or non-), only 5 have him as an agent external to the process in question.55 In other words, in all clauses but 5, the processes in which Moses is involved are represented as self-engendered, as in ἔφυγεν δὲ Μωϋσῆς (process + medium, 7:29), rather than engendered by himself, as in πατάξας τὸν Αἰγύπτιον (agent + process + medium, 7:24).56 Lastly, as he did with Joseph in the previous paragraphs, the writer focuses on one specific aspect of Moses’ life, to the detriment of almost everything else: Moses, rejected by men, but blessed and chosen by God: ‘This very Moses, whom they denied … even this man God has sent as both ruler and deliverer …’ (7:35).

The question asked of Stephen by the high priest in 7:1 was simply this: ‘Are these things so?’ that is, are you blaspheming against Moses and God? (6:11), are you indeed announcing that Jesus will alter the traditions we have received from Moses, and what concerns the

                                                            51 ἦν ἀστεῖος τῷ θεῷ (7:20); ἦν δὲ δυνατός (7:22); ἰδών τινα ἀδικούμενον (7:24); ἠμύνατο (7:24); ἐποίησεν ἐκδίκησιν (7:24); πατάξας τὸν Αἰγύπτιον (7:24); ἐνόμιζεν δὲ συνιέναι τοὺς ἀδελφούς (7:25); ὤφθη αὐτοῖς μαχομένοις (7:26); συνήλλασσεν αὐτοὺς εἰς εἰρήνην (7:26); εἰπών (7:26); ἀνεῖλες ἐχθὲς τὸν Αἰγύπτιον (7:28); ἐγένετο πάροικος (7:29); ἐγέννησεν υἱοὺς δύο (7:29); ἰδὼν (7:31); οὐκ ἐτόλμα κατανοῆσαι (7:32); ποιήσας τέρατα καὶ σημεῖα (7:36). 

52 ἀνείλατο αὐτόν (7:21); ἀνεθρέψατο αὐτόν (7:21); ὡς δὲ ἐπληροῦτο αὐτῷ (7:23); ἀπώσατο αὐτόν (7:27); τίς σε κατέστησεν (7:27); ὤφθη αὐτῷ (7:30); εἶπεν δὲ αὐτῷ ὁ κύριος (7:33); ἀποστείλω σε (7:34); τοῦτον τὸν Μωϋσῆν … ὁ θεὸς [καὶ] ἄρχοντα καὶ λυτρωτὴν ἀπέσταλκεν (7:35); ὃν ἠρνήσαντο (7:35); τοῦτον ὁ θεὸς … (7:35); ἀγγέλου τοῦ ὀφθέντος αὐτῷ (7:35); τοῦ ἀγγέλου τοῦ λαλοῦντος αὐτῷ (7:38); ᾧ οὐκ ἠθέλησαν ὑπήκοοι γενέσθαι (7:39); τί ἐγένετο αὐτῷ (7:40); ὁ λαλῶν τῷ Μωϋσῇ (7:44). 

53 Barrett, Acts, p. 338. Similarly, Munck has written that in Stephen’s speech one finds ‘the highest 

appreciation of Moses that we meet in the New Testament’ (Acts, p. 221 n. 1). 

54 Cf., e.g., 7:20–22 with Luke’s account of the birth and childhood of Jesus in Lk. 2. 

55 Of these five clauses where Moses is an agent, three refer to the same event, namely his killing of 

the Egyptian: μὴ ἀνελεῖν με σὺ θέλεις (7:28); ἐποίησεν ἐκδίκησιν (7:24); πατάξας τὸν Αἰγύπτιον 

(7:24). 

56 See Halliday, Functional Grammar, p. 147. 

175  

temple in particular? (6:13–14).57 I wish to argue that Stephen’s speech, culminating so poignantly with the Moses account, represents a strong and well-argued reply to these charges.58 There are only two kinds of Israelite, argues the martyr, God’s faithful envoys together with those who submit to their message, and those who, throughout the history of the nation, have opposed, persecuted and murdered them. The high praise that Stephen directs at Moses and the law that was given through his mediation (7:38, 44) makes clear where he stands with respect to this great divide: It is not he who has rejected Moses and the law, but ‘your fathers’ (7:39), of whom Stephen’s audience, in their prejudiced and malevolent opposition to him, are showing themselves to be true heirs.

The Moses theme leads to the disquisition on the temple in vv. 44–50. In this regard, Barrett is incorrect when he affirms that, in the speech, ‘the temple is treated with no respect at all’.59 Quite the contrary, Stephen begins his brief commentary on Israel’s place of worship by connecting it, in its original form, to Moses himself (7:44), to whom the command to build (ποιῆσαι) such a structure was originally given. Secondly, David’s desire to find (εὑρεῖν) a temple (σκήνωμα60) is described approvingly by the writer, and it is only Solomon’s construction (οἰκοδόμησεν) of a house (οἶκον) that receives the harsh condemnation of the writer. A sounder reading of this section of the speech is offered by Kilgallen61 and Sylva,62 for it is a particular view on the temple, rather than the temple per se, that is under attack in this section of the speech. The ‘Most High’ affirms the martyr, using a divine name that highlights God’s transcendence, which is not contained by structures made with hands (χειροποιήτοις, v. 48), for he himself is the maker of all things. The writer’s account of Israel’s worship progresses as follows:

                                                            57 Sylva is on target when he argues that vv. 6:13–14 are an elaboration and specification of the one 

fundamental charge, namely to have spoken blasphemies against Moses and God. See Sylva, ‘Acts 

7:46–50’, p. 268. Similarly, Barrett sees the charges of 6:13–14 as a second, more formal stage of the 

accusation, rather than a separate indictment. Furthermore, he argues that ‘to speak against Moses 

and to speak against the law are … synonymous’ (Acts, p. 327). 

58 Supporting this view see also Kilgallen, The Stephen Speech, esp. pp. 107–19; Sylva, ‘Acts 7.46–50’, 

p. 263; F.F. Bruce, The Acts of the Apostles (London: Tyndale, 4th edn, 1956), pp. 160–61. Against this 

view see Colmenero‐Atienza, ‘Hechos’, pp. 38–39; Haenchen, Acts, p. 288; F.J. Foakes‐Jackson, The 

Acts of the Apostles (London: Hodder & Stoughton, 6th edn, 1945), p. 61. Barrett’s may be 

considered an intermediate position regarding this fundamental issue in the study of the speech, 

since he argues that the speech is most likely ‘a qualified kind of answer’ to the charges (Acts, p. 

335). 

59 Barrett, Acts, p. 338. See also Haenchen, Acts, p. 290. The basic problem with this view is that while 

matters related to the law did concern early Christian communities in cities such as Jerusalem and 

Antioch, there is no historical evidence to suggest that the temple was a point of similar controversy. 

See Hill, ‘Division or Diversity?’, p. 143. 

60 The term is used in the Septuagint of both the tabernacle and the temple. See Sylva, ‘Acts 7:46–50’, 

p. 264. 

61 Kilgallen, The Stephen Speech, pp. 87–92. 

62 Sylva, ‘Acts 7:46–50’. See also Dahl, ‘The Story of Abraham’, pp. 145–46. 

176  

• False worship by the fathers involving a σκηνή (7:43) • True worship by Moses and David involving σκηνή and σκήνωμα (44–46) • False worship by Solomon involving οἶκος (47–50) • The invective: you are just like your fathers: they killed the prophets and did not keep the

law (51–53) Thus, the thematic links of this brief section to both the previous material and the invective are clear: failure to understand the true purpose and meaning of the temple is presented as one last defining characteristic of those who persecuted the prophets63 and did not keep the law.

When presenting arguments for the literary unity of the Stephen episode, or for its unity in relation to the rest of Acts, it has been customary since Harnack’s time to focus primarily on lexis. Yet, Harnack notwithstanding, lexical features alone may not be considered ‘proof’ of unity of style and cohesiveness in Luke–Acts, as the famous debate between Harnack and Cadbury showed.64 Without denying the relative value of lexis in determining cohesion in our text or in Luke–Acts as a whole, I have turned my attention to other linguistic features of the Stephen story that seem to be both consistent, and purposefully placed, and may therefore be considered contributive to its texture.65 Specifically, the transitivity choices made by the writer in the description of Stephen, Joseph and Moses, the consistency of clause structure (I shall have more to say on this in the next section), and the repeated theme of ‘human opposition-divine blessing’ make the literary unity of the Stephen episode very difficult to dispute. A method of participant reference and clause structure analysis such as the one offered here allows us to record these observations in a manner that is concrete, linguistically principled, and replicable by other scholars using the same method. 3. Foregrounded Syntax in Acts 7:35: Its Relation to the Speech, the Episode and the Book of

Acts

In the previous section I have attempted to show that the literary structure of the Stephen episode centers around the biographical sketches of Stephen, Joseph and Moses. More specifically, I pointed out that the writer is focusing on the status of these three men as relatively passive victims of unrighteous opposition, who nonetheless bear the seal of God’s approval and are blessed and chosen by him. I now turn my attention to a particular passage in

                                                            63 This is to be expected, since one of the most recurrent themes in the prophetic writings is precisely 

to warn Israel against placing her trust in the physical structure of the temple, while at the same time 

neglecting the commandments. See esp. Jer. 7:4–15. 

64 Harnack concluded his lexical analysis with the words ‘the proof is now complete’ (Luke the 

Physician, p. 81). But see Cadbury’s reply in, among others, The Making of Luke–Acts (London: SPCK, 

3rd edn, 1968). For the problems inherent in ‘counting words’ in order to prove cohesiveness, 

authorship or unity of style, see G.U. Yule, The Statistical Study of Literary Vocabulary (Cambridge: 

Cambridge University Press, 1944), esp. Chapter 6, ‘Word Distributions from Different Works of the 

Same Author: Macaulay and Bunyan’, esp. p. 133. 

65 In Hallidayan functional grammar, texture, or the cluster of properties that separate a text from 

non‐text, involves three essential elements: cohesion (both grammatical and lexical), theme and 

information systems (pertaining to the internal structure of the clause), and the macrostructure of 

the text or its genre, such as narrative or conversation. See M.A.K. Halliday and R. Hasan, Cohesion in 

English (London: Longman, 6th edn, 1984), esp. pp. 324–28. 

177  

this story, namely 7:35, that is used by the writer to culminate and bring this theme to the fore with unparalleled impetus and clarity.

τοῦτον τὸν Μωϋσῆν, ὃ ἠρνήσαντο εἰπόντες, τίς σε κατέστησεν ἄρχοντα καὶ δικαστήν; τοῦτον ὁ θεὸς [καὶ] ἄρχοντα καὶ λυτρωτὴν ἀπέσταλκεν σὺν χειρὶ ἀγγέλου τοῦ ὀφθέντος αὐτῷ ἐν τῇ βάτῳ. Two features of this passage, missed almost entirely by the major commentaries, call for some detailed discussion. a. The Adjectival-Demonstrative Use of οὗτος in Acts Haenchen,66 Bruce67 and Barrett68 have all argued that οὗτος is ‘emphatic’ in 7:35–38. Unfortunately, however, such an affirmation does not carry conviction for two fundamental reasons: no linguistic arguments are offered to back such an assertion (see my comments under section 1 above), and no proper differentiation is made between the adjectival use of the pronoun in v. 35, and its nominal use in vv. 36–38. Yet, this distinction is significant enough to merit separate treatment. In its nominal usage, the demonstrative pronoun οὗτος functions most often as an anaphoric and, less often, cataphoric referential tie. That is, it points, without mentioning it esplicitly, to a referent present either before or after in the text, as in Acts 2:15: οὐ γὰρ ὡς ὑμεῖς ὑπολαμβάνετε οὗτοι μεθύουσιν (for these men are not drunk, as you suppose), or Acts 7:36: οὗτος ἐξῃγαγεν αὐτούς (This man led them out). The adjectival function of οὗτος, however, is the qualification, description or highlighting of the noun to which it is attached. The use of the adjectival demonstrative pronoun οὗτος (henceforth ADP οὗτος) in Acts seems rather peculiar to this book and calls for specific comment.

The 28 instances of the ADP οὗτος when used of human participants in Acts are distributed as follows:

                                                            66 See Haenchen, Acts, p. 282. 

67 Bruce, Acts, p. 171. 

68 Barrett, Acts, p. 364. For Barrett, the ‘emphatic’ use of the pronoun apparently begins in v. 36. 

ADP djectival demonstrative pronoun 

178  

10 qualifying Paul69 6 qualifying the apostles70 5 qualifying Jesus71 2 qualifying Moses72 2 qualifying Paul’s nephew73 2 qualifying other Christians74 1 qualifying Stephen75

                                                            69 ἀπεκρίθη δὲ Ἁνανίας· κύριε, ἤκουσα ἀπὸ πολλῶν περὶ τοῦ ἀνδρὸς τούτου ὅσα κακὰ τοῖς ἁγίοις σου ἐποίησεν ἐν Ἰερουσαλήμ (9:13); τινὲς δὲ καὶ τῶν Ἐπικουρείων καὶ Στοϊκῶν φιλοσόφων συνέβαλλον αὐτῷ, καί τινες ἔλεγον· τί ἂν θέλοι ὁ σπερμολογος οὗτος λέγειν (17:18); καὶ θεωρεῖτε καὶ ἀκούετε ὅτι οὐ μόνον Ἐφέσου ἀλλὰ σχεδὸν πάσης τῆς Ἀσίας ὁ Παῦλος οὗτος πείσας μετέστησεν ἱκανὸν ὄχλον (19:26); ἀκούσας δὲ ὁ ἑκατοντάρχης προσελθὼν τῷ χιλιάρχῳ ἀπήγγειλεν λέγων· τί μέλλεις ποιεῖν; ὁ γὰρ ἄνθρωπος οὗτος Ῥωμαῖός ἐστιν (22:26); ἀναστάντες τινὲς τῶν γραμματέων τοῦ μέρους τῶν Φαρισαίων διεμάχοντο λέγοντες· οὐδὲν κακὸν εὑρίσκομεν ἐν τῷ ἀνθρώπῳ τούτῳ (23:9); τὸν ἄνδρα τοῦτον συλλημφθέντα ὑπὸ τῶν Ἰουδαίων καὶ μέλλοντα ἀναιρεῖσθαι ὑπʼ αὐτῶν (23:27); εὑρόντες γὰρ τὸν ἄνδρα τοῦτον λοιμὸν καὶ κινοῦντα στάσεις πᾶσιν τοῖς Ἰουδαίοις (24:5); καὶ ἀναχωρήσαντες ἐλάλουν πρὸς ἀλλήλους λέγοντες ὅτι οὐδὲν θανάτου ἢ δεσμῶν ἄξίον [τι] πράσσει ὁ ἄνθρωπος οὗτος (26:31); Ἀγρίππας δὲ τῷ Φήστῳ ἔφη· ἀπολελύσθαι ἐδύνατο ὁ ἄνθρωπος οὗτος εἰ μὴ ἐπεκέκλητο Καίσαρα (26:32); πρὸς ἀλλήλους ἔλεγον· πάντως φονεύς ἐστιν ὁ ἄνθρωπος οὗτος (28:4). 

70 τί ποιήσωμεν τοῖς ἀνθρώποις τούτοις; (4:16); εἶπέν τε πρὸς αὐτούς· ἄνδρες Ἰσραηλῖται, προσέχετε ἑαυτοῖς ἐπὶ τοῖς ἀνθρώποις τούτοις τί μέλλετε πράσσειν (5:35); καὶ τὰ νῦν λέγω ὑμῖν, ἀπόστητε ἀπὸ τῶν ἀνθρώπων τούτων καὶ ἄφετε αὐτούς (5:38); αὕτη κατακολουθοῦσα τῷ Παύλῳ καὶ ἡμῖν ἔκραζεν λέγουσα· οὗτοι οἱ ἄνθρωποι δοῦλοι τοῦ θεοῦ τοῦ ὑψίστου εἰσίν (16:17); οὗτοι οἱ ἄνθρωποι ἐκταράσσουσιν ἡμῶν τὴν πόλιν, Ἰουδαῖοι ὑπάρχοντες (16:20); ἠγάγετε γὰρ τοὺς ἄνδρας τούτους οὔτε ἱεροσύλους οὔτε βλασφημοῦντας τὴν θεὸν ἡμῶν (19:37). 

71 οὗτος ὁ Ἰησούς ὁ ἀναλημφθεὶς ἀφʼ ὑμῶν εἰς τὸν οὐρανὸν οὕτως ἐλεύσεται ὃν τρόπον ἐθεάσασθε (1:11); τοῦτον τὸν Ἰησοῦν ἀνέστησεν ὁ θεός, οὗ πάντες ἡμεῖς ἐσμεν μάρτυρες (2:32); καὶ κύριον αὐτὸν καὶ χριστὸν ἐποίησεν ὁ θεός, τοῦτον τὸν Ἰησοῦν ὃν ὑμεῖς ἐσταυρώσατε (2:36); βούλεσθε ἐπαγαγεῖν ἐφʼ ἡμᾶς τὸ αἷμα τοῦ ἀνθρώπου τούτου (5:28); ἀκηκόαμεν γὰρ αὐτοῦ λέγοντος ὅτι Ἰησοῦς ὁ Ναζωραῖος οὗτος καταλύσει τὸν τόπον (6:14). 

72 τούτον τὸν Μωϋσῆν ὃν ἠρνήσαντο εἰπόντες· τίς σε κατέστησεν ἄρχοντα καὶ δικαστήν; τοῦτον ὁ θεὸς [καὶ] ἄρχοντα καὶ λυτρωτὴν ἀπέσταλκεν (7:35); ὁ γὰρ Μωϋσὴς οὗτος, ὃς ἐξήγαγεν ἡμᾶς ἐκ γῆς Αἰγύπτου, οὐκ οἴδαμεν τί ἐγένετο αὐτῷ (7:40). 

73 τὸν νεανίαν τοῦτον ἀπάγαγε πρὸς τὸν χιλίαρχον (23:17); ἠρώτησεν τοῦτον τὸν νεανίσκον ἀγαγεῖν πρὸς σέ (23:18). 

74 σὺ κύριε καρδιογνῶστα πάντων, ἀνάδειξον ὃν ἐξελέξω ἐκ τούτων τῶν δύο ἕνα (1:24); ἦλθον δὲ σὺν ἐμοὶ καὶ οἱ ἓξ ἀδελφοὶ οὗτοι καὶ εἰσήλθομεν εἰς τὸν οἶκον τοῦ ἀνδρός (11:12). 

75 ἔστησάν τε μάρτυρας ψευδεῖς λέγοντας· ὁ ἄνθρωπος οὗτος οὐ παύεται λαλῶν ῥήματα κατὰ τοῦ τόπου τοῦ ἁγίου (6:13). 

179  

In 26 out of the above 28 instances (93%), the ADP οὗτος is used to qualify a central Christian figure or Moses, and it is used exclusively of Christians or Moses. Further, the participants that are singled out for reference by means of ADP οὗτος are already ‘on stage’ or under the spotlight due most often (in 23 out of 28 cases) to their being under investigation or accusation by a hostile audience or tribunal. It is perhaps on the basis of some of these

examples that Blass writes that ‘οὗτος appears to be used in a contemptuous sense (like iste) of a person present: Lk. 15:30’.76 Barrett’s comment on οὗτος in Acts 6:13 and 14 is taken directly from Blass, a grammar he is heavily dependent on.77 But the above list shows that the meaning of the ADP οὗτος in Acts is more complex than Blass and Barrett have supposed. I wish to argue that in Acts this particular form of participant reference is used consistently by the writer to mark either key Christian figures or Moses, who are typically found under unjust persecution. These references occur in the words either of an accuser or opponent, or of a Christian speaker or the narrator for specific comment: this very one(s) whom … you murdered, you bring before me, the fathers rejected … we now preach, you must leave alone, is not worthy of death, etc.

The use of the ADP οὗτος in Acts 7:35 is consistent with the above explanation, and is particularly significant, in light of the fact that it is also used of both Stephen and Jesus in 6:13 and 14 respectively. The ADP οὗτος, therefore, appears to be a linguistic means used by the writer to highlight further the characteristic his three protagonists78 have in common in the Stephen episode: ‘this very one whom men reject God has both chosen and blessed’. b. COMPLEMENT-SUBJECT-PREDICATE: A Strikingly Rare Clause Type in Acts 7:35 In commenting upon 7:35, Barrett writes that ‘the object of ἀπέσταλκεν is brought to the beginning of the sentence for emphasis’,79 once more providing no reasons for such an assertion. All that we can gather from a thorough analysis of clause structure in the Stephen episode and elsewhere in Acts, is that the structure COMP-SUBJ-PRED (τοῦτον τὸν Μωϋσῆν, ὃν ἠρνήσαντο … τοῦτον ὁ θεὸς [καὶ] ἄρχοντα καὶ λυτρωτὴν ἀπέσταλκεν) is very rare, indeed it is the most unusual clause type there is in Acts. In independent clauses, the complement is not likely to come first in either clauses with explicit subject (0 instances in the narrative section

                                                            76 BDF, p. 151 [§290 (6)]. Similarly, Turner writes, somewhat more helpfully: ‘οὗτος is very frequent in papyri and as in earlier Greek refers to someone actually present (often contemptuously, Lk 15:30, ὁ υἱός σου οὗτος, 18:11 …) not necessarily referring to the noun which is nearest, but to the noun 

which is most vividly in the writer’s mind (deictic)’ (Nigel Turner, A Grammar of New Testament 

Greek. III. Syntax [ed. J.H. Moulton; Edinburgh: T. & T. Clark, 1963], pp. 44–45). It is important to note 

that neither Blass nor Turner makes a distinction between the nominal and adjectival uses of οὗτος. 

77 See Barrett, Acts, pp. 327–28, where he writes that ὁ ἄνθρωπος οὗτος in 6:13 ‘gives a derogatory tone to the reference’, and that Ἰησοῦς ὁ Ναζωραῖος οὗτος in 6:14 ‘is contemptuous, as in v. 13’. For 

Barrett’s dependence on Blass, see also Barrett, Acts, pp. 346, 348, 349, 350, 353, 355, 359, 361, 362, 

364, etc. That a commentary published in 1994 be so dependent on a century‐old grammar, largely 

unaffected by the findings of modern linguistics, seems at best difficult to accept. See Stanley E. 

Porter and Jeffrey T. Reed, ‘Greek Grammar since BDF: A Retrospective and Prospective Analysis’, FN 

4.7 (1991), pp. 143–64. 

78 Though Joseph is not referred to by means of the ADP οὗτος, compare 7:35 with 7:9. 

79 Barrett, Acts, p. 363. 

180  

of 6:1–7:1, and 5 [4.8%] in the speech) or those without it (1 in the narrative section [3.5%] and 6 [5.8%] in the speech).80 In isolation from other linguistic or literary data about our text, however, this observation is of limited value. If one wishes to argue, as Barrett does, that a particular clause type is purposely being used by the writer as an indicator of prominence in this story, one must be able to show how that prominence relates to the ‘big picture’ of the text, that is, to the writer’s theme(s) or agenda(s) in the Stephen episode, and in the Acts of the Apostles. In his classic essay on linguistic foregrounding, Michael Halliday has this to say:

Foregrounding, as I understand it, is prominence that is motivated. It is not difficult to find patterns of prominence in a poem or prose text … that stand out in some way, or may be brought out by careful reading; and one may often be led in this way towards a new insight, through finding that such prominence contributes to the writer’s total meaning. But unless it does, it will seem to lack motivation; a feature that is brought into prominence will be ‘foregrounded’ only if it relates to the meaning of the text as a whole [emphasis mine].81

The real significance of Acts 7:35 is not the rarity of its Greek syntax or grammar. Rather, it is that these unusual features of language are used by the writer to highlight and drive home, that is, to foreground, the theme he has been focusing on from the beginning of the Stephen episode. The key protagonist of Stephen’s speech appears once more in the complement slot of the clause: this very Moses whom they denied … even this man God has sent as both ruler and deliverer. The presence of the ADP οὗτος, together with the writer’s choice of clause structure, has the effect of capturing the reader’s attention and bringing the central theme of the Stephen episode to a climax.

The relevance of this theme to the speech, the episode and the book of Acts seems clear. It is relevant to the speech because by means of the feature of ‘rejected by men/chosen by God’ as embodied in the lexico-grammar and explicitly stated at key points of the text, the careers of Joseph and Moses are related and their shared significance explained. It is relevant to the episode as a whole, because in relating Stephen to Joseph and Moses, and his persecutors to theirs, the speech is shown to be a reply to the charges, and the narrative and speech portions are seen to possess both coherence and cohesiveness. Lastly, the theme that we see foregrounded in Acts 7:35 is relevant to the ‘big picture’ of the Acts of the Apostles, because it clearly reflects what appears to be one of the author’s primary concerns, namely to narrate the sovereign unfolding of the ‘plan of God’ (βουλὴ τοῦ θεοῦ) in the face of human and demonic opposition.82

My findings in Acts 6–7 confirm my previous work on Acts 2783 and bear out Hopper and Thompson’s transitivity hypothesis, namely that narrative characters who do not initiate                                                             80 The most frequent clause structure in both speech and narrative is PREDCOMP. The data here 

presented are roughly equivalent to that gathered by myself from the sample texts of chs. 1 and 27 

of Acts. 

81 Halliday, ‘Linguistic Function’, p. 112. 

82 See John T. Squires, The Plan of God in Luke–Acts (SNTSMS, 76; Cambridge: Cambridge University 

Press, 1993). 

83 Gustavo Martín‐Asensio, ‘A Survey of Foregrounding and its Relevance for Interpretation and 

Translation, with Acts 27 as a Case Study’, in Stanley E. Porter and Richard S. Hess (eds.), Translating 

the Bible: Problems and Prospects (JSNTSup, 173; Sheffield: Sheffield Academic Press, 1999), pp. 189–

223. 

181  

events and appear in clauses with a low transitivity coefficient are likely to receive a backgrounded interpretation.84 In Hallidayan terms, participants who are represented consistently as mediums in non-ergative clauses appear to readers as incapacitated or sidelined vis-à-vis any participants represented as agents in ergative clauses.

Having said this, the Acts narrative I have examined thus far is uniquely interesting for two reasons. First, it is the main characters, the veritable heroes of the story, who are being backgrounded through the linguistic means I have described. In the shipwreck narrative, the highly dynamic ‘they’ subject (most often the soldiers and sailors on the ship) appears foregrounded in contrast with the rather passive ‘we’ subject and Paul, yet only in order to underline the absolute futility of their actions, as all their efforts to oppose ‘what is necessary’ fail. The message is clear: though seemingly engaged in the bringing about of events, the soldiers and sailors are in fact as overwhelmed by them as the rest of the ship’s passengers. In the Stephen episode, the account of the martyr’s interrupted career is placed alongside those of Joseph and Moses, in order to highlight the feature held in common by the three: an apparent passivity and helplessness in the face of human opposition that is in each case accompanied by God’s approval and blessing. This leads to my second point. The writer of Acts seems quite content with backgrounding his key characters, without providing a starkly contrasting foreground that would answer the question of causation in each case. This fact continually prompts the reader to ask who or what—if not Jesus, Moses, Paul or the apostles—is the initiator or final cause behind the events recounted in this book. Though an answer is provided explicitly at various points (notably Paul’s words to the crew in 27:21–26, and the highly impacting statement of 7:35 in the Stephen episode), the writer’s forerounding scheme has the effect of creating a measure of suspense and inciting the reader to read on, thus making the fact of God’s sovereign (if often invisible to the human eye) hand all the more impacting.

4. Conclusion

In discussing the major trends in twentieth-century Acts scholarship, and research into the speeches in particular, C.K. Barrett exposes the reductionism that the well-known ‘sources’ versus ‘free composition’ debate has often led to. Instead of the somewhat doctrinaire and necessarily restricted approaches of Bultmann (whom Barrett proposes as the champion of sources) and Haenchen (whom Barrett considers the champion of free composition), Barret has issued a timely call for a greater openness to all relevant methods in the study of Acts.

Although significant ground has been gained in various types of historical study of the Acts, the potential insights that modern linguistics offers remain largely unexplored. This essay is meant partly as a corrective to this limitation in past Acts scholarship. Functional grammar reminds us that in interpreting a text, we are primarily involved in a response to language, and this must encompass lexis, as well as syntax and style in the widest sense. Foregrounding theory, especially when understood within the framework of functional grammar, serves as a link between the lexico-grammatical features of a text, and the effects that those features have upon the reader. As such, it is capable of yielding important clues into the author’s agenda in writing, and thus begin to offer linguistically grounded solutions to such age-old problems as the literary unity (or lack thereof) of the Stephen episode, or the connection (or lack thereof) of the speech to the charges.                                                             84 ‘[T]he likelihood that a clause will receive a foregrounded interpretation is proportional to the 

height of that clause on the scale of transitivity. From the performer’s viewpoint, the decision to 

foreground a clause will be reflected in the decision to encode more (rather than fewer) transitivity 

features in the clause’ (Hopper and Thompson, ‘Transitivity’, p. 284). 

182  

NAKED AND WOUNDED: FOREGROUNDING, RELEVANCE AND SITUATION IN ACTS 19:13–20 Todd Klutz

The growing body of literature in discourse analysis and related disciplines has made available whole new vocabularies for the description and interpretation of texts. Three concepts frequently discussed in this literature are employed below to reinterpret the relationship between Acts 19:13–20 (the story about the humiliation of a group of Jewish itinerant exorcists in Ephesus) and its context.

First of all, an attempt will be made to identify which features of this text exemplify ‘foregrounding’, which is understood here as the special highlighting of certain linguistic features against the background of the rest of the text.1 Secondly, after the text’s most prominent features have been identified, an effort will be made to reconstruct certain aspects of the story’s original ‘context of situation’, which for the purposes of this essay is understood as the complex of environmental factors—both verbal and non-verbal—that shaped the author’s selection of ideas and wordings.2 And thirdly, underlying and permeating the entire discussion is an assumption about why foregrounded items in the text can be interpreted as clues to the text’s original but now predominantly lost context of situation; namely one corollary of the ‘maxim of relation’, which maintains that participants in discourse normally expect each other’s utterances to be relevant to the situation at hand,3 is that any situation a historian might reconstruct for an ancient, written, decontextualized text ought to be one for which the text would have possessed a high degree of relevance.

1. Foregrounding in Acts 19:13–20

Undoubtedly, the numerous instances of foregrounding in this passage could be categorized in more than one way. The scheme of classification implicit in the headings below should therefore be understood as only one of many possible approaches to organizing the relevant data. a. Repetition The mere fact that a particular word or morpheme in a text is repeated, or even repeated often, is not adequate justification for concluding it is foregrounded; for the relative prominence of any given instance of repetition depends on various types of local factors (i.e. features of the context in which the repeated item occurs). Nevertheless, even after local factors are given due consideration, at least two instances of repetition in Acts 19:13–20 invite special attention.

The word Ἰουδαῖος, for instance, occurs three times in this story, functioning as a modifier in vv. 13 and 14, where it characterizes the family of itinerant exorcists as Jewish, and functioning as a substantive in v. 17, where it denotes the Jewish population of Ephesus. Now by itself this frequency of occurrence is scarcely remarkable, but two local considerations cause this repetition to look conspicuous. First, in Luke’s other episodes

                                                            1 For a useful, concise discussion of foregrounding, see R. Fowler, Linguistic Criticism (Oxford: Oxford 

University Press, 1986), pp. 71–74. 

2 For an understanding of ‘situation’ similar to that used here, see J. Lyons, Language, Meaning and 

Context (Fontana Linguistics; London; Fontana, 1981), pp. 205–206. 

3 On the ‘maxim of relation’, see G. Brown and G. Yule, Discourse Analysis (CTL; Cambridge: 

Cambridge University Press, 1983), pp. 32–33. 

183  

dealing with demons or exorcism (e.g. Lk. 4:33–37; 8:26–39; 9:37–43a; Acts 16:16–18), Ἰουδαῖος occurs not once, so that unlike, say, δαιμόνιον, the word is not essential to the genre of the text. And secondly, since the occurrence of Ἰουδαῖος in v. 13 unambiguously identifies the itinerants as Jews, the use of the same term in v. 14 to identify their father (i.e. Scaeva) as a Jew is, in terms of its information value, completely redundant. It is therefore difficult not to conclude that the second occurrence is designed to foreground the Jewishness of these hapless competitors, whose exorcistic caper contrasts sharply with the miraculous accomplishments of Luke’s hero (19:11–12).

To recognize that the itinerants’ Jewishness is foregrounded, however, is not to explain why this is so. As to this latter question, one might be tempted to accuse Luke of anti-Semitism if Luke himself were not so insistent on the Jewish origins and loyalties of his heroes.4 Fortunately, better explanations—and certainly less anachronistic ones—can be found. And one worth mentioning here is that Luke’s desire to lampoon Jewish charismatics in particular might well reflect a situation in which such persons constituted both his stiffest competition and the most important cause of the doubts and uncertainties experienced by his Christian audience; the strengths of this explanation will become increasingly apparent below, especially in my treatment of the text’s situational context.

Like Ἰουδαῖος, the form ὄνομα occurs three times in this episode, twice as a noun (vv. 13, 17) and also in the cognate verb ὀνομάξειν (v. 13). In v. 13 the Jewish itinerants try ‘to name the name of the Lord Jesus’ over those possessed by evil spirits; and in v. 17, due to spreading reports of the itinerants’ failure, ‘the name of the Lord Jesus’ is magnified by the whole city of Ephesus. Now because ὄνομα, not surprisingly, has occurred in a couple of Luke’s earlier narratives of exorcism (Lk. 8:30; Acts 16:18), its appearance in the present story hardly seems abnormal. Nevertheless, two considerations encourage the reader to attribute special salience to this usage. First, in none of Luke’s previous episodes of exorcism does ὄνομα occur more than once. And secondly, the three occurrences of ὄνομα in Acts 19:13–20, when considered in the light of other features of this same episode, beg to be seen as a component of a larger complex of semantically interrelated items, that is, as part of a collocational pattern which also includes the appellations Ἰησοῦς (vv. 13, 15, 17), Παῦλος (vv. 13, 15) and Σκευᾶς (v. 14). In this light it is not merely ὄνομα that is being foregrounded, but a much larger, socioreligious issue concerning whose name possesses enough power to protect people from the demons that haunt the universe of early Judaism and Christianity; indeed, as will become evident below, this repetition of ὄνομα, together with the references to the specific names mentioned above, realizes a higher level semiotic consisting of socioreligious and economic competition between the author of this story and his non-Christian Jewish opponents. b. Parallelism By highlighting the semantic equivalence (or opposition) of two or more elements in a text, the linguistic phenomenon of parallelism often contributes to foregrounding.5 Although this type of feature occurs far less frequently in prose narrative than in poetic texts, it is by no means foreign to writings such as Luke–Acts. Indeed, when such features are encountered in works of prose, they can be extremely arresting, precisely because they deviate from the

                                                            4 See especially R. Brawley, Luke–Acts and the Jews: Conflict, Apology, and Conciliation (SBLMS, 33; 

Atlanta: Scholars Press, 1987), pp. 68–83; and M. Salmon, ‘Insider or Outsider? Luke’s Relationship 

with Judaism’, in J.B. Tyson (ed.), Luke–Acts and the Jewish People (Minneapolis: Augsburg, 1988), 

pp. 76–83 (79–81). 

5 Fowler, Linguistic Criticism, pp. 71–72. 

184  

stylistic norms embodied in their co-text. In the present story there is only one noteworthy instance of this phenomenon; but in view of its weighty contribution to the overall force of the narrative, it deserves far more attention than it has received from most of the commentators who have discussed this unit.6

The parallelism is found in v. 15, where, in response to the exorcistic edict pronounced by the itinerants, the evil spirit responds:

τὸν μὲν Ἰησοῦν γινώσκω καὶ τὸν Παῦλον ἐπίσταμαι, ὑμεῖς δὲ τίνες ἐστέ;

Several features of this construction merit comment. First of all, the first two lines stand in synonymous parallelism to one another, highlighting correspondences not only between γινώσκω and ἐπίσταμαι but also between Ἰησοῦν and Παῦλον. Secondly, the propositional agreement between these two lines is strengthened by their phonaesthetic and metrical qualities: the repetition of the masculine article τόν, the phonemic correspondence between the masculine accusative singular endings of Ἰησοῦν and Παῦλον, and the conformity of each line to an eight-syllable measure powerfully reinforce the impression of unity and harmony.7 And finally, while the third line diverges from the first two both metrically (it has seven syllables rather than eight) and grammatically (it is a question rather than declaration), it still ought to be taken with the other two as part of a unified tristich; for the divergence just noted does not represent a merely ordinary or arbitrary difference, but materially reinforces a semantic antithesis between the third line and the previous two. To put this another way, since the demon’s discourse implies not only an equivalence between Paul and Jesus (lines one and two) but also an opposition between the Jesus-Paul solidarity and the itinerants (line three), and since this combination of equivalence and antithesis is engendered not only by the propositional sense of the words but also by the distinctive combination of formal agreements and disagreements between all three lines, the speech of the demon ought to be read as a single tristichic unit.8

Viewed strictly within the boundaries of this brief story, these lines scarcely inspire a confident reconstruction of the situational factors that originally conditioned the text. However, when seen in the light of certain prominent themes found elsewhere in Luke–Acts, these same lines plead to be understood as part of a larger scheme of macro-structural

                                                            6 It is passed over without comment, e.g., in H. Conzelmann, Acts of the Apostles (Hermeneia; 

Philadelphia: Fortress Press, 1987), p. 164; and F.F. Bruce, The Book of Acts (NICNT; Grand Rapids: 

Eerdmans, rev. edn, 1988), pp. 368–69. 

7 Regarding the measure of these lines, it is important to remember that Greek poetry, unlike English, 

produces rhythm not by means of accent but rather by means of metre, that is, the actual quantity of 

syllables in the verse; see W.W. Goodwin, A Greek Grammar (London: Macmillan, rev. edn, 1894), pp. 

348–49. 

8 This impression of identification and opposition is strengthened still further when the sorts of 

processes represented by the verbs in these three lines are taken into consideration; while both of 

the verbs in the first two lines represent processes of the ‘mental’ variety, the verb in the third line 

represents a ‘relational process’. For further discussion of the various categories of processes 

represented by verbs, see M.A.K. Halliday, An Introduction to Functional Grammar (London: Edward 

Arnold, 1985), pp. 102–31. 

185  

parallelism and foregrounding that offers one of the best clues we have to the Lukan writings’ situational context; to be more precise, in Acts 19:15 both the concord between Jesus and Paul and the discord between this duo and the itinerants reinforces Luke’s repeated efforts (1) to create images of continuity and harmony between Jesus and the Christian heroes of Acts, and (2) to contrast this harmonious band of protagonists with non-Christian intermediaries and ‘magicians’.9 As to what these emphases might suggest about the Lukan writings’ original context of production, more will be said below, under situational context. c. Verbal Aspect Recent study of the tense-forms in New Testament Greek has emphasized that differences in verbal aspect can (and often do) correspond to differences in semantic prominence.10 More specifically, S.E. Porter has suggested that most selections of tense-form in Hellenistic Greek can be interpreted as signifying one of three alternative degrees of prominence, the aorist, for instance, functioning as a ‘background’ tense, the present and the imperfect serving as ‘foreground’ tenses, and the tense-forms that are stative in aspect (i.e. the perfect and the pluperfect) functioning as ‘frontground’ devices.11 While this hypothesis has not yet won universal acceptance among authorities on the language of the New Testament, it certainly merits further testing and therefore serves as a tool of experimentation in the ensuing analysis.

Of the 155 words that make up Acts 19:13–20, 35 have a tense-form. Of this latter group, 13 are in the aorist, each of the 13 appearing on the lips of the narrator, and all but one (τῶν πραξάντων, v. 19) presenting narrative events in the order of their imagined occurrence. In this unit, then, as in many others, the aorist helps to create a backdrop against which other features of greater significance can be spotlighted or emphasized. As this pattern of usage is typical of the aorist’s backgrounding function, however, this set of verbs requires no further comment.

The verbs in the aorist having been accounted for, there still remain 22 verbs that have potential to highlight the processes or participants they denote. And since a text that foregrounded everything would be a text that foregrounded nothing, not all of the remaining 22 are likely to indicate special prominence. Three of these, in fact, can be dismissed on account of their aspectual vagueness: the verb εἰμί, which occurs once in the present (v. 15) and twice in the imperfect (vv. 14 and 16), is one of a small group of verbs that do not have a full set of tense-forms and that therefore cannot encode a meaningful selection between one verbal aspect and another.12

Among the remaining 19 verbs, however, 2 carry so much semantic weight that there can be little doubt about their prominence; in each case, moreover, the special salience is indicated by selection of a perfect tense-form. The first to occur is the participle                                                             9 On the macro‐structural parallels between the heroes of Luke–Acts, see especially W. Radl, Paulus 

und Jesus im lukanischen Doppelwerk: Untersuchungen zu Parallelmotiven im Lukasevangelium und 

in der Apostelgeschichte (Europäische Hochschulschriften, 49; Bern: Lang, 1975), p. 44; R.C. Tannehill, 

The Narrative Unity of Luke–Acts. II. The Acts of the Apostles (Minneapolis: Fortress Press, 1990), pp. 

49–50, 74–77, 162, 237; and D. Moessner, ‘ “The Christ Must Suffer”: New Light on the Jesus‐Peter, 

Stephen, Paul Parallels in Luke–Acts’, NovT 28 (1986), pp. 220–56 (221–27). 

10 S.E. Porter, Idioms of the Greek New Testament (Biblical Languages: Greek, 2; Sheffield: JSOT Press, 

1992), pp. 22–24, 302–303. 

11 Porter, Idioms, pp. 22–24, 302–303. 

12 Regarding aspectually vague verbs, see Porter, Idioms, pp. 24–25. 

186  

τετραυματισμένους (v. 16), which along with γυμνός characterizes the manner in which the overwhelmed itinerants left the site of their humiliation. ‘They fled out of that house’, says the narrator, ‘naked and wounded’ (RSV). By choosing the perfect tense-form where another tense of the same verb or even an adjective could have been selected, the narrator heavily emphasizes that the itinerants not only failed to expel the spirit and help the poor man, but actually lacked power to prevent themselves from falling victim to the malevolent being. Moreover, the itinerants’ affected role in the transitivity structure of τετραυματισμένους—they are the ‘goal’, or the affected object of the wounding, whereas the evil spirit is the implied ‘actor’, or the one who does the deed—echoes their affected position in the previous three clauses (ἐφαλόμενος ὁ ἄνθρωπος ἐπʼ αὐτούς, κατακυριεύσας ἀμφοτέρων, ἴσχυσεν κατʼ αὐτῶν), drawing yet further attention to their impotence vis-à-vis the demon.

The other instance of the perfect occurs in v. 18, where those who confess their entanglement in magical practices are denoted by the substantival participle τῶν πεπιστευκότων. While this occurrence of the perfect undoubtedly places special emphasis on this group of believers, the interpreter may wonder whether this occurrence has the additional effect of implying that the action of believing was antecedent to the actions represented by the main verb and participles that follow (i.e. ἤρχοντο ἐξομολογούμενοι καὶ ἀναγγέλλοντες). The question is of no small importance since, if the perfect does convey this additional meaning, the narrator would be implying that some people in the Christian community at Ephesus had practised ‘magic’ not only before their conversion but also afterwards;13 furthermore, since this same group of clauses (i.e. v. 18), with its relatively complex syntax and progressive verbs (one in the imperfect and two in the present), functions as a species of narrative ‘internal evaluation’ and thus helps to establish the story’s relevance for its audience, the verse could be interpreted as evidence that the receptors of Luke–Acts had continued, like the Ephesians in this episode, to practise ‘magic’ after their conversion and perhaps even to finance those who specialized in the art.14

Although the tense-forms of substantival participles often possess no temporal significance,15 a couple of the features in the co-text of τῶν πεπιστευκότων support the temporal interpretation suggested above. First of all, the amount of attention Luke devotes to opponents who are either implied or asserted to be ‘magicians’ (cf. Acts 8:9–24; 13:4–12) is more than adequate to warrant the conjecture that he considered such figures to be a threat to the faith of his audience, who almost certainly had received some form of Christian instruction before Luke gave them his narratives (cf. Lk. 1:4); in other words, if Luke’s emphasis on ‘magic’ and ‘magicians’ can be assumed to have been relevant to his original audience, then the involvement of Christians in ‘magical’ belief and practice may well have

                                                            RSV Revised Standard Version 

13 Conzelmann, Acts, p. 164, rightly acknowledges this semantic potential of the perfect, but implies 

Luke’s choice at this point was unconscious or inept; as should become evident below, Luke probably 

knew exactly what he was doing, his selection of the perfect having been guided by considerations of 

relevance. 

14 On the various types of ‘internal evaluation’ in narratives and on their role in establishing 

contextual relevance, see M. Toolan, Narrative: A Critical Linguistic Introduction (Interface; London: 

Routledge, 1988), pp. 156, 160–61. (Toolan’s treatment of this topic relies heavily on W. Labov, 

Language in the Inner City [Philadelphia: University of Pennsylvania Press, 1972].) 

15 Porter, Idioms, pp. 187–88. 

187  

been an important social and religious reality in Luke’s milieu. And secondly, in an earlier episode that has a striking number of links to this one, Simon the ‘magician’ is portrayed as having ‘believed’ in the good news prior to his attempt to purchase the apostolic power to bestow the Spirit (Acts 8:13, 18–20);16 the Lukan conception of what it means to ‘believe’, therefore, by no means precludes the possibility of post-conversion entanglement in magic and related practices. In summary, then, as Luke saw ‘magic’ as a prominent menace to his audience and knew of people who, even after receiving the gospel, continued to be enchanted by its charms, his decision to use the perfect tense-form in this context should be understood to frontground—to make as emphatic as possible—the problem which the supramundane events of this story allegedly corrected: after accepting the Christian gospel, a significant number of people in Ephesus had continued to buy and use apotropaic formulae and devices that were incompatible with Luke’s own ideological point of view.

As hinted above, however, this story is not simply a record of facts about an allegedly supernatural occurrence in first-century CE Ephesus; on the contrary, as has been repeatedly implied above, this story appears to have been powerfully conditioned by the constraints of relevance. And as practitioners of pragmatics and discourse analysis have tirelessly demonstrated, statements and propositions—indeed, entire stories and histories—often function as indirect instructions, requests or directions.17 Accordingly, for some if not many of the people in Luke’s audience, this little narrative about the past would have communicated a simple but important set of instructions concerning the present. However implicit, the instructions are clear in the conspicuously idealized reaction of the Ephesian believers to the spreading gossip about the itinerants’ failure: ‘Many of those who had believed’, stresses the narrator, ‘were coming to confess and disclose their magical practices’ (trans. mine, v. 18). The Ephesians’ praiseworthy response is foregrounded, moreover, by the imperfective aspect of ἤρχοντο, ἐξομολογούμενοι and ἀναγγέλλοντες, as if to leave no room for question about what those in Luke’s own audience ought to do.

Several other verbs in this passage are imperfective in aspect and could be understood to highlight one semantic feature of the text or another; but only one of these forms significantly affects the overall force of the story. Following the string of imperfective forms discussed above comes, in v. 19, another narrative clause that functions in the interests of relevance and tellability: ‘A considerable number of those who practised magical arts collected their books and were burning them up in the presence of everyone’ (trans. mine, v. 19). The Greek verb translated here as ‘were burning up’, κατέκαιον, stands out not only because it is in a foreground tense-form (imperfect) but also because it is embedded in a nest of background tense-forms—the two participles that immediately precede it are, like the two verbs that directly follow it, in the aorist.

As in v. 18, therefore, so also here in v. 19 the narrator’s use of the imperfective tense-form(s) results in prominence, heightens interest, and signals relevance; and since these same effects are strengthened still further by means of the exaggerative quantifier πάντων—the flameworthy books were not merely burned up, but burned up publicly, ‘in the presence of

                                                            16 For a nice summary of the links between Acts 19:11–20 and 8:4–24, see S.R. Garrett, The Demise of 

the Devil: Magic and the Demonic in Luke’s Writings (Philadelphia: Fortress Press, 1989), p. 91. 

17 See, e.g., T.A. van Dijk, Text and Context: Explorations in the Semantics and Pragmatics of Discourse 

(Longman Linguistics Library, 21; London: Longman, 1977), pp. 233–46; G.N. Leech, The Principles of 

Pragmatics (Longman Linguistics Library; London: Longman, 1983), pp. 37–40, 97–99; K. Wales, 

‘Indirect Speech Acts’, in idem, A Dictionary of Stylistics (Studies in Language and Linguistics; London: 

Longman, 1989), pp. 247–48; and Toolan, Narrative, p. 227. 

188  

all’—the illocutionary force of this clause could not have been missed by Luke’s original audience: those who had purchased exorcistic incantations or similar formulae that were at odds with Luke’s version of the truth would do well both to recognize the inferiority of their ‘magical’ devices to the power of the ‘Lord Jesus’, and to externalize this recognition symbolically by incinerating their books in public.18 Thus, while we may never know whether this account of book-burnings accurately represents events that took place in ancient Ephesus, we need only a modicum of historical imagination to see that any serious effort on the part of Luke’s audience to imitate the example provided in Acts 19:19 would have deprived Luke’s magicoreligious competitors of a considerable source of revenue and liberated it for Christian uses; undoubtedly, it also would have facilitated the Christian community’s sense of separateness and uniqueness, which, in giving its members an impressive clarity in their self-definition and identity, would have enhanced the new faith’s appeal to many outsiders. d. Word Order In the linear flow of words and clauses that form this unit, only one construction exemplifies a ‘marked’ order (i.e. a sequencing of grammatical elements that is peculiar enough to require explanation). That construction is the last sentence of the story, where the prepositional phrase κατὰ κράτος τοῦ κυρίου is placed at the beginning of its clause. The placement of a prepositional phrase at the head of its clause, of course, is not very unusual in itself; but a number of contextual considerations cause it to stand out in this particular instance. First, an analysis of all the passages in Luke’s Gospel and Acts 1:1–19:9 in which κατά is used with the accusative case indicates that while phrases governed by this preposition not infrequently precede the verbs they qualify (they do so in approximately 40% of the instances),19 this order still has the potential to highlight one feature or another of the prepositional phrase;20 by contrast, in all the instances that exemplify the more common order, neither the prepositional phrase as a whole nor any of its constituents ever appear to be emphasized. Secondly, the separation of κατὰ κράτος τοῦ κυρίου from the verbal construction it qualifies (ηὔξανεν καὶ ἴσχυεν)—between the two is interposed the clause’s grammatical subject, ὁ λόγος—causes special emphasis to fall on the former.21 And thirdly, because the expressed subject ὁ λόγος signals a shift in topic, it might have been expected to stand at the beginning of the clause, so that its unexpected placement after κατὰ κράτος τοῦ κυρίου causes the front position of the latter to stand out still further.22

Therefore, while ‘the Lord’ himself might not be portrayed as materially present in the happenings of the immediately preceding verses, this last clause of the story emphasizes that                                                             18 On the use of exaggerative quantifiers to heighten a story’s contextual significance or relevance, 

see Toolan, Narrative, p. 160. 

19 In the portions of Luke–Acts cited above, there are 72 instances in which κατά with an accusative noun qualifies an expressed verb, 43 of these having the prepositional phrase after the verb and 29 

having it before the verb. 

20 See, e.g., Lk. 6:23, 26; 8:39; 17:30; 21:11; 22:22, 53; Acts 3:17; 13:23, 27; 15:21; 17:11, 22. 

21 On the effect of a word in a Greek clause being separated from the word or words most closely 

related to it, see G. Zuntz, Greek: A Course in Classical and Post‐Classical Greek Grammar from 

Original Texts, II (Biblical Languages: Greek, 4; Sheffield: Sheffield Academic Press, 1994), p. 45. 

22 On the tendency of the expressed subject in Greek to signal a shift in topic, and thus to stand at the 

front of the clause so as to minimize ambiguity, see Porter, Idioms, pp. 295–96. 

189  

both the humiliating failure of the Jewish itinerants and the subsequent growth and victory of the Christian message were consequences of the Lord’s might.23

2. Situational Context

On the basis of the preceding analysis, and as a foundation for the ensuing discussion, the main emphases of Acts 19:13–20 may be summarized as follows: (1) the Jewishness of the itinerants, (2) the inability of the itinerants to manage the evil spirit, (3) the unity of Jesus and Paul, (3) the contrast between the demon’s respectful acknowledgment of Jesus and Paul and his disrespectful treatment of the itinerants, (4) the presence of the Lord’s might in both the humiliation of the itinerants and the consequent spread of the gospel, (5) the socioreligious reality of believers continuing to rely on ideologically inadequate persons and practices to defend themselves against demons, and (6) the Ephesian Christians’ exemplary behaviour in confessing their ‘magical’ practices and burning their ‘magical’ formulae.

Although a comprehensive discussion of this narrative’s situational context cannot be offered in the space available here, something can and should be said about what type of situation this particular combination of emphases would have been relevant to. Before efforts are made to outline the extratextual factors that conditioned this story, however, at least brief attention should be paid to the text’s co-text (i.e. its verbal surroundings, or the larger linguistic event of which it is a part). a. Co-text In attempting to ascertain how the situational meanings of Acts 19:13–20 might be affected by the passage’s co-text, the interpreter would do well to begin by making a simple distinction between two classes of verbal surroundings. One of these consists of the text’s nearest environment, or the words, clauses and sentences that come immediately before and after it. Just prior to the story under consideration, for example, readers encounter several motifs that unquestionably affect their response to the story about Scaeva’s sons. Perhaps the most noteworthy of these is the narration of the spectacular healings and exorcisms God performed through ‘the hands of Paul’ (διὰ τῶν χειρῶν Παύλου), vv. 11–12), whose mighty wonders stand in bold contrast with the ineffectual attempts of the Jewish itinerants.24 This contrast, moreover, is reinforced by another striking difference between the characters: whereas Paul is portrayed as remaining in one location or another for considerable stretches of time (19:8, 10, 22), the sons of Scaeva are implicitly characterized as rambling from one geographical area to another (περιερχομένων, v. 13). And since a mildly pejorative overtone should probably be heard in the description of the Jewish competitors as περιερχομένων, the attitude of the narrator toward the figures of the contrast is hardly a mystery.25

In view of the prominence in v. 18 of believers (τῶν πεπιστευκότων) whose understanding of the Christian message had not been entirely adequate, the repeated use of                                                             23 While the link between the Lord’s power and the victorious spread of the Christian gospel is part of 

the overt message of this clause, the connection between the same power and the humiliating 

failure of the itinerants is conveyed only indirectly, by means of the inferential particle οὕτως, which 

indicates that what follows in v. 20 is based on the events narrated in the preceding co‐text. 

24 Cf. W. Kirschläger, Jesu exorzistisches Wirken aus der Sicht des Lukas: Eine Beitrag zur lukanischen 

Redaktion (ÖBS, 3; Klosterneuberg: Österreichisches Katholisches Bibelwerk, 1981), p. 258. 

25 On the pejorative overtones of περιέρχομαι, see especially 1 Tim. 5:13, where the same term is 

used in a disparaging description of young widows as ‘roaming’ from house to house. 

190  

πιστεύω in the preceding co-text (19:2, 4)—once again, in connection with believers who needed instruction—is also noteworthy. The events narrated in both of these passages, moreover, are explicitly set in Ephesus. To be sure, the shortcomings of the Ephesian believers mentioned in 19:1–7 (they know nothing about the Holy Spirit and have not been baptized in the Holy Spirit) differ from those of the believers pictured in v. 18; but the two episodes in which these characters play a role share a sizeable group of co-occurring lexemes and motifs, the most significant of these (other than πιστεύω) being πνεῦμα (vv. 2, 6, 13, 15, 16), τὸ ὄνομα τοῦ κυρίου Ἰησοῦ (vv. 5, 13, 17), and the impressive authority and power of Paul (vv. 6, 11–12, 15).

Reading vv. 13–20 in the light of vv. 1–7 inevitably generates meanings and effects that would not be produced by reading the former episode in isolation from its co-text. One important effect of this sequencing is the close relationship it creates between the ministry of Paul, the Christian rite of baptism, the presence and activity of the Holy Spirit, and ‘the name of the Lord Jesus’. In view of this network of tightly interconnected motifs, moreover, the reader might infer that the incompetence and humiliation of Scaeva’s sons can be attributed to two particular deficiencies: first, their effort to use Jesus’ name as a source of exorcistic power was neither preceded by nor rooted in the initiatory rite of Christian baptism, which in Luke’s view would have been essential both for an intimate union with ‘the Lord Jesus’ and for a proper use of Jesus’ name; and secondly, since the itinerants’ use of Jesus’ name was not grounded in the proper experiential context, they also would have been understood to lack the Holy Spirit, whose presence and power would have been regarded by Luke as virtually indispensable to apotropaic and exorcistic success (Lk. 11:13, 24–26).

As the society inhabited by the first readers of Luke–Acts undoubtedly valued community opinion more (and intellectual autonomy of the individual less) than most contemporary westerners do, Luke’s references to how large groups of people responded to the words and deeds of his heroes would have had special relevance for his first audience.26 Interpreters should not fail to note, therefore, that Luke’s employment of the phrase Ἰουδαίοις τε καὶ Ἕλλησιν (‘both Jews and Gentiles’) to denote those who heard about the humiliation of Scaeva’s sons and made the right response to it echoes a line from the immediately preceding co-text, where the same construction is used to denote those who accepted Paul’s message in Asia (i.e. everyone, v. 10). Thus, while a reading of vv. 13–20 apart from its co-text would not result in Ἰουδαίοις τε καὶ Ἕλλησιν receiving special prominence, a recognition of the phrase’s occurrence in the immediately preceding unit suggests that it probably has more than minor relevance to Luke’s situation; more specially, Luke’s use of this construction may well be another instance of prescriptive iconicity, or of narrated events communicating indirectly what the author wants the audience to do (i.e. interpret the Christian gospel as a means of transcending ethnic and cultural divisions). This suggestion increases in plausibility, moreover, when occurrences of the same construction in other parts of Acts 14–20 are taken into consideration (14:1; 18:4; 20:21).

Only one other feature of the near co-text significantly impacts the situational interpretation of Acts 19:13–20. Having decided to stay in Asia for a while, the Lukan Paul encounters strong opposition from an association of Ephesian craftsmen whose profits depended heavily on the local popularity of the goddess Artemis (19:22–34). By means of a speech attributed to one of the craftsmen, Luke implies that the opposition to Paul is grounded

                                                            26 On the especially strong tendency of other‐directed, dyadic personalities to imitate the observed or 

narrated actions of others, cf. J. Rockwell, ‘A Theory of Literature and Society’, in J. Routh and J. 

Wolff (eds.), The Sociology of Literature: Theoretical Approaches (Sociological Review Monograph, 

25; Keele, Staffordshire: University of Keele Press, 1977), pp. 32–42 (37–40). 

191  

in one of the basest of motives, the love of money, no small sum of which the craftsmen were in danger of losing on account of Paul’s outstanding success in converting the Ephesian populace (vv. 25–27). Thus, as in vv. 13–20, where the ruination of an apparently lucrative commerce in ‘magical’ texts is tacitly associated with the public disgrace of Scaeva’s sons, so in this subsequent episode Paul’s competitors are portrayed as economic parasites, who in exchange for silver give their deluded hosts the pathetic and unnecessary comforts of false belief.

As hinted at the outset of this section, the co-text of Acts 19:13–20 consists not only of the units immediately surrounding it, but also of Luke–Acts as a whole. Of course it is neither possible nor necessary to explore, within the confines of this study, every meaningful relation between the story about the failure of Scaeva’s sons and this wider level of co-text. Instead, the interpreter might find valuable insights simply by asking what the most significant relations might be. A familiarity with the major patterns and themes of Luke–Acts as a whole, I believe, supports the following suggestions.

First of all, the contrast between Paul’s divinely empowered ministry of healing and the itinerants’ tacit involvement in ‘magic’ recalls a similar contrast, found in Acts 8:4–24, between the healing ministry of Philip and the ‘magical’ activity of Simon the Samaritan. The juxtaposition of Christian miracle and Jewish folly involving the reputation of high priests, moreover, replicates a pattern met back in Acts 5:12–21, where a summary of Peter’s healings and exorcisms is followed immediately by a story about the failure of the high priests and Sadducees to keep the apostles in prison. Furthermore, the emphasis in 19:11–12 on the healing power of Paul’s hands, which highlights by contrast the impotence of the itinerants, parallels the distinctively Lukan reference to the healing power of Jesus’ hands in Lk. 4:40. Thus, in addition to creating the impression that the narrator is concerned about various types of Jewish opposition and competition, this network of parallels strengthens the macrostructural cohesion of Luke–Acts as a whole and reinforces the already prominent message conveyed through the demon in Acts 19:15 (i.e. concerning the unity of Paul and Jesus and the superiority of their reputation to that of the itinerants).

If the features discussed above failed to make the contrast between Luke’s heroes and the itinerants so clear that his audience could not overlook it, the existence of any other co-textual detail that might fortify the contrast surely would have made it unmistakably plain. There is, in fact, another detail that fits this description. As the story in Luke’s Gospel about the demoniac of Gerasa (Lk. 8:26–39) produces echoes in Acts 16:16–18, Luke’s readers should not be surprised to hear reverberations of the same story in Luke’s next episode on exorcism.27 The features in Acts 19:13–20 that might encourage the reader to make this connection are the transformation of the itinerants into naked fugitives, and the ratio of exorcists to evil spirit. To be more precise, whereas Jesus overpowers an entire legion of demons on his own (Lk. 8:30–33) and transforms the naked demoniac into a clothed disciple (Lk. 8:27, 35), the seven itinerants cannot manage a solitary demon, despite their numerical advantage, or even prevent the fiend from wounding them and tearing their clothes off (Acts 19:16).

As the greatness of Luke’s heroes and the impotence of their competitors could hardly be more pronounced, these themes need not receive further attention here. One other topic, however, does merit comment. In the preceding discussion of verbal aspect, it was suggested                                                             27 The most important component of the correspondence is the phrase ‘the Most High God’ (ὁ θεὸς ὁ ὕψιστος), which is placed on the lips of each of the demoniacs in these passages (Lk. 8:28; Acts 

16:17); for further discussion of this connection, see J. Roloff, Die Apostelgeschichte (NTD, 5; 

Göttingen: Vandenhoeck & Ruprecht, 1981), p. 245; and W. Schmithals, Die Apostelgeschichte des 

Lukas (Zürcher Bibelkommentare; Zürich: Theologischer Verlag, 1982), p. 150. 

192  

that the perfect tense participle τῶν πεπιστευκότων highlights an intra-narrative situation that probably possessed special relevance to the extratextual circumstances of Luke’s original audience; but most important for the present discussion, one implication of this suggestion is that Luke’s intended audience, like the Ephesians pictured in Acts 19:18, would have consisted of persons who, while they apparently needed further Christian instruction, had already accepted the basic Christian message. This understanding of the nature of Luke’s audience, though not uncontroversial, coheres impressively with a wide range of semantic data found in this episode’s co-text; but perhaps the weightiest support for this interpretation is that found in the preface to Luke’s first volume, where the words used by Luke to indicate his purpose—ἵνα ἐπιγνῷς περὶ ὧν κατηχήθης λόγων τὴν ἀσφάλειαν (that you might find out for certain about the things in which you have been instructed)—presuppose a reader who has already received training in the faith.28 b. Extratextual Dynamics While the preceding analysis of Acts 19:13–20 by no means provides enough evidence to justify dogmatic claims about the extratextual situation of Luke–Acts as a whole, it does uncover a combination of emphases and meanings that titillate a reconstructive imagination and support some hypotheses more than others. Clearly, for instance, in view of the semantic data collected above, the notion that Luke was addressing outsiders—whether potential converts or pagan political officials—can be laid to rest. Luke almost certainly was addressing an audience that had already made a favourable response to the Christian gospel.

However, in the light of the aforementioned agreement between the purpose clause at the end of the Lukan Gospel’s preface (1:4) and the relevance-enhancing emphasis on the believers who finally stop relying on ‘magic’ (Acts 19:18), the believing audience of Luke–Acts was hardly mature or firm in its faith. To say they were entangled in ‘magic’, of course, would be to accept the value-laden vocabulary of Luke himself, as if ‘magic’ and similar words might be capable of denoting some objective phenomenon distinguishable from Luke’s own religious practices.29 The closest we can come to understanding what lay behind Luke’s rhetoric about magic, probably, is to imagine a network of popular assumptions, ideas, sayings and practices which, despite their blatant similarity to the supernatural aspects of early Christianity, were evaluated by Luke as having closer ties to Christianity’s religious competitors than to Christian faith itself. In other words, Luke’s talk of magic, magicians, magic books and magical practices looks much more like a rhetorical disparagement of people and practices he needs to discredit than like a detached characterization of the phenomena;30 and the reason he needs to discredit them, not just in the present episode but also in Acts 8:4–24 and 13:4–12, is that they symbolize the religious influences that resist the flowering of Christian ideology in his own context.                                                             28 For excellent discussion and defense of this particular understanding of the Lukan preface’s last 

clause, see L. Alexander, The Preface to Luke’s Gospel: Literary Convention and Social Context in Luke 

1:1–4 and Acts 1:1 (SNTSMS, 78; Cambridge: Cambridge University Press, 1993), p. 122. 

29 On the pitfalls and disadvantages of essentialistic interpretations of ancient ‘magic’, see especially 

A.F. Segal, ‘Hellenistic Magic: Some Questions of Definition’, in R. Van den Broek and M.J. 

Vermaseren (eds.), Studies in Gnosticism and Hellenistic Religions (Festschrift G. Quispel; Leiden: E.J. 

Brill, 1981), pp. 349–75 (359, 367–70); J. Speigl, ‘Die Rolle der Wunder im vorkonstantinischen 

Christentum’, ZTK 92 (1970), pp. 287–312 (299); and Garrett, The Demise of the Devil, pp. 4–5, 18–19, 

24–31. 

30 Cf. Garrett, The Demise of the Devil, pp. 4–5. 

193  

The implicit link between the itinerants and ‘magic’, therefore, discloses little more about these sorts of people than that they trafficked in the miraculous and that Luke was against them. His emphasis on their Jewishness, however, is an altogether different matter; for in this detail Luke leaves behind valuable clues concerning whom he needed to discredit, what kinds of people had been preventing his audience from advancing in the Christian faith, and what kinds of things these people had been saying. As for the first two of these questions, since Luke undoubtedly wanted his portraits of the itinerants and similar figures to represent human types his audience would recognize his characterization of them as exemplifying a thoroughly Jewish system of culture and displaying a flare for the miraculous strongly suggests that Luke’s own antagonists had similar traits.

These same two characteristics, however, are also prominent features of the Lukan writings’ heroes. Because many of the fiercest social conflicts arise between groups that have much in common, this similarity should not occasion surprise. Still, the resemblance raises questions. Perhaps most important, in view of the cultural and psychological similarities between the two groups, one might wonder why Luke’s heroes and the Jewish charismatics are opposed to rather than allied with each other. Luke’s own answer, not surprisingly, is coloured by his own ideological point of view: all the opponents of his heroes, Luke repeatedly implies, are energized by Satan and motivated either by love of money or by jealousy of the Lukan heroes’ missionary success.31

On the other hand, hints that might provide material for a less biased answer are available. And in brief, one highly plausible explanation of the conflict is that the opponents have rejected the Lukan heroes’ claims to authority.

As a considerable portion of Luke’s second volume indicates, Luke’s most problematic hero is Paul. Paul’s chief difficulty, moreover, from Luke’s perspective at least, is not difficult to identify. Again and again, the narrator of Acts portrays Paul as the victim of unjustified slander, of dishonest accusations to the effect that he teaches Jews who live among the Gentiles to forsake Jewish custom and that he acts not only against Moses but also against Caesar.32 Those who level charges of this kind, of course, are invariably non-Christian Jews; yet the anti-Paulinist polemic of non-Christian Jews has not failed to have an effect among Jewish Christians. In Acts 21:17–26, for instance, James and the Jewish Christian elders of the church in Jerusalem anxiously refer to a large group of Jewish believers who ‘have been informed’ (κατηχήθησαν) that Paul teaches Jews to abandon Moses and the customs. The passive voice of κατηχήθησαν, along with the deletion of the process’s agent (i.e. the ‘sayer’ or informer), is striking, since its communicative effect is to leave unknown the identity of those spreading the unflattering reports. In view of the patterns of accusation outlined above, however, both the identity of the slanderers and the path of the communication can be ascertained easily enough: non-Christian Jews are telling their Jewish-Christian kinsmen that Paul is an antinomian. A comparable pattern of communication, slander and religious contamination can be discerned in Acts 20:29–30, where the Lukan Paul warns about ‘fierce wolves’ coming into the community and spreading a vicious virus of disaffection.

If, as seems likely, the features summarized above owe as much to the dynamics of Luke’s own situational context as to the historical events he purports to narrate, a hypothesis concerning Luke’s original audience can be offered that might account for a high proportion of the emphases discussed in the preceding pages. In brief, Luke probably was addressing a mixed community of Jewish Christians and god-fearers whose understanding of themselves as members of the one messianic people of God was being severely tested by non-Christian

                                                            31 Cf. Lk. 16:14; 22:3, 53; Acts 5:1–6; 8:18–23; 13:6–10; 16:16–18. 

32 Cf. Acts 17:5–8; 18:12–13; 21:20–21, 27–28; 23:27–29; 24:1–9; 25:1–8, 18, 24–25; 26:2–3, 21–23. 

194  

Jewish opposition and calumny.33 In the social world of an extratextual Christian community like this, as in the intratextual relations between non-Christian Jews and Jewish Christians, many Jewish Christians unavoidably would have continued to associate with non-Christian Jews; and certainly, any Jewish Christians who participated in a communion that included uncircumcised Gentiles (even if they happened to be god-fearers) as uncontaminated equals would have rankled at least a few of their non-Christian Jewish neighbours. The means by which this egalitarian inclusion, moreover, had been achieved—the relaxation of Jewish laws concerning circumcision, diet, sabbath—would have made this and many other Christian communities vulnerable to the sorts of changes which in Luke’s narrative are repeatedly levelled against Paul.

Understandably, a Christian community of this composition and in this type of situation would have continued to use the apotropaic formulae and incantations on which they had relied before their conversion; like most Jews and pagans in antiquity, they would have attributed many of life’s misfortunes to demonic activity and therefore would have continued to live in fear of demonic attack.34 Furthermore, without instruction on this and related matters, they would not necessarily have known what they were to relinquish from their past and what they could continue to embrace. And to continue embracing as much culture from their Judaistic past as they could would have restrained the friction they were already experiencing with their own history and with their Jewish kinsmen. Accordingly, they held on to their Judaeo-Hellenistic formulae, which, in addition to providing a sense of security against the spirits of misfortune, would have given their users a much needed sense of integration with their history and neighbours.

As was implicit in the above discussion of co-text, however, the difficulties reflected in Acts 19:18–19 (i.e. concerning the books and practices) constitute only one piece of a larger circumstantial puzzle. Indeed, the single episode in which this note is found, when interpreted in the light of its co-text, strongly suggests that Luke’s misgivings about ‘magic’ were of no greater relevance to his situation than were his worries regarding the reputation of Paul. Since both concerns, moreover, were apparently related to the same set of extratextual dynamics, a key to the environmental forces that determined Luke’s whole narrative discourse might be discovered by asking how these two issues were interrelated.

Unfortunately, a comprehensive exploration into this question cannot be offered here. The following conjectures, therefore, are offered chiefly as a stimulus to further inquiry. Luke’s description of the unsuccessful exorcists as ‘itinerants’, like his emphasis on their Jewishness, is motivated by the need to make his opponents recognizable to his audience. His addressees having already been unsettled by the opponents’ reports about Paul, Luke responds by attacking the credibility of the opponents and by providing a flattering redescription of Paul. In his castigation of the opponents, Luke implicitly portrays them as dependent on magic (and thus, tacitly, on Satan), motivated by ‘silver’, and powerless to help a world congested with

                                                            33 Cf. J. Jervell, Luke and the People of God: A New Look at Luke–Acts (Minneapolis: Augsburg, 1972), 

pp. 146–47; and P.F. Esler, Community and Gospel in Luke–Acts: The Social and Political Motivations 

of Lucan Theology (SNTSMS, 57; Cambridge: Cambridge University Press, 1987), pp. 30–45. 

34 Cf. O. Böcher, Das Neue Testament und die dämonischen Mächte (SBS, 58; Stuttgart: Katholisches 

Bibelwerk, 1972), p. 13; E.R. Dodds, Pagan and Christian in an Age of Anxiety: Some Aspects of 

Religious Experience from Marcus Aurelius to Constantine (Cambridge: Cambridge University Press, 

1965), pp. 14–16, 37–38; and R. MacMullen, Paganism in the Roman Empire (New Haven: Yale 

University Press, 1981), pp. 50–51, 79–83. 

195  

demons. Indeed, their failure is, like the progress of Luke’s own ideology, a result of the might and will of the Lord (v. 20).

This whole programme of invective also contributes to Luke’s defence of Paul, whose powerful healings and close relationship with the Holy Spirit look all the more impressive vis-à-vis the unsuccessful magic of the itinerants. In a stroke of dramatic irony that borders on comic absurdity, Paul’s superiority to the ‘magicians’ is advertised even by one of the spiritual powers behind the ‘magicians’, that is, by an evil spirit. By means of the same demonic utterance, moreover, Luke reinforces an emphasis which the Jewish Christians in his audience especially needed to hear: Paul, as the macrostructural parallels between Luke’s heroes in Acts repeatedly imply, stood in perfect unity, harmony and continuity with Jesus, whom Luke has already depicted in his first volume as a Torah-observing, Scripture-quoting, prophecy-fulfilling Jew.

Finally, in a situation that required such vigorous defence of Paul, Luke’s repeated depiction of ‘both Jews and Greeks’ as responding to events in an ideal manner (i.e. they accept Paul’s preaching and magnify ‘the name of the Lord Jesus’) would have been entirely appropriate. For the non-Christian Jewish criticism of Paul, like the consequent Jewish-Christian reservations about him, would almost certainly have involved his interrelated teachings on the law and Jew-Gentile unity in Christ. Accordingly, to an audience strongly influenced by reputation, or by what is said and done by others, Luke’s insistence that large groups of Jews and Gentiles had already united in an enthusiastic acceptance of Paul’s Jesus would have functioned as a directive: what was already a social and historical fact, and what God himself had already approved in a variety of extraordinary ways, ought to be nurtured, not undermined.

Of course, Luke’s defence of Paul differs significantly from Paul’s own self-representation in the apostle’s undisputed letters, in ways that do not need to be repeated here. Nevertheless, contrary to a considerable and prestigious body of scholarly opinion, this difference by no means warrants the conclusion that Luke knew little about Paul and wrote Acts long after Paul died. Indeed, if the pragmatic maxim of relation (i.e. relevance) constrained Luke’s composition of Acts the way it appears to constrain almost all known forms of human communication, then the sorts of problems to which Luke–Acts was a response revolved around issues very similar to those that Paul himself had to address. In conclusion, unless we require Luke to have been truthful, and despite the notable tensions that exist between the Lukan Paul and the Paul of the letters, the situational correspondences between Luke–Acts and Paul’s own writings (especially the Corinthian correspondence) support the hypothesis that Luke was addressing a church that had been strongly influenced by Paul but had found its identity being undermined not long after the apostle died.

196  

Part III DISCOURSE ANALYSIS AND THE PAULINE CORPUS

THE DAMNED AND THE JUSTIFIED IN ROMANS 5:12–21: AN ANALYSIS OF SEMANTIC

STRUCTURE Richard J. Erickson

1. Introduction

Contemporary biblical scholarship devotes much of its energies to the analysis of texts in the light of ancient rhetorical science, modern sociological and literary theories, and special interest hermeneutics such as feminist and Marxist approaches. All of these approaches and others have important roles to play in the work of letting the ancient Hebrew and Christian Scriptures speak meaningfully to the world at the end of the twentieth century. a. Kaiser, Fee and Louw Parallel to these developments and at a more fundamental level in terms of the text itself, systematic methods for analyzing the syntactic and semantic structure of texts have been developed that offer advances on the traditional approaches to Greek or Hebrew grammar. Two decades ago, the science of ‘textlinguistics’ was sufficiently underway that an annotated bibliography devoted to it could be published.1 However, because progress in interdisciplinary applications necessarily lags behind the progress in the separate disciplines they seek to relate, use of textlinguistics—or discourse analysis, as it is sometimes2 called—has been slow to make itself widely felt in the field of biblical exegesis.3 I say ‘widely’ felt, because in fact Bible translators4 have been working in this area from the beginning, and the number of

                                                            1 W. Dressler and S.J. Schmidt, Textlinguistik: Kommentierte Bibliographie (Munich: Fink, 1973). 

2 ‘Sometimes’, in that recent discussion has applied the term discourse analysis more broadly than 

merely to textlinguistics; it is now used to speak of the larger context of discourse, linguistic and 

pragmatic, defining a communication event. For a convenient orientation to this ‘larger application’ 

of the term, see Joel B. Green, ‘Discourse Analysis and New Testament Interpretation’, in idem (ed.), 

Hearing the New Testament: Strategies for Interpretation (Grand Rapids: Eerdmans, 1995), pp. 175–

96. The present essay uses the term in its narrower sense. 

3 See the highly rhetorical essay by E. Güttgemanns, ‘Sprache des Glaubens‐Sprache der Menschen: 

Probleme einer theologischen Linguistik’, VF14 (1969), pp. 86–115, as well as J.F.A. Sawyer’s ‘Context 

of Situation and Sitz im Leben: Some Questions concerning Meaning in Classical Hebrew’, 

Proceedings of the University of Newcastle upon Tyne Philosophical Society 1 (1967), pp. 137–47, for 

early discussion of the communication lag between biblical studies and linguistics. 

4 Under the auspices of various Bible societies and of the Summer Institute of Linguistics, the 

research arm of Wycliffe Bible Translators. No attempt will be made here to provide a bibliography 

on discourse research, but see various references in the following notes. A recent article by Stanley E. 

Porter and Jeffrey T. Reed, though aimed primarily at making suggestions for a new BDF, provides a 

helpful summary of discourse analysis and its inroads into the study of biblical Greek (‘Greek 

Grammar since BDF: A Retrospective and Prospective Analysis’, FN 4.8 [1991], pp. 143–64, esp. 156–

197  

exegetes using these methods has begun to increase of late.5 But compared with the current attention given by scholars to the topics mentioned in the preceding paragraph, the systematic pursuits of syntactical and semantic analysis of texts are still rather ignored as scholarly subjects in New Testament studies. Journals such as Filología neotestamentaria, Linguistica biblica, and the Journal of Translation and Textlinguistics doubtless go a long way in correcting this deficiency.

Among biblical students, the better known of these newer approaches to textual analysis, at least in North America (and presumably South Africa), are perhaps those represented by Walter Kaiser, Gordon Fee and Johannes Louw (see the references in n. 4). Both Kaiser and Fee call their methods ‘syntactical’ or ‘structural’ analyses, ‘sentence-flows’, or the like. They focus on grammatical structures and interrelationships as indicators of theological content. Louw’s concern is with what he calls ‘colon’ analysis, taking his cue from the ancient grammatical unit of the colon rather than dealing with the later notion of a sentence. A sentence may contain one colon or several, whereas the reverse is not true. Louw’s point is that the semantic content of a text is best discerned on the basis of viewing the colon as the fundamental unit of analysis.

What makes these methods an advance on the traditional grammatical study of biblical texts is that they recognize the disadvantage of a word-bound, clause-bound or even sentence-bound approach, and attempt to remedy it. Traditional grammatical exegesis, productive though it is in its own way, tends to prevent an adequate appraisal of the effect of context on the form and meaning of a text. Not that biblical exegetes only recently began to take context into consideration! But until recently we lacked adequate tools and adequately modeled categories with which to study it systematically. b. The SIL SSA Series Yet there are two main disadvantages of Kaiser’s, Fee’s and Louw’s methods for laying out the syntactic or semantic relationships of a text. One is that they are too dependent on the text’s surface structures. Admittedly the surface structure is, in one sense, all we have to work with, and to pay close attention to it is not a vice but a virtue. But anyone who works with

                                                                                                                                                                                          63). And see now Stephen H. Levinsohn, Discourse Features in New Testament Greek (Dallas: Summer 

Institute of Linguistics, 1993). 

5 As a sampling, consider Robert E. Longacre, ‘The Discourse Structure of the Flood Narrative’, JAAR 

47.1 (1979), pp. 89–133; Walter C. Kaiser, Jr, Toward an Exegetical Theology: Biblical Exegesis for 

Preaching and Teaching (Grand Rapids: Baker Book House, 1981), esp. Chapters 5 and 8; Gordon D. 

Fee, New Testament Exegesis: A Handbook for Students and Pastors (Louisville, KY: 

Westminster/John Knox Press, 1993); Johannes P. Louw, Semantics of New Testament Greek 

(Philadelphia: Fortress Press; Chico, CA: Scholars Press, 1982); David A. Black, ‘Hebrews 1:1–4: A 

Study in Discourse Analysis’, WTJ 49 (1987), pp. 185–94, and ‘A Note on the Structure of Hebrews 12, 

1–2’, Bib 68 (1987), pp. 543–51; William L. Lane, Hebrews 1–8 (WBC, 47A; Dallas: Word Books, 1991), 

esp. pp. lxxv–xcviii; Peter Cotterell and Max Turner, Linguistics and Biblical Interpretation (Downers 

Grove: IVP, 1989); and the recently published collection of 14 essays by a variety of scholars, edited 

by David Black with Katharine Barnwell and Stephen Levinsohn, Linguistics and New Testament 

Interpretation: Essays on Discourse Analysis (Nashville: Broadman Press, 1992). It is interesting to 

note the appreciation of professional linguist Joseph Grimes for the pioneering ‘discourse’ approach 

of Daniel P. Fuller, The Inductive Method of Bible Study (mimeo; Pasadena: Fuller Theological 

Seminary, 1959); see Grimes, The Thread of Discourse (The Hague: Mouton, 1975), p. 7, passim. 

198  

grammar and semantics knows that there is not always a one-to-one correspondence between the two. Kaiser, Fee and Louw are, of course, all aware of this skewing factor, but their use of the surface text to display its semantic structure hampers them in allowing for possible discrepancies. The other disadvantage is that even when their displays are successful in presenting the semantics of a text, it is not easy from the display to see what the relationships among semantic units actually are. Both these problems are addressed in the method of semantic analysis developed by members of the Summer Institute of Linguistics (SIL).

The SIL method of semantic structure analysis for biblical texts is explained in detail in a handbook written by John Beekman, John Callow and Michael Kopesec.6 It is being applied to individual New Testament books in a series of publications known as SSAs (for ‘Semantic Structure Analysis’).7 Earlier volumes were edited for SIL by Kopesec; the current series editor is John Banker. The principal purpose of these materials is to aid field translators in the preparation of translations of the Bible into the less-known languages of the world. But their importance and the importance of the method itself in the work of New Testament exegesis is far-reaching. The method provides a supple and powerful tool for analyzing the semantic structure of a biblical text, an analysis which, of course, needs to form the foundation for any further exegetical and hermeneutical study.

While faithful to the surface text as the last court of appeal, the SIL approach seeks to identify the underlying semantic relationships which come to expression on the surface and to display them labeled in a clear and straightforward manner. By taking account of the features of marked and unmarked prominence in the surface text, the method identifies thematic material for the progressively higher level units of the discourse, from a concept all the way up to the level of an entire book, thereby extracting its semantic core. Thus, the entire literary context of a passage, from concept level to book level, can be displayed in such a way that its significance for the semantics of any portion of the discourse becomes obvious.

The remainder of this essay will be devoted to analyzing the semantic structure of Rom. 5:12–21, SIL style, and to drawing some conclusions about its meaning.

2. Semantic Structure Analysis of Romans 5:12–21

a. Method The method used here consists of several steps that lead to an exposition of the semantic relationships and central theme of a selected text. These steps may be described briefly as follows.

The first determines from various linguistic features what are the natural external boundaries of the text. These features include signals of transition in thought and logic (or scenes, for example, in narrative) with respect to the surrounding co-text, as well as marks of internal coherence. Once the external boundaries and internal coherence are discerned, the

                                                            6 John Beekman, John Callow and Michael Kopesec, The Semantic Structure of Written 

Communication (Dallas: SIL, 1981). A slightly more ‘user‐friendly’ version of this method is described 

by Cotterell and Turner in Linguistics and Biblical Interpretation, Chapters 6–8. 

7 For example, John Callow, A Semantic Structure Analysis of Colossians (Dallas: SIL, 1983). Several 

SSAs are currently available through the SIL bookstore: 7500 W. Camp Wisdom Road, Dallas, TX 

75236, USA. Especially relevant for this article is Ellis Deibler’s A Semantic Structure Analysis of 

Romans (forthcoming), which, at this writing, is being edited for publication; I wish to thank Mr 

Deibler for kindly making available to me in prepublication form his material on Rom. 5:12–21. 

199  

second step identifies in a similar way the text’s internal divisions and any marks of coherence that define subunits of the larger passage.

The reason for doing this is the conviction that coherent texts are built up, level by level, in a series of ‘head term-support term’ relationships in which discrete units either support or are supported by other units of the same level. These ‘head-support’ relationships form clusters which themselves become discrete units forming similar relationships on a higher level (sentences to paragraphs to sections, etc.).8 In order to arrive at a plausible analysis of these relationships, it is essential to identify the relevant units as accurately as possible. Hence, the importance of the first two steps.

The third step involves reducing the surface text (for this essay, of course, the Greek text of Rom. 5:12–21) to ‘nuclear sentences’. These are stripped-down versions of surface sentences or fragments of them, and are designed to render unambiguously the semantic relationships operating within them. From this data and with constant reference to the surface text, the analyst discerns, fourthly, the ‘communication relations’ (usually binary relations) existing between the discrete nuclear sentences, and between clusters of related nuclear sentences, for each of the subsections (paragraphs perhaps) of the text in question.

By a predetermined but flexible scheme of prominence within binary communication relations, the fifth step deduces the semantic theme of each unit. In a relation of ‘condition-CONSEQUENCE’, for example, the assumption is that usually the condition supports the consequence; in other words, the ‘head term’ and most prominent feature—the focus—is the consequence and not the condition (and in the SIL method the designator for the head term of a given relation is written in all capital letters). There may be instances where this ‘usual’ polarity is reversed, and such instances are linguistically ‘marked’ to indicate that. Detecting such marked reversals of prominence is often a matter of debate, however. All this leads, sixthly, to determining in an identical way the most prominent of the interrelated themes. This becomes the overarching semantic theme of the passage under study and can be used to relate the passage to its co-text in a progressive synthesis of the larger discourse, of the book of Romans for instance.

With this brief description of SIL’s general method of discourse analysis, we may proceed to considering Rom. 5:12–21. b. Boundaries and Coherence Since, on all levels of language, any unit of meaningful communication is indeed a unity, that is, a prominent central core with peripheral elements in support of it, a study of the semantic structure of a pericope must justify that pericope as a unit.

The unity of 5:12–21 as a section within the structure of the book of Romans is in fact widely recognized and needs little arguing. But the following features of the text support formal justification for drawing the boundaries at 5:12 and 6:1. Only the common transitional phrase διὰ τοῦτο marks the upper boundary of 5:12–21; but it is reinforced by the immediate onset of the ὡς/ὥσπερ plus οὕτως construction which permeates the entire passage and gives it some of its coherence and character. While there are numerous other comparisons made throughout the letter (employing ὡς, ὥσπερ, καθώς, οὕτως, etc.), the only other occurrences of this highly intentional and formal construction, outside of 5:12–21, are found at 6:4, 19 and 11:30–31. The last of these bears a thematic resemblance to those in 5:12–21 but is physically too far removed from them to have any direct bearing on the unity of 5:12–21; the two instances in ch. 6 are close enough to raise the question of whether or not 5:21 should constitute the lower boundary for the pericope, but their themes are quite different from those

                                                            8 Wilbur N. Pickering provides a useful discussion of this dynamic in A Framework for Discourse 

Analysis (Dallas: SIL; Arlington: University of Texas, 1980), pp. 14–15. 

200  

of ch. 5. In addition, 6:1 employs the rhetorical opener τί οὖν ἐροῦμεν, which is repeated at 6:15 and 7:7, suggesting the beginning of a new division. Besides the use of no fewer than six formal comparisons (counting the elliptical one at v. 16a), the internal coherence of 5:12–21 is also shown (1) by the focus on the theme of the two men Adam and Christ, marked as it is with the frequent repetition (12 times) of the term εἷς and the relationship of each of the two men to ‘the many’, ‘all humanity’, and (2) by a high concentration of a variety of terms for sin, judgment, grace and righteousness. c. Internal Boundaries and Coherence Within 5:12–21 three boundaries divide the section into four paragraphs. A major signal of internal division occurs at v. 15 (ἀλλά), where the subject matter shifts from the effect of Adam’s sin on the world to the contrast between the effect of that sin and the effect of the work of Christ. The inferential and transitional ἄρα οὖν at v. 18 marks a second major division, introducing the conclusion drawn from the foregoing argument, a conclusion demonstrating the similarity between Adam and Christ. The final division, at v. 20, is less obvious, being formally indicated only by δέ; but the recapitulation in vv. 20–21 of the argument in vv. 12–19, as well as the reintroduction of the law (last mentioned, explicitly, in v. 13, implicitly in v. 14) makes these final two verses a paragraph unto themselves. Thus, the internal paragraphing for 5:12–21 is as follows: vv. 12–14, 15–17, 18–19 and 20–21.9 d. Discerning Nuclear Sentences With the text’s boundaries defined, we can begin the analysis by classifying the elements of the surface structure (the Greek words and phrases) into the four major semantic categories of Things, Events, Abstracts and Relations. This enables us to formulate ‘nuclear sentences’ for the text, on the basis of which the semantic relationships within the text are determined. For instance, a ‘TEAR’ analysis of 5:12 might look like Because, for example, ἡ ἁμαρτία is understood to be a nominalization of the ‘event’ of sinning, it is analyzed as ‘E’ and thus becomes the core verb for a ‘nuclear sentence’ representing the semantic content of the opening phrase of v. 12: ‘One man sinned’. Theoretically, this sort of procedure could produce a bewildering array of very delicately analyzed nuclear sentences, but at some point the returns on the invested effort begin to diminish.10 Hence in the final display of nuclear sentences and their relationships to one another, the analysis is often not taken—and need not be taken—as far as it could be.

                                                            9 Deibler’s division is different (Semantic, pp. 115–17, 122). He sees the text as consisting of three 

claims (at v. 12—implied, v. 16 and v. 18) together with their respective justifications (vv. 13–15, v. 

17 and v. 19). Verses 20–21 forms a summary thesis statement. Thus his (sub) paragraphs are vv. 12–

15, 16–17, 18–19 and 20–21. Part of the rationale for this analysis is his assumption of an implied 

claim at v. 12, parallel to that in v. 18, based on the well‐known ‘incomplete comparison’ in v. 12. But 

see below, where this assumption is challenged. 

10 Where this point of diminishing returns is reached depends on the purpose for which the analysis is 

being prepared. The translation‐minded SSA books published by SIL, for example, will carry the 

analysis to finer distinctions than are necessary here. Compare the displays in this article with those 

for the same passage in Deibler. 

 

e. SemaParagraof 5:12–semanticonfigu

Somreasons concessCONSE

Capcase labpromine(12a–b)even thoconsideone-man

At a

οὕτως r

antic Relatioaph 1: verse–21) are diic relations

urations, andme of the l

of space: cs indicates aEQ refers to pitalized lababels indicaence and the) is labeled ough ordinared prominn’ theme wh

a few pointrather than t

ons among Nes 12–14. Tisplayed ons binding d so at last inlabels usedcompar refera concessiothe consequ

bels indicateate the supeme. In onein capital l

arily the conent here behich lies at

s the analysthe expecte

Nuclear SenThe nuclear n the right,

them ‘binnto a cohes

for indicars to a relat

on, which isuence of a (e the head opporting hae instance, tletters beca

omparison wecause of thits center.

sis is probleed οὕτως κα

Figure 1

ntences in 5sentences eoverleaf (F

narily’ withive unit.

ating semantionship of cs linked to (fulfilled) coor supportedalf. This dthe relation ause it is cowill be the she importan

Figure 2a

ematic. Theαί, as a resp

5:12–21 extracted forFigure 2a).h each oth

ntic relationcomparisona contraexpondition. d half of a

distinction iof 12a–b to

onsidered tosupportive r

nce for the e

e text at 12ponse to the

r vv. 12–14 To the lef

her into c

ns have bebetween on

pectation, ab

binary relatis importan

o 12c–d, theo be ‘markerather than entire perico

d reads, as ὥσπερ in 1

(the first paft are indic

clusters and

een abbrevine unit and

abbreviated

tionship, thnt for detee ‘comparised’ for promthe head teope of the

is well kno12a (cf. vv.

201 

aragraph cated the d larger

iated for another; CTRAX.

he lower-ermining on’ term minence, erm. It is ‘through

own, καὶ 18–19).

202  

Thus the relation of 12c–d to 12a–b is unclear. Is 12c–d a mere amplification of 12a–b, or should the phrase καὶ οὕτως be translated ‘indeed in this way’ (i.e. through one man), rather than ‘and thus’, and be understood to signal a complete comparison between 12a–b and the HEAD ‘term’ 12c–d? My suspicion is that too much is made of the unusual construction, that a genuine comparison is in fact intended, and that the supposedly reversed order of οὕτως and καί is merely a rhetorical device marking 12c as prominent and focusing the comparison on the role of the ‘one man’ by substituting οὕτως in the apodosis for the emphatic διʼ ἑνὸς ἀνθρώπου in the protasis (as J.T. Kirby argues11).

A second problem, also in v. 12, is likewise well known: What is the sense of ἐφʼ ᾧ? If it is taken as an indicator of cause, as is usual in translations, then the clause it introduces seems to contradict the entire argument that Paul is building in this passage. Various solutions have been offered to this difficulty,12 but its ordinary use may not be quite as contradictory here as it first appears. For the moment, although it is tagged with a question mark in the display in Figure 2a, I will propose 12d as the ‘reason’ for 12c, and argue its appropriateness below.

Three concession-CONTRAEXPECTATION relations (‘although X, nevertheless or surprisingly Y’) appear in the display for vv. 12–14. Two of them are signaled by adversatives, δέ (13a–b/13c–d) and ἀλλά (13a–d/14a–d), which one would expect. The other, however, is marked by καί (14a/14b–d), obviously in the sense of ‘even’, suggesting surprise, that is, contraexpectation.

                                                            11 Cf. J.T. Kirby, ‘The Syntax of Romans 5.12: A Rhetorical Approach’, NTS 33 (1987), pp. 283–86. 

James D.G. Dunn (Romans 1–8 [WBC, 38A; Dallas: Word Books, 1988], p. 273) rejects Kirby’s 

argument because it does not explain the sequence of thought from v. 12 to v. 13. On the contrary, 

however, the sequence seems quite clear on Kirby’s reading: after stressing the role of the one man 

Adam in the coming of sin and, especially, death upon all humanity, Paul simply goes on to explain 

his rationale in vv. 13–14. ‘How else’, he in effect argues, ‘can we account for the deaths of people 

who had no law to sin against?’ Cf. Stanley E. Porter, who finds the view that the construction is a 

completed comparison ‘plausible conceptually but probably not grammatically’ (‘The Pauline 

Concept of Original Sin, in Light of Rabbinic Background’, TynBul 41.1. [1990], pp. 3–30, quotation p. 

21). 

12 See the summary by Dunn, Romans 1–8, p. 273, who concludes that ‘The classical debate on the 

meaning of ἐφʼ ᾧ has more or less been settled in favor of the meaning “for this reason that, 

because” … “in view of the fact that” ’. A recent contribution to the debate is Joseph A. Fitzmyer’s 

1992 SNTS Presidential Address, ‘The Consecutive Meaning of ἐφʼ ᾧ in Romans 5:12’ (NTS 39 [1993], 

pp. 321–39). Fitzmyer analyzes the occurrences of the phrase generated by a search of the Thesaurus 

Linguae Graecae and concludes that a majority of these, especially when connected with an aorist 

verb, carry a consecutive sense; thus the phrase in Rom. 5:12 implies that the sin of all humanity is 

the result of the mortal condition brought about through Adam’s transgression. However, majority 

does not necessarily rule in linguistic usage, and as will be seen presently, there is contextual, or 

discourse, rationale for preferring the more traditional, causal interpretation here. The difficulty 

Fitzmyer (p. 327) cites against the causal sense, namely that it contradicts the rest of v. 12 by 

ascribing death also to human acts, is resolved, it seems to me, when v. 12d is taken into 

consideration with v. 17 and the entire discourse. See below. 

203  

A ‘reading’ (necessarily clumsy) of the analysis as displayed in Figure 2a would sound perhaps like this: ‘On the following grounds,13 I, Paul, conclude that just as sin came into the world through one man and death came into the world through that one man’s sin, in the same way, all people died through that one man, because [or and moreover?] all people sinned. The grounds are these: God does not reckon sin where there is no law, even though people sinned in the world before the law came; yet in spite of that, everyone died between Adam and Moses, even those who did not sin the way Adam did [meaning, presumably, “against a known law”].14 And Adam is a type of the Coming One.’ Paragraph 2: verses 15–17. The next three paragraphs can be set out in similar displays. Figure 2a, opposite, shows vv. 15–17. Abbreviations in this display are clear. But attention may be drawn to the ‘identification-CONCEPT’ relation in v. 15 and the ‘description-CONCEPT’ relation in v. 17. In both these relations the connection is not between two full units, two full nuclear sentences, but between a nuclear sentence or cluster of nuclear sentences and a single element within another nuclear sentence. In v. 15, for example, 15d–f constitutes the identification only for the ‘concept’ grace in 15g, not for 15g in its entirety. Likewise, 17c and 17d form the descriptions for the ‘concepts’ grace and righteousness in 17e.

Some further comments are in order. There has been some discussion on whether Rom. 5:15–17 compares or contrasts Adam and Christ.15 Naturally, they could neither be contrasted if they were not in some way comparable, nor compared if they did not contrast with each other in some respect. But the construction εἰ + πολλῷ μᾶλλον most naturally implies a contrast, an intent to indicate a difference. The nature of a contrast, as described by Beekman, Callow and Kopesec,16 confirms this view of vv. 15–17: a contrast between two items requires two points of dissimilarity and one of similarity. In v. 15, for example, the point of similarity is ‘the many’; the two points of dissimilarity are (a) the transgression of Adam vs.

                                                            13 This phrase is not a rendering of the διὰ τοῦτο at the beginning of v. 12, which has anaphoric 

reference, not cataphoric as would be suggested here. The phrase ‘on the following grounds’ 

interprets the internal structure of vv. 12–14; διὰ τοῦτο, not represented here, connects the entire 

passage 5:12–21 to what precedes it. 

14 Porter (‘Pauline Concept of Original Sin’, p. 26) puzzles over the clash between Rom. 5:14ff. and 

Paul’s statement in v. 13 that ‘until the law sin was in the world, but sin is not counted where there is 

no law’; he then makes the puzzling statement ‘The fact that sin was in the world is attested by the 

fact that humans died before the giving of the law …’ I do not think that is Paul’s point at all; as I read 

the passage, Paul is arguing not the fact that there was sin before the giving of the law or that 

humans died then—which are truisms; rather his point is that death before the coming of the law 

and the consequent ‘reckoning’ of sin testifies to the fact that death is a result of Adam’s sin, and not 

of the sins of those who died between Adam and Moses. 

15 Cf. Chrys C. Caragounis, ‘Romans 5.15–16 in the Context of 5.12–21: Contrast or Comparison?’, NTS 

31 (1985), pp. 142–48, who suggests that the οὐχ in each of the first lines of vv. 15 and 16 actually 

introduces a question that expects an affirmative answer, thus in each case setting up a comparison 

rather than a contrast. The proposal Caragounis makes is ingenious and, in spite of Dunn’s dismissal 

(Romans 1–8, p. 271), does make better sense of the entire passage on the assumption that the 

comparison in v. 12 is incomplete. But as long as v. 12 is understood as a complete comparison, vv. 

15–16 make good sense as contrasts in the context of the section. 

16 Semantic Structure, pp. 97, 100. 

 

the gracin grace

The

nuclear

ce and free e.

addition osentence 1

gift of Chri

of the word5a is based

ist, and (b)

ds ‘result od on the fac

the fact tha

Figure 2b

f Christ’s’ ct that the f

at one cause

and ‘that [feature in te

es death and

[sc. result] erms of whi

d the other

of Adam’sich the παρ

204 

abounds

s’ in the ράπτωμα

205  

and the χάρισμα are dissimilar is indeed their end results: death and abundance of grace. The contrast in v. 16, however, is between the fact that the judgment and condemnation came to all humanity through just one man’s sin, and the fact that the free gift leading to justification for all humanity came in response to many transgressions, the point presumably being that grace is that much more powerfully effective than condemnation. In this way v. 16 amplifies the contrast, first set up in v. 15, between what Adam did and what Christ did.17

Ellis Deibler understands the second halves of vv. 15 and 16 to provide further explanation of what Paul means in their respective first halves; the relation, indicated by γάρ, is one of specification rather than of grounds.18 This analysis is, of course, possible; but since the content of v. 17 (which Deibler himself takes as the justification for the claim made in v. 16) and the content of both 15b–g and 16b–f are so similar, it seems better to take them all as providing the grounds on which the conclusions in vv. 15 and 16 are drawn.

We may now ‘read’ the display of this second paragraph:

I, Paul, conclude that the result of Christ’s gift is unlike the result of Adam’s transgression. I conclude this on the grounds that whereas the many died because the one man sinned, far more did God’s grace and the gracious gift of Jesus Christ abound to the many. Moreover, the effectiveness of Christ’s gift is not like (in fact it exceeds) that of the one man’s sin; I conclude this on the grounds that whereas God condemned all people by judging just the one man’s sin, nevertheless, much more, even though people sinned many times, God freely makes all people right with himself. That is to say, I conclude all this because, whereas all people died because of the one man’s sin, far more will all people live through the one man Jesus Christ, provided they receive this grace which abounds toward them and this righteousness God so freely gives them.

Paragraph 3: verses 18–19. Figure 2c displays the third paragraph:

                                                            17 Deibler (Semantic, p. 120) sees the contrast in v. 16 as essentially the same as that in v. 15, 

between the results of what Adam and Christ have done. But the numerous additions that he feels 

are necessary to make sense of the highly elliptical phraseology seem to me to argue that he is 

mistaken. Caragounis (‘Romans 5:15–16’, p. 145), approaching the two verses as comparisons, 

believes that v. 16 focuses on the agents responsible for the effects that were compared in v. 15. But 

while the initial proposition of v. 16 does indeed read more easily as a question framing a 

comparison, the explanatory material in the second portion of the verse does not properly answer 

such a question of agent. The contrast in v. 16 seems far more easily understood as a contrast 

between the power or degree of effectiveness: even many transgressions could not nullify the effect 

of the free gift. This view fits well with the εἰ + πολλῷ μᾶλλον constructions in vv. 15 and 17. Dunn’s 

setting up the contrast between the free gift and the one man who sinned (Romans 1–8, p. 280) 

certainly takes 16a as it appears in surface form, but again the rest of the verse does not seem to 

support it, and ultimately in fact he sees the contrast in v. 16 as between the two epochs in terms of 

their beginnings and ends (p. 295). 

18 Semantic, pp. 119, 120; cf. pp. 115–16. 

 

Renderi19d as causalit

The a synonof HEArestatem17 reinf

Thu Just justimighmanthe o

Paragra

ing the term‘counted as

ty within thi relationshi

nymous restAD-contractiment of v. 1forces the ‘gus the displa

as God conified all peopht live. To pu

n Adam disobone man Chr

aph 4: verse

ms κατεστάs’ and ‘willis verb. Thep between atement of ion, and the8 in v. 19 hgrounds’ poay may be re

ndemned all ple because tut it another beyed him, inrist obeyed h

es 20–21. F

θησαν and l count as’, e result is an18a–e and 118a–d. Onle γάρ must

has the effecortions of vvead:

people becthe one manway, just as n the same wim.

igure 2d dis

Figure 2c

κατασταθήrespectivel

n act of God19a–d is sigly 18e is lefbe taken as

ct of reinforv. 15 and 16

ause the onn Christ acte God counte

way God wil

splays the fi

Figure2d

ήσονται in tly, represend. gnalled by γft unrestateds continuatircing its poin6.

e man Adamd righteously

ed the many all count the m

final paragra

the nuclear nts the sema

γάρ, but the d in v. 19. Hional. The (nt, much in

m sinned, iny. God did tas sinners simmany as righ

aph:

r sentences antic comp

content of Hence a rela(slightly conn the same w

n the same wthis so that amply becaus

hteous simply

206 

19b and onent of

19a–d is ationship ntracted) way as v.

way God all people se the one y because

 

The one‘reason-20c–d inrepetitiolife, I crelation

The Alththeypeopdiedmigh

f. PromAs menrelationrelationare) appnature owhen cbecome

It ismakes uthereforway thamaterialin quest

e thing to no-RESULT’ ren this paragon of the veconclude thnship in 21a– display of

hough God sy transgress, ple sinned abd, so likewiseht live foreve

inence and ntioned abovnal display nship is conspended as of languageonsidered o

e part of the s essential up the core re, consolidat those conl is indentedtion (here at

otice here iselationship graph). Howerb βασιλεύat the usua–b and 21c–the final pa

ent the law still, surprisibundantly. He, he might ber through C

Theme Dispve, the use is intendedsidered in fonon-promin

e structure, on another non-prominto analyze of a unit. B

date the nucnsidered prod to the right the level o

s that accordis on the r

wever, becaύειν and theal ‘prominen–e. ragraph ma

into the woringly, God b

His purpose be gracious thrist.

play of capitaliz

d to indicatocus, prominent, suppowhat is healevel. For

nent ‘purpoprominenc

But it is notclear sentenominent areht, with respof paragraph

ding to SIL result, the rause of the e subordinance polarity

ay be read as

rld in order became supein this was to all people

zed and lowte which ofinent, the ‘hortive matead material

example, tose’ for v. 20ce in order t easy to ‘s

nces into clue brought oupect to the th) is extracte

theory the reason beingway in whi

ated positiony’ is revers

s follows:

that people rabundantly that just as

e by justifyin

wer-case labf two (or mhead’ term, erial. Howe

on one levthe promin0. to determi

ee’ it in a rusters whicut to a centtext it suppoed from its

usual focusg in the supich v. 21 is ns of the nosed for the

might knowgracious preall people sing them, wit

bels on the more) membto which thver, becaus

vel may be nent ‘reason

ine the themrelational dich are then ter line andorts. Then thprominent p

s or prominepportive rostructured,

otions of deoccurrence

w more and mecisely becauinned and asth the result

left-hand smbers of a she other is (ose of the rsupportive

ns’ in 21a

matic mateisplay. Figuarranged in

d the non-prhe theme ofproposition

207 

ence in a le (as in with its eath and es of the

more that use/when s a result that they

side of a semantic or others recursive material and 21c

erial that ures 3–4, n such a rominent f the unit

ns.

 

THEME

THEMEeffectiveother re

E: All huma

E: The resuleness of Adaesult in life f

anity has bee

lt and effectam’s sin: wfor the many

en subjected

Figure

tiveness of Cwhile the one

y.

d to death b

e 3a: Parag

Christ’s freee resulted in

because of t

graph 1

e gift is not n death for t

he sin of the

like the resthe many, fa

e one man A

sult and far more wil

208 

Adam.

ll the

 

THEMEeffectiveother re

THEMEChrist’s

E: The resuleness of Adaesult in life f

E: Just as ths righteous a

lt and effectam’s sin: wfor the many

he one man act resulted

Figure

tiveness of Cwhile the one

y.

Figure

Adam’s sind in universa

e 3b: Parag

Christ’s freee resulted in

e 3b: Parag

n resulted inal justificat

graph 2

e gift is notn death for t

graph 2

n universal cion.

like the resthe many, fa

condemnati

sult and far more wil

ion, so the o

209 

ll the

one man

 

THEMEmight li

g. RelatThe fininterrelacase thesame warranged

This dis

Therone diffe

E: God has ive.

tions amongal step in aationships ae four parag

way as wered according

splay for the

refore (διὰ τοman Adam,

erent from th

become sup

g Theme Staanalyzing thamong the tgraphs of Roe the units og to promine

e entire sect

οῦτο), althou, surprisingl

he result of A

Figure

perabundan

Figure

atements he internal hemes of thom. 5:12–2on the levelence.

Figure 4

tion, then, m

ugh all humay (ἀλλά) Ch

Adam’s sin: f

e 3c: Parag

ntly graciou

e 3d: Parag

semantic she compone21. Their rell of nuclear

4: Romans

may be read

anity has beehrist has frefor while the

graph 3

s, in order t

graph 4

tructure of ents makinglations to onr sentences.

5:12–21

d as follows:

en subjected eely given se one resulte

that through

a passage g up the pasne another a. Figure 4 p

:

to death becomething thd in death fo

h Christ all

is to determssage, in theare describepresents the

cause of the she result of or the many,

210 

l sinners

mine the e present ed in the e results,

sin of the which is far more

211  

will the other result in life for the many. So then on these grounds (ἄρα οὖν), I, Paul, conclude that just as the one man Adam’s sin resulted in universal condemnation, so likewise the one man Christ’s righteous act resulted in universal justification. In summary then (δέ), God has become superabundantly gracious, in order that through Christ all sinners might live. The theme for the section, based on prominence, can be stated, Just as the one man

Adam’s sin resulted in universal condemnation, so the one man Christ’s righteous act resulted in universal justification.

3. Interpretation

a. Relation to External Structure This theme, then, becomes the proposition to which the surrounding context of 5:12–21 is related, according to the same principles operating in the analysis of the section’s internal structure. And while it is outside the scope of this essay to analyze, or even to discuss, the structure of the entire epistle, a few comments may be made here to demonstrate how the ‘thematized’ version of 5:12–21 fits into the whole.

Romans may be said to have as its theme proposition the following statement, which can also be thought of as Paul’s gospel: ‘In Christ, God has received and united all the peoples of the world, Jew and Gentile alike’. Section 5:12–21 falls in the first of the epistle’s three major parts (1:16–8:39), the theme of which can be stated as ‘The Gentile, like the Jew, is made part of God’s people by faith’. The other two major parts of Romans, 9:1–11:36 and 12:1–15:13, concern themselves, respectively, with the fact that the Jews remain God’s people despite the addition of the Gentiles, and that this receiving and uniting of all God’s people in Christ has implications for Christian behavior.

The first major part (1:16–8:39) consists of two subparts (1:16–5:21, 6:1–8:30) and a triumphant conclusion (8:31–39). In the first subpart of Part 1, Paul argues that righteousness comes by faith in Christ for both Jew and Greek; he eliminates every escape from the wrath of God but one: God’s mercy. The second subpart takes up the obvious question of how this doctrine of righteousness by faith and mercy squares with the clear necessity to be obedient to God’s law. Paul argues that this grace is not an excuse to sin (6:1–7:6), nor is the law itself sinful because it ‘increases’ sin (7:7–25), but indeed, that those who are in Christ, filled with the Spirit, are anything but lawless in their behavior (8:1–30).

In Part 1, Subpart 1, Paul argues first of all that under the law, all people are condemned; there is no escape (1:16–3:20). In the second half of Subpart 1 (3:21–5:21), he demonstrates that under grace all people are made righteous. He argues first that righteousness comes by faith apart from the law, as even the law itself testifies (3:21–4:22). In fact, it comes by Jesus Christ (4:23–5:21); through him we have peace and reconciliation with God, even as God’s enemies (4:23–5:11), and though him and him alone—through one single individual—the entire human race has been justified. It is this last point that Paul explains and demonstrates in 5:12–21. b. Corporate Solidarity and Personal Responsibility If this analysis of the semantic core of Rom. 5:12–21 and its relation to the overall thrust of the book is sound, then we can conclude that for Paul’s presentation of the way God deals with humanity it is essential that the entire matter come down to this: that all people collectively are represented before him by one person, whether for condemnation or for justification. If it were otherwise, then each individual must stand on his or her own merit, or lack of merit, before God; the system of righteousness by the law would prevail after all. It could then be argued that some persons are more deserving than others, and the reunification of humanity—a prime concern in Romans—would thus be undermined. But where all people

212  

are subsumed under the corporate headship of one solitary person, there are no grounds on which to plead individual cases. The one man Adam, in his disobedience, has effectively and finally represented the race he fathered. It is in this sense that he is a ‘type’ of Christ19 the Coming One (14d): both Adam and Christ, individually, effectively represent the entire human race.20 That is, if a person who has been condemned in Adam, along with everyone else, is to be justified before God, it must also be in solidarity with everyone else, with the mass of humanity, under the effective and final corporate headship of Christ. That is, it must be done solely on the terms that God himself, in sovereign mercy, dictates.

Paul applies the same principle of corporate solidarity to death and resurrection in 1 Cor. 15:20–22: just as Adam represents all humanity in the matter of death, so Christ represents them all in the matter of resurrection to new life. And again, a little further along in the same chapter (vv. 45–50), he applies the principle to the subject of the kind of bodies he believes we humans have and shall have, whether of perishable dust or of imperishable heavenly substance. We have the former because we belong to Adam’s race and shall have the latter because we bear the image of Christ.

This notion of solidarity or corporate headship is, of course, not original with Paul; it is well known from the Old Testament (as, e.g., in the incident with Achan, Joshua 7) and other Jewish literature. Eduard Lohse sees it as common among apocalypticists, whose musings trace the evil in the world back to Adam (4 Ezra 4:30, Syr. Bar. 23.4).21 Similarly, 4 Ezra 3.21–22, 26; and 7:118: ‘O Adam! What have you done? Your sin was not your fall alone; it was ours also, the fall of all your descendants.’ Nor, on the other hand, is the notion of solidarity in the matter of belonging to God’s people original with Paul. It is doubtless at the root of the Pharisees’ confidence in their descent from Abraham, which prompted a sarcastic remark from John the Baptist (Lk. 3:7–8). Paul himself subscribes to this Pharisaic ideal of descent from Abraham, except that, somewhat like the Baptist, he redefines descent in terms of the imitation of Abraham’s faith (Rom. 4:16–17).

In other words, Paul’s theme of universal condemnation in Adam and universal justification in Christ, if offensive to his Jewish (and perhaps Gentile) readers, was offensive not in terms of its operating principle of solidarity, which apparently would have made perfect sense to them, but rather in terms of its application to all human beings. For many modern readers in the West, it is perhaps the other way around. The application to all humanity, Gentiles as well as Jews, is perfectly acceptable. The offense comes more in the issue of solidarity and representation by corporate headship; a Western reader might be tempted to object that his or her fate ought not to be determined on the basis of some other person’s behavior.

Almost as if he were anticipating those later Western objections, Paul has buried in the text of 5:12–21 two provisos that raise the question of personal responsibility. I say ‘buried’ because these two items belong to non-prominent, non-thematic material. The puzzling ἐφʼ ᾧ πάντες ἥμαρτον in v. 12, taken in the causal sense usual in translations, supplies the first of these. Stanley Porter neatly summarizes three possible interpretations that can be placed on the ‘causal’ sense of v. 12d: the Pelagian (all have sinned following Adam’s example), the realist (all sinned when Adam sinned), and the federalist (all are constituted sinners because

                                                            19 Not a type of humanity generally, or of Moses, or of ‘man under the law’, etc. Cf. Dunn, Romans 1–

8, p. 277. That Christ is meant as ‘the Coming One’ is thoroughly appropriate to this discourse. 

20 Nothing is implied here regarding doctrines of the origin of the human soul, viz. creationism, 

infusionism or traducianism. 

21 Eduard Lohse, The New Testament Environment (Nashville: Abingdon Press, 1976), pp. 56–57. 

213  

Adam legally represents them). Paul, says Porter, is not specific on the issue at this stage and in fact breaks off his discussion in order to clarify several points in vv. 15–18.22 It seems to me, however, that while Paul eventually explains himself in essentially ‘federalist’ terms, none of the three causal possibilities mentioned by Porter applies to v. 12d. Rather, the function of ἐφʼ ᾧ πάντες ἥμαρτον within the discourse is to signal to the readers—in response to their potential objections to Adam’s corporately representing them—that they have indeed ratified the arrangement by the very fact that they themselves have sinned. On this reading, all humanity somehow suffers death both because of Adam’s transgression and because they, too, have actually sinned.23 Lohse finds a similar notion represented in apocalyptic thought: ‘If Adam first sinned and brought premature death upon all, still everyone of those who descended from him has incurred future torment upon himself, (Syr. Bar. 54.15).24

On the other hand, according to a second proviso, actually ‘reigning in life’ on the basis of the universal justification in Christ is an experience apparently enjoyed only by those who receive the abundant grace and the gift of righteousness (v. 17). One might expect that Paul would have parallelled the second half of v. 17 with the first half, so that the entire verse would read: ‘If, because of one man’s trespass, death reigned through that one man, much more, through the abundance of grace and the free gift of righteousness, life shall reign through the one man Jesus Christ’. However, the restructuring of the second half of the verse, so as to introduce the notion of a human response, parallels instead the human responsibility introduced into the matter of condemnation in v. 12.

                                                            22 Porter, ‘Pauline Concept of Original Sin’, pp. 24–26. 

23 Deibler (Semantic, p. 117) gives the usual sense of ‘because’ to ἐφʼ ᾧ but rejects the application of the clause to ‘actual sins by everyone’. He believes that because ἥμαρτον is aorist it must refer to a 

‘one time action’, and therefore would make no sense as a reference to actual sins, since Paul’s point 

in that case would have been the habitual nature of the sins. Deibler therefore understands the 

aorist to refer to a ‘once‐for‐all’ action in the past, namely Adam’s; it is ‘as though’ all people sinned 

when Adam sinned (i.e. Deibler takes a ‘realist’ view of the matter). But the traditional interpretation 

of the aorist aspect as indicating a one‐time or once‐for‐all ‘kind of’ action has long been known to 

be false. As early as 1900, Ernest De Witt Burton was saying so (Syntax of the Moods and Tenses in 

New Testament Greek [repr.; Grand Rapids: Kregel, 1976], p. 16), and even the venerable grammar 

by J. Gresham Machen (New Testament Greek for Beginners [New York: Macmillan, 1923]) is 

consistently careful not to say any more than that the aorist in the various moods ‘refers to action 

without saying anything about its continuance or repetition’ (§283), or that the aorist refers to the 

action ‘in no … special way’ (§299). The aorist may indeed be used to refer to a single action, but that 

is not its only function and certainly not its nature as an aspect. Rather, as the indefinite, 

noncommital aspect, it can be applied much more widely than traditional New Testament Greek 

grammar has led us to believe. See the lively discussions by Frank Stagg, ‘The Abused Aorist’, JBL 91 

(1972), pp. 222–31, and D.A. Carson, Exegetical Fallacies (Grand Rapids: Baker Book House, 1984), 

pp. 69–75. Porter and Reed, ‘Greek Grammar since BDF’, contains a helpful section on recent 

developments in studies of aspect (pp. 149–56). The point here is that ἐφʼ ᾧ πάντες ἥμαρτον can 

legitimately be taken to refer to the actual sins of all human beings, and in this text, should be. 

24 Lohse, Environment, p. 57. 

214  

c. The Damned and the Justified Who then, according to Paul, are the damned, the condemned? And who are the justified, that is, those who have been, or will be, made alive, as 1 Cor. 15:22 puts it? Quite simply, the damned constitute all humanity, every last human being; and the justified likewise constitute all humanity, every last human being. The damned are justified and the justified are damned. No one is preferred; no one is excluded. The pronouncement is made in both directions without the consent of any single person; it is done strictly in accordance with the sovereign will of God.25

The two provisos do not in any way alter this arrangement. In Paul’s judgment, God has determined to view the entire race of humanity—whether en masse or one by one—through the two Corporate Heads, the two all-encompassing representatives of Human Solidarity: Adam and Christ. The provisos enable Paul to portray the situation also from the perspective of humanity. To anyone who, as a sovereign individual, objects to God’s cavalierly casting his or her lot with ‘the many’ under that sinner Adam, Paul (or presumably—in Paul’s view—God himself!) can reply, ‘Well, then, what about you? Have you sinned?’ Yes, indeed! All have sinned. And in so doing, all sinners have individually and collectively ratified God’s decision to make Adam their representative before him. Whether considered on their own merits, or on Adam’s, all human beings stand condemned before God. All human beings, even the justified.

The second proviso, likewise, does not change the fact that God has included all humanity under Christ’s justifying representation. But like the first, it also views the matter from ‘below’, from the human vantage point. The implication of v. 17 is that while no one is excluded from coverage under God’s provision in Christ, just as no one is excluded from sharing Adam’s condemnation, nevertheless, only those who ‘receive’ the gift of grace that God provides in Christ will enjoy the benefits of it. That is, although Paul is teaching universal justification, corresponding to universal condemnation, he is not teaching universalism. Grace is provided abundantly for all people; some will respond to it and others will not.

This idea can be schematized (as in Figure 5) in the form of two partially overlapping right triangles whose bases coincide but whose ninety degree angles are on opposite ends of the base. The apex of one triangle represents Adam, that of the other stands for Christ. The base represents the human race, and individual points along it represent individual human beings. Each point along the base aims an arrow back toward the apex representing Adam, indicating that all people ratify with their own sins Adam’s corporate headship over them (v. 12d). Only some of these points, however, also send an arrow toward the apex representing

                                                            25 Deibler (Semantic, pp. 116, 121) softens the bald statement of universal justification in v. 18 by 

phrasing the relevant nuclear sentences as follows: ‘the fact that one man (Jesus) acted righteously 

(when he died) resulted in all people’s (or, all who believe) being (able to be) acquitted and live 

eternally’. He supplies the words all who believe and able to be ‘in order to keep Paul’s exposition of 

his main theme in 3:22 consistent and free from self‐contradiction’. But in doing so he saps the 

driving power from the passage, the very power that enables Paul to say what he says in 3:22, and 

indeed to preach the gospel at all. For if the damned are not acquitted before they have anything to 

offer, then there must be something to offer in exchange for acquittal, even if it is ‘faith’ or 

‘acceptance’ of acquittal. On the present interpretation, however, faith and acceptance enable 

people to lay hold of and enjoy what God has already fully provided without consulting them; faith 

and acceptance are not the medium of exchange that finally convinces God to acquit an individual. 

There is in fact no contradiction between 3:22 and the non‐nuanced interpretation of 5:18. 

 

Christ, headship

Understthe pivostructurone manact resufor arguthat undIn this cPaul adidentificthe Jewtrue, theone acchis peopclear in by theirare indedamned

Not determi5:14–15all; therthemsel

indicating p over them

tood in this ot point. Thre analysis, n Adam’s sulted in univuing both thder grace, alcomparison

ddresses in cation in 6:1

wish nation wen what advording to wple must bechs. 12–15

r own choiceed justifiedd is to be on

surprisinglnative not o

5: ‘For the lrefore all halves but for

that only sm (v. 17e).

way, Rom.e theme stais in fact th

sin resulted versal justifhat under thll people, Jen, the pith o

the three 1–4 proceedwithin the pvantage is th

which God vehave towa

5. For, to pue, being me

d, they are snce again ouly, the notioonly for Palove of Chrave died. Ahim who fo

some ‘rece

4

. 5:12–21 isatement for he bedrock

d in universafication. Lohe law all pews and Genof the letterchapters th

ds from it. Iplan of Godhere in bein

views humaards others, ut it bluntly,eanwhile fulo only beca

utside of graon of Adam

aul’s gospelrist controls

And he died or their sake

eive’ the gr

Figure 5

4. Conclusio

s fundamenthe section,on which thal condemnocked withinpeople, Genntiles alike,r, are the grhat follow It is, as welld (chs. 9–1ng an Israelanity, then it

both insid, it means thlly justified

ause they arace. m and Chril but for hiss us, becaus

d for all, thae died and w

race he pro

on

tal for Paul, deduced thhe entire st

nation, so thn this form

ntiles and Je, are made mrounds for 5:12–21; t

l, the reason1); for if thlite? And fint has tremene and outsihat if the da

d in Christ. Are also fully

st as corpos mission ase we are coat those whowas raised.’

ovides them

’s argumenhrough the pructure is fohe one manal comparis

ews alike, amembers ofthe fears othe argumen he must whis interpretnally, if thindous impliide the churamned go toAnd converdamned; to

rate heads s well. As honvinced tho live migh

m by his c

nt in the epiprocess of s

founded: Jusn Christ’s rson are the are condemnf the people

of libertinisment from b

work out the tation of hiis plan is inications for

urch, as Pauo hell, they rsely, if the o be other th

for all humhe puts it in

hat one has ht live no lo

215 

corporate

stle; it is semantic st as the

righteous grounds ned, and

e of God. m which aptismal place of

s plan is deed the the way

ul makes go there justified

han fully

manity is n 2 Cor. died for

onger for

216  

A DISCOURSE READING OF EPHESIANS 1:3–14 Johannes P. Louw

The title of this essay seems to be essentially redundant since it tends to emphasize the obvious. Should one not read any passage with due attention to the discourse? Indeed, yet the general practice of traditional Bible translation, with its strict focus on words and phrases to ensure a ‘close rendering’ that will produce a ‘careful and thorough reproduction of the original text’, often fails to do just that. The crux of the matter lies in the surprising fact that reading a text involves various levels of meaning: the meaning of a sentence is not merely the sum total of the meanings of the words constituting the sentence, and likewise, the meaning of a paragraph is more than what is contributed by the sentences comprising the paragraph. In addition, the structure of a discourse contributes considerably to what is linked or contrasted, what is foregrounded or backgrounded, to what is fact or speech act, what is informative or emotive, what is history or parable, etc. Even more surprising is to recognize that a text may allow for more than one reading. This is quite the case with fine literature, especially poetry.

Reading is always a complex process. Therefore, discourse analysis is not an end in itself; it does not provide a final reading void of any alternatives or even subjective notions. But, on the other hand, one should not overstate the issue since that would make all reading futile. A discourse reading is rather an account of how a text as a whole is, or can be, read. It is a valuable tool enabling a reader to comprehend longer stretches of communication, especially long complex sentences or paragraphs.

The Greek text of Eph. 1:3–14 can be considered a single sentence as printed in most editions of the Greek New Testament. Recent Nestle-Aland editions print the section as four sentences: 3–6, 7–10, 11–12, 13–14 taking the phrase ἐν ᾧ in vv. 7, 11 and 13 as introducing a break. Bible translations regularly divide the long Greek sentence of Eph. 1:3–14 into a number of sentences and even introduce paragraph breaks. It is almost inevitable that any reading of this passage should cluster the contents in some way since the style of the passage is tense; it is language stretched to its limit.

The Jerusalem Bible (JB) maintains the passage as one paragraph, but divides it into six sentences: (1) v. 3; (2) vv. 4–7b; (3) vv. 7c–8; (4) vv. 9–10; (5) vv. 11–12; (6) vv. 13–14. The New English Bible (NEB) has eight sentences: (1) v. 3; (2) vv. 4–6; (3) vv. 7a–b; (4) vv. 7c–8; (5) vv. 9–10; (6) vv. 11a–b; (7) vv. 11c–12; (8) vv. 13–14. The Revised English Bible (REB) maintains the two paragraphs and the eight sentences, though the second sentence (vv. 4–6) is divided into 4–5c and 5d–6, while the sixth and seventh sentence of the NEB becomes one in the REB (vv. 11–12). Such divisions are probably not to be taken as randomly made, but seem rather to depend on how the stringed elements of the Greek sentence, which is rather disproportionate, are clustered. In some instances it seems that discourse markers have suggested the clustering of the translated text, but in others the content to be translated seems to have been influenced by current English sentence patterns. However, these are merely hunches since the layout of the translated text does not easily suggest a particular pattern. In the end it may even be the result of taking the Greek text ‘as it stands’ and rendering the flow of argument as best as English style would allow. Nevertheless, many readers will find it difficult to comprehend the gist of the communication, especially when different translations are compared.

The New International Version (NIV) follows the verse divisions of the Greek text more closely, but also renders the passage by eight sentences: (1) v. 3; (2) v. 4; (3) vv. 5–6; (4) vv. 7–8; (5) vv. 9–10; (6) vv. 11–12; (7) vv. 13a; (8) vv. 13b–14. Yet the NIV, while saying in the Study Bible edition that Paul speaks of God’s blessings as through the Father (v. 3), the Son

                                                            NIV New International Version 

217  

(vv. 4–13a) and the Holy Spirit (vv. 13b–14),1 adheres to a twofold paragraph division in the translated text, namely vv. 3–10 and vv. 11–14 similar to that of the NEB/REB.

The New Evangelical Translation (NET) has 11 sentences, almost a new sentence for each verse of the Greek text: (1) v. 3; (2) v. 4; (3) v. 5; (4) v. 6; (5) v. 7; (6) vv. 8–9a; (7) vv. 9b–10; (8) v. 11; (9) v. 12; (10) v. 13; (11) v. 14. However, the argument is arranged in four paragraphs: 1 (v. 3); 2 (vv. 4–9a); 3 (vv. 9b–10); 4 (vv. 11–14).

The Good News Bible (TEV, 1993 British usage) has 15 sentences divided into 5 paragraphs: (1) vv. 3–4a; (2) vv. 4b–8a; (3) vv. 8b–10; (4) vv. 11–12; (5) vv. 13–14.

The number of sentences used in different translations is not as crucial as the number of paragraphs since the number of sentences depends more on style, while the paragraphs involve the clustering of the contents. It is, however, rarely clear why some translations having two paragraphs would break at v. 11, while others such as the Dutch Canisius translation, for example, breaks at v. 7. The New American Bible (NAB) and the Contemporary English Version (CEV) have three paragraphs, namely vv. 3–6, 7–10, 11–14. In the NAB these paragraphs have headings that help the reader to understand the breaks: The Father’s Plan of Salvation (vv. 3–6), Fulfillment through Christ (vv. 7–10), and Blessing for Jew and Gentile (vv. 11–14).

A long sentence strung together as in Eph. 1:3–14 logically allows, or perhaps suggests, that multiple readings are possible due to different foci. This is clearly discernible from the way in which different translations render the passage. One reading of Eph. 1:3–14, quite popular during the fifties and sixties of the twentieth century, but also earlier as well as lately from time to time, is to understand the passage as a hymn. Much has been written on the nature of Christian hymns, and Eph. 1:3–14 was generally part of the discussion.2 But, whether the text as we now have it, is literally a hymn, remains an open question since it is easily likely that Eph. 1:3–14 may merely contain phraseology derived from a hymn or hymns, or may even have been the source from which a hymn developed at a later time. It is, therefore, stretching the point to arrange the text of Eph. 1:3–14 word by word into a strophic pattern since one would then be bound by line lengths that would be required to fit the melody. Many of the proposed suggestions3 have lines that differ so much in length and

                                                            1 For a trinitarian scheme (vv. 4–6, 7–10, 11–14 or 7–12, 13–14), see T. Innitzer, ‘Der Hymnus im 

Epheserbrief 1, 3–14’, ZKT 28 (1904), pp. 612–21; J. Cambier, ‘La bénédiction d’Eph 1, 3–14’, ZNW 54 

(1963), pp. 58–104. 

TEV Today’s English Version 

2 See, among others, H. Coppieters, ‘La doxologie de la Lettre aux Eph’, RB 6 (1909), pp. 74–88; E. 

Lohmeyer, ‘Das Proömium des Ephesersbriefes’, TBl 5 (1926), pp. 120–25; N.A. Dahl, ‘Adresse und 

Proömium des Ephesersbriefes’, TZ 7 (1951), pp. 141–264; C. Maurer, ‘Der Hymnus von Epheser 1 als 

Schlüssel zum ganzen Briefe’, EvT 11 (1951), pp. 151–72; J. Coutts, ‘Ephesians 1, 3–14 and 1 Peter 1, 

3–12’, NTS 3 (1956–57), pp. 115–27; J.M. Robinson, ‘Die Hodajot‐Formel in Gebet und Hymnus des 

Frühchristentums’, in W. Eltester and F.H. Kettler (eds.), Apophoreta (Festschrift E. Haenchen; Berlin: 

Alfred Töpelmann, 1964), pp. 194–235; G. Schille, Früchristliche Hymnen (Berlin: Evangelische 

Verlagsanstalt, 1965); J.T. Sanders, ‘Hymnic Elements in Ephesians 1–3’, ZNW 56 (1965), pp. 214–32; 

R. Deichgräber, Gotteshymnus und Christushymnus in der frühen Christenheit: Untersuchungen zu 

Form, Sprache und Stil der frühchristlichen Hymnen (Göttingen: Vandenhoeck & Ruprecht, 1967), pp. 

65–76; C. Schalk, Key Words in Church Music: New Testament Hymnology (St Louis: Concordia, 1978). 

3 G. Bouman, De Brief aan de Efeziërs (Bussum, 1974). 

 

number if one assumptthat are

A dpays clobe usedTherefofrom thestructurcommunsemanti

they areπατήρ—obscurediscoursverse nu

r from versewould, in

tions for whbeyond the

discourse reaose attentiond as indicatoore, though ie text by hi

re be mappenication foic elements.

e rather mo—Ἰησοῦ Χριe the main ise pointers umbers of th

e to verse thaddition,

hich we hare text. ading, hown to the struors to at leit is not a fiighlighting ted to show

or that matt. Though th

ore significaστοῦ are noissues. The and highlighe Greek te

hat each verassume tha

rdly have ad

ever, is a mucture of theast give accinal readingthe crucial ithe links a

ter. The foese element

ant if the grot contrastioutline giv

ghts the relatxt are retain

rse would hat ancient dequate data

more controe discourse count of wh

g or the onlyindicators.

and dependeollowing outs may be p

rid is not toing elementven below mtionships bened only as

have had to melody w

a. A hymnic

olled readinand to varihy a passag

y reading, itSuch a readencies of thutline of Eresented in

oo fine sincts, and thermarks the cetween the sa reference

have a sepwas differenc reading ca

g in terms ous discourge is read iis a reading

ding requirehe segmentsEph. 1:3–14

a finer grid

ce the itemrefore too fcrucial elemstructural un help.

parate melodnt, it still alls for assu

of the text rse markers in a particug that justifes that the ds of inform4 maps thed as, for exa

ms in the strfine a grid ments recogunits. The tra

218 

dy. Even requires

umptions

itself. It that can

ular way. fies itself discourse ation, or

e crucial ample,

retch καὶ tends to

gnized as aditional

 

219 

220  

The phrase εὐλογητὸς ὁ θεός at the beginning of v. 3 constitutes the main sentence unit to which everything that follows up to the end of v. 14 is syntactically linked as indicated by καθώς that can be read as a marker of cause or reason implying a comparison (LN 89.34).4 That is to say: what follows in vv. 4–14 is justifying why God must be praised, and is in essence similar to what is said in v. 3, namely that God blessed us with every heavenly spiritual blessing because of what Christ did for us—and for that we must praise him. The contents of v. 3 can therefore be understood as a heading to vv. 4–14 in which this content is elaborated on.

A threefold repetition of εὐλογητός by means of εἰς ἔπαινον (vv. 6, 12, 14) continues the main statement, and is echoed in the sound of another meaning of εὐλογέω (act kindly toward, LN 88.69) expressed in v. 3 by εὐλογήσας and εὐλογίᾳ highlighting the notion ‘God must be praised for his kindness’, and reiterated in εἰς ἔπαινον—αὐτοῦ (v. 6) and εἰς ἔπαινον (τῆς) δόξης αὐτοῦ (vv. 12, 14) thus marking the discourse as a type of doxology. Therefore, the overall discourse pattern can be divided into a heading with a threefold repetition of the content of the heading, and expanded by additions (marked 1 + 2, 3 + 4 + 5 and 6 + 7 in the final translated text at the end of this article).

Another remarkable feature of the discourse of Eph. 1:3–14 is the linking of the phrases ἐν ἀγάπη (vv. 4–5) and ἐν πάσῃ σοφίᾳ καὶ φρονήσει (vv. 8–9) occurring with προορίσας and γνωρίσας, participles depending upon ἐξελέξατο in v. 4. These phrases can be read with what precedes as well as with what follows. In the layout of the Greek text above, these phrases are repeated to emphasize that they are links reinforcing the cohesion of the discourse, and followed by a third counterpart in ἐν αὐτῷ (vv. 10–11). Thus we have another threefold link emphasizing the unity of the discourse, and also pointing to the fact that the total sentence (vv. 3–14) is a continuous repetition, with expansions, of the same content introduced by v. 3 and elaborated on by vv. 4–14. The phrase ἐν ἀγάπῃ (vv. 4–5) constitutes the link between the sections marked 1 and 2 in the translated text at the end of the article, and is the basis for grouping 1 and 2 as a larger unit. The same applies to ἐν πάσῃ φρονήσει and ἐν αὐτῷ linking 3, 4 and 5 in the translated text.

The phrase ἐν αὐτῷ (vv. 10–11) is not only part of this triad, but also of a series beginning with ἐν Χριστῷ (v. 3) and followed by ἐν αὐτῷ (v. 4), διὰ Ἰησοῦ Χριστοῦ (v. 5), τῷ ἠγαπημένῳ (v. 6), ἐν ᾧ (v. 7), ἐν αὐτῷ (v. 9), ἐν τῷ Χριστῷ, ἐν αὐτῷ (v. 10), ἐν αὐτῷ, ἐν ᾧ (v. 11), ἐν τᾧ Χριστῷ (v. 12), ἐν ᾧ, ἐν ᾧ (v. 13). The structure of the discourse clearly expands the reiterated theme: ‘God must be praised for his kindness’ by referring to Christ as the central and most important base of God’s kindness, bestowed upon us through our union with Christ, and attested by the Holy Spirit (v. 14) which is the guarantee of the blessings in Christ promised by God. The total discourse of Eph. 1:3–14 is a laudatory elaboration on God’s involvement with us. Each of the seven demarcated units in the following translation defines the initial statement from a different, though related, perspective—a concise credo expressed in one long ingeniously constructed sentence:5 Let us praise God the Father of our Lord Jesus Christ. He was kind to us by giving us every spiritual blessing in heaven through our union with Christ.

                                                            4 J.P. Louw and E. Nida, Greek‐English Lexicon of the New Testament Based on Semantic Domains 

(New York: United Bible Societies, 1988). Domain and sub‐domain numbers are cited as LN in the 

text. 

5 Contra E. Norden, Agnostos Theos (Leipzig: Teubner, 1913), p. 253 n. 1 referring to this sentence as 

a ‘monströseste Satzconglomerat—von einer Periode kann man gar nicht mehr reden’. 

221  

1. For he chose us through our union with Christ, before the world was made, to be devoted to him and without blame in his sight because he loved us.

2. Because of his love he decided to make us his children through Jesus Christ for it pleased

him to want to do it, so that we can praise him for his glorious kindness which he graciously bestowed upon us through our union with his beloved Son.

3. Through our union with Christ we are set free by his death, that is, our sins are forgiven

because of his great kindness which he abundantly bestowed upon us in all wisdom and insight.

4. In all wisdom and insight he made known to us his secret plan because it pleased him to do

it. This he purposed, by means of Christ, to accomplish when the time is right to bring everything in heaven and earth together in union with Christ as head.

5. It is in union with Christ that God also gave us our heritage as he had already decided to do

because of what he had purposed since he does everything according to the plan he has decided on, so that we may praise his glory, we who were the first to have hope through our union with Christ.

6. Through Christ you also heard the true message, that is, the Good News that brought you

salvation. 7. Through Christ God also put his stamp of ownership upon you who believed by means of the

Holy Spirit he had promised, which is the guarantee that we shall receive our inheritance, so that we may experience freedom and praise his glory.

222  

SOME CONSTRAINTS ON DISCOURSE DEVELOPMENT IN THE PASTORAL EPISTLES Stephen H. Levinsohn

As the reader of a text such as one of the Pastoral Epistles (hereafter, PE) encounters each new sentence, one must decide how the author intended the information contained in the sentence to be related to its context. Was the author making a new point, or was he continuing with the previous one? If he was making a new point, was he developing it in some way from the previous one and, if so, how? Alternatively, did he view the two sentences as basically independent of each other? If, in contrast, he was continuing with the previous point, how did he intend the new sentence to be related to what he had already presented?

Although many discourse analysts address these questions, this essay focuses on one particular approach to the issue, namely, that of Blakemore1 who, working within the framework of Sperber and Wilson’s ‘Relevance Theory’,2 proposed that particles such as ‘so’ and ‘also’ guide the hearer as to how to process an utterance. As Blass puts it, ‘if a speaker wants to be sure that the hearer will arrive at the intended interpretation of an utterance, it is in his interest to make the intended context immediately accessible. The particles … enable him to direct the hearer to a particular set of assumptions in an economical way.’3

Titus 1:12 (εἶπέν τις ἐξ αὐτῶν ἴδιος αὐτῶν προφήτης κ.τ.λ. ‘It was one of them, their very own prophet, who said …’) illustrates this point.4 The preferred reading has asyndeton, but variants include the particles γάρ or δέ. If the preferred reading is followed, the reader theoretically has to rely on the context to decide whether the quotation cited makes a new point about Cretans5 or is intended to strengthen the assertion of the previous sentence (vv. 10–11)6 by citing evidence about the character of Cretans. The variant readings have the effect of disambiguating (or changing)7 the intended relationship between vv. 10–11 and v.

                                                            1 D. Blakemore, Semantic Constraints on Relevance (Oxford: Basil Blackwell, 1987). 

2 D. Sperber and D. Wilson, Relevance: Communication and Cognition (Oxford: Basil Blackwell, 1986). 

3 R. Blass, Constraints on Relevance in Koiné Greek in the Pauline Letters (Nairobi: Summer Institute of 

Linguistics, Exegetical Seminar, 29 May to 19 June, 1993), p. 3. Whereas H.P. Grice (‘Logic and 

Conversation’, in P. Cole and J.L. Morgan [eds.], Syntax and Semantics, III [New York: Academic 

Press], pp. 41–58) suggested that particles such as ‘therefore’ and ‘but’ involve the truth‐functional 

logical operator ‘&’ plus an additional ‘conventional implicature’ resulting from the content of the 

lexical item, Blakemore considers that the particles have both non‐truth conditional content (they 

have no effect on the semantic content of the proposition) and semantic content. 

4 The Greek text used is that of the United Bible Societies’ Third Corrected Edition (1983). The English 

translation is based on the New Revised Standard Version, New Testament (1990), but has been 

changed, where necessary, to reflect the Greek more closely. 

5 For instance, J. Banker (Semantic Structure Analysis of Titus [Dallas: Summer Institute of Linguistics, 

1987], p. 55) considers that v. 12 begins the second ground for the exhortation that begins in v. 13b. 

Cf. also n. 7. 

6 ‘The apostle supports his argument by appealing to a venerated Cretan critic of the Cretan 

character’ (D. Guthrie, The Pastoral Epistles [London: Tyndale Press, 1957], p. 200). 

7 Banker’s analysis (Semantic Structure Analysis, p. 65) treats vv. 12–13a as parallel to vv. 10–11, 

rather than as developing from it. This analysis is consistent with the use of asyndeton (S.H. 

223  

12. If γάρ is read, the quotation strengthens the characterization of the ‘many’ who ‘must be silenced’. If δέ is read, the quotation makes a new point about Cretans, which nevertheless represents a development from the previous point in the argument, enabling the author to shift from the need to silence the rebellious deceivers (v. 10) to the exhortation to reprove the Christians (v. 13).

This essay builds on Heckert’s8 demonstration that γάρ, δέ, conjunctive καί and οὖν (plus ἀλλά) have a single basic function. It considers them not to convey a variety of senses,9 but rather always to place the same semantic or pragmatic constraint10 on the reader’s processing. The different senses that are discerned by grammarians are then viewed as ‘contextual effects’,11 that is, the effects of constraining the reader’s processing of the material in particular contexts.

The constraints placed on the reader’s processing by the particles considered in this paper are characterized as follows:

γάρ the material concerned supports previous material (section 1)

δέ the material concerned is a new ‘discourse unit’ that develops from previous material (section 2)

καί the material concerned is added to and associated with previous material (generally, the current ‘discourse unit’; section 3)

οὖν the material concerned is a new point on the main topic under consideration that develops from previous material (section 4).12

                                                                                                                                                                                          Levinsohn, Discourse Features of New Testament Greek: A Coursebook [Dallas: Summer Institute of 

Linguistics, 1992], p. 62; cf. also section 5 of this essay). 

8 J.K. Heckert, Discourse Function of Conjoiners in the Pastoral Epistles (Dallas: Summer Institute of 

Linguistics, 1996). 

9 Cf. S.E. Porter’s definition of a particle (Idioms of the Greek New Testament [Sheffield: JSOT Press, 

1992], p. 204): ‘a word of set form … used for the purpose of introducing subjective semantic 

nuances … to a clause or to the relationship between clauses’. 

10 I question whether ‘developmental’ can be described as a semantic constraint, since it closely 

relates to the author’s purpose in presenting the argument (Levinsohn, Discourse Features, p. 31). 

11 Sperber and Wilson, Relevance, p. 108. 

12 Heckert (Discourse function, p. 9) describes ἀλλά as ‘a marker of global contrast’ that constrains 

the material concerned to ‘weaken or eliminate assumptions’ (Sperber and Wilson, Relevance, p. 

115). 

224  

A ‘discourse unit’ consists of one or more propositions and ‘presents a new development in the discourse’.13 In reasoned argument, for example, it may informally be thought of as the next step or point in the argument.

Strictly speaking, asyndeton should indicate the absence of any constraints on the reader’s processing of the material concerned with reference to previous material. However, I observe in section 5 that, in practice, it tends to signify that the material concerned is not to be viewed as supportive of, added to or developing from previous material.

Although δέ and καί are both found sentence-internally with the same function as that described in this essay, the essay concentrates on the intersentential function of the particles. Relative pronouns and adjectives are considered to function intrasententially, even when editions of the Greek text are so punctuated that they occur sentence initial (e.g. διʼ ἣν αἰτίαν ‘for which cause’, 2 Tim. 1:6). ὅτι is also considered to function only within the sentence (cf. 1 Tim. 4:4). Conjunctive expressions which only occur once or twice in the PE (e.g. διὰ τοῦτο in 2 Tim. 2:10; μέντοι in 2 Tim. 2:19) are not considered in this essay.

Although the conclusions presented in this essay appear to apply to much of the New Testament, it is not claimed that they reflect the way every author uses them. In particular, the use of these particles and asyndeton in John’s Gospel differ significantly in certain respects from that described for the PE.14

1. γάρ

Postpositive γάρ constrains the material associated with it to be processed as supporting previous material. Such a statement is more generic than classifications of γάρ in terms of ‘reason’ or ‘explanation’;15 these senses are considered to derive from the context, rather than to be an inherent part of the function of γάρ.16

In fact, Heckert17 argues that nearly all of the material introduced in connection with γάρ in the PE does provide a reason for previous material (e.g. 1 Tim. 4:8) or for something implied by the previous material (e.g. 1 Tim. 3:13). Nevertheless, ‘occasionally even “reason” seems too strong a relationship to describe the function of the proposition introduced by γάρʼ.18 In support of this claim Heckert cites 1 Tim. 4:10 (εἰς τοῦτο γὰρ κοπιῶμεν καὶ ἀγωνιζόμεθα); ‘The fact “to this end of our toiling and striving” is evidence for the sureness of the saying’ of v. 9 that ‘godliness is of value in every way’ (v. 8b), but scarcely the reason for it.19 It, therefore, seems appropriate to view ‘reason’ as a specific type of support.

                                                            13 S.H. Levinsohn, Textual Connections in Acts (Atlanta: Scholars Press, 1987), p. 179, where it is called 

a ‘development unit’. Cf. pp. 82–85 for discussion of these units in the narrative of Acts. 

14 Cf. Levinsohn, Discourse Features, pp. 39–46. 

15 R.A. Edwards, ‘Narrative Implications of Gar in Matthew’, CBQ 52 (1990), pp. 636–55. 

16 Porter (Idioms, p. 207) mentions an inferential sense which is found frequently in a question. In the 

example he cites (Mt. 27:23), γάρ constrains the question to be processed as an INVITATION to supply 

material that will support the previous demand (‘Let him be crucified!’, v. 22). 

17 Heckert, Discourse Function, pp. 32–36. 

18 Heckert, Discourse Function, p. 36. 

19 Heckert, Discourse Function, p. 36. 

225  

Although γάρ constrains the material concerned to be processed as supporting previous material, rather than introducing a new point, the sentence containing the γάρ may be constructed in such a way that it does contain a new point. Such is the case with Tit. 1:10–11 (εἰσὶν γὰρ πολλοὶ [καὶ] ἀνυπότακτοι κ.τ.λ., οὓς δεῖ ἐπιστομίζειν κ.τ.λ. ‘For there are [also] many rebellious people … who must be silenced …’). Banker20 considers v. 10 to begin the second major section of the book (1:10–3:11) rather than to provide supporting evidence for v. 9, in spite of the presence of γάρ. The ‘continuative’ non-restrictive relative clause21 of v. 11 is the cause of the problem. Characteristically, in sentences which contain such clauses, ‘the information preceding the relative pronoun becomes the ground for what follows’.22 Thus, v. 10 provides material which both supports v. 923 and provides the ground for v. 11.

2. δέ

This section proposes a unitary treatment of postpositive δέ as a particle that constrains the material associated with it to be processed as a new discourse unit that builds on and develops from the previous material.24

Traditional descriptions of δέ distinguish two major functions: an ‘adversative’ or contrastive usage and a ‘connective’ or continuative one.25 However, the very existence of a ‘connective’ usage implies that it cannot be the presence of δέ per se that signals an adversative relationship; other factors must also enter in if the relationship is to be so viewed (cf. below). This leads Winer to propose that δέ indicates movement ‘to something new, different and distinct from what precedes’.26 The adversative and connective senses attributed to δέ then become specific instances of ‘something new, different and distinct’.

Heckert shows that, whether the context suggests an adversative or a connective overtone in the PE, there is always development in the author’s argument from the material that precedes the clause or sentence containing δέ to that of the material containing the particle.27

Heckert cites 1 Tim. 4:8b as a typical context in which δέ occurs and contrast is involved; I produce his reasoning more or less verbatim.28 The use of γάρ at the beginning of v. 8

                                                            20 Banker, Semantic Structure Analysis, p. 48. 

21 G.B. Winer, A Treatise on the Grammar of New Testament Greek (Edinburgh: T. & T. Clark, 1877), p. 

680. 

22 Levinsohn, Discourse Features, p. 186. 

23 ‘The reason why men so highly qualified for spiritual office are especially necessary in Crete is now 

stated’ (W. Hendriksen, I–II Timothy and Titus [Ann Arbor: Cushing‐Malloy, 1957], p. 350). 

24 Cf. Levinsohn, Textual Connections, pp. 87–96, for a detailed discussion on the unitary nature of δέ, 

where it is described as a ‘developmental conjunction’. 

25 Porter, Idioms, p. 208. Porter also mentions an emphatic usage. However, the example he cites 

(Rom. 3:22) is readily analysed as developmental: after introducing ‘a righteousness of God attested 

by the law and the prophets’, the next step in the author’s reasoning is that that righteousness is 

‘through faith in Jesus Christ to all who believe’. 

26 Winer, Treatise, p. 552. 

27 Heckert, Discourse Function, pp. 42–57. 

226  

constrains the verse to be processed as supporting previous material, namely, v. 7b (γύμναζε δὲ σεαυτὸν πρὸς εὐσέβειαν ‘but train yourself in godliness’). The supporting material itself consists of two parts, linked by δέ: ἡ γὰρ σωματικὴ γυμνασία πρὸς ὀλίγον ἐστὶν ὠφέλιμος ‘for bodily training is of some value’ (v. 8a) and ἡ δὲ εὐσέβεια πρὸς πάντα ὠφέλιμός ἐστιν κ.τ.λ. ‘but godliness is of value in every way’ (v. 8b). The first part does not address the topic of ‘godliness’ found in v. 7b. Rather, it provides a ‘foil’29 with which the second part can then contrast. This part, in turn, returns to the topic of v. 7b which v. 8 is supporting. Thus, within the supporting material of v. 8, there is development from less relevant to more relevant material. The presence of δέ, in turn, constrains v. 8b to be processed as a new discourse unit that develops from the previous material.

Heckert cites 2 Tim. 1:5 as a clear instance in which δέ is taken to have a connective, rather than an adversative sense.30 The clause ὑπόμνησιν λαβὼν τῆς ἐν σοὶ ἀνυποκρίτου πίστεως ‘being reminded of your sincere faith’ is modified by a complex relative clause, the two parts of which are linked by δέ: ἥτις ἐνῴκησεν πρῶτον ἐν τῇ μάμμῃ σου Λωΐδι καὶ τῇ μητρί σου Εὐνίκῃ ‘that dwelt first in your grandmother Lois and your mother Eunice’ and πέπισμαι δὲ ὅτι καὶ ἐν σοί ‘and I am sure, [dwells] also in you’.31 Of these two clauses, it is the second that relates directly to the initial clause of v. 5. Once again, the presence of δέ constrains the last clause of v. 5 to be processed as developing from the previous material.

If δέ is not an adversative marker, yet occurs most frequently in contexts in which contrast can be discerned,32 what is it that conveys the adversative sense? Mann and Thompson33 state that two propositions are prototypically in contrast when they are (1) perceived as the same in certain respects, (2) perceived as different in certain respects, and (3) compared with respect to one or more of these differences. Longacre, in turn, considers that ‘there must be at least two opposed pairs of lexical items’ for true contrast to be present.34

In the case of 1 Tim. 4:8, the common element between the two propositions is ἐστὶν ὠφέλιμος ‘is useful’, while two opposed pairs of lexical items occur: ἡ σωματικὴ γυμνασία ‘bodily exercise’ versus ἡ εὐσέβεια ‘godliness’, and πρὸς ὀλίγον ‘for little’ versus πρὸς πάντα ‘for all’. Thus, this verse is a prototypical instance of contrast, whether or not δέ is

                                                                                                                                                                                          28 Heckert, Discourse Function, pp. 47–48. 

29 ‘A “foil” is a constituent that is presented for the purpose of being contrasted or added to in the 

following material’ (Levinsohn, Discourse Features, p. 84). 

30 Heckert, Discourse Function, p. 43. 

31 As Heckert observes (Discourse Function, p. 43), the presence of adverbial καί tends to ensure that δέ is interpreted in a connective, rather than an adversative sense. 

32 According to Heckert (Discourse Function, p. 53), ‘Out of 59 occurrences of δέ in the PE, 38 are 

contrastive’. 

33 W.C. Mann and S.A. Thompson, Antithesis: A Study in Clause Ordering and Discourse Structure 

(Marina del Rey: Information Sciences Institute, 1987), p. 8. 

34 R.E. Longacre, An Anatomy of Speech Notions (Lisse: The Peter de Ridder Press, 1976), p. 104. 

227  

present. Furthermore, such instances of contrast in Greek typically involve topicalization35 of at least one member of one of the opposing pairs, which itself is a signal of a ‘switch’36 to the topicalized element from a corresponding one (stated or implied). In the case of 1 Tim. 4:8, the topicalization of ἡ εὐσέβεια signals a switch from the corresponding element ἡ σωματικὴ γυμνασία. I therefore conclude that, in the context of prototypical contrast, δέ is either redundant or conveys something other than contrast, namely, development.

Once one or more of the characteristics of prototypical contrast are absent, commentators often disagree as to whether δέ is used in an adversative or a connective sense. 2 Timothy 2:23 illustrates this (τὰς δὲ μωρὰς καὶ ἀπαιδεύτους ζητήσεις παραιτοῦ ‘and/but have nothing to do with stupid and senseless controversies’). This exhortation to avoid negative behaviour follows an exhortation to positive behaviour (δίωκε δὲ δικαιοσύνην κ.τ.λ. ‘and/but pursue righteousness …’), which leads Alford to claim that δέ indicates a contrast.37 However, the difficulty (among other factors) of identifying two opposed pairs of lexical items between vv. 22b and 23 leads other commentators to view δέ as continuative, introducing a second admonitio,38 in parallel with that of v. 22a or vv. 20–22.39 A developmental treatment of δέ readily fits the context. Within the general topic of behaviours to practise and to avoid (vv. 14–26), the topicalization of τὰς μωρὰς καὶ ἀπαιδεύτους ζητήσεις indicates a switch to the next subtopic, in the development of the overall topic.

Some commentators interpret δέ in 1 Tim. 4:140 as adversative, and discern a contrast between the material presented in 3:15 and/or 3:16 and the Spirit’s announcement of apostasy: τὸ δὲ πνεῦμα ῥητῶς λέγει ὅτι ἐν ὑστέροις καιροῖς ἀποστήσονταί τινες τῆς πίστεως ‘now/yet the Spirit expressly says that in later times some will renounce the faith’. Other commentators treat δέ as continuative, while many do not translate it at all. The presence of δέ constrains the material concerned to be processed as developing from the previous material. The topicalization of τὸ πνεῦμα also indicates that the primary basis for relating 4:1ff. to its context is a switch to ‘the Spirit’. If the associative nature of καί in 3:16 is given its due weight,41 then 4:1ff. is to be interpreted as developing from some or all of the unit 3:14–16, which in the first instance concerns the author’s motivation for writing. δέ in 4:1 might then mark the development in the argument from the motivation for writing to the urgency of writing, namely, because of the revelation recorded in 4:1–3 that the Spirit has provided.

                                                            35 Topicalization involves the placement of a constituent ‘at the beginning of a clause or sentence … 

to set a “domain” for what follows … and to provide the primary basis for relating what follows to the 

context’ (Levinsohn, Discourse Features, p. 18). 

36 A. Andrews, ‘The Major Functions of the Noun Phrase’, in T. Shopen (ed.), Language Typology and 

Syntactic Description, I (Cambridge: Cambridge University Press, 1985), pp. 62–154 (78). 

37 H. Alford, Alford’s Greek Testament, III (Grand Rapids: Baker Book House, 1980 [1871]), p. 387. 

38 Cf., e.g., Hendriksen, I–II Timothy and Titus, p. 274. 

39 This admonition itself is viewed by some as being a specific example of the exhortation of vv. 14–

15, in parallel with vv. 6–9; cf. R.E. Smith and J. Beekman, A Literary‐Semantic Analysis of Second 

Timothy (Dallas: Summer Institute of Linguistics, 1981), pp. 62–63. 

40 Cf. also Heckert, Discourse Function, p. 46, for discussion of δέ in this verse. 

41 Cf. the discussion of καί in this verse in section 4. 

228  

Although a majority of commentators take δέ in 2 Tim. 2:2042 to be adversative (ἐν μεγάλῃ δὲ οἰκίᾳ οὐκ ἔστιν μόνον σκεύη χρυσᾶ καὶ ἀργυρᾶ ἀλλὰ καὶ ξύλινα καὶ ὀστράκινα ‘now in a great house there are not only vessels of gold and silver but also of wood and earthenware’), the points of contrast with the context are not explicit, and some writers therefore take it to be connective. Once again, a developmental interpretation of δέ satisfactorily fits the context. Having stated (v. 19) that ‘God’s firm foundation stands, bearing this inscription: “The Lord knows those who are his” and “Let everyone who calls on the name of the Lord turn away from wickedness”,’ δέ constrains v. 20 to be processed as a new discourse unit that builds on and develops from the previous one. What the nature of the development is becomes clear when the author moves from the illustration of v. 20 itself to the application in v. 21:43 ἐὰν οὖν τις ἐκκαθάρῃ ἑαυτὸν ἀπὸ τούτων ‘so if anyone cleanses himself from these things’ (cf. ἀποστήτω ἀπὸ ἀδικίας ‘let [him] turn away from wickedness’), ἔσται σκεῦος εἰς τιμήν, ἡγιασμένον, εὔχρηστον τῷ δεσπότῃ ‘he will be a special utensil, dedicated and useful to the master’.

As Heckert notes,44 δέ in 1 Tim. 3:5 occurs in connection with a parenthetical comment that interrupts the list of qualifications for an overseer: εἰ δέ τις τοῦ ἰδίου οἴκου προστῆναι οὐκ οἶδεν, πῶς ἐκκλησίας θεοῦ ἐπιμελήσεται ‘if someone does not know how to manage his own household, how can he take care of God’s church?’ δέ is developmental45 since the material concerned builds on previous material (τοῦ ἰδίου οἴκου καλῶς προϊστάμενον ‘managing his own household well’, v. 4) and makes a new point. Although some versions translate δέ as ‘for’, thus interpreting v. 5 as stating a reason for the injunction of v. 4,46 failure to use γάρ means that Fairbairn is right in observing, ‘it introduces parenthetically a statement which forms an antithesis to the one immediately preceding, yet an antithesis which at the same time constitutes a reason …’47 In other words, the selection of δέ, rather than γάρ, constrains v. 5 to be processed as a new discourse unit that develops from the previous material, even though it happens also to support that material.

The developmental nature of δέ can be seen also by comparing its usage with that of ἀλλά in similar contexts. When ἀλλά links a negative characteristic or proposition with a following positive one in the PE, both are equally significant. When δέ is used, however, the characteristic or proposition associated with δέ is more significant.

A comparison of 1 Tim. 6:17 with 1 Tim. 1:9 illustrates this. In 6:17 (παράγγελλε μὴ ὑψηλοφρονεῖν μηδὲ ἠλπικέναι ἐπὶ πλούτου ἀδηλότητι ἀλλʼ ἐπὶ θεῷ ‘charge them not to be

                                                            42 Cf. also Heckert, Discourse Function, p. 45, for discussion of δέ in this verse. 

43 Cf. section 4 for a discussion of οὖν in v. 21. 

44 Heckert, Discourse Functions, p. 50. 

45 ‘Antidevelopmental’ might be a better term when δέ occurs in connection with parenthetical 

information; cf. Levinsohn, Discourse Features, p. 31 n. 1. It cannot really be said that parenthetical 

material ‘makes a distinct contribution to the argument’ (p. 64). 

46 Cf., e.g., D.E. Hiebert, First Timothy (Chicago: Moody Press, 1957), p. 66: ‘In verse 4 Paul gives the 

requirement as to his domestic relations, while verse 5 presents the reason for it’. 

47 P. Fairbairn, Commentary on the Pastoral Epistles (Grand Rapids: Zondervan, 1956 [1874]), p. 142. 

Cf. also Guthrie, The Pastoral Epistles, p. 93: ‘The parenthetical question in verse 5’ gives ‘rhetorical 

support to the point just made’. 

229  

haughty, or to set their hopes on the uncertainty of riches, but on God’), the rich are both to avoid putting their confidence in riches and to set it on God.48 In contrast, in 1:9 (εἰδὼς τοῦτο, ὅτι δικαίῳ νόμος οὐ κεῖται, ἀνόμοις δὲ καὶ κ.τ.λ. ‘knowing this, that the law is laid down not for the innocent but for the lawless and …’), δικαίῳ is only a foil for the list of types of people that is associated with δέ.49 δέ constrains the list to be processed as developing from the preceding foil.

3. καί

Conjunctive καί, which ‘is used to link items of equal status’,50 constrains the material it introduces to be processed as being added to and associated with previous material. In contrast with δέ, the material it introduces does not represent a new development with respect to the context.51

For most cases of conjunctive καί in the PE, it is pedantically obvious that it associates together the material it conjoins. For instance, in 1 Tim. 2:13–14 (Ἀδὰμ γὰρ πρῶτος ἐπλάσθη, εἷτα Εὕα. καὶ Ἀδὰμ οὐκ ἠπατήθη, ἠ δὲ52 γυνὴ ἐξαπατηθεῖσα ἐν παραβάσει γέγονεν ‘for Adam was formed first, then Eve; and Adam was not deceived, but the woman was deceived and became a transgressor’), it conjoins two sentences which, introduced by γάρ, together support the previous verse.

The final καί of 2 Tim. 4:17 (ὁ δὲ κύριός μοι παρέστη καὶ ἐνεδυνάμωσέν με, ἵνα διʼ ἐμοῦ τὸ κήρυγμα πληροφορηθῇ καὶ ἀκούσωσιν πάντα τὰ ἔθνη, καὶ ἐρρύσθην ἐκ στόματος λέοντος ‘but the Lord stood by me and gave me strength, so that through me the message might be fully proclaimed and all the Gentiles might hear it, and I was rescued from the lion’s mouth’) has the effect of associating the result of the first two actions of the verse with those actions and, developmentally, of treating the verse as a single unit.

Many commentators interpret the καί that introduces the last clause of 2 Tim. 3:11 (καὶ ἐκ πάντων με ἐρρύσατο ὁ κύριος ‘and/yet the Lord rescued me from all of them’) as adversative.53 This cannot mean that καί itself conveys an adversative sense, but rather that καί conjoins propositions that are in an adversative relationship and which under other circumstances might have been linked with δέ. Why, then, was καί used in this passage (associating the propositions together), rather than δέ (treating them as distinct discourse units)? The reason is found in the flow of the argument. This is from οἵους διωγμοὐς ὑπήνεγκα ‘what persecutions I endured’ (v. 11c) to καὶ πάντες δὲ οἱ θέλοντες εὐσεβῶς ζῆν ἐν

                                                            48 Cf. also 1 Tim. 3:3 and 2 Tim. 2:24. 

49 As Levinsohn notes (Discourse Features, p. 84), foils are brought into temporary focus by fronting, 

in anticipation of a switch of attention to a corresponding constituent. 

50 Porter, Idioms, p. 211. 

51 Cf. Levinsohn, Discourse Features, p. 32. Cf. below on distinguishing conjunctive καί from adverbial 

καί. 

52 Within the sentence introduced by καί, δέ in v. 14b functions developmentally. 

53 Hendriksen (I–II Timothy and Titus, p. 292), for instance, says, ‘καί is undoubtedly adversative 

here’. 

230  

Χριστῷ Ἰησοῦ διωχθήσονται54 ‘and all who want to live a godly life in Christ Jesus will also be persecuted’ (v. 12), which is constrained by the presence of δέ to be processed as a new discourse unit that develops from the previous material. Since v. 12 does not develop from the last clause of v. 11, but from an earlier part of the verse, καί is the appropriate particle to associate the verse non-developmentally with the rest of that verse.55

From a discourse perspective, the καί that introduces 1 Tim. 3:16 (καὶ ὁμολογουμένως μέγα ἐστὶν τὸ τῆς εὐσεβείας μυστήριον ‘and without any doubt, the mystery of our religion is great’) is well motivated, without it being necessary to claim that it ‘heightens the force of the predication’.56 The presence of δέ in 4:1 constrains the material associated with it to be processed as developing from the discourse unit that preceded it, a unit which, in the first instance, gives the motivation for writing (3:15) and then, having emphasized the truth of the gospel, expands on the grandeur of the gospel (3:16).57

Distinguishing conjunctive καί from adverbial καί, when the latter begins a clause and no postpositive particle is present, is not always easy.58 To help to differentiate them, Heckert notes that ‘As a conjunction, καί almost invariably links contiguous constituents’, whereas adverbial καί ‘links noncontiguous constituents across clause or sentence boundaries’ and ‘constrains an immediately following word, phrase, or clause to parallel processing with a preceding word, phrase, or clause, whether stated or implied’.59 As far as the development of the discourse is concerned, however, the implications of starting a sentence with καί are very similar, whether it is conjunctive or adverbial; in either case, the material it introduces is to be associated with previous material, rather than representing a new development in the argument.

2 Timothy 1:18b (καὶ ὅσα ἐν Ἐφέσῳ διηκόνησεν, βέλτιον σὺ γινώσκεις ‘and [?] you know very well how much service he also [?] rendered in Ephesus’) illustrates this point. If this καί is viewed as conjunctive, then the sentence is to be conjoined to vv. 16–17 (πολλάκις με ἀνέψυξεν καὶ τὴν ἅλυσίν μου οὐκ ἐπαισχύνθη, ἀλλὰ γενόμενος ἐν Ῥώμῃ σπουδαίως ἐζήτησέν με καὶ εὗρεν ‘he often refreshed me and was not ashamed of my chain but, when he arrived in Rome, eagerly searched for me and found me’), even though they are not contiguous to v. 18b. If it is viewed as adverbial, then ὅσα ἐν Ἐφέσῳ διηκόνησεν is constrained to parallel processing with his service in Rome. Either way, the material introduced by καί is to be associated at least with v. 17.60                                                             54 Adverbial καί constrains πάντες to parallel processing with ‘I’ (Paul). 

55 ‘Καί links sentences W and X if the following events Y do not build upon the event described in X’, 

but on W (Levinsohn, Textual Connections, p. 112). 

56 C.J. Ellicott, The Pastoral Epistles of St Paul (London: Longmans, 1883), p. 194. 

57 G.W. Knight, Commentary on the Pastoral Epistles (Grand Rapids: Eerdmans, 1992), p. 182. 

58 Cf., for instance, 1 Tim. 5:7 and 2 Tim. 3:5b, which R.K. Brown and P.W. Comfort (The New GREEK‐

ENGLISH Interlinear New Testament [Wheaton: Tyndale House Publishers, 1990], pp. 732, 742) gloss 

as ‘and’, but which Heckert (Discourse Function, p. 107) argues are adverbial. 

59 Heckert, Discourse Function, pp. 58, 64. 

60 The presence of καί in Tit. 3:8 (πιστὸς ὁ λόγος· καὶ περὶ τούτων βούλομαί σε διαβεβαιοῦσθαι κ.τ.λ., 

‘the saying is sure and I desire that you insist on these things …’) makes the NRSV paragraphing, which 

introduces a break between the two clauses, very suspect. 

231  

4. οὖν

Heckert identifies two functions of postpositive οὖν in the PE: an inferential one, which he considers to be primary, and a continuative one, in connection with the resumption of an earlier topic, which may or may not have inferential overtones.61 This section proposes a slightly more unitary treatment of οὖν, namely, that its presence always constrains the material concerned to be processed as a new point on the main topic under consideration, which builds on and develops from previous material. Whether there are inferential overtones or not, however, continues to depend on the context. οὖν is used in two specific contexts in the PE: following material of an introductory

nature and following material of a digressional nature, such as that introduced by γάρ to support earlier material on the main topic. I consider these contexts in turn.

The clearest example of οὖν following material of an introductory nature in the PE is found in 1 Tim. 3:2. Verse 1 introduces the topic: πιστὸς ὁ λόγος. εἴ τις ἐπισκοπῆς ὀρέγεται, καλοῦ ἔργου ἐπιθυμεῖ ‘the saying is sure: whoever aspires to the office of bishop desires a noble task’. οὖν then introduces the inferences that are to be drawn from this fact: δεῖ οὖν τὸν ἐπίσκοπον ἀνεπίλημπτον εἷναι κ.τ.λ. ‘a bishop must therefore be above reproach …’ οὖν constrains the material with which it is associated to be processed as a new point on the topic introduced in v. 1, and which develops from it in an inferential way. οὖν in 2 Tim. 2:21 may be interpreted similarly.62 Alternatively, v. 20 may be viewed as

parenthetical (digressional),63 in which case v. 21 resumes the topic of v. 19 (cf. below). In the narrative of John’s Gospel, οὖν is commonly used in connection with a return to the

storyline following a parenthetical comment,64 ‘provided that the associated event represents a new development, as far as the author’s purpose is concerned’.65 A cursory glance at the Pauline epistles reveals that οὖν commonly occurs after supportive material associated with γάρ. In the PE, this is the case in 1 Tim. 2:8, 1 Tim. 5:14 and 2 Tim. 1:8. In each instance, Heckert concludes that, in such situations, the topic that was under consideration before the supportive material occurred is resumed and advanced, but οὖν retains some ‘inferential force; that is, it draws a conclusion from the supportive material introduced by γάρ’.66

Heckert also classifies οὖν in 1 Tim. 2:1 and 2 Tim. 2:1 as resumptive, returning to a main theme following a digression. For instance, concerning 1 Tim. 2:1 (παρακαλῶ οὖν πρῶτον πάντων ποιεῖσθαι δεήσεις προσευχὰς ἐντεύξεις εὐχαριστίας ὑπὲρ πάντων ἀνθρώπων ‘first of all, then, I urge that supplications, prayers, intercessions and thanksgivings be made for

                                                            61 Heckert, Discourse Function, p. 96. 

62 Cf. Heckert, Discourse Function, p. 97. 

63 Cf. R.C.H. Lenski, The Interpretation of St Paul’s Epistles to the Colossians, to the Thessalonians, to 

Timothy, to Titus and to Philemon (Columbus: Wartburg, 1937), p. 807: ‘∆έ is not adversative, for it 

does not introduce a contrary thought; it is parenthetical and introduces a preliminary thought that 

is preparatory to what follows’. 

64 Cf., e.g., V.S. Poythress, ‘The Use of the Intersentence Conjunctions δέ, οὖν, καί and Asyndeton in 

the Gospel of John’, NovT 26 (1984), pp. 312–40 esp. p. 327. 

65 Levinsohn, Discourse Features, p. 44. 

66 Heckert, Discourse Function, p. 122. 

232  

everyone’), he comments, ‘It is not apparent that the function of οὖν in this text is inferential’, that is, that it relates inferentially to the comments of 1:19–20 about those who have suffered shipwreck in the faith. ‘Not only does 2:1 resume the theme of what Timothy is directed to do, but it also develops it by moving to a new point’.67

In such instances, then, the presence of οὖν constrains the material with which it is associated to be processed as a new point on the main topic under consideration, building on and developing from that earlier material, but without inferential overtones.

It could be argued that, when οὖν follows supportive material associated with γάρ, the main topic has never really been left, since the material immediately preceding οὖν supports an aspect of that topic. It could then be generalized that οὖν constrains an inferential relationship if and only if the main topic is still under consideration in the material immediately preceding the use of the particle.

The use of οὖν following supportive material associated with γάρ may be contrasted with the use of δέ in the same situation. Each instance in the PE in which δέ follows supportive material associated with γάρ also involves topicalization of a different subject from that of the previous material. In other words, following this supportive material, the epistle develops in connection with a switch from the previous subject or topic to a different one.

This is illustrated in 2 Tim. 3:1068 (σὐ δὲ παρηκολούθησάς μου τῇ διδασκλίᾳ κ.τ.λ. ‘but you have observed my teaching …’). This verse follows comments about the fate of ‘those who oppose the truth’: οὐ προκόψουσιν ἐπὶ πλεῖον, ἡ γὰρ ἄνοια αὐτῶν ἔκδηλος ἔσται πᾶσιν ‘they will not make much progress, because their folly will become plain to everyone’. Following this supportive sentence, it is not the topic of ‘those who oppose the truth’ that is resumed. Rather, δέ constrains v. 10 to be processed as a new discourse unit which develops from the previous material, while the topicalization of σύ and the content of the verse make it clear that that development is contrastive: ‘you’ is opposed to ‘those who oppose the truth’ and ‘observed my teaching’ is opposed to ‘oppose the truth’ (v. 8).

I conclude that, whereas both δέ and οὖν constrain the material they are associated with to be processed as a new discourse unit that develops from previous material, they differ in that, when οὖν is used, the previous main topic continues to be considered, whereas no such constraint applies to δέ.

5. Asyndeton

Strictly speaking, the absence of any connecting particle between sentences in the Greek text should imply only that the author offered no processing constraint on how the following material is to be related to its context. In practice, however, the author of the PE tends to use a particle whenever an associative, developmental or supporting relationship is to be discerned. Consequently, asyndeton tends to signify ‘not associative, not developmental, not supporting’.

One place where asyndeton is commonly found is between paragraphs with different topics,69 when the topic of the new paragraph is not considered to develop from, be parallel to or support that of the previous one. Thus, the transition from the opening salutation to the

                                                            67 Heckert, Discourse Function, p. 99. 

68 Other examples include 1 Tim. 6:9, 11 and 2 Tim. 4:5. 

69 Cf. Levinsohn, Discourse Features, p. 62. 

233  

body of each letter is characterized by asyndeton,70 as is the transition from the body of the letter to its closure.71

Similarly, both major and minor topic changes in the PE are often characterized by asyndeton. In 1 Tim. 3:1, for instance, the absence of a constraining particle implies that no developmental connection is to be made between ‘Directions with respect to Public Worship’ (ch. 2) and ‘Directions with respect to the Institution of the Offices’.72

In turn, topic changes that nevertheless involve a degree of parallelism may be characterized by asyndeton plus ὡσαύτως ‘similarly’.73

Asyndeton is found also between individual propositions that are viewed as independent of each other. Thus, in Tit. 3:15 (ἀσπάζουταί σε οἱ μετʼ ἐμοῦ πάντες. ἄσπασαι τοὺς φιλοῦντας ἡμᾶς ἐν πίστει ‘all who are with me send greetings to you. Greet those who love us in the faith’), though both sentences involve greetings, they are not strictly parallel, and the second does not support or develop from the first.

Other relations between propositions or groups of propositions that are characterized by asyndeton include:74

• Evaluations of previous material, commonly involving the topicalization of the demonstrative οὗτος: τοῦτο75 καλὸν καὶ ἀπόδεκτον ἐνώπιον τοῦ σωτῆρος ἡμῶν θεοῦ ‘this is right and is acceptable in the sight of God our Saviour’ (1 Tim. 2:3). Summaries of previous material are similar: ταῦτα λάλει κ.τ.λ. ‘speak these things …’ (Tit. 2:15).

• Orienter-Head relationships, in which the first proposition introduces the second: εἶπέν τις ἐξ αὐτῶν ἴδιος αὐτῶν προφήτης, Κρῆτες ἀεὶ ψεῦσται, κ.τ.λ. ‘it was one of them, their very own prophet, who said, “Cretans are always liars …” ’ (Tit. 1:12).

• Head-Specific relationships, in which the second proposition gives a specific instance of the

more generic proposition that precedes it: σὺ δὲ λάλει ἃ πρέπει τῇ ὑγιαινούσῃ διδασκαλίᾳ. πρεσβύτας νηφαλίους εἷναι, κ.τ.λ. ‘but as for you, teach what is consistent with sound doctrine. The older men are to be temperate …’ (Tit. 2:1–2).76

In general, then, asyndeton is associated with specific relationships between propositions

or groups of propositions. On occasion, however, propositions on the same theme are

                                                            70 1 Tim. 1:3; 2 Tim. 1:3; Tit. 1:5. 

71 1 Tim. 6:21b; 2 Tim. 4:19, 22; Tit. 3:15. 

72 Hendriksen, I–II Timothy and Titus, pp. 90, 116. 

73 1 Tim. 2:9; 3:8; Tit. 2:3, 6; contrast Tit. 2:9, in which the author does not discern significant 

parallelism between the instructions to slaves and those to earlier groups. The combination of 

ὡσαύτως plus adverbial καί, however, implies a close association between the constituents that are 

to be processed in parallel; cf. 1 Tim. 5:25. 

74 Cf. also Levinsohn, Discourse Features, pp. 61–64; this list is not exhaustive. 

75 γάρ is a variant, in which case 1 Tim. 2:3 would be constrained to be processed as supporting vv. 1–

2. 

76 For a detailed analysis of asyndeton in Tit. 1:10–16, cf. Levinsohn, Discourse Features, pp. 62–63. 

234  

juxtaposed and treated formally as independent of each other, when a constraining particle might appear to be appropriate. Such is the case in 1 Tim. 5:20, for which a variant with δέ exists: τοὺς ἁμαρτάνοντας [δὲ] ἐνώπιον πάντων ἔλεγχε ‘as for those who persist in sin, rebuke them in the presence of all’. Concerning vv. 17–25, White observes that, whereas they ‘form one section, marked by one prominent topic, the relation of Timothy to presbyters … the sequence of thought in these concluding verses of the chapter is not formal and deliberate’.77 The effect of adding δέ is to constrain v. 20 to be processed as developing from the previous material, thus excluding the possibility that τοὺς ἁμαρτάνοντας might refer to sinners in the congregation as a whole, rather than just to presbyters.

2 Timothy 4:18 (ῥύσεταί με ὁ κύριος ἀπὸ παντὸς ἔργου πονηροῦ ‘the Lord will rescue me from every evil attack’) is another instance in which groups of propositions are treated as formally independent of each other, yet their content suggests a semantic relationship between them that might have merited a constraining particle. A number of commentators consider this verse to be ‘a conclusion drawn from the preceding verse’,78 and καί might have been expected (had it not introduced the previous clause, perhaps), as its inclusion in the Textus Receptus suggests. Nevertheless, the relationship between the verses is left implicit; no constraining particle occurs.

I conclude that, whereas the author of the PE tends to juxtapose sentences to indicate ‘not associative, not developmental, not supporting’, the possibility still exists, in individual passages, for asyndeton to imply solely an absence of a constraint on processing, in which case the reader must use the general principles of relevance to interpret the relationship between each new proposition and its context.

6. Conclusion

This essay has sought to show, not only that describing the functions of γάρ, δέ, καί and οὖν in terms of constraints on processing gives meaningful insights into the discourse development of the Pastoral Epistles, but also that approaching the particles from the point of view of different senses in which they may be used is misdirected. Thus, for example, it is a mistake to talk of an adversative versus connective sense for καί or for δέ. Rather, since either may link propositions in an adversative or a non-adversative relationship, their functions need to be described in different terms, such as the distinct constraints that they put on the processing of the material with which they are associated.

In summary, δέ and οὖν differ from γάρ and καί in that the first pair is developmental in nature, whereas the second pair is not. γάρ and καί constrain the material concerned to be processed as respectively supporting versus adding to previous material (usually, the current discourse unit). δέ and οὖν both constrain the material concerned to be processed as a new discourse unit that develops from previous material; οὖν adds the additional constraint that the new unit or point is on the main topic that was previously under consideration.

                                                            77 N.J.D. White, ‘The First and Second Epistles to Timothy and the Epistle to Titus’, in W.R. Nicoll (ed.), 

The Expositor’s Greek Testament, IV (Grand Rapids: Eerdmans, 1970 [1909]), p. 136. 

78 E.E. Minor, An Exegetical Summary of 2 Timothy (Dallas: Summer Institute of Linguistics, 1992), p. 

128. Thus, Hendriksen (I–II Timothy and Titus, p. 327) observes, ‘From this experience of the past the 

apostle draws encouragement’. 

235  

‘LET NO ONE DISREGARD YOU!’ (TITUS 2:15): CHURCH DISCIPLINE AND THE CONSTRUCTION

OF DISCOURSE IN A PERSONAL, ‘PASTORAL’ EPISTLE Ernst R. Wendland

1. Introduction: Why Should Titus Be ‘Disregarded’?

The following composite quotation pretty well summarizes the opinion of many contemporary scholars with regard to the relative literary merits and theological quality of the epistle headed ΠΡΟΣ ΤΙΤΟΝ:1 The author of the Pastorals writes good koinē Greek, but his prose is pedestrian … The Pastorals are made up of a miscellaneous collection of material. They have no unifying theme; there is no development of thought … There seems to be nothing very distinctive about Titus … This is why one is led to suspect that Titus was written last of all and that the author was beginning to run short of material … If they are Pauline, [the Pastorals] represent a dismal conclusion to Paul’s writings … In addition to its allegedly ‘monotonous style’ in common with the other Pastorals, Titus (like Jude) is often discriminated against because of its size and content, that is, it is ‘viewed as a miniature 1 Timothy’, hence hardly worth studying on its own.2 No doubt this is why one does not find many articles devoted to Titus in recent journals of biblical study. Why waste time on topics written in such lackluster prose that are already found elsewhere in the ‘genuine’ Pauline corpus? But as I will demonstrate in this essay, while Paul’s style on the microlevel of discourse may not be especially noteworthy—though it is by no means as dull as some have characterized it—his manner of writing on the macrolevel is exceptional. This textual construct embeds the implicit illocutionary and theological framework for the entire composition, which on closer scrutiny reveals an extremely felicitous welding together of subject and structure in the common interest of communicative effectiveness.3

                                                            1 A.T. Hanson, The Pastoral Epistles (NCB; Grand Rapids: Eerdmans, 1982), pp. 3, 42, 47, 51. Hanson 

feels that ‘the adaptation of Pauline material for use in his own day is one of the author’s leading 

literary characteristics … [but he] falls far short of Paul in almost every respect’ (pp. 28, 47). In the 

absence of what I consider to be sufficient credible evidence to the contrary (e.g. Hanson, Pastoral 

Epistles, pp. 2–6), I retain the traditional view of Pauline authorship, as cogently defended, for 

example, in G.W. Knight III, Commentary on the Pastoral Epistles (NIGTC; Grand Rapids: Eerdmans, 

1992), pp. 21–52, and in E.E. Ellis, ‘Pastoral Letters’, in G. Hawthorne, R. Martin and D. Reid (eds.), 

Dictionary of Paul and his Letters (Downers Grove: InterVarsity Press, 1993), pp. 659–61. I will 

therefore refer to the author of this epistle as ‘Paul’ and designate ‘Titus’ as his primary, ‘implied’ 

addressee. 

2 G.D. Fee, 1 and 2 Timothy, Titus (NIBC; Peabody, MA: Hendrickson, 1988), pp. 24, 11. In addition to 

the supposed monotony of ‘repetition’ (cf. the following analysis), C.B. Puskas detects a ‘prudish 

concern for preciseness’ (The Letters of Paul: An Introduction [Collegeville, MN: Liturgical Press, 

1993], p. 174). 

3 No scholar pays much attention to the larger text organization of Titus. Instead, questions of 

authorship and supposed compositional types take up much of the discussion in commentaries and 

other studies of the letter. Hanson calls attention to a ‘peculiar feature which runs through the three 

epistles’, namely, ‘his habit of beginning a theme, dropping it, and then resuming it’ (Pastoral 

Epistles, p. 46). However, he does not see the structural patterning or the thematic significance that 

is inherent in this important device. Puskas outlines Titus analytically as an instance of classical 

‘deliberative’ rhetoric (Letters of Paul, pp. 184–86), but this approach fails to reveal the many 

236  

To be sure, Paul’s manner of writing may have been limited to a certain degree stylistically by his larger didactic, denunciatory and relational purposes. Such concerns were perhaps made more pressing due to the immediacy with which they had to be effected with respect to the current situation on Crete. However, the genius of this personal letter is reflected in the masterful way in which the apostle puts the whole piece together to reinforce his message of heartfelt advice to a ‘true son’ in their ‘common faith’ (1:4). Here we have a theologically and pastorally vital appeal that is, as so often the case in the Pauline corpus, concisely summarized in the opening words of the text. This initial epistolary segment not only ‘set[s] the social and interpersonal context of the entire discourse’,4 but it also provides us with a neat topical summary. Thus Paul’s encouragement to Titus revolves around the dynamic relationship that integrates a tight cluster of key terms found concentrated in the first two verses: πίστιν ‘faith’, ἐπίγνωσιν ‘knowledge’, ἀληθείας ‘truth’, εὐσέβειαν ‘godliness’, ἐλπίδι ‘hope’ and ζωῆς αἰωνίου ‘eternal life’. The significant feature of this epistle is the artful manner in which its structure is utilized to heighten the impelling rhetorical power of this thematic nucleus of content-generating notions.5 The letter exemplifies literary excellence in the service of religious meaningfulness (i.e. consecrated Christian commitment) that has lost neither its urgency nor its relevance to the people of God.

I will necessarily have to be selective in my analysis of the many interesting stylistic features that characterize the text of Titus. I will therefore focus my attention on the macrostructural level of composition and the functional-thematic implications of this aspect of the epistle’s overall organization. Within the general framework of an opening, body and closing that was ‘obligatory to the Hellenistic letter’,6 we find manifested a dual serial and chiastic format that is forged primarily by the lexical recursion of key concepts both to demarcate and also to distinguish certain focal points in the discourse development. I will apply a basic rhetorical-structural methodology to reveal how these larger patterns of textual arrangement serve to highlight the epistle’s gospel-centered foundation and to reinforce its principal paraenetic (admonitory-hortatory) and didactic purposes. It is hoped that such a procedure will provide a new perspective on this underrated, oft ‘disregarded’, personal and pastoral letter. Thus I do not view Titus as being an example of ‘pedestrian prose’ or some ‘ingenious pastiche’.7 I rather see it as a very carefully constructed and cohesive text in which the author has utilized certain features of style and structure to accentuate his essential message and to facilitate the attainment of his rhetorical objectives.

                                                                                                                                                                                          meaning‐bearing intratextual compositional patterns and correspondences (similarities as well as 

contrasts with respect to form and/or content). 

4 J.T. Reed, ‘Discourse Features in New Testament Letters with Special Reference to the Structure of 1 

Timothy’, JOTT 6.3 (1993), pp. 228–52 (233). 

5 For a diagram of the intricate structuring that is manifested in 1:1–4, see J. Bateman, ‘Context‐

Based Analysis as a Basis for Translation’ (Part One), Notes on Translation 9 (1995), pp. 37–60, esp. p. 

41. 

6 Reed, ‘Discourse Features’, p. 233. 

7 Hanson, Pastoral Epistles, pp. 3, 47. 

237  

2. A Top-to-Bottom Structural Overview of Titus

I will begin with a summary of my preferred method of doing discourse analysis.8 It is an approach that is quite simple in its essentials, but one that may be complemented eclectically by a number of other text-oriented studies to provide a more complete overview of a given biblical pericope or book.9 The core of the approach involves a careful examination of the composition to detect any and all types of verbal repetition, whether this is phonological, lexical or morpho-syntactic in nature. Again, to limit the scope of this investigation, I will concentrate upon word-level recursion since this is the easiest to discern and to apply in the process of interpretation. However, other forms of reiteration, including the more abstract conceptual and pragmatic varieties, will be noted when they appear to be important to the author’s overall plan and purpose.

Significant instances of repetition (i.e. not just any word or two) thus serve to outline the basic structural and semantic framework of a text, at least in its broad, initial contours. This provisional hierarchical framework consisting of various strata of compositional wholes and parts must next be either confirmed or revised through the application of additional diagnostic discourse procedures. One such method is to note any special concentration of literary or rhetorical microfeatures, such as figurative language, alterations in the expected word order, the insertion of intensifiers (e.g. vocatives, qualifying adjectives and adverbs), rhetorical questions, irony, allusion, a set of synonyms or antonyms, and so forth. Other distinctive linguistic elements found within a particular passage may also be suggestive of a sectional beginning (i.e. aperture), ending (closure) or high point (peak) in the writer’s development, for example: a transitional conjunction, temporal phrase, imperative verb, asyndeton, or text citation (‘intertextuality’). Another common marker of the presence of a discourse boundary involves a shift in features such as the addressee(s) or the prevailing topic, tone or functional purpose of the text (e.g. a switch from exhortation to instruction, which occurs a number of times in Titus). Of course, the overall paradigmatic (i.e. pertaining to lexical distribution), pragmatic (e.g. imperatival versus optative mood), syntactic (sentential), and semantic (propositional) structures of the composition must also be taken into consideration in order to promote an approach that is flexible, holistic, balanced and integrated. Thus the analytical procedures may be modified as needed according to the particular interests of the analyst and the nature or genius of the text under study. a. Titus 2:15—The Compositional Core The methodology as outlined above was applied to Titus in its entirety. The results are as follows, beginning from the ‘top’ (higher levels) of the text and proceeding to the smaller, included units (i.e. the ‘bottom’).10 The fundamental structural ‘hinge’ or functional

                                                            8 For a more detailed description of this methodology, see my articles, ‘Rhetoric of the Word: An 

Interactional Discourse Analysis of the Lord’s Prayer of John 17 and its Communicative Implications’, 

Neot 26 (1992), pp. 59–88, esp. pp. 70–76; ‘Cohesion in Colossians: A Structural‐Thematic Outline’, 

Notes on Translation 6 (1992), pp. 28–62, esp. pp. 30–32; and ‘ “Dear Children” Versus the 

“Antichrists”: The Rhetoric of Reassurance in First John’, JOTT 11 (1998), pp. 43–68. 

9 For example, see Reed (‘Discourse Features’) for a structural analysis of 1 Timothy in terms of its 

obligatory and optional epistolary elements, in particular, its characteristic compositional formulas. 

10 Of course it is impossible to divorce completely ‘top‐down’ and ‘bottom‐up’ textual operations, for 

in most types of analysis there is a continual shifting and interaction between the two. But the major 

238  

‘keystone’ of the letter is found in 2:15, a verse which is distinguished semantically by the seemingly incidental (and sometimes regarded as ‘misplaced’) remark by Paul concerning the self-image of his younger colleague: ‘Don’t let anybody disregard (“disparage” or “despise”—περιφρονείτω) you!’ With this brief personal and pastoral remark (undoubtedly intended also for the congregations that Titus served) Paul intimates something of the urgency of his various epistolary instructions (‘as I commanded you’, 1:5). These are encapsulated in the three initial imperative verbs of this same verse: ‘teach’ (λάλει, lit. ‘speak’, but cf. 2:1), ‘encourage’ (παρακάλει) and ‘rebuke’ (ἔλεγχε). Here we have a neat synopsis of the functional import of the entire letter—that is, the primary interpersonal illocutions which Paul desired to have accomplished in relation to the Cretan Christians through the agency of Titus’s ministry among them.

The manner in which each of these three communicative objectives was to be carried out is clearly in focus: ‘with all authority’ (μετα πάς ἐπιταγῆς), an injunction that allows for no laxness or leeway. The personalized final command, ‘let none of them look down on you’ (TEV), is in effect a negative restatement of the preceding. It is intended to reinforce the ‘authority’ (+ responsibility) that Titus was to wield as he dealt decisively with the potentially volatile disciplinary situation on Crete—that is, to ‘set things in order’ there (ἐπιδιορθώσῃ, 1:5). The issue of ecclesiastical authority was of considerable importance since it reflected upon both the reliability of the teaching that Titus was to give (2:1) and also the credibility of the example that he was to live (2:7). Indeed, the pastoral magisterium of Titus was closely bound up with the apostolic eminence of Paul himself as it is clearly established at the very outset of the epistle (1:1–3). The derived authority of both apostle and protégé rested ultimately upon the divinely sanctioned and delivered foundation of ‘sound [healthful] doctrine’ which they were commissioned to proclaim (ἐν τῇ διδασκαλίᾳ τῇ ὑγιαινούσῃ, 1:9; cf. 1:3; 2:1–2, 10b, 15; 3:8). b. Titus 1:1–2:15 The pivotal nature of 2:15 within the epistle as a whole is signaled by the various structural and thematic links that it has with other key points in the discourse.11 To begin with, the particular ‘authority’ being spoken about is certainly not human with regard to source, sanction or scope: it is patently spiritual in nature. That is shown by the term’s initial occurrence at the climactic close of Paul’s introductory proclamation at the start of his letter to Titus. The apostle’s entire ministry—and that included his overseeing of younger colleagues such as Titus and Timothy—was conducted ‘according to the authority/command (κατʼ ἐπιταγήν) of our Savior, God’ (1:3). This prescription encompasses the triad of pastoral activities referred to above: In short, there was to be a general ‘teaching (διδαχήν) of a trustworthy message’ that would serve both to ‘encourage’ (παρακαλεῖν) the faithful and, on the other hand, to ‘refute’ (ἐλέγχειν) all opponents of the truth (1:9). It is important from a compositional perspective to note that v. 9 of ch. 1 comes as the conclusion and motivating

                                                                                                                                                                                          focus of my attention did begin with the macrofeatures and the larger segments of the letter and did 

move towards the lesser compositional constituents that were incorporated within these. 

TEV Today’s English Version 

11 Several linguistic markers act in concert to signal the break between vv. 14 and 15, for example: a 

shift to the imperative mood, the present tense, a second‐person singular focus, asyndeton, and the 

syntactic fronting of ταῦτα ‘these things’. Asyndeton plus a clear shift in topic follows v. 15. 

239  

grounds for Paul’s first set of instructions to Titus, that is, concerning the appointing of theologically and behaviorally qualified ‘elders’ (1:5–9).12

The notion of ‘instruction’ (διδασκαλία, 1:9) is prominent throughout the lengthy paraenetic portion of ch. 2, that is, 2:1–10, where it structurally ‘bounds’ the unit (at the beginning and ending, i.e. inclusio). In addition, this crucial concept demarcates externally, that is, by enclosure (exclusio), the preceding segment (i.e. 1:10–16) in conjunction with the qualitative term ὑγιαινούσῃ ‘healthy’ (in chiastic order, 1:9; 2:1). ‘Instruction’ also partially circumscribes the following compositional unit, 2:11–14, by means of the contextually synonymous verbals, παιδεύουτα ‘train, discipline’ (v. 12) and λάλει ‘speak about’ (v. 15). Through such patient and persistent teaching, individual members as well as the body as a whole would be ‘encouraged’ (παρακάλει, 2:6) and enemies would be given no grounds to ‘disparage’ the word (βλασφημῆται, 2:5, the latter term in the present context being a synonym of ‘refute’ and an antonym of ‘encourage’).13 The enjoined disciplinary action of ‘sharp rebuking’ (ἔλεγχε ἀποτόμώ 1:13) is fore-grounded at the center of the prior descriptive section in which the apostle confronts the ‘rebellious’, ‘deceitful’, ‘denying’ and ‘disobedient’ leaders who were threatening to destroy all his gospel-based work on Crete and that of Titus as well (1:10, 16, i.e. another sectional inclusio). How?—’by teaching (διδάσκοντες) what they ought not [to teach]’(1:11).14 c. Titus 2:15–3:15 Moving then to the opposite textual side of 2:15, we observe several other major conceptual correspondences which support the decision to treat this verse as the structural pivot (‘fulcrum’ or ‘midpoint’) of the entire epistle. First of all, in 3:1 we encounter a synonymous imperative in ὑπομίμνῃσκε ‘remind’, which combines the notions of ‘teach’ and ‘encourage’. That is coupled with a corresponding apical exhortation at the close of this new principal discourse unit in 3:8 (i.e. inclusio with 3:1 + epiphora with 2:15): ‘and I want you to speak confidently (διαβεβαιοῦσθαι, “teach!”) concerning these things’. ‘These things’, namely whatever I am telling you in this letter (but perhaps more specifically the ‘reliable teaching’ of 3:3–7), is reiterated later in v. 8 to form another transitional overlap to the next segment (i.e. 3:8–11, anadiplosis).

In other words, contrary to all versions and most commentators (plus the punctuation of the standard Greek text),15 I view the statement starting with ταῦτα in 3:8c (fronted as in 2:15

                                                            12 To supplement the well‐known literary‐structural concept of inclusio, I use the terms anaphora and 

epiphora to designate the occurrence of points of significant lexical‐thematic correspondence that 

are found at the respective opening and closing portions of distinct compositional units within a text 

(Wendland, ‘Rhetoric’, p. 74; ‘First John’, p. 49), e.g. 1:10; 3:9 = anaphora and 1:9; 2:15 = epiphora. 

13 These two contrastive terms appear at the compositional midpoint of the four segments that 

comprise 2:1–10, i.e. ‘old(er) men’ (v. 2), ‘old(er) women’ (v. 3), ‘younger men’ (v. 6), and ‘slaves’ (v. 

9). A significant topical convergence in such a position of ‘tail‐head’ structural overlap (i.e. vv. 5b–6a) 

is termed anadiplosis (Wendland, ‘Rhetoric’, p. 74; ‘First John’, p. 48). 

14 For a good overview of the legalistic, ultra‐Judaistic, opportunistic and schismatic dogma that Titus 

was encouraged to ‘refute/rebuke’, see Knight, Pastoral Epistles, pp. 11–12, 25–28, 295–96. Knight 

lists some of the major intertextual parallels between Titus and 1 Timothy on pp. 256 and 353–54. 

15 A notable exception is J. Banker, Semantic Structural Analysis of Titus (Dallas: Summer Institute of 

Linguistics, 1987), pp. 111–12. 

240  

for emphasis and to serve as a compositional marker) to be the contrastive beginning of a second paragraph in which the false teachers on Crete are both described and proscribed. This summary unit corresponds to its more detailed development in 1:10–16, one that begins in like fashion (i.e. structural anaphora) with a contrastive reference to the false teachings of the troublemakers (e.g. ματαιολόγοι ‘useless talkers’, 1:10).16 That section, like this one, is similarly preceded by a reference to ‘the reliable message’ (πιστὸς ὁ λόγος) which thematically undergirds and emphatically rounds out structurally parallel admonitions to sound teaching and good living (1:9; 3:8a–b, i.e. epiphora; cf. a similar juxtaposition of thematic notions in 2:14b–15). 3:8a–b also corresponds with 2:15 in appearing as the third element in a larger parallel formation that consists of an exhortation to teach (2:1–10; 3:1–2), an expression of the evangelical motivation for such action (2:11–14; 3:3–7), and a concluding summary admonition (2:15; 3:8a–b). At this juncture, I would call attention to the proficient manner in which such compositional parallels and topical resonances are established through reiteration within the discourse. They all serve to confirm and to complement one another to give us a progressively more accurate picture of the main points and purposes of this text in terms of the whole and its parts.

Moving back to 3:1 we notice another expression that is of vital significance to this letter, for it complements the various terms involving some manner of ‘speaking’ that have been discussed thus far. This is the imperative that Christians must ‘be ready [to do] every kind of good work (ἔργον ἀγαθόν)’. A similar phrase (but in the plural) appears in the verse immediately preceding 2:15: ‘zealous [to do] admirable works (καλῶν ἔργων)’, thus bracketing and thereby setting off (i.e. exclusio) our focal ‘hinge’ passage. Another instance of this same phrase is found in 3:8b (καλῶν ἔργων) to form an inclusio with 3:1. Further examination reveals how important this exhortation to ‘do good’ is to the entire epistle, semantically as well as structurally. In fact, it forms the other essential half of the thematic equation that summarizes the pragmatic component of the discourse and serves as the rhetorical-structural ‘backbone’ which gives a sense of overall shape and direction to the message as a whole.

In other words, the pastoral activities of ‘teaching’ (plus ‘encouraging’) the ‘truth’ and ‘rebuking’ the things ‘that really ought not be done/taught’ (ἃ μὴ δεῖ, 1:11) can never be credible or convincing unless they are explicitly confirmed by one’s own lifestyle. In the case of all believers, this may be summed up simply as ‘godliness’ (εὐσέβειαν, 1:1), which is equivalent to what is ‘admirable’ (2:14) or ‘good’ (3:1) in the eyes of God and one’s fellow human beings as well, including religious outsiders (an important concern of the letter to Titus; cf. 2:5b, 8b, 10b; 3:2b).17 This primarily epiphoric (unit-final) concept recurs for emphasis at the close of the ‘body’ of the epistle in 3:14 (καλῶν ἔργων προστᾶσθαι; cf. v. 8) to forge a double inclusio around its median in 2:15. An obvious equivalent also appears in a prominent contrastive conclusion at the end of ch. 1 (πᾶν ἔργον ἀγαθόν, 1:16) and as a verbal expression in 1:8 (φιλάγαθον), while an interesting combination with ‘faith’ occurs at the outer compositional border in 2:10 (i.e. πᾶσαν πίστιν ἐνδεικνυμένους ἀγαθήν).                                                             16 The phonological similarity here may help to draw aural attention to the conceptual 

correspondence. Although these two passages are relatively far removed from each other in the text, 

their similar content and reference coupled with strong language would tend to make them more 

memorable, hence conceptually relatable. 

17 T. Schreiner makes the important observation that the major segments that constitute the 

exhortation of 2:1–10 each conclude with a purpose clause that demonstrates Paul’s concern ‘about 

the witness of believers in the world’ (Interpreting the Pauline Epistles [Grand Rapids: Baker Book 

House, 1990], p. 125). 

241  

d. 2:15 and the Message of Titus as a Whole All the patterning described above is not simply an instance of random distribution or a fortuitous placement of lexical items, for considered as a group along with other, similarly positioned key terms, a significant arrangement—perhaps even a deliberate plan—emerges, one that encompasses the entire letter. Neither is this mere literary flourish or an attempt to achieve some subtle artistic or rhetorical effect. On the contrary, these repetitions, coupled with other important compositional markers and stylistic devices, help readers and listeners alike to follow the author’s argument and highlight the manner in which he has given the necessary definition and distinction to the discourse so that his main communicative objectives are accomplished.18 ‘Faith’ and ‘good works’ (a typical Pauline theme)—this was the essence of what Titus was to ‘teach’, ‘exhort’ and ‘reprove’ about (2:15) as he endeavored to establish a more dynamic, God-pleasing, good-producing and gospel-centered fellowship of all those ‘in faith’ (1:1–3; 3:14–15) on the island of Crete (1:5). There was in turn a dual focus of application for Paul’s personal appeal, namely internal—that is, to strengthen the Christian community (e.g. 1:9), and external—to present a positive witness to the outside world (e.g. 2:5). It would seem that this same twofold message is every bit as relevant and necessary today as the Church prepares to embark upon a new millennium in a world that is filled with just as many ‘empty talkers and deceivers’ (1:10), both within and without the professed membership.

With reference to this essential gospel (or ‘truth that leads to godliness’ ἀληθείας τῆς κατʼ εὐσέβειαν, 1:1 [NIV]), we observe once again how its periodic synoptic expression is also strategically situated to act as the grounds and motivation for all of Paul’s pastoral directives concerning church ‘discipline’ (i.e. ‘God-pleasing’ character training). An evangelical summary occurs at the beginning, middle and ending of the epistle: 1:1–4; 2:11–14; 3:3–7—each successive segment being somewhat more elaborately stated than its predecessor. All three key passages stress the life-changing significance of the ‘manifestation/appearing’ of our ‘God’ and ‘Savior’, ‘Jesus Christ’, to give us the ‘blessed hope’ of ‘[eternal] life’ (τὴν μακαρίαν ἐλπίδα, 2:13). In each of these focal pericopes an essential contrast is drawn with respect to time (e.g. ποτε ‘formerly’// ὅτε ‘when’ [= afterwards] // ἐλπίδα ‘hope’ [= future], 3:3–7) and spiritual condition (e.g. ἀσέβειαν ‘ungodliness’// εὐσεβῶς ‘in a godly manner’, 2:12).19 There is also a fundamental linkage established between faith/doctrine and life/works in that the former is viewed as inevitably producing the latter, whether for ‘good’—as stressed in these passages—or for ‘evil’, as demonstrated in the teachings and practices of the ‘deceptive’ ‘legalists’ (1:10–16; 3:9–11). This unique kernel of ‘saving’ ‘truth’ (1:1, 3) is thus

                                                            18 W. Doty emphasizes the purpose of Greco‐Roman epistles to convey the parousia or personal 

‘presence’ of the writer as a means of maintaining contact between two or more individuals (the so‐

called relational function of communication). ‘Unable to be present in person, his letters were a 

direct substitute, and were to be accorded weight equal to Paul’s physical presence’ (Letters in 

Primitive Christianity [Philadelphia: Fortress Press, 1973], p. 36). Accordingly, they had an ‘oral 

character’ (p. 44), including the presence of formulae, repetition, exclamations, etc., which rendered 

the text not only easier to comprehend but also stylistically more natural as it was being aurally 

processed. 

NIV New International Version 

19 The latter contrast is implicit in the first pericope, e.g. in the term ἐκλεκτῶν ‘of the elect’ (1:1). 

242  

built into the very textual fabric of the discourse,20 both to highlight the structural dimensions of this potent little epistle and also to accentuate its divine ‘proclamation [that] was entrusted (ἐπιστεύθην)’ to Paul (1:3). Indeed, this was the basis for the ‘common faith’ (κοινὴν πίστιν) which he shared not only with Titus (1:4), but with other beloved readers/hearers as well (1:1; 3:15, another text-spanning inclusio)—an ‘elect’ community that includes all like-minded receptors of the present day. 3. A Display of Some Text-Traversing Tectonic Patterns in Titus The ‘significantly’ (in a semiotic sense) symmetrical way in which this letter is demarcated into its major and minor constituent units is visually displayed by means of the following series of schematic charts. Viewed in conjunction with one another, they give at least a partial indication of the intricate manner whereby the text of Titus is ‘traversed’, that is, criss-crossed, by a number of interlocking structural-thematic and rhetorical-pragmatic arrangements which variously pertain to form, content and function. These operate in concurrence to render the composition more effective in terms of formal cohesion, topical coherence and ethical persuasion.

Figure 1 gives a general overview of the interlocking constituent segments of the complete discourse organization of Titus as it was selectively outlined with respect to 2:15 in the preceding section. We observe how these units, both large and small, are linked in paradigmatic as well as syntagmatic fashion to reiterate the essence of the apostolic instruction from similar, yet varied perspectives. As shown in the following diagram, the text ‘turns’, as it were, on its paraenetic midpoint at 2:15, a passage that incorporates a crucial thematic dimension by virtue of the anaphoric demonstrative ταῦτα, which refers to the preceding creedal and hortatory synopsis of 2:11–14.

                                                            20 ‘Savior’ σωτήρ occurs more often in Titus (with reference to ‘God’ in general [1:3] and/or ‘Jesus 

Christ’ [1:4]) than in any other New Testament book, i.e. seven times; cf. also 3:5 and its central verb 

ἔσωσεν ‘he saved us’. 

 

The prismoothlcoupledsuccessialternati(imperacontrastconnectthat is, witness

Figu

ncipal consly integrate

d with subtlive composion based

ative/appeal)tive and ction one main the perio’ to the surr

ure 1. A For

stitutive eleed into onee formal, sesitional union topical ) and motivorrelative t

ay note the uodic explicitrounding so

rmal, Seman

ements of the another inemantic, ants. These uantithesis

vation (ratitransitional underlying mt and implicciety.

ntic and Fu

he main bon linear se

nd logical counobtrusive

as well asionale/purpo

features amissionary cit expressi

unctional Ov

ody of the lequential faonnective d

cohesive ls the paireose). Some are shown emphasis th

ions of conc

verview of T

letter (as inshion by m

devices on thlinkages feaed relations

of the main Figure

hat runs throcern for an

Titus

ndicated abmeans of rehe borders ature a proship of exhain aspects

2 below. roughout theappealing ‘

243 

ove) are epetition between

ogressive hortation of these In this

e epistle, ‘external

 

Figure 2. The Lineear Sequencce of Logicaal and Themmatic Links iin Titus

244 

245  

I will conclude this illustration of the multifaceted architectural patterning that characterizes the discourse organization of Titus with two samples of such purposeful construction that occur within a discrete constituent section of the epistle. The first (1:10–16, Figure 3) reflects the confused self-centered theology of the Cretan false teachers by means of a chiastic pattern of reversal, one that focuses upon a centrally placed and strongly worded admonition along with its double corrective aim. The cohesive links that integrate this segment into the larger framework of Titus have already been pointed out.

A 1:10–11 = characterization: the false ‘teaching’ of the ‘rebellious’ ‘vain talkers’ (ματαιολόγοι) B 1:12 = support: a proverbial and hyperbolic saying critical of such individuals (not necessarily all

Cretans!) with respect to their outward words/deeds C 1:13 = injunction (based on the preceding ‘true’ characterization): ‘rebuke them!’—(a) for the

sake of their individual ‘faith’ (πίστει) C’ 1:14 = injunction (implied)—(b) for the sake of the communal ‘truth’ (ἀλήθειαν)

B’ 1:15 = support: a religious maxim/commonplace (e.g. Mt. 7:17–19; 15:11, 17–20) critical of all those who are ‘faithless’ and inwardly ‘defiled’

A’ 1:16 = characterization: the ungodly ‘works’ of the ‘disobedient’ false ‘confessrs’ (ὁμολογοῦσιν)

Figure 3. An ‘Inverted’ Compositional Structure Describing Un-converted People

Another elaborate pattern appears in the climactic gospel proclamation of 3:4–7. This description of the ‘purified’ people of God stands in sharp contrast both to what ‘we’ either were or would be without Christ (3:3) and to all who produce divisions within the church, either by their teaching or their wicked lives (3:9–11). This trinitarian segment also acts as the motivating grounds for our godly (= ‘good’) living in the world (3:1–2 and 8). In this case two synonymous thematic panels reiterate in a roughly parallel, sequential fashion the key themes in our salvation (note the centrally positioned principal verb ἔσωσεν ‘he saved us’, which expresses the logical result of vv. 4–5 and the summary purpose of vv. 6–7). This bipartite construction, which is marked by a significant chiastic shift (perhaps to create a bounding temporal framework) at the middle boundary, is shown in literal translation in Figure 4 below:21

                                                            21 The common claim that this pericope is derived from some earlier ‘credal statement’ or 

‘confession’ (e.g. Puskas, Letters of Paul, p. 179) is difficult to prove in the absence of any substantial 

prototypes. Whatever the case, Paul has skillfully (re)worked the material, whether traditional or 

original with him, to blend in with the letter’s overall structure, theme and purpose. 

 

The salimagnifithe priothese twapostle’was to m

This brstructurepistle iaction reliabili‘encour

Sevediscours(cf. 1:1athe divpersona

               22 P.H. To

23 It is im

type of m

F

ient correspicently craftor doctrinal wo similarly’s practical minister ‘wi

4. Conclus

rief investigre-functionais organizedwould reflity—of his raging’ and eral synoptse to be a da/3b, 4b)—winely comm

al letter.23 T

                      owner, 1–2 T

mportant to r

moral and re

igure 4. The

pondences bted reminde(+ ethical)

y placed secadvice and

ith all autho

sion: On the

gation has al nucleus od and givenect badly gospel me

‘correctingtic, synchro

dynamic funwith Paul—missioned i

The elaborat

                       Timothy and 

ecognize tha

eligious beha

e Parallelis

between thier of what ‘summary s

ctions functd admonitioority’ (2:15)

e Dispositio

shown thaof Titus andn shape. Indon his pa

essage and ’ the congreonic displaynctional ‘org

—with Titus—instruction te interweav

Titus (Down

at Titus is no

vior; nor is it

sm of Salvifi

is ‘single, d‘sound teacstatement otion in tand

on to Titus a).

on of Discou

at the key d the ‘keysdeed, no on

astoral ‘auththe associ

egations on ys of the cganism’ tha—with the that is set

vings of the

ners Grove: I

t just a para

t one that is 

fic Content i

densely paching’ was (

of 2:11–14 hdem as the as well as t

urse and Di

passage in stone’ accorne was to ‘hority’—as iated apostoCrete.

complete teat links the TCretan Chrt forth in te hierarchica

nterVarsity P

enetic epistl

concerned p

n Titus 3:4–

ked sentenc(and is) all have alreadymotivating the congreg

scipline in t

2:15 serverding to whdisregard’ well as t

olic impera

ext thus revTriune God—istians, boththis compaally arrange

Press, 1994),

e, intended t

primarily wit

–7

ce of theoloabout (cf. 2

dy been notebasis for a

gations to w

the Church

es as the shich the resTitus becauthe credibiative of ‘te

veal this epd—with Jesuh good and act, but coned structura

, p. 254. 

to urge a cer

th promoting

246 

ogy’,22 a 2:1), and ed. Thus all of the whom he

semantic st of the use such lity and

eaching’,

pistolary us Christ bad, via

nvincing al panels,

rtain 

g 

247  

which comprise several distinct compositional schemata, complement one another and act as an ‘isomorphic’ text-tectonic equivalent of the work’s central theme.24 Together they set forth the integral and indispensable interconnection involving the ‘discipline’ of faith (teaching) and life (practice) as the foundation of the theological ‘truth’ (1:1) about ‘salvation’ (e.g. 1:2; 2:14; 3:4) in relation to believers (e.g. 1:5) and outsiders as well (e.g. 2:5, 8, 10; 3:1–2). During the course of this epistle and through the eyes of Titus,25 we are given a ‘slice-of-life’ picture of a developing, and apparently rather undisciplined (1:5), young Christian church—that is, how, when, why and for what purpose it was formed according to the eternal purpose (1:2) and grace of a merciful Savior-God (1:3–4; 2:11, 13; 3:4–7). Though obviously dated, this concise ecclesiastical portrait is just as instructive and relevant to the current worldwide fellowship of ‘God’s elect’ (1:1).

The rhetorical-structural method of discourse analysis that is applied in the present investigation focuses upon the organizational function of several types of unit-delineating repetition, especially inclusio, epiphora, anaphora and anadiplosis. The prominent device of lexical and conceptual recursion is supported by selective concatenations of such familiar literary features as nominal (topic) advancement, conjunction and asyndeton, emphatic pronominal insertion, epistolary conventions, and the alteration of subject, tense, mood, tone, addressee, etc. This text-governed approach has several pertinent applications with respect to current biblical studies.

In the first place, it is one way of providing analysts with a reliable overview of how a particular literary work is put together as a compositional unity that consists of a hierarchy of subsections that are related to one another both syntagmatically and paradigmatically within the totality. A clearer perception of the linear and/or concentric structural framework of the complete text helps in turn to give a better idea of the central message that original authors wanted to get across to their readers or hearers. This includes the principal communicative purpose(s) which are conveyed by and through the discourse with respect to its stated and/or supposed initial contextual (including the historical) setting, or ‘rhetorical situation’.

Additional hermeneutical insights would apply to the interpretation of problematic areas within individual verses and discourse segments, for the whole and its parts are assumed to illuminate one another.26 A fuller understanding of the macrostructure of a biblical text also                                                                                                                                                                                           practical organizational and administrative affairs. Such a misleading impression is often given by the 

summary thematic statements and topical outlines that are provided for this text, for example, that 

of Puskas: ‘the thesis of Titus is concerned with directives on appointing worthy leaders to withstand 

opposition’ (Letters of Paul, p. 184). The important didactic‐mnemonic‐structural function of the 

three crucial gospel‐oriented synopses should not be overlooked when assessing either the message 

or the value of the letter as a whole. 

24 On the notion of an ‘isomorphic equivalent’, or structural metaphor, between different semiotic 

sign systems (one of these being literary discourse), see J. de Waard and E.A. Nida, From One 

Language to Another: Functional Equivalence in Bible Translating (Nashville: Thomas Nelson, 1986), 

pp. 63–68. 

25 Cf. Reed, ‘Discourse Features’, p. 248. 

26 One important exegetical application concerns the referential range and direction of certain key 

demonstratives. I have, for example, interpreted on the basis of this analysis that the initial ταῦτα of 

3:8c is anaphoric with reference to the total content of 3:3–7. It also begins a new, contrastive 

‘paragraph’ unit (cf. 2:15) and stands ‘balanced’ as it were by anadiplosis with the preceding τούτων 

in 3:8a. 

248  

enables one to make a more accurate assessment of its stylistic quality on the microstructural level of composition. In this regard, it was shown that Titus may not be as ‘drab’, ‘boring’ and ‘pedestrian’ as it is often attributed to be in the scholarly literature. And finally, the discourse-oriented methodology applied here is not limited to epistolary writing, but may be profitably utilized in greater or lesser detail and scope to analyze any type of text in the Bible.27

We have no indication in ‘Titus’ itself, or elsewhere in the epistles, or even in the book of Acts (sometimes linked to the Pastorals via ‘Luke’)28 whether Paul’s indirect admonition warning people not to ‘disregard’ the primary addressee (2:15c) was heeded. Neither do we know whether his ‘true son in their common faith’ (1:4) was actually able to utilize the apostle’s rhetoric to fully convict and convince their common opponents (e.g. 1:9–11; 2:1, 15a–b) or to accomplish his assigned practical mission on Crete successfully (e.g. 1:5; 3:1–2, 14). However, it is certain that a greater measure of attention devoted today to a disciplined study of the style and structure of literary discourse will prevent interpreters from making the serious mistake of ‘disregarding’ this valuable little epistle that bears Titus’s name.

                                                            27 For a recent application of this method of text study to the closest functional equivalent of the 

epistles in the Old Testament, see E.R. Wendland, The Discourse Analysis of Hebrew Prophetic 

Literature: Determining the Larger Textual Units of Hosea and Joel (Lewiston, NY: Mellen Biblical 

Press, 1995). 

28 See, e.g., Fee, 1 and 2 Timothy, Titus, p. 26. 

249  

Part IV DISCOURSE ANALYSIS AND THE GENERAL EPISTLES

HEBREWS 6:4–6: FROM A SEMIOTIC DISCOURSE PERSPECTIVE Andries H. Snyman

1. Introduction

New Testament scholarship (and especially hermeneutics) is in a process of methodological definition and redefinition. A common aspect of all the discussions on methodology is that they fall under one of the three poles in communication theory, namely the sender (author), the text and the receptor (the reader). The focus in this essay will be on the text, and the approach adopted a specific type of discourse analysis developed within the New Testament Society of South Africa.

The first part of the essay will deal with the setting or background from which this approach has emerged, followed by a description of the method itself. In the second part the approach will be applied to the hotly debated Heb. 6:4–6, which seems to provide a basis for denying the possibility of repentance for Christians who apostatize.

2. Methodological Considerations

a. The Setting of Discourse Analysis All disciplines interested in communication regard discourse analysis as part of their field: sociology, psychology, semiotics, speech act theory, etc. As far as New Testament scholarship is concerned, it developed within a structural approach to linguistics. Text-linguistic discourse analysis, as it is practised today, reached its height in the seventies and early eighties and coincides with an important shift in the field of biblical studies: the movement away from a historical and towards a text-immanent approach to the New Testament.

At the heart of this shift lie two different philosophies. The historical approach was based on positivism, according to which the understanding of a text was sought in a genetic-causal process of origin and development. In historical exegesis the authors of the New Testament were seen as exponents of their time and the situations in which they lived; therefore their writings were the result of a process of development, which must be studied historically. The literary work was seen as a historical document of a specific time, which needed to be described in terms of causality. The real object of study was not the literary work as such, but its origin, its genesis.1

In the seventies this historical approach was challenged by various text-immanent approaches, with their focus on the text as a literary work. The underlying philosophy of these approaches is phenomenology, where the text is seen as a phenomenon that must be understood and explained in terms of itself. Not the history of the text but the text itself, not the situation behind the text but the text as an autonomous object of study, not the intention of the author but the intention of the text comes to the fore.2 The result of this shift was that

                                                            1 C.J. Den Heyer, Exegetische methoden in discussie: Een analyse van Markus 10:46–13:37 (Kampen: 

Kok, 1979), pp. 89–90. 

2 W.S. Vorster, ‘De struktuuranalyse’, in A.F.G. Klijn (ed.), Inleiding tot de studie van het Niewe 

Testament (Kampen: Kok, 1982), pp. 127–52, esp. p. 128. 

250  

various text-immanent approaches came into existence among linguists and literators. They may be classified in three groups, each with its own subdivisions: French structuralism, literary criticism and discourse analysis.3

The last approach has received enthusiastic attention and support within South African New Testament studies. Virtually all scholars in this country who study the Greek text use it in their research. Discourse analysis is the one field where South Africa has made a significant contribution to New Testament scholarship in general. b. Discourse Analysis as Method The type of analysis referred to in the above section was originally called semantic discourse analysis.4 The shift from word to sentence and from sentence to paragraph in modern text linguistics paved the way for this method of analysis. The basic premise is that meaningful relations not only exist between the words in a sentence, but also between larger parts of a text such as sentences, groups of sentences (clusters), pericopes and chapters. It is therefore important to grasp these relations in order to follow the trend of the argument and to understand the meaning of the text. As the method developed, it became clear that a semiotic approach to discourse analysis was necessary, since discourse involves more than words and sentences.5 It involves the recognition of multiple signs, which are interrelated and allow for more than one level of meaning. These signs function as multiple layers, which may be grouped into three significant macro levels: the declarative, structural and intentional. Each communicates certain information, but together they convey the real purpose and sense of what a text has to say.6

The declarative level entails the division of a text into cola. A colon is a syntactic unit, which has a central matrix consisting of a nominal element (subject) and a verbal element (predicate), each having the possibility of extended features. As long as all these features can                                                             3 For a discussion see Vorster, ‘De struktuuranalyse’, pp. 127–52; A.H. Snyman, ‘A Semantic Discourse 

Analysis of the Letter to Philemon’, in P.J. Hartin and J.H. Petzer (eds.), Text and Interpretation 

(Leiden: E.J. Brill, 1991), pp. 83–99. 

4 In 1973 J.P. Louw published his ‘Discourse Analysis and the Greek New Testament’, BibTrans 24 

(1973), pp. 108–18. Since then numerous publications have emerged: H.J.B. Combrink, Structural 

Analysis of Acts 6:8–8:3 (Cape Town: DRC Publishers, 1979); A.B. du Toit, ‘Die praktyk van eksegese in 

die lig van nuwe wetenskaplike ontwikkeling’, in idem, Die nuwe‐testamentiese wetenskap vandag 

(Pretoria: University of Pretoria, 1980), pp. 119–36; idem, ‘Strukturale eksegese en die Suid‐

Afrikaanse redevoeringsanalise’, Skrif en Kerk 2 (1981), pp. 3–14; J.P. Louw, Semantics of New 

Testament Greek (Philadelphia: Fortress Press, 1982); N.S.L. Fryer, Discourse Analysis and Exegesis 

(Kwadlangezwa: University of Zululand, 1984), etc. Also editions of Neotestamentica, the official 

journal of the New Testament Society of South Africa: Volume 8 (1974), 11 (1977), 13 (1979) and 16 

(1982). The method has been successfully applied to argumentative material like the letters of Paul 

and long discourses in the Gospels. The best examples—among many others—are the two volumes 

by J.P. Louw on Romans, A Semantic Discourse Analysis of Romans (Pretoria: University of Pretoria, 

1979), and the dissertation by H.C. van Zyl, ‘Matteus 18:15–20: ‘n diachroniese en sinchroniese 

ondersoek met besondere verwysing na kerklike dissipline’ (University of Pretoria, 1987). 

5 J.P. Louw, ‘Macrolevels of Meaning in Lk 7:36–50’, in P.J. Hartin and J.H. Petzer (eds.), A South 

African Perspective on the New Testament (Leiden: E.J. Brill, 1986), p. 129. 

6 Louw, ‘Macrolevels’, p. 131. 

251  

be grouped under one N plus V, it forms one colon.7 A colon constitutes the smallest semantic unit. The analysis of a text’s syntactic features is important, since they constitute ways in which basic relationships between fundamental units are clearly marked by the text itself. The declarative is the level of the mere statement, where the bare facts are lexically and syntactically predicated.

Another level consists of the grouping of cola into clusters or pericopes, mainly on semantic considerations. In doing this, logical and stylistic markers are also taken into account, because they contribute in various ways to the demarcation of a pericope and the formulation of its theme.8 Louw calls this the structural level, since compositional features suggest particular groupings or clusters of information which, along with their mutual interrelationship, signify the focus of a discourse.9

A third level of meaning is the intentional. It refers to the purpose of the discourse, ‘that is, what the author, or perhaps rather what the text seems to convey within a particular frame of reference—in short, what is the message’.10 On this level cognizance is taken of all possible signs (linguistic, paralinguistic and extralinguistic) that may be relevant to understanding and interpreting a text. This intentional level is very prominent in the works of Eugene A. Nida11 and by incorporating it into the method, the scope and applicability of the latter are extended considerably.

The type of discourse analysis described above has, of course, its limitations. As is the case with many types of discourse analysis up to the present, it has not yet developed a comprehensive theory of its own. Nor can it be presented as an exegetical method complete in itself. It is liable to the subjectivity of the reader, no matter how objectively he has been led by instructions from the text. It only explains how the reader understands the text.12 On the other hand, the analyst must guard against philosophical trends like structuralism and phenomenology, according to which the meaning of a text lies completely within the text itself.13 Despite these limitations, however, this type of discourse analysis has proved to be a viable method in demarcating pericopes, in describing the structure or layout of a text and in following the trend of the argument. It also provides a sound framework for discussing the various interpretations of a text.

                                                            7 For more information see H.C. du Toit, ‘What Is a Colon?’, Neot 11 (1977), pp. 1–10. 

8 An indispensable aid in this regard is the publication of E.A. Nida, J.P. Louw, A.H. Snyman and J.V. 

Cronjé, Style and Discourse, with Special Reference to the Text of the Greek New Testament (Cape 

Town: Bible Society of South Africa, 1983). 

9 Louw, ‘Macrolevels’, p. 131. 

10 Louw, ‘Macrolevels’, p. 131. 

11 See for example his Signs, Sense, Translation (Cape Town: Bible Society of South Africa, 1984), pp. 

3–5, as well as the above‐mentioned Style and Discourse, pp. 145–48. 

12 F.E. Deist, ‘Ope vrae aan die diskoersanalise’, Nederduitse‐Gereformeerde Teologiese Tydskrif 19 

(1978), pp. 260–64. 

13 G.J.C. Jordaan, ‘’n kritiese evaluering van die Suid‐Afrikaanse diskoersanalise’, Koers 51 (1986), p. 

415. 

252  

3. An Application: Hebrews 6:4–6

a. A Crux Interpretum Hebrews 6:4–6 has been selected for illustration because it has presented a crux interpretum for exegetes and theologians since early times. A literal translation of these verses reads:

For it is impossible for those who were once enlightened, and have tasted of the heavenly gift, and were made partakers of the Holy Ghost, and have tasted the good word of God, and the powers of the world to come, if they shall fall away, to renew them again unto repentance; seeing they crucify to themselves the Son of God afresh, and put him to an open shame (AV). These verses seem to contradict many others, which declare that ‘God will keep his

promise: he will not allow you to be tested beyond your power to remain firm; at the time you are put to the test, he will give you the strength to endure it, and so provide you with a way out’ (1 Cor. 10:13); or Jn 10:28: ‘I will give them eternal life, and they shall never die. No one can snatch them away from me’; or the promise that God, who has begun a good work in us, will perform it until the very end (1 Cor. 1:8; Phil. 1:6).

The question is how Heb. 6:4–6 must be understood in terms of all these promises and assurances in the rest of the New Testament. What must be done with this text, which seems to contradict numerous promises and denies repentance for those Christians who have ‘fallen away’? In an attempt to answer this question I first want to contextualize 5:11–6:20 within the argument of the letter, then divide the relevant passage (6:1–8) into cola and finally discuss the proposals of Michel14 and Sabourin15 in the light of this information. A semiotic discourse analysis will conclude this part of the essay. b. The Argument of the Letter According to Van Zyl, Hebrews can be divided into two phases.16 The first is the overall argument, in which the author wants to encourage his readers to persist in their faith by focusing on the elevated position of Christ as the only and eternal High Priest. The High Priest theme is first mentioned in 4:14–5:10, and Christ is presented as the perfect one (7:28), the one to be followed (2:10; 12:23). This line of argument, however, is interrupted by 5:11–6:20, because the author first wants to attend to the readers’ reluctance to grow spiritually. This reluctance is a stumbling block in his teaching about the High Priesthood of Christ.

The interruption in 5:11–6:20 constitutes the second phase in the argument. The author warns his readers against their inability to consume the solid food of Christian doctrine; they are like small children who still have to drink milk (5:11–14). Nevertheless, the author wants to go forward to mature teaching, leaving behind the first lessons of the Christian message (6:1–3).

In pursuing this aim there may be a serious obstacle that prevents him from achieving this purpose: the fact that his readers could abandon their faith to such an extent that a second repentance becomes impossible—in spite of their former participation in Christ’s salvation (6:4–6). With an illustration from nature the author continues to explain the impossibility of a                                                             AV Authorized Version 

14 O. Michel, Der Brief an die Hebräer (Göttingen: Vandenhoeck & Ruprecht, 1975). 

15 L. Sabourin, ‘Crucifying Afresh for One’s Repentance (Hebr. 6:4–6)’, BTB 6 (1976), pp. 264–71. 

16 H.C. van Zyl, ‘God se genade verliesbaar?—Nog eens Hebreërs 6:4–6’, Nederduitse Gereformeerde 

Teologiese Tydskrif 31 (1990), pp. 336–77. 

253  

second repentance from apostasy (6:7–8). He is, however, convinced that they have not deteriorated to such an extent; therefore, he urges them to follow in the footsteps of their predecessors and to persist in their faith (6:9–12).

In 6:13–20 the author describes their predecessors’ faith and perseverance, as well as the basis thereof, especially the fact that Christ acts as our eternal High Priest. And with this theme the author returns to the basic argument, the High Priesthood of Christ.

Hebrews 6:4–6 forms part of the interruption in the main argument, namely 5:11–6:20. It must be understood within this context. The exposition of 5:11–6:20 above can be summarized as follows:

Theme: Go forward to perfection

A 5:11–14 Babes in the faith

B (i)

6:1–3 Aim: perfection

(ii)

6:4–6 Obstacles

(iii)

6:7–8 Illustration from nature

C

6:9–12 Wish for the readers: diligence, faith, perseverance

D 6:13–20 Examples of perseverance c. A Colon Analysis of 6:1–8 Since Heb. 6:4–6 forms part of a distinct unit (B above), I will restrict myself to an analysis of this unit, followed by a literal English translation (see opposite). This will enable us to follow the proposals of Michel and Sabourin. These two scholars were selected from many others17 because Michel wrote a standard exegetical commentary on Hebrews, while Sabourin approached the problem from a point of view closely related to the type of discourse analysis proposed in this article. Just to repeat the main problem in this passage: It seems as if 4 above, the impossibility that those who abandon their faith can be brought back to repent again, is irreconcilable with the central message of the gospel. If this interpretation of 6:4–6 is correct, it creates a serious problem for exegetes and theologians, in fact for all believers, since it would mean that God’s grace can be irretrievably lost. d. Explanations Offered by Michel and Sabourin The impossibility of a second baptism. From the fifth century to the Middle Ages the church interpreted ἀδύνατον as referring to a second baptism. Those who had fallen away were not allowed to be baptized again, although their sins could still be forgiven by penance.18 What is impossible is not to repent again, but to be baptized for a second time.

The basis for this interpretation is the meaning of φωτισθέντας in colon 4.1. Justin understood it as referring to baptism—a strong possibility, since the adverb ἅπαξ qualifies it as a specific, unique event. Furthermore: ἀνακαινίζειν is taken as referring back to φωτίζω

                                                            17 Van Zyl, ‘God’, pp. 339–43. 

18 Michel, Der Brief, p. 247. 

 

and mebaptizedby a sec

The ἀνακαινmeaning

(1) ‘t

t(2) ‘t

m The sechowevewhich isto bring

Anoφωτίζωnot only(or its im

               19 Miche

20 J.P. Lo

(New Yo

21 Louw a

eaning ‘to bd once (τοὺcond baptism

main probνίζω. Accogs:

o cause lighto shine upoo cause sommake know

cond meaniner, that the s described

g back.21 other probleω and ἀνακαy to colon 4mpossibility

                      l, Der Brief, p

uw and E.A. 

ork: United B

and Nida, Lex

be renewedὺς ἅπαξ φωm (πάλιν ἀν

blem with ording to th

ht to shine uon; mething to

wn, to make

ng is applicaterm refersas ‘to caus

em with thαινίζω. Acc4.1 (τοὺς ἅπy) is not onl

                       p. 241. 

Nida, Greek‐

ible Societie

xicon, p. 157

d by baptismωτισθέντας) νακαινίζειν

this argumhe Greek-En

upon some o

be fully kplain, to rev

able in this to baptisme a change

his argumencording to παξ φωτισθly meant for

‐English Lexi

s, 1988), pp.

7. 

m’.19 The and fallen ν).

ment centrenglish Lexic

object, in th

known by rveal, to brin

context. Thm. The same

to a previou

nt is the dimy analysiθέντας), butr those who

icon of the N

. 173, 338. 

argument iaway may

s around tcon of Lou

he sense of i

evealing clng to the lig

here is no eve applies tous, preferab

irect link dis above, πάt to cola 4.1o have been

New Testame

s then: thonot be rene

he meaninguw and Nid

illuminating

early and ight, to disclo

vidence in to the meanible state’—t

drawn betwάλιν ἀνακα1–4.5 as a wenlightened

ent Based on 

ose who haewed to rep

gs of φωτda φωτίζω

g it’—to illu

in some deose.20

the New Teing of ἀνακto renew, to

ween the twαινίζειν refwhole. The d, but also f

Semantic Do

254 

ave been pentance

τίζω and has two

uminate,

tail’—to

estament, καινίζω,

o restore,

wo terms fers back

renewal for those

omains 

255  

who have tasted the heavenly gift, who were made partakers of the Holy Ghost, etc., and have fallen away.

The proposal of Michel and others does not solve our problem, since it only deals with the technical point whether the Church has the right to rebaptize those who have apostatized. In this way the problem is not solved, but explained away. The point of 6:4–6 is that it is impossible for believers who have abandoned their faith to be renewed unto repentance. And this is the meaning that must be understood within the context of 5:11–6:20. An alternative translation. In an interesting article Sabourin offers an alternative translation by regrouping the words in 6:4–6. His analysis of these verses corresponds to my proposal above, except for the division between cola 4.5–7. According to Sabourin εἰς μετάνοιαν does not qualify πάλιν ἀνακαινίζειν (colon 4.5), but ἀνασταυροῦντας ἑαυτοῖς τὸν υἱὸν τοῦ θεοῦ in colon 4.6.

His analysis of cola 4.5–7 reads:

καὶ παραπεσόντας πάλιν ἀνακαινίζειν εἰς μετάνοιαν ἀνασταυροῦντας ἑαυτοῖς τὸν υἱὸν τοῦ θεοῦ και παραδειγματίζοντας22

This gives him the translation:

and having fallen, [it is impossible] to restore a second time, crucifying afresh for oneself the Son of God with a view to repentance, and mocking [him].23 This translation presupposes the following: (1) There are two groups of persons in colon 4 (vv. 4–6): those who are described by the

five aorist participles and who apostatized, and those who would make themselves new a second time.

(2) ἀνακαινίζειν is a transitive verb, meaning ‘to renew’ or ‘to restore’. It must not be read

with εἰς μετάνοιαν; this will give to the verb the weak sense of ‘renovate spiritually’. ἀνακαινίζειν is strong enough to stand on its own, as is proved by the participles in the accusative.

(3) The focus is on the definitiveness of Christ’s saving work. His crucifixion cannot be

repeated, since his sacrifice was offered once and for all. If anyone renounces Christ’s salvation, he cannot count on another one to rescue him from his fall.

Although one appreciates the attention Sabourin pays to the Greek text and its syntax, the following can be brought in against his proposal:

                                                            22 Sabourin, ‘Crucifying’, p. 264. 

23 Sabourin, ‘Crucifying’, p. 265. 

256  

(1) It is grammatically incorrect to distinguish two groups of persons in colon 4. All the

participles in the accusative are direct objects of the infinitive ἀνακαινίζειν, and refer to the same persons. The only difference between the participles in 4:1–5 and 4:6–7 is that ἀνασταυροῦντας and παραδειγματίζοντας in 4:6–7 provide the reason for the impossibility of bringing those who have fallen away back to repentance.

(2) The grouping of εἰς μετάνοιαν with ἀνασταυροῦντας … instead of with ἀνακαινίζειν

is not supported by any edition of the Greek New Testament and complicates the translation considerably (see point 4 below).

(3) Focusing on the uniqueness and definitiveness of Christ’s saving work is in order,

except that it does not fit the trend of the argument. Van Zyl correctly points out that Sabourin removes these verses from the locus of perseverance to that of soteriology. And soteriology is not the point of the argument here.24

(4) Sabourin’s translation at the end of his article is highly problematic. It reads:

‘For it is impossible to crucify afresh the Son of God for the sake of one’s repentance, mocking him, so as to restore a second time those who have once been enlightened … and have apostatized.’ From a grammatical point of view, the only mood that can follow ἀδύνατον is an infinitive (ἀνακαινίζειν), not a participle (ἀνασταυροῦντας).

Sabourin based his article on one by P. Proulx and L. Alonso Schökel.25 Their attempt to solve the problem of 6:4–6 by offering an alternative translation, however, is not successful. Difficult passages like 6:4–6 must not be dogmatized or translated away, but understood within their context. This is one of the main principles of discourse analysis. e. A Semiotic Discourse Approach to Hebrews 6:4–6 Declarative level. The first level of reading a text, the declarative, deals with the facts of the communication. This has already been described in the above section on the argument of the letter, where we dealt with the second phase in the argument of the letter (5:11–6:20). Also important on this level are the lexical terms used and their meanings. These received attention in the discussion of Michel’s and Sabourin’s proposals (see above).

What it boils down to is that Heb. 6:4–6 must not be softened or put in the so-called correct dogmatic perspective. These verses must be understood with the relations and meanings described above, without an attempt to change the grouping of the words or explain their meaning by referring to other contexts in the New Testament. Hebrews 6:4–6 was used in this context for a specific purpose, and it is our task to determine that purpose. On the declarative level the meaning is clear—those who have tasted of the heavenly gift, etc., and then abandoned their faith cannot be brought back to repentance again. Structural level. A structural analysis reveals the focus of the discourse. The structure of 5:11–6:20 was summarized above. For our purpose now, the first two components are important:

                                                            24 Van Zyl, ‘God’, p. 341. 

25 P. Proulx and L. Alonso Schökel, ‘Heb 6:4–6, eis metanoian anastaurountas’, Bib 56 (1975), pp. 193–

209. 

257  

A Babes in the faith (5:11–14)

B (i) Go forward to perfection (6:1–3)

(ii) Obstacles (6:4–6)

(iii) Illustration from nature (6:7–8) A prepares the readers for the warning in B. B (i)–(ii) warns them against apostasy and B (iii) illustrates the disastrous results of not bearing fruit. B (i)–(ii) is thus the central or pivotal point of the passage.

Put in more detail: The trend of the argument is the concern of the author about the reluctance of the readers to grow in their faith. In spite of this, he urges them to a fuller understanding of the High Priesthood of Christ. But before doing so, the readers must be warned against the real danger of apostasy. Intentional level. In reading the letter to the Hebrews one must distinguish clearly between the expository or dogmatic, and the paraenetic or hortatory, sections. They are to be read differently. In Heb. 5:11–6:20 we do not have an exposition of the Christian faith, but a warning against abandoning it. The purpose of the discourse is not to build the readers’ knowledge of a specific topic, but to challenge them to right action by eliciting an emotional response from them.26 An important feature of paraenetic sections such as this one is the use of forceful expressions and style. One such expression is to be found in 6:4–6 (repentance impossible for apostates). Hebrews 6:4–6 must be understood within this context. It serves the purpose of the discourse, that is, what the text seems to convey within a particular frame of reference.27 It serves the message of 5:11–6:20.

Other features strengthening the impact of 6:4–6 are the following:

(1) ἀδύνατον is placed first in colon 4 to emphasize the impossibility of a second repentance. According to Longinus the first word in a sentence carries more weight than the rest.28

(2) The use of ἀδύνατον seems to be an instance of hyperbole, in view of the warnings against apostasy.29

(3) Michel argues that there is a ‘Steigerung’ from cola 4.1 to 4.4, with 4.4 carrying the most weight: ‘to taste the good word of God and the powers of the world to come’.30 The lengthening of commata is an impressive stylistic technique.31

                                                            26 G.H. Guthrie, The Structure of Hebrews: A Text‐Linguistic Analysis (Leiden: E.J. Brill, 1994), p. 139. 

27 Louw, ‘Macrolevels’, p. 131. 

28 Longinus, Peri Hupsoes 38.3. 

29 Louw and Nida, Lexicon, p. 669. 

30 Michel, Der Brief, p. 242. 

31 Cf. Demetrius, Peri Hermeneias 1.18. 

258  

(4) There is a hyperbaton with καλὸν … ῥῆμα in 4.4. Stacy regards this stylistic figure also as forceful.32

(5) Two metaphors strengthen the impact of this passage: φωτισθέντας (‘to be enlightened’) in 4.1 and γευσαμένους (‘to taste’ the good word of God and the powers of the world to come) in 4.4. Longinus, inter alia, refers to the powerful effect of metaphors.33

(6) In cola 4.1–5 we have a syntactic structure known as periodos. The reader must wait until the end of the sentence to know what is impossible. The function of a periodos is to hold the reader’s attention.34

(7) πάλιν ἀνακαινίζειν in 4.5 is tautological, but used for a purpose. It emphasizes what is impossible: to be renewed again.

(8) παραδειγματίζειν in colon 4.7 is a stronger term than δειγματίζω. Although Louw and Nida draw no distinction between them,35 Newman translates δειγματίζω as ‘disgrace, expose’,36 while παραδειγματίζω means ‘to expose to public ridicule’.37 A preposition attached to a verb may strenghten its meaning, especially in a fervant discourse.

(9) Cola 5 and 6 are linked to 4 by γάρ. It is an illustration from nature that one who perseveres will be blessed, but one who apostatizes will be punished. This is a form of amplificatio, known as comparatio. Amplificatio is used when an author wants to emphasize a point or stress the importance of the argument.38

Hebrews 6:4–6 thus forms part of a paraenetic section, characterized by forceful expressions and stylistic techniques in order to (1) underline the warning that the author wants to convey, and (2) exhort his readers to take the right action, which is to endure in their faith.

                                                            32 R.H. Stacy, Defamiliarization in Language and Literature (Syracuse, NY: Syracuse University Press, 

1977), pp. 60–61. 

33 In his Peri Hupsoes 32.1–8. See also J.V. Cronjé, ‘Demetrius on Forcible Style’, Acta Patristica et 

Byzantina 4 (1993), p. 40. 

34 For a discussion of the structure and function of a periodos, see Demetrius, Peri Hermeneias 1.10–

26. 

35 Louw and Nida, Lexicon, p. 311. 

36 B.M. Newman, A Concise Greek‐English Dictionary of the New Testament (London: United Bible 

Societies, 1971), p. 40. 

37 Newman, Dictionary, p. 132. 

38 Cf. J.V. Cronjé, ‘Αὔξησις as styltegniek in die tyd van die Nuwe Testament’, in L. Cilliers and A.H. 

Snyman (eds.), Varia Studia in honorem W.J. Richards (Acta Academica, 22; Bloemfontein: University 

of the Orange Free State, 1987), pp. 204–207. Cronjé refers to Quintilian, who distinguished four 

types of amplificatio: incrementum, comparatio, ratiocinatio and congeries. In all of them the author 

uses his knowledge of everyday life to strengthen his argument by relying on the fact that certain 

things are generally accepted, or that his readers have consensus on them, especially as far as their 

emotional experience thereof are concerned. 

259  

Besides the point and on a dogmatic level: Is it possible for a Christian who apostatizes to be brought back to repentance or not? Michel39 and Van Zyl40 both distinguish between the warning as such, and the final judgment, which is God’s. Christians must know that they can lose God’s grace. On the other hand: God speaks the final word. The Christian’s responsibility is to endure until the end, whatever comes his or her way. Evidence for this distinction is found in other parts of Hebrews, as well as in 6:3.41

4. Conclusion

Semiotic discourse analysis has proved to be a viable method in describing the layout of a text and in following the trend of the argument. It is also useful to describe the cohesion of a text and to determine its purpose, that is, the message that it intends to convey.

Hebrews 5:11–6:20 is a warning against apostasy. As such, the problem created by 6:4–6 must not be explained or translated away, but its function must be understood within its immediate context. It is a forceful expression which, together with other techniques, strengthens the impact of this paraenetic section.

                                                            39 Michel, Der Brief, pp. 246–47. 

40 Van Zyl, ‘God’, pp. 346–47. 

41 Van Zyl, ‘God’, p. 346. 

260  

FIRST JOHN: DISCOURSE ANALYSES AND INTERPRETATIONS*

Birger Olsson

In the early 1970s, Wolfgang Dressler published a short, introductory presentation of textlinguistics, listing at the beginning a number of important questions within the new discipline and concluding with the following words: ‘Textlinguistics approaches these problems with the presupposition that text constitutes the primary linguistic sign, the fundamental unit of language, i.e., it is not in sentences but in texts that people write or speak’ (Dressier 1972: 3).

The text is the basic linguistic unit. This assertion no longer wakens any great surprise among biblical scholars. It is commonly accepted, even if not always practised in individual analyses. Gradually, since the early 1970s, we have seen within interpretive studies a general shift away from a focus on the word and the phrase to a focus on the clause, and from the clause to the larger units, and finally to the whole document itself. Reflection on what actually constitutes a text has influenced biblical interpretation in our time perhaps more than any other consideration (Thiselton 1992: 55–79).

In Sweden, at Lund, René Kieffer has played a part in this worldwide change of perspective. His first contribution to the subject appeared the same year as Dressler’s book and was followed by several others (Kieffer 1972, 1975, 1982a; Kieffer and Olsson 1983; and later works). The inspiration for this movement came from France and the analyses had above all a textsemiotic foundation (Kieffer 1982b: 139–40, 145). A more textlinguistic orientation characterizes some works from Uppsala (Olsson 1985).

The background for what follows is mostly practical in nature. In writing a commentary on the Johannine letters, I have encountered from the outset questions about how to view the text as a whole, questions exemplified in this essay by reference to 1 John. More broadly applied, the subject has many names: the text’s structure, composition, literary form, arrangement or content; sometimes it can be combined with segmentation, thought sequence, argumentation, purpose, text-type, rhetorical genus, genre or message. There is a plethora of articles and monographs on this subject listed in the larger commentaries and surveys of research (see, e.g., Beutler 1988: 3779–80; Klauck 1991a: 59–87). Do we in fact understand how we read a text? Do we understand why we read a text? The subject of discourse analysis has more to do with such questions as these than with the traditional question: Do we understand what we are reading? And for this reason the answers are legion, and not least of all with 1 John.

Discourse analysis can be defined in several ways (Winther-Nielsen 1995: 79–96 and the literature cited there). I use the word in an elementary sense for the analysis of the text as a whole, the text being determined syntactially, semantically and pragmatically. How do such analyses of 1 John look? And why do we carry out such analyses? Several aims are discernible:

• to show how the text first came to be or how it was redacted • to show how the text coheres as a unity of some kind • to identify the reading instructions given in the text itself • to describe how the text functions • to grasp the author’s intention or purpose • to determine the text’s genre • to describe the text’s argument

                                                            * Translated from Swedish original, ‘Första Johannesbrevet: Diskursanalyser och tolkningar’, SEÅ 60 

(1995), pp. 141–63, by Richard J. Erickson, Fuller Theological Seminary, Seattle, Washington, USA. 

261  

• to show how the text reflects non-textual conditions • to summarize the text’s theme or fundamental thought • to demonstrate the relevance of the text for readers of later times • to draw up an interpretation of the text for others The list could be enlarged. But already, just from this it can be seen that structural

analyses often deviate from each other. This is certainly true with respect to 1 John. Various analyses seek to answer different questions. So far as I know, there is as yet no methodical analysis of the connections between various discourse analyses, proposed aims and interpretations of a text.

The structural descriptions of 1 John are, as I have said, distressingly many (see the surveys in, inter alia, Marshall 1978: 22–27; Miehle 1981: 28–31; Brown 1982: 116–29, 764; Klauck 1991a: 59–87). In the following pages I consider a few commentaries, particularly Brown (1982) and Klauck (1991b); an instructive example from the so-called Rome school, Malatesta (1978); a comprehensive analysis from South Africa, du Rand (1981); two discourse analyses from scholars affiliated with Wycliffe Bible Translators, Miehle (1981) and Longacre (1992); and several rhetorical descriptions, Vouga (1990); Watson (1989, 1993); and Morland (1992).

It should be clear that one of my chief interests is in detecting the reading instructions to be found in the text itself (Olsson 1985), which of course is not to say that these signals alone are sufficient to yield an interpretation of the text on which everyone can agree. The most exciting thing here is the interplay between the perspectives and questions which I as interpreter put to the text out of my own context, on the one hand, and on the other, the various kinds of reading instructions which the text itself provides (Olsson 1994). Discourse analysis is undeniably the highest level of interpretation for a particular text and is also the linguistic level that comes closest to the basic non-linguistic questions of sender, receptor, place, time, external circumstances, etc., questions most appropriate if we wish to pursue a historical understanding of a text (Porter 1992: 300). Thus, in spite of the troubling multiplicity of senses for ‘discourse analysis’, which some people consider a reason for avoiding any serious analyses of texts as wholes, discourse analysis is and will continue to be decisive for how we interpret many text-types (which is not yet to say all text-types!) Discourse analysis and interpretation are intimately connected and ought to be critically examined much more than has been the case till now.

1. Discourse Analysis in Commentaries: Brown 1982; Klauck 1991b

Anyone acquainted with the many descriptions exegetes have proposed for the structure of 1 John as a whole may well be tempted, with A. Jülicher’s words, to dismiss all attempts at its analysis: ‘The countless attempts to find a well thought-out arrangement in 1 John have the merit of cancelling each other out’ (cited in Klauck 1991a: 60–61). Several writers have surveyed these many attempts (Segovia 1982: 37–38; Brown 1982: 764; Klauck 1991a: 62–63), and there are even whole dissertations on the subject (Miehle 1982; Van Staden 1988). Some see the letter as a completely unstructured collection of independent units. After 1 Jn 2:27, for example, Bultmann (1973: 44) finds nothing but a string of smaller pieces, sketches or meditations: ‘Indeed’, he says, ‘one may be so bold as to inquire whether they are not reports from sessions of a seminar’. Others have produced the most ingenious arrangements, as, for example, Lohmeyer (1928), who builds his analysis on the number seven. Most commentators find themselves somewhere between these two extremes.

At one time, one read 1 John as a collection of aphorisms; later more thematic arrangements were preferred, sometimes with a trinitarian scheme. The shift between dogmatic/christological and ethical sections has characterized many proposals, as has the

262  

view that 1 John provides the reader with a series of tests for Christian life, a structual model that facilitates the transfer to the situation of the modern reader. Most often the beginning and the ending, 1:1–4 and 5:13–21, are seen as independent units, prologue and epilogue. The remainder is then divided into anywhere from two to seven parts. Most common is the analysis of 1:5–5:12 into three sections. The majority of scholars seem to be agreed that 4:1–6 and 2:18–27 form independent units, though with a lively discussion about where the boundary at 2:28 actually belongs. A similar accord prevails with respect to the limits of 5:1–12. Proposals vary widely, however, for the analysis of the more ethical sections 1:5–2:17; 2:28–3:24; and 4:7–21. And certain verses are perhaps best understood as transition passages: 2:27, 28, 29; 3:1, 18, 24; 5:4–5.

Klauck keeps to the middle of the road when he chooses a triple partition of 1:5–5:12. His reasons for dividing the passage into three parts are mostly text-external or content-motivated: an old and very common structural form; something very natural in the ordering of one’s thoughts; the appearance of groups of three in 1 John (1:6–10; 2:3–11; 2:12–14, 16; 5:6–8); balance and symmetry in the division of the text; a pattern for a step-by-step enlarging and deepening of the message of love. The result is as follows:

A. Prologue: Regarding the Word of Life (1:1–4) B. Body: Instruction in the Reality of Love (1:5–5:12)

I. The Community of God and the Confession of God (1:5–2:17) 1. Live in the Light (1:5–2:2) 2. Obey the Commandments (2:3–11) 3. The Certainty of Faith and Ethical Responsibility (2:12–17)

II. Facing the Demand of the Final Hour (2:18–3:24) 1. Confessing the Son as a Criterion (2:18–27) 2. Final Expectation and Sinlessness in the Children of God (2:28–

3:10) 3. Practising the Love Command (3:11–24)

III. Faith and Love Put to the Test (4:1–5:12) 1. Distinguishing the Spirits (4:1–6) 2. Love’s Song of Songs (4:7–21) 3. Witness for the Faith (5:1–12)

C. Epilogue: Eternal Life (5:13–21) The division of each of the three main parts of the body into yet another three parts may

be, according to Klauck, merely coincidental. Below these subdivisions, moreover, no overarching organizational principle can be found. Some commentators who make these same triple divisions interpret them in a trinitarian fashion: Father, Son and Holy Spirit, respectively; but on this point Klauck is more restrained.

This description of 1 John in its wholeness attempts to integrate as many textual observations as possible (Klauck 1991a: 66). The above-mentioned rationales are, however, mostly text-external. The analysis is carried out from above and has a dominating thematic character. References to how the first recipients can be thought to have read the text are nowhere to be found, that is, the discourse analysis is oriented to text-content alone. The main motivational reference seems to be the ‘love command’, thematized into three parts in a staged enlargement and deepening. The synopsis thus shows how the letter’s sender presumably wishes to arrange his message, or more likely, how the interpreter wants to summarize that message.

The fundamental perspective in Brown’s discourse analysis is not the love command but the relationship of 1 John to the Gospel of John and to the prevailing situation of conflict. Both the Gospel and the Revelation of John are clearly structured, and therefore 1 John ought

263  

to be as well. The letter is thus interpreted primarily from within the developing Johannine context, as a commentary on the Johannine tradition. It was written to protect the author’s followers within the Johannine community against further influence from the secessionist teachers. At a median level there are many similarities with Klauck, but not at a more overarching level. With the Gospel of John as a model, the body of the letter is divided into two parts, each with its own character. The result is 11 units (Brown 1982: 765), combined in the following manner:

1 PROLOGUE (1:1–4): Reflections on the Gospel of John

Prologue: ‘In the beginning was the Word …’ PART ONE (1:5–3:10): The obligation of walking in light in response to the gospel of God as light, a response that divides the secessionist Antichrists from the author’s Little Children.

2 1:5 ‘This is the gospel: God is light and in him there is no darkness at all’. 1:6–2:2 Three boasts and three opposite hypotheses, reflecting different understandings of the gospel.

3 2:3–11 Three claims of intimate knowledge of God, which are to be tested by the way one walks.

4 2:12–17 Admonitions to believers who have conquered the Evil One and so must resist the world.

5 2:18–27 Warning against the secessionist Antichrists who deny the Son and the Father.

6 2:28–3:10 In face of the coming encounter with Christ and God, the contrast between God’s children and the devil’s children. PART TWO (3:11–5:12) The obligation of loving in deeds in response to the gospel that we should love one another according to the example of Jesus as Christ come in the flesh.

7 3:11 ‘This is the gospel: We should love one another’. 3:12–24 Admonitions to the author’s Brothers and Little Children on the need to show love in deeds.

8 4:1–6 The Spirits of Truth and Deceit, governing respectively the author’s beloved adherents who belong to God and the secessionists who belong to the world.

9 4:7–5:4a The absolute necessity of loving one another in order to love God. 10 5:4b–12 Faith as the conqueror of the world and the believer’s relation to testimony. 11 CONCLUSION (5:13–21) A statement of the author’s purpose.

Besides reference to the structure of the Gospel of John, individual units are often

provided with a detailed argument for their delimitation and coherence. The reasons brought forth are syntactic and semantic: syntactic patterns, antitheses, triple groupings, chiasms, inclusio, other repetitions, changes in theme, and transitional verses. Overall, according to Brown, 1 John is a well-structured text.

It is possible that the editing of the Gospel of John and 1 John have something in common. The greatest influence can be observed in the prologue. But while the seam in the middle is very obvious in the Gospel, it is not so in the letter. The letter’s conclusion has a character different from that in the Gospel. Still, if 1 John is indeed redacted material, there is a certain demonstrable influence from the Gospel of John.

But a common redactional pattern does not need to determine completely the way in which the letter functioned as a communication. It is worth observing that Brown’s

264  

commentary contains very little consideration of the pragmatic. This is true to such an extent that sections which clearly refer to the communication situation (2:1, 7–8, 21, 26; 3:1–3; etc.) become mere parentheses in Brown’s analysis or are dropped from the discourse analysis. Thus the result becomes content-oriented, as was the case with Klauck’s analysis, but more defined than Klauck’s by the letter’s historical context as Brown sees it. The differences appear clearly in the terms that are used in the two synopses. Yet like Klauck’s, Brown’s description is actually a summary of how he himself interprets the letter.

2. Discourse Analyses in Rome: Malatesta 1978

A number of Catholic scholars have for some time been especially interested in the literary form of texts and have worked out a ‘stylistic’ method for arriving at a full description of a text as a whole (Olsson 1974: 6–7 and the literature cited there). Edward Malatesta has produced such an analysis for 1 John. His intention above all is to contribute to ‘a sharper awareness and better understanding of the peculiar style and rhythm of these letters’ (Malatesta 1973: 4). The analysis begins at the level of the clause and notes especially certain formal and explicit features in the text: thematic words, antitheses, parallellisms, contrasts, concentric patterns, inclusio, formal linking words, etc. From these characteristics the text is then grouped into various units on various levels. In his 1978 work, Malatesta provides a comprehensive analysis of 1 John on at least three levels.

The result, laid out in a typographically well-structured Greek text, cannot be reproduced here. The overall structure, however, is of particular methodological interest, and has been formulated as follows: PROLOGUE: Apostolic witness to life and communion (1:1–4)

I. FIRST EXPOSITION OF CRITERIA OF NEW COVENANT COMMUNION WITH GOD (1:5–2:28) Perspective: God is light (1:5) This communion considered in terms of light No explicit indication of the connection between love and faith

II. SECOND EXPOSITION OF CRITERIA OF NEW COVENANT COMMUNION WITH GOD (2:29–4:6) Perspective: God is just (2:29) This communion considered in terms of sonship Mention of the connection between faith and love (3:23)

III. THIRD EXPOSITION OF CRITERIA OF NEW COVENANT COMMUNION WITH GOD (4:7–5:13) Perpective: God is love (4:8, 16) This communion considered in terms of love Development of the relationship between love and faith EPILOGUE: Prayer, summary of letter (5:14–21)

As is clear even from this basic structure, Malatesta wants to emphasize the progression in

the text, the movement and deepening of the train of thought. The author’s thought is unified, but it must be expressed in stages. One stage leads to another. Section by section, one and the same subject is treated and developed (Malatesta 1978: 40). If we drop to the next level,

265  

which is divided into elements A, B and C (except that the third movement has only elements B and C), we observe again this staged development:

A Walking in Light and Freedom from Sin (1:5–2:2) A′ Doing Right and Avoiding Sin (2:29–3:10) B Knowledge of Communion with God and Observance of the New Commandment of

Love (2:3–11) B′ Love: Its Nature, Exigencies and Signs (3:11–24) B″ Love comes from God and is Rooted in Faith (4:7–21) C Believers Contrasted with the World and with Antichrists (2:12–28) C Discernment of Spirits (4:1–6) C″ Faith in the Son of God is the Root of Love (5:1–13)

Malatesta’s distinct conception of the letter taken as an entirety depends on a combination

of ‘the literary structure’ and ‘the theological structure’. Each of these is handled separately in his introduction (Malatesta 1978: 37–41, 77–79). In a striking way, each is analyzed in turn before being unified in a well-defined pattern. The rationale for this is based on the literary features in the text, the author’s personality or mentality, his particular message, and the letter’s aim and genre.

Even if Malatesta emphasizes the text’s literary features, the most decisive factor for his structural analysis lies in his own conception of the author and his experience. ‘Perhaps no other single writing of the Bible places so forcefully and so explicitly its entire contents under the sign of an experience the readers are invited both to share and to discern’ (1978: 39). ‘The author approaches his expositions in the light of his contemplation of the mystery of God’s love in Jesus Christ’ (1978: 40). His theology is an experience he has deeply felt, an experience of fellowship with God—Father, Son and Spirit—in the new covenant, an experience that he wishes to share with his readers. Thus he offers them criteria for distinguishing a genuine fellowship with God from the deviant behavior and the un-Christian attitudes of some who falsely claim to be united with God. ‘He is writing from heart to heart. This appeal to an awareness of the Christian mystery in faith and love provides the specific, conscious framework of all his thought’ (1978: 78).

Malatesta’s literary analyses of the text and his discussion of the deepening of covenantal thought in 1 John are of lasting value, but his way of bringing them together in a description of the text as a whole is unfortunate. It is clear that from a pragmatic perspective such a structure functions not at all (as for example in describing the function of 2:29; 5:13, 18–21). Literary issues are restricted for the most part to surface expressions (syntactic patterns, formal features of various kinds) and only to that which is explicitly expressed in the text. There is good reason to begin with the surface text, but numerous semantic and pragmatic qualities are expressed only implicitly. Malatesta’s results, as now presented, are strongly thematic and wholly determined by the central theological idea in 1 John as he sees it.

3. Discourse Analysis in South Africa: Du Rand 1981

Since the beginning of the 1970s, researchers in South Africa have worked with discourse analyses of individual documents or pericopes from the New Testament (Louw 1973, 1992; Combrink 1979). Du Rand published ‘a discourse analysis of 1 John’ in 1981. According to him, all earlier analyses of the letter are merely divisions of the text along certain aesthetic or dogmatic lines and are not a structural analyses at all. He calls for a more coherent and thoroughgoing analysis which begins with the individual statements in the text. One cannot

266  

stop with stylistic observations but ‘must attempt—in addition—to penetrate the extreme depth of the total structure to determine the coherence of the whole document. The coherence of the various semantic motifs, according to the syntax of structure, must lead to the “depth”, namely the coherent meaning (of the writer) of the whole document’ (1981: 2).

The orientation of du Rand’s method is strongly semantic. ‘The aim must be to present a semantic methodical exposition of the sequence of thoughts of 1 John’ (1981: 2–3) To reach this goal, the Greek text is first divided into cola, that is, syntactic units which as a rule contain a nominal and a verbal component (roughly, subject and predicate) each of which can be expanded with various kinds of qualifications (Louw 1982: 91–158). 1 John contains 196 such units. The hierarchical relations among them (co-ordinate or subordinate) are determined, producing groups of cola, which in turn form ‘pericopes’. These pericopes combine to form sections, which then form the larger ‘divisions’. Thus there result six levels in the text: colon, groupings of cola, ‘pericopes’, sections, divisions and the whole text.

It is primarily the content of the various units, though also certain parallellisms and other patterns, that determines the relations between units. The frequency and distribution of certain words reveal the theme of a pericope, section or division. The result, on the level of divisions, looks like this:

1. Introduction 1:1–4 Witness about Jesus Christ, the life which aims at fellowship

2. Fellowship 1:5–2:17 with God (who is Light)—walk in the light

3. Filiation 2:18–4:6 of God (identity)—identify yourself (through conduct) as children of God

4. Love 4:7–5:5 of God (who is love)—walk in the love

5. Conclusion and Résumé 5:6–21 Witness about Jesus Christ, the life which aims at certaintyof possession of life

These five divisions can be summarized in one theme which appears as the leading motif

in the whole letter and shows up especially in Divisions 1 and 5 (in 1 Jn 1:3 and 5:13 with the words ‘have fellowship with’ and ‘know’). At stake here is the receptors’ identity, which is built upon their certainty of possessing eternal life. This certainty is the subjective aspect, the fellowship the objective aspect, of the same basic motif. And in 1 John this certainty has a single source: the incarnation of Jesus Christ. It can be experienced in the form of fellowship (Division 2), sonship (Division 3), and love (Division 4). Du Rand (1981: 29–30) formulates the theme for 1 John as follows: ‘We know that we possess eternal life. We must show this in our way of life. The source for this certainty lies in the incarnation of Jesus Christ. The forms of certainty are fellowship, filiation, and love.’

Even on the levels of sections and pericopes, each unit is presented through its theme, which is almost always determined by the frequency of the more significant words in that unit. ‘A total survey of 1 John’s structure’ accounts for the three highest levels: divisions, sections and pericopes (du Rand 1981: 30–34). The three central divisions have sections that are quite in line with the character of the respective divisions:

2:1 1:5–2:2 Foundation of fellowship

2:2 2:3–11 Criterion for fellowship

2:3 2:12–17 Test for fellowship

3:1 2:18–27 Identification test for filiation

3:2 2:28–3:3 Foundation of filiation

 

3:3

3:4

3:5

3:6

4:1

4:2

4:3 Eve

gives ‘arecur inbrotherl4:2). Thsuffice drawinglove-hateternal lappears WE KN

The

carries semantiprogresshowevedimensidown’? levels in

3:4–10

3:11–18

3:19–24

4:1–6

4:7–11

4:12–18

4:19–5:5

n though tha semantic cn a concently love (2:2hese themeto give a c

g attention tte, etc. The life can be on its own

NOW THAT

C

strength oit out, the

ic orientatios on these

er, several qion? Must n

What is thn the analys

Crite

Crite

Foun

Iden

Foun

Crite

Visib

his descriptcontinuationtric pattern:, 3:4, and 4

es are impocomprehensto a fundamimportant mexpanded apage oppos

T WE POSS

CERTAINT

f du Rand’e clear andon (with itspoints thanquestions a

not a ‘bottomhe rationalesis: the mea

erion (negati

erion (positiv

ndation of ce

ntification tes

ndation of lo

eria for the fu

ble text for lo

ion is the on in the line a christoc

4:2), and tesortant for thsive, coheremental pattemotif in 1 Jaccording tosite.

SESS ETER

Y THAT O

s analysis d ‘from-the starting po

n any earliearise: Wherm-up’ analye for the maanings of co

ive) for filiat

ve) for filiati

ertainty (filia

st for relation

ove

fulfilment of

ove

one in du Re of thoughtcentric witnsts for indenhe letter’s ent picture ern of antithJohn regardio a diagram

RNAL LIFE

ONE POSSE

of 1 John ie-bottom-upoint in the er analysis re in the anysis be suppany interprola, the rela

ion

ion

ation)

nship (filiatio

love

Rand’s treatt’, he notes

ness (items ntifying antiprogressionof the lette

hesis in 1 Joing the we-g

m furnished

E AND MU

ESSES ETE

is to be foup’ analyticasyntactic cohas. From nalysis do plemented wetive choic

ations betwe

on)

ment that mthat certain1 and 5:1),ichrists/falsn of thoughr as a who

ohn: God-degroup’s certby du Rand

ST SHOW

ERNAL LIF

und in the al process, ola-system)the point owe find th

with a treatmes that du een cola, etc

most authorn themes an, sin (2:1 ase prophets ht, but theyole. It is alsevil, light-dtainty of pod (1981: 36

IT IN OUR

FE

methodicaland the d

). He makesof view of he text’s prment ‘from-Rand makec.? The form

267 

ritatively nd motifs and 3:3), (3:1 and

y do not so worth darkness, ossessing 6), which

R LIVES

way he doggedly s greater method,

ragmatic -the-top-es on all m of the

268  

analysis gives the impression of objectivity, but it clearly has a subjective character. It amounts to an account of how du Rand reads the text but without naming all the reasons for it. The overall interpretation actually turns out to be quite (Christianly) traditional. One might ask why.

J.P. Louw, Professor in Greek at Pretoria and the foremost spokesperson for the South African perspective, published an article in 1992 which both expands and delimits this type of discourse analysis. To begin with, the argument is directed at those who believe they have understood a message once they have understood the words or sometimes also the sentences. ‘Reading a text involves far more than reading words and sentences. This implies that a primary reading of a text is not necessarily the “correct” reading in terms of the original author’s intention’ (1992: 18). He lists the many features that determine the interpretation of a text: extralinguistic, paralinguistic and linguistic. Reading is actually a very complex process. It is here that discourse analysis enters in as a conscious procedure. It reveals how we read and understand a text. Discourse analysis is not intended to be a recipe for ensuring a final reading of a passage, void of any subjective element. ‘It is rather a demonstration, a displaying or showing, first of all to oneself, how a text is being read, then giving account to others how the text is read and used to eventually come to an understanding of the text. In short, it is revealed reading; it charts the course of the reading process’ (1992: 18).

To further underscore the complexity of the reading process, Louw briefly describes the various aspects that can be involved in a discourse analysis: syntax, semantics, pragmatics, as well as text-typology, cognitive processes and individual styles. ‘There are numerous aspects to be recognized if one intends to read closely. Many readers, however, though they think they read closely, will rarely stop and check whether all discourse features have been considered. Yet the more one considers, the more one can expect to infer from a text’ (1992: 19). But there are so many pitfalls here that Louw wonders whether a comprehensive reading is generally possible. In the end, what is decisive is the possibility of giving reasons for what has been understood. And here he warns against one of the chief sins of exegetes: overinterpetation.

This sensible discussion of text interpretation and discourse analysis considerably broadens the perspectives found in du Rand. Yet what Louw gives with one hand he takes back with the other. When he tells us what discourse analysis is, he immediately places a restriction on it: ‘It gives an account of how a person understands the syntax and semantics of a text’ (1992: 18). It deals with a text’s syntax and semantics, especially the latter. It is necessary in interpretation to go behind a text (to deal with paralinguistic and extralinguistic circumstances), but Louw warns against going beyond the text: Going beyond a text involves a refusal to recognize the discourse constraints of a text. The closest one can get to this ideal is to take the linguistic syntax, which is perhaps the most objective feature of a text, as the point of departure that will constrain the overall process of discourse analysis (1992: 19–20).

In spite of everything Louw says about the many aspects in a discourse analysis, he limits them here to the syntactic-semantic and to a single perspective on the text: the message which the author intended. Thus, for all practical purposes, Louw ends up at the same point as du Rand begins in 1981. Du Rand’s analysis reveals much about 1 John, but it is difficult to accept that his above-cited theme for the letter captures the whole.

4. Discourse Analysis in Texas: Miehle 1981; Longacre 1992

For anyone who wants to produce a good translation of an ancient text, discourse analyses are important. No wonder that several discourse analysts have been and still are affiliated with the United Bible Societies or with Wycliffe Bible Translators. To the latter belong Kenneth L. Pike, John Beekman, Kathleen Callow, Helen Louise Miehle, Stephen H. Levinsohn and

269  

Robert Longacre. In 1981, Helen Louise Miehle defended her doctoral dissertation ‘Theme in Greek Hortatory Discourse: Van Dijk and Beekman-Callow Approaches Applied to 1 John’ at the University of Texas at Arlington. The work is dedicated to John Beekman; Robert Longacre, Professor of Linguistics at UTA, had been her mentor. Longacre himself published an article in 1992 on the same subject: ‘Towards an Exegesis of 1 John Based on the Discourse Analysis of the Greek Text’.

Using 1 John as a test case, Miehle (1981) tries out two different linguistic models for text analysis: van Dijk’s more information-oriented model (see, for example, van Dijk 1977) and Beekman’s and Callow’s analytical method, which is more literarily and semantically determined. She does this by particularly checking them against four aspects in a discourse analysis:

(1) In what way do van Dijk and Beekman-Callow include factors of the communication

situation? (2) How do they describe each linguistic unit in an analysis ‘from the bottom up’? (3) How do they determine the relations between units and motifs in the text? (4) How do they derive and formulate the themes for each unit from the text?

Miehle presents the two models, applies them to 1 John, and evaluates the results. This

leads in turn to a scarcely successful attempt to combine Van Dijk and Beekman-Callow and to a very productive analysis of the perlocutionary function in 1 John. ‘1 John is shown to be a hortatory (not simply expository) text with the perlocutionary function of persuasion’ (1981: ix).

An analysis following Beekman’s and Callow’s model—which is carried out in detail in the dissertation—has much in common with du Rand’s semantic analysis but builds upon so-called propositions and uses a series of logical qualifiers to determine the hierarchical build-up of units on various levels. The result, however, is rather different. The overarching theme is taken from the most important sections of the corpus, namely 4:1–6, 7–10, and 5:6–12, as well as from the letter’s beginning and ending. It is summarized as follows:

Introduction 1:1–10 I am writing about true things I have witnessed first hand about God through his Son Jesus Christ so that you can BELIEVE, OBEY, COMMUNE WITH God and with those who believe the truth.

Body 2:1–5:12 BELIEVE the truth, LOVE God and others who believe the truth, BELIEVE and RECEIVE eternal life.

Closure 5:13–21 Because you BELIEVE the truth, OBEY the true God. It is apparent that a bottom-up semantic analysis with a fixed thematic, hierarchical structure as its goal can yield differing results, though the use of logical qualifiers for the relations between units makes the hierarchy firmer with Beekman-Callow/Miehle than with du Rand. Among other things, this indicates that implicit interpretation is woven into the apparently objective procedure.

The most important thing about Miehle’s treatment is that she compares Beekman-Callow with another, more communication-situation oriented model, and that forces her to go further than she otherwise might have. She evaluates the practical applicability of the Beekman-Callow model and concludes that it can be used effectively to segment a text into its individual elements, to describe relations between various units, and to distill the theme from the most prominent statements. At the same time, she identifies the greatest weakness of this system: ‘its scanty treatment of the communication situation and how this communication situation influences the shape of the discourse in specific ways’ (1981: 114). We would put it

270  

this way: it lacks the pragmatic dimension. In Miehle’s terms, this means, among other things, ‘the omission of perlocutionary function from the propositions in the display’ (1981: 115).

The chapter on perlocutionary function in 1 John thus becomes the most important contribution of Miehle’s work, marking a fundamental shift in the history of discourse analysis. The restriction to the syntactic and semantic dimensions and to the merely explicit begins to be relaxed. By perlocutionary, Miehle means the special way in which a speaker tries to influence the hearers’ behaviour. What does the ‘influence-structure’ look like in 1 John? Miehle investigates the various forms for exhortation found in the letter: imperatives, the word ἐντολή with ἵνα-clauses, ὀφείλω, common phrases of the type πᾶς plus participle, and ἐάν/ὅταν-clauses. If one considers all the expressions for exhortation, including the more indirect, it can be shown that 1 John is more exhortative in its character than informative. ‘1 John was written primarily to persuade its readers to act consistently with what they say they believed, rather than to inform them about what was desirable to believe’ (1981: 178). The way in which the author seeks to persuade (I have attempted to describe 3 John in this respect in my 1986 article) is as worthy of continued study, with an eye to the analysis of the letter as a whole, as is any other subject.

Longacre (1992) builds upon his student’s investigations, but without mentioning her by name. The issue of pragmatics becomes much clearer as he moves beyond the semantic-thematic restrictions characterizing the South African and Beekmanian models and concentrates on the question of how the text can be thought to have functioned. Thus he leaves behind, too, the merely explicit and the counting of content-words in the text, which is so common in many discourse analyses.

Like Miehle, Longacre divides the text into many smaller pieces (she 18, he 16 paragraphs) and argues for an introduction that is longer than the usual 1:1–4. The first step in Longacre’s analysis is to determine ‘structural paragraphs based on the surface structure of the book’ (Longacre 1992: 272–76). The analysis, which is not recounted in its entirety in the article, is based primarily on the use of vocatives in the letter and on certain changes in the text. The result is an outline with 16 sections—‘a string of natural paragraphs’—of quite different character and extent: 1:1–4, 1:5–10, 2:1–6, 2:7–11, 2:12–17, 2:18–27, 2:28–29, 3:1–6, 3:7–12, 3:13–18, 3:19–24, 4:1–6, 4:7–10, 4:11–21, 5:1–12 and 5:13–21. These paragraphs are not summarized with themes as is the case with du Rand and Miehle, nor is there any description of the flow of thought from one to the next.

In the second stage of the analysis, Longacre seeks a larger grouping of the paragraphs and finds a criterion for this in the distribution of performative verbs in the text, especially the verb ‘write’. This verb occurs abundantly in the first seven paragraphs, namely 1:1–2:29, which Longacre describes as an introduction. It is an unusually long introduction and contains nearly all the themes of the letter. The author is anxious to indicate why he writes. Here the pragmatic, communication situation comes to the fore. The paragraph 2:28–29 functions as a conclusion. At we meet again, quite unexpectedly, the verb ‘write’, and thus the paragraph 5:13–21 is considered the final conclusion. The letter’s main body, then, turns out to comprehend eight paragraphs, namely 3:1–5:12.

Longacre’s third step treats the question of the type of text we have in 1 John. Is the letter ‘expository’ or ‘hortatory’? Miehle, as we have seen, had argued for the latter. Longacre arrives at the answer by investigating the verbs in the text. The static and relational verbs (be, have, remain, know …) dominate in sheer numbers, but forms for exhortation—which, as Miehle demonstrated, are of quite different kinds in 1 John—have a much greater functional significance. They occur at marked locations in the text which Longacre calls ‘peaks’. On the surface, then, the letter has an expository, informational character, but seen at a deeper level it has a clearly exhortational function.

In a fourth step, Longacre identifies what he calls the peaks of the book, ‘points of cumulative development’ (Longacre 1992: 279–80). For 1 John, they reveal themselves to be

271  

dual in character. In the introduction there is an ethical peak in 2:12–17 (love not the world) and a dogmatic peak in 2:18–27 (remain in what you have heard/in Jesus Christ). These two peaks recur in more pronounced form in the main body of the letter and in reversed order, the dogmatic in 4:1–6 and the ethical in 4:7–21. The latter occurs as a triple thesis in 4:7–10, which is then repeated, expanded and strengthened in 4:11–21. These twin ‘main peaks’ are strengthened also in the preceding and following paragraphs—to be precise, in 3:23 (‘to believe in the name of Jesus Christ and love one another’) and in 5:1 (‘everyone who believes that … everyone who loves …’). In this way the basic structure of 1 John is settled.

From this structure the letter’s macrostructure, as defined by Van Dijk, can be determined, ‘a summary or précis or abstract of a text which is stated so that it gives the central thrust of the whole work’ (Longacre 1992: 280). We need be in no doubt about what this message in fact is. It is explicitly marked in the peaks in the introduction, in the peaks in the main body, and in strategically placed passages immediately before and after these main body peaks. Longacre (1992: 284) formulates it thus: To be a follower of the Christian way requires doctrinal and ethical commitments. We must believe in the central fact that God sent His son, that God came incarnate in human flesh; and we must take this as a central doctrine around which other doctrines of the Christian faith are developed. We must love God and our fellow men, and this requirement is equally stringent.

Faith and love are inextricably interwoven in 1 John. Anyone who separates them from one another forfeits his or her Christian identity.

There are many questions which Longacre’s analysis raises—unfortunately there is not enough room to discuss them in this essay—but it does demonstrate convincingly how fruitful the pragmatic dimension can be for a discourse analysis. The questions one puts to the text should concern not only the what, but the how and the why as well. An analytical outlook oriented to the communication situation and the pragmatic function of a text is essential, at least in certain investigative perspectives.

5. Rhetorical Analyses

In connection with what has just been said, it is natural to think of rhetorical analyses of 1 John. I know of no such studies done in any completeness. Duane F. Watson approached 1 Jn 2:12–14 from a rhetorical perspective in 1989, and Kjell Arne Morland (1992) used rhetorical concepts in his analysis of Christology in 1 Jn 4:1–6. It is symptomatic that precisely these passages first drew the attention of rhetorical critics. They contain several problems which have been difficult for exegetes to solve in the usual ways, and they are central in the communication situation of 1 John. Both studies show how fruitful it is to consider rhetorical issues in an analysis of these passages.

The only scholar to describe 1 John in its entirety from a rhetorical point of view is F. Vouga (1990). In his commentary he divides the text—apart from the letter’s prescript and epilogue—into six rhetorical units:

1. captatio benevolentiae 1:5–2:17

2. narratio 2:18–27

3. propositio 2:28–29

4. probatio 3:1–24

5 exhortatio 4:1–21

272  

.

6. peroratio 5:1–2

It is easy to criticize this division of the letter (see Klauck 1990), as it is not built on any

detailed working with the text and shows obvious signs of ‘labeling’. It is, however, worth noting that Vouga observes metalinguistic elements in the text and that he argues the letter has an unusually long introduction. Not until 3:1ff. does the actual ‘argument’ begin.

In an article from 1993, D.F. Watson, too, describes a particular group of rhetorical figures in 1 John. In general, he takes Quintilian’s classification of amplificatio and R.E. Brown’s descriptions of the historical situation and applies them to 1 John. But also here there is no careful, conscientious treatment of the Johannine text. Many of the stylistic features in 1 John get Latin names in Watson’s investigation, but perhaps the most important point with respect to viewing the letter in its wholeness is his description of 1 John as epideictic (1993: 118–23; Vouga assigns 1 John to the deliberative genre). In the words of the concluding abstract:

The repetitive and amplified nature of 1 John is an integral part of the author’s use of epideictic rhetoric to increase his audience’s adherence to the traditional truths of the Johannine Community in the face of secessionist challenges. Amplification clarifies the author’s themes and topics, as well as the distinctions between Johannine tradition and its aberrant forms (1993: 123). In a separate article from 1990, Klauck evaluates rhetorical analyses of 1 John and comes

to a rather negative result: They do not tell us much about the letter’s macrostructure. The advantages for a microstructural analysis are greater, however. His low opinion of a rhetorical perspective on 1 John hangs together with his interest in paying attention to all exegetical aspects, an orientation that often results in an emphasis on content and the theological message. 1 John is primarily interpreted as an exposition and development of the message of love.

Brown interprets the letter more as an act in a definite situation. It is anchored in a Johannine development—being clearly designed according to the arrangement of the Fourth Gospel—and plays a part in a carefully described historical conflict. In spite of the letter’s aspect as an act in history, Brown ignores as a rule the influence that the communication situation has on the structure of the letter. The communicationally significant parts of the text tend to become mere parentheses in his description of the letter as a whole.

Malatesta brings more literary aspects into the discourse analysis, but at the same time forces a vigorous connection with the author’s personality and his deep experiences of the Christian mystery, faith and love. A fundamental theological idea, birthed and nourished in the author through these experiences, becomes decisive for describing the letter as a whole. It is easy to see how the discourse analysis summarizes Malatesta’s interpretation of 1 John.

Du Rand’s study introduces a thorough and methodical analysis on several levels, from the bottom up, and with stress on the semantic and on a linear development of the thought. From each unit a theme is derived, mostly drawn from repetitions of important words in the text, and the whole letter is described hierarchically and thematically. For Louw, du Rand’s fellow South African, discourse analysis is first of all a way of revealing to oneself—and to others—how a text is being read and how it is used to arrive at an understanding of it (1992: 18). This can easily be seen in du Rand’s work. His rigorously applied analysis, however, conceals numerous interpretive steps and is seriously limited by the basic syntactic-semantic perspective.

273  

Miehle has much in common with du Rand in her semantic analysis à la Beekman-Callow, but at the same time, she makes a serious attempt to take into account the influence of the communication situation on the discourse. The pragmatic, communication dimension is developed later by Longacre, without his losing sight of the historical context. That loss does occur in Neufeld (1994), who reads certain verses in 1 John in light of Speech Act Theory. ‘Language is a form of action and power. Discourse becomes responsible for creating reality and not merely reflecting it’ (1994: 133). This shows clearly how the description of a text in its wholeness becomes dependent on which perspective on the text one chooses.

Discourse analyses are thus heavily dependent on text definitions, research perspectives, and overall interpretations. But the narrow reasoning of Adolf Jülicher, cited at the beginning of this article, that ‘The countless attempts to find a well thought-out arrangement in 1 John have the merit of cancelling each other out’, simply will not do.

REFERENCES

Beutler, J.

1988 ‘Die Johannesbriefe in der neuesten Literatur (1978–1985)’, ANRW, II, 25.5: 3773–90. Brown, R.E.

1982 The Epistles of John (AB, 30; Garden City, NY: Doubleday). Bultmann, R.

1973 The Johannine Epistles (Philadelphia: Fortress Press). Combrink, H.J.B.

1979 Structural Analysis of Acts 6:8–8:3 (Stellenbosch Theological Studies, 4; Cape Town). Dijk, T.A. van

1977 Text and Context: Explorations in Semantics and Pragmatics of Discourse (London: Longman). Dressler, W.

1972 Einführung in die Textlinguistik (Tübingen: Max Niemeyer). Kieffer, R.

1972 Essais de méthodologie néo-testamentaire (ConBNT, 4; Lund: C.W.K. Gleerup). 1975 Le primat de l’amour: Commentaire épistémologique de 1 Corinthiens 13 (Paris: Cerf). 1982a Foi et justification à Antioche: Interprétation d’un conflit (Ga 2, 14–21) (Paris: Cerf). 1982b ‘Lingvistiken och nya testamentet: En forskningsöversikt’, STK 58: 139–47.

Kieffer, R., and B. Olsson (eds.)

1983 Exegetik idag (Religio, 11; Lund: Teologiska institutionen i Lund).

                                                            ANRW Hildegard Temporini and Wolfgang Haase (eds.), Aufstieg und Niedergang der römischen 

Welt: Geschichte und Kultur Roms im Spiegel der neueren Forschung (Berlin: W. de Gruyter, 1972–) 

AB Anchor Bible 

ConBNT Coniectanea biblica, New Testament 

STK Svensk teologisk kvartalskrift 

274  

Klauck, H.-J. 1990 ‘Zur rhetorischen Analyse der Johannesbriefe’, ZNW 81: 205–24. 1991a Die Johannesbriefe (Erträge der Forschung, 276; Darmstadt: Wissenschaftliche

Buchgesellschaft). 1991b Der erste Johannesbrief (EKKNT, 23.1; Zürich: Benziger Verlag; Neukirchen-Vluyn:

Neukirchener Verlag). Lohmeyer, E.

1928 ‘Über Aufbau und Gliederung des ersten Johannesbriefes’, ZNW 27: 225–63. Longacre, R.E.

1992 ‘Towards an Exegesis of 1 John Based on the Discourse Analysis of the Greek Text’, in D.A. Black et al. (eds.), Linguistics and New Testament Interpretation: Essays on Discourse Analysis (Nashville: Broadman Press): 271–86. Louw, J.P.

1973 ‘Discourse Analysis and the Greek New Testament’, BibTrans 24: 101–18. 1982 Semantics of New Testament Greek (Chico, CA: Scholars Press). 1992 ‘Reading a Text as Discourse’, in D.A. Black et al. (eds.), Linguistics and New Testament

Interpretation: Essays on Discourse Analysis (Nashville: Broadman Press): 17–30. Malatesta, E.

1973 The Epistles of St John: Greek Text and English Translation Schematically Arranged (Rome: Pontifical Gregorian University).

1978 Interiority and Covenant (AnBib, 69; Rome: Pontifical Biblical Institute). Marshall, I.H.

1978 The Epistles of John (NICNT; Grand Rapids: Eerdmans). Miehle, H.L.

1981 ‘Theme in Greek Hortatory Discourse: Van Dijk and Beekman-Callow Approaches Applied to 1 John’ (unpublished dissertation; University of Texas at Arlington). Morland, K.A.

1992 ‘Om å avsløre og avsondre vranglærere’, TTK 63: 95–112. Neufeld, N.

1994 Reconceiving Texts as Speech Acts: An Analysis of 1 John (BIS, 7; Leiden: E.J. Brill).

                                                            ZNW Zeitschrift für die neutestamentliche Wissenschaft 

EKKNT Evangelisch‐katholischer Kommentar zum Neuen Testament 

BibTrans The Bible Translator 

AnBib Analecta biblica 

NICNT New International Commentary on the New Testament 

TTK Tidskriftfor teologie og kirke 

BIS Biblical Interpretation Series 

275  

Olsson, B.

1974 Structure and Meaning in the Fourth Gospel: A Text-Linguistic Analysis of John 2:1–11 and 4:1–42 (ConBNT, 6; Lund: C.W.K. Gleerup).

1985 ‘A Decade of Textlinguistic Analyses of Biblical Texts at Uppsala’, StudTheol 39: 107–26. 1986 ‘Structural Analyses in Handbooks for Translators’, BibTrans 37: 117–27. 1994 ‘Förstår du hur du läser?’, in S. Hidal, L. Haikola and S. Norin (eds.), Mellan tid och evighet

(Festschrift Bo Johnson; Lund: Teologiska institutionen i Lund): 169–84. Porter, S.E.

1992 Idioms of the Greek New Testament (Sheffield: JSOT Press). Rand, J.A. du

1981 ‘A Discourse Analysis of 1 John’, Neot 13 (1979): 1–42. Segovia, F.S.

1982 Love Relationships in the Johannine Tradition (Chico, CA: Scholars Press). Thiselton, A.C.

1992 New Horizons in Hermeneutics (Grand Rapids: Zondervan). Van Staden, P.J.

1988 ‘Die struktuur van die eerste Johannesbrief’, (unpublished dissertation; University of Pretoria). Vouga, F.

1990 Die Johannesbriefe (HNT, 15.3; Tübingen: J.C.B. Mohr). Watson, D.F.

1989 ‘1 John 2.12–14 as Distributio, Conduplicatio, and Expolitio: A Rhetorical Understanding’, JSNT 35: 97–110.

1993 ‘Amplification Techniques in 1 John: The Interaction of Rhetorical Style and Invention’, JSNT 51: 99–123. Winther-Nielsen, N.

1995 A Functional Discourse Grammar of Joshua: A Computer-Assisted Rhetorical Structure Analysis (ConBOT, 40; Stockholm: Almqvist & Wiksell).

                                                            StudTheol Studia Theologica 

Neot Neotestamentica 

HNT Handbuch zum Neuen Testament 

JSNT Journal for the Study of the New Testament 

ConBOT Coniectanea biblica, Old Testament 

276  

WHERE DOES 1 JOHN 1 END?

John Callow

1. Introduction

Traditionally, the first chapter of John’s first epistle ends with v. 10. And that is a tradition still widely maintained in translations. Figure 1 shows paragraph divisions and headings (where used) in ten modern translations, from the 1946 Revised Standard Version to the 1990 New American Bible and the Revised English Bible. Only two of the ten do not have a paragraph break at the end of v. 10, and of the seven that use headings, three start ch. 2 with a new heading.

The two versions (New Jerusalem Bible, New King James Version) that do not have a paragraph break at 1:10 replace it, effectually, by one at the end of 2:2, and they both have a (new) heading at that point. Five other translations support a break at the end of 2:2. This evidence, then, raises the question of whether the flow of thought initiated at 1:5 (there is no dispute over 1:1–4 being a separate ‘introductory’ unit) continues as far as 1:10 only, or as far as 2:2.

Among Bible translators, it is widely known that versions and commentators often do not agree with each other in the matter of boundaries. Perusal of the commentaries shows that this is the case here. The main interest of the commentators is directed towards the question of whether it is possible to divide the whole epistle into a (relatively) few major parts. In general, there has been considerable scepticism about this. In 1912, Brooke could say, ‘While some agreement is found with regard to the possible division of the First Epistle into paragraphs, no analysis of the Epistle has been generally accepted … perhaps the attempt to analyse the Epistle should be abandoned as useless.’1 Much more recently (1982), Brown says, ‘In 1 John it is virtually impossible to detect a structured sequence of thought …’2 He then goes on to say, however, ‘there may be structure in 1 John … and, even if hesitantly, most commentators have proposed a division of the work that may correspond to the author’s intent’.3 Schnackenburg is somewhat more positive: ‘The beginnings and conclusions constitute recognizable units of thought. Consequently, there should be some consensus among interpreters as to where the main caesuras lie.’4

                                                            1 A.E. Brooke, The Johannine Epistles (Edinburgh: T. & T. Clark, 1912), p. xxxii. 

2 Raymond E. Brown, The Epistles of John (London: Geoffrey Chapman, 1982), p. x. 

3 Brown, Epistles, p. 117. 

4 Rudolf Schnackenburg, The Johannine Epistles (ET; Tunbridge Wells, Kent: Burnes & Oates, 1992), p. 

13. 

NAB

PH

RSV NRSV

LB GNB

NIV REB

NJB

 

The hori Identific

B N J

RV N T

B G N T

U T

izontal lines

cation of th

New AmericJ.B. Phillips1971). Revised StaNew ReviseThe Living BGood NewsNew InternaThe RevisedUniversity PThe New Je

mark paragr

e translation

can Bible (Ns, The New

andard Versed Standard Bible (Whe

s Bible (Newational Versd English BPress, 1990erusalem Bib

aph boundar

ns used, fro

New York: Testament i

sion (New Yd Version (Neaton, IL: Tyw York: Thesion (Grandible (Oxford). ble (London

ries.

om left to rig

Oxford Uniin Modern E

York: NelsoNew York: Oyndale Houe Bible Socd Rapids: Zod: Oxford U

n: Darton, L

ght:

iversity PreEnglish (Ne

n, 1946). Oxford Univuse Publisheieties/Collinondervan B

University P

Longman &

ss, 1990). ew York: M

versity Presers, 1971). ns/Fontana,ible Publish

Press; Camb

& Todd, 198

Macmillan, r

ss, 1989).

, 1976). hers, 1984).bridge: Cam

5).

277 

rev. edn,

. mbridge

278  

NKJV The New King James Bible (New York: Nelson, 1979).

Figure 1. Chart of Translations of 1 John 1:5–2:17

With this background, then, do the commentators consider 1:5–10 to be a ‘recognizable unit of thought’, or is their preference for 1:5–2:2? As might be expected, there is a division of opinion, as among the versions, but I think it would be accurate to say that, since the war, there has been a distinct move towards treating 1:5–2:2 as a unit in the structure of the epistle. Thus, an older commentator, such as Brooke, proposes 1:5–2:6 as a larger unit, and divides it into the two subdivisions of 1:5–10 and 2:1–6.5 Much later, C.H. Dodd is still following Brooke’s division of 1:5–2:6.6 It is generally agreed that the two most influential commentaries on 1 John since the war have been Schnackenburg’s, with German editions since 1965, and Brown’s, in 1982. Both of these commentators recognize a unit beginning at 1:6 and running to 2:2.

However, not all post-war commentators agree. Grayston has his first major unit as 1:5–2:11, and while he recognizes the connection of 2:1–2 with what precedes, he does not see a significant break at 2:27. And evidently other commentators, such as Stott, reached the same conclusion as Schnackenburg and Brown but independently of them.8

Before presenting my own analysis, I want to look briefly at three analyses produced by linguistic colleagues in SIL. The first of these is Hansford’s.9 His main thesis is that 1 John is a type of structured poetry, showing obvious affinities with Hebrew poetry, such as the use of parallelism and chiasmus. In addition to the poetic structure, he also divides John’s epistle up into 18 ‘strophes’, and identifies Strophe 2 as consisting of 1:5–10 and Strophe 3 as 2:1–6, thus reflecting the more ‘traditional’ view of the versions and older commentaries.

The second analysis based on discourse principles is one by Longacre.10 He also comes to the conclusion that the first two units, after the opening four verses, are 1:5–10 and 2:1–6. His reasons for these decisions will be discussed in the next section of this essay, as they illustrate clearly a key issue in discourse analysis: how to evaluate the data—which are the same for everyone—in arriving at an analysis.

The third analysis is that of Sherman and Tuggy.11 They divide 1 John into ‘semantic units’ of varying size. Unlike Hansford and Long-acre, they analyse 1:5–2:2 as constituting a semantic unit, a ‘paragraph’ in their terminology.                                                             5 Brooke, Johannine Epistles, p. xxxiv. 

6 C.H. Dodd, The Johannine Epistles (London: Hodder & Stoughton, 1946), p. lx. 

7 Kenneth Grayston, The Johannine Epistles (London: Marshall, Morgan & Scott, 1984), p. 4. 

8 Stott clearly identifies 1:5–2:2 as a unit in the epistle, giving it a distinctive heading. John R.W. Stott, 

The Epistles of John (Leicester: Inter‐Varsity Press, 1960), p. 55. 

9 Keir L. Hansford, ‘The Underlying Poetic Structure of 1 John’, Journal of Translation and 

Textlinguistics 5.2 (1992), pp. 126–74. 

10 Robert E. Longacre, ‘Towards an Exegesis of 1 John Based on the Discourse Analysis of the Greek 

Text’, in David A. Black ex al. (eds.), Linguistics and New Testament Interpretation: Essays on 

Discourse Analysis (Nashville: Broadman Press, 1992), pp. 271–86. 

11 Grace E. Sherman and John C. Tuggy, A Semantic and Structural Analysis of the Johannine Epistles 

(Dallas: Summer Institute of Linguistics, 1994), p. 9. 

279  

In the following section, then, I will look at the linguistic data presented by 1:5–2:2 and will seek to evaluate it so as to reach a conclusion about the ‘structure’, or ‘unit(s) of thought’ in this part of John’s first epistle.

2. The Analysis

a. The Structure The Greek text12 of 1:5–2:2 is presented, along with its structural analysis, in Figure 2, as this saves a considerable amount of explanation.13 Certain features immediately stand out: (1) There is a clear threefold pattern, labelled as Units 1, 2 and 3. (2) Each ‘Unit’ consists of two protasis + apodosis constructions, each protasis (i.e. six in

all) being introduced by ἐάν. (3) Each apodosis is double in form (labelled ‘x’ and ‘y’), and in each case the second half

is introduced by καί. (Thus, again, six in all.) And in three cases, 6e, 8d and 10d, the καί clause is negatived by οὐ(κ).

(4) Verse 5 stands outside this threefold patterning. (5) 2:1a and 1b break the pattern, which, if strictly regular, would have started at 1c.

ref. Greek text structure

1:5a καὶ ἔστιν αὕτη ἡ ἀγγελία orienter

1:5b ἣν ἀκηκόαμεν ἀπʼ αὐτοῦ

1:5c καὶ ἀναγγέλλομεν ὑμῖν ὅτι

1:5d ὁ θεὸς φῶς ἐστιν SETTING

1:5e καὶ σκοία ἐν αὐτῷ οὐκ ἔστιν οὐδεμία.

UNIT 1

1:6a ἐὰν εἴπωμεν ὅτι protasis

1:6b κοινωνίαν ἔχομεν μετʼ αὐτοῦ

1:6c καὶ ἐν τῷ σκότει περιπατῶμεν, A1

1:6d ψευδόμεθα apodosis (x)

                                                            12 The text used is that of the United Bible Societies’ Greek New Testament, 4th edition, 1993. So far 

as this passage is concerned, there are no significant differences from the text of The Greek New 

Testament According to the Majority Text (New York: Thomas Nelson Publishers, 1982), so the text is 

considered well established. 

13 The ‘lines’ of the Greek text correspond to grammatical clauses, i.e. each contains one (and only 

one) verb, except for the last two lines, where there is no verb present. Each such clause is lettered 

successively a, b, c, etc., within its traditional verse number. The resultant reference system will be 

used throughout this article, without the preceding chapter reference, when there is no ambiguity. 

280  

1:6e καὶ οὐ ποιοῦμεν τὴν ἀλήθειαν. apodosis (y)

1:7a ἐὰν δὲ ἐν τῷ φωτὶ περιπατῶμεν protasis

1:7b ὡς αὐτός ἐστιν ἐν τῷ φωτί,

1:7c κοινωνίαν ἔχομεν μετʼ ἀλλήλων

1:7d καὶ τὸ αἷμα Ἰησοῦ τοῦ υἱοῦ αὐτοῦ apodosis (x) B1

καθαρίζει ἡμᾶς ἀπὸ πάσης ἁμαρτίας. apodosis (x)

UNIT 2

1:8a ἐὰν εἴπωμεν ὅτι protasis

1:8b ἁμαρτίαν οὐκ ἔχομεν, A2

1:8c ἑαυτοὺς πλανῶμεν apodosis (x)

1:8d καὶ ἡ ἀλήθεια οὐκ ἔστιν ἐν ἡμῖν. apodosis (y)

1:9a ἐὰν ὁμολογῶμεν τὰς ἁμαρτίας ἡμῶν protasis

1:9b πιστός ἐστιν καὶ δίκαιος apodosis (x) B2

1:9c ἵνα ἀφῇ ἡμῖν τὰς ἁμαρτίας

1:9d καὶ καθαρίσῃ ἡμᾶς ἀπὸ πάσης ἀδικίας. apodosis (y)

UNIT 3

1:10a ἐὰν εἴπωμεν ὅτι protasis

1:10b οὐχ ἡμαρτήκαμεν, A3

1:10c ψεύστην ποιοῦμεν αὐτὸν apodosis (x)

1:10d καὶ ὁ λόγος αὐτοῦ οὐκ ἔστιν ἐν ἡυῖν. apodosis (y)

2:1a τεκνία μου, ταῦτα γράφω ὑμῖν orienter

2:1b ἵνα μὴ ἁμάρτητε.

2:1c καὶ ἐάν τις ἁμάρτῃ, protasis

2:1d παράκλητον ἔχομεν πρὸς τὸν Πατέρα apodosis (x) B3

Ἰησοῦν Χριστὸν, δίκαιον.

2:2a καὶ αὐτὸς ἱλασμός ἐστιν περὶ τῶν ἁμαρτιῶν ἡμῶν, apodosis (y)

2:2b οὐ περὶ τῶν ἡμετέων δὲ μόνον,

2:2c ἀλλὰ καὶ περὶ ὅλου τοῦ κόσμου

Figure 2. Chart of the Structure of 1 John 1:5–2:2

281  

We have here, then, a highly structured piece of writing, the sort of structure that led Hansford to consider it to be a form of poetry.14 In addition to the formal structure, there are numerous lexical links:

(1) φῶς and σκοτία/σκότος in 5d, 5e, 6c, 7a and 7b; (2) κοινωνίαν ἔχομεν in 6b and 7c; (3) περιπατῶμεν in 6c and 7a; (4) καθαρίζει/καθαρίσῃ ἡμᾶς ἀπὸ πάσης ἁμαρτίας/ἀδικίας in 7d and 9d; (5) (οὐκ) ἔστιν ἐν in 5e, 7b, 8d, 10d (that is, in all three Units); (6) ἁμαρτία in 7d, 8b, 9a, 9c and 2a, ἁμαρτάνω in 10b, 1b and 1c (again, in all three Units). This evidence means that not only is this a strongly structured piece of writing, but it is also lexically cohesive. Further, when it is observed from the text that φῶς and κοινωνία are closely connected in thought (see 6b and 6c, 7a to 7c), it can be seen that the author has linked v. 5 to Unit 1, Unit 1 to Unit 2, and Unit 2 to Unit 3; 1:5 to 2:2 forms a cohesive whole. I consider, therefore, that this detailed ‘discourse’ analysis supports those versions and commentators who see 1:5–2:2 as constituting ‘a recognizable unit of thought’. b. The Role It is not enough, however, simply to establish that such a unit exists within the larger body of the epistle. The question must also be asked concerning its role. To put it in more straightforward terms: What was the author seeking to communicate to his readers by this ‘unit of thought’? Headings in versions, and in structural outlines in commentaries, make it clear that both translators and commentators accept that 1:5–2:2 is communicating a basic thought, and such headings state what that thought is considered to be. When such headings are compared with each other, however, there are obvious differences:

NJB 1:5–7 To Walk in the Light

1:8–2:2

First condition: to break with sin

NKJV 1:5–2:2 The Basis of Fellowship with Him

Brown15 1:5 This is the gospel: God is light and in Him there is no darkness at all

1:6–2:2

Three boasts and three opposite hypotheses, reflecting different understandings of the gospel

Jones16 1:5–2:27 God is light                                                             14 It is of some interest that the New Jerusalem Bible prints the text in lines, as poetry. 

15 Brown, Epistles, p. 765 (Chart Six of Appendix 1). 

16 P.R. Jones, ‘A Structural Analysis of 1 John’, Review and Expositor 67 (1970), pp. 433–44, cited by I. 

Howard Marshall, who has a very helpful review of different structural analyses, The Epistles of John 

(Grand Rapids: Eerdmans, 1978), p. 23. 

282  

1:5–2:2

Communion with God and confession of sin

Feuillet17 1:5–2:28 (29) The demands of communion with God who is light

1:5–2:2 Avoid sin

Schnackenburg18 1:6–2:2 Fellowship with God and Sin

Sherman and Tuggy19

1:5 God is absolutely morally pure and in no way morally impure

1:5–2:2

We should continue to be morally pure, since God is morally pure and is able to forgive our sins on the basis that Jesus died to acquit us of the guilt of all our sins.

Two questions, from a discourse perspective, arise from a consideration of these headings: (1) Clearly, 1:5 is distinctive, and provides a theological context for what follows. But does

it provide a context for 1:6–2:2 only? And if not, how far does it continue to be relevant? And should it therefore have a separate heading from 1:6–2:2?

(2) Is there a single, major ‘thought’ presented in 1:6–2:2? If so, on what basis is it to be established? How can it be ascertained?

Two theoretical issues are involved in any attempt to answer these questions. One is: How can the ‘scope’ of a general statement, such as 1:5, be determined? The other is: How can the more important information in a unit of text be distinguished from the less important? I will discuss these two issues in the reverse order. c. Establishing What Is Important It is intuitively obvious that in a written text some of what is communicated is more important, and some less important, relatively speaking. The fact that children can be taught to read a passage and summarize it shows that they are able to respond to the most important information provided by the writer. Discourse analysis seeks to identify the most important information by a careful study of the form of the text. The problem for the analyst is that the devices used for signalling important information are very varied, both within a language and between languages. They include phonological phenomena (such as stress and intonation in English), morphological and syntactic phenomena (special affixes, the order of the

                                                            17 ‘Etude structurale de la première Epître de Saint Jean’, in H. Baltensweiler and B. Reicke (eds.), 

Neues Testament und Geschichte (Festschrift O. Cullmann; Zürich: Zwingli, 1972), pp. 307–27, cited 

by Marshall, Epistles, p. 25. 

18 Schnackenburg, Epistles, p. 13. 

19 Sherman and Tuggy, Johannine Epistles, p. 10. 

283  

constituents, etc.), special words (ἰδού), structural patterns (such as parallelism and chiasmus in Hebrew), figurative language, repetition, etc.

This being so, can the important information in 1:6–2:2 be identified? Without wishing to be dogmatic in an area of theory that is by no means fully investigated as yet, the following factors are considered to be significant.

First, it is clear from the grammar that an apodosis is more important than the protasis to which it is related, since the protasis is a subordinate clause, whereas the apodosis is a main one.

Secondly, it seems a reasonable conclusion that B1, B2 and B3 are relatively more important than (the corresponding) A1, A2 and A3. The author has made his starting point clear in v. 5, and B1, B2 and B3 are the ‘correct’ views and practices in the light of 5d and 5e. Put in different terms, A1, A2 and A3 are exposing error, whereas B1, B2 and B3 are describing the true way to behave. Unless the author clearly indicates to the contrary, ‘truth’ (from the author’s standpoint) would rank higher in the scale of importance than ‘error’.

At this point, the somewhat controverted matter of the ‘topic’ needs to be considered. In general, the ‘topic’ is the concept round which the communication, or some unit within the communication, is organized. Put in less technical terms, at the end of the (unit of) communication, you know more about the topic than you did at the beginning. Or, if a reader is asked, ‘What was that about?’ the answer will typically be the topic concept.

Topics can, like prominence, be signalled in many different ways, but a common method is simply that of repeated reference, that is, the topic is referred to regularly throughout the unit of communication that is being considered. The form of the reference will almost certainly be varied, as many languages reject, as bad style, identical repetitions. Thus a noun phrase, a single noun, a synonym, a pronoun and a verbal affix are all common devices for referring to the topic.

In 1:6 to 2:2, the only concept that meets the above criteria for a topic is the concept ‘sin’, formally introduced in 7d with the noun ἁμαρτία (in the phrase ἀπὸ πάσης ἁμαρτίας). This noun is repeated in 8b, 9a, 9c and 2a; the corresponding verb is used in 10b, 1b and 1c; and the synonym ἀδικία is used in 9d. And although in 2:2 the noun is used only once, the περί phrases that are used in 2b and 2c clearly presuppose the ἁμαρτιῶν of 2a.

This evidence indicates, then, that certainly from 7d onwards, the topic concept is ‘sin’. This conclusion is further reinforced by 2:1b. Any statement of an author’s purpose in writing is inherently prominent, since that purpose determines what an author chooses to communicate.20 In 1b, John tells his readers that he is writing to them ἵνα μὴ ἁμάρτητε. Since 1:6–2:2 is such a close-knit, highly structured piece of writing, it is reasonable to assume that this purpose refers to the purpose of this unit, not to the epistle as a whole,21 and that since the purpose refers to ‘sin’ this confirms that ‘sin’ is the topic concept for this unit. Further, it is

                                                            20 Cf. Stephen S. Smalley, commenting on 2:1, 2: ‘John now turns to answer the third version of the 

basic error … In so doing he declares one of his chief objects in writing the document, and comes to 

the heart of his purpose in this section’ (1, 2, 3 John [Waco, TX: Word Books, 1984], p. 34). Note also 

Marshall (Epistles, p. 22), speaking more generally: ‘The structure of 1 John is determined by its 

author’s purpose …’ 

21 Contrast ἀπαγγέλλομεν καὶ ὑμῖν ἵνα καὶ ὑμεῖς κοινωνίαν ἔχητε μεθʼ ἡμῶν (1:3) and καὶ ταῦτα γράφομεν ἡμεῖς ἵνα ἡ χαρὰ ἡμῶν ᾖ πεπληρωμένη (1:4). Since these statements of purpose are found 

in the Introduction or Prologue to the letter, I would assume they state the author’s purposes for the 

whole letter, not just for part of it. 

284  

concluded that the use of the vocative τεκνία μου, and the performative γράφω ὑμῖν, focuses attention on the purpose statement, and so serves to give it added prominence.22

To return now to the question of prominent information in 1:6–2:2, it has been argued that the ‘B halves’ of the three Units are more important than the ‘A halves’, and that ‘sin’ is the topic concept for this unit. Bringing these two points together, the conclusion is drawn that the ‘y’ apodoses, that is, 7d, 9d and 2a–c are more prominent than their ‘x’ counterparts, because in each case the ‘y’ apodosis presents to the readers the ‘true’ answer to the problem of sin. This conclusion is reinforced by the explicit mention of Jesus Christ and his atoning death in 7d and 2a–c, and it is clearly assumed in 9d, since the statement there is equivalent to that in 7d.23

To draw this discussion of prominence to a conclusion, then, the author’s concern in 1:6–2:2 is with sin, and the most important information he gives about sin is that God ‘cleanses’ sin by means of the blood of Jesus, since he is the ἱλασμός for our sins. Any heading, or summary statement for these verses, should contain these prominent elements. d. The ‘Scope’ of 1:5d, e Verse 5 follows immediately on the opening unit of vv. 1–4. Is the conclusion therefore to be drawn that the theological statement of 5d and e that God is light establishes a context, ‘sets the scene’ as it were, for the rest of the epistle? How would a discourse analyst attempt to answer such a question?

It should be clear by now that the way would be to look for evidence, in the epistle itself, that could be considered relevant. But what sort of evidence would that be?

The most significant evidence is that the contrasting concepts of τὸ φῶς and ἡ σκοτία are reintroduced in 2:8–11, occurring at least once in each of these four verses; and that neither τὸ φῶς nor ἡ σκοτία is used again in the epistle. This means that vv. 8–11 form an inclusio with 1:5–7. Also, 2:12 is the first of three very distinctive verses, with their six-times repeated γράφω/ἔγραψα ὑμῖν plus vocative plus ὅτι. Thus there is a clear boundary between 2:11 and 2:12, and it is only in the span 1:5–2:11 that τὸ φῶς and ἡ σκοτία are used. The implication of this data is that the ‘scope of significance’ of the light/darkness contrast is 1:6–2:11.

Looking again at the chart of translations, we note that only the NRSV has no break at 2:11. However, of the seven translations that use headings, only four have a new heading following 2:11; the other three place their headings after 2:17 (GNB, REB) or 2:15 (NIV—which also has a heading following 2:17). It appears, then, that no version regards 1:5/6 to 2:11 as a major unit, with 2:12 initiating the next major unit.

Such decisions in translations are often influenced by the commentators. What, then, do they say? Brown24 divides the epistle into two ‘parts’, with ‘part one’ extending to 3:10.

                                                            22 It is at this point that I would differ from Longacre, and others who see a major break after 1:10. 

They interpret the vocative + performative as signalling a new start; because of the triple pattern of 

the three Units, I interpret it as having a ‘prominence’ function, not an initiating one, in this 

particular context. 

23 Expressed more technically, 7d states the means of cleansing from sin, viz. τὸ αἷμα Ἰησοῦ, whereas 

9d gives the agent of the cleansing action. In 7d the agent is implicit; in 9d the means. 

NRSV New Revised Standard Version 

NIV New International Version 

24 Brown, Epistles, p. 765. 

285  

Within this, 1:6–2:2 and 2:3–11 are simply two of the constituent units. Schnackenburg25 considers that the first ‘part’ ends at 2:17, and gives it the heading ‘Fellowship with God means walking in the light: its realization in the world’. This would seem to imply that the concept of God being light is relevant as far as 2:17. Analyses differing from these are those of Malatesta26 who treats 1:5–2:28 as a major unit, under the heading ‘First exposition of criteria of the new covenant communion with God’; of Feuillet,27 who also analyses 1:5–2:28 (or 29) as a major unit, but has the heading ‘The demands of communion with God who is light’; and of Smalley,28 who heads 1:5–2:29 ‘Live in the light’. In fact, the view that 1:5–2:28 forms the first major unit of 1 John goes back as far as Law,29 who speaks of a ‘First Cycle’ of 1:5–2:28 and gives it the heading ‘The Christian life as fellowship with God, conditioned and tested by walking in the light’.

The only commentator that I am aware of who handles 1:5–2:11 as a distinct major unit is Grayston30 who says ‘1:5–2:11 is chiefly concerned with the moral consequences’.

The purpose of this essay is to present the structure and role of 1:6–2:2. To pursue the discussion beyond 2:2 would extend the discourse analysis of 1 John well beyond the confines of this essay. However, the following are some of the factors that I would consider relevant in extending the analysis beyond 2:2.

First, the bracketing or inclusio of 1:5/6–2:8–11 has to be accounted for in the analysis. Secondly, there are some obvious structural and lexical parallels between 1:6–2:2 and

2:3–11. Corresponding to the triple use of ἐὰν εἴπωμεν, there are the three uses of ὁ λέγων in 2:4, 6 and 9. There are very similar statements, such as ψεύστης ἐστίν (2:4) and ψευδόμεθα (1:6d); καὶ ἐν τούτῳ ἡ ἀλήθεια οὐκ ἔστιν (2:4) and καὶ ἡ ἀλήθεια οὐκ ἔστιν ἐν ἡμῖν (1:8d). Note also the use of περιπατέω in both units (it is not used again in the epistle); of a vocative medially within the unit as in 1:6–2:2;31 of ὁ λόγος (2:5, 7; cf. 1:10). These correspondences, together with the inclusio mentioned above, provide strong discourse evidence that 1:6–2:11 is a larger semantic unit, consisting of two smaller units, 1:6–2:2 and 2:3–11, and with 1:5 providing the theological background for what is said.

                                                            25 Schnackenburg, Epistles, p. 13. 

26 As cited by Marshall, Epistles, p. 23. 

27 As cited by Marshall, Epistles, p. 23. 

28 Smalley, 1, 2, 3 John, p. 17. 

29 R. Law, The Tests of Life (Edinburgh: T. & T. Clark, 1909), cited by Brooke, Johannine Epistles, p. 

xxxvii. 

30 Grayston, Epistles, p. 4. 

31 The occurrence of the vocative in 2:7, plus the reference to a new commandment and the use of 

γράφω ὑμῖν is undoubtedly why eight of the translations have a break after v. 6, and two have a new 

heading. A major division between vv. 6 and 7 was also favoured by older commentators, such as 

Brooke (Johannine Epistles, p. xxxiv). My proposed analysis would be that, as in 2:1, the vocative 

indicates important (prominent) material to follow. At this point the author switches from generic 

statements about all commandments, ‘being in’, and ‘walking’, to the specific commandment to love 

one another. 

286  

One final observation can be made. As has already been observed, 2:12–14 is quite distinctive in form, often being printed as poetry in translations. Two things can be said about these verses:

(a) They, too, like 1:6–2:2 and 2:3–11 make use of threefold patterns (doubled, in fact, in

this case). (b) The concepts used, as reflected in the vocabulary of the author, are both retrospective

(i.e. looking back to 1:6–2:11) and prospective (i.e. looking forward to their use later in the epistle; e.g. τὸ ὅνομα αὐτοῦ [2:12; cf. 3:23, 5:13]; νικάω [2:13, 14; cf. 4:4, 5:4, 5]; ὁ πονηρός [2:13, 14; cf. 3:12, 5:18, 19]).

It seems reasonable, then, to put forward the hypothesis that 2:12–14 might constitute a ‘transition unit’,32 providing a bridge between 1:5–2:11 and what follows.

3. The Theory and Some Implications

The preceding analysis of 1 Jn 1:5–2:2 has made use of units, both larger and smaller, and has sought to study what is the main thrust of those units by ascertaining the most important information in each unit.

The emphases of this approach reflect the underlying assumptions of the theory that is being used, some of which are discussed below.33

(a) 1 John, and the individual parts of it, are an example of human communication, that is,

the author had a message that he wished to communicate to his readers and 1 John is the result. It is not just a linguistic document, existing in its own right, in vacuo, as it were. It is the record of one human being passing on something from his own mind, by means of written language, to others, and so to their minds.

(b) Because it comes from a human mind, it shows some of the fundamental characteristics of human minds. One of these is a strong tendency to ‘group’ information—as is shown clearly in printed communications, with their full stops, commas, paragraph breaks, headings, etc. This innate tendency to group things also shows up in such non-verbal activities as music and art. Such ‘groupings’, when they occur in language, are characterized by ‘coherence’, that is, the parts are compatible with each other and are appropriately related to each other; and by ‘prominence’, that is, some of the information conveyed by the unit is relatively more important than the rest of the information, and will be signalled as such in the data in a variety of ways.

(c) Such groupings, and their prominent information, relate closely to the author’s overall purpose and to his or her more ‘local’ purposes as the communication develops.34

                                                            32 Cf. Grayston, Epistles, p. 4: ‘2.12–14 is a transition unit from the statement to the writer’s 

development of it’. 

33 A fuller discussion of the theory can be found in Kathleen Callow and John C. Callow, ‘Text as 

Purposive Communication: A Meaning‐Based Analysis’, in William C. Mann and Sandra A. Thompson 

(eds.), Discourse Description: Diverse Linguistic Analyses of a Fund‐raising Text (Amsterdam: John 

Benjamins, 1992), pp. 5–37. An application to a particular New Testament text is found in Kathleen 

Callow, ‘Patterns of Thematic Development in 1 Corinthians 5:1–13’, in Black et al. (eds.), Linguistics 

and New Testament Interpretation, pp. 194–206. 

287  

(d) The form of these units, what determines their boundaries and their internal prominence, is strongly affected by the overall type and ‘genre’ of the writing—narrative, teaching (doctrinal, hortatory), etc. A narrative unit is typically characterized by consistency of time, location and participants; a ‘doctrinal’ unit by the development of a topic. Further, a narrative unit is essentially ‘factual’ (even in fiction it is presented as ‘factual’, that is, in the same way as a historical record would be), whereas a ‘doctrinal’ unit is likely to be ‘evaluative’, that is, presenting a claim, or statement, as true or false, and giving arguments in support of the author’s evaluation, from reason, experience, etc.

(e) Finally, as language is a ‘living’ thing, there will always be ‘anomalies’, that is, communications that differ from normal patterns, that stretch the boundaries of both analysis and theory, such as the now-recognized phenomena of ‘fuzzy’ borders, ‘transition’ units, etc.

While such issues, and the analysis of both spoken material and written texts, are of

interest to linguists, anthropologists, sociologists, etc., it may well be asked what interest can they have for a biblical scholar? In response, it can be said that much of biblical scholarship is concerned with exegesis, in the sense of enquiring what the original writer sought to communicate to his readers. The type of discourse analysis outlined above relates very readily to exegetical studies, since it works from the language data to the author’s purpose(s) and message. But it must be emphasized that ‘analyses are accountable to the data’:35 if the analysis contradicts the data, it is wrong; if it does not account for all the data, it is still inadequate; if it goes beyond the data, that is, it is not supported by the data, it is speculative. The number and diversity of the structures proposed for 1 John36 show only too clearly that the data are not yet being handled adequately from a discourse analysis perspective. It is doubtless naive to hope that the insights of discourse analysis will mean that in, say, ten years time, there will be a generally agreed structural analysis of 1 John. But maybe it is not naive to hope that discussion of differing structures will be solidly based on the linguistic data, and accepted, rejected or modified on that basis.

                                                                                                                                                                                          34 Cf. Marshall, Epistles, p. 22: ‘The structure of 1 John is determined by its author’s purpose’. Note 

also Longacre, ‘1 John’, p. 271: ‘I raise the possibility that discourse analysis can suggest a natural 

outline for the book, and beginning with such a natural outline of the book enables us to grasp the 

fundamental thrust of the whole and to understand better what the book is saying’. 

35 Deborah Schiffrin, Approaches to Discourse (Oxford: Basil Blackwell, 1994), p. 416. 

36 Admittedly, a very difficult book to analyse—but would other biblical books and texts fare all that 

much better? Cf. Johannes P. Louw, ‘In Romans 8:1–17 one would think that the paragraphing should 

be fairly obvious, but if we compare the following chart [of seven translations] with the structural 

analysis discussed above, it is obvious that these translations differ in this respect’ (‘A Receptor’s 

Understanding of a Reasoned Discourse: Romans 8.1–17’, in Meaningful Translation: Its Implications 

for the Reader [UBS Monograph Series, 5; United Bible Societies, 1991], p. 102). 

288  

INDEXES INDEX OF REFERENCES

OLD TESTAMENT Genesis 7 142 9 142 19:17 203 33:5–10 224 41:14–45 142 41:32 163 Deuteronomy 5:20 204 8:2–3 129 13:2 204 13:6 204 Joshua 6:26 226 7 302 Ruth 3:9 228 1 Samuel 17 25 24:21 227 2 Samuel 21:19 25 24:1 25 24:9 26 1 Kings 16:34 226 2 Kings 3:6–8:16 226 14:27 227 15:28 226 1 Chronicles 20:5 25 21:1 25 21:5 26 2 Chronicles 15:6 201 Psalms 9:5–6 227 28:3 244

289  

29:3 244 41:5 226 44:20 227 83:4 227 108:13 228 109:13 227 110:1 159 137:9 227 Isaiah 7:9 226 13:10 205 14:22 226 34:4 205 40:3 147 53:8 226 53:10 226 Jeremiah 7:4–15 249 11:19 227 23:27 227 34:4 205 Daniel 2:28–29 201 2:45 201 7:13 205 7:14 205 12:1 204 12:7 200 12:11 203 Joel 2:10 220 3:5 202 4:15 220 Micah 7:6 202 Zephaniah 1:4 227 Zechariah 2:10 205 Malachi 3:1 147

290  

2 Esdras 20:38 199 22:44 199 Matthew 1:1–4:11 129 2:9 36 2:13 121, 128 2:19 121, 128 3 128, 129 3:1 121, 128, 129 3:13 127, 128 3:15 127, 128 4 128, 129 4:1–11 120, 124, 129–31, 134, 139 4:1–4 134 4:1 130, 131, 133 4:2–10 131 4:2 130 4:3 131, 133 4:4 131 4:5 127, 128, 131, 134 4:6 131 4:7 131, 133 4:8 127, 128, 131 4:9 131, 135 4:10 131, 133, 135 4:11 127, 128, 131–33, 135 4:16 129 4:17 128 4:18 137 7:1–11 148 7:17–19 347 7:29 137 8:8 133 8:14–15 133 9:14 121, 128, 129 9:21 137 10:2 125 11:1 90, 91 14:3 137 14:4 137 14:8 121, 133 14:19 188 14:24 40, 137 15:1 121, 128, 129 15:11 347 15:17–20 347 16:21 128 17:1 127, 128 17:26 133

291  

19:21 133 19:22 137 21:27 133 22:16 121, 128, 129 22:37 133 26 128 26:3–46 124 26:30–35 138 26:34 133 26:36–46 120, 124, 129, 135, 136, 138, 139 26:36 127, 128, 135, 136, 138 26:37 135, 136, 138 26:38 136 26:39 136 26:40–41 136 26:40 127, 128, 136, 138 26:41 136, 138 26:43 136, 137 26:44 136 26:45–46 136 26:45 127, 128, 136 26:57–27:31 229 26:69 231 26:71 231 27:11 133 27:18 137 27:22 319 27:23 133, 319 27:38 127, 128 27:46 125 27:65 133 28:2 137 28:19–20 178 Mark 1:1 147, 164, 194, 197 1:2–8 147, 164, 194 1:2 197 1:3–7 198 1:7–8 198 1:7 147 1:8 147 1:9–13 147, 148, 164, 170, 194 1:9–11 147, 148 1:9 217 1:10 148 1:11 166 1:12–13 147, 148 1:12 131 1:13 131, 205 1:14–5:43 147, 149, 164, 165, 170

292  

1:14–15 164 1:14 194, 202 1:16–45 150, 164 1:16–20 150, 164, 217 1:21–28 150, 164 1:29–34 150, 164 1:31 133 1:33 221 1:35–39 150, 164 1:37 33 1:38 202 1:40–45 150, 164 1:44 202 2:1–3:6 165 2:1–12 150, 165 2:2 221 2:13–17 150, 165 2:18–22 151, 165 2:22 199 2:23–27 151, 165 3:1–6 151, 165 3:6 205 3:7–12 165 3:7 202, 205 3:11 218 3:13–19 165 3:14 205 3:15 153, 206 3:18 200 3:20–30 165 3:24 201 3:27 203 3:31–35 165 4 149, 163, 194 4:1–41 165 4:1–34 165 4:11–12 165 4:17 204 4:26 206 4:28 221 4:29 205 4:31 206 4:32 205 4:35–41 150 4:35–40 165 4:36 150 5 17, 149, 151, 152, 165, 169–71, 179, 194 5:1–43 169 5:1–20 165, 169, 171, 174, 194 5:1–5 171, 174 5:1 181

293  

5:2 181 5:3–5 176, 177 5:3 172 5:4 172 5:5 172 5:6–8 171, 172, 175 5:8 176 5:9–10 171, 172, 175 5:11–14 171, 172, 175 5:11 176 5:14 203 5:15–20 171, 173, 176 5:18 176, 181 5:20 174 5:21–43 165, 170, 179 5:21–24 183 5:21 180, 183 5:22–24 180, 181 5:24–34 185, 187 5:24 187 5:25–28 187 5:27 149 5:29–34 187 5:30–34 187 5:31 187 5:34–43 183 5:35–36 180, 181, 184 5:37–43 182, 184 5:37–40 180 5:38–40 195 5:38 199 5:39 207 5:40–41 182 6:1–8:26 147, 151, 166 6:1–7:37 151 6:1–6 151, 166 6:7–13 151, 166 6:7 206 6:8 199 6:11 202 6:14–29 148, 151, 166 6:14–27 201 6:14 153 6:23 201 6:24 152 6:30–56 151, 166 6:30–44 151, 166 6:31 152 6:32 152 6:45–52 151, 166 6:48 166, 205, 206

294  

6:51 205 6:53–56 151, 166 7:1–23 151, 163, 166 7:1–9 166 7:11–12 198 7:18 203, 205 7:24–37 152 7:24–30 166 7:31–37 163, 166 7:32 166 8:1–9 152 8:1 203, 217 8:5 217 8:10–12 152, 166 8:13–21 152, 166 8:17 201, 203 8:19–20 152 8:21 201 8:22–26 152, 163, 166 8:24–9:50 154 8:27–9:50 147, 148, 153, 163, 166 8:27–9:1 153, 166 8:29 148 8:33 203 8:35 199, 202 8:38 205, 206 9 154 9:1 214 9:2–13 153, 166 9:3 204 9:4 205 9:7 144, 148 9:9 203, 205 9:11 201 9:14–29 166 9:18–26 195 9:19 200 9:20 205 9:25 166 9:28 200 9:30–32 153, 166 9:33–50 167 9:33 154 9:36 167 9:37–41 154 9:37 201 9:38 167, 201 9:39–41 167 9:39 201 9:41 199, 201, 214 9:42–47 199

295  

9:48 203 10 154, 167 10:1–52 147, 154 10:1 154, 155, 167, 203 10:2–12 155, 167 10:3 206 10:6 201, 204 10:9 206 10:13–16 155, 167 10:14 204 10:15 214 10:17–31 155, 167 10:17 154 10:19 204 10:21 199 10:25 199 10:26 202 10:28 200 10:29 202, 206 10:30 204 10:32–45 155, 163 10:32–34 155, 167 10:32 154, 155, 200 10:33–34 155 10:33 201, 202 10:35–45 155, 167 10:37 205 10:41–45 155 10:41 200 10:42–43 200 10:42 199, 201 10:43 205 10:45 155, 205 10:46–52 155, 163, 167 10:46 154 10:47 200 10:49 167 10:52 202 11:1–16:8 147, 156, 167 11:1–13:37 156 11:1–11 147, 156, 157, 167 11:1 157, 200 11:2 157, 199, 200 11:3 200, 204 11:4 157, 205, 221 11:5 202, 204 11:7–8 203 11:7 157, 199 11:8 203 11:9 206 11:11 203, 205

296  

11:12–13:37 147, 157, 163, 167 11:12–26 163 11:12–14 167 11:12 157 11:13 205, 207 11:14 199 11:15–19 167 11:15 158, 198, 203 11:17 201, 202 11:19 200, 218 11:20–26 167 11:20–21 205 11:20 207 11:21 200 11:23–24 204 11:23 199, 202, 203 11:25 206, 218 11:27–33 158, 167 11:31 204, 206 11:33 199, 206 12:1–12 158, 167 12:1 200, 202, 205, 206 12:2 200, 206 12:3 201 12:4 206 12:5 201 12:10 200, 203 12:12 158, 205 12:13–17 158, 167 12:13 205 12:14 200 12:18–27 158, 167 12:18 199 12:19 200 12:21 199 12:22 200 12:24 205, 206 12:25 205 12:26 203 12:27 206 12:28–34 158, 167 12:31 206, 221 12:32 200 12:33 199 12:34 200 12:35–37 159, 167 12:35 200 12:36 199 12:37 199 12:38–40 159, 167, 214 12:38 200, 201

297  

12:40 199, 203 12:41–13:37 198 12:41–44 159, 167, 198 12:41 199–201 12:42 199, 201 12:43 199, 200, 202, 214 12:44 199–201 13 17, 198, 210, 211, 213, 215, 217 13:1–2 159, 168, 199, 200 13:1 200, 203, 204, 214, 218 13:2 200, 204, 206, 214 13:3–37 200 13:3–36 159 13:3–4 200, 205 13:3 200 13:4 200, 201, 204–206, 217 13:5–37 200, 217 13:5–27 217 13:5–23 200 13:5–22 205 13:5–13 213 13:5–6 200, 204 13:5 200–202, 204, 206, 207, 218 13:6–36 217 13:6–27 217 13:6 201, 202, 205, 207 13:7–8 201, 202 13:7 201, 202, 203, 207, 218, 222 13:8–9 207 13:8 201, 202, 204–207, 222 13:9–13 201, 220 13:9 201–204, 207, 218 13:10–11 201 13:10 200, 202–204, 207 13:11 202–204, 206, 207, 218 13:12–13 207 13:12 202 13:13 202, 204, 207, 218 13:14–20 201, 202, 207 13:14 200, 202–205, 207, 218, 219 13:15 203, 206 13:16 203, 204 13:17 204, 205, 207, 217, 219 13:18 199, 203, 204, 206, 207, 219 13:19–20 202, 205, 207, 217 13:19 200, 202–205, 214 13:20 202, 204, 205, 207, 219–21 13:21–22 200, 204 13:21 199, 204, 205, 207 13:22 204–207, 221, 222 13:23 200, 204–206, 218

298  

13:24–27 204, 207, 213 13:24–25 220 13:24 200, 205, 206, 217 13:25 201, 205 13:26–27 219 13:26 205, 206 13:27 200, 201, 205, 206, 221 13:28–36 217 13:28–31 205, 207 13:28–29 218 13:28 205, 207, 218 13:29–30 202 13:29 205–207, 218 13:30 199, 205–207, 214 13:31 206, 207 13:32–37 206, 207 13:32 200, 206, 207, 222 13:33 199, 206, 207, 218, 222 13:34–35 218 13:34 204, 206 13:35 200, 204, 206, 207 13:36 199, 202, 204, 207 13:37 199, 206, 207, 218 13:39 201 13:40 203 13:43 200, 205 14 159, 168 14:1–16:8 147, 156 14:1–11 158–60, 163 14:1–2 159, 168 14:3–9 159, 168 14:7–21 159 14:10–11 159, 168 14:12–31 159, 168 14:12–16 159, 168 14:17–21 168 14:22–26 159, 168 14:25 214 14:27–31 159, 168, 195 14:30 206 14:31 214 14:32–42 136, 146, 159, 160, 168 14:33 136 14:35 136 14:36 136 14:43–52 160, 168 14:44 161 14:51–52 160 14:52 168 14:53–15:20 229 14:53–65 161, 168

299  

14:59 200 14:61–62 148 14:61 153 14:66–72 161, 168 14:67 231 14:72 206 15 159, 168 15:1–15 161, 168 15:16–32 161, 168, 195 15:16 161 15:17 161 15:20 161 15:21 161 15:22 161 15:24 161 15:27 161 15:29–32 161 15:29 200 15:33–41 161, 168 15:42–47 161, 168 16:1–8 159, 161, 168, 195 16:2 161 16:4 161 16:6 161 16:8 198, 203 Luke 1:1–4 237 1:4 265 3:7–8 302 4–20 231 4–5 231 4:1–13 131 4:1–12 148 4:1–4 131 4:5–8 131 4:9–12 131 4:13 131 4:33–37 259 4:39 133 4:40 272 6:13 252 6:23 268 6:26 268 6:32 226 8–9 231 8:26–39 259, 272 8:27 273 8:28 272 8:30–33 273 8:30 260

300  

8:39 268 9:37–43 259 10:37 232 11:13 270 11:24–26 270 15 23 15:11–32 21 15:30 252 16:14 275 17:20 215 17:30 268 21:11 268 22 17 22:3 275 22:22 268 22:52 230 22:53 268, 275 22:54–23:25 229 22:54–62 229 22:54 230 22:56 230 22:57 230 22:59 230 22:61 224 22:63–65 229 22:63–64 224, 225, 229, 230, 233 22:63 225 22:66–67 229 22:66 230 23:1 230 23:2 230 23:3 230 23:4 230 23:6 230 23:7 230 23:8 230, 231 23:9 230 23:11 230 23:14 230 23:15 230 23:16 230 23:18 230 23:20 231 23:21 230 23:22 230 23:23 230 23:25 231 23:26 230 23:47 84 Q 3.17 205

301  

4.10 206 6.22 202 7.21 202 7.24 205 7.27 205 10.1 205 10.3 205 10.13 203 10.21 205, 206 10.22 206 11.17–18 201 11.30–32 205 11.30 205 11.51 205 12.11 201 12.12 202 12.23 199 12.37 199, 205 12.39–40 206 12.39 206 12.40 205 12.42–43 206 12.45–46 206 12.53 202 13.34–43 200 13.34–35 200 13.34 205 13.38 205 16.17 205, 206 17.23 204 17.24 205 17.26 205 19.12 206 19.12.21 206 19.13 206 19.15 206 19.24 206 John 5 163 6 163 7:36 76 8:19 228 8:41 228 9:29 228 10:28 358 11:27 33 13 27 17 27 18:12–19:16 229 20:22 27

302  

Acts 1 254 1:1–19:9 267 1:1 238 1:4 274 1:11 252 1:24 252 2:15 251 2:32 252 2:36 252 3:17 268 4:10 232 4:12 232 4:16 251 5:1–6 275 5:12–21 272, 303 5:12 303 5:15–18 303 5:28 232, 252 5:35 251 5:38 252 6–7 239, 255 6:1–7:1 254 6:1–2 241 6:1 241 6:3–5 241 6:3 243 6:5–7:2 241 6:5–7:1 242 6:5 242, 243 6:6 243 6:7 241 6:8 242 6:9–10 241 6:9 242 6:10 242, 243 6:11 241, 242, 247 6:12 242 6:13–14 238, 247 6:13 241, 242, 252, 253 6:14 241, 242, 246, 252, 253 7 17 7:1 241, 247 7:2–8 239, 241, 244 7:2 238, 242, 244 7:3 244 7:4 244 7:5 244 7:6 244 7:7 244 7:8–9 241

303  

7:8 244, 245 7:9–18 241 7:9–16 239 7:9–14 245 7:9–10 246 7:9 245, 246, 253 7:10 245 7:11–12 246 7:13–16 246 7:13 245 7:14 245 7:17–43 239 7:18–19 241 7:18 245 7:20–44 241 7:20–40 246 7:20–22 247 7:20 246 7:21 241, 246 7:22 246 7:23 241 7:24 241, 246, 247 7:25 246 7:26 246 7:27 246 7:28 241, 246, 247 7:29 246, 247 7:31 246 7:32 246 7:33 246 7:34 246 7:35–38 239, 251 7:35 239, 246, 247, 250–56 7:36–38 251 7:36 246, 250, 251 7:37 241, 246 7:38 246, 248 7:39 241, 246, 248 7:40 241, 246, 252 7:43 249 7:44–50 239, 248 7:44–46 249 7:44 246, 248 7:46 241 7:47–50 249 7:48 241, 249 7:51–53 239, 243, 249 7:51 243 7:52 241 7:54–60 241 7:55–60 242

304  

7:55 242, 243 7:56 242 7:58 241 7:59 242, 243 7:60 242 8:4–24 272, 274 8:9–24 265 8:13 265 8:18–23 275 8:18–20 265 9:13 251 11:12 252 13:4–12 265, 274 13:6–10 275 13:23 268 13:27 268 14–20 271 14:1–18:4 271 15 53 15:21 268 16:11–21 103 16:16–18 259, 272, 275 16:17 252, 272 16:18 260 16:20 252 17:5–8 275 17:11 268 17:18 251 17:22 238, 268 18:11 252 18:12–13 275 19 17 19:1–7 270 19:2 270 19:4 270 19:5 270 19:6 270 19:8 269 19:10 269, 271 19:11–12 259, 269, 270, 272 19:13–20 258–60, 263, 268, 271–73 19:13 259, 260, 270 19:14 259, 260, 263 19:15 260–63, 270, 272 19:16 263, 264, 270, 273 19:17 259, 260, 270 19:18–19 277 19:18 264, 266, 270, 273, 274 19:19 263, 266, 267 19:20 268, 278 19:22–34 271

305  

19:22 269 19:25–27 271 19:26 251 19:35 263 19:37 252 20:5–15 103 20:21 271 20:29–30 276 21:1–25 103 21:17–26 275 21:20–21 275 21:27–28 275 22:26 251 23:9 251 23:17 252 23:18 252 23:27–29 275 23:27 251 24:1–9 275 24:5 251 25:1–8 275 25:11 33 25:18 275 25:24–25 275 26:2–3 275 26:21–23 275 26:31 251 26:32 251 27 254, 255 27:1–29 103 27:21–26 256 28:4 251 Romans 1:5 202 1:16–8:39 300 1:16–5:21 300 1:16–3:20 301 1:18–32 101 1:20 203 1:21 205 1:30 202 2:1–29 101 2:12 41 3:20 204 3:21–5:21 301 3:21–4:22 301 4:16–17 302 4:23–5:21 301 4:23–5:11 301 5 287

306  

5:12–21 17, 285–89, 291, 299–301 5:12–18 288 5:12–15 288 5:12–14 288, 289, 291, 299 5:12 287–91, 297, 304 5:13–15 288 5:13–14 290 5:13 288, 290–92, 297 5:14 288, 290–92, 297, 301 5:15–17 288, 292, 294, 299 5:15–16 292 5:15 288, 292–95, 298 5:16–17 288 5:16 288, 292–95, 298 5:17 288, 293–95, 298, 304, 305 5:18–19 288, 290, 295, 299 5:18 288, 295, 298, 305 5:19 288, 295, 298 5:20–21 288, 296, 299 5:20 288, 296, 297, 299 5:21 296, 297, 299 6 287 6:1–8:30 300 6:1–7:6 301 6:1 287 6:4 287 6:15 287 6:19 287 7:7–25 301 7:7 287 8:1–30 301 8:1–17 406 8:33 204 8:36 202 9:1–11:36 300 10:13 202 11:10 205 11:30–31 287 12–15 306 12:1–15:13 300 12:12 202 12:18 204 13 56 13:11 206 15:31 203 16:17 205 16:21–23 58 1 Corinthians 1:8 358 1:27–28 204

307  

3:5 206 3:13 206 3:22 304, 305 4:2 207 4:5 206 4:6 205 5:5 202 5:11 39 7:21 56 7:25 39 7:29 203 7:31 203 8:9 200 10:12 200 10:13 358 10:28 204 12:16 38 14:31 205 14:35 205 15:20–22 302 15:22 304 15:33 200 15:41 205 15:45–50 302 16:13 206 2 Corinthians 1:13 203 3:2 203 4:11 201, 202 4:13 202 5:1 200 5:3 207 5:14–15 306 6:5 206 8:2 199 8:4 199 8:9 199 8:12 199 8:14 199 11:20 201 11:27 206 13:10 206 Galatians 2:2 202 2:6 206 2:9 78 2:16 204 2:18 200 3:8 202

308  

3:14 202 3:28 55 4:18 204 4:19 201, 205 4:27 201 5:15 200 6:5 206 Ephesians 1:1 312 1:2 312 1:3–14 17, 308, 310–14 1:3–10 309 1:3–6 308, 310 1:3–4 310 1:3 309, 310, 312–14 1:4–14 313 1:4–13 309 1:4–9 310 1:4–8 310 1:4–7 309 1:4–6 309 1:4–5 309, 313 1:4 309, 312–14 1:5–6 309 1:5 309, 312, 314 1:6 309, 312–14 1:7–12 309 1:7–10 308–10 1:7–8 309 1:7 308–10, 312, 314 1:8–10 310 1:8–9 309, 313 1:8 312 1:9–10 309, 310 1:9 312, 314 1:10–11 313 1:10 312, 314 1:11–14 309, 310 1:11–12 308–10 1:11 308–10, 312, 314 1:12 309, 312–14 1:13–14 308–10 1:13 308–10, 312, 314 1:14 309, 312–14 2:20 219 3:5 219 6:12 148 Philippians 1:5 204

309  

1:6 358 1:17 201 1:27 217, 218 1:30 217, 218 2:4 41 2:26 217, 218 2:28 218 3:2 201 4:9 217, 218 4:12 199 1 Thessalonians 5:3 203 5:6–7 207 5:27 203 2 Thessalonians 2:2 218 1 Timothy 1:3 331 1:9 325 1:17 328 1:19–20 329 2:1–2 331 2:1 329 2:3 331 2:8 329 2:9 331 2:13–14 326 3:1 328, 331 3:2 328 3:3 325 3:4 324 3:5 324, 325 3:8 331 3:13 319 3:14–16 323 3:14–15 323 3:15 323, 327 3:16 323, 326, 327 4:1–3 323 4:1 323, 327 4:4 319 4:7 321 4:8 319, 321, 322 4:9 319 4:10 319 4:13 39 5:7 327 5:13 269

310  

5:14 329 5:17–25 332 5:20 332 5:25 331 6:9 330 6:11 330 6:17 325 6:21 331 2 Timothy 1:3 331 1:5 321, 322 1:8 329 1:16–17 327 1:18 327 2:10 319 2:14–26 323 2:19 319, 329 2:20–22 323 2:20 324, 328 2:21 324, 328 2:22 323 2:23 323 2:24 325 3:5 327 3:8 330 3:10 330 3:11 326 3:12 326 4:5 330 4:17 326 4:18 332 4:19 331 4:22 331 Titus 1 342 1:1–2:15 339 1:1–4 336, 343, 346 1:1–3 339, 343, 345 1:1 342–45, 349, 350 1:2 349 1:3–4 349 1:3 339, 344, 349 1:4 335, 344, 345, 349, 351 1:5–9 339, 345 1:5–6 346 1:5 331, 338, 343, 349, 351 1:7–8 346 1:8 342 1:9–11 351

311  

1:9 320, 339–41, 343, 346 1:10–3:11 320 1:10–16 332, 340, 341, 344, 345, 347 1:10–12 346 1:10–11 317, 320, 347 1:10 317, 320, 339–41, 343 1:11 320, 340, 342 1:12–13 317 1:12 317, 332, 347 1:13–15 346 1:13 317, 347 1:14 347 1:15 347 1:16 340, 342, 346, 347 2 340 2:1–10 340–42, 345 2:1–2 332, 339 2:1 338–40, 346, 348, 351 2:2 340, 346 2:3–5 346 2:3 331, 340 2:5–6 340 2:5 340, 342, 343, 349 2:6–8 346 2:6 331, 340 2:7 339 2:8 342, 346, 349 2:9 331, 340, 346 2:10 339, 342, 346, 349 2:11–14 340, 341, 343, 345, 348 2:11–13 346 2:11 349 2:12 340, 344 2:13 343, 349 2:14 339, 341, 342, 346, 349 2:15–3:15 340 2:15 331, 338–46, 348, 350, 351 3:1–2 341, 345, 347, 349, 351 3:1 340–42, 346 3:2 342, 346 3:3–7 341, 343–45, 350 3:3 346, 347 3:4–7 346–49 3:4–5 347, 348 3:4 349 3:5 344 3:6–7 347, 348 3:8–11 341, 345 3:8 328, 339–42, 345–47, 350 3:9–11 344, 347 3:9–10 346

312  

3:9 339 3:11 346 3:12–13 345, 346 3:14–15 343 3:14 342, 345, 346, 351 3:15 331, 344–46 Philemon 1 59, 61 2 57, 59, 64 3 56 4 57 5 57 6 57 7 59 8 57, 59 9 57, 59, 61 10 59, 61 13 60 14 59 16 57, 59, 61 17 59 18 59 19 57, 59, 60 20 57, 59 22 60 23 57 24 57, 59 25 57 Hebrews 2:10 359 4:14–5:10 359 5:11–6:20 358, 359, 362, 364, 365, 368 5:11–14 359, 360, 365 6:1–8 358, 360, 361 6:1–3 359, 360, 365 6:3 367 6:4–6 18, 358–60, 362–68 6:5–6 362 6:7–8 359, 360, 365 6:9–12 359, 360 6:13–20 359, 360 7:28 359 12:23 359 James 1:26 33

313  

1 John 1 18 1:1–2:29 385 1:1–10 383 1:1–4 372, 374, 375, 378, 385, 392, 402 1:2 400 1:3 378, 401 1:4 401 1:5–5:12 372 1:5–3:10 374 1:5–2:29 403 1:5–2:28 375, 398, 403 1:5–2:27 398 1:5–2:17 372, 378, 387 1:5–2:8–11 403 1:5–2:11 394, 402–404 1:5–2:6 394 1:5–2:2 372, 376, 379, 394, 395, 397–99, 404 1:5–10 385, 394, 395 1:5–7 398, 402 1:5 374, 375, 392, 393, 396–400, 402, 404 1:6–2:22 403 1:6–2:8–11 403 1:6–2:11 402, 404 1:6–2:2 374, 398, 399, 401–404 1:6–10 372 1:6 393, 396–98, 400, 402, 403 1:7 393, 396–98, 400, 401 1:8–2:2 398 1:8 393, 396, 397, 400, 403 1:9 393, 397, 400, 401 1:10 392, 393, 396, 397, 400, 401, 403 2 392 2:1–5:12 383 2:1–6 385, 394, 395 2:1–2 394 2:1 375, 393, 396, 397, 400, 401, 403 2:2 392–94, 397, 398, 400, 401, 403 2:3–11 372, 374, 376, 379, 402–404 2:3 393 2:4 393, 403 2:5 393, 403 2:6 393, 403 2:7–11 385 2:7–8 375 2:7 393, 403 2:8–11 402 2:8 393 2:9 393, 403 2:10 393 2:11 393, 402

314  

2:12–28 376 2:12–17 372, 374, 379, 385, 386 2:12–14 372, 387, 404 2:12 393, 402, 404 2:13 393, 404 2:14 393, 404 2:15 393, 402 2:16 372, 393 2:17 393, 402 2:18–4:6 378 2:18–3:24 372 2:18–27 372, 374, 379, 385–87 2:21 375 2:26 375 2:27 372 2:28–3:24 372 2:28–3:10 372, 374 2:28–3:3 379 2:28–29 385, 387 2:28 372 2:29–4:6 376 2:29–3:10 376 2:29 372, 376, 377 3:1–24 387 3:1–6 385 3:1–3 375 3:1 372, 387 3:4–10 379 3:7–12 385 3:10 402 3:11–5:12 374, 385 3:11–24 373, 376 3:11–18 379 3:11 374 3:12–24 374 3:12 404 3:13–18 385 3:18 372 3:19–24 379, 385 3:23 376, 386, 404 3:24 372 4:1–5:12 373 4:1–21 387 4:1–6 372–74, 376, 379, 383, 385–87 4:4 404 4:7–5:13 376 4:7–5:5 378 4:7–5:4 374 4:7–21 372, 373, 376, 386 4:7–11 379 4:7–10 383, 385, 386

315  

4:8 376 4:11–21 385, 386 4:12–18 379 4:16 376 4:19–5:5 379 5:1–13 376 5:1–12 372, 373, 385 5:1–2 387 5:1 386 5:4–12 374 5:4–5 372 5:4 404 5:5 404 5:6–21 378 5:6–12 383 5:6–8 372 5:13–21 372–74, 384, 385 5:13 377, 378, 385, 404 5:14–21 376 5:18–21 377 5:18 404 5:19 404 Jude 1 99, 106, 108, 112–16 2 99, 104, 106, 108, 115 3 95, 99, 106–108, 114, 115 4 99, 108, 109, 114–17 5 99, 106, 108, 112–16 6 116 7 99 8 99, 105, 109, 110, 117 9 99, 101, 106 10 99, 109, 112, 113, 117 11–14 106, 107, 112 11 99, 107, 109, 110, 117 12 99, 106, 107, 115, 117 13 99, 106, 117 14–16 101 14 99, 116 15 99, 107, 112, 113, 116 16 99, 101, 106, 117 17–24 102 17–23 101, 103 17 99, 101, 103, 106, 110, 112, 114, 115 18–24 102 18 99, 101, 106, 115 19 99, 101, 105, 117 20 99, 101, 105, 106, 112, 113, 115 21 99, 101, 106, 110, 114–16 22 99, 101, 110, 115

316  

23 99, 101, 106, 110, 115 24 99, 101, 106, 115, 116 25 99, 112–16

OTHER ANCIENT REFERENCES Pseudepigrapha 3 Baruch 9.7 219 1 Enoch 72–82 216 91.11–17 216, 221 91.16 221 93.1–10 221 93.3–10 216 4 Ezra 3:21–22 302 3:24 302 4:30 302 7:118 302 13:16 203 Qumran 1QM 11.7–8 219 4Q521 2.2.1 219 8.9 219 9.3 219 CD 2.12–13 219 6.1 219 Philo Somn. 1.23 220 Josephus War 5.200 199 7.123–57 220 Classical Demetrius Peri Hermeneias 1.10–26 366 1.18 366 Longinus

317  

Peri Hupsoes 32.1–8 366 Pliny Nat. Hist. 2.18–19 220 Ps.-Xenophon Cyneget 5.4 220 Suetonius Nero 49 214 Thucydides History of the Poloponnesian War 1.22.1 236 Syr. Bar. 23.4 302 54.15 304

INDEX OF AUTHORS

Aijmer, K. 96 Aland, B. 80, 145, 160, 180, 219, 220, 222 Aland, K. 80, 122, 145, 160, 180, 219, 220, 222, 308 Alexander, L. 273 Alford, H. 323 Allen, W.C. 138 Allison, D.C. 129, 130, 134 Alonso Schökel, L. 364 Altenberg, B. 96 Althaus, H. 216 Althusser, L. 49 Andrews, A. 322 Arnold, G. 211 Aschbach, A. 208 Aune, D.E. 211, 213 Bal, M. 210, 211 Baltensweiler, H. 398 Banker, J. 285, 317, 320, 341 Barnett, R. 92, 93 Barnwell, K. 16, 283 Barrett, C.K. 238, 239, 247, 248, 250, 252–54, 256 Bateman, J. 336 Bauer, W. 219, 220, 222 Beasley-Murray, G.R. 210, 211 Beekman, J. 285, 293, 323, 382–85, 389 Berge, D. la 141

318  

Berger, K. 213, 215 Beutler, J. 370 Biber, D. 73, 96, 97 Black, C.C. 198, 212, 213, 217 Black, D.A. 16, 78, 239, 283, 395, 404 Black, E. 83 Black, S.L. 17 Blakemore, D. 316 Blass, F. 218, 222, 252, 253, 316 Bobrow, D.G. 140 Böcher, O, 277 Boos, D. 127 Botley, S. 89–91 Bouman, G. 311 Brandenburger, E. 210, 213, 217, 218 Brawley, R. 260 Breytenbach, C. 213 Broek, R. van den 274 Brooke, A.E. 392, 394, 403 Brown, G. 50, 82, 225, 258 Brown, H. 94 Brown, R. 223 Brown, R.E. 371, 373, 375, 387, 388, 392, 394, 398, 402 Brown, R.K. 327 Bruce, F.F. 232, 248, 250, 261 Bühler, K. 208 Bultmann, R. 256, 372 Burnard, L. 93 Burridge, R.A. 210 Burton, E.D.W. 303 Bushnell, M. 98 Buth, R. 126, 127 Cadbury, H.J. 224, 225, 235, 236, 249 Caldas-Coulthard, C.R. 47 Callahan, A.D. 56 Callinan, L. 194 Callow, J. 18, 127, 285, 293, 404 Callow, K. 382–84, 389, 404 Cambier, J. 309 Caragounis, C.C. 292, 294 Carmignac, J. 214, 215 Carnap, R. 210 Caron, J. 45 Carroll R., M.D. 85 Carson, D.A. 16, 45, 79, 84, 88, 100, 105, 304 Carter, R. 239 Chilton, B. 80 Chomsky, N. 84 Cilliers, L. 367 Cole, P. 316 Collins, A.M. 141

319  

Collins, J.J. 215 Collins, P. 96 Colmenero-Atienza, J. 236, 238, 239, 244, 248 Combrink, H.J.B. 355, 377 Comfort, P.W. 327 Conrad, S. 73, 96, 97 Conzelmann, H. 261, 264 Coppieters, H. 310 Cotterell, P. 283, 285 Coulthard, M. 47 Coutts, J. 310 Craft, R.A. 80 Cranfield, C.E.B. 147, 150, 154, 156, 157, 161, 162 Cronjé, J.V. 357, 366, 367 Cruse, D. 86 Dahl, N.A. 245, 248, 310 Darr, J.A. 221 Dautzenberg, G. 218 Davies, W.D. 129, 130, 134 Debrunner, A. 218, 222 Deibler, E. 285, 288, 289, 294, 303, 304 Deichgräber, R. 310 Deist, F.E. 357 Denaux, A. 209 Dexinger, F. 221 Dibelius, M. 235, 236, 239 Dijk, T.A. van 50, 196, 266, 383, 386 Dodd, C.H. 394 Dodds, E.R. 277 Donahue, J.R. 211 Doty, W. 343 Downing, F.G. 210 Dressler, W. 37, 282, 369 Dry, H.A. 93 Dunn, J.D.G. 212, 290–92, 294, 301 Dupont, J. 212, 218, 238 Eberhard, K. 208 Edwards, R.A. 319 Ehrman, B.D. 80 Ellicott, C.J. 327 Ellis, E.E. 334 Eltester, W. 310 Enos, R. 146, 164, 170 Erickson, R.J. 17 Esler, P.F. 276 Evans, C.A. 80 Eyes, E. 83–85 Fairbairn, P. 324, 325 Fairclough, N. 47–52, 54, 55, 60, 62, 63, 66, 67 Fanning, B.M. 121, 123, 126, 132 Fee, G.D. 283, 284, 334, 351

320  

Feuillet, A. 398 Fitzmyer, J.A. 291 Fligelstone, S. 89–91 Foakes-Jackson, F.J. 231, 235, 248 Forell, U. 214 Fornberg, T. 211 Foucault, M. 49, 50 Fowler, R. 47, 258, 261 Friberg, B. 72, 94, 98, 99 Friberg, T. 72, 94, 98, 99 Fryer, N.S.L. 355 Fuller, D.P. 283 García-Miguel, J.M. 241 Garr, W. 224 Garrett, S.R. 221, 265, 274 Garside, R. 74, 79, 83, 86, 89–92, 110 Geddert, T.J. 212, 218 Gempf, C. 237 Gerke, E.-O. 208 Gill, D. 237 Gilman, A. 223 Givón, T. 37 Glasson, T.F. 212, 215 Gnilka, J. 213 Goethe, J.W. 208 Goldfarb, C.F. 93 Goodwin, W.W. 261 Grayston, K. 210, 394, 403, 404 Green, J.B. 282 Greenwood, D.C. 208 Grice, H.P. 30, 316 Grimes, J. 283 Grishman, R. 83 Grundmann, W. 213 Gumperz, J. 46 Gundry, R.H. 132, 136 Guthrie, D. 231, 317, 325 Guthrie, G.H. 88, 104, 365 Güttgemanns, E. 210, 283 Habermas, J. 49 Habicht, C. 220 Haenchen, E. 211, 235, 239, 248, 250, 256 Hage, W. 219 Hagner, D.A. 130, 134, 135 Hahn, F. 213, 218 Hahne, H. 94, 95 Halliday, M.A.K. 17, 29, 30, 32–34, 36, 38–41, 43, 44, 50, 85, 88–91, 100, 223, 230, 237, 240,

241, 247, 250, 254, 262 Hansford, K.L. 395, 397 Hanson, A.T. 334–36 Harnack, A. 235, 249

321  

Hartin, P.J. 355, 356 Hartman, L. 212 Hasan, R. 30, 32, 36, 38, 41, 43, 44, 88–91, 100, 230, 250 Hauck, F. 221 Hawkins, J.C. 122–24 Hawthorne, G. 334 Hayes, M. 98 Healey, A. 177 Healey, P. 177 Heckert, J.K. 317–19, 321, 323, 324, 327–29 Hegel, G.F.W. 221 Hellholm, D. 211, 215 Helm, J. 140 Hendriksen, W. 320, 323, 326, 331, 332 Hengel, M. 235 Hess, R.S. 255 Heyer, C.J. den 355 Hibbard, A. 140 Hiebart, D.E. 324 Hill, C.C. 238, 248 Hirst, G. 87 Hock, R. 213 Hockey, S. 93 Hodge, R. 47 Hoey, M. 88, 106 Holenstein, E. 208 Holman, H. 140 Holmes, M.W. 80 Hölscher, G. 213 Hooker, M. 218 Hopper, P.J. 105, 241, 255 Howard, G. 232 Howard, W.F. 78 Hudson, R. 110 Hwang, S.J.J. 37, 146 Iersel, B.M.F. van 220 Innis, R.E. 208 Innitzer, T. 309 Jacquier, E. 237 Jervell, J. 276 Johnson, E.S. 221 Johnson, L.T. 149 Jones, P.R. 398 Joos, M. 20 Jordaan, G.J.C. 357 Jülicher, A. 389 Kahrel, P. 92, 93 Kaiser, W.C. Jr 225, 283, 284 Kampen, K. van 80 Keck, L.E. 245 Kempson, R.M. 86

322  

Kennedy, G. 92, 93 Kenyon, F.G. 79 Kertelge, K. 218 Kettler, F.H. 310 Kieffer, R. 369 Kilgallen, J. 238, 244, 246, 248 Kilpatrick, E. 194 Kingsbury, J.D. 128, 138 Kirby, J.T. 290 Kirschläger, W. 269 Klauck, H.-J. 370–73, 375, 387, 388 Klein, L. 228 Klijn, A.F.J. 238, 355 Klutz, T. 17 Knight, G.W. 327, 334, 340 Knipp, M.A. 215 Knox, J. 57 Koch, K. 221 Kopesec, M. 285, 293 Kress, G. 47 Kühschelm, R. 218 Kümmel, W.G. 212, 213 Labov, W. 28, 140, 265 Lake, K. 231, 235 Lambrecht, J. 211–13, 218 Lane, W.L. 157, 161, 283 Larsen, S.E. 208 Law, R. 403 Lawler, J. 93 Lee, J.A.L. 214, 219, 221 Leech, G.N. 74, 76, 79, 82–86, 89, 92, 93, 110, 222, 266 Leitner, G. 91 Lenski, R.C.H. 328 Levinsohn, S.H. 16, 17, 120, 121, 127, 133, 134, 146, 170, 239–41, 283, 317–21, 324–26, 329–

32, 382 Levinson, S. 38, 40 Lewandowski, T. 211 Lohmeyer, E. 212, 304, 310, 372 Lohse, E. 302, 303 Longacre, R.E. 17, 140–44, 149, 169, 171, 172, 174, 177, 179, 180, 188, 194, 283, 322, 371, 382,

383, 385, 386, 395, 401 Louw, J.P. 17, 86, 87, 104–107, 283, 284, 313, 355–57, 362, 365, 366, 377, 381, 382, 388, 406 Luz, U. 124, 125 Lyons, J. 78, 100, 126, 258 MacMullen, R. 277 Machen, J.G. 303 Mack, B.L. 213 Maier, J. 216 Malatesta, E. 371, 375–77, 388 Malbon, E.S. 211, 214, 221 Malina, B.J. 229

323  

Mann, W.C. 143, 322, 404 Mansen, K. 174 Mansen, R. 174 Marshall, I.H. 225, 370, 398, 401–403, 405 Martin, R. 334 Martín-Asensio, G. 17, 85, 255 Martyn, J.L. 245 Marxsen, W. 198, 211 Matera, F. 211 Matin, V. 75 Matthews, P. 78 Maurer, C. 310 Mayser, E. 76, 80 McCoy, W.F. 236 McEnery, A. 73, 74, 79, 86, 89, 92, 95, 96, 110 McKay, K.L. 121 McKendrick, S. 80 McKnight, E.V. 211, 214, 221 Menne, A. 208 Merrifield, W.R. 37, 146 Michel, O. 358, 360, 362, 364, 366, 367 Miehle, H.L. 370, 371, 382–85, 389 Miller, C. 228 Miller, P.A. 93 Minor, E.E. 332 Moessner, D. 262 Morgan, J.L. 316 Morland, K.A. 371, 387 Moulton, J.H. 54, 78 Mounce, W.D. 78, 79 Müller, K. 214, 215 Munck, J. 237, 247 Neirynck, F. 213 Nestle, E. 122, 219, 222, 308 Neufeld, N. 389 Newman, B.M. 366, 367 Neyrey, J.H. 229 Nicoll, W.R. 332 Nida, E.A. 16, 86, 87, 104–107, 313, 349, 357, 362, 366 Norden, E. 314 North, H.F. 237 O’Donnell, M.B. 17, 80, 98, 100, 111 Oehler, K. 208 Olsson, B. 18, 369, 371 O’Neil, E.N. 213 Orange, H.P. L’ 220 O’Toole, R.F. 243, 267 Palmer, M.W. 84 Paulston, C. 223 Pearson, B.W.R. 56, 62 Peirce, C.S. 24, 197, 208

324  

Perrin, N. 211 Pesch, R. 210, 213, 217, 219, 221 Petersen, N.R. 52, 56, 210, 212, 217 Petersen, W.L. 212 Petzer, J.H. 355, 356 Phillips, J.B. 393 Pickering, W.N. 286 Pike, E. 144 Pike, K.L. 144, 382 Pirsig, R. 28 Popper, K. 49 Porter, S.E. 16, 17, 45, 47, 75, 79, 80, 82, 84–86, 88, 89, 94, 98, 100, 101, 103, 105, 109–11, 123,

124, 132, 137, 138, 223, 232, 236, 237, 240, 244, 253, 255, 263, 265, 268, 283, 290, 292, 303, 304, 317, 319, 320, 325, 371

Poswick, F.R. 94 Poythress, V.S. 329 Proulx, P. 364 Puskas, C.B. 335, 348, 349 Rader, W. 208 Radke, G. 220 Radl, W. 262 Rand, A. 142 Rand, J.A. du 371, 377–79, 381, 382, 384, 385, 388, 389 Reed, J.T. 16, 17, 36, 45, 47, 49, 52, 54, 63, 75, 78, 79, 82, 85, 86, 88, 89, 94, 100–104, 106, 253,

283, 304, 335–37, 349 Reese, R.A. 101–103 Rehkopf, F. 218, 222 Reicke, B. 398 Reid, D. 334 Reppen, R. 73, 96, 97 Richard, E. 238, 245 Ricoeur, P. 140, 145 Robbins, V.K. 210, 213–15 Robertson, A.T. 120, 123, 133 Robinson, J.M. 310 Rockwell, J. 271 Roloff, J. 272 Routh, J. 271 Rowland, C. 217 Rumelhart, D.E. 140 Rusterholz, P. 219 Sabourin, L. 358, 360, 362–64 Salmon, M. 260 Samuels, S.J. 141 Sanders, J.T. 310 Sawyer, J.F.A. 283 Schalk, C. 310 Schenk, W. 17, 209, 212 Schiffrin, D. 89, 406 Schille, G. 310 Schiller, F. 208

325  

Schmidt, D.D. 84 Schmidt, J.M. 214 Schmidt, S.J. 282 Schmidt, T.E. 220 Schmithals, W. 214, 272 Schmitter, P. 209 Schmitz, W.H. 209 Schmökel, H. 221 Schmoll, H. 76 Schnackenburg, R. 213, 394, 398, 402 Schnider, F. 64 Schoeps, H.J. 213 Schottroff, L. 211, 218–20 Schreiber, J. 211 Schreiner, T. 342 Schüling, J. 212 Schüssler Fiorenza, E. 215 Schweizer, E. 213 Scott, M.P. 144 Segal, A.F. 274 Segbroeck, F. van 210, 212 Segovia, F.S. 371 Shelley, B.L. 233 Sherman, G.E. 395, 399 Shopen, T. 322 Short, M. 96 Silva, M. 86 Simon, E. 220 Sinclair, J. 73 Smalley, S.S. 401, 403 Smith, R.E. 323 Snyman, A.H. 18, 355, 357, 367 Soards, M.L. 237, 238, 244 Speigl, J. 274 Sperber, D. 29, 316, 318 Sperberg-McQueen, CM. 93 Squires, J.T. 255 Stacy, R.H. 366 Staden, P.J. van 371 Stagg, F. 304 Stamps, D.L. 85 Stegemann, H. 215, 216 Stegemann, W. 217 Stenger, W. 64 Stenström, A. 96 Stott, J.R.W. 394 Strecker, G. 210 Ströker, E. 208 Stubbs, M. 96 Summers, D. 20 Sylva, D.D. 238, 243, 247, 248

326  

Szondi, P. 219 Tagawa, K. 211 Tannehill, R.C. 262 Tarski, A. 210 Taylor, V. 212 Telford, W.R. 210 Theissen, G. 213 Thiselton, A.C. 369 Thomas, J. 86, 96 Thompson, S. 143, 241, 255, 322, 404 Thrall, W. 140 Tietzmann, M. 208 Toit, A.B. du 355, 356 Tombs, D. 85, 98 Toolan, M. 265 Towner, P.H. 348 Tröger, K.-W. 209 Tuggy, J.C. 395, 399 Turner, E.G. 77 Turner, M. 283, 285 Turner, N. 120, 121, 123, 124, 128, 252, 253 Tyson, J.B. 260 Uhlig, S. 221 VanderKam, J.C. 216 Verheyden, J. 212, 213 Vermaseren, M.J. 274 Via, D.O. 221 Vorster, W.S. 212, 355 Vouga, F. 371, 387, 388 Waard, J. de 349 Wales, K. 266 Waletzky, J. 140 Wallace, D.B. 98 Wallace, S. 105 Wallis, E. 144 Walter, N. 216 Wanke, J. 218 Watson, D.F. 212, 371, 387 Watt, J.M. 17 Wegener, M. 162 Wendland, E.R. 17, 339, 351 Wenham, D. 210, 212 Wenham, G.J. 144 Werth, P. 32, 37, 38, 42, 43 White, N.J.D. 332 Wick, P. 209 Wilcox, M. 236 Willi-Plein, I. 216 Wilson, A. 56, 73, 74, 86, 95, 96 Wilson, D. 29, 316, 318 Winer, G.B. 320, 321

327  

Winther-Nielson, N. 370 Wirthe, J.R. 141 Witherington, B., III 236, 238 Wodak, R. 47–52, 54, 55, 60, 62, 63, 66, 67 Wohlenberg, G. 219 Wolff, J. 271 Woods, F. 174 Yarbro-Collins, A. 212, 213, 217, 221 Yule, G.U. 50, 82, 225, 249, 258 Zuntz, G. 268 Zyl, H.C. van 356, 359, 360, 367