NEA NIKOMEDEIA I: THE EXCAVATION OF AN EARLY NEOLITHIC VILLAGE IN NORTHERNGREECE 1961-1964. THE EXCAVATION AND THE CERAMIC ASSEMBLAGEAuthor(s): R. J. RODDEN, GILLIAN PYKE, PARASKEVI YIOUNI and K. A. WARDLEReviewed work(s):Source: The British School at Athens. Supplementary Volumes, No. 25, NEA NIKOMEDEIA I:THE EXCAVATION OF AN EARLY NEOLITHIC VILLAGE IN NORTHERN GREECE 1961-1964. THEEXCAVATION AND THE CERAMIC ASSEMBLAGE (1996), pp. iii-xx, 1-212Published by: British School at AthensStable URL: http://www.jstor.org/stable/40856021 .Accessed: 21/08/2012 04:10

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

.

British School at Athens is collaborating with JSTOR to digitize, preserve and extend access to The BritishSchool at Athens. Supplementary Volumes.

http://www.jstor.org

ΝΕΑ NIKOMEDEIA I: THE EXCAVATION OF AN EARLY NEOLITHIC VILLAGE IN NORTHERN GREECE 1961-1964

DIRECTED BY R. J. RODDEN

THE EXCAVATION AND THE CERAMIC ASSEMBLAGE

GILLIAN PYKE AND PARASKEVI YIOUNI

EDITED BY K. A. WARDLE

SUPPLEMENTARY VOLUME NO. 25 Published by

THE BRITISH SCHOOL AT ATHENS

1996

Published and distributed by The British School at Athens

31-34 Gordon Square, London WC1H OPY

© The Managing Committee, The British School at Athens ISBN 0 904887 19 7

Printed at Alden Press Limited, Oxford and Northampton, Great Britain

Contents

List of Figures ix List of Tables xiii List of Plates xv List of Abbreviations xvii

Preface, by K. A. Wardle xix

Chapter ι Introduction and Acknowledgements, by R. J. Rodden ι 1.1 Preamble ι 1.2 Character and Location of the Site 2 1.3 Aims and Methods of the Excavations 5 Acknowledgements 6

Chapter 2 Stratigraphy, by Gillian Pyke 9 2.1 Introduction 9 2.2 Vertical Associations 16

2.2.1 Group 1 16 2.2.2 Group 2 18 2.2.3 Group 3 20 2.2.4 Group 4 22 2.2.5 Group 5 22 2.2.6 Group 6 24 2.2.7 Group 7 26 2.2.8 Group 8 27 2.2.9 Group 9 29

2.2.10 Summary 30 2.3 Horizontal Associations 30

2.3.1 Group 1 and Group 2 30 2.3.2 Group 3 30 2.3.3 Group 4 32 2.3.4 Group 5 32 2.3.5 Group 6 32 2.3.6 Group 7 33 2.3.7 Group 8 33 2.3.8 Group 9 34 2.3.9 Summary 34

Chapter 3 Structures and Architecture, by Gillian Pyke 39 3.1 Introduction 39 3.2 Buildings 39

3.2.1 Walls 39 3.2.2 Floors 41 3.2.3 The account of F6/1 41 3.2.4 Construction 42 3.2.5 Size and shape of buildings 44 3.2.6 Settlement 47

ν

vi CONTENTS

3.2.7 Summary 48 3.3 Features 49

3.3.1 Pits 49 3.3.2 Ovens and hearths 50 3.3.3 Storage and granaries 52 3.3.4 Ditches 52

Chapter 4 The Early Neolithic Pottery: Technology, by Paraskevi Yiouni 55 4.1 Introduction 55 4.? Methodology 58

4.2.1 Macroscopic examination of the material 59 4.2.2 Microscopic examination of thin sections 60

4.3 Manufacture of Vessels 60 4.3.1 Coiling 60 4.3.2 Pinching 61 4.3.3 Mat impressions 61 4.3.4 Scraping 62 4.3.5 Application of bases and lugs 62

4.4 Surface Coating 63 4.4.1 Red-brown slip 63 4.4.2 Pink coating 65

4.5 Firing 69 4.6 Pétrographie Examination 71

4.6.1 Introduction 71 4.6.2 Fabric A 72 4.6.3 Fabric B-i 73 4.6.4 Fabric B-2 73 4.6.5 Fabric C 73 4.6.6 Fabric D 74 4.6.7 Fabric Ε 75 4.6.8 Pétrographie analysis of painted pottery 75

4.7 Discussion of the Results of Pétrographie Analysis 76 4.7.1 Diversity of fabrics 76 4.7.2 Source location of non-plastic inclusions and clay 76 4.7.3 Addition of non-plastic inclusions 78 4.74 Refining of clay 78

Chapter 5 The Early Neolithic Pottery: Typology, by Paraskevi Yiouni 81 5.1 Introduction 81 5.2 Sampling of the Plain Pottery 82 5.3 Surface Treatment 84

5.3.1 Plain pottery 84 5.3.2 Decorated pottery 86

5.4 Analysis of the Shapes 92 5.4.1 Methodology 92 5.4.2 Neck jars 94 5.4.3 Hole-mouthed jars 95 5.4.4 Askoid vessels 95 5.4.5 Slightly closed vessels 96 5.4.6 Open vessels with rim angle from 85-900 96 5.4.7 Open vessels with rim angle from 60-840 97 5.4.8 Open vessels with rim angle smaller than 6o° 97

CONTENTS vii

5.4.9 Handles 97 5.4.10 Bases 99 5.4.11 Legs 100 5.4.12 Comparison of shapes of plain and decorated vessels 101 5.4.13 Comparison of fabric types and quality of paste 102

5.5 Intra-site Development 103 5.6 Conclusion 104

Chapter 6 The Early Neolithic Pottery: Functional Analysis, by Paraskevi Yiouni 181 6.1 Introduction 181 6.2 Estimation of the Number of Pots from Nea Nikomedeia 181

6.2.1 Recovery rate 184 6.2.2 Rate of production 184

6.3 Function of Vessels 186 6.3.1 Cooking vessels 186 6.3.2 Storage vessels 191

Appendices A The Radiocarbon dating of Nea Nikomedeia 195

Περίληψη (Summary in Greek), by Paraskevi Yiouni 197 Bibliography 209 Plates

List of Figures

1.1 Map of Macedonia showing the location of Nea Nikomedeia. 2 1.2 Plan of the mound showing the extent of excavation. 3 2.1 A preliminary plan of the 1963 excavations of the Early Neolithic building levels 10

at Nea Nikomedeia. 2.2 Plan showing the structural groups. 11 2.3 Plan of the area excavated in 1961: features recorded at the base of each spit. 12 2.4 Plan of the area excavated in 1963/64: features recorded at the base of spit 1. 13 2.5 Plan of the area excavated in 1963/64: features recorded at the base of spit 2. 14 2.6 Plan of the area excavated in 1963/64: features recorded at the base of spit 3 15

and below. 2.7 Plan of Group 1 by phase. 18 2.8 Plan of Group 2 by phase. 20 2.9 Plan of Group 3 by phase. 21

2.10 Plan of Group 4 by phase. 23 2.11 Plan of Group 5 by phase. 24 2.12 Plan of Group 6 by phase. 25 2.13 Plan of Group 7 by phase. 27 2.14 Plan of Group 8 by phase. 28 2.15 Plan of Group 9 by phase. 29 2.16 Grid systems used during the excavations at Nea Nikomedeia 31 2.17 Plan of structural phase 1 . 36 2.18 Plan of structural phase 2. 37 2.19 Plan of structural phase 3. 38 3.1 Reconstruction of a Nea Nikomedeia house. 43 3.2 Plan of the 1961 excavation showing the location of the two ovens. 50 3.3 Plan of the ditches. 53 4.1 Geological map of the area surrounding Nea Nikomedeia. 68 5.1 Plain pottery: relative frequency of burnished vessels and vessels with a 105

surface coating. 5.2 Relative frequency of the different types of decorated pottery. 105 5.3 Relative frequency of the different types of impressed decoration. 105 5.4 Neck jars. 106 5.5 Neck jars. 107 5.6 Neck jars. 108 5.7 Closed vessels with rim angle larger than 115o (hole-mouthed jars). 109 5.8 Closed vessels with rim angle larger than 115o (hole-mouthed jars). 110 5.9 Askoid vessels. 111

5.10 Slightly closed vessels with rim angle from 91-11 40. 112 5.1 1 Slightly closed vessels with rim angle from 91-11 40. 113 5.12 Slightly closed vessels with rim angle from 91-1 14o. 114 5.13 Slightly closed vessels with rim angle from 91-11 40. 115 5.14 Open vessels with rim angle from 85-900. 116 5.15 Open vessels with rim angle from 85-900. 117 5.16 Open vessels with rim angle from 60-840. 118 5.17 Open vessels with rim angle from 60-840. 119 5.18 Shallow vessels with rim angle from 60-840. 120

ix

χ LIST OF FIGURES

5.19 Open vessels with rim angle less than 6o°. 121 5.20 Open vessels with rim angle less than 6o°. 122 5.21 Ring based vessels. 123 5.22 Flat based vessels. 124 5.23 Round bottomed vessels (1-3); pedestal bases (4-6). 125 5.24 Legs from polypod vessels (i-6);unpierced lugs (7-9); 126

Z-shaped lugs (10-11); spindle-whorl (12); sherd discs (13-14). 5.25 Distribution of rim diameters for neck jars. 127 5.26 Distribution of rim diameters for closed vessels with rim angle 127

larger than 115o (hole-mouthed jars) 5.27 Distribution of rim diameters for closed vessels with rim angle from 91-11 40. 128 5.28 Distribution of rim diameters for open vessels with rim angle from 85-900. 128 5.29 Distribution of rim diameters for open vessels with rim angle from 60-840. 129 5.30 Distribution of rim diameters for open vessels with rim angle less than 6o°. 129 5.31 Distribution of rim diameters of painted vessels. 130 5.32 Distribution of rim diameters of impressed vessels. 130 5.33 Distribution of basal diameters of plain vessels. 131 5.34 Painted vessels with rim angle larger than 115o (1-12); Painted neck jar (13). 132 5-35 Slightly closed painted vessels with rim angle from 91-1 14o. 134 5.36 Slightly closed painted vessels with rim angle from 91-1 14o. 136 5-37 Slightly closed painted vessels with rim angle from 91-1 14o. 138 5.38 Slightly closed painted vessels with rim angle from 91-1 14o. 140 5.39 Slightly closed painted vessels with rim angle from 91-1 14o. 142 5.40 Slightly closed painted vessels with rim angle from 91-1 14o. 144 5.41 Slightly closed painted vessels with rim angle from 91-1 14o. 146 5.42 Slightly closed painted vessels with rim angle from 91-1 14o. 148 5.43 Open painted vessels with rim angle from 85-900. 150 5.44 Open painted vessels with rim angle from 85-900. 152 5.45 Open painted vessels with rim angle from 85-900. 154 5.46 Open painted vessels with rim angle from 85-900 (1-7) and from 60-840 (8-14). 156 5.47 Open painted vessels with rim angle from 60-840. 158 5.48 Open painted vessels with rim angle from 60-840. 160 5.49 Open painted vessels with rim angle from 60-840. 162 5.50 Open painted vessels with rim angle less than 6o°. 164 5.51 Painted body sherds. 166 5.52 Painted base sherds. 168 5.53 Applied decoration: sherds with raised bands. 170 5.54 Applied decoration: sherds with raised bands and blobs. 171 5.55 Impressed decoration: vessels decorated with finger pinchings (1-4); 172

finger nail impressions (5-8) and instruments (9-11). 5.56 Bases decorated with sharp pointed instrument. 173 5.57 Plain pottery: correlation of surface treatment and fabric types. 174 5.58 Plain pottery: correlation of surface treatment and paste quality. 175 5.59 Relative frequency of fabric types in plain, impressed and applied pottery. 176 5.60 Relative frequency of surface coating in plain, impressed and applied pottery. 177 5.61 Relative frequency of fine, medium and coarse vessels in plain, painted, 177

impressed and applied pottery. 5.62 Relative frequency of vessel shapes in plain, painted and impressed pottery. 178 5.63 Relative frequency of open, closed pots and neck jars in first and second 179

building phases. 5.64 Relative frequency of different vessel forms from the three building phases. 179

LIST OF FIGURES xi

5.65 Relative frequency of surface coating in the three building phases. 180 5.66 Relative frequency of fabric types in the three building phases. 180 6.1 Calculation of the surface area of a vessel with rim angle 85-900. 183 6.2 Correlation of fabric type and vessel form. 187 6.3 Correlation of quality of texture and vessel form. 188 6.4 Correlation of surface treatment and vessel form. 189 6.5 Correlation of surface treatment and vessel form. 190

List of Tables 2 .1 Depths of the spits for each group. 35 3.1 Building dimensions: phase 1. 45 3.2 Building dimensions: phase 2. 46 3.3 Building dimensions: phase 3. 47 4.1 List of thin sections examined with the polarizing microscope. 56 4.2 Chemical composition of slips and coatings of the Nea Nikomedeia vessels, 57

studied by SEM. 4.3 List of sherds refired at 8oo°C for 40 minutes. 58 4.4 Surface colour of the plain vessels from Nea Nikomedeia. 64 5.1 Sampled excavation squares from the main excavation grid of Nea Nikomedeia 84

(1963 season). 5.2 Relative frequency (%) of the lip-types represented on the Nea Nikomedeia plain 93

vessels. 6.1 Average surface area (cm8), volume (cm3), occurrence and the diameter/height 183

ratio of the different vessel forms from Nea Nikomedeia. 6.2 Thickness of occupation deposits of the Thessalian sites with early decorated 185

wares. 6.3 Measurement of the weight of 1 litre of cereals and pulses. 192

xiii

List of Plates ι (a) The plain of Macedonia showing the low mound of Nea Nikomedeia and the

surrounding foothills. (b) Foreman Yiannis Papadopoulos (left) with other members of the Greek workforce.

2 (a) At work clearing the site in 1963. (b) Excavation of the same area.

3 (a) & (b) Aerial views of the eastern half of the settlement mound, taken from the north west, 1963. Structures of Groups 1, 2, 5 and 6 are visible on the left; Group 4 is the excavated structure in the centre. (Photographs courtesy of the Greek Army Air Corps.)

4 (a) Group 2 Structure 3, squares A6-A8 and B6-B8. (b) Close-up of Group 2 Structure 3, squares A6-A8 and B6-B8.

5 (a) Group 4, from south east. (b) Group 4, from north east.

6 (a) Group 4 Structures 1 and 2, from east. (b) Group 5 showing one wall of Structure 1 (right), the corner of Structure 2 (bottom left)

and Late Neolithic Trench (across centre). 7 (a) Group 6 Structure 2.

[b) Group 9 in squares TB1-2 and Tgi-2 with Late Neolithic Trench in foreground. 8 {a) Miniature vessel with traces of pinching.

(b) Sherds showing the method of base attachment. 9 (a) Thrust lugs'.

(b) Red brown slip (crossed polars 170X). 10 (a) Pink coating (crossed polars 170X).

(b) Pink coating (crossed polars 42 ox). 11 (a) Fabric A (crossed polars 170X).

(b) Fabric A (crossed polars 170X). 12 {a) Fabric B-i (crossed polars 170X).

(b) Fabric B-s (crossed polars 170X). 13 [a) Fabric C (crossed polars 170X).

(b) Fabric D (crossed polars 170X). 14 (a) Fabric Ε (crossed polars 170X).

(b) Charred plant inclusions (crossed polars 170X). 15 (a) White on Red painted ware.

(b) Sherds impressed with finger tips. 16 {a) Sherds impressed with finger pinching and finger nails.

(b) Sherds impressed with finger pinching. 17 (a) Sherds impressed with finger pinching.

(b) Sherds impressed with finger nails.

XV

List of Abbreviations

AAA Αρχαιολογικά Ανάλεκτα εξ Αθηνών ΑΕ Αρχαιολογική Εφημερίς Am. Anthr. American Anthropologist Am. Antiq. American Antiquity BAR British Archaeological Reports BCH Bulletin de Correspondance Hellénique BEFG Bulletin of the Experimental Firing Group ΒΙΑ Bulletin of the Institute of Archaeology, University of London PPL Plane polarised light PPS Proceedings of the Prehistoric Society R/W Red on White Ware SEM Scanning electron microscope WA World Archaeology W/R White on Red Ware

xvii

Preface

The excavation of the Greek Early Neolithic site of Nea Nikomedeia, directed for the British School at Athens by Dr. R.J. (Bob) Rodden with the support of Professor Grahame Clark between 1961 and 1964, was an important landmark in the development of prehistoric archaeology in Greece. It was clearly one of the earliest farming settlements in Europe - although perhaps not as early as the first C14 determinations suggested - and one of the first extensive excavations of a settlement site in Greece. The collection of faunal remains and paJaeobotanical samples was integral to the project from its conception and the study of the changing landscape and environment an essential adjunct. The early years of the ensuing study were also exemplary - the pottery drawings, for example, included in this volume, were done by Dr. David Biernoff and accompanied by a detailed catalogue including Munsell colour numbers for the body and paint colour of each sherd. The preliminary reports in the Proceedings of the Prehistoric Society (1962, 28: 267-88), The Illustrated London News (April 11 1964, 564-7; April 18 1964, 604-7) ^d Scientific American (1965, 212 (4): 83-91) aroused immediate interest and continuing debate. Was it going to be possible to determine whether European farming developed in Europe or was imported from further east? How far did the settlement represent continuity from preceding Epi-palaeolithic activity in Greece? - one of Rodden's chief interests before the excavation was the early chipped stone industries of the area. It is no exaggeration to say that the full publication of this site was among one of the most eagerly awaited in Greek archaeology: appetites were further whetted by the excellent publications by van Zeist and Bottema on the plant remains and physical environment [Acta Botânica Neerlandica 1971, 20 (5): 524-38).

Sadly, Rodden's poor health prevented him concentrating on the essential detail of the study which must precede reasoned publication, although much meticulous work was completed. As the years passed, new discoveries and new techniques robbed Nea Nikomedeia and its excavator of the impact of novelty and to some extent of its significance, but he and his loyal colleagues never lost sight of the goal of full publication when this could be brought about. I am privileged to have helped in a small way to bring this publication to press - at Rodden's request and with the encouragement of Professor Robert Cook, Professor Colin Renfrew and Dr. Sebastian Payne - and especially of Judith Rodden. Late in 1991, I was asked to coordinate work on the excavation and finds record as a first step towards publication.

Funds granted in 1992-3 by the British Academy, the British School at Athens and University of Cambridge enabled us to support a postgraduate student for a year to complete the study of the stratigraphy and architecture at the site, as well as to prepare an interim finds catalogue as essential prerequisites to the specialist study of the different classes of object. We were especially fortunate that Gillian Pyke, who graduated in Ancient History and Archaeology at the University of Birmingham in July 1992, was able to undertake the task and to see it to a successful conclusion, as demonstrated by her contribution to this volume. Despite the difficulties of grappling with thirty year old records and understanding the stratigraphy of an excavation which she could see only as represented in the abundant photographic record, she has managed to produce a coherent

xix

χχ Κ. Α. WARDLE

picture on which the remaining studies can be based. In the meantime, Paraskevi Yiouni had been studying the Early Neolithic pottery for a Ph.D. dissertation, under the supervision of Dr. John Nandris at the Institute of Archaeology in London, which placed Nea Nikomedeia securely in its SE European context. These two elements, the stratigraphy and the Early Neolithic pottery, form the core of the present volume. Bob Rodden has, of course, provided considerable assistance in interpreting the records where there could be no substitute for autopsy, and has written an introduction setting out the background to the excavation and recording his own thanks to his excavation team. Especial thanks are due to Rayna Andrew for incorporating all the different elements and preparing the final version of the text camera-ready. We are also grateful to Nigel Dodds and Harry Buglass for help with plans and drawings, and to Nicola Wardle for typing up the pottery catalogue and preparing the decorated pottery figures for publication.

This volume has been prepared on an Escom PC using Aldus Pagemaker and Microsoft Word for Windows. The text is set in Baskerville and printed on a Hewlett Packard LaserJet using Postscript.

The second stage of the work has been to allocate groups of finds for specialist study and we are grateful to the British Academy, the British School at Athens and the Institute for Aegean Prehistory for funds granted in 1993 to support these aspects of the work. The second volume will contain the full finds catalogues and finds distribution prepared by Gillian Pyke, together with summary accounts of some categories of finds such as the ground stone and clay objects. Rosemary Payne will contribute a study of the numerous bone tools, one of the largest assemblages from SE Europe, and Antiklea Agrafioti will present her analysis of the chipped stone tools which are potentially so important for understanding the origins of the Neolithic in Greece. John Nandris is preparing his account of the varied range of human and animal figurines from the site and Paraskevi Yiouni will conclude the volume with a study of Nea Nikomedeia in its Greek and Balkan context. A consolidated index for both volumes will also appear in this second volume.

We expect that further studies will take place of the burials and human remains (Theya Molleson) and animal bones (Sebastian Payne) but these are not yet sufficiently advanced to allow the form of the eventual publication to be determined.

The staff of the 17th Ephoria of Prehistoric and Classical Antiquities, based in Pella, have given us every assistance and encouragement to complete the various aspects of the study. I can only hope that they feel their exemplary patience has been sufficiently rewarded in this publication which Bob Rodden inspired and to which so many colleagues have gladly devoted their efforts. I am particularly proud to have been able to assist: on my first extended study visit to Macedonia in 1969 Bob made me welcome in Verroia - as he did so many others - introduced me to Greek colleagues in the Ephoria, and set me on the right track.

K. A. W.

Chapter i

Introduction

(plates 1-3)

1.1 PREAMBLE

This volume describes the excavation of the Early Neolithic site of Nea Nikomedeia, located roughly between mainland Greece to the south and the Balkan heartland and the gateway to Europe to the north (fig. 1.1). In the context of recent Aegean and Balkan prehistoric studies, the excavation of the earliest levels at Nea Nikomedeia, consistently radiocarbon dated to the sixth millennium bc (uncalibrated), was particularly important for two specific reasons. First, because of the large area of the very early neolithic settlements exposed and in the house plans recovered (sometimes incompletely); and secondly in the exceptionally large quantities of pottery, animal bones and small finds in their spatial and stratigraphie context.

In successive years of excavation, more than 1 900 square metres of the Early Neolithic settlements were excavated. For much of this 'window into the prehistoric past' all remains were studied and recorded, thus allowing for the possible future testing of various sampling techniques and questioning the future viability of 'sondages' or trial trenches, the usual technique applied to multi-period prehistoric sites. At the time of the Nea Nikomedeia excavations, 1961-4, a question of sampling procedure had particular importance for the recognition of a Tre-pottery neolithic' at the base of neolithic mounds in Thessaly, Northern Greece.

Nea Nikomedeia is just such a multi-period neolithic settlement mound. Composed of successive layers of broken, deteriorated and weathered remains of clay and mud-walled structures, such sites in Macedonia are traditionally known as 'toumbas' (including Nea Nikomedeia, by which term the site is now known locally). In Thessaly to the south, such sites and features of the landscape are known as 'magoulas'; in the southern Balkans (Bulgaria) as 'mogila'; and in the Near East (where mud-brick is the traditional building material used), Hell' is the term generally applied. Here, in due course, we shall first consider the character and location of the site, its importance and finally, the aims of the excavations.

1

2 R. J. RODDEN

1 Nea Nikomedeia 2 Verroia 3 Thessaloniki

fig. ι . ι : Map of Macedonia showing the location of Nea Nikomedeia.

1.2 CHARACTER AND LOCATION OF THE SITE

The major dimensions of the Nea Nikomedeia settlement mound are approximately 22OX nom; the area exposed therefore represents roughly one-twelfth of the total area of settlement (fig. 1.2). The archaeological evidence indicates that the Early Neolithic settlement extended to the 'skirt' or edge of the mound. The mound at present rises a mere 2.50-1 .50 m above the surrounding plain or some 10.50 m above present sea level (plate 1 a). On site, the surrounding plain slopes from the south-west, from a surface c. 9.25 m above sea level to the north-west of the mound and to a contour of 8.00 m above sea level.1 There are two high points - summits is hardly the word - the northernmost serving as the triangulation point for what is both possibly the earliest detailed topographical

1 These figures are a compilation of a Greek Ministry of Agriculture contour map of the area and a detailed topographical survey of the mound undertaken by

R. J. Rodden and keyed into the closest national topographical triangulation reference point as of 1963.

INTRODUCTION 3

/^

fig. 1.2: Plan of the mound showing the extent of excavation.

map of the region and one of the latest prior to the drainage of the Plain of Macedon. The 'toumba' of Nea Nikomedeia is located approximately 10.5 km North 30o East of Verroia, the major town in the area; or East Longitude 22o 151 41.5", North Latitude 40o 361 14". The subject of the environment of the Plain of Macedon at the time of the Early Neolithic occupation of Nea Nikomedeia is a complex and problematic one, but obviously of critical importance in understanding events at the time of occupation. As both the plain and the site are in an area of tectonic instability as well as of varying sea-level and, in the long term, of climatic change, no geologist, geomorphologist, palaeobotanist or historian would, in this circumstance, take 'the present as the key to the past' as fixed and for granted.

This having been said, at this juncture we can only definitively portray Nea Nikomedeia in relation to the Plain of Macedon at the turn of the century, i.e. preceding the drainage of the plain in the 1920's where, at least, topography, lake and marshland are historically documentable. The reference point we have taken is a widely available, essentially military map of 1910, accurate in topographical detail and showing the precise and not so precise borders of the marshland surrounding Lake Yiannitson (Lake Ludias).2 In relation to the

2 Topographical detail of the Plain of Macedon can be

found in Greece III, G B Admiralty Handbook, Naval Intelligence Division, 1945, which contains maps showing the area prior to and after drainage.

4 R. J. RODDEN

local streams and regional rivers with large catchment areas in the Vermion range and bordering uplands, particularly noteworthy is the course of the River Tripotamas. This passes through Verroia from the uplands behind, then enters the plain and passes the site of Nea Nikomedeia, before flowing into the marshy and ill-defined south-western border of the lake, 'rich in wildfowl'. Nea Nikomedeia also lay close to the south-western border of the Lake Ludias marshland.5 However, while in the uplands the courses remained unchanged for the last few millennia, the behaviour of the same streams upon entering the plain and that of the lake/inlet of the sea would be much more sensitive to the relatively small changes in the land/sea levels in the Thermaic Gulf, not to mention minor tectonic movement beneath the plain itself south of Yiannitsa and Pella.

During the 1920's, major changes in the physical appearance of the plain, in the number and ethnic composition of its inhabitants and in the village and other 'place-names' took place. First to happen was the influx into Greece of hundreds and thousands of Greek- speaking orthodox Christians in 1922-23 in one of the major repatriations of modern history. On the plain of Macedon, this involved (in part) the resettlement of Greeks native to the Pontos - it is said the descendants of ancient Greek colonies on the southern and eastern Turkish/Russian shores of the Black Sea reputed to reach back to Classical times. We named the site after the nearby village, now known as Nea Nikomedeia;4 the earliest maps show that both the mound and the village were referred to as Branyiates.5

After the influx of population, an enormous project of river canalization and land drainage of the plain was undertaken, reported in The Illustrated London News of 1925. The waters of the Moglenitsa and other rivers of the Pelles region which previously were large contributors to Lake Ludias, were diverted into the major diversionary canal on the plain; it passes along the western edge of the plain to join the Haliakmon River on the southernmost border of the plain and thence into the Thermaic Gulf or Gulf of Thessaloniki. At what must have been roughly the same time, what remained of the lake and streams draining the surrounding plain were channelled into a dendritic pattern of smaller canals used both for drainage and irrigation before entering the gulf further to the north of the Haliakmon by route of the former River Ludias.

The mound, lying several metres above the level of the surrounding plain, was the obvious choice of road-fill for causeways leading into the plain and behind the site onto higher ground. To judge from our excavations this was done by a forked heavy earth-moving bulldozer, clearing material down to - but not below - the level of the surrounding plain; and fortunately, not deep enough to remove all of the Early Neolithic deposits.

At the time, the finds came to the attention of Professor Petsas, at the Ephoria of Antiquities for Western Macedonia, who appreciated their possible significance for Greek prehistory and stopped further destruction of the mound. In the summer of i960, he directed the late Dr. David Clarke and myself to the site, as one deserving further investigation. Thorough examination of the mound indicated the great promise of what came to be known as the

5 That the Early Neolithic site lay in close proximity to marshlands is also reflected in the extensive importance of reeds in building construction.

Nikomedeia was an important city in northwesternmost Turkey: a contender for the capital of Byzantium under Constantine and a major cross-roads for

the Roman road system in Asia Minor. Branyiates, Nea Nikomedeia's previous name, is

possibly Bulgarian (after the verb 'to defend*) in origin: no surprise, here, for Bulgarian cultivators of Turkish chiftliks being the rule on the 'lower plain' of Macedonia as late as the nineteenth century.

INTRODUCTION 5

'lower site' for exposing over a large area an occupation belonging to the very early neolithic of mainland Greece.

1.3 AIMS AND METHODS OF THE EXCAVATIONS

The excavations of Nea Nikomedeia were undertaken during the summers and autumns of 1961, 1963 (the main season) and 1964 with the invaluable assistance mainly of students - most of whom came from the universities of Cambridge and Harvard - and of an enthusiastic and soon rapidly experienced team of Greek workmen from the village of Nea Nikomedeia, headed by Mr. Yiannis Papadopoulos (plate 1 b and plate 2).

From the outset the main thrust of the excavations was on the lower site and towards the exposure of the earlier neolithic structural remains, whilst keeping in mind the distribution of contemporary pottery and animal bones and the accurate location of small finds. Towards these ends I think - in the main - we were successful.

The method employed was that used by Professors Waterbolk and Modderman in the excavation of the Early Neolithic Bandkeramik sites in the Netherlands during the late 1950's and using the same equipment. Once the disturbed soil or 'plough'6 had been removed, at Nea Nikomedeia's lower site a comparatively large area, generally 32 or 64 m2, would be excavated to one spit's thickness (approximately 20 cm) at a time, followed by scraping or shaving the excavated surface with Dutch shovels specially made for the purpose. Properly used these were designed to reveal discolorations - as at Nea Nikomedeia wall-slots, post-holes and the outlines of refuse pits. Some structural remains did appear during excavation: to be precise, the two structures belonging to the original settlement that had been destroyed by fire. Other structural remains were generally discovered by this scraping technique, following excavation of a second or third spit, scraping and looking for patterns of discoloration in an otherwise uniformly fawn-coloured culture fill. By excavating and scraping large areas - viewed from a 6-7 m high ladder tower which could be moved from area to area as required - it was possible from above to see emerging patterns of postholes and wallslots which would be quite indecipherable at ground level, or if smaller areas were exposed (plate 3). Over much of the lower site excavated in late 1963-early 1964 (i.e. area TX and TY 1-12) there was a scant single spit's thickness of deposit with which to work, but the contrast between the structural features and the virgin soil was clearly defined. Closer in towards the centre of the mound, deposits reached 70 cm in thickness and the task of differentiating features from fill in the upper spits was difficult indeed: interpretation could depend upon the time of day, whether it was cloudy or clear and how long since the last rain or latest scraping.

The recording system used during the 1 963 and 1 964 seasons was based upon a grid of 4 χ 4 m squares. Several, or more likely four, squares would be excavated and scraped simultaneously and then mapped. The pottery and animal bone were collected by these units and cleaned and the pottery classified in the laboratory as it arrived from the excavations. The pottery from adjacent units was laid out for classification - according to a standard reference system - to look for joins, particularly from squares next to one another and thereby providing yet another means for correlating contemporary deposits.

6 Since bulldozing, the lower site had been 'shallow-ploughed* by draught animal methods.

6 R. J. RODDEN

Alas, such joins were few and far between.7 'Small finds' were classified in the field, located within an accuracy of ι ο cm and entered into the field notebooks before being sent to the laboratory.

Meaningful vertical sections were few and far between and only those which had been left to weather for three weeks or longer ordinarily revealed natural stratigraphie differences within the culture fill. Those studied and drawn were of baulks defining major intervals in the site grid and which had been left standing. Excavation of individual squares always kept a vertical face where any variation from the normal uniform fill was noted and a section drawn in the site-book for the square. In the last analysis, it was what might be called 'horizontal stratigraphy' - the demonstrable intersection of one structure by another - which enabled successive building periods to be distinguished one from another, together with the direction or orientation of buildings and the materials and methods used in their construction.

By contrast, the 1 96 1 excavations of the lower site were less effective and the exposure and mapping of structural features made more difficult, if only because a grid system of 2 x 2 m squares was used and a much smaller area of the earlier neolithic deposits excavated at one time. In sum, although a large area was excavated, only in the smaller area excavated down to virgin soil were structural features clearly defined and mapped, as reported in Rodden 1962. As in 1963-64, the pottery and animal bones from the 1961 excavations were identified by square and spit. 'Small finds' however were not located so precisely as in the 1963 and 1964 seasons and the closest find spot was by square.

Immediately adjacent to the 1961 excavations of the lower site was Area L, a comparatively small section through all of the deposits of the mound, also excavated in 1961 . Area L measured only 5 χ ίο m in area and was excavated (unnecessarily) in 20 cm spits once the rather thick overburden of disturbed Late Neolithic deposits was removed.8 The pottery and small finds from Area L have not been studied in detail. But as in later seasons, a documented record with drawings of the 'small finds' was kept by the site laboratory. The stratigraphie studies that have been made indicate an earlier neolithic deposit, perhaps as much as 1.50 m in thickness, overlain by the deposits of a much later Late Neolithic occupation, the major structural feature of which was a series of concentric 'defensive' ditches, the innermost of which cut through the earlier neolithic deposits in the Area L sections. It, and the other Late Neolithic ditches at the site, were picked up as clearly intrusive features during the excavations of the lower site.

ACKNOWLEDGEMENTS

It has been many years since the excavation of the Early Neolithic settlement site of Nea Nikomedeia began. Despite the intervening years it is with great gratitude that acknowledgment is made to those individuals and institutions who made the excavations such a success and still of contemporary relevance to Greek and Balkan Early Neolithic studies.

7 The smaller number of such joins also suggests that 'treadage' and 'scuffage* at the time of occupation played little role in the distribution of objects at Nea Nikomedeia.

8 This disturbance was due to deep, heavy tractor-drawn ploughing of the top of the mound.

INTRODUCTION η

Because of their importance to the success, indeed the very existence of the excavations, it is a particular pleasure at the outset to acknowledge those people whose assistance was crucial from the beginning to the end of the dig. Especial mention must be made of Professor Sir Grahame Clark for his intellectual guidance and advice; of the late Dr. Hugh Hencken, without whose help the major financial support of the United States National Science Foundation would not have been obtained; of Professor Photios Petsas, at the time representing the Greek Archaeological Service in Western Macedonia, whose day-to-day and year-to-year help extended far beyond his formal duties; and finally, but not least, of my wife, Judith Rodden, who took on the sometimes unexciting but vitally important task of directing the excavation laboratory.

International financial support made the Nea Nikomedeia project possible. The excavations of 1 96 1 were made possible by 'seed' grants from the Crowther-Benyon Fund, Faculty of Archaeology and Anthropology, University of Cambridge, the British Academy and the Wenner-Gren Foundation for Anthropological Research. The major excavation and laboratory seasons were in 1963 and 1964. This work, including the laboratory study of the finds during 1 965, was sponsored by grants from the United States National Science Foundation, awarded to and administered by Harvard University. Subsequent grants for the completion of the study and recording of all categories of finds from the Nea Nikomedeia excavations were made by the British Academy, Ford Foundation Traineeships and the University of California at Berkeley.

The excavation seasons at Nea Nikomedeia were undertaken under the auspices of the British School at Athens. The work would not have been possible without the administrative support of the School and its Directors throughout the duration of the excavations and later study periods. The Greek Archaeological Service provided unfailing support and guidance. I prefer to think of the very great help afforded by both institutions in terms of individuals and to acknowledge the late Dr. J. Papadimitriou and, locally in Verroia, Professor Photios Petsas, Dr. Katie Romiopoulou, Dr. Yiannis Touratsoglou and the late Dr. Maria Siganidou and, on behalf of the British School, the Director, Mr. Sinclair Hood, the Assistant Director, the late Dr. Philip Sherrard and the subsequent Director, Mr. A. H. S. Megaw.

Professor Sir Grahame and Lady Clark and the late Mr. R. F. W. ('Squire') Hutchinson took a most active and valuable part in the 1961 and 1963 excavation seasons. The main part of assisting in the archaeological excavations, on site or in the laboratory, was undertaken by students, mainly from Harvard and Cambridge Universities, the efforts of all of whom contributed a very great deal to the success of the project. The names of those who took part in the several excavation seasons are listed below.

1961 1963 Antoniou, Savvas J. (Peterhouse) Campbell, Cathleen M. (Radcliffe) Blackman, David J. (Trinity) Chapman, Angela M. (New Hall) Brockbank, Rona (Newnham) Choderow, Nancy (Radcliffe) Clark, Margaret Clark, Margaret Higham, Charles F. W. (St. Catherine's) Clegg, Ian M. I. (Queen's) Movius, Geoffrey H. (Harvard) Cohen, Mark N. (Harvard) Nandris, John G. (Trinity) Cooper, J. Christopher (Clare) Oakley, Robin E. (King's) Cramm, C. Leslie (Peterhouse) Renfrew, Colin (St. John's) Fleming, Andrew M. (Christ's) Whinney, Sybil M. (New Hall) Jones, Margaret

8 R. J. RODDEN

1963 continued Molleson, Theya (Edinburgh) Jones, Tom Poole, Judith G. (St. Martin's School of Art) Masters, Lionel J. (St. John's) - deceased Moore, C. Nicholas (Peterhouse) Ronhovde, Nora M. (Radcliffe) Murray, Georgina (Edinburgh) Wilson, Aimée M. (Radcliffe) Petsas, Angeliki (Thessaloniki) Yealland, Dr. Susan Renfrew, Colin (St. John's) von Strachwitz, Count Rupert 1965-1969 (laboratory work only) White, J. Peter (St. John's) Mock, Renita E. (Berkeley and UCLA)

Ridley, The Hon. Cressida (University of 1 964 London Institute of Archaeology) Britton, Jane S. (Radcliffe) - deceased. Villa, Paola (Berkeley) McNeil, David O. (Antioch College)

Mr. Yiannis Papadopoulos of the village of Nea Nikomedeia, who acted as Foreman, deserves particular mention for his ability to co-ordinate the workforce and for his quick grasp of the requirements of the excavation. He headed a team of up to twenty men from the village, whose knowledge as farmers of the local soil conditions quickly proved invaluable in an archaeological context.

This is also an unique opportunity to thank the 22nd Squadron of the Greek Army Air Corps, the senior officers and the pilots, Capt. Christos Skarzas, Capt. Nikolas Tountas and Capt. Andreas Kolliopoulos, for their assistance in taking aerial photographs of the excavations at a crucial stage, namely, on the 22nd and 24th August 1963.

Valuable medical supplies were donated for the 1961 season by Boots Pure Drug Co. Ltd., the British Drug Houses Ltd., CIVA Laboratories Ltd., Burroughs Welcome & Co., Johnson & Johnson (Gt. Britain) Ltd., Smith & Nephew Ltd. and Parke, David & Co. Dr. L. Hawtrey May, University of Cambridge Health Service Officer, assisted in the preparation of a list of necessary medical supplies.

The personal responsibility for important, but unpublished, work has been a considerable burden over a number of years. It was in this context that I recently approached close friends with a view to handing over publication to an ad hoc committee comprising Professor Lord Renfrew, Professor R M. Cook and Dr. Sebastian Payne. To my immense satisfaction Dr. Kenneth Wardle accepted the responsibility of o ver seeing publication: an archaeologist who probably knows as much about prehistoric Macedonia as W. A. Heurtley did in his own time several generations ago.

R. J. R.

Chapter 2 Stratigraphy

(plates 4-7)

2.1 INTRODUCTION

In view of the early date of the site and its importance for the history of human settlement in Greece an estimate of the duration of the site's occupation and the continuity of occupation are vital considerations. The establishment of a stratigraphie sequence was considered the first step in the present study of the site.

A preliminary study of the stratigraphy was published by Rodden in 1 964 in relation to the 1963 excavation (Rodden 1964 c). This concentrated on the structures which Rodden assigned to the Early Neolithic. The accompanying plan of the Early Neolithic structures (fig. 2.1) shows two building phases, each divided into two subphases. Features considered to be Late Neolithic were omitted from this plan.

In order to clarify further this interpretation of the stratigraphy it was decided to go back to the original data and produce an independent account of the phasing.

The accounts and sketch plans in the excavation notebooks, the site plans and the overall site plan showed that the foundation trenches could be interpreted as structures or parts of structures and that some of the foundation trenches cut through others. The foundation trenches formed clusters, here termed as groups, where structures appear to have been built in succession (fig. 2.2).

The first task was to establish the sequence of structures within each group. The plan of each structure within a group was defined by examining the site plans and the account in the excavation notebook for each square. The site plan which I prepared to show the features found in each spit (ç^fig. 2.3, the 1961 excavation; figs. 2.4, 2.5 and 2.6 which show the features in each spit separately for 1963/64) was used to find out in which spits the foundation trenches were observed. Where the same foundation trench occurred in several spits it was necessary to determine, where possible, whether the spits were dug at different levels, the foundation trench was dug to different levels, or there was a slope in the original topography. The vertical relationship between the structures of a group was investigated by considering the instances in which one foundation trench cut through another.

The second task was to determine the horizontal relationship of the structures across the site. In squares in which foundation trenches were found the level of each spit was noted. This was done using the section drawings in the excavation notebooks with the depths recorded in relationship to the site datum (6.85 m above sea level).

9

ίο GILLIAN PYKE

Π~Ί_ ^^■■■'"vL^ # ΝΕΑ NIKOMEDEIA 1963

/Ni · f^f^ö I I ? ■ ■ r

/^^^ "^^rfll -^^C' ■ istBuiidingperiod

fig. 2.1. A preliminary plan of the 1963 excavations of the Early Neolithic building levels at Nea Nikomedeia.1

1 After Rodden 1964 c, 1 15 fig. 2. Storage and refuse pits and intrusive Late Neolithic features have been

omitted for the sake of clarity.

STRATIGRAPHY 1 1

Ou

' ν ff ·°Α

V >Α ο-

ι ^ // γ °

(Λ

Ι "Β ι CO

•4-»

f Ο CO

s ε ci Ο

12 GILLIAN PYKE

0 12 3 4 5m / I A l ι ι ι ι l r /!l

fig. 2.3. Plan of the area excavated in 1961: features recorded at the base of each spit.2

The section drawings were also used to check that the plans of the structures were credible, in cases where their foundation trenches appeared in different spits in different squares. The depths at which the spits were dug and their thickness sometimes explained the location of a foundation trench in different spits in different squares; the foundation trench being at a consistent level, but the spits being excavated at different levels.

This procedure enabled the comparison of the depths of the spits in which the earliest structures of each group were found. It was hoped that the earliest structures would appear at a similar level, which would support their belonging to the same building phase. The same was done with the later structures. The groups which consisted of less than three structures were investigated to see if it was possible to identify which phase was missing, and whether it was the same phase in each case.

2 After Rodden 1962, facing 272, Plate XXXVIII.

STRATIGRAPHY 13

,^. Γη ^ '·; r^ ο .. (| :_

S

Ό

1 s 3 ci

CO

ιΗ

#g

I 8 I S ι ci Ü

i4 GILLIAN PYKE

' ' * °* * ι - ι °"

vus .'*· ,o ~Ja ι

I If rC vi rJl· ' Vi

ci

■&

'S

J ■s

α> "S ο u s CO

S

«S '-Φ

co %

fH

#c

Β cd

i 'S 'S 1 ei

i

STRATIGRAPHY 15

UmfTïïïiTÏÏïnT]

II '' ^1

^2

% CO

Í Xi

*

T3

Ή 8 Si CO

S

'-# CO

CD σ)

#g

aj U

α*

S ■a ο

S s oi Ü

i6 GILLIAN PYKE

The main problem encountered in attempting to establish the relationships between the groups was that full sections across the site were not drawn during excavation because the excavators believed they were dealing with a single phase site, and test pits dug in each square gave no indication to the contrary.5 Therefore the only sections which were available were those which were sketched in the excavation notebooks by the excavators of each square, several of which had not been completed. In the areas M5-8, L5-8, K5-8, J5-8, H5-8 and TX1-6, TY1-6 only a single spit was dug, therefore it is not possible to relate the structures within these groups with absolute certainty.

2.2 VERTICAL ASSOCIATIONS

The vertical association of the structures within each individual group established the dimensions and shape of each structure. It is clear that during the excavation there was uncertainty over the delineation of the house plans in some cases. This resulted in the site plans being drawn with this in mind and the corners of the structures and the overlaps of walls being left ambiguous.

The information in the site plans was combined with the written account and sketch plans in the excavation notebooks in order to assemble the fullest possible account of the structural sequence within each group.

2.2.1 Group 1 This group of buildings (fig. 2.7) appears in squares B8-9, C7-10, D7-10 and E7-8. There are three overlying structures in this area. Structure 1 is square with three complete walls and part of the fourth. Its dimensions are approximately 7.40 x 6.50 m. Structure 2 is probably rectangular with a single complete wall and part of the adjoining walls preserved and is at least 7.40 x 6.80 m. Structure 3 is rectangular with one complete wall and part of the two adjoining walls preserved. The dimensions of this building are 8.80 x 7.70 m.

structure 1 The earliest structure is that of E7-8, D7-9, C7-9. The excavation notebook gives no account of the structural remains for squares C7 and C9. Wall A in C8/2 is mentioned as being cleaned down to base. This implies that it is earlier than the other wall which appears in this square. In square D7/2, Feature A is described as a wall with burning. Shortly after this feature is mentioned, the final 5 cm of the spit is dug, suggesting that the wall appears fairly low in spit 2. In square D8, wall A appears at the base of spit 2. It was sectioned, and this section is labelled in the excavation notebook as 'spit 3', lending weight to the interpretation that the structure is the earliest. In square D9 there is a definite distinction between the lower wall', belonging to Structure 1 and the 'upper wall', which runs diagonally across it and belongs to a later structure. The excavation notebook account also states that the 'lower wall' continues and joins Pit C in square C9. The sketch plan for E8/2 in the excavation notebook shows that the walls of Structure 1 belong to this spit rather than spit 1 , which is more consistent with the general appearance of the structure in the other squares.

Rodden 1993 pers. comm.

STRATIGRAPHY 17

It is also noted in the excavation notebook that these walls would line up with those of D8/ 2. A hard white layer c. 5 cm thick was observed within the angle of the walls, which may be the floor.

STRUCTURE 2 The second structure of this group lies in squares D8-10, C8-10, B8-9. The excavation notebook account of the walls in D9/2 show that the wall of Structure 2 was named the 'upper wall'. It was interpreted at the time of excavation as being higher in the spit (and was therefore later) than the wall of Structure 1 , which was labelled the 'lower wall'. The consistent appearance of the 'upper wall' in both spit 1 and spit 2 in the site plans of squares Dg and D 10 supports the interpretation of Structure 2 as later than Structure 1, although the other two walls are firmly in spit 2. In the excavation notebook entry for square D 10 there is no description of structures but the foundation trench does appear on a sketch of spit 2. Similarly, the 'upper wall' of square D9 is not shown in the sketch of spit 1 in the excavation notebook but does appear in the sketch of spit 2. There is no structural account for squares Cio, B8 and Bg in the excavation notebook.

In comparing the plans of Structures 1 and 2 it is clear that they have very different orientations, and that Structure 2 is significantly larger than Structure 1, although exactly how much larger is impossible to tell because the south east wall probably lies in the unexcavated area.

STRUCTURE 3 The latest structure of this group is found in squares D8-g, C8-g and B8-g. There is no structural account in the excavation notebook for squares C8, D8 and Cg. In the sketch of B8/2 the wall of Structure 3 is indicated as having a hard white fill, while that of Structure 2 had the usual culture fill. Similarly, in the sketch of Bg/i, the fill of the foundation trench of Structure 3 is described as consisting of hard yellow-white material, while that of a nearby trench was a softer grey material. This supports the interpretation that both these walls belong to the same structure. In the sketch of Dg/2 the wall of Structure 3 is indicated as overlying that of Structure 2 ('upper wall'), although both appear in the same spit. The inconsistency of the appearance of the foundation trench running across squares D8, Dg, Cg and C10 in both spits 1 and 2 can be partly explained by the section D7~Dg in which the top of spit 1 is at the same level, but the base of spit 1 in square Dg is about 8 cm lower than that of square D 8, so that the foundation trench could appear towards the base of spit 1 in square Dg but at the top of spit 2 in square D8. The same is true of squares B8 and Bg, and is confirmed by the section drawings from the excavation notebook.

Only three walls of this structure can safely be identified, but it may be that the second short wall was not noted during excavation or was obscured among the various pits which appear at that end of the structure. It is possible that these pits represent the final wall, which could have been simply a line of posts supporting a light wall. Similarly the wall may never have existed and the structure could have had one open end. It seems likely that as the two parallel walls are lost at the same point, this marks the end of these walls.

This interpretation (fig. 2.7) agrees in general terms with Rodden's plan (fig. 2.1 above) in which he identifies several building phases and assigns the three structures in this group in a comparable way.

i8 GILLIAN PYKE

fig. 2.7. Plan of Group 1 by phase.

I N

E7 M Wall A E8

MV I Feature A/1 λΛ'1 I V^ Upper Wall

λΛ/ Upper Wall λΛ'1

ti D7 D8 / Π ^^^('y Ρ9! '' D1°

£L^ C8 ^^ 1 1 {jf' CW

[^X] Structure 1

0 1 2 3 4 5m | I Structure 2

^H Structure 3

2.2.2 Group 2 This group covers squares B6-8, A5-8 and O6-7. There appear to be at least two structures (fig. 2.8), but the site plans are confused as to the sequence and layout of these buildings. Structure 1 is a rectangular building, all four walls of which are preserved. It measures approximately 6.48 x 6.80 m. Structure 2 is of a similar size and shape to Structure 1, with all four walls preserved. Its dimensions are 5.92 x 7. 1 2 m. Structure 3 (plate 4) is probably square and has two adjoining walls preserved, one with a possible entrance. A short length of a third wall is preserved. It is possible that the remainder of this wall and the adjoining wall, of which no foundation trench was found, were composed of a light stake barrier. If

STRATIGRAPHY 19

the post holes in squares A6, B6 and B7 are taken to represent a less substantial wall, the structure measures approximately 4.20 x 4.70 m.

In his interim report (Rodden 1964 c) Rodden interprets the group in terms of a single structure relating to the Early Neolithic, clearly regarding Structure 3 as a feature of the Late Neolithic and therefore omitting it. The remaining complex of foundation trenches he interprets as a single structure unevenly divided into a large square central compartment flanked by two smaller rectangular compartments. This complex form of structure is also seen in group 4/2.

The ground plans of Structures 1 and 2 were established by examining the areas of burning in spit 1, as shown incompletely on the site plans. This survey was originally undertaken to investigate the nature of the destruction of the structures and the possible location of ovens and hearths, but was abandoned when it was discovered that no consistent key was used on the site plans.

When the areas of burning which could be safely identified using the site plans and the excavation notebook entries were drawn out on a plan of the site, they showed the plan of the later of the two overlying structures, which was apparently destroyed by fire.

STRUCTURE 1 The earliest structure was located in squares A6-7 and B6-7. The west wall was not well defined by excavation, and is very narrow and fragmentary. The east wall is also fragmentary with narrow sections in the northern part. There was evidently confusion over the relationship between this wall and that of Structure 3 when the plans were drawn, as the earlier wall is shown as overlying the later one.

structure 2 The second structure appeared slightly farther east in squares A6-8, B6-8 and O7. It is similar in size, shape and orientation to the earlier structure. Parts of the north and south walls lie over those of Structure 1, and the plans show that it was difficult to distinguish between the two during excavation.

STRUCTURE 3 (PLATE 4) This structure was located in squares A7-8 and O6-7. Two and a half walls can clearly be identified, one with a possible doorway marked by a pair of posts. Half of the wall opposite this is also preserved, and it seems possible that the angle of the wall changes at the point where the foundation trench is lost, to become a light form of post wall, which also forms the adjacent wall. This is represented on the ground plan as post/stake holes. The wall which is wholly composed of a post wall, is opposite the only solid wall of the structure. The site plans indicate that the structure was found in spit 1 .

2O GILLIAN PYKE

/[ B6 / · B7 /S')y £> / B8 -I ^

( / * · μ Γ Γ'Ν ^■s f^^^^ m k'l Staicture 1

Structure 2

0 1 2 3 4 5m H Structure 3 ι ι . ι | j | Oo | 07 ^BB

fig. 2.8. Plan of Group 2 by phase.

2.2.3 Group 3 This group (fig. 2.9) consists of a pair of structures in squares E00-E2, D00-D2 and Coo- C2. Neither of these structures have been completely revealed by excavation. Both seem to have been located in spits 2/3. Structure 1 may be rectangular with sides measuring 9.00 m and 6.80 m, if the tentative corner in square Coo is correct. Its southern wall lies within the unexcavated area. Structure 2 is rectangular. From the two virtually complete walls the dimensions of the building are 10.90 x 7.50 m.

STRUCTURE 1 This structure appears in squares Do, Di, Coo and possibly Ci and C2. From the site plans this structure appears to be lying under Structure 2. Although there are no sketches or structural accounts for squares Coo, C 1 and D 1 regarding this structure, the information in the excavation notebooks for squares Do and Coo suggests that this structure appears in spit 3. In the account of D0/3 there is a clear sketch of the angle of the wall of Structure 1, and mention is made of the clearing of wall 1 in spit 3. In C1/3 there is a description and sketch of Wall A. The wall is described as 'plaster/clay lined', and this was shown by a section of the foundation trench taken on the C0-C1 line.

The structure consists of two definite walls, which join in the corner located in square Do. Another possible corner is tentatively marked at the end of the wall passing through squares Do, Co and Coo, which leads off into an area which was not excavated. It seems feasible that the wall located in C 1-2 is the one running parallel to wall Do-Coo, the only drawback being that it appears excessively long. It is unfortunate that the squares Ao-Aoo and Bo-Boo were not excavated as these might have clarified the problem. It is clear that some explanation is necessary for the presence of this foundation trench as it appears to be

STRATIGRAPHY 21

at the same general level as the definite walls of Structure 1. Rodden omits this foundation trench from his plan (fig. 2.1).

STRUCTURE 2 This is located in squares E00-E2, D 1-2 and C00-C1. There are three walls fairly clearly preserved, which the site plans indicate are overlying Structure 1 . The final wall is probably located in the unexcavated area bordering Coo-Eoo.

There are no structural accounts or sketches for squares Di, D2, Ei or E2. In the sketch of C1/2, Wall A of Structure 2 is shown as having a whitish clay fill. The sketch of E0/2 shows a wall which is said to be a continuation of that of square Eoo and is recorded as having been cleared. There is no account of Eoo but the wall of Structure 2 did appear on the sketch of spit 2. Structure 2 appears in the sketch of the base of spit 2, with the added note that the foundation trench traces line up with those of Ci, but were too faint to be indicated on the site plan. For square Coo, the foundation trench of Structure 2 appear on the sketch of spit 1. It appears that this structure appears mainly in spit 2, although it also appears in spit 1 towards the eastern side of the site.

The site plans show that the wall of Structure 2 which lies in Coo, Co and Ci appears in several spits. The section drawings in the excavation notebook show that the structure could appear in the base of square C00/1 and the top of square C0/2 and, allowing for some inaccuracy in the drawing and some tilt of the building, could also appear in Co/ 2 and C1/3. An estimate of + 40 cm at the east end and + 30-35 cm at the west end would explain this apparent disparity.

Γ td EO ^^>v E2_ / / ^^

/ Ρ°° 'õ DO ^D.Wy D2_

/ / ^f^Wa"A ^^ ^^ / //if3 / /'/WaUA / Ay ^^ ^^ / / /'/WaUA / COO Ay ^- I CO^xl ̂ r^ L PJ[ Ç2_ °L- * I * I 1 I 1

4 1 -j

*" Π fl F33 k ' 1

Structure. * * 1 I I 1 1 -j Π k ' 1

f'? QH Structure 2

fig. 2.9. Plan of Group 3 by phase.

22 GILLIAN PYKE

2.2.4 Group 4 This group (fig. 2.10, plates 5, 6 a) forms the largest area of the site, covering squares H6, G6, F5-8, E4-8, D4-8 and C4-8. The main problem with the interpretation of this group is that the foundation trench of Structures 1 and 2 were all found in the same spit, spit 2. This means that the site plans do not give a clear indication of whether any phasing was clear at the time of excavation. The interpretation given by the excavator appears to be a logical view of the evidence that remains (Rodden 1964 c).

Structure 1 is a large square building which measures 1 1.80 x 13.60 m. Structure 2 is a large rectangular building with two internal east-west partitions. Its dimensions are at least 10.90 x 9.10 m. Very little of Structure 3 was uncovered during the excavation, so its shape is unknown, and it is not possible to estimate its size. The corner of a structure was also found in squares D3-4/1, which might suggest a possible fourth building phase.

STRUCTURE 1 This structure is located in squares F7-8, E5 and E8, D5-8 and C5-7. The foundation trenches are difficult to trace but form a large squarish building whose northwestern corner lies in the unexcavated area.

STRUCTURE 2 This structure is located in squares F6-7, E4 and E7 and D4-7. It partly overlies the earlier structure. This structure was divided into three compartments by two cross walls running east-west. The jumbled southern foundation trench may be the result of its having been reçut.

STRUCTURE 3 (PLATE 6 d) This structure appears in squares H6, G6 and F5. It overlies both Structure 1 and Structure 2 and appears in spit 1, therefore leading to the conclusion that it is the latest construction of this group. One wall has been preserved to its full extent, with one corner and a short stretch of the perpendicular wall visible in square F5. The opposite corner is lost due to the presence of the Late Neolithic ditch which cuts through the structure at that point in squares H6-G6.

The section of building in squares D4-5 have not been considered by Rodden but may belong to the same phase as the corner which is visible in square F5, as it overlies both of the main structures. It is also possible that this structure represents a fourth building phase. Very little of the structure has been preserved so it is impossible to judge its size or shape.

2.2.5 Group 5 This group (fig. 2.1 1, plate 6 b) lies in squares H6-8 and G6-8. It consists of two structures, which each appear to have two compartments. Structure 1 is a rectangular structure with a north-south partition. It is at least 3.80 x 6.80 m. Structure 2 is represented by only a corner of the building, so it is only possible to determine that it was partitioned. The interpretation given below agrees with that of Rodden who assigns both structures to the second building phase (Rodden 1964 c).

STRATIGRAPHY 23

__ F4 F5Tr~YS ' F6 Λ F7_

M / E5 I

■ Qr* E3 fl E4 E5 / E6 E7_

D3 D^^^^^^^

^^ ^ D5 / D6 M

D7

Ç3 C4 1 C5 C6 C7_

hs x j Structure 1

0 1 2 3 4 5m I j III; | I | Structure 2

^^B Structure 3

fig. 2.1O. Plan of Group 4 by phase.

24 GILLIAN PYKE

STRUCTURE 1 This is located in squares H6-8 and G7-8. It has two long parallel walls, while the eastern end of the north wall is lost beneath the unexcavated area. The western side wall appears in square H6 and a parallel internal wall (G7) marks the partition of the structure. Parts of these two walls are obscured by the later digging of the Late Neolithic ditch, but there is sufficient remaining of these walls to assume that the south west corner of the structure lies in square G6.

STRUCTURE 2 The building overlying Structure 1 can be seen in square G8. This consists of a corner and short stretches of the two walls leading off it. This structure also appears to have internal divisions of unequal size, although it is not possible to determine how many as part of the building remains in the unexcavated area and no further traces were identified in squares F7-8.

GóJ G7J G8_

0 1 2 3 4 5m Γ'Ν Structure I 1

1 I

2 I

3 I

4 l=j

5m

^ Structure 2

fig. 2.11. Plan of Group 5 by phase.

2.2.6 Group 6 This is formed by three buildings, one of which has two compartments (fig. 2.12, plate 7 a). The group covers squares M5-8, L5-8, K5~8,J6-8, TY1-2, and TX1-3. Structure 1 is fragmentary but probably square with dimensions of at least 5.90 x 7.80 m. Structure 2 is rectangular with an internal north-south partition. It measures at least 15.00 x 6.80 m. Structure 3 is a fragmentary rectangular structure, a substantial part of which lies outside the excavated area. It measures at least 10.20 χ 7.90 m.

In squares L4-5 and M5 is the possible corner of a structure. According to the site plan this lies beneath Structure 2, indicating that there may in fact have been four structures in this area. Unfortunately the plan is not clear enough for certainty.

STRATIGRAPHY 25

MS //~^^

M6 Κ 1 M7 ' MH r^^^^' ' F 3 AS"A

K5^' ^~^l / J K6 K7 ^φ / K8 ^^^^ Ν

(7 ^^^ j Stmc.ure 1

J J6 1 Π^^. 1 J8j | St^tur^ 0 1 2 3 4 5m l-.--l-.-,.},,,.,.k.,,-_4-.-.-.-.-J

fig. 2.12. Plan of Group 6 by phase.

STRUCTURE 1 This lies in squares M6-8, L6 and L8, and K6-8. The entire length of the wall lying in M6-K6 has been recovered, including both corners and at least half of the two perpendicular walls. Only a faint trace of the opposite wall was found, which is located in square L8.

STRUCTURE 2 (PLATE η û) The second structure is located in squares M5, L5-8, K5-8 and J6-7. It is clearly divided into two compartments, which lie at a slight angle to each other. The western part of the structure was labelled 'house 1 ' in the excavation notebook, and is very well preserved. A gap in the south wall of the eastern section could mark an entrance, although it is fairly narrow (< 1 m). Most of the southern wall of this part of the building was identified during excavation, and the greater part of the northern wall. The eastern wall appears to degenerate into a large straggling feature which seems to cover the probable position of this foundation trench.

26 GILLIAN PYKE

STRUCTURE 3 The fragmentary remains of this building appear in squares M7, L7-8, K7-8, TY1-2 and TX2-3. From the site plans it is clear that this structure overlies both Structure 1 and Structure 2. Parts of one short foundation trench can be traced in squares L7-8 and K7, where there is a corner. The long foundation trench which joins at this point is fragmentary, and only identified in some places. This foundation trench was not observed in square J8 but may have continued into the unexcavated area to the east. The longer foundation trench, running parallel to the fragmentary remains in K7-8, is located in squares TY1-2 and TX2-3. In contrast with the other, the foundation trench was well preserved, which would probably form a corner with the short wall (L7-8, K7) in square M8.

Rodden sees Structure 1 of this group as belonging to his first building phase of the Early Neolithic (Rodden 1964 c). Both Structures 2 and 3 are assigned to the second building phase, with Structure 2 earlier than Structure 3 within this phase.

2.2.7 Group 7 Group 7 (fig. 2.13) appears in squares TY2-8 and TX3-7. It consists of three structures, one of which has compartments. All of these structures were found in spit 1. These seem to be aligned on generally the same axis and overlap each other. A further wall, aligned with the partition walls of these structures is located in squares TY5-6 and TX6-7, which may be part of a fourth structure.

STRUCTURE 1 This is seen in squares TY2-4 and TX3-4. The southern foundation trench, running diagonally through from square TY2 to TX4, is the most complete and has both corners intact. From these corners, foundation trenches run diagonally, parallel with each other, in a generally north east direction into the unexcavated area. Towards the southern foundation trench, and parallel with it, is a narrower partition wall which separates off a small section of the building, to which there is no visible entrance.

structure 2 This building is located in squares TY3 and TY5 and TX4-5. From the site plans, this structure overlies that of TY2-4/TX3-4 (Structure 1). As in the case of the earlier structure, the south foundation trench (TY3-4) is virtually complete, with only the western corner obscured by the cutting of a Late Neolithic ditch. A short portion of the western foundation trench is present; no doubt the rest lies in the unexcavated area. Most of the eastern foundation trench (TX4-5, TY5) survives, including both corners. A small portion of the northern foundation trench is visible but it is likely that it is obscured by a later trench.

structure 3 The latest structure of this group is sited in squares TY5-8 and TX5-6. The orientation of this building is similar to that of the earlier two. The west and south foundation trenches are almost complete and corner in square TX6. The far corner of the west foundation trench is visible in square TY5 and the east corner of the south and minimal part of east

STRATIGRAPHY 27

foundation trenches can also be identified; the rest has not been excavated. This structure overlies Structure 2 in square TY5.

TY2/^ ^^ TY3 '^λ V4 ^/Ί^ίν TY6 ̂^r^ - Ti%

^ ^y TX2 TO TX4 /Γξ TX5 | TX6 | TX7

[''j Structure 1

Structure 2

| I I I I -I ^m ^H Structure 3

fig. 2.13: Plan of Group 7 by phase.

2.2.8 Group 8 This group (fig. 2.14) consists of one separate and two overlapping structures in squares TX7-9 and TY7-9. Too little of Structure 1 was found to estimate its shape and dimensions. Structure 2 was a small square building with dimensions of 4.08 x 4.48 m. Only a corner of Structure 3 was found, which has a possible partition, and a stake barrier joining the ends of the two walls, which are about 9.20 and 3.40 m in length. The corner of a possible fourth structure is located in square TYg.

STRUCTURE 1 The site plan does not clarify the relationship between the two structures which appear in squares TY7-8 and TX7-9. The lines marking the edges of the foundation trenches cross each other in an unlikely manner. As that part of the structure located in squares TX7-8 appears mainly in spit 2, it is likely that this is the earlier structure. It consists of two narrow foundation trenches which form a corner in square TX8.

STRUCTURE 2 This is located in squares TY7-8 and TX7-9. Its appearance in spit 1 leads to its interpretation as the later of the two. The north and west foundation trenches are preserved

28 GILLIAN PYKE

to almost their full extent, and three of the four corners are accounted for. The south east corner was not within the excavated area. This structure, like the earlier one, consists of narrow foundation trenches. Whether by an accident of preservation, or design, the north wall seems set apart from its two adjoining walls and does not actually make contact with them. In square TX7 there is a short section of wall which does not join the structure but aligns with the west foundation trench, giving rise to the possibility that this is a compartmented structure.

STRUCTURE 3 Although not overlapping either of the previous two, this is included in the group because of its proximity to them. It was found in squares TY8-9 and TX9, in spit 1, which indicates that it is possibly later than Structure 1 , although this is not necessarily the case as only one spit was excavated in these squares. The one visible corner appears in TX9. Both these walls appear to terminate with no evidence of corners, the shorter foundation trench ending in ΊΎ9 with a post hole. This could indicate the presence of a doorway or could simply mark the end of the wall of a wind break or animal pen. The presence of a line of post holes running between the ends of the two foundation trenches in TY8-9 might be unrelated to the structure, but could be interpreted as a fence designed to form the third wall of an irregular triangular structure.

™ / [ VK™^^j' /τΥ9 y(

// ,s> η

[' m Structure 1

j Structure 2 0 1 2 3 4 5m ' '

' ' ' ■ ' ==^ ^H Structure 3

Ο pits

fig. 2.14. Plan of Group 8 by phase.

STRATIGRAPHY 29

2.2.9 Group g This is not really a group but a single structure which lies in the squares TB1-2 and Tgi- 2 (fig. 2.15, plate 7 b). It was probably rectangular with two partition walls forming a corridor around two of the sides. It was at least 4.00 x 2.00 m. The site plans show a number of ditches in this area, in squares TBi and Tgi-2, which the site plans indicate belong to the Early Neolithic. The site plans also show that the structure is overlying these ditches, and therefore must post date them. The sequence of the ditches indicates that there were several successive phases. The central ditch overlies both the one lying east of it (TBi-Tgi) and the one to its west (TB2-Tg2). The earlier ditches are parallel and may belong to the same period, or could denote separate episodes. The structure was built over both the central ditch and its western neighbour.

Parts of three of the outer foundation trenches of the structure were discovered during excavation, and the site plans also show that there were interior walls parallel to the west and south foundation trenches. The western interior foundation trench also appears to have two adjoining post holes which are positioned roughly equidistant from the centre of the wall and the two ends. This suggests that their purpose was to support the wall. The southern exterior wall is fragmentary and only a short section of the foundation trench in Tg2 was identified, possibly continuing in TB. The western exterior wall is apparently truncated by the later digging of a large pit, and no trace of the foundation trench was located on the southern side of this feature. The eastern exterior foundation trench is missing, but the adjoining foundation trenches appear to end in a roughly parallel position and the edge of the underlying trench is unusually straight at this point. This suggests that it is not the original edge of the trench that was detected in excavation, but the edge of the eastern foundation trench.

ffe( ?#ΓΒ1 I >S TB2 -X

I Tgi ς. 1 af Tg2

1'^n Structure 1

° ; ]

3 * f A Early Ditch

Β Late Ditch

fig. 2.15. Plan of Group 9 by phase.

3O GILLIAN PYKE

2.2. 10 Summary Examination of the overall site plan showed that there were groups of structures which had been built in succession in specific areas. These groups contained between one and three clear structures, with the possibility of a fourth in groups 4, 7, 8 and 9. The establishment of the plans of the structures allows conclusions to be drawn about their size and complexity, which is the subject of later discussion (see Chapter 3).

2.3 HORIZONTAL ASSOCIATIONS

Following the definition of the plans of the structures within each group, it is possible to investigate the relationship between all the structures on the site. The fact that several of the groups consist of three structures suggests that there might have been three phases of occupation. The location of the earliest structure of each group at a similar level would support this theory.

The relationship of the structures horizontally across the site is based on the matching of the artificial spits in which the building remains were found. Due to the fact that the depths of the foundation trench were not recorded, it is necessary to rely on the plans of each spit to register in which spit the foundation trenches appeared. The fact that it was the base of the spit that was planned must be borne in mind. Due to the constraints of the recorded data, it was only possible to relate the structures in a general way.

2.3.1 Group 1 and Group 2 The sections O7-A7-B7 (along grid line 7 for squares O7, A7 and B7, see fig. 2.16 for the grid systems used), B9-C9-D9 and B8-C8-D8-E8 were examined to relate the structures of groups 1 and 2. Although the sections across group 1 do not provide the information necessary to make certain correlations between the two groups, it appears that the spits in these two areas were dug to roughly the same total level, and therefore it is justifiable to assume that spit 1 and spit 2 were the same approximate depth. In this case the three structures in each group relate to the same three building phases as there is agreement as to where in the spits these structures were found. Structure 1 occurs towards the base of spit 2 and appears more in this spit than in spit 1 ; Structure 2 lies at the base of spit 1 and the top of spit 2 and therefore appears in both spits; Structure 3 is found almost entirely in spit 1 .

2.3.2 Group 3 Group 3 was related to groups 1 and 2 by examining section Eo-Do-Co. Structure 1 of this group lies primarily in spit 3, but also appears in spit 2, probably towards the base. This places it within the range c. + 1 2-24 cm above site datum, approximately corresponding to spit 2 in the other two groups. This suggests that it may belong to the same building phase.

The second structure appears mainly in spit 2 but also in spit 1 in one area. This concurs with the stratigraphie position of Structure 2 in both groups 1 and 2. The location of the

STRATIGRAPHY 3i

12 3 4 5

Β Β I LXI r-r- 1 LI I XLI I A Β Β Β c Γ Π . . . c

ο

Ι . Ι . Ι . "Π I 1 d ε ι it ZlZlllIj ε

Fl^ ' 23 45 LII

12 3 4 5 °^2ηη

Grid system for the 1961 excavation (after Yiouni 1991, Fig. 3.5).

Γ '^'' '^>* 0 12 3 4 5 6

0 j I ' I

Η

^f| Ί

J M I I I 1 I I I l~t~ 000 00 I 1 1 1

j 1 1 1 1 1

Β

a I I I I I 0 12 3 4

T~~^

Q| Ι Γ °_ "" _ ~~ _J 10 5 6 7 8 9 10

"" ~~ _ "^"tres

Grid system for the 1963/64 excavation

fig. 2.16: Grid systems used during the excavations at Nea Nikomedeia.

32 GILLIAN PYKE

group 3 structure at + 24-45 cm would generally overlap with spit 2 and the lowest part of spit 1 in groups 1 and 2. This confirms that Structure 2 in all these groups belongs to the same building phase. In group 3 there are two structures, rather than the three found in groups 1 and 2. From the sections it appears that no structure was built which corresponds with the latest structure of the groups 1 and 2 (Structure 3/spit 1).

2.3.3 Group 4 The relationship of the group 4 structures and those of groups 1-3 could not be established from the D5-E5-F5 section. Section D4-E4-F4 showed that spit 1 was not excavated in all areas in this group, as there was a great difference in surface height. For example, in square E5 the surface of the plough soil corresponds with the level of the base of spit 1 in square D5, and therefore the single spit that was dug in square E5 was labelled spit V by the excavator as its surface level corresponds with the base of spit 1 in square D5. The excavation notebook makes it clear that this is the situation in squares E4, E5, and E6. In square F6, spit 1 is described as being taken down to the level of spit V in E6.

As both structures 1 and 2 occur in spit 2, it is not possible to state their exact relationship with the corresponding structures in groups 1-3. However, it is possible to state that as the range 0-42 cm matches the range of those of Structures 1 and 2 in all these groups, they probably belong to the same building phases. In the same way, the position of Structure 3 in spit 1, approximately matching spit 1 of the other groups, indicates that it belongs to the same building phase.

2.3.4 Group 5 The two structures in this group appear in squares G6-8 and H6-8, an area in which only one spit was excavated. As the section diagrams in the excavation notebook were not completed, and the most information that any of them contain is the depth of the plough soil, it is not possible to make conclusive observations on the relationship between this group and the others on the site. However, it is noticeable that the surface of spit 1 in squares G6-8 (varying from + 37-45 cm above site datum) corresponds with the surface of spit 2 in the other groups, and that of squares H6-8 with that of spit 1. It seems likely that these structures correspond with phases 1 and 2.

2.3.5 Group 6 The three structures of this group are located in squares J6-8, K4-8, L4-8, M5-8, TX2-3 and TY1-3, which is also an area in which a single spit was excavated. It seems likely, however, that these three structures relate to the three structural phases. Again there is the same situation concerning the section diagrams in the excavation notebook.

In squares J6-8 the base of the plough soil can lie at a height of as much as + 68 cm or as little as + 42 cm, but averages at about + 60 cm. In squares K6-8 the highest level of the base of plough soil is at + 50 cm, while the lowest is at + 35 cm, the average in this case being around + 40 cm. The highest level of the base of the plough soil in squares L6-8 is + 46 cm, while the lowest is + 28 cm, the average seems to be around + 35 cm, but the level does fluctuate a great deal in these squares. The section drawings for squares TX2-3

STRATIGRAPHY 33

are much more detailed and show the levels of both the surface and base of spit 1. However, the surface of spit 1 ranges from + 26-50 cm, which must be kept in mind even though the average level is about + 40 cm, while the lower level is fairly consistent at + 14-16 cm. There are no section drawings for square TYi. There is a single complete drawing for square TY2, and another showing the top surface of spit i/base of plough soil. The single section for TY3 also shows only the top of spit 1 /base of plough soil. In this area, from the limited information available, it seems that the upper surface of spit 1 varies between + 40 cm and + 14 cm and a tentative average can be given as + 30 cm. The lower level, where recorded, varies from 18 cm above site datum to 2 cm, with a possible average at around 10 cm.

In spite of the lack of information for these squares, the levels for this spit do show a general agreement with those of the other groups, allowing the structures within this group to be provisionally considered in terms of the same phasing as has been demonstrated for groups 1-4.

2.3.6 Group 7 This group of three structures also appears in an area where only a single spit was excavated, but the correspondence to the three structural phases can be implied.

In the section diagrams for TX3-7 in the excavation notebook only the base of the plough soil is indicated. This varies between 0-45 cm, but a general trend towards about + 25 cm is distinguishable. The recording of sections is more complete for squares TY2-8, but still shows a great deal of variation in the spit 1 surface height (+ 42 cm to - 04 cm), with the average lying at + 18 cm. The poorly recorded base height of spit 1 lies between 2-28 cm. This, however, may give an unrealistic view of the depth of the spit as it would appear that it was therefore only 2 cm thick, whereas in fact it could be anything up to 40 cm thick. The fact that the upper surface of the spit has in some cases been recorded at - 04 cm but the lowest value recorded for the base of the spit is + 2 cm indicates that there may be some inconsistency, although this may simply be due to the lack of detailed recording of the base of spit 1 .

In comparison with the other groups, the section shows that spit 1 lies at a fairly low level, comparable with spit 2 in group 4 and in group 2, and spit 3 in group 3. However it is comparable with the values given for the TY squares of group 6, therefore possibly indicating a trend not only of the spit being at a lower level, but the plough soil is also at a lower level, indicating a slope between the TX and TY squares. This means that the whole section appears at a lower level than in other groups.

2.3.7 Group 8 A single spit was dug in squares TX7. Again, the section diagrams give an incomplete record of the spits. In squares TY8 and TXg the finds list in the excavation notebook shows that 2 spits were excavated but they do not appear on the section diagrams. For squares TX7-8 only the base of the plough soil is indicated. In the TX squares there is a limited range of values for the base of the plough soil, the most frequently recorded levels lie between 10-30 cm, averaging at + 20 cm. The base of spit 1 is recorded in only one section drawing (TX9-TX8) and the highest level recorded is + 02 cm, while the lowest is

34 GILLIAN PYKE

- 04 cm; the section shows that the average is around - 02 cm. In squares TY7-9 only the base of the plough soil is shown on the section drawings. The heights given for the base of the plough soil vary from 6-20 cm.

2.3.8 Group 9 The single structure of group 9 is located in squares TB1-2 and Tgi-2. Five spits were excavated in this area. The section drawings for squares Tgi-2 are limited in that for square Tgi only part of the square has the section drawn in and the levels are ruled in, suggesting a schematic representation of the spits. There is little variation in the levels of the surfaces and bases of the spits in these diagrams. In the case of square TB 1 only the base of spits 2 and 3 are shown. There is no information concerning spit 1. There is no information available for square TB2 as the section drawings were not completed.

The single structure in this group was located in spit 2, which agrees in broad terms with the level of the single spit in which the structures are located in groups 7 and 8. The ditches were found in spit 3, which is at a much lower level than any spit in which structures have occurred.

These ditches were marked as being Early Neolithic on the site plans by the excavator in 1964. However, when recently discussing the structural sequence, Dr. Rodden felt that the assignment of the ditches associated with the group 9 structure to the Early Neolithic was not certain.4

It is difficult to correlate the group 9 structure with the general phasing on the site. The fact that it does not appear in the first spit may argue that it does not belong to the very latest building phase, which tends to appear at a higher level and in the first spit. The interpretation is, however, complicated by the presence of the three ditches, which are apparently Early Neolithic. If, as suspected, there are two phases of ditch activity, these may correspond with the first two phases of building construction over the rest of the site. This would suggest that the structure of group 9 does in fact belong to the final building phase, and its appearance in spit 2 is due to the depth of colluvial deposit from the slopes of the mound which had accumulated on top of this area.

2.3.9 Summary The present study suggests there were three consistent building phases of Early Neolithic occupation of the site. The structures of the earliest phase (fig. 2.17) were found in the lower part of spit 2 and were also located in spit 3 in some cases, in areas in which this spit was dug. The second phase (fig. 2.18) contains structures which appeared in the upper part of spit 2 and the lower part of spit 1, while those of the third phase and the traces of the possible fourth phase (fig. 2.19) appear solely in spit 1.

G. P.

4 Rodden 1993 pers. comm.

STRATIGRAPHY 35

Table. 2.1. Depths of the spits for each group.

Spit 1 (cm) Spit 2 (cm) Spit 3 (cm)

Group 1 +68 - +40? +40 - +20 (base 20) - +62 - +40 (square D8)

Group 2 +60 - +34 +34 - +04 -

Group 3 +78 - +45 +45 - +24 +24 - +1 2 Group 4 +65 - +42 +42 - 00 -

Group 5 +45 - ? (G squares) - - +60 - ? (H squares)

Group 6 +60 - ? (J squares) - -

+40 - ? (K squares) +35 - ? (L squares) +40 - +15 (TX squares) +30 - +10 (TY squares)

Group 7 +25 - ? (TX squares) +18 - 16 (TY squares) - -

Group 8 +20 - 02 (TX squares) - -

+13 -? (Τ Υ squares) Group 9 +33 - +20 (Tgi - west) +20 - +15 (Tgi - west) +15 - 04 (Tgi - west)

+35 - +15 (Tg1 - east) +15 - -°5 (Tg1 - east) +15 - °5 (Tg1 - east) +36 - +20 (Tg2 - west) +20 - +15 (Tg2 - west) +15 - 02 (Tg2 - west) +32 - +18 (Tg2 - east) +18 - +08 (Tg2 - east) +08 - -05 (Tg2 - east) ? - ? (TB squares) ? - -06 (TB squares) -06 - -27 (TB squares)

36 GILLIAN PYKE

'A ■

* ι - I

fH

1 Ή 3

GO

'S S ε κ» ci

STRATIGRAPHY 37

' A

v' VA _ΡΊπ 4? Hf τ

rp-. J

•S. •a I

ε 66 »Η

ei Ü

38 GILLIAN PYKE

V '

' " '

ρ

Ly * r

i ■g

Ο

Ο ΪΖ

Chapter 3 Structures and Architecture

3.1 INTRODUCTION

The buildings and features of the site are only briefly described in Rodden's preliminary report (Rodden 1962) and survey of the Plain of Macedon (Rodden 1964 c). A detailed study of all the architectural remains and other features is therefore crucial to the understanding of the site. Thus the excavation notebooks for the 1 963 and 1 964 seasons and the site plans were used as a source to compile every reference to building materials and features.

The excavation notebooks provide descriptions of features as they were excavated and therefore do not represent a complete account. The representations on site plans were drawn after each spit was completely excavated, hence these may not directly correlate with descriptions in the notebooks. The most detailed information comes from the notebooks of the 1963 and 1964 seasons and therefore this account is primarily based on these.

Foundation trenches, post and stake-holes appear on the site plans, which record the plan at the base of each spit in most cases, but do not describe building materials. The site plans do, however, provide valuable information as to the width and the depth of the foundation trenches. In addition, there are references to post and stake-holes in the excavation notebook, which also appear on the general plan of the excavated area. These were clearly assumed at the time of excavation to be structural remains.

3.2 BUILDINGS

A total of 24 recognisable structures were found during the excavations at Nea Nikomedeia. These were located as foundation trenches and post-holes which showed up as discoloured areas. In all cases, with one possible exception, the structures were rectangular or square. Several ground plans and construction styles are represented, with no one style characterising a particular phase.

3.2.1 Walls The physical evidence for the walls of structures consists principally of foundation trenches. Post-holes and stake-holes, building material, 'rubble' and 'structural carbon' were also found.

39

4o GILLIAN PYKE

FOUNDATION TRENCHES The remains of walls on the site of Nea Nikomedeia appear consistently as foundation trenches, referred to as 'wall-slots' in the excavation notebook and which are often described as 'white' or 'plastered'. For example, the wall at E6 1.00 x 0.50 m in spit 1 of square E7, is reported as being a 'white wall'. The walls of square C2 in spit 2 are described as plastered/ white walls, as is wall A of square C4, and a wall in square D4.

The foundation trenches were probably fairly deep so that the walls were stable and the roof could be supported, but there is no evidence to suggest that they were as much as a yard in depth as Rodden originally believed.1 The foundation trenches were often associated with post-holes. Rodden states that the foundation trenches found in the 1961 excavation were "roughly U-shaped [in profile] and average a width of 60 cm. Down the centre of each wall-slot posts (8-20 cm in diameter) were driven into the subsoil at regular metre to metre-and-a-half intervals, providing a sturdy framework for the dried mud walls" (Rodden 1962, 269).

BUTTRESSES Rodden states "Sockets for at least two sturdy buttress posts were noted on the inside of the long wall [of group 3/2], evidently to support cross-beams" (Rodden 1962, 270). The sketch plan of square C1/2 shows foundation trench A with whitish clay fill and a possible buttress is labelled.

POST-HOLES AND STAKE-HOLES There are references to post and stake-holes in the excavation notebook, and they also appear on the general plan of the excavated area. As many more post-holes appear on the site plans than are described, it seems that they were only noted in the excavation notebooks if they were plaster lined. Examples of these are Feature C in square A3/1, which is recorded as a 'plaster post-hole', and Feature A in square B4 which was a plaster lined post-hole noted to continue through spits 2 and 3. The post-holes were assumed at the time of excavation to be structural remains.

Rodden notes that "The rectangular arrangement of post-holes uncovered outside the north wall of the principal structure ... [group 3/2] describes a subsidiary timber enclosure. In plan, it is not unlike an open-ended square: two of the sides were built of freestanding rows of posts (m-p and p-u), the third side incorporated the existing north wall of the house. The post-holes vary from c. 20-30 cm in diameter, and were sunk up to 65 cm into the sub-soil. To judge from the size and depth of the post-holes, the surround must have been a sturdy and fairly permanent structure" (Rodden 1962, 270).

Post-holes filled with burnt material were also recorded. These include several post-holes which appear on the site plan of square B6 and were filled with burnt rubble. The description of square O7/1 notes a possible plaster-lined post-hole (Feature A) and two stake-holes (Features Β and C) with sooty fill.

1 In reference to the structures excavated in 1961, Rodden describes the foundation trenches as cutting "34- 50 centimetres into virgin subsoil* (Rodden 1962, 269). In a later report he states that the foundation trenches were "a yard or so deep, evidently to ensure that the buildings would not be affected by frost heave or by the

wetness of the waterfront soil** (Rodden 1965, 97). The climate of Macedonia at that time was probably not such that frost heave would be a consideration, and it seems unlikely that a structure would be built on extremely wet ground.

STRUCTURES AND ARCHITECTURE 41

'rubble' and other building material There are several references to the presence of bubble' in pits close to foundation trenches, which were presumably filled as a result of building collapse. This 'rubble' was originally interpreted as representing the clay plaster or daub of wall tumble, but subsequent evidence from the experimental burning of a similar house suggests that such rubble can also come from the ceiling (Bankoff and Winter 1979).

The excavation notebook account for spit 1 mentions 'rubble' from the walls and possibly ceilings of some of the structures. These sometimes have 'wattle impressions', such as in square D9/1, where a piece of building material with 'wattle impression' was found in debris (unlocated) and at D8 2.00 x 3.75 m (Feature A). However, it is not clear what these descriptions refer to. The presence of rubble would suggest that mudbrick or stone were used for building, which was not the case at Nea Nikomedeia. The initial observation that lumps of clay with wattle impressions (implying the weaving of branches) were preserved appears to have been revised as Rodden's later description of the construction techniques does not include the use of wattle and daub (Rodden 1964 a).

In the entry for spit 1 in square TXg, fragments of burnt bricks are recorded as being found in the Late Neolithic Trench. A second 'brick' was reported to be found at C5 3.80 x 3.80 m in square C6/2. Mudbrick might have been used in the Late Neolithic at Nea Nikomedeia, but as no structures have survived it is impossible to be certain.

'structural carbon' It seems that the term 'structural carbon' was applied to charcoal and represents the remains of posts which were burnt when the structures were destroyed by fire. Several samples of this material were taken in spits 1, 2 and 3. Rodden says that "Carbonized remains of wood indicate that the frames of the houses were made of oak", presumably on the basis of these samples (Rodden 1965, 84).

3.2.2 Floors The presence of floors within the Nea Nikomedeia structures is attested by areas of hard white clay. An example of this is a platform of hard clay which was noted as a possible floor in square A5/1. In the second sketch plan of square A6/2 there was a floor of hard white clay with yellowish discoloration.

There is at least one instance of a floor constructed of pebbles and clay, including a yellow plaster and pebble floor in square B4/ 1 which appears in a sketch in the excavation notebook. No floors were found in spit 3.

3.2.3 The Account of F6/1 There is a longer description in the excavation notebook of the structural remains found in square F6/ 1 . The fact that there is an unusually long account of the excavation of this square suggests that the excavator regarded the discoveries as particularly important. It seems likely that the account was written while the excavation was in progress; in places the description is confusing and the exact location of the features being described and their relationship with each other uncertain.

42 GILLIAN PYKE

It appears that a large section of collapsed wall was found, which sealed an oven and some objects, including five figurines. The figurines were apparently found on a floor surface, which is associated in some way with the base of a possible oven. The remains of the wall were consistently called 'chaff-like material' or 'chaff tumble', which suggests that the clay used to coat the walls was mixed with chaff temper. The description of 'large lumps of red- black chaff-filled material' might indicate that the wall fragments were burnt, which resulted from the destruction of the building. It is not clear, however, if the description refers to wall debris or part of the oven, which would contain burnt material. Another description of an area of heavily burnt lumps of 'chaff-like' material, located at F5 3.20 x 1 .00 m, was interpreted as a collapsed wall. Below the large red lumps were found two pieces of what was possibly the base of a clay oven.

The location given for the area of heavily burnt wall debris suggests that the wall in question may be that of group 4/1, as the wall appears to have fallen into the interior of the building rather than outside it. This means that the oven and the small finds are probably associated with group 4/ 1 .

3.2.4 Construction It is possible to suggest construction methods from the architectural evidence found during the excavation. Rodden proposes that the same techniques were used in both the building periods which he identified (Rodden 1964 a), and there is a diagram of a hypothetical structure (fig. 3.1) in one of his preliminary reports (Rodden 1965). The present examination of the evidence supports this interpretation. The same kind of information appears in the excavation notebooks and on site plans for each spit, indicating continuity in materials and techniques.

Foundation trenches were dug for the walls of each structure. Sometimes only part of a foundation trench was found (group 6/3); in some cases it seems that there was no foundation trench and the wall was simply constructed by sinking posts directly into the ground (group 2/3).2

Posts were sunk down the centre of the foundation trenches to form "a sturdy framework of vertical timbers (probably oak)" (Rodden 1964 a, 564). This can be seen clearly in the north wall of group 3/2 and the south cross wall of group 4/2. Buttresses were sometimes added, possibly to support the cross beams for the roof (Rodden 1962). These can be seen as post-holes placed close to the foundation trench on the inside of the structure (such as on the cross walls of group 4/2).

According to Rodden the main part of the wall consisted of "stalks of reeds or rushes, about a centimetre in diameter, standing between the posts" (Rodden 1964 a, 564). This view may be based on the pieces of 'building material' with 'wattle impressions' which are recorded in the excavation notebooks, which may in fact have had reed impressions. These reeds would have formed the continuous surface of the wall, which was "completed with the plastering of mud onto both outer and inner surfaces" (Rodden 1964 a, 564). This clay

2 Rodden suggests that, in order to keep the walls dry, *The foundation trenches for the walls were dug to an unusual depth and either packed with dry, permeable soil, or, more commonly, lined with clay-marl, which was left

to dry to cement-hardness in the sun before the walls were built.** (Rodden 1964 a, 564). A more likely interpretation is that the clay (?) lining was for stability.

STRUCTURES AND ARCHITECTURE 43

fig. 3.1: Reconstruction of a Nea Nikomedeia house.3

could have been derived from the material from the foundation trenches, or pits might have been dug as a source. It is also suggested that "clay mixed with chaff" was sometimes used to coat the walls (Rodden 1964 a, 564) probably due to the 'chaff-like material' and 'chaff wall tumble' mentioned in the excavation notebook entry for square F6/1.

The use of wood and reeds at Nea Nikomedeia was probably due to their local availability. The presence of reeds close to the settlement suggests that the local environment may have been marshy. It is probable that the oak that was used for the framework of the walls came from a different source from the reeds as they grow on drier soil.

The floors were made of compacted clay or clay with pebbles. In particular reference to the Original 'shrine" (group 4/1), Rodden describes the floor as being "made of mud,

3 After Rodden 1965, 87.

44 GILLIAN PYKE

which had been plastered onto a 'matting' of broad-leaved marsh grasses or reeds laid on the clay subsoil" (Rodden 1964 a, 564). Unfortunately, there is no mention of the discovery of any organic 'matting' in the excavation notebooks.

No evidence for the construction of the roofs of the structures was discovered during the excavation. As Clark points out, heavy rainfall necessitates houses having pitched roofs, so that the rainwater would run off rather than come in through the roof (Clark 1952). In addition to this consideration, Rodden suggests that "in order to keep the walls dry ... it also seems probable that ... the buildings had roofs which were fairly steeply pitched, with wide eaves" (Rodden 1964 a, 564). Clay house models found at sites in Greece and the Balkans, such as at Porodin and an unprovenanced find from Rumania show that pitched roofs were used (Theocharis 1973).

It is probable that the technique of roof construction was similar to that used for the walls in the use of a wooden framework and reeds. Bankoff and Winter note that the ceiling of the building that they used for their burning experiment was daubed (Bankoff and Winter 1979). Clay was probably used to waterproof and insulate the roofs of the structures at Nea Nikomedeia.

3.2.5 Size and Shape of Buildings The size and shape of the buildings was investigated by phase to try to establish if any patterns were noticeable. It was hoped that it would be possible to determine if any differences between the phases represented trends in the development of the settlement.

PHASE ONE The first structural phase had a mixture of square and rectangular buildings, all of which are represented by foundation trenches. Two of the rectangular structures had internal partitions. Group 5/ 1 had a partition offset to the east of a possible entrance, dividing the structure into a square room to the east and a more rectangular one to the west. The exact size of the structure is uncertain because the west foundation trench and the western ends of the north and south foundation trenches were not within the excavated area. The other example (group 7/1) was partitioned in such a way as to create a very narrow rectangular southern room, and a much larger northern room. The exact shape of the building is unknown. A break in the foundation trench at about its mid point may represent an entrance.

The size of the building of this phase (table. 3.1) was fairly uniform, the average dimensions being 8.37 x 6.66 m, taking the lengths of the complete walls but omitting those of group 4/1 due to their unusual length. The square structure (group 4/1) was very poorly preserved but considerably larger than all the other phase 1 buildings. The difference in size suggests that it may have had a special function, possibly as a communal building. The buildings of this phase are generally orientated east-west.

STRUCTURES AND ARCHITECTURE 45

Table. 3.1: Building dimensions: phase 1.

Group Shape Dimensions (m) Trench Width Area m2 Comments (n/s χ e/w) (average m) (Ρ) = preserved length

1 square 7.41x6.51 0.47 48.43 2 rectangle 6.48 χ 6.8ο 0.40 44.06 clear stakeholes 3 rectangle 8.99x6.82 0.51 61.31 shape irregular 4 square 11.78x13.64 0.44 160.68 difficult to trace 5 rectangle 6.82 (P) x 3.75 (P) 0.54 >25-58 partitioned 6 square 5.89 (Ρ) χ 7.75 (P) 0.46 >45·^5 difficult to trace 7 unknown 8.68 χ 7.13 (P) 0.47 >6i.89 8 unknown - - - partitioned 9 no structure - - -

PHASE TWO There are both square and rectangular structures in this phase (table. 3.2), several of which have partitions. There are examples of partitions which divide the structure into rooms of equal sizes and of those which divide the structure into a small rectangular room and a much larger room. In the case of group 4/2 there appear to be two partitions, following Rodden's reconstruction at the time of excavation (Rodden 1964 c). It is possible, however, that there were in fact two structures: a large square building, and a later rectangular building with its shorter east and west walls overlying those of the larger structure.

The rectangular structure must postdate the square one, as the rectangular inner structure (or area of the larger structure) has many obvious post-holes within the foundation trench.

If we are dealing with two structures, the visible post-holes of the later rectangular structure must penetrate the earlier foundation trench, as the foundation trench could not be cut through the post-holes without destroying them.

As was the case in phase 1 , the lengths of the walls are fairly consistent when they are preserved to their full extent. Once again, the exception is the group 4 structure. If the size of the building indicates a special function, then this must be true of group 4/2, which may have been built to replace group 4/1 and therefore had the same function. The orientation of the buildings of this phase is generally east-west.

46 GILLIAN PYKE

Table. 3.2: Building dimensions: phase 2.

Group Shape Dimensions (m) Trench Width Area m2 Comments (n/s χ e/w) (average m) (P) = preserved length

1 unknown 7.44 (Ρ) χ 6.82 (P) 0.45 >5Ο·74 ι complete wall 2 rectangle 5.92x7.12 0.62 42.15 3 rectangle 10.85 χ 7.50 (Ρ) Ο·48 >8i.s8 3 walls 4 rectangle 10.85 (Ρ) χ 9.05 °·5° >98·ΐ9 partition 5 unknown - 0.55 - corner, partition 6 rectangle 15.04 (Ρ) χ 6.76 0.48 M01.67 partition 7 square 8.43x6.08 0.47 51-*5 8 square 4.08x4.48 0.25 18.28 9 no structure - - -

PHASE THREE

Only six structures were found which belonged to phase 3 (table. 3.3). Two of these were rectangular, one was a square structure and one possibly a square structure. The other two lie partially in the unexcavated area, and their shape is impossible to assess.

The structures of groups 1 and 2 and possibly those of group 8 appear to be built using two construction techniques. The foundation trench and post-hole technique was used for most of the building, but the foundation trench is absent for at least one of the walls in each case.

The presence of a series of posts along the line where a foundation trench would be expected suggests that simply posts were used on their own without the trenches.

This feature appears for the first time during this phase. It is not certain whether this represents simply a change in building technique, perhaps reflecting that the foundation trenches were no longer considered as important, or a change in the use of part of the building or the whole building.

The interpretation of structure group 8/3 in terms of a post-walled building is tentative, but the line of pits or post-holes between the ends of the two walls gives the impression that the building was triangular with two sturdy walls - with foundation trenches - and a less substantial cross wall. The trench leading out of the apex of the triangle could indicate that the building was partitioned, possibly for the stalling of animals. The buildings of phase 3, like those of the earlier phases, are orientated east-west.

STRUCTURES AND ARCHITECTURE 47

Table. 3.3: Building dimensions: phase 3.

Group Shape Dimensions (m) Trench Width Area m2 Comments (n/s χ e/w) (average m) (P) = preserved length

1 rectangle 8.84x7.74 0.45 68.42 post wall? 2 square 4.19x4.74 0.19 19.86 post wall? 3 none - - - no structure 4 unknown - - - traces 5 none - - - no structure 6 rectangle 10.23 (P) x 7-91 (Ρ) Ο·49 >8o.92 post wall? 7 unknown 7.04x8.22 0.50 57.87 corner 8 unknown 9.15 χ 3.35 0.34 30.65 unknown 9 unknown 4.03 (Ρ) χ 2.02 (P) 0.34 >8.i4 partition

3.2.6 Settlement Any estimation of the size of the settlement in terms of the total number of structures per phase and the total number of inhabitants per phase can only be tentative and based on the information available from the excavated area. However, having established a provisional stratigraphy and examined the structures of each phase, it might be possible to estimate the actual population of Nea Nikomedeia.

Considerations such as variable density of occupation, the presence of possible communal buildings, the possibility of structures lying empty and uncertainties about the area covered by the settlement must be borne in mind when assessing the number of structures in each phase. Having considered these points, if the excavated area is approximately 1690 m2 and the total mound area 24,200 m2, then the excavated area accounts for around 8% of the mound.

It was decided to include the possible communal structures in the total of excavated structures as it is unknown how many such structures there might be within the settlement. In phase 1, the eight structures of the excavated area would imply a total of one hundred for the entire site. In phase 2 there would also be one hundred structures and in phase 3 there would be seventy five.

When estimating the population of the settlement for each phase, it is important to remember that the procedure is necessarily based on a guess about the number of people per house. We cannot know whether the structures housed nuclear, extended or another variant on a family group. It is possible to base the population estimate on the floor area within a structure required by each individual, which Whitelaw estimates at 10 m2 in his study of Early Minoan Myrtos (Whitelaw 1981). Cross-cultural studies suggest that each household consisted of 4-6 individuals.

48 GILLIAN PYKE

The average floor area per structure in phase ι is approximately 67 m2, which gives an average of 6-7 people per structure, which is comparable to Whitelaw's estimate, considering the unusual size of group 4/ 1 . If this structure was omitted from the calculations, then the number of people per structure would be 5. The population of the whole settlement might have numbered from 500-700. In phase 2 the average area of each structure is about 68 m2, which is very similar to the estimate for phase 1 . This gives a similar total population for the settlement. The average area of each structure in phase 3 is 44 m2, giving a total of 4-5 people per structure. If there were 75 structures in the settlement, then there would be a population of 300-75.

3.2.7 Summary The stratigraphie investigation, using the available evidence, tentatively suggests three stratigraphie phases. The lack of evidence for any occupation during the Middle Neolithic indicates that there must have been a break in occupation between the Early ajid Late Neolithic periods.

The three phases recognised here represent only the Early Neolithic occupation of the site and appear to follow in relatively swift succession. There is no break or change in the pottery sequence (Yiouni 1991) or the small finds.

There is no archive evidence for abandonment between the structural phases, such as the development of a humic layer, represented by a dark band within the archaeological deposits.4

From the accounts in the excavation notebooks and the plans, the structures appear to overlap in such a way that it was difficult to define the walls of each individual structure. This suggests that they were at a similar stratigraphie level and that there was little time for the accumulation of debris before another structure was built. From the amount of burnt material that was found, it is possible to deduce that several structures may have been destroyed by fire. The fact that there is such noticeable clustering of the buildings which overlie each other, and yet avoid the structures of other groups, suggests that there was a degree of spatial awareness in the siting of structures.

An interpretation of the site based on the hypothesis of a succession of buildings which were occupied over a relatively short period of time is supported by the building techniques. The plans of the structures over the three phases are fairly homogeneous, with variations (such as the use of partitions) possibly reflecting functional variation within the structure (which does not show up in the archaeological record) or simply personal preference. The use of foundation trenches and pisé walls was ubiquitous throughout the Early Neolithic and was supplemented by the possible use of walls which did not have foundation trenches in phase 3.

The variation in the size of structures is not great, but the presence of an unusually large structure must be explained. The first two structures of group 4 are both considerably

4 In his preliminary reports, Rodden appears uncertain on this point as in one report he states "... there are neither marked strati graphical breaks in the accumulation of deposit nor obvious changes in the material culture. *

(Rodden, 1964 a, 564), while in a later article he states

"[the] two building periods ... are separated in places by a deposit of what appears to be the beginning of a humus soil, so that the second building period evidently represents a reoccupation of the site after a period of abandonment." (Rodden, 1965, 96).

STRUCTURES AND ARCHITECTURE 49

larger than their contemporaries. This may be because the building had a special function, such as a 'shrine', as is suggested by Rodden (1964 a and 1964 b). It is also possible to suggest other special functions such as a meeting house for elders or other special groups or the house of a chief.

It is not possible to determine the size of the Early Neolithic settlement during each phase. The number of structures found within the excavated area from phases 1 and 2 is relatively consistent. This suggests that the size of the settlement remained approximately the same. On the basis of the surviving evidence, the settlement was 25% smaller in phase 3, although Rodden originally believed that his earlier phase was a smaller settlement than his later (Rodden 1965).

3.3 FEATURES

Every reference to building materials, ovens and clay lined pits for every spit across the whole site was gathered from the excavation notebooks for the 1963 and 1964 seasons and the site plans. The information from these sources was compared with Rodden's descriptions of the features in the preliminary reports.

3.3.1 Pits The preliminary survey described above concentrated on clay lined pits, as these appear from the descriptions in the excavation notebook to be a definite group. This may, however, be a false impression, and the clay lined pits could have been used for a variety of functions. The purpose of the use of clay in the lining of these pits is not certain.

A further survey of all the pits, noting all the available information from the excavation notebooks and the site plans was attempted, with the hope of distinguishing those used for rubbish disposal, storage, or other purposes and thus estimate whether a distribution pattern of each type was apparent. The account of the pits in each square usually took the form of brief observations on the texture and colour of the fill; not all pits were described and no samples were taken. Some pits had drawn sections showing their shape, but as these were produced after the pit had been completely excavated they could not be used to obtain information regarding the fill.

The overall site plan showed that the pits were scattered randomly across the site, sometimes cutting through foundation trenches. There were no concentrations of pits associated with particular structures. The frequency of pits appears to decrease towards the eastern side of the site, and there are few close to the ditches at the eastern limit of the excavated area.

In some cases the pits cut through the foundation trenches of structures of the latest phase of occupation. These might be a product of the activity of Late Neolithic people. Pit A in square D9 was identified at the time of excavation as being of Late Neolithic date, but it is unclear whether this is based on the character of objects or pottery found in the pit or simply due to the fact that it cuts through the foundation trench.

Sections of some of the pits were drawn, and some of those completed during the 1 96 1 excavation appear in the 1962 preliminary report. Rodden suggests that the three largest pits, which were relatively shallow and flat bottomed, were originally used as a source of clay for wall construction and later for storage (Rodden 1962). He also identifies rubbish

5o GILLIAN PYKE

pits, which were deep and irregular with rounded bottoms. These had a dark fill composed of animal bones, ash and charcoal, all of which were absent from the storage pits, which he says were relatively clean.

3.3.2 Ovens and Hearths At the time of excavation, distinctions between ovens, which are roofed constructions, hearths, which are unroofed constructions and cooking areas, which have no constructed element, would have been difficult to make.

In his 1962 report, Rodden describes what the two 'ovens' found in 1962 probably looked like (fig. 3.2) "... open on top, roughly cylindrical in form, set in or on a basin scooped out of the sub-soil" (Rodden 1962, 270-1).

1 ^ A/

/ ο ] Ν·- -' ; I / ο '

' r v. *s,-« 0/,'

0 1 2 3 4 5m / X ^ Ovens 1 I I I I I Y /íl

fig. 3.2. Plan of the 1961 excavation showing the location of the two ovens.

STRUCTURES AND ARCHITECTURE 51

He continues to describe the material from which the oven was made "The fragments of oven wall found in the oven debris were made of baked clay with a large admixture of straw and vegetable temper. Besides the wall fragments the partly collapsed ovens contained burnt ash and rubble, but very little charcoal" (Rodden 1962, 271).

Despite the fact that there is no detailed description of these features of the two ovens found in 1961 in the site notebook, clearly their excavation provided a great deal of information about the size, shape and construction of ovens used at Nea Nikomedeia.

The problem of the accurate description of features is also highlighted by Rodden's account of hearths found in the 1 963 excavation. He states that "Clay-lined basins, filled with ashes and carbonized wood, were found inside two . . . structures [of Rodden's building phase 2]. The discovery of carbonized cereal grains scattered around one of these basins suggests that they may have been hearth basins used for parching grain - a procedure necessary in a temperate climate in order to prevent it from germinating" (Rodden 1964 a, 564)·

It is unclear from this account which structure these hearths were found in as there is no description in the excavation notebook of these features. Clearly, Rodden subsequently distinguished between ovens and hearths, but the preliminary accounts in the excavation notebooks do not do so.

Possible ovens and hearths were found in spits 1, 2 and 3, although the majority of notebook entries concern spit 1. The entries in the excavation notebook are not always clear as to the size, shape and exact location of the feature being described, hence it is impossible to reconstruct exactly what they are describing, especially as a definite identification could not be made. The presence of burning, charcoal and some kind of building material may be interpreted as indicating the presence of an oven, but may also result from the preservation of building materials such as posts and plaster due to the destruction of a building by fire.

Ovens' were identified through the presence of burnt areas and structural material. The excavation notebook records the presence of several Ovens', some of which are also represented on the site plans. Rodden notes the presence of two 'ovens' which were found in the first season of excavation, which are located between the two structures, and may be associated with them (fig. 3.1) (Rodden 1962). These do not appear on the overall site plan.

The excavation of a possible oven is described in the excavation notebook entry for square F6/1, but does not appear on a plan. It may have been located in the corner of structure group 4/1. In square B8/1, Feature A is a possible Oven' (according to the spit 1 sketch in the excavation notebook) in the corner of group 2/2. It appears to be a pit which was lined with burnt brown material which may originally have been clay. There is another 'oven', in spit 1 in the same square (Feature B), located between group 2/2 and group 1/2. This 'oven' was similarly lined. No structural remains are mentioned in either case, which suggests that the provisional identification as Ovens' may be incorrect. The excavator notes that the 'cooking area' (previously unmentioned) in this square was demolished, although it is not clear what form this 'cooking area' took.

Hearths are not specifically mentioned. Areas of burning are also recorded in the excavation notebooks, and to a limited extent on the site plans. As areas of burning could only be identified when a note was written on the site plan, it was not possible to determine the size or shape of the burnt area. An example of this is Feature Ε in square D4/2, which

52 GILLIAN PYKE

is described in the excavation notebook as consisting of an 'area of earth burned red- brown with flecks of charcoal'. Some discoloration of the surrounding black earth under the 'white border' was noted on the surface. The feature was also said to contain fragments of 'burnt plaster', and rested on a 'clayey grey deposit'. The presence of charcoal, burnt plaster and general burning may suggest that this was an oven, but as the dimensions of the feature are not available and it does not appear on a site plan, such an interpretation remains hypothetical.

3.3.3 Storage and Granaries There is a single reference to a storage jar in the excavation notebook, which is in square B7/0. It was located at B6 1.00 m x 0.75 m and consisted of fragments in a burnt area with burnt daub and charcoal adhering to the outside of them. However, there is no indication of any architectural features used specifically for storage purposes.

No granaries have been specifically mentioned in the excavation notebooks. The presence of carbonized grain on the site indicates that storage of grain was undertaken.

3.3.4 Ditches The main ditches on the site (fig. 3.3) are those lying on the far east of the excavation, at the edge of the mound. According to the pottery and objects associated with these ditches, they were identified as belonging to the Late Neolithic phase of occupation.5 It is possible that some of the ditches are from the Early Neolithic period.6 The ditches in question are those which appear as two parallel narrow ditches, over which lies the single structure of group 9, with a wider ditch cut across them, located in squares TB1-2 and Tgi-2. This interpretation is, however, far from certain.

G. P.

5 Rodden 1993 pers. comm. 6 Rodden 1993 pers. comm.

STRUCTURES AND ARCHITECTURE 53

I ^*- 1

1 i

' ' S

' ' s

Wà* i 1 1

^ ι

CO

•s Ο

ε CO

CO

Ü

Chapter 4 The Early Neolithic Pottery: Technology

Analysis and Comparison with other Neolithic Ceramic Materials

(plates 8-14)

4.1 INTRODUCTION

Information on the technology of the Nea Nikomedeia vessels was obtained by three methods: macroscopic examination, microscopic analysis, and a retiring test.

Macroscopic examination: During macroscopic examination the fabrics used for the manufacture of the vessels

were separated into different groups on the basis of colour and hardness of clay, the amount, type and size of non-plastic inclusions. Sherds were carefully examined for any marks left from the manufacturing techniques and for information on the method of attachment of lugs and bases. Cross sections and surfaces of vessels were examined in order to ascertain the firing conditions.

Miaoscopic examination: A number of sherds were selected for further analysis at the Institute of Archaeology,

University College, London. The following analyses were undertaken: pétrographie study of fifty four thin sections (table 4.1) and analysis of five thin sections with the scanning electron microscope (table 4.2).

Pétrographie examination: The aim behind the pétrographie examination was to confirm information on the

technology of vessels obtained by macroscopic examination, and to relate to provenance studies (see section 4.7.2). The sizes, shapes and proportions of non-plastic inclusions present in the clay matrix of ceramics can be determined from cut sections (1 mm thick) viewed under a polarising microscope. Two types of light are used for the examination of thin sections: plane polarised light (PPL) and light with crossed polars (XP). Minerals and ceramic substances exhibit a number of properties in both types of light that are characteristic and allow their identification. For the Nea Nikomedeia material thin sections were taken from all fabric groups distinguished by macroscopic examination. The major categories of non- plastic inclusions were identified and much information on clay preparation was obtained by examining the size and contour of non-plastic inclusions. The presence and nature of surface coatings was investigated.

By using a detailed geological map of the area, the author was able to compare the non- plastic inclusions present in the clays, with the local rock outcrops. This comparison gave information on the availability of the non-plastic inclusions in the surrounding area. The composition of red-brown slipped sherds was the main objective for the scanning electron

55

56 PARASKEVI YIOUNI

Table 4. 1 : list of thin sections examined with the polarizing microscope.1

Reference Decoration Treatment Munsell Colour Fabric

D4/2, R13 Plain Slipped 2.5YR 4/6 A D4/2, R32 Plain Slipped 2.5YR 4/6 A A3/3, R16 Plain Slipped 2.5YR 4/6 A Β 1/3, R23 Plain Slipped 5YR 4/4 A D0/1, R18 Plain Slipped 2.5YR 6/6 A C3/1, B21 Plain Uncoated 10YR 6/3 A B4/2, R5 Plain Slipped 2.5YR 3/6 A L8/1, B31 Plain Uncoated 10YR 6/3 A B4/1 Impressed Uncoated ioyr 6/3 A C6/1 Impressed Uncoated 7.5YR 7/4 A C9/1, Ri Plain Slipped 5YR 4/6 B-i D4/1, R11 Plain Slipped 2.5YR 6/6 B-i A3/3, R2 Plain Slipped 2.5YR 4/6 B-i K5/1, R16 Plain Slipped 2.5YR 6/6 B-i D0/1, R23 Plain Uncoated 5YR 4/6 B-i C9/1, R35 Plain Slipped 5YR 3/4 B-i M5/1, R25 Plain Slipped 5YR 5/4 B-i K5/1, R5 Plain Slipped 2.5YR 4/4 B-i D4/2, R45 Plain Slipped 2.5YR 4/6 B-i G8/1, R7 Plain Slipped 2.5YR 5/4 B-i A5/2, R12 Plain Slipped 2.5YR 4/8 B-i C10/1, R67 Plain Slipped 5YR 3/3 B-i D0/1, Ri Plain Slipped 2.5YR 4/4 B-i D6/2, R15 Plain Slipped 2.5YR 3/3 B-i C7/1, R27 Plain Slipped 2.5YR 4/6 B-i C00/2, B19 Plain Uncoated 7.5YR6/4 B-2 G8/1, R33 Plain Slipped 2.5YR 6/6 B-2 A6/2, B7 Plain Uncoated 7.5YR6/2 B-2 D2/2, Rio Impressed Slipped 2.5YR 4/6 B-2 D0/2 B3 Impressed Uncoated 7.5YR6/4 B-2 B9/3, B2 Plain Uncoated 7-5™ 7/4 C C10/1, P25 Plain Coated 10R 4/6 C D4/2, R13 Plain Slipped 2.5YR 6/6 C G8/1, P3 Plain Coated 10R 5/6 C D4/2, P2 Plain Coated 7.5R 5/6 C A3/3> P5 plain Coated 7.5R 5/6 C C10/1, Pi Plain Coated 10R 5/6 C C5/2, P22 Plain Coated 10R 5/4 C M7/i,P4 Plain Coated 1 or 5/6 C

1 Key to tables 4. 1 and 4.3: Reference: Excavation square/spit number, followed by the number given to sherds during macroscopic examination (R: Red; P: Pink; B: Beige); Decoration: Plain/ Decorated sherd;

Treatment: Surface Treatment - Slipped/Coated/ Uncoated; Munsell Colour: Munsell Soil Colour Chart notations of the exterior surface; Fabric: Fabric types, as described in section 4.4.6.

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 57

Table 4.1 continued

D 10/2, Pi 9 Plain Coated 10R 5/4 C K5/1, P24 Plain Coated 7.5R 5/6 C Β 1/3, P8 Plain Coated 10R 5/6 C F7/1, P6 Impressed Coated 10R 4/6 C J6/1, R3 Plain Uncoated 2.5YR 4/6 D A2/3, R49 Plain Uncoated 2.5YR 4/8 D B9/2, R45 Plain Uncoated 5YR 4/6 D B6/1, P3 Plain Coated 7.5R 5/4 D B9/3, R12 Applied Uncoated 2.5YR 6/6 D E7/1, R3 Impressed Uncoated 5YR 4/6 D B3/1, R12 Impressed Uncoated 2.5YR 5/6 D A1/3, B4 Plain Uncoated 10YR 6/3 Ε B2/1, B33 Plain Uncoated 10YR 7/3 Ε F4/1, Bi7 Plain Uncoated 7.5R 7/4 Ε C3/2, B2 2 Plain Uncoated 10YR 6/3 Ε

Table 4.2: Chemical composition of slips and coatings of the Nea Nikomedeia vessels, studied by SEM.

% Oxide Samples ist 2nd 3rd 4th 5th

SiO2 68.032 60.102 63.237 63.620 57.OI9 AI203 18.032 21.073 18.922 18.247 14·497 Na2oi .649 1.038 1·394 ι.88ο Κ 2Ο1 2.464 4.308 3.770 3.651 3.565 Mgioi 0.666 1.086 !-795 1-33° 4·297 Caioi !-572 1-P7^3 2.820 2.005 Fe2O3 7·679 9.106 7-39° 8.191 20.071 Mmo2 .102 .181 .230 .171 Cr2O3 .148 .168 .000 .129 T1102 .479 .699 .736 .920 .558 Cuioi .000 .026 -255 .000 S 101 .115 .080 .150 .115 Ρ 205 .ooo .000 .000 .000 Ci 102 .000 .028 .279 .743

TOTAL 99982 99978 99.979 100.001 100.003

Key to samples (an explanation of the data given can be found in footnote 1, above): ist Red slip (Reference: D4/1, R11; Munsell Colour: 2.5YR 6/6; Fabric: B-i). 2nd Red micaceous slip (Reference: K5/1, R16; Munsell Colour: 2.5YR 6/6; Fabric: B-i). 3rd Brown slip (Reference: C9/1, R35; Munsell Colour: 5YR 3/4; Fabric: B-i). 4th Clay matrix (Reference: K5/1, R16; Fabric: B-i). 5th Pink coated sherd (Reference: C10/1, P25; Munsell Colour: ior 4/6; Fabric: C).

58 PARASKEVI YIOUNI

microscope analysis. Scanning electron microscopy is a technique used to gain information on the microtopography of the surface of a wide range of materials. The excellence of the image is built up by observing the intensity of emitted and back-scattered electrons at each point while scanning the beam over the surface. The chemical composition of the material can be deduced from the back-scattered electrons. A number of sherds were refired (table 4.3), under oxidizing conditions, at 8oo° C for 40 minutes. The refiring test was carried out in order to establish whether the pink coating was applied before or after firing. Brown and red-brown sherds were also refired, for checking their reaction in firing under oxidizing conditions.

Table 4.3: List of sherds refired at 8oo°C for 40 minutes.2

Reference Decoration Treatment Munsell Colour Fabric

A3/3, R16 Plain Slipped 2.5YR 4/6 A D4/2, R45 Plain Slipped 2.5YR 4/6 B-i G8/1, P3 Plain Coated ior 5/6 C M7/1, P4 Plain Coated ior 5/6 C B9/3, R12 Applied Uncoated 2-5** 6/6 D

This Chapter can be separated into three parts. In the first part the method followed in the macroscopic and microscopic analysis of the ceramic material is described. The second part is concerned with the description of manufacturing techniques, surface coatings and firing of the vessels. The third part starts with the presentation of the fabrics used for the manufacture of the vessels, following which questions concerning the origin of the non- plastic inclusions, addition of temper and refining of the clay are discussed. In each section, a comparative description of technological aspects from other Greek sites and sites of the Early Neolithic of the Balkan Peninsula is included.5

4.2 METHODOLOGY

Study of the ceramic material from Nea Nikomedeia involved two seasons of fieldwork (four and five months respectively) at the Verroia Museum, where the material is stored. The Early Neolithic pottery from the following sites was also examined: Achilleion (Larisa Museum), Soufli Magoula (Larisa Museum), Magoulitsa (Volos Museum), Nessonis I and Nessonis II (Volos Museum), Servia-Varythimidis (Fiorina Museum), Anzabegovo (Anza) and Vrsnik (Stip and Skopje Museums), as well as a selection of whole vessels from Veluska and Porodin (Bitola Museum).

2 For an explanation of the headings used in TABLE 4.3 see n. 1, above.

3 For Early Neolithic sites of the Balkans see Nandris 1971; 1977, fig. 4.

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 59

4.2.1 Macroscopic Examination of the Material

During macroscopic examination of the material at the Verroia Museum the following form was completed for each sherd.

1. Registration of sherd: Excavation square/spit; Ware category: red/pink/beige/applied/impressed; Number of sherd: each sherd was given an individual number.

2. Shape: Drawing: a right-hand section was drawn for each sherd; Rim/base diameter: it was calculated by adjusting the sherd to a set of concentric circles (the radius was increasing 0.5cm).

3. Surface colour: The Munsell Soil Colour Chart notations of both interior and exterior surfaces were recorded.

4. Fabric: Hardness: the surface was scratched with nail/glass/steel point; The colour of sub-surface was recorded; Frequency of inclusions: sparse/common/abundant; Size of inclusions: very small (not visible with the naked eye)/ medium (smaller than 1 mm)/large (larger than 1 mm); Sorting: well sorted/ill-assorted; Identification of inclusions: the following types of inclusions were identified during the field-work: calcite/quartz/mica/shell/grog/plant inclusions. For the other types of inclusions their colour and shape was recorded. Finally, the reaction of the fabric with dilute hydrochloric acid was recorded.

5. Firing: Presence/absence of black core in the cross-section; Presence/absence of firing clouds on the exterior surface/interior surface/base.

6. Surface coating: Present/absent; Area covered: all surface/upper part; Quality of coverage: adequate/poor/wash-like cover; Mica particles: few/many/absent; Traces for the method of application of coating: wiping/pouring/ dipping; The same information was recorded for both exterior and interior surfaces.

7. Burnishing: Area covered with lustre: whole surface/upper part; Quality of lustre: good/medium/poor; Direction of burnishing strokes: horizontal/diagonal/vertical/ absent; The same information was recorded for both exterior and interior surfaces.

8. Decoration: Apart from the description of the different types of decoration, typical examples of impressed and applied decoration were drawn.

9. Manufacture: All sherds were examined for any traces of the primary manufacturing techniques and the methods used for the application of bases and lugs.

6o PARASKEVI YIOUNI

4.2.2 Microscopic Examination of Thin Sections In examining the sections a fairly simple procedure was followed. The colour of the matrix in plane polarizing light (PPL) and the number of voids were noted. Differences in colour between core and surface were also briefly noted.

In examining the inclusions the aim was to note the major classes with a general percentage per total fabric using observational comparisons with percentage charts such as are used in sedimentary studies. For all classes of inclusions the identification was made with a note as to its relative abundance, size, angularity and sphericity. The latter two qualities were judged using the guide given in the Atlas of Sedimentary Rods (Adams, Mackenzie and Guilford 1984).

Special attention was given to the technological data that can be ascertained from the examination of thin sections. In more detail, the following aspects were examined:

1. The processes involved in the preparation of the clay body prior to manufacture. In order to investigate this, the relative amount as well as the shape and size of rock, mineral, argillaceous and organic inclusions, were examined.

2. The method used in the manufacture of ceramics. In order to investigate this, the presence of voids, their orientation as well as the orientation of non-plastic inclusions were recorded (see Rye 1976, 1981; Woods 1982 b, 1984-5).

3. The nature of surface treatments. The characteristics of surface treatments as seen under the microscope, were recorded.

4.3 MANUFACTURE OF VESSELS

The vessels from Nea Nikomedeia were carefully finished, hence marks of the primary manufacturing techniques were usually obliterated by subsequent treatment. From the large sample studied, only a small number of sherds have such marks. Apart from two miniature vessels which were fired without being scraped, marks of primary manufacturing techniques are restricted to the interior surface of the vessels. Sherds with fresh breaks can give information on the manufacturing techniques used, and on the method of attachment of later additions such as bases and lugs. Macroscopic examination has shown that two primary manufacturing techniques were used: coiling and pinching.

4.3.1 Coiling The majority of coiling marks are found on rim sherds (20 sherds); only one base and one body sherd with lug are included in the sample. In all cases two or three superimposed coils, 10-20 mm wide, were visible on the interior surface of the sherd. Coils were thoroughly connected to one another, i.e. no traces of division between the coils were visible. This is attested by the rarity of sherds with horizontal fracture along the smooth junction of coils. In all, only two such sherds have been found.

Coil building may be detected in thin sections, the most obvious signs being join voids resulting from inadequate bonding (Woods 1982 b, photo 22). Orientation of inclusions may also be indicative of the manufacturing techniques (Woods 1984-5). Despite the careful examination of the thin sections, coil joins were not present, nor was the orientation of inclusions diagnostic.

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 61

4.3.2 Pinching

Evidence that pinching was used as a primary manufacturing technique comes from two miniature beige, unslipped vessels. Both vessels were left unscraped, so that their surfaces are marked with adjacent grooves (plate 8 0). In one of them, impressions of finger nails are visible on the interior wall. Since the pots were not properly finished, they were less adapted to stand the firing. One of them was cracked vertically into two parts; one exterior and one interior.

Despite the fact that these miniature vessels were the only examples with pinching marks it seems probable that pinching, a simple and fast manufacturing technique, was more widely used. The paucity of sherds with manufacturing traces can be explained by the fact that pinching marks are more easily obliterated than those left by coiling (Shepard 1 956, 184).

Ethnographic research has shown that pinching is a widespread method, used for the manufacture of whole pots or for the modelling of the lower part of larger vessels; the upper part being made by coiling (Rye 1981, 70]. It seems that a similar process was used by the Nea Nikomedeia potters. This is suggested by the fact that coiling marks were found only in one base sherd, in contrast with the 20 rim sherds which had such marks.

Technological analysis of the pottery from other sites has shown that both coiling and pinching were known, and were mainly used in combination for the manufacture of the Neolithic vessels. Among the sites where the combination technique was used for vessel manufacture are Early Neolithic Sesklo (Wijnen 1981), Middle Neolithic Sesklo (Kotsakis 1981), Middle Neolithic Lerna and Franchthi (Vitelli 1974). It should be pointed out however that as far as coiling is concerned, two different techniques can be distinguished. In Nea Nikomedeia potters were placing the coils one on top of the other. The same technique has been used by the potters of Vrsnik,* and the potters of the Middle Neolithic East Macedonian sites (Yiouni 1994). In the other technique the coils were placed not on top of, but next to each other. This technique was used in Early and Middle Neolithic Sesklo, Middle Neolithic Lerna and Franchthi and in Middle Neolithic Makri in Thrace (Yiouni 1994).

4.3.3 Mat Impressions A number of the Nea Nikomedeia bases have mat impressions on their exterior surface. These impressions could have been made when the finished vessels were left to dry on mats. Alternatively, such mats could have been used for supporting and rotating the vases, in order to facilitate their building (Vitelli 1984, 1 19; Rye 1981, 63). This is because when a pot becomes too large to be shaped in the potter's hands, it has to be set on a surface. Porous surfaces, such as mats, are preferable because when a pot is rotated against a hard plane, its bottom is weakened by friction and may crack or bend. Similar mat impressions have been found at a number of sites: Early Neolithic Servia (Wijnen 1979), Anzabegovo (Anza) I & II (Gimbutas 1976, fig. 64), in sites from the Veluska-Porodin complex (Simoska and Sanev 1976, figs. 59 & 58), at Sitagroi (Renfrew 1973, fig. 122) and Makri (Yiouni 1994)·

4 Observation of the author.

62 PARASKEVI YIOUNI

4.3.4 Scraping After the primary building had finished, the pots were scraped smooth. Scraping marks are present in only a few sherds. These are series of parallel lines, sometimes forming adjacent groups with different orientation. Despite the rarity of these marks, scraping is recognized by the quality of contour and the uniform thickness of the vessels' walls. Pots were scraped when they were still plastic: grit drag-marks, formed by dragging grains with a scraping tool on a hard clay surface, are absent. Scraping can be done with smooth-edged tools such as bone and wood, hard-edged tools such as pieces of flint or other rocks, or with a toothed or serrated device such as a shell or pot-sherd (Rice 1987). Some of the scraping marks, seen on the Nea Nikomedeia vessels, look as though they were made with a comb-like tool, probably shell or a fragment of another pot.

Ethnographic research has shown that the paddle and anvil technique is sometimes employed to complete the union of coils. This technique produces overlapping or closely spaced depressions, left by the anvil, on the interior of the pot (Rye 1981, fig. 70). Such marks were not present in the Nea Nikomedeia vessels and it seems that this technique was not used by the potters, unless the characteristic depressions were obliterated by subsequent scraping.

4.3.5 Application of Bases and Lugs The majority of Nea Nikomedeia vases have ring or flat bases; only a few of them are round bottomed (see figs. 5.2 1-5.23). After the vessel was built, a round base was modelled by scraping away the excess clay and smoothing the surface. Some flat and a few ring bases were modelled in a similar way.

Most often however, for the manufacture of ring and flat bases, a disk or a coil was added to the bottom of the vessel. These additions are visible in many broken sherds (plate 8 b). The coil was restricted to a number of well made and easily distinguished ring bases, whereas the disk was encountered in both flat and ring bases. Actually, it is this disk which is modelled in order to make a ring or flat base. The fact that the disk is thicker than the bottom of the vessel, suggests that it was added partly for modelling the base, and partly for strengthening it in order to sustain the weight of the vessel. All three methods described above (smoothing, addition of coil or disk), were encountered even in the earlier pots from Thessaly (Wijnen 1981) and the Péloponnèse (Phelps 1975).

Lugs were applied as thick rounded bulbs of clay which were pressed onto the surface of the vessel. Additional wet clay was smeared on to cover and seal the joint. The bulb was then pierced. In all pots, lugs were made from the same fabric used for the manufacture of the whole pot.

For the larger lugs, extra precaution was taken to secure them safely onto the body. An oblong plug was added to the bottom surface of the lug, which went through the wall of the vase (plate 9 0). In all, five such examples have been found. A similarly attached handle has been found in Elateia.5 This method of handle attachment was very common in Middle Neolithic Makri (Yiouni 1994). Similar 'thrust handles' are known from the Early Bronze Age in Greece and Anatolia (Weinberg 1962, 175).

5 In the stratum with the early painted pottery - Weinberg 1962, 175.

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 63

Lugs were added after the vessels had been scraped. This is attested by the pots in which lugs have broken off the body at the point of attachment. In such pots, the surface below the lug is well smoothed. Bases were also added at a similar stage (after scraping). After the addition of lugs and bases, the unslipped pots were burnished so as to acquire the characteristic compact and smooth surface. They were burnished when the surface was still yielding but not too plastic. This is apparent from the fact that the medium and large sized inclusions are usually levelled on the surface, being visible only in the fresh breaks. Troughs made by the burnishing tool are rarely visible.

4.4 SURFACE COATING

Two different surface coatings were used by the Nea Nikomedeia potters: red-brown slip and pink coating. Their characteristics are described below.

4.4.1 Red-Brown Slip The red-brown slip adheres well to the body since flaking or peeling is almost absent. Generally it has a sufficient covering power to conceal the body, but vessels with a thin slip (wash) are also present. Slipped surfaces are carefully finished as no runs are visible, and the margin line of the slip, on the interior surface of closed pots, is more or less regular. Fine grooves produced when the slip is applied by wiping are absent. So the slip was applied either by dipping or pouring. Since dipping allows more control over coverage (Rye 1 98 1 , 4 1), it seems that this was the most widespread method, with pouring limited to the larger vases. After the application, the slip was burnished.

In all thin sections of red-brown slipped sherds, the slip appears as a straight, fine grained line, of uniform thickness in a single sherd (plate 9 b). It has a distinctive difference in colour from the underlying body and there is a clear demarcation between body and surface. The slip of painted sherds has a similar appearance under the microscope.

VARIATION IN COLOUR: Red-brown slipped vessels show a great variation in colour from one pot to the other. Colours range from red through red-brown, to brown; the lighter shades being more common. The Munsell Soil Colour Chart notations are given in table 4.4. These differences in colour could be the result of using different slips, or the result of firing the vessels under varying conditions. Thus, the brown colour of a slip can be attributed either to manganese, which acquires a brown colour when fired, or to iron oxides fired in a non oxidizing atmosphere (Shepard 1956, 39).

In order to clarify this, three thin sections were examined with the SEM. Two of the sections belong to red coloured sherds, the third comes from a brown pot. The results of the analysis are presented in table 4.2, ist, 2nd and 3rd samples. As can be seen, red and brown slips have the same composition, and the amount of manganese is negligible in all cases.

To verify the above, brown and red slipped sherds were refired in oxidizing conditions for 40 minutes at 8oo° C. After the refiring, all sherds had a clear red colour, and it was

64 PARASKEVI YIOUNI

Table 4.4: Surface colour of the plain vessels from Nea Nikomedeia.

Pottery Munsell Colour

Red-brown slipped 2.5YR 5/6; 2.5YR 5/4; 2.5YR 4/8; 2.5YR 4/6; 2.5YR 4/4; 2.5YR 3/6; 2.5YR 3/4; 5YR 5/6; 5YR 5/4; 5YR 4/6; 5YR 4/4; 5YR 4/3; 5YR 3/4; 5YR 3/3; 5YR 3/2

Pink coated 10R 5/6; 10R 5/4; ior 4/6; ior 4/4; 7.5R 5/6; 7-5R 5/4

Beige unslipped ioyr 7/3; ioyr 7/4; ioyr 6/3; ioyr 7/4; 7.5YR 7/4; 7.5YR 6/4; 7.5YR 6/2

Red-brown unslipped 2.5YR 4/6; 2.5YR 5/4; 2.5YR 4/4; 5YR 6/6; 5YR 6/4; 5YR 5/6;5yr 5/4; 5YR 4/6; 5YR 4/4

impossible to distinguish the previously brown sherds. It seems safe then to conclude that brown and red sherds were coated with similar slips; the difference in colour being the result of firing. Since there is a great variation in shades from one pot to the other, it seems that these differences were the result of varying conditions from one firing to the other, rather than being intentional attempts by the potters to produce light red or darker vessels.

The slipped surface of each vessel is fairly uniform and consistent in colour, although some darker or lighter spots, due to firing clouds, are present. A few sherds (1% of red- slipped sample) however, show greater variation in colour. The exterior surface of each sherd appears red, light brown, dark brown and black in places. The interior surface has either the same mottled appearance, or its slip has one of the exterior shades.

It is difficult to determine if this variation was intentional. The different coloured patches are mixed and they do not follow any decorative pattern. It should be noticed however that such variations, on a single vessel, are absent among the other wares (i.e. pink coated, painted, impressed and uncoated vessels).

This group of sherds should be differentiated from the characteristic Variegated' or 'buntpolierte' vessels, which are commonly found in the Thessalian sites (Chourmouziadis 1971, figs. 26d and 25b). These vessels have triangular, tongue-like and other decorative motifs, which have been made by covering parts of their surface during firing. The mottled vessels from Nea Nikomedeia appear buff and monotonous, when compared with these Thessalian examples.

A final remark should be made about the red-brown mottled vessels. These, in the 1962 report (Rodden 1962, 284), were referred to as "brown, self coloured, burnished ware" (groups D and Ε in Rodden's classification). During the fieldwork it was not easy to distinguish, by macroscopic examination alone, whether these vessels were slipped or not. Their surface was covered in large parts by firing clouds. For this reason, a number of thin sections were examined with the polarizing microscope. All examined sections were slipped.

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 65

PREPARATION OF SLIP:

Analysis of the three thin sections with the SEM has shown that slips have a significant clay content (table 4.2, ist, 2nd and 3rd samples), in contrast with the small amount of iron oxide. The clay matrix of one sherd, made from fabric B-i (see section 4.6.3), was also examined (table 4.2, 4th sample). The amount of iron oxides present in the matrix is comparable with that found in the slip (9% in the slip; 8.2% in the clay matrix). This suggests that the slip was made from a ferruginous clay, similar but more fine-grained than that used for the manufacture of the pot. Fabric B-i, when fired, acquires colours ranging from brown-beige to red-brown. Since no colouring material has been added to the slip, the difference in colour between slip and clay matrix can be explained as resulting from the fine particle size of clay minerals and other inclusions in the slip. Moreover, the iron oxides present in the slip were better oxidized during firing, and acquired clearer colours. No other clay matrix was examined under the SEM. However, pétrographie analysis has shown that all clays used for the manufacture of Nea Nikomedeia vessels are highly ferruginous. It is probable then, that it was a general practice for the slip to be made from clays used for the manufacture of the vessels, without the addition of colourings. The same technique was used by the potters in Middle Neolithic Sesklo (Kotsakis 1981). Use of a slip which does not contain added colorant is not uncommon (Shepard 1956, 67).

Red-brown slips contain mica grains. Their amount however, is not constant in all vessels. Almost 80% of the studied sherds have slips with a few mica grains. The remaining sherds are coated with a mica dusted slip which has a glittering appearance.

There are two alternative explanations for the presence of plentiful mica in the micaceous slip. Either ground mica was added to the slip, or mica occurred naturally in the raw material used for slip. Taking into account the suggestion that slip was made from the same clays used for the manufacture of vessels, one would expect the clay used for the manufacture of pots with the micaceous slip to be richer in mica. Comparative study of thin sections of sherds with micaceous and non-micaceous slip, has shown no differences in the amount of mica present in the clay matrix (nb. this study is based on an estimation of the inclusions present in the matrix, and not on a point count analysis). Thus, the addition of ground mica seems to be the most probable explanation. Judging even from broken sherds, the addition of plentiful mica had a highly decorative effect.

4.4.2 Pink Coating In all thin sections of pink coated sherds, the pink coating appears as a thin irregular line of non-uniform thickness (plate 10). The coating consists of rounded and subrounded dark red particles of varying size, connected with a 'fluid', fine grained paste, most probably clay. Apart from the rounded particles and the clay, the coating looks free from other inclusions. There is not a distinct demarcation between body and surface. Instead, the red particles are pressed into the matrix, making grooves in it.

According to the ceramic literature, slip is a suspension of fine clay particles in water, with or without the addition of colouring material (Shepard 1956, 67; Rye 1981, 41). Seen under the microscope, slip appears as aline of uniform thickness, with no obvious particles of the colouring material, but with a clear demarcation line between surface and clay matrix (Woods 1982 h, 16, photo 28). Although the red-brown slip compares well with this description, it is clear that the pink coating cannot be identified as a slip. In fact, the features

66 PARASKEVI YIOUNI

of the pink layer, as seen under the microscope, distinguish it as a mineral coating (pigment) made from crushed colorant and clay (Woods 1982 b, photo 30).

Macroscopically the pink coating is difficult to distinguish from a slip. It adheres well to the body, and it is not crusty. Only the following features can differentiate pink coated and red-brown slipped vessels. First, few large (up to 1 mm across) particles of colorant are visible with the naked eye in most of the pink coated sherds. Second, in vessels which have lost part of their burnishing lustre, the pink coating appears powdery and rough. It is only occasionally that colour comes off when the hand is rubbed on the surface of these vessels. It should be stressed though, that it is under microscopic examination that the pink coating is clearly distinguished from a slip.

IDENTIFICATION OF COLORANT:

Compared with the Munsell Soil Colour Chart, the colour of pink coated sherds ranges from 7.5R 5/6-5/4 to 10R 5/6-5/4 (see fig. 4.4). This 'cherry-red' colour is usually connected with haematite (red iron oxide: Fe2O5). One thin section of a pink coated sherd has been examined with the SEM. As can be seen from table 4.2, 5th sample, haematite has been detected amounting to 20%, but the high percentage of clay (Slo2, A12o3) - amounting to 7 1 % - shows that pure haematite has not been used for the pink coating. Instead, it seems that a ferruginous sandstone has been used as a colorant.

METHOD OF APPLICATION: Most pigments are mixtures of colorants, fine clay, water and a binder (Shepard 1 956, 71). Clays slow the settling of particles in the mixture and enhance the flow and adhesion of pigment. The binder is an organic medium which improves spreading quality and binds the paint. The use of a binder is a common practice, but not universal with pre-wheel potters.

Pigments can be applied either before or after firing. The pink coating of the Nea Nikomedeia vessels has been applied before firing. This can be recognized from the thin sections where the particles of colorant are pressed, by burnishing, into the clay matrix, thus making grooves in it. This would have been impossible on a hard fired surface. Moreover, refiring of two sherds under oxidizing conditions, for 40 minutes at 800o C, did not alter their appearance. As the pink coating was applied before firing, the presence of a binder cannot be detected because, if present, it is burned out during the firing.

After applying the pigment, the ground colorant will remain soft and powdery unless it is burnished before firing, or sintered during firing (Shepard 1956, 32). Sintering in iron oxides occurs at temperatures between 920o C and 950o C. Pink coated vessels from Nea Nikomedeia were fired at temperatures lower than 800o C (see section 4.4.5). Therefore, the permanence of pink coating can be attributed to burnishing. The quality of burnishing lustre of the pink coated sherds varies from high to poor.

In a recent study of haematite-coating' in Late Bronze/Early Iron Age vessels from South England (Middleton 1987), two methods were used for the application of coating in modern replicas. First, a thick slurry of crusted haematite was burnished into the leather hard surface. Second, crushed haematite was applied as a dry powder to the wet clay surface followed by burnishing at the leather hard stage. Both methods produced coatings which microscopically are similar to the Nea Nikomedeia pink coating (Woods 1982 b, photo 30; see also Rigby et al. 1989, figs. 5 & 7).

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 67

USE OF MINERAL COATINGS:

Apart from the graphite-painted pottery, other vessels with mineral coatings are the so- called Crusted Ware, reported from the Late-Final Neolithic period. The term Crusted Ware is applied to all forms of decoration employing a thick red or white crusty pigment which adheres badly to the vessel (Phelps 1975, 310). It is generally considered as a post- firing decoration, and the red coating has been identified as haematite or as iron rich (Jones 1986, table 9.6a).

The pink coated ware from Nea Nikomedeia is not comparable with the Crusted Ware. The pink coating has been applied before firing, it adheres well to the body and it is not crusty. Moreover, it covers the whole surface of the vessels, whereas in Crusted Ware the pigment is usually used for decorative patterns.

A coating technique similar to that used by the Nea Nikomedeia potters, has not been reported from any other early Neolithic site of the Balkan area. Gardner (1978) reports that haematite has been identified, by chemical analysis, in the slip of Early Neolithic vessels from Achilleion. The surface of such vessels is described as being powdery and the colour as fugitive. A number of these sherds were examined, macroscopically, by the present author. None of them had the powdery/rough surface of the Nea Nikomedeia pink coated sherds, and haematite particles were not visible with the naked eye. Thus the macroscopic analysis suggested that the Achilleion potters were using a clay slip which contained very fine particles of haematite, rather than coating the vessels with a crushed mineral pigment.

In the 1962 report for Nea Nikomedeia the pink coated sherds were referred to as "the Pink Slipped Ware, characterized by a uniformly pink slip applied as a wash to the surface" (Rodden 1962). Pink slipped sherds were also reported by French for a number of Neolithic sites in West Macedonia (French 1967, 1970). The results of the macroscopic examination, by the present author, of the surface material collected by French from the sites which are situated near to Nea Nikomedeia are presented below. The surface collection is kept at the University of Thessaloniki.

Trilophos: One pink sherd from this site is very similar to the pink coated sherds from Nea Nikomedeia. Its surface has the same colour (10R5/6) and the same powdery appearance. Even its fabric looks, macroscopically, very similar to fabric group C from Nea Nikomedeia (see section 4.6.5). Thin sectioning of the sherd was impossible since it belongs to a reference collection. Trilophos and Nea Nikomedeia are situated 8 km apart and they share the same geology (see fig. 4.1). It is highly probable then that the potters from the two settlements were using the same clay beds and were employing similar techniques. The possibility that some kind of pottery exchange existed between the two settlements cannot be excluded.

Rizarion, near Edessa: A number of red coloured sherds (colour of surface: 2.5YR 6/8) have a very powdery surface and their colour comes off very easily. It seems that a similar technique was used for the production of this coating, but the colouring material and fabric appear different from those used in Nea Nikomedeia. Rizarion and Nea Nikomedeia are situated 30-35 km apart.

Pink slipped sherds are also reported from Yiannitsa Β (Chrisostomou and Chrisostomou 1993), a site which is situated 30-35 km ne of Nea Nikomedeia. Macroscopic examination

68 PARASKEVI YIOUNI

<υ

cd

.si

11

11

2S cd ^

cd ̂ O Ο Ο

|1 S3 S3 co en υ c •&^ Ο C *o ®

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 69

of the material indicates that their colour was produced by slipping rather than by covering with a mineral coating. In contrast, some of the Red-Pink coloured vessels and Red on Red painted pots from Early Neolithic Servia appear similar to the Nea Nikomedeia coated pots. Turning further north, similar sherds were noticed, by the present author, among the Early Neolithic material from Anzabegovo (Anza) and Vrsnik. Moreover, in Lakavica (near Stip in south Yugoslavia), an animal figurine has been found which shows typological and stylistic similarities with figurines from Nea Nikomedeia (Nandris 1968). It is covered with a pink paint, which macroscopically is very similar to the pink coating from Nea Nikomedeia.6

Thus, macroscopic examination indicates that, apart from Nea Nikomedeia, the technique of covering the vessels with a mineral coating was also used in other Early Neolithic sites. The available information suggests that this technique was employed in sites in West Macedonia and South Yugoslavia, being absent from the Thessalian region. This conclusion of course has to be verified by microscopic examination of a sample from the sites mentioned above.

4.5 FIRING

It seems that the Nea Nikomedeia pots were fired in open bonfires. This is recognized by the appearance of the pottery, which has all the signs of a bonfire, uneven in temperature and atmosphere: i.e. firing clouds, and a range of surface colours, both among different pots, or even on a single vessel.

Apart from the appearance of the pottery itself, open bonfire firing is implied by the absence of any kiln structures. Remains of bonfires when preserved, may be indistinguishable from remains of fires lit for heating or cooking, unless they are accompanied by a large amount of ceramic wasters. Vitelli reports that after an experimental firing, only a slight grey ash stain remained (Vitelli 1974, 2, 29) After two years, nothing remained at all.

At Achilleion large-sized circular hearths situated in courtyards, are reported as having been used for the firing of pottery on a communal basis (Gimbutas 1989). Such hearths have been found in almost every excavated level (apart from phase I and Ha). It should be noted however, that firing wasters are reported only for the hearth in phase Illb (Gimbutas 1989, figs. 4.25b and 4.29).

At Nea Nikomedeia a large number of firing remains and hearths have been found which, apart from cooking and heating, could have been used for firing vessels. It is also possible that some of the pits, which contain ashes and black earth, had served a similar purpose.

The atmosphere during the firing usually supported incomplete oxidation. The majority of the fresh breaks in pots show a dark core between two light coloured layers. The relative proportions of core and surface layers vary, sometimes within one vessel. In most cases however, the light coloured margins are very thin (1-2 mm). In monochrome, impressed and pottery with applied decoration, completely oxidized cores are rare. In painted vessels however, oxidized cores are more common, and firing clouds are almost absent.

6 Nandris pers. comm.

7O PARASKEVI YIOUNI

ESTIMATION OF TEMPERATURE:

Temperatures attained by bonfire firing generally range between 650-9000 C, but higher temperatures may occasionally be reached (Rice 1987; Woods 1982 a).

A rough estimation of the maximum temperature at which the vessels from Nea Nikomedeia were fired, can be obtained from the clays containing calcium carbonate grains. Calcium carbonate when heated at temperatures between 700o C and 850o C, begins to decompose into CO2 and CaO. Pottery with calcitic inclusions, fired above 700o C, is subject to disintegration, due to the hydration of CaO. Calcium hydroxide, which has a larger volume, exerts pressure on the fabric and causes a cone-shaped piece to spall from the wall leaving a white grain at the apex (Rye 1981, 114).

Among the clays used by the Nea Nikomedeia potters, fabrics A and C are rich in calcitic inclusions; their size varying from fine to coarse (see sections 4.6.2 and 4.6.5). Both fabrics were commonly used: 72% of the plain vessels were made from them. Lime spalling however, is absent, even in the pots containing large calcitic inclusions. It should be pointed out that, after refiring at 800o C, lime spalling was noticed in all sherds (3 sherds) made from fabrics A and C. Therefore, it can be suggested that the vessels from Nea Nikomedeia were fired at temperatures not exceeding 750-8000 C. This temperature range can be taken only as a rough estimation, since firing is a complicated process and the final result is affected by many factors: temperature, duration of firing, atmosphere of firing, type of clay, pre-drying of pots etc.

Turning for comparison to the pottery from other Greek and Early Neolithic sites, it can be seen that the estimated temperatures range from 750- 1050o C; 750-8500 C being the consensus (Jones 1986, table 9.6b).

The relatively low firing temperatures are in accordance with the scarce evidence for kiln structures in the archaeological record. One kiln is reported from Late Neolithic Olynthos, in Chalkidiki (Mylonas 1929). The identification of this structure as a kiln however, is questioned on chronological and technological grounds (Vitelli 1974; Jones 1986, 777).

Another example of a pottery-firing site is reported from Late Neolithic Dimini (Chourmouziadis 1979). This is alow, circular structure (1.15 m diameter), built with clay, thick tiles and flat schist fragments. The structure preserves its original height (30 cm), so it is clear that it was not a domed oven but a fire pit. The excavator suggests that this pit has been used by a specialized workshop, since inside and outside the pit (in a radius of 3-4 m) abundant sherds with stylistic similarities have been found.

In contrast with the negative evidence for kiln structures from Greek neolithic sites, kilns have been found in other Balkan countries. The earliest examples so far, are dated to the last stages of the Starõevo culture (Starõevo IV). Four kilns have been found at Círcea (Nica 1978). They consist of a single fuel chamber with an opening, served by a lateral fuel chamber (Ellis 1984, fig. 48). One kiln had two adjoining firing chambers, both served by the same fuel chamber. Similar simple kiln structures have been found at a number of Late Neolithic sites.7

7 VinCa, Turdas, Dude§ti, Boian Cultures: see Ellis 1984.

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 71

4.6 PETROGRAPHIC EXAMINATION

4.6.1 Introduction A sample of fifty-four thin sections from monochrome and impressed and pottery with applied decoration, has been examined (table 4. 1). Painted sherds were studied by Hodges (Biernoff 1969, Appendix A). In order to give a coherent picture of the pottery from Nea Nikomedeia, a summary of the pétrographie examination of the painted vessels will be included at the end. Most of the thin sections of the painted vessels were available at the Institute of Archaeology, University College, London, so that microscopic comparisons with sections of monochrome and impressed pottery were possible.

One of the aims of carrying out the pétrographie analysis, was to confirm information obtained by macroscopic examination. It has been proven that a macroscopic analysis using general features such as colour, hardness of clay, size, amount and colour of non- plastic inclusions and their reaction with dilute hydrochloric acid, can be a reliable method for a general separation of different fabric categories. Based on these features, the fabrics were separated, macroscopically, into five different types (fabrics A-E).

With only one exception, pétrographie analysis confirmed this separation. The exception concerns fabric B. Twenty thin sections, which according to macroscopic examination were considered to be made from fabric B, were examined with the polarizing microscope. From these, fifteen were clearly made from the same fabric. The remaining five however, had to be separated as a different fabric type. In the following description, the first group is called fabric B-i and the second fabric B-2. The relative frequency of the different fabric types, used for the manufacture of Nea Nikomedeia vessels, has been calculated from data obtained from the macroscopic analysis. Therefore in all subsequent comparative analyses, fabrics B-i and B-2 are not distinguished as different, but they are considered as a single group (fabric type B).

SIZE VARIATION OF NON-PLASTIC INCLUSIONS: The following size gradation of the non-plastic inclusions was used in this work: very fine inclusions are <o.i25 mm; fine XX125 mm and <o«3 mm; medium >o.3 mm and <o«5 mm; coarse >o.5 mm and <i mm; very coarse >i mm.

Before we proceed with the description of the different fabric types, a remark concerning the size of the non-plastic inclusions should be made. Every fabric type contains a variety of classes of non-plastic inclusions. Thin sections belonging to the same fabric group contain the same variety of inclusions, in comparable amounts. Moreover, in thin sections of the same fabric type, every class of inclusions shows a uniformity in shape and contour. However a variation on the size of the inclusions has been noticed among thin sections of the same fabric type. Some thin sections contain inclusions showing a great variation of particle size, ranging from very fine to coarse, whereas in others most of the inclusions are smaller than 0.3 mm. Based on this feature, thin sections can be separated into two categories: fine textured and medium-coarse textured.

It should be stressed that within these categories, a variation in the size of inclusions is still noticeable. For example amongst the fine textured sections of fabric type A, the following variations have been observed:

72 PARASKEVI YIOUNI

ι. The majority of inclusions are very fine, few fine inclusions are present; 2. The majority of inclusions are very fine, few fine and few medium are present; 3. Many inclusions are very fine, many are fine, some are medium. Equally among the

medium-coarse textured sections of the same fabric type (fabric A), there are examples with few coarse inclusions, in others the coarse inclusions are more numerous, or some very coarse grains are present.

Apart from fabric A, the same variation was noticed in fabric B-i and fabric E. The same variation in the size of inclusions was noticed in fabric C, but the fine textured sherds contained more medium sized grains than those of the other fabrics. All thin sections from fabric B-2 and fabric D are medium-coarse textured.

During macroscopic examination, sherds were distinguished, according to the size and frequency of their larger non-plastic inclusions, into fine, medium and coarse textured groups. Thus, there is a slight discrepancy between the microscopic analysis, where only two groups were distinguished, and the macroscopic examination. Although macroscopic analysis is not a very detailed and precise technique, it can be applied on a large number of sherds and gives information on general trends in the sample. The results of the macroscopic analysis on the quality of paste, will be incorporated in the description of the different pottery wares and vessel forms.

4.6.2 Fabric A This is the most common fabric type, being used for the manufacture of 60% of the plain vessels (fig. 5.40). It is characterized by a soft (2.5-3 Moh's scale), calcareous beige-grey matrix. Non-plastic inclusions amount to 25-30% of the matrix.

Subrounded quartzite and limestone, and subangular to subrounded calcite fragments, constitute the major inclusions (plate 11 a). A few molluscan fragments are also present. Among the quartzite fragments, examples with a little mica (mainly muscovite) and iron mineral inclusions, are quite common; occasionally, hornblende is present as well. Quite often, quartzite inclusions are surrounded by a calcitic rim. Inclusions with an overgrowth of calcite are rare. All inclusions mentioned so far, apart from shell fragments, are usually of low sphericity, and show a variation in particle size ranging from very fine to very coarse. Shelly molluscan inclusions are usually of medium size.

Subangular quartz fragments of low sphericity are abundant; the majority measures less than 0.50 mm. Rounded iron mineral inclusions, of high sphericity, are frequent (<o.O5 mm). They often appear in association with quartz and mica inclusions.

Some subrounded to rounded serpentine and greywacke fragments are scattered throughout the matrix (plate 1 1 b: one serpentine fragment can be seen on the left part of the field; one greywacke fragment on the right). They measure up to 1 mm across, but generally are less than 0.05 mm. Greywacke fragments consist of abundant polycrystalline quartz, a little mica and a few iron mineral inclusions. Occasionally, feldspar inclusions, mainly microcline, are also present. The matrix between the inclusions is not clearly visible (200 x).

Rounded schist and subangular to subrounded slate fragments, are present as minor inclusions. Schist fragments, brown and yellow in colour, are heavily stained by iron minerals (plate 13 a: the large inclusion on the right part of the field). They have a high content of

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 73

mica inclusions, parallel to the foliation. Due to the very fine size of the individual grains, no other minerals could be distinguished (200 x), apart from a few quartz and feldspars. Slate fragments consist of numerous tiny quartz grains and some iron mineral inclusions oriented parallel with the laminated structure.

Of the finer inclusions, mica fragments (mainly biotite) are quite common. Plagioclase and microcline feldspar, and pyroxene inclusions, are present in minor amounts. In a few thin sections, occasional grog-like inclusions are visible.

4.6.3 Fabric B-i Fabric B-i (and B-2) has been used for the manufacture of 20% of the Nea Nikomedeia plain vessels (fig. 5.40). It is a hard fabric (4.5-5.5 in Moh's scale), characterized by a red- brown, compact matrix. Inclusions amount to 25-30% of the matrix. Subangular to subrounded feldspar and quartz fragments, of low sphericity, constitute the major inclusions (plate g b). They can measure up to 1 mm, but usually are less than 0.5 mm across. Among the feldspar inclusions, plagioclase predominates, but some microcline grains are present as well. Fine grained quartzite inclusions (fine to medium) are common.

Rounded to subrounded fragments of andésite (0.125-2 mm), often surrounded with a ferruginous rim, are common (plate 1 2 a: the two large inclusions in the centre of the field). They consist of angular and subangular quartz and plagioclase feldspar, angular and subrounded biotite and hornblende; the inclusions are set in a very fine grained felspathic groundmass.

Rounded iron mineral inclusions of high sphericity (0.125-0.5 mm), are frequent. They often appear in association with quartz, feldspar and abundant mica inclusions.

The finer inclusions consist of abundant mica (mainly biotite), less abundant hornblende, and relatively rare pyroxene grains. Occasionally, medium-sized biotite (lath-like) fragments are present.

4.6.4 Fabric B-2 This fabric type is characterized by a hard (4.5-5.5 in Moh's scale), red-brown clay matrix. Inclusions amount to 25-30% of the matrix. Subangular to subrounded quartzite fragments of low sphericity, predominate (plate 12 b). They range in size from fine to very coarse. Among the quartzite fragments, examples with very few mica and iron mineral inclusions are present. Subangular quartz fragments are the second most common inclusion (usually <o.O5 mm across). Some plagioclase and microcline feldspars are present. Rounded iron mineral inclusions, in close association with quartz, mica and feldspar fragments, are scattered throughout the matrix (plate 12 b: on the left corner of the field). In some cases, a clear halo effect is noticeable.

The smaller inclusions comprise pyroxene, hornblende, muscovite and biotite. Occasionally, large biotite lath-like fragments are also present.

4.6.5 Fabric C Fabric C has been used for the manufacture of 12% of the Nea Nikomedeia plain vessels (fig. 5.40). It is characterized by a soft (2.5 in Moh's scale), beige-red clay matrix. Inclusions amount to 25% of the matrix. The fabric contains abundant subrounded limestone and

74 PARASKEVI YIOUNI

subangular to subrounded quartzite and quartz fragments (plate 10 ö). Calcite is virtually absent, and one shell fragment was found only in one thin section. Among the quartzite fragments, examples with little mica and iron mineral inclusions are present. Quartzite and limestone inclusions show a great variation in size, ranging between 0.125 and 2 mm; quartz fragments are usually smaller than 0.05 mm.

Rounded shale, schist and subangular to subrounded greywacke fragments are also present. Their frequency is not constant in all thin sections. In nine of the fifteen sections, only a few fragments are present. The remaining thin sections contain abundant fragments of these rock types. Despite these differences shale fragments are the most common, in all cases.

Shale fragments are heavily stained with iron minerals (dark brown in colour), and they contain a large amount of mica inclusions. Greywacke fragments consist of numerous quartz and biotite inclusions and relatively infrequent calcite and plagioclase inclusions (plate 13 b: inclusions in the left corner of the field). They are yellow and brown in colour, and they have a characteristic streaking appearance, due to a ferruginous cement between the individual inclusions. They are different from those described in fabric A: they contain monocrystalline quartz grains, instead of polycrystalline ones. Moreover greywacke fragments from fabric A do not contain calcite inclusions, they lack the streaking appearance and the matrix between the individual inclusions is not as clearly visible as in the greywacke fragments of fabric C. Schist fragments are similar to those described in fabric A (large inclusions in the middle of the field in plate 13 û).

Rounded iron mineral inclusions are frequently found, sometimes in association with subangular quartz and mica inclusions. The smaller inclusions consist of abundant mica grains and a few feldspars.

4.6.6 Fabric D Fabric D has been used for the manufacture of 3% of the Nea Nikomedeia plain vessels (fig. 5.40). It is characterized by a hard (5.5 in Moh's scale) red clay matrix. Inclusions amount to 20% of the matrix.

Rounded schist and subrounded quartzite fragments, sometimes accompanied by a few mica and iron mineral grains, constitute the major inclusions (plate 13 b). Subangular to subrounded quartz inclusions are also abundant. Subangular to subrounded greywacke fragments are common. A few rounded shale fragments are scattered throughout the matrix. Rounded sandstone fragments, resembling the greywacke, are present as minor inclusions in some of the thin sections. Schist, greywacke and shale fragments are similar, in appearance and composition, to those described in fabric C.

All inclusions, apart from quartz, show a great variation in size ranging from fine to very coarse. Quartz fragments are usually less than 0.05 mm.

Some rounded iron mineral inclusions (<o.5 mm), sometimes in close association with quartz grains, are present in Fabric C. Of the smaller inclusions, mica grains (mainly muscovite) are abundant. A few fine-sized feldspars are also present.

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 75

4.6.7 Fabric Ε Fabric Ε has been used for the manufacture of 4% of the Nea Nikomedeia plain vessels. It is characterized by a soft (2.5-3 *n Moh's scale), grey-cream matrix. Inclusions amount to 25-30% of the matrix.

Rounded serpentine and subrounded quartzite fragments are the major inclusions. Subangular quartz (<o.o5 mm) grains are common (plate 140). Serpentine and quartzite fragments range in size from fine to coarse. Among the quartzite fragments, examples with a little mica, iron minerals and rare hornblende inclusions, are found. Some subrounded greywacke and a few schist fragments, similar in appearance and composition to those described in fabric A, are present.

Among the smaller inclusions, fine mica fragments (mainly biotite) are abundant. Some pyroxene and a few feldspars are scattered throughout the matrix. One of the thin sections contained a few subrounded calcite inclusions.

4.6.8 Petrographic Analysis of Painted Pottery In all, 66 thin sections were studied by Hodges and his analysis is included in BiernofPs thesis which had as its subject the painted pottery from Nea Nikomedeia (Biernoff 1 969, Appendix A). The material was separated into three groups: Groups A to C. Here they will be referred to as Groups 1 to 3, in order to avoid confusion with the fabric types described above. A few of the thin sections of painted pottery could not be included in any of these groups (Hodges 1969), and they were reported as non-conformers.

GROUP 1 (42 THIN SECTIONS): This resembles fabric type A, the fine textured category. Included in this group are all sections of the Standard Ware, ten sections of the Porcelain Ware and four sections of the White on Red Ware. The description of these Painted Wares can be found in the following Chapter (section 5.3.2).

GROUP 2 (7 THIN SECTIONS): This fabric type has not been encountered among the monochrome, impressed and pottery with applied decoration. It is extremely fine textured, and appears to be almost without inclusions. In a highest magnification (200 x), quartz, quartzite, mica, a few limestone and pyroxene inclusions are visible. A few shell fragments were present in some of the sections. All thin sections from Group 2 belong to the Porcelain Ware.

group 3 (8 thin sections): This fabric type has not been encountered among the plain, impressed and pottery with applied decoration (figs. 5.40, 5.42). The predominant inclusions are quartzite; quartz and limestone inclusions are also common. A few feldspar, pyroxene and mica inclusions are present. The inclusions range in size from very fine to medium. From the eight thin sections of Group 3, seven belong to the White on Red Ware and one to the Porcelain Ware.

From the eight thin sections considered as non-conformers, four were available at the Institute of Archaeology. From these, two were identified, by the present author, as belonging to the medium-coarse textured fabric A, one to fabric B-i and one to fabric D.

76 PARASKEVI YIOUNI

4.7 DISCUSSION OF THE RESULTS OF PETROGRAPHIC ANALYSIS

4.7.1 Diversity of Fabrics Pétrographie analysis has shown that a variety of fabrics were used by the Nea Nikomedeia potters. In all, eight different fabrics were distinguished.

This picture is in accordance with the situation encountered in most Greek Neolithic sites. At Servia, chemical analysis has shown that there is at least as much variation in composition among the Early Neolithic coarse wares, as there is among the remainder of the Early Neolithic material (Jones 1986). Analysis of the pottery from Franchthi,8 Corycian cave,9 Kitsos cave10 and Makri11 has produced similar results. Exceptions exist, as for example at Vassilika (Sikalidis et al 1983) where the pottery was made from only two different clay fabrics.

4.7.2 Source Location of Non-Plastic Inclusions and Clay It should be pointed out that, despite the variety of fabrics used at Nea Nikomedeia, 62% of the plain pottery and 90% of the painted were made from the same fabric (fabric A). Macroscopic examination suggests that fabric A was also used for the manufacture of stamps, altars and the majority of figurines. This preference implies that this clay source was near the settlement, and/or that it was the most commonly encountered over a wide area. It is also probable that this clay was more workable.

Location of the clay sources requires a detailed geological survey of the area. Considering the dramatic changes of the landscape surrounding Nea Nikomedeia, due to extensive alluvium accumulation, it was thought that a survey of the surface clay sources was highly unlikely to give any results of relevance to the conditions at that time. Similar attempts at other Neolithic sites, with more stable geomorphology, are rather disappointing. With few exceptions,12 the collected clays did not compare with the archaeological material.

By using a detailed geological map of the area, the non-plastic inclusions present in the clays were compared with the local rock outcrops. All inclusions present in the fabrics are consistent with the prevailing geology of the area which is comprised of igneous, metamorphic and sedimentary outcrops. From the geological map (fig. 4.1), it can be seen that within a minimum distance of 7 km from the site, there are abundant calcareous deposits, outcrops of serpentine, igneous rocks (andésite and trachyte), red and yellow schists, calcareous conglomerates and sandstones. Similar types of minerals and rocks were found in all thin sections of fabrics used by the Nea Nikomedeia potters. It should be stressed that in order to confirm the similarity of the minerals and rocks present in the vessels' fabrics with those available locally, future petrological research of the local rock fragments has to be carried out. For the present, all that can be said with certainty is that in the archaeological material there are no inclusions inconsistent with the local geology.

8 Early-Final Neolithic: Tones 1086. 9 Late Neolithic: Courtois and Dimou iq8i. 10 Middle-Final Neolithic: Courtois 1981.

11 Middle Neolithic: Yiouni 1Q04. 12 Vassilika: Sikalidis et al 1983; Sesklo: Overweel 1981.

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 77

Fabric B-i is the weathering product of andésite,15 since it contains inclusions of andésite, abundant feldspar and quartz grains, numerous fine mica, hornblende and pyroxene inclusions. Moreover, the absence of calcareous inclusions, serpentine, schist and greywacke particles, imply that this clay source was deposited near the parental rock. At a distance of 7 km north-west of the site, there are outcrops of trachyte and andésite. Therefore, one may speculate that fabric B- 1 was collected near this area.

Of special interest is the fact that two different fabrics, not encountered among the plain pottery, were used for the manufacture of Porcelain and White on Red painted ware. Both pottery groups are rare, each amounting to 4% the of painted pottery (see section 5.3.2). Moreover, White on Red pottery shows similarities, in decorative motifs and syntax of decoration,14 with the painted pottery from the Early Neolithic sites and the pottery from the recently excavated site of Yiannitsa Β (Chrisostomou and Chrisostomou 1993), a site which is situated 35 km ne of Nea Nikomedeia.

It has been stated above, that all non-plastic inclusions present in the fabrics of Nea Nikomedeia are compatible with the local geology. The low frequency of these wares though, and their different fabric, raises the question of imports. The proximity of Nea Nikomedeia and Yiannitsa Β supports such an explanation. A pétrographie examination of the ceramic material from Yiannitsa Β is necessary in order to check the validity of this hypothesis. However, the following arguments are against this explanation:

1. Preliminary macroscopic examination, by the present author, of the vessels from Yiannitsa Β indicated that their fabrics differ from those used in Nea Nikomedeia.

2. Three legged White on Red painted vessels which are present in Yiannitsa Β (Chrisostomou 1994, figs. 7 & 8) are lacking from the Nea Nikomedeia painted examples.

3. White on Red and Porcelain Wares were also made from fabrics common in plain, Red on White, impressed and pottery with applied decoration. Moreover, they were in use throughout the occupation of Nea Nikomedeia and were evenly distributed in the settlement.15

Thus, on the basis of the available data the White on Red and Porcelain vessels from Nea Nikomedeia should be considered as local products. The different fabric groups used for their manufacture are in accordance with the practice of the Neolithic potters to use, simultaneously, a variety of fabric types.

The use of fabric Group 2 for the manufacture of Porcelain Ware could have a technological explanation. Porcelain vessels have a careful finish and a superior quality of burnishing. It could be suggested that potters from Nea Nikomedeia selected a fine textured clay exclusively for the manufacture of this superior quality ware. The difficulty of working with such a clay should be stressed. Many fine textured clays are too plastic for manufacturing vessels and they tend to crack during firing and drying (Rye 1981, 31).

15 I. Freestone, pers. comm. 14 K. A. Wardle (Ed.) Nea Nikomedeia II: The Finds

and the Place of Nea Nikomedeia in its SE European

Context (BSA Suòòl. forthcoming). 15 Ibid.

78 PARASKEVI YIOUNI

4.7.3 Addition of Non-Plastic Inclusions Inclusions found in a fabric may be naturally present in the parent clay utilized, or they may be added by the potters as filler or temper. Addition of temper could facilitate the building and firing of vessels (Rye 1981, 31). Moreover, different types of temper are better suited for the varying functions of pots (see section 6.3.1). A few types of inclusions such as grog and abundant plant fragments, are generally considered as deliberate additions. For the other categories of non-plastic inclusions it is difficult to determine, due to the complexity of the geology, whether they were naturally present in the clay or added. On this basis, three characteristics can be used to determine their origin: size, contour and amount of inclusions. Deliberately added inclusions tend to be of uniform size and are usually large, they generally have an angular contour and their proportions can vary from one pot to the next.

As has been seen above (sections 4.6.2-4.6.7), all inclusions of Nea Nikomedeia fabrics are subrounded to rounded, and they show a great variation in particle size. Moreover, different types of inclusions are present in similar amounts in all thin sections of the same fabric. The only exception concerns the schist fragments present in fabric C, whose amount is not constant in all sections. However, since schist fragments are subrounded in contour and they show a great variation in particle size, they cannot easily be considered as deliberate additions. Shell fragments present in fabric A, cannot be regarded as temper either, because they are always present in small amounts. In most cases where shell is added as temper, it is present in relatively large quantities.16 The occasional grog fragments found in fabric A show that the addition of this temper type was limited.

During macroscopic examination of the material, charred plant inclusions were noticed in all fabrics. Such examples were limited, not exceeding 1-2% of the sample. This percentage was similar in all pottery groups: slipped, coated, uncoated, impressed and pottery with applied decoration. Also, fabrics with vegetable matter were not restricted to any particular vessel form or to coarse textured fabrics. The amount of plant fragments is usually small, but some heavily tempered sherds are present as well.

During microscopic examination of the thin sections, charred plant fragments were noticed in three of them (1 from fabric Α-i and 2 from fabric B-i). Two of the sections contain a few plant fragments, whereas the third was full of elongated voids and vesicles (plate 14 b). In painted pottery, no plant inclusions were found (Hodges 1969).

Thus, examination of thin sections of all fabric types indicates that tempering was very limited among the Nea Nikomedeia potters. Although this result should be verified by more detailed textural analysis (Le. point count of particle size distribution, and calculation of average grain size), it seems unlikely that such analysis will alter this conclusion.

4.7.4 Refining of Clay

Despite the fact that tempering was very limited, some of the fabrics show variation in the size of their inclusions. Since the larger inclusions were not added, it seems that their absence from the fine textured categories is the result of a clay refining process. Ethnographic research

16 20-30% Rice 1987, 410.

THE EARLY NEOLITHIC POTTERY: TECHNOLOGY 79

shows that for the removal of the coarser particles present in the clay, two methods are usually used: sieving and settling (Rye 1981, 37). For settling, the clay is mixed with water to a sufficiently thin consistency to permit the coarser particles to settle to the bottom of the pit. By taking the clay from different levels of the pit (upper or lower), fine or medium- coarse textured vessels can be made from the same clay.

It seems that sieving was not used by the Nea Nikomedeia potters. Clays refined by sieving contain inclusions which have a uniform size. This is not the case with the Nea Nikomedeia vessels since, as has been seen in section 4.6.1, a clear distinction into fine, medium and coarse textured sherds is difficult to make. Such a situation could be explained by a settling procedure, which allows the broad separation into fine and medium/coarse textured batches of clay, but some mixing should be expected. The very fine textured clay, at the top of the pit, could have been used for slipping the pots.

The only thin sections which contain inclusions of a rather uniform size, are those of the Porcelain Ware (group 2). In this case however, it is difficult to distinguish if this is the result of a refining procedure, or whether the potters were using a naturally fine textured clay source.

P. Y.

Chapter 5 The Early Neolithic Pottery: Typology

(plates 15-17)

5.1 INTRODUCTION

The ceramic assemblage of Nea Nikomedeia includes decorated and plain monochrome vessels. Plain vessels form the majority of pottery, amounting to 96% of the ceramics found at Nea Nikomedeia. Decorated pottery, which amounts to 4% of the ceramic material, can be separated into pottery with painted, impressed and applied motifs. Painted pottery is considerably more common, forming 88% of the decorated vessels (fig. 5.2). Impressed pottery and pottery with applied decoration amounts to 9% and 3% respectively (fig. 5.2). All three decorative types were in use simultaneously.

Painted pottery has been studied by Biernoff and Washburn. In his Ph.D. thesis, Biernoff presents the shapes, fabrics and decorative patterns of the pottery (Biernoff 1969). Moreover, Biernoff has made drawings of every painted sherd. Washburn's study concerns the organization of the decorative motifs depicted on the painted vessels (Washburn 1983, 1984). A preliminary study of the decorated and plain pottery was also published by Dr. Rodden (Rodden 1962).

In the pottery with applied decoration the decorative patterns range from complicated modelling of human faces, silhouettes and animal snouts to simple linear motifs. All face vessels have been studied by Nandris in his Ph.D. thesis (1968). Apart from the description of the decoration, information on the shapes, fabrics and spatial distribution is also included. This information will be presented by him in the next volume.1

Plain, impressed pottery and vessels with simple applied motifs were studied by the present author. For the classification of the material the following attributes were used: composition of paste (fabric type), surface treatment and shape. Using one attribute at a time the pottery was divided into three main classes; smaller subsets were subsequently distinguished within each class. For example, the criterion for surface treatment in plain pottery was the presence of surface coating. Pottery was initially divided into coated and uncoated vessels and on the basis of surface colour each set was divided into smaller groups. Similarly, the other two classes were divided into smaller subsets.

As for the painted pottery and the vessels with complicated applied motifs, the information provided by the works presented above2 was treated with the same methodology used for the other pottery categories.

1 K. A. Wardle (Ed.), Nea Nikomedeia II: The Finds and the Place of Nea Nikomedeia in its se European Context (BSA Suppl. forthcoming).

2 i.e. Nandris 1968; Biernoff 1969 and Washburn 1983, 1984.

81

82 PARASKEVI YIOUNI

Chapter 5 has the following structure. Initially, the method employed for the sampling of plain pottery is presented. It is followed by the description of the ceramic material (plain, painted, pottery with applied decoration and impressed vessels) according to the surface treatment. Subsequently, the shapes of the vessels from Nea Nikomedeia are presented. Since sherds of plain pottery are far more numerous, the shapes of this pottery category are analytically described; then the vessel forms of the other pottery categories are comparatively presented. The description of the different fabric types used for the manufacture of the Nea Nikomedeia vessels has already been given in Chapter 4. In this Chapter a comparative study of the quality of paste and the percentages of the fabric types used for the manufacture of the different pottery categories is included. Finally, the chronological development of the pottery from Nea Nikomedeia is examined.

5.2 SAMPLING OF THE PLAIN POTTERY

During the excavations at Nea Nikomedeia, a large amount of Early Neolithic pottery was recovered. In fact the ceramic material found in the main excavation grid (squares Ai- M8) alone, amounts to some 140,000 sherds, the great majority of them being plain sherds (monochrome). Thus, it was necessary that a sample be taken from the plain pottery. In contrast with the plain pottery where a sample was taken, all recovered sherds with impressed, applied and painted decoration were studied.

The ceramic material collected during the 1 96 1 campaign was washed and the feature sherds (rims, bases and lugs) of the monochrome vessels were stored in the 'ferroia Museum, according to their excavation square/spit designations. From the decorated vessels, all sherds were stored.

The ceramic material revealed in the main excavation grid (squares A1-M8) during the 1 963 campaign, was washed and possible joins were made. The monochrome sherds were divided by Rodden into three main groups: red-brown (fine and coarse), pink (fine and coarse) and beige (fine and coarse). For every excavation square/spit, the sherds from each group were counted and their surface area was measured by the excavator. Subsequently, only the feature sherds were stored in the Museum according to their excavation square/ spit designations. The number of painted and impressed sherds present in each excavation square/spit as well as their surface area was also recorded. In contrast with the monochrome pottery, all decorated sherds were stored in the Museum. The ceramic material recovered in trenches TX, TY, TA, TB and TC was stored, unwashed, in the Verroia Museum.

In order to use the preliminary work carried out by the excavator, it was decided to take a sample from the pottery recovered in the main excavation grid (A1-M8). Sampling was not extended to the pottery recovered during the 1 96 1 campaign for the following reasons. First, the stratigraphie associations/correspondence of the ceramic material revealed during the 1961 (mainly Area L) and the 1963-64 campaigns were not clear.5 Secondly, due to the fact that in the 1961 and 1963-64 campaigns two different grid reference systems were employed (2 x 2 m square units in 1961; 4 x 4 m in 1963 and 1964), it was decided

3 Rodden pers. comm.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 83

to use the material from the 1963 main excavation grid only so that the sampling was based on units of comparable size.

The main excavation grid extends over a large part of the site, where plans of superimposed houses were uncovered. Sampling and comparative study of the pottery from different building periods was important for the examination of any changes through time. The ceramic material found in Trenches TX, ΊΎ, TA, TB and TC was not included in the sampling because the stratigraphic/cultural associations of their deposits were still under study by the excavator.

After consulting Mr. C. Orton, it was decided that Selective Sampling (or Stratified Random Sampling) was the most appropriate method for the case study. Stratification is a way of using prior knowledge to improve the precision of estimates, usually by minimizing the variability within each stratum. The results can be combined to give estimates for the population as a whole (Torrence 1978).

In Selective or Stratified Sampling the population is divided into a number of sub- populations (or strata). Then, a random sample is taken from each of these strata. The main excavation grid from Nea Nikomedeia was divided into 7 strata. Each area occupied by a building or a group of superimposed buildings was considered as a separate stratum. The excavation squares included in each stratum as well as the excavation squares sampled from each stratum, are given in table 5.1. The amount of sampled squares correspond to 25% of the total.

It is obvious that, in many cases, the boundaries of the strata are not clear cut, since a single excavation square is shared by two strata {e.g. excavation squares D2, C5, B8 etc.). This is common in cases where a square is bisected by a house-wall. In these cases the square was incorporated into one of the two strata. This decision is not so arbitrary, if we take into consideration that during the excavations at Nea Nikomedeia there was no distinction among the sherds found inside or outside a house-wall; all sherds present in an excavation square were referred to as bulk material from that particular square.

In sampling the excavation squares, a fairly simple procedure was followed. After examining the ceramic material stored at the Verroia Museum, a few excavation squares (B7, A5, Co) were eliminated from the sampling. These were squares that did not have clear designation indexes (i.e. not clear or missing spit designations). The ceramic material from the excavation squares Eo, Ei, E2, E3 and E4 was also excluded from the sampling, because parts of these squares were excavated during the 1961 campaign and their excavation was completed during the 1 963 season.

After this preliminary procedure, the excavation squares from each stratum were separated into two groups. One group consisted of excavation grids where, according to the excavator's measurement (number of sherds and their surface area), a large amount of pottery has been found. The second group included squares where less pottery has been recovered. One or two random samples, depending on the size of the stratum, were selected from each group. Each sampled grid was studied vertically (the material from all spits present in it). The distinction of the excavation squares into two groups was made in order to make sure that the sample will include grids which will facilitate subsequent spatial, functional and chronological comparisons of the ceramic material.

It should be pointed out that apart from the sherds from the sampled excavation squares, which were studied analytically, the ceramic material from all excavation squares of the main grid was examined during the field-work.

84 PARASKEVI YIOUNI

Table 5.1: Sampled excavation squares from the main excavation grid of Nea Nikomedeia (1963 season).4

Stratum Excavation squares Number Sampled grids

1 Coo, Co, Ci, Doo 8 (2): Coo, Do Do, Di, Eoo, Eo

2 Ai, A3, A2, A4, A5 13 (4): Bi, A3, C3, B4 Bi, B2, B3, B4, B5 C2, C3, C4,

3 A6, A7, A8, B6, B7 7 (3): A6, A7, B6 O6,O7

4 C7, C8, C9, Cio, D7 1 1 (3): B9, Cio, C7 D8, Dg, D 10, B9, E8 E9

5 D3, D4, D5, D6, E3 14 (4): D4, F6, D6, C5 E4, E5, E6, F4, F5 F6, B6, C5, C6

6 F7, F8, G6, G7, G8 8 (2): M7, G8 H6, H7, H8

7 J6, J7 , J8, K4, K5 17 (4): L4, K5, J6, L8 K6, K7, K8, L4, L5 L6, L7, L8, M5, M6 M7, M8

5.3 SURFACE TREATMENT

5.3.1 Plain Pottery On the basis of surface treatment, plain vessels can be separated into two groups: vessels with a surface coating and vessels without a surface coating (uncoated).

VESSELS WITH A SURFACE COATING

These vessels amount to 73% of the plain pottery (fig. 5.1). In open vessels both the interior and exterior surfaces have the same treatment. In closed pots the interior surface is uncoated; the exterior surface coating stops either at the rim, or it continues on the interior surface for a few centimetres below the rim. As has been seen in Chapter 4, two different surface coatings were used by the Nea Nikomedeia potters. These comprised a red-brown slip made from very fine textured clay and a pink coating made from crushed shale, which was burnished on to the vessel's surface. The range of surface colours of red-brown slipped and pink coated vessels is given in table 4.4.

4 The first column gives the number of the stratum; the second and third columns give the name and number of

the excavation squares present in each stratum; the fourth column gives the sampled excavation squares.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 85

Red-brown slipped ware: Red-brown vessels form the commonest category of the plain pottery, amounting to 46% of the sample (fig. 5.1). The pots are slipped in shades varying from red through red- brown to brown, with lighter shades being more common. Dr. Rodden (1962) distinguished the following groups based on these differences in colour: Light-coloured Slipped Ware, Red Slipped Ware, Dark Burnished and Brown Burnished Ware. However, as these differences in colour are the result of firing the vessels under varying conditions (see section 4.4.1), it was decided to incorporate all Rodden's groups into one category.

For the manufacture of red-brown slipped vessels all types of fabric were used, but in different proportions: 49% of the vessels were made from fabric A, 36% from fabric B, 5% from fabric C, 4% from fabric D and 5% from fabric Ε (fig. 5.57). It should be noted that fabric Β was mainly used for the manufacture of red-brown slipped vessels: 57% of the pots made from this fabric have a red-brown slip. As far as the quality of the paste is concerned, red-brown slipped vessels from all the plain pottery groups have the highest percentage of fine textured sherds at 69%, with 29% medium and 6% coarse textured (fig. 5.58).

All vessels were burnished. The quality of burnish ranges from moderate to good. Some of the fine textured sherds were heavily burnished.

Pink coated ware: Vessels with a pink coating amount to 27% of the plain pottery (fig. 5.1). This category coincides with Dr. Rodden's Pink Slipped Ware (Rodden 1962). For the manufacture of pink coated vessels all fabric-types were used. Their relative frequency was as follows: 55% of the sherds were made from fabric A, 5% from fabric B, 33% from fabric C, 5% from fabric D and 2% from fabric Ε (fig. 5.57). It should be pointed out that fabric C was mainly used for the manufacture of pink coated vessels since 73% of the pots made from this fabric have a pink coating. The quality of paste varies with 55% of the sherds being fine textured, 40% being medium textured and 5% being coarse textured (fig. 5.58).

All vessels were burnished. It seems however that the pink coating does not hold the burnishing lustre as well as the red-brown slip. In many cases only parts of the burnishing lustre are preserved.

UNCOATED MONOCHROME VESSELS Uncoated monochrome vessels amount to 27% of the plain pottery (fig. 5.1). The exterior and interior surface of open vessels as well as the exterior surface of closed pots are carefully smoothed. The interior surface of closed pots is roughly smoothed, apart from a strip below the rim which is more carefully finished. Most vessels are burnished. The quality of the lustre varies from moderate to poor and is inferior to that on coated pots. The great majority of uncoated vessels have a light beige colour. The remaining sherds are red-brown coloured. The Munsell Soil Colour Chart notations for beige and red-brown sherds are given in table 4.4. Separation of uncoated vessels into beige and red-brown groups is based on the colour of the fabric-types used for their manufacture.

86 PARASKEVI YIOUNI

Beige uncoated vesseh: Beige vessels amount to 24% of the plain pottery. Their surfaces have a clear light colour, except for some reduced grey patches. This category coincides with Dr. Rodden's (1962) Grey-Beige Monochrome Ware. The vessels are mainly made from fabric A at 88% of the sample, with 3% of fabric B, 3% of fabric C and 5% of fabric Ε (fig. 5.57). Well levigated pastes were regularly used. As can be seen from fig. 5.58, fine and medium textured vessels amount to 48% and 40% respectively. However, by comparison with the coated vessels, there is an increase in the proportion of coarse textured sherds (at 12% and 5% respectively).

Red-brown uncoated vessels: Only 3% of the plain pottery belongs to this category, which coincides with Dr. Rodden's (1962) Coarse Ware. Vessels are made from fabric D and fabric Β at 66% and 34% of the sample respectively (fig. 5.57). Fabric D, when fired, attains colours varying between red, red-brown, light brown and dark brown. Fabric B, when fired under oxidizing conditions, has a beige colour, but when fired in a reducing atmosphere attains a light brown colour. The proportion of fine and medium textured vessels is substantial at 50% and 30% respectively (fig. 5.58), but the proportion of coarse textured examples is larger than in any other ware, at 20%.

5.3.2 Decorated Pottery

painted pottery Painted sherds amount to 88% of the decorated pottery. On the basis of the colour of the decoration, two groups of painted pottery were distinguished. One group includes sherds with red patterns set on a white background. Sherds of the second group have white patterns on a red surface. Red on White (R/W) vessels form the great majority as they amount to 96% of the painted pottery (fig. 5.2).

Redon White Ware: According to the quality of finish, two varieties within this group were distinguished: the Standard R/W Ware and the R/W Porcelain Ware (Rodden 1962; Biernoff 1969). Standard Ware is considerably more common. It amounts to 92% of the painted pottery whereas only 4% belongs to the Porcelain Ware (fig. 5.2). The remaining 4% of the painted pottery belongs to the White on Red (W/R) Ware.

In Übe Standard R/W Ware the exterior surface is usually covered by a cream-beige slip, but in a few cases red patterns are applied directly to the light coloured fabric surface. The interior surface, which was painted only very occasionally, is covered with the reddish slip. The surfaces are usually burnished. It seems that the red paint retains its gloss better than the white, which now has a matt finish (Biernoff 1969). Both fine and medium textured sherds are included in this ware.

Porcelain Ware is differentiated by both its superior quality of burnish (porcelain-like finish) and the superior quality of the fabric used for the manufacture of a number of the vessels. As has been seen in Chapter 4, seven out of the eighteen thin sections of Porcelain Ware belonged to a very fine textured fabric group (see section 4.6.8). Porcelain vessels are

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 87

painted only on the exterior surface but, in contrast with the Standard Ware, both their exterior and interior surfaces are covered with a cream coloured slip.

White on Red Ware: (PLATE 1 5 a) Vessels in this group have both their exterior and interior surfaces covered with a red slip. Decoration is restricted to the exterior surface where the designs are painted with a white paint.

Among the painted sherds one polychrome example is also included (fig. 5.34, 1 2). The sherd is decorated with white and red paint over the light surface of the vessel in order to produce a three-coloured design (Biernoff 1969). There are also a few examples (10 sherds among the 1000 painted body sherds) which are decorated with a pink paint (ior 5/3 or 10R 5/4-6) on a lighter overall slip or wash (fig. 5.70, 8).

Syntax of decoration: Solid and linear elements were both used for the decoration of the painted sherds. Solid elements include triangles, squares and bands. Linear elements include straight, zig-zag and wavy lines and simple curved lines. The syntax of the Nea Nikomedeia decoration is difficult to reconstruct as the majority of sherds are fragmented. Washburn (1984) has reconstructed eight patterns (Classes A-H) which all come from the Standard R/W Ware. In the other two categories (Porcelain Ware and White on Red Ware), only fragmentary combinations of motifs could be recognized. It is clear, however, that differences in the selection and combination of motifs is evident among the three painted wares.5

Standard ware: This ware is unified by the bold, simple character of the decorative motifs. Eight types of pattern were reconstructed by Washburn comprising: two triangles (simple and double) placed back to back (Class A: fig. 5.21, 1; fig. 5.45, 6; fig. 5.46, 2; fig. 5.47, 4; fig. 5.49, 13 & 14; fig. 5.50, 3), a series of isosceles triangles pendant from the rim line (Class B: fig. 5.37, 1 1; fig. 5.50, 5), two rows of interlocking isosceles triangles (Class C and Class D: fig. 5.35, 4; fig. 5.35, 10; fig. 5.37, 2 & 4; fig. 5.45, 9; fig. 5.45, 13), interlocking diamonds (Class E: fig. 5.45, 3; fig. 5.36, 1; fig. 5.45, 2; fig. 5.46, 1 & 4) and 'step' designs descending from both right and left (Class F: fig. 5.36, 10; fig. 5.39, 1; fig. 5.45, 4; fig. 5.36, 8). Linear elements which are less common than the solid ones, include series of vertical zig-zags extending from the rim to base (Class G: see Washburn 1984, fig. III. 10) and series of wavy lines extending vertically or diagonally from the rim to base (Class H: fig. 5.38, 16; fig. 5.40, 2; fig. 5.40, 12; fig. 5.44, 12 & 13; fig. 5.45, 11; fig. 5.61, 1). It is clear that the decorative patterns are not restricted to any particular vessel form.

From the fragmentary elements, which are probably parts of the patterns presented above, the following should be mentioned: wedge-shaped elements which are attached to rims in several different ways (Group I: fig. 5.35, 1 1 & 12; fig. 5.36, 2; fig. 5.41, 2; fig. 5.45, 10; fig. 5.48, 8; fig. 5.49, 1 2), bold elements extending from rim or angled toward rim (Groups II & III: fig. 5.34, 1 1; fig. 5.36, 3; fig. £.37, 6; fig. 5.37, 10; fig. 5.38, 15; fig. 5.41, 10; fig. 5.43, 3; fig. 5.43, 13). Groups IV (fig. 5.45, 14; fig. 5.49, 15), V (fig. 5.51, 1), VI and VII (fig. 5.46, 9) are considered by Washburn as being parts of Class A. Other sherds

5 See also Biernoff 1969.

88 PARASKEVI YIOUNI

have horizontal lines (Group VIII fig. 5.41, 6; fig. 5.36, 7) or curved units, perhaps spirals (Groups X & XI: fig. 5.35, 6; fig. 5.41, 1).

Porcelain ware. Linear elements seem to be more common in this ware. They include thin lines around the rim (Group XII: fig. 5.34, 9), thin lines crossed at right and acute angles associated with square and triangular designs (Group XIII: fig. 5.35, 7; fig. 5.51, 4), thin parallel lines in series frequently bent at right angles (Group XIV: fig. 5.38, 1; fig. 5.51, 2 & 3) and series of open and cut-out designs below the rim (Group XI: fig. 5.34, 6; fig. 5.51, 5 & 6). Curvilinear elements are also present (fig. 5.51, 7).

White on Red: Instead of being negative copies of the Red on White pottery, White on Red sherds have a different character. Both linear and solid elements were commonly used in the decoration. They include diagonal parallel lines from the rim to base (Group XV: fig. 5.50, 7; fig. 5.50, 9), groups of wavy, parallel lines or two-three wavy lines, usually angled, sometimes attached to a triangle (Groups XVI and XVIII: fig. 5.39, 7, 1 1 & 13; fig. 5.40, 6 & 9; fig. 5.42, 1; fig. 5.35, 2), triangles and cut-out triangles (Group XVII: fig. 5-35» 5; FIG· 5·49> 14)· Combinations of straight and curved lines are also present (Group XIX: fig. 5.34, 13; fig. 5.38, 2; fig. 5.44, 6; fig. 5.46, 10).

APPLIED DECORATION In this section, pottery is presented according to the subject of the decoration. Vessels with human faces, animal snouts and human silhouettes will be analytically presented by Dr. Nandris in the next volume.6 However, in order to give a coherent picture of the pottery with applied decoration, a summary description is included in this section. Vases with simple linear motifs, sometimes related to the face vessels, are presented separately.

Face vessels: Ten sherds with representations of human faces, three with animal snouts and two with human silhouettes have been found. Most of them are very fragmented so that only parts of the decoration are shown.

In vessels with human faces the modelling starts a few centimetres below the rim and the facial features regularly shown are eyes, nose and ears. The mouth was only occasionally modelled and, in one sherd, arms were added. At least three of the examples are representations of male faces, since a beard has been added.

Two of the zoomorphic vessels have a snout modelled just below an everted rim. The third example has a prominent nose, a marked bump on the forehead and eyebrows which are curving out over the eyes. One of the sherds with human silhouettes carries a representation of a female figure; its upper part is missing. The pubic triangle and the navel are incised, whereas the legs are applied strips which are bent outwards at an angle. On the other sherd, the head and upper part of a torso are represented. The arms are zig-zag strips.

Surface coating is common among the face vessels. There are four red-brown slipped and two pink coated examples. All have traces of medium to poor burnishing.

Faces were sometimes modelled on small, thin-walled vessels (e.g. with a wall thickness of 2.5 mm), but the majority of them belong to medium-large, thick-walled vessels. The height of one such vessel was calculated to be at least 40 cm.

6 K. A. Wardle (Ed.), Nea Nikomedeia II: The Finds and the Place of Nea Nikomedeia in its SE European

Context (BSA SuppL forthcoming).

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 89

Vessels with simple applied decoration: Vessels with simple applied decoration were rather scarce. During the excavations of the main grid, only a small number of sherds (85) have been found. They are decorated with simple linear and solid elements. Linear elements include straight or curved raised bands, right or acute angled lines and more rarely, zig-zags (fig. 5.53). Solid elements include simple blobs and pellets (fig. 5.54).

The syntax of the decoration is difficult to reconstruct, since all the examples are fragmented. By comparison with the sherds with human silhouettes discussed above, it seems probable that some of the angled bands and zig-zags were parts of a similar decoration (fig. 5.53, 1 & 7). Combinations of linear and solid elements are suggested by two sherds, which have a blob set below a straight band (fig. 5.54, 5-6). Groups of blobs form another decorative pattern: on a number of sherds (5), two or three blobs are placed next to each other (fig. 5.54, 2-4). Moreover, sherds with single blobs and pellets are commonly found (fig. 5.34, 7, 13).

Decorative elements were usually placed somewhere on the belly of the vessels. This is suggested from the fact that among the 85 sherds, there are only seven rim sherds and four concave sherds coming from near the rim line. The remaining examples belong to body sherds. From the rim sherds it can be seen that linear motifs were usually set horizontally a few centimetres below the rim line (fig. 5.54, 7). In one case, however, a raised band starts at the lip and continues vertically along the collar of a neck jar (fig. 5.53, 2).

Red-brown slipped and pink coated examples are common among the vessels with simple, applied decoration, amounting to 34% and 22% of the pottery (fig. 5.60). They are usually burnished and the quality of burnishing ranges from poor to medium, but only traces of the burnishing are usually visible. Uncoated vessels are always smoothed but only a few of them have traces of burnishing.

IMPRESSED DECORATION Vessels with impressed decoration amount to 9% of the decorated pottery (fig. 5.2). Decoration is restricted to the exterior surface of the vessels. It consists of impressions of finger-tips and finger-nails and also of impressions made either by finger-pinching or with an instrument. Each vessel usually carries one type of decoration, but some sherds with a combination of two types of decoration have also been found.

Decoration was executed when the clay was still soft or leather hard. A number of vessels have a surface coating on their exterior impressed surface. In fact, 8% of impressed sherds are red-brown slipped and 5% have a pink coating (fig. 5.60). All examples have a low relief, so that the coating adequately covers the impressions. Depending on the vessel's contour (open/closed), the interior surface can be coated and burnished, smoothed and slightly burnished, or only smoothed. It should be pointed out that only medium and poorly burnished surfaces are encountered among the impressed sherds.

Syntax of decoration: Although most of the sherds are fragmented, enough information is still available to reconstruct the syntax of the decoration. The most common practice, which is encountered in all types of impressed decoration, is for the surface of the vessel to be subdivided into three zones: one on the belly of the vessel, which is covered with impressions and two horizontal zones around the rim and foot, which are left undecorated. By slightly burnishing

go PARASKEVI YIOUNI

the smoothed zones, the decorated belly stands out against a slightly darker surrounding area.

The smooth zone below the rim can be considered as a rule since it is present in 90% of the rim sherds. It ranges from 1.5 cm to 10 cm, with the most common group being between 2 cm and 4 cm (fig. 5.55, 2-9). An undecorated zone measuring up to 4 cm, is found in half of the base sherds. In the other sherds, the impressions continue up to the base and in a few cases, continue even below the base. Moreover, two sherds should be mentioned which are decorated only below the base (fig. 5.56).

An undecorated zone was sometimes left at the belly of the vessel as well. On a small number of sherds (3 examples) finger-pinched impressions surround an undecorated area. It is probable that this zone was restricted to this type of decoration, which produces the most rough and densely decorated surfaces and it is against such a background that an interplay of smooth and rough areas would have a more decorative effect.

Pottery decorated with finger-tip impressions. This is a rather rare type of decoration, since it is encountered in only 5% of the impressed vessels. The decoration consists of shallow circular, or slightly oval, impressions made when the finger-tips are pressed on the soft clay (plate 1 5 b). In a few cases the impression of the nail is also visible. Impressions are arranged slightly apart in parallel vertical and horizontal rows. In all cases the decoration is very carefully executed and the exterior surface is rather smooth. Vessels with a surface coating are common and almost half of them are red-brown slipped, or have a pink coating.

Pottery decorated with finger pinching. This is the most common type of decoration and it is found in 71% of the impressed pottery. The impressions were made with the thumb and fore-finger which are pressed together so that the clay was squeezed into bulges (plate 16 û, sherd in the lower left corner and plate 16 b' fig. 5.55, 1-5). Often one finger was pressed into the clay and then slightly moved to one side. Thus, a small depression, with a bulge at its side, was made (plate 17 a). It is probable that some of these impressions were made with instruments.

All vessels have a surface covered with dense horizontal and vertical rows of impressions. Some pieces were carefully decorated with the rows of finger-pinching being nearly horizontal or vertical (plate 16 û); other pieces were decorated with less care (plate 16 b). The size of impressions varies considerably; there are sherds with very small depressions and a coffee-bean-like relief and others with a very rough relief. In a few cases the finger- pinching of impressions is very small and appears to be made by a child. Similar impressions, however, can be produced by instruments. Vessels with a surface coating amount to 13% of this group.

Pottery decorated with finger-nail impressions: Sherds decorated with finger-nail impressions amount to 18% of the impressed pottery. They can be separated into two groups. The first group includes sherds decorated with repeated impressions of single finger-nails (fig. 5.55, 6-7; plate 17 b). On a number of sherds impressions were made with the finger-nail and with part of the tip of the finger as well. Thus, these impressions are wider and have an oblong shape. Sherds decorated with single finger-nail impressions amount to 13% of the impressed pottery. Impressions are usually carefully set in horizontal and vertical rows, but sometimes they are scattered randomly.

In the second group, the nails of both the thumb and fore-finger were pressed into the clay (fig. 5.55, 6; plate 16 û, the two sherds on the right). The pattern is similar to that made by finger pinching, but since the nails are not pressed together, the sherds look

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 91

smoother and lighter than the densely decorated, finger-pinched examples. A few sherds have an intermediate decoration, as one finger and one nail are pressed into the clay (fig. 5·55> 5)· Decorating the vessels with double nails was not common and only 5% of the impressed pottery belongs to this group. In all sherds the impressions are carefully executed. There are two pieces with very small impressions which were made either by a child or with an instrument.

Finger-nail impressions (single or double) are sometimes indistinguishable from impressions made with an instrument which has a straight edge. The latter are usually straight and longer, but a differentiation is not always possible.

Vessels with a surface coating amount to 33% of the vessels with double impressions and to 19% of those with single nail impressions.

Pottery decorated with the help of an instrument Sherds decorated with instrument impressions amount to 6% of the impressed pottery. The decoration ranges from simple scratches to circular, oblong, triangular and short, straight impressions. The instruments used for the decoration were probably bones, edged stones, reeds or dried plants. Sherds with a surface coating amount to 1 1 % of the sample.

Sherds decorated with scratches amount to 4% of the impressed pottery. This decoration is characterized by a poor selection of short, straight parallel lines, angled lines, crossed lines and zig-zags. The scratches are scattered on the vessel's surface without following any pattern. On one sherd for example, a few crossed lines appear 1 cm below the rim and below them, short, straight lines are sparsely distributed (fig. 5.55, 10-1 1).

The remaining sherds are very few (9 examples). From these, three examples are decorated with circular impressions, one with semi-circular and two with oblong impressions (fig. 5.55, 9). All examples were carefully decorated, having their impressions set in vertical and horizontal rows. In two cases, circular impressions were made with an instrument which had long parallel scratches on the exterior surface, probably a bone or reed. From the remaining sherds, two were probably decorated with small, triangular impressions, but their form is not clearly distinguishable. Lastly, one sherd with oblong impressions combined with angled lines has been found.

Finally, two bases with impressed decoration should be described. Their decoration is confined to areas which are not usually visible, as the impressions are set on the exterior surface of the base. Both examples are broken and almost half of the decoration is missing. However, it is clear that the decoration was not extended to the vessel's body (fig. 5.56). The designs are made up of very small impressions the size of a pin-head.

In the first base the decoration consists of a central floral motif surrounded by one circular and two arch-shaped designs. Three thin girdles are set on the edge of the surface (fig. 5.56, 1). The bold motifs are made from numerous impressions, while the thin girdles comprise a single row of impressions. The designs are carelessly executed. The decorative pattern of the second base is more simple. It consists of a number of straight lines radiating from a central point (fig. 5.56, 2). All the lines are made up of a single row of impressions.

Both bases are short and flat and their diameters 7 cm and 5 cm respectively. The impressed surfaces are smoothed, but the exterior walls of the vases are coated with a red- brown slip.

92 PARASKEVI YIOUNI

5.4 ANALYSIS OF THE SHAPES

5.4.1 Methodology The vessels were divided into three structural classes.

(A) Closed Vessek: These are vessels with a rim angle larger than go°. They have converging walls which are carried beyond the point of greatest diameter. Rim angle, which is defined as the angle between the rim line and the upper wall of the vessel (see also Phelps 1975), measures the slope of the upper wall. In examples where the vessels have strongly inverted, or everted rims, the change in the slope at the rim line is ignored.

(B) Open Vessek: These are vessels with rim angle equal to or smaller than go°. Their walls vary from spreading to vertical and stop at the point of greatest diameter.

(C) Ned jars: These are vessels with an inflection or corner point above the major diameter (Shepard 1956, 93)·

Each of these classes were divided into a number of subcategories. Closed vessels were divided into slightly closed bowls, whose rim angle ranges from 91-114° and into hole- mouthed jars where rim angle is larger than 115°. A third division is into askoid vessels which have an asymmetrical form, intermediate to slightly closed pots and neck jars.

Open vessels were divided into three subcategories of bowls: first, those whose rim angle ranges from 85-90°, secondly to those with rim angle from 60-84° ̂ d thirdly to bowls with rim angle smaller than 6o°.

Neck jars were divided, on the basis of the height of their neck, into jars with medium and high collars. The height of medium collars varies between 2.5 cm and 4.5 cm; high collars are taller than 4.5 cm.

Although the dividing line between the different subcategories is to some extent arbitrary, the study of the Nea Nikomedeia pottery has shown that vessels from different subcategories have different shapes. The span of rim angles in each vessel form is quite large (15-20°) in order to compensate for variations which are to be expected in hand-made pottery.

Rim diameter and the ratio of rim diameter to total height, are the two parameters used in this work to describe the size of the vessels. The ratio of rim diameter to total height is estimated only from whole or reconstructible vessels. Reconstructible vessels are rim fragments which preserve at least two thirds of the original height.

In the present work, pots with a diameter equal to or smaller than 10/10 of their height are characterized as deep, those with a ratio between 20/10 and 10/10 as medium and those with a diameter greater than 20/10 of the height, shallow. This ratio seems to have a chronological value, since studies of Early Neolithic material from the Péloponnèse (Phelps 1975) and Central Greece (Weinberg 1962) indicate that deep pots are more characteristic of the later stages of this period.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 93

There are three types of bases (ring, flat and rounded) and two main types of handles (string-hole and unpierced lug). These types are analytically presented in this Chapter, sections 5.4.10 and 5.4.9 respectively.

The Nea Nikomedeia vessels have direct rims. In a few cases slightly everted rims appear but they usually follow the direction of the upper walls, which are concave (fig. 5.12, 1; fig. 5.16, 5 & 8; fig. 5.35, 1; fig. 5.45, 4; fig. 5.46, 8 & 12; fig. 5.47, 4 & 6; fig. 5.47, 11; fig. 5.48, 7 & 12). Such examples are more common among the painted vessels. Lips vary in shape. They can be rounded, thinned either at the interior or on both sides, thickened at the interior or they are flat. The relative frequency of the different lip-types found in each vessel form of the plain pottery is given in table 5.2. Rounded lips are the most common lip-type, whereas thickened or flat rims are rare. It should be pointed out however, that in many cases a single rim sherd has more than one lip-type. Decorated vessels have similar lip-types. Moreover painted vessels occasionally have beaded lips (mainly the R/W Porcelain pots with rim angle from 91-1 14o: fig. 5.34, 5; fig. 5.36, 4; fig. 5.35, 7; fig. 5.35, 9; fig. 5.35, 13; fig. 5.43,9).

Before describing the different vessel forms, it should be noted that lids are absent among the pottery fragments. A few (1-4 examples) round-oval sherds, formed from broken vessels, are present in every excavation square. Their greatest diameter ranges from 4 cm to 6 cm and only scarcely reaches 9 cm (fig. 5.24, 13-14). As the Nea Nikomedeia vessels (excluding miniatures) have rim diameters larger than 6 cm (fig. 5.26 -5.3 1), it is clear that the round- oval sherds were not used as lids. It has, therefore, to be assumed that for the protection of the contents of pots other objects such as larger pottery fragments, stones, cloth or mats were used.

The use of the round-oval sherds is not clear as they do not have any obvious wear damage. They should be differentiated from similar shaped items which have a perforation hole at the centre and which were most probably used as spindle-whorls (fig. 5.24, 12).

Table 5.2: Relative frequency (%) of the lip-types represented on the Nea Nikomedeia plain vessels.

SHAPES TYPES OF UPS %

A B C D E

rim-angle > 115o 49% 29% 16% 3% rim-angle 91-1 14o 45% 19% 33% 3% 4% rim-angle 85-900 50% 21% 29% rim-angle 60-840 51% 17% 28%

rim-angle <6o° 69% 8% 27%

Key: Type A: Round lip Type B: Lip thinned from the interior Type C: Lip thinned from both sides Type D: Lip thickened from the interior Type Ε Flat lip

94 PARASKEVI YIOUNI

5.4.2 Neck Jars Neck jars amount to 9% (255 examples) of the plain pottery (fig. 5.62). The sample includes three reconstructible vessels. They are deep pots and their form resembles an ellipsoid which has the greatest diameter horizontal (fig. 5.5, 4). One of the pots has four vertically pierced lugs placed below the greatest circumference (fig. 5.5, 4). The rim diameter of neck jars ranges from 5-29 cm with the most common group being between 9-17 cm (fig. 5.25).

Height and shape of neck, as well as the way the neck is joined to the body, are the two criteria usually used for subdividing neck jars (MilojCiC 1971; Phelps 1975). Almost 70% of the sherds (178 examples) included in the sample of neck jars have their collars intact and attached to the vessel walls. These were separated, according to the height of their collars, into medium and high neck jars. The collars and upper walls are usually connected with a smooth curve. There are only two examples with an angular join (fig. 5.5, 1). Internally thickened angular joins which are so characteristic of the Early Neolithic Thessalian neck jars (MilojCió 1971, Figs. 3, 6, 7, 8 and Fig. 5; Gimbutas 1989, Fig. 5.46, 4) are not present.

The remaining sherds are broken and it is often impossible to attribute them to either of the two subcategories. As the data is fragmentary, the exact percentages of the occurrence of each subcategory cannot be considered as representative. For this reason only general estimates of their frequency will be given.

medium-collars: The height of collar ranges from 2.5-4.5 cm. Shorter necks have not been identified among the sherds which have their collars intact. They can be separated into two groups: Ai and A2.

Ai: Vessels included in this group have internally thickened, cylindrical necks. The beginning of the neck is mainly marked by a thickening of the wall and not by a clear angle or curve (fig. 5.4, 1-3). Necks are usually 0.3-0.5 cm thicker than the upper walls of the pots. These pots are intermediate in shape to the hole-mouthed jars and the neck jars. They are less common than the vessels belonging to the A2 subcategory.

A2: These are vessels with a definite neck (fig. 5.4, 4; fig. 5.5). Their collars are either concave or conical with the latter being more rare. Two examples of conical necks have an angular articulation with the body, while all concave necks are smoothly connected with the upper walls.

high collars: The height of collar ranges from 5-10 cm. More common are vessels with necks between 5 cm and 7 cm tall. The collars can be concave, conical or cylindrical with concave examples being the more frequent (fig. 5.5). In all, vessels with high necks are less numerous than those with medium necks.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 95

Among the 255 rim sherds included in the sample of neck jars, there are 27 broken examples which could be from either high collared neck jars, or small open vessels with flaring walls (fig. 5.6, 7-8). Indeed their shape is that of a cone with flaring walls but they are broken in such a way that it is impossible to distinguish to which type of vessel they belong.

If they are open vessels their profile is very similar to that of Late Neolithic cups. Their fabrics and surface coatings, however, are no different from those of Early Neolithic pots. Macroscopic examination of the fabrics of the Late Neolithic painted and monochrome pottery indicates that they are different in colour, sorting and firing from those used in the Early Neolithic Period. Three similar sherds are reported from the Early Neolithic Otzaki Magoula. They are classified as neck jars7 and are described as having "high rims of non- typical shape". It is therefore considered that these 27 sherds are open vessels or neck jars from the Early Neolithic levels and not intrusions from later strata.

5.4.3 Hole-Mouthed Jars (Closed Pots with Rim Angle Larger than 1 15o) Hole-mouthed jars are rare, amounting to 3% (85 examples) of the plain pottery (fig. 5.62). Their rim angles range from 1 15- 140o with the most common group ranging from 115-1250.

Four whole vessels have been recovered. Their contour is similar to that of neck jars (figs. 5.7 and 5.8). The transition from the lower to the upper part is rather abrupt. Their upper walls can be straight, slightly concave, or convex. String-hole lugs are present in two whole vessels (figs. 5.8, 5-6). In one case, the lugs are set vertically far below the greatest circumference. In the other vase, they are set horizontally above mid-height.

Rim diameters range from 10 cm to 32 cm with the most common group being between 13 cm and 24 cm (fig. 5.26). Basal diameters measure between 3.5 cm and 1 1 cm.

5.4.4 Askoid Vessels Askoid vessels have an asymmetrical shape resembling the leather containers used for carrying liquids, from which they take their name. Their upper part has a form intermediate to that of neck jars and slightly closed vessels, as one side converges and the other forms a short neck (fig. 5.9).

During the excavations at Nea Nikomedeia three whole vessels have been found, two of them in the central building (excavation square F6/1) and the third in square K7/1. A broken miniature vessel has also been found in square A6/1 (fig. 5.9, 3). Two of the pots have oval bases (fig. 5.9, 1), while the other two are supported by round bases.

Despite the relatively large sample of whole vessels, during the study of the pottery no rim or base fragments have been found which could be clearly attributed to this vessel category. This is due to the fact that, unless a very large part of the vessel is preserved, the fragment cannot be distinguished from a slightly closed pot or a neck jar. A number of oval bases have been found (see section 5.4.10), but as these bases are not restricted to askoid vessels they cannot be considered as definite evidence for the presence of askoi.

7 See MilqjCi<f 1971: Rim Form II, 3.

96 PARASKEVI YIOUNI

Askoid vessels are rarely found in Early Neolithic sites. The only other examples come from the Péloponnèse.8 In contrast with the Nea Nikomedeia vessels which are uncoated, the examples from these sites have painted decoration. Moreover, their shape is less asymmetrical and they usually have a strap-handle.

5.4.5 Slightly Closed Vessels (Closed Vessels with Rim Angle from 91-1 14o)

Slightly closed vessels constitute the commonest vessel form, comprising 33% (935 sherds) of the studied sample (fig. 5.62). Among them, seven whole vessels and sixty-nine reconstructible pots are included. Two of the whole vessels are miniatures (fig. 5.10, 1).

The rim angle of slightly closed vessels ranges from 91-1 14o. The greatest concentration of values is towards the lower limit: pots with rim angle varying between 91o and 110o amount to 28% of the total sample of monochrome pottery. Vases with rim angle between iii° and 114o amount to only 5%. As hole-mouthed jars (rim angle larger than 115o) constitute only 3% of the sample, sherds with a rim angle larger than 11 1° could have been included in the same category. However, vessels with rim angles up to 114o have similar shapes. Changes in form occur in pots with a rim angle greater than 1 1 50.

Slightly closed vessels usually have a spherical lower part which reaches the greatest diameter by mid-height (fig. 5.1 1). The inclining upper walls continue straight, or slightly concave, up to the rim. In a few sherds (9%) the upper walls are convex. It seems that lugs are common accessories to this shape, since there are seven examples with string-hole lugs set just below the greatest diameter (fig. 5.10, 6 and fig. 5.1 1, 3).

If whole and reconstructible vessels are considered as a representative sample of the slightly closed pots, then deep forms (fig. 5.10, 6-7; fig. 5.12, 4; fig. 5.13, 2) seem to be quite common, amounting to 42% of the sample. Apart from being more elongated, deep vessels are similar in shape to the medium ones.

Rim diameters range from 8-35 cm with the most common group being between 1 2 cm and 24 cm (fig. 5.27). Basal diameters measure between 3.5 cm and 1 1 cm.

5.4.6 Open Vessels with Rim Angle from 85-900 Despite the fact that the span of rim angle is restricted, this vessel category is very common, amounting to 25% of the monochrome pottery (708 rim sherds) (fig. 5.62). The sample includes four whole vessels and eighty-three reconstructible sherds.

The basic shape consists of straight upper walls and a spherical bottom (fig 5.15, 2). Less often the lower part is oval (fig. 5. 1 5, 4). The transition from the lower to upper body is usually gradual and smooth. In some examples however, the join of the two sections is marked by an abrupt bend in the wall (fig. 5.14, 1-3). Such examples are more common among the shallow vessels. Another variation of the basic shape is seen in a few vessels which have their upper walls concave (fig. 5.14, 5). Painted vessels offer more complete examples of this shape (fig. 5.45, 9; fig. 5.46, 5).

8 Nemea: Biegen 1975 Plate 63; Akrata: Phelps 1975.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 97

Vessels with medium depth predominate in the sample of reconstructible pots at 67%. Deep vessels (fig. 5.15, 4) are slightly more common than the shallow ones (fig. 5.14, 3-4) with proportions of 18% and 14% respectively. String-hole lugs have been found in only two vessels (fig. 5.15, 3). In both cases they are set around the greatest diameter.

Rim diameters range from 6 cm to 30 cm, the most common group being between 10 cm and 20 cm (fig. 5.28). Basal diameters range from 3 cm to 8 cm.

5.4.7 Open Vessels with Rim Angle from 60-840 This is a common vessel form amounting to 23% of the monochrome pottery (680 sherds) (fig. 5.62). The rim angle ranges from 60-840 but the greatest concentration of values occurs towards the upper limit: only 30% of the sample have rim angles smaller than 70o. The sample includes 120 reconstructible sherds and one whole vessel.

Vessels show a variety of shapes. Forms resembling sections of sphere, ovoid, ellipsoid and cone have all been found. Rounded forms are more common and only 9% of the vessels are conical. Despite the low frequency of the conical forms, the single whole vessel has a conical shape. It is a large pot supported by a flat angular base (fig. 5.17, 3).

Among the round vessels, ovoid bowls are the most frequent (fig. 5.16, 1,3, 5-8). A few of them have concave upper walls (fig. 5.16, 5-8). Ellipsoid forms are usually confined to shallow vessels (fig. 5.18).

Lugs are absent from the monochrome and decorated sherds of this vessel form. Nevertheless, their use should not be completely discounted, at least for the deep and medium vessels. Medium and shallow vessels predominate in the sample, at 58% and 40% respectively. Deep vases are rare, at 2%.

The rim diameter ranges from 8-30 cm with the most common group being between 10 cm and 26 cm (fig. 5.29).

5.4.8 Open Vessels with Rim Angle Smaller than 6o° This is a fairly uncommon vessel category amounting to 6% (170 sherds) of the plain pottery (fig. 5.62). The rim angle ranges from 30-590 with the most common group being between 55o and 40o. A large number of reconstructible sherds has been recovered (130), but only one whole vessel. The large number of reconstructible sherds is explained by the fact that as the great majority of vessels are shallow (89% of the sample), their original height can be estimated even from smaller fragments.

Apart from a few (12) conical examples (fig. 5.19, 2, 5, 10), all sherds have a round profile. Their shape resembles the shallow section of an ellipsoid (fig. 5.19, 1, 3-4, 6-9). A number of them have concave walls (fig. 5.20, 1-2, 4-6). Since there is not a single vessel with lugs and most of the pots are shallow, it seems that this vessel form did not have lugs as accessories.

The rim diameters range from 6-32 cm with the most common group being between 1 2 cm and 22 cm (fig. 5.30).

5.4.9 Handles Handles are present on a number of whole or reconstructible plain vessels. Most often, however, lugs have been broken off the vessel where they were attached. The great majority

98 PARASKEVI YIOUNI

of handles are simple string-hole lugs. Unpierced, knob-like and ledge lugs (pellets) were only occasionally used. Strap handles are absent.

Examination of the material showed that there is no correlation between the use of handles and the texture of clay fabrics. Instead, handles are commonly found on fine, medium and coarser textured pots. String-hole and knob lugs are used on both coated and uncoated vessels. In fact, the relative frequencies of coated/uncoated string-hole lugs are the same as those of coated/uncoated vessels. Red-brown slipped lugs amount to 44%, pink coated to 27%, beige unslipped to 29% and red-brown unslipped to 2%. All examples of ledge lugs (4 sherds) come from coated pots.

STRING-HOLE LUGS:

Depending on the size of the vessel, lugs range from tiny pellet-handles to large ones. The majority of them are of medium size. Most lugs have circular or oval sections, but a few examples with a conical section are also present.

Among the fifteen examples of whole/reconstructible vessels with lugs, only two open pots are included. This suggests that lugs were mainly attached to closed vessels. In fourteen of the reconstructible examples, four equally spaced lugs were set around the greatest circumference. In one vessel, only two lugs were used. Lugs were either vertically or horizontally pierced. The studied sample of monochrome pottery included 339 string-hole lugs and 1227 base sherds. From these it can be estimated that lugs were attached to 25- 30% of the Nea Nikomedeia vessels.

UNPIERCED LUGS: In this group knob-like and ledge lugs (pellets) are included. These are rare features of the Nea Nikomedeia vessels. From the main excavation area (main grid) only thirty six sherds with single knobs, or pellets, have been found (fig. 5.24, 7-9).

Most of them are small and as they are applied to thick-walled vessels it seems that they were decorative rather than utilitarian. In Nea Nikomedeia, groups of knobs or knobs associated with raised strips are common motifs of the applied decoration (section 5.3.2; fig. 5.37). Similar examples have also been found at other Neolithic sites (Milojflc 1971, Fig. XII, 7; Weinberg 1962 Plate 52a & 53c, 5; Gimbutas 1989, Fig. 5.65, 12-14).

Thus, from the sample of thirty six sherds, only five knobs and four pellets are thick enough to have been used as lugs. From these, four knobs have a circular section and the fifth is triangular. Pellets are 2-6 cm long and have an almost conical section.

SPECIALIZED HANDLES: In this category four strangely shaped handles are included (fig. 5.24, 10-11). They consist of a thick core which has the shape of a section of cone. This is covered by an external layer, whose top surface has an oblong slot which exposes the central core (fig. 5.24, 11). Two slots are cut along the centre of the two side surfaces.

In two of the handles the surface layer is pink coated, while the third example is red- brown slipped. In all cases, the coloured layer stands out against the uncoated core. One of the examples is attached to a rim sherd showing that the handle was set just below the rim (fig. 5.24, 10). Handles similar to those described above are not reported from any other Neolithic site of the Balkan area.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 99

5.4. 10 Bases

Nea Nikomedeia vessels usually have low ring and flat bases. Round bottomed examples are rare and only a few pedestal bases have been found.

ring base: Ring bases amount to 47% of the sample of base sherds. The height of ring bases ranges from 0.5-1 cm and only rarely reaches 2 cm (fig. 5.21). The concavity of the bottom is usually pronounced apart from some examples which look like flat bases with a slightly concave centre (fig. 5.21, 2, 6, 8).

flat base: 45% of the vessels have a low flat base. Two varieties of this type can be distinguished. In the first, vessels are supported by a low foot whose height ranges from 0.5-4 cm (fig. 5.22, 1-5). Seen only from the exterior (not a cross section), it cannot be distinguished from a ring base. In the second variety, the transition from base to body is marked by a slight concavity on the profile of the vessel (fig. 5.22, 6-10). This variety is slightly more common than the first one.

ROUNDED BASE: True round bottomed vases are almost absent (fig. 5.23). Most of the examples incorporated in this group are rather flat bottomed vessels with a rounded transition from the base to body. This type of base was encountered in 8% of the vases.

Ring, flat and rounded bases are found in all vessel forms. The base thickness ranges from 9-30 mm. It has been noticed that flat bases are usually thicker than ring bases, whereas rounded bases are comparatively thin, having the same thickness as the body walls. The majority of bases are almost circular, but a few oval and rectangular examples are also present (10 oval and 7 rectangular bases; fig. 5.9).

The diameter of circular bases ranges from 3 cm to 14 cm, the most common group being between 4 cm and 10 cm (fig. 5.33). The longest diameter of oval bases measures between 6.5-14 cm. All rectangular examples are broken, but seem likely to be longer than 5-6 cm.

Simple, short bases, such as those described above, are the common feature of all Neolithic sites of the Balkan region. Rare oval bases have been found in the Thessalian sites (Theocharis 1967, plate XXIVB; Milojclc' 1971, Fig. XIV, 1-4), Servia (Wijnen 1979, 193) and Vrsnik (Garasanin 1982, 90). It should be noticed however that vessels from the Early Neolithic sites often have elaborate bases with a wavy contour, quatrefoil and cruciform shape. Such examples are common in Anzabegovo (Anza) (Gimbutas 1976, Fig. 18; Garasanin 1971, Figs. 4 & 5), Círcea (Nica 1977, Fig. 7) and the Bulgarian sites.9 Rare examples have also been found at Podgorie (Lera 1983 Plate 1,6). Such bases are absent from Nea Nikomedeia, Servia and the Thessalian sites.

9 i.e. Asmaska, Muldava: Fortier 1981, Figs. 75, 85 & 88.

loo PARASKEVI YIOUNI

PEDESTAL BASES: Pedestal bases are a rare feature of the Nea Nikomedeia vessels. Only fifteen examples have been found during the excavations. These include two miniature 'egg-cups' which are the only examples where both the conical base and the supported vessel are preserved (Rodden 1962, 2, Fig. 15).

The remaining pedestals are broken off the vessels. The majority of them (1 1 examples) have compact cylindrical stems which open to a conical foot at the base (fig. 5.23, 4-6). There are only two whole pedestals and their heights are 6 cm and 8 cm respectively. One of the broken bases reaches 8 cm, suggesting that taller bases were used as well. Basal diameters range from 5-8 cm, with 5 cm and 6 cm being the most common.

Two short, carelessly made examples are also present. Their heights are 2.5 cm and 4 cm respectively. They look as though a round lump of clay has been pressed at the sides; a round opening at the base has been made by pressing in the thumb. Their basal diameters are 3 cm and 4.5 cm.

On the basis of the miniature 'egg-cups' and by comparison with examples from other sites, it seems that the pedestals supported open, carefully made vessels. The miniature vessels and two of the larger examples are unslipped. All the other pedestals are red-brown slipped.

Pedestal bases similar to those from Nea Nikomedeia have been found at Early Neolithic Otzaki (Milojdc' 1971, Fig. 15, 10-12) and Soufli Magoula (Gallis 1982, Fig. 2, 13) in Thessaly. At both sites, as in Nea Nikomedeia, such examples are rare.

5.4. 1 1 Legs A small number of clay legs (10 examples) have been found during the excavations at Nea Nikomedeia. They are usually cylindrical in form, sometimes tapering towards the base (fig. 5.24, i~6). Their height ranges from 2 cm to 5 cm and their larger diameter measures between 2 cm and 4 cm. Although none of the legs are attached to a pottery fragment, it seems more probable that they were supports for polypod vases or small 'altars', rather than being parts of figurines. Figurines from Nea Nikomedeia have plump legs which are modelled together with the buttocks (Nandris 1968). Moreover, figurines are usually unslipped, whereas most of the legs included in the sample have a surface coating. Red- brown slipped and pink coated examples are present, with the latter being more rare.

One miniature vase on four legs (two legs are missing) shows that polypod pots were in use at Nea Nikomedeia and that the legs, described above, could have been used in a similar way. It is probable, however, that some of them could have been used as supports of small 'altars'. At Achilleion, throughout the Neolithic occupation of the site, small legs similar to those from Nea Nikomedeia were used for supports of small triangular or rectangular 'altars' (Gimbutas 1989, Figs. 7.63-7.67). Similar 'altars' have been found at Prodromos (Theocharis 1973, Figs. 12 & 214).

A few legs, similar to those from Nea Nikomedeia, have been found at Otzaki (MilojCié 1971, Fig. X, 28). Similar broken off legs have also been found at Servia (Wijnen 1979, 194) and Vashtëmi (Korkuti 1982, Plate III, 1-10).

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 101

Polypod vessels are present in the Early Neolithic sites: examples have been found at Vrsnik (Garasanin 1982, 90), Porodin (Grbi(f et al. i960, Plate XXI 2, 4, 5) and Drenovac (VeiniC 1972, Fig. 1, 1 2). A miniature vessel similar to that from Nea Nikomedeia has been found at Anzabegovo (Anza) (Gimbutas 1976, Fig. 23, 1).

5.4.12 Comparison of Shapes of Plain and Decorated Vessels From the drawings of the painted vessels, made by Biernoff, the relative frequency of each vessel form has been estimated (fig. 5.62). In the same figure the relative frequency of the shapes of impressed and plain vessels is presented. For the vessels with applied decoration, percentages could not be estimated since the sample of rim sherds is very small, consisting of four rim sherds with face vessels and seven sherds with simple linear motifs. The remaining examples belong to body sherds.

Comparing these figures it can be seen that the decorated and plain vessels have a comparable repertory of shapes, but some differences do exist. In the case of the painted pottery, the differences concern the relative frequencies of neck and hole-mouthed jars. Hole-mouthed jars are slightly more numerous, being twice as frequent in painted pottery than in plain. Neck-jars, in contrast, are virtually absent amongst the painted vessels (only one example).

In the case of impressed pottery, an increase of the slightly closed vessels (rim angle from 91-1 14o) and dish-like pots (rim angle smaller than 6o°) can be seen. These shapes were favoured at the expense of neck and hole-mouthed jars, which are absent. A decrease in the number of open vessels with straight walls (rim angle from 85-900) is also noticed. The absence of neck jars among the impressed pots (and probably also among the painted) could be explained by the fact that the decoration was not extended up to their collars. Whole neck jars from other sites {e.g. examples from Porodin at Bitola Museum) show that impressed decoration was restricted to the body of the jars.

The few rim sherds with applied decoration belong, almost exclusively, to neck jars and open vessels (rim angle from 60-840) with concave upper walls. Five out of the seven rim sherds with simple linear motifs and one face vessel, belong to neck jars with high concave collars. The remaining face vessels (3 sherds) and one sherd with linear motifs, are from open pots (rim angle from 60-800). There is only one example of a slightly closed vessel. Neck jars and open vessels with concave walls must have been common among the pots with applied decoration, since many body sherds from near the rim-line show a pronounced concavity (fig. 5.54, 4-6). This indicates that, in the case of vessels with applied decoration, there is a correlation between vessel form and decoration.

Apart from the differences in the amount of some vessel forms, painted and impressed vessels have similar contours to those of the plain vessels. Their upper walls are usually straight, but slightly concave or convex examples are also present. The latter examples (concave upper walls) are more common among the painted vessels (fig. 5.36, 1; fig. 5.36, 8 & IO; FIG. 5.39, 6; FIG. 5.39, 10; FIG. 5.43, 13; FIG. 5.44, 7-9; FIG. 5.45, 9; FIG. 5.46, 5- 7; fig. 5.46, 8, 1 1 & 12; fig. 5.47, 10-13; fig. 5.48, 16). The lower parts of the vessels are round or oval.

However, a difference in the size of the vessels is noticeable. In painted vessels the rim diameters range from 5 cm to 27 cm, the most common group being between 13 cm and 19 cm (fig. 5.31). Examples of the lower and upper values of diameters are rare. The

1O2 PARASKEVI YIOUNI

distribution of the rim diameters for each vessel form is as follows: vessels with rim angle larger than 115o: rim diameters range from 12-18 cm; vessels with rim angle from 91- 1 14o: 8-27 cm; vessels with rim angle from 85-900: 5-25 cm; vessels with rim angle from 60-840: 7-21 cm; vessels with rim angle smaller than 6o°: 13-21 cm. Large pots (rim diameter up to 32 cm) are absent among the painted vessels. Such examples, although rare, are present in the plain pottery.

In contrast, it seems that impressions were not applied to small vessels, since examples with rim diameters smaller than 1 2 cm are absent among the impressed pots. Their rim diameters range from 1 2 cm to 32 cm, the most common group being between 18 cm and 30 cm (fig. 5.32).

Among the painted vessels, lugs are present in two hole mouthed jars (fig. 5.34, 6; fig. 5.42, 4) and in a limited number of body sherds (in 40 examples among the 1000 painted body sherds). Also, none of the impressed sherds have lugs or bear scars made from the attachment of lugs. It would seem, therefore, that lugs were restricted mainly to plain vessels. However, it should be pointed out that the majority of lugs are broken off the vessels; with those lugs that are not decorated (painted or impressed), it is impossible to distinguish whether they belong to plain or decorated vessels.

Painted pots are usually supported by short ring bases. Their diameters range from 3- 6.5 cm. Impressed vessels are supported by ring and flat bases; a few round bottomed examples have also been found. Their diameters range from 6-10 cm. One sherd with diameter around 16 cm is probably from an oval base. A similar oval example has been found among the painted vessels (Biernoff 1969).

5.4.13 Comparison of Fabric Types and Quality of Paste The fabrics used for the manufacture of painted pottery have been presented in Chapter 4. It was seen in section 4.6.8 that, although the majority of painted vessels were made from the same fabrics used for the plain pots (mainly fabric A), two different fabrics (Groups 2 and 3) were used exclusively for the manufacture of painted vessels.

Impressed pottery and pottery with applied decoration were made from the same fabrics used for the manufacture of plain vessels. Their relative frequencies are given in fig. 5.59. From fig. 5.59 it can be seen that a considerably larger amount of impressed and pottery with applied decoration were made from fabrics D and E. Indeed, 2 1 % of the impressed and 27% of the pottery with applied decoration, was made from fabric D. This fabric was used in only 3% of the plain vessels. Similarly, 16% of the impressed and 19% of the applied vessels, were made from fabric E, which was used for only 4% of the plain pottery. In impressed pottery, fabrics D and Ε were used at the expense of fabrics Β and C. In the pottery with applied decoration, it is fabric A which is less frequent.

In both impressed and applied pottery, the increase in the frequency of fabric D is accompanied by a rise in the number of red-brown unslipped vessels (fig. 5.43). Such examples amount to 15% of the impressed and 26% of pottery with applied decoration, whereas red-brown unslipped vessels comprise only 3% of the plain pottery. A similar correlation between fabric types and decoration has been noticed at Prodromos (Chourmouziadis 1971).

When comparing the quality of the paste of plain pottery and pottery with applied decoration (fig. 5.61), it can be seen that the latter contains a considerably larger proportion

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 103

of coarse sherds (42%) as opposed to only 6% of the plain vessels. In contrast, impressed decoration has not been used widely on coarse textured sherds as 52% of the sherds are fine textured, 26% are medium and 22% coarse (fig. 5.61). These proportions are similar to those obtained from the plain unslipped sherds (fig. 5.58). The great majority of the painted vessels are fine textured (see section 4.6.8).

5.5 INTRA-SITE DEVELOPMENT

Re-examination of the excavation data from Nea Nikomedeia suggests that there were three building phases during the Early Neolithic occupation of the site. Study of the plain pottery shows that the same fabric types, coated/uncoated ware and vessel forms were in use in all building periods.

However, it might be expected that the ceramic material from the different phases could show differences in the relative frequencies of the fabrics, shapes and surface treatment. These differences would reflect altered preferences among the inhabitants of the three phases due to changes in the economy, function of vessels, availability of raw material and/or the state of technological development. In order to check whether any such differences were discernible, the plain pottery from the different periods has been comparatively studied.

It is clear that only a rough comparative study can be attempted since each spit (with the exception of spit 3) contained the remains of more than one building phase. Thus, the ceramic material was divided into two broad categories: one category consists of the squares/ spits with pottery from the first and second building phases and the other contains the material from the second and third building phases. The small number of sherds coming from spit 3 were incorporated into the first category (first and second building phases). The relative frequencies of vessel forms, coated/uncoated ware and fabric types, has been calculated for each category. It should be noticed that this comparative study is based on the ceramic material from the squares/spits belonging to Structural Groups 1-6.

The percentage of red-brown slipped, pink coated, beige and red-brown uncoated vessels as well as the relative frequency of the fabric types present in each category, are given in fig. 5.65 and fig. 5.66 respectively. From fig. 5.65 it can be seen that the ratio of coated and uncoated vessels remains constant. Similarly, the proportion of fabric types is the same for the two categories. The frequency of fine, medium and coarse textured vessels does not change from one category to the other. Moreover, macroscopic examination of the vessels shows that there is no difference in the firing of the vessels.

The relative frequencies of the different vessel forms are given in fig. 5.63. As can be seen from this figure, the proportion of the vessel forms is similar in the two categories. Only one difference is noticed in the amount of neck jars which seem to have been used more frequently during the later phases of occupation, at the expense of the hole-mouthed jars. Variations in the profile {e.g. concave/convex/straight walls) of the vessels from the two categories are not observable. The amount of deep, medium and shallow vessels seems to remain constant.

Comparative study of the decorated pottery shows that all the types of decoration (painted: R/W & W/R, impressed and applied) and the same decorative motifs, were in use throughout the occupation of the site. The relative frequencies of occurrence of the different types of decoration were not calculated separately for each category (first and second; second and third building phases) because the quantity of ceramic material is very small.

io4 PARASKEVI YIOUNI

5.6 CONCLUSION

It is clear from the above analysis that the pottery from the three building phases is rather homogeneous. Despite the fact that the sample is mixed (one category includes pottery from the first and second building phases whereas the other includes pottery from the second and third phases) the relative frequencies of surface treatment and fabric types, present in the two categories, show a striking similarity.

As for the relative frequencies of the vessel forms, the observed difference in the proportion of neck and hole-mouthed jars is 5%, which seems to be above the experimental error. Even so, the difference is small and is not accompanied by any changes in the proportion of the other vessel forms, or by any technological development. On the contrary, the Nea Nikomedeia potters show an adherence to their potting tradition.

Study of the early monochrome pottery from Sesklo (Wijnen 1981) concludes that the amount of fine and medium textured vessels increases with time, at the expense of those with a coarse texture. This was accompanied by an increase in the amount of open vessels and an improvement in the firing of the pots. At Early Neolithic Achilleion, in the strata with decorated pottery10, a development in the repertory of vessels is noticed as high necked jars and vessels with S-profiles appear at a later stage. Moreover, an improvement in slipping, firing and levigating the paste of the vessels is apparent, with time. A similar technological improvement is noted between Anzabegovo (Anza) la and Ib (Gardner 1976). This is accompanied by the introduction of a new type of pottery decoration (impressed).

The studied ceramic material from Nea Nikomedeia does not show any distinct differences that could be related to an economic, functional or technological change. Instead, the uniformity of the material suggests that the excavated part of the Nea Nikomedeia settlement represents only a short time span. A similar conclusion was reached from the study of the architectural remains and building techniques.

The Figures figs. 5.1-5.33, 553-566 have all been drawn and prepared by Paraskevi Yiouni using MacDraw II on an Apple Macintosh computer. For figs. 5.4-5.24 the information given for each vessel lists the excavation square/spit number (-/- indicates no known provenance); Feature (F.); the fabric colour (R: Red; P: Pink; B: Beige); and finally, the number given to each sherd during macroscopic examination.

figs. 5.34-5.52 have all been prepared from the archive of pottery drawings produced by David Biernoff, and a catalogue of the sherds illustrated accompanies them. The catalogue lists the following data:

Sherd feature followed by decoration (R/Ws: Red on White Standard Ware; R/Wp: Red on White Porcelain Ware; W/R: White on Red Ware); Area Code, listing the excavation square/spit number (-/- indicates no known provenance), Feature (F.), followed by the number given to sherds during macroscopic examination; Munsell Soil Colour Chart notations for both interior and exterior surfaces.

10 Achilleion Ib- II: Gimbutas 1989.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 105

fig. 5.1: Plain pottery: relative frequency of burnished vessels and vessels with a surface coating.

fig. 5.2: Relative frequency of the different types of decorated pottery.

fig. 5.3: Relative frequency of the different types of impressed decoration.

io6 PARASKEVI YIOUNI

Πλ 1 4

fig. 5.4: Neck jars. Intermediary between hole-mouthed jars and neck jars (Ai): 1) B6/1 Pi 2, 2) D6/2 R26, 3) B6/2 Bi; Neck jar with medium collar (As): 4) C5/1 R25.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 107

1 4

fig. 5.5: Neck jars with medium collar (As): 1) C5/2 P33, 2) B4/2 Ri?, 3) B7/1 R22, 4) E6/1 R.

io8 PARASKEVI YIOUNI

τπΓΠ m

1 4

fig. 5.6: Neck jars with high collar: 1) Cio/i P21, 2) C00/2 R2, 3) B9/1 P28, 4) B2/3 Rio, 5) D6/2 P21, 6) A7/1 R31; Probable neck jars: 7) C3/1 B3, 8) L8/1 R6.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 109

s 1 '

ΓΓΛ '

fig. 5.7: Closed vessels with rim angle larger than 1 15o (hole-mouthed jars): 1) B1/3 P5, 2) B6/1 Pi, 3) B1/3 B25, 4) B1/3 B11, 5) B1/3 R5, 6) -/-.

no PARASKEVI YIOUNI

Γ^ '

fig. 5-8: Closed vessels with rim angle larger than 115o (hole-mouthed jars): 1) -/-, a) B1/2 Ra, 3) B6/s» Ri, 4) Tgs/i B, 5) A6/1/2, 6) B6/2 F.X. R.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 1 1 1

L Jy */ *--»*. */ 1 4

fig. 5.9: Askoid vessels: 1) A/i B, 2) K7/1 B, 3) F6/1 F.C. B.

112 PARASKEVI YIOUNI

©α? ππ m

™ J

Φ. fig. 5.10: Slightly closed vessels with rim angle from 91-1 14o: 1) B6/1, 2) A6/0, 3) Β 1/3 P6, 4) D4/3 Rs>3, 5) B6/2, 6) B6/1 Ris», 7) F7/1 EC.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 113

to to

fig. 5.11: Slightly closed vessels with rim angle from 91-1 14o: 1) H6/1, 2) C5/1 Pio, 3) D4/1 P.

ii4 PARASKEVI YIOUNI

ΓΊ 1 I I 1

2

i~' '

Φ 1 4

fig. 5.12: Slightly closed vessels with rim angle from 91-1 14o: 1) C3/1 Pu, 2) Coo/s? R28, 3) D6/s B6, 4) -/-.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 115

1 4

fig. 5.13: Slightly closed vessels with rim angle from 91-1 14o: 1) A6/1, s>) A8/1.

ii6 PARASKEVI YIOUNI

6

fig. 5.14: Open vessels with rim angle from 85-900: 1) D6/1 R13, 00 2) B5-C6 Baulk, *) D6/1 Rk, O' 4) D6/2 R20, 5) D 1/3 R6, 6) A6-B6/2 Pis>, 7) M8/1.

00 O'

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 117

A 3 CO '

1 4

fig. 5.15: Open vessels with rim angle from 85-900: 1) F6/1 EC, 2) A6/0, 3) F6/1 F.C., 4) surface find.

ii8 PARASKEVI YIOUNI

^7 '7^7

ίΤΊΓτ7 ' · ! /

1 4

fig. 5.16: Open vessels with rim angle from 60-840: 1) C00/1 R25, 2) A7/1 Rio, 3) C00/2 R22, 4) A2/3 R?5, 5) C3/1 Psi, 6) C5/2 B2, 7) D4/1 R6, 8) C5/2 P34, 9) D4/2 Bi?.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 1 19

vn 1 4

fig. 5.17: Open vessels with rim angle from 60-840: 1) D3/3 Β 12, 2) Coo/2 R3, 3) A2/3.

12O PARASKEVI YIOUNI

l~l 7 V~l

7

1 4

fig. 5.18: Shallow vessels with rim angle from 60-840: 1) A6/1 B12, 2) B6/2 P8, 3) D6/1 Rio, 4) B1/3 R18, 5) D6/1 R4.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 121

V7"h 7 STTy7

1 4

fig. 5.19: Open vessels with rim angle less than 6o°: 1) A2/3 B4, 2) C5/2 B13, 3) D1/3 R32, 4) C00/2 R7, 5) B9/1 R27, 6) B9/1 R22, 7) D1/3 B2, 8) D1/3 P32, 9) B1/3 Bs, 10) B4/2 R23.

122 PARASKEVI YIOUNI

V^ 7 V^ I

7

V^ 7 V^n 7

"s ; - ι ζ

^ ; - ι 7 1 4

fig. 5.20: Open vessels with rim angle less than 6o°: 1) B2/3 R39, 2) C5/2 R23, 3) D6/2 Bi, 4) B9/1 R6, 5) B1/3 B5, 6) D1/3 Bs, 7) C5/2 R20.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 123

^^^^ 8

^/

^^^. 9 >/

^ / 1 4

fig. 5.21: Ring based vessels: 1) A6-B6/2 R11, 2) K5/1 B45, 3) Β 1/3 Ris>, 4) K5/1 B37, 5) A7/1 R21, 6) L4/1 B25, 7) A7/1 R17, 8) A6/1 R16, 9) B6/1 P17.

124 PARASKEVI YIOUNI

1 4

fig. 5.22: Flat based vessels including vessels with a low foot (1-5): 1) K5/1 R38, 2) D4/2 B28, 3) C5/1 B34, 4) A7/1 R18, 5) A6-B6/2 P8, 6) K5/1 R52, 7) D4/3 R38, 8) C5/2 R24, 9) C00/1 B27, 10) A7/1 R32.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 125

^^^. 2 /

^^^^ 3 /

17 4 ^U œZZZZa &77ZZZPZPZZZZZZZZZZ*

r ' j * ι - ^

1 4

fig. 5.23: Round bottomed vessels: 1) B6/2 B32, 2) A3/1 B23, 3) L4/1 K30; Pedestal bases: 4) NN LE 3/7, 5) NN Es>/2, 6) D6/1.

126 PARASKEVI YIOUNI

I já -J slipped

,9 ' core

/Γ~-- >^ core B^' /Γ~-- >^ 13 Ι Λ f - ' 14 V Ι Λ O f - ' 14 o

1 4

fig. 5.24: Legs from polypod vessels: 1) -/-, 2) C5/1, 3) D9/2 FA, 4) B1/1, 5) Β?/?, 6) B4/2; Unpierced lugs: 7) B9/1, 8) D6/1, 9) B3/1; Z-shaped lugs: 10) B5/1, 11) B2/2; Spindle-whorl: 12) L8/1 R21; Round sherds: 13) B2/3 R56, 14) B1/3.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 127

fig. 5.25: Distribution of rim diameters for neck jars.

fig. 5.26: Distribution of rim diameters for closed vessels with rim angle larger than 115o (hole- mouthed jars).

128 PARASKEVI YIOUNI

fig. 5.27: Distribution of rim diameters for closed vessels with rim angle from 91-114°.

fig. 5.28: Distribution of rim diameters for open vessels with rim angle from 85-90°.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 129

fig. 5.29: Distribution of rim diameters for open vessels with rim angle from 60-840.

fig. 5.30: Distribution of rim diameters for open vessels with rim angle less than 6o°.

i3o PARASKEVI YIOUNI

fig. 5.31: Distribution of rim diameters of painted vessels.

fig. 5.32: Distribution of rim diameters of impressed vessels.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 131

fig. 5.33: Distribution of basal diameters of plain vessels.

i32 PARASKEVI YIOUNI

/ ΚΤΛ <^^^

* w^' J [J '

fig. 5.34: 1-1?) Painted vessels with rim angle larger than 115o; 13) Painted neck jar.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 133

FIG. 5.34, 1 rim sherd; R/Ws Area Code: LI C4/11, 5, F. A. Colour: Int: 10R4/4 | 10YR8/3-4; Ext: 10R4/4 | 10YR8/3-4 Diam: 16 cm Porcelain ware. Red paint over white slip interior and exterior, medium gloss finish.

fig. 5.34, 2 rim sherd; R/Ws Area Code: D6/2, 1 Colour: Int: 10YR8/4; Ext: 2.5YR4/4 | 10YR8/3 Diam: 12 cm Porcelain ware. White slip interior. Red paint over white slip exterior. Medium gloss finish both surfaces.

fig. 5.34, 3 rim sherd; R/Ws Area Code: E8/2, 3 Colour: Int: 2.5YR4/5-6; Ext: 2.5YR4/5 | ioyr6- 7/2 Diam: 15 cm

fig. 5.34, 4 rim sherd; R/Ws Area Code: Τ Υ 2/0, 1 Colour: Int: 2.5YR5/4-5; Ext: 10R4/4-5 I 10YR7/ 4-5 Diam: 11-12 cm

fig. 5.34, 5 rim sherd; R/Wp Area Code: C2/2, 1 Colour: Int: 7.5YR7 -8/4; Ext: 10R3-4/4 | 7.5YR8/4 Diam: 18 cm Porcelain ware. White slip interior. Red paint over white slip exterior. Both surfaces medium to high gloss finish.

fig. 5.34, 6 rim sherd; R/Wp Area Code: J8/1, 4 Colour: Int: 10YR6-7/3-4; Ext: 2.5YR2-4/4-5 | 10YR6-7/3-4 Diam: 17 cm Porcelain ware, light slip interior. Red paint over light slip exterior. Both surfaces medium gloss finish. Small round knob on lower exterior of sherd.

fig. 5.34, 7 rim sherd; W/R Area Code: -/-, 27 Colour: Int: 10R3/5; Ext: 10YR8/3 | 10R3/5 Diam: 15 cm

FIG. 5.34, 8 rim sherd; R/Ws Area Code: C5/1, 2 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR7/4 Diam: 16 cm?

fig. 5.34, 9 rim sherd; R/Wp Area Code: C1/1, 2 Colour: Int: 7. 5YR7/4; Ext: 2.5YR4/5-6 | 7.5YR7/4 Diam: 17 cm Porcelain ware. Light slip interior. Red paint over light slip exterior. Both surfaces medium gloss finish.

fig. 5.34, 10 rim sherd; R/Ws Area Code: ΙΧΙΌ$/ι, ι Colour: Int: 10R4/4; Ext: ioyr8/i | 10R4/4 Diam: 15 cm Red slip interior and exterior, medium gloss. White paint exterior over slip, matt.

fig. 5.34, 1 1 rim sherd; R/Ws Area Code: LXI D5/2, 2 Colour: Int: 5 YR3-5/1 -3 ; Ext: 5 YR3 -4/ 1-3 | 10YR4- 6/1-2 Diam: 18 cm Red slip interior, low gloss. Red paint over light slip exterior, low to medium gloss finish.

fig. 5.34, 12 rim sherd; Polychrome vessel Area Code: D 5/1, 1 Colour: Int: 5YR6/4; Ext: 10R4/3-4, 10YR8/3 | 5YR6/4 Diam: 3.5 cm Unique sherd. Red and white paint over body. Interior unslipped. Neither surface burnished. Paint is very smeared and unevenly applied.

fig. 5.34, 13 rim sherd; W/R Area Code: L6/1 F. G, 1 Colour: Int: 2.5YR3-4/6 | 7.5YR6/4; Ext: 10YR5/4 | 2.5YR4/6 Diam: 1 1 cm Interior red slip, rim band 3 cm, over body. Exterior white paint over red slip; medium to high gloss finish.

134 PARASKEVI YIOUNI

A 'J )

ιλ ■ il "- I I Q 4 'L ψ "- ,- + Ι ■■ΖΖβΚΙΖΙ

FIG· 5·35: Slightly closed painted vessels with rim angle from 91-1 14o.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 135

fig. 5.35, 1 rim sherd; R/Ws Area Code: C8/1, 2 Colour: Int: 2.5YR4/5-6; Ext: 2.5YR4/5-6 | 10YR7/ 4, 7.5YR7/4 D iam: 14 cm Unusual design. Medium gloss finish interior and exterior.

fig. 5.35, 2 rim sherd; W/R Area Code. TY/10 N. Baulk, 1 Colour: Int: 2.5YR4/3-5; Ext: 10YR7-8/3-4 | 2.5YR4/3-5 Diam: 15 cm? Red slip interior. White paint over red slip exterior. medium gloss finish.

fig. 5.35, 3 rim sherd; R/Ws Area Code: D 4/1, 2; D6/1, 1 Colour: Int: 2.5YR3-4/2-4 | 7.5YR5-4/2, 5YR5-4/ 2; Ext: 5YR4/4, 2.5YR4/4-5 I 7-5YR7/4-5 Diam: 16 cm Unevenly slipped interior

fig. 5.35, 4 rim sherd; R/Ws Area Code: D-E8, 1 Colour: Int: 2.5YR4/5-6 | 7.5YR5/4; Ext: 2.5YR4/5 | 10YR8/4, 7.5YR8/4 Diam: 18 cm Only faintest traces of slip towards bottom of sherd interior.

fig. 5·35> 5 rim sherd; W/R Area Code: K8/ i , 7 Colour: Int: 2.5YR4/6; Ext: 10YR8/5 | 2.5YR4/6 Diam: 16 cm Exterior surface and design partially destroyed. Red slip interior. Light paint over red slip exterior, low to medium gloss finish. Originally very fine ware.

fig. 5.35, 6 rim sherd; R/Ws Area Code: C8/1, 1 Colour: Int: 10YR4-5/1-2; Ext: 2.5YR2-4/2-4 | 10YR8/2-3 Diam: 17 cm? Unusual ware. Interior slip? Smooth matt finish. Exterior thin red paint over matt eroded white slip. High gloss finish to red paint.

fig. 5.35, 7 rim sherd; R/Wp Area Code: J8/1, 5 Colour: Int: 10YR8/3-4; Ext: 5YR4-5/6 | ioyr8/ 3-4 Diam: 14 cm Porcelain ware. Light slip interior, low to medium gloss finish. Red paint over white slip exterior, high gloss finish.

fig. 5.35, 8 rim sherd; R/Ws Area Code: LXI E5/1, 1 Colour: Int: 5YR6/2-5; Ext: 5YR4/1, 5YR5/2-4 | 10YR7/1-4 Diam: 15 cm Red slip interior smoothed to matt finish. Red paint over light slip exterior applied using outline and infill technique; low to medium gloss finish.

fig. 5.35, 9 rim sherd; R/Wp Area Code: B4/2, 1 Colour: Int: 10YR8/3-4; Ext: 10R4/4-5 | 10YR8/3 Diam: 14 cm Porcelain ware. White slip interior, low to medium gloss. Red paint over white slip exterior. Medium to high gloss.

fig. 5.35, 10 rim sherd; R/Ws Area Code: A,B6/2 X, 6 Colour: Int: 10R4/4; Ext: 10R4/4 I 10YR7/4, 7.5YR 7/4 Diam: 12 x 14 cm May be oval vessel

fig. 5.35, 11 rim sherd; R/Ws Area Code: -/-, 57 Colour: Int: 10R4/4; Ext: 10R3-5/4 | 10YR6/4 Diam: 20 cm

fig. 5.35, 12 rim sherd; R/Ws Area Code: LXI D5/3, LXI E5/F.A., LXI D5/4, 3 Colour: Int: 2.5YR6-5/5-6; Ext: 2.5YR5/5 I 10YR7- 8/3 Diam: 16 χ 20 cm Oval vessel. Interior red slip matt to low gloss finish. Exterior red paint over white slip, low to medium gloss finish. Paint thin and evenly applied.

i36 PARASKEVI YIOUNI

ι τ^τη ,1 Ρ Ι

.ΓΡΙ .ΠΓι

fig. 5.36: Slightly closed painted vessels with rim angle from 91-1 14o.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 137

FIG. 5.36, 1 rim sherd; R/Ws Area Code: B2/3, 2 Colour: Int: 2.5YR3-4/2-6; Ext: 2.5YR4/5-6 | 10YR7/2-3 Diam: 16.5 cm

fig. 5.36, 2 rim sherd; R/Ws Area Code: HS/ ιά, 2 Colour: Int: 5YR4/4; Ext: 2.5YR4/5-6 | 7.5YR7/5 Diam: 15 cm

fig. 5.36, 3 rim sherd; R/Ws Area Code: O7/1, 8 Colour: Int: 2.5YR5/5; Ext: 10R4/4 | 10YR8/4 Diam: 15 cm Red slip interior. Red paint over light slip exterior. Applied using outline and infill technique. Outline overlapped by infill.

fig. 5.36, 4 rim sherd; R/Ws Area Code: F8/1?, 5 Colour: Int: 7.5YR8/ 2-4, 5YR4/4-5 | 7.5YR4-5/0-1; Ext: 7.5YR5/2-4 I 10YR5-6/1 Unusual rim elaboration. Ware close to porcelain. Interior painted rim band ( 1 cm) over body.

fig. 5.36, 5 rim sherd; R/Ws Area Code:B6/i, A,B6/2 X, 1 Colour: Int: 10R5/5; Ext: 10R4/4 | 10YR6-7/4, 7.5YR7/4-5 Diam: 17 cm

fig. 5.36, 6 rim sherd; R/Ws Area Code: C0-C1, Do-Di F. B, 1 Colour: Int: 2.5YR4-5/4; Ext: 2.5YR4/4-5 | ioyr 7/3 Diam: 20 cm

fig. 5.36, 7 rim sherd; R/Ws Area Code: -/-, 76 Colour: Int: 10R4/4-5; Ext: 10R4/3-4 | 10YR6-7/ 2-3 Diam: 1 1 cm

fig. 5.36, 8 rim sherd; R/Ws Area Code: -/-, 42 Colour: Int: 2.5YR4/4; Ext: 2.5YR4/5 | 10YR7/2 Diam: 13 cm

fig. 5.36, 9 rim sherd; R/Ws Area Code: C7/1, 1 Colour: Int: 2.5YR4-5/5; Ext: 2.5YR4/5 | 10YR7/4 Diam: 16 cm

fig. 5.36, 10 rim sherd; R/Ws Area Code: LXI E4/1, 3; LXI E5/3, 1 Colour: Int: 2.5YR2-3/2-4; Ext: 2.5YR3-4/2-4 | 10YR5-6/2 Diam: 15-16 cm Red slip interior, uneven in thickness and finish, matt to low gloss. Underlying body shows through in places. Exterior, red paint over white slip. Paint thin and applied using outline and infill technique, low to medium gloss finish.

138 PARASKEVI YIOUNI

.Γ~ΡΊ ,/ ^Ί

10 I ^W I 0 4

fig. 5.37: Slightly closed painted vessels with rim angle from 91-1 14o.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 139

fig. 5.37, 1 rim sherd; R/Ws Area Code: D 2/2, 2 Colour: Int: 2.5YR4/5 | 5YR3/4; Ext: 2.5YR4/5 | 7-5YR7/5 Diam: 17 cm Interior slip in broad band below rim (3 cm), then zone of partial coverage and smearing. Below unslipped. Exterior paint applied using outline and infill technique.

fig. 5.37, 2 rim sherd; R/Ws Area Code: A8/2 F. A, 1 Colour: Int: 10R4/4-5; Ext: 10R4/4 | 10YR7/4 Diam: 18 cm

fig. 5-37» 3 rim sherd; R/Ws Area Code: A3/3, 3 Colour: Int: 2.5YR5/4; Ext: 2.5YR5/5 | 10YR7/3 Diam: 20-22 cm? Red slip interior, matt to low gloss. Red paint over light slip exterior. Low to medium gloss finish.

fig. 5.37, 4 & 5 rim sherd; R/Ws Area Code: L8/iy 1 & 3 Colour: Int: 2.5YR3-4/6; Ext: 10R4/5 | 10YR7/4-5 Diam: 18-20 cm

fig. 5.37, 6 rim sherd; R/Ws Area Code: B8/1, 2 Colour: Int: 2.5YR4-5/6; Ext: 2.5YR4/5-6 | ioyr 7/4 Diam: 13 cm

fig. 5.37, 7 rim sherd; R/Ws Area Code: By /i, 1 Colour: Int: 2.5YR5/5; Ext: 2.5YR4/5 | 10YR7/4-5 Diam: 16 cm

fig. 5.37, 8 rim sherd; R/Ws Area Code: ΎΧ$/ι, 2 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/4-5, 10R4/4-5 I 10YR7/4 Diam: 18 cm

fig. 5.37, 9 rim sherd; R/Ws Area Code: B3/2, 3 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5-6 | 10R8/5 Diam: 18 cm

fig. 5.37, 10 rim sherd; R/Ws Area Code: O6/ i, 2 Colour: Int: 2.5YR5/4; Ext: 2.5YR4/6 | 10YR7-8/4 Diam: 18 cm? Red slip interior. Red paint on white slip exterior.

fig. 5.37, 11 rim sherd; R/Ws Area Code:O$/i, 5; D3/2, 2 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 7-5YR7/4~5 Diam: 15 χ ι8 cm Oval vessel.

Mo PARASKEVI YIOUNI

' m ι J en 11 ^

15 % ^^ / 16

fig. 5.38: Slightly closed painted vessels with rim angle from 91-1 14o.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 141

FIG. 5.38, 1 rim sherd; R/Wp Area Code: L6/K, 9 Colour: Int: 10YR8/3-4; Ext: 2.5YR4/5 | 10YR6-7/ 3-4 Diam: 17-18 cm Porcelain ware.

fig. 5.38, 2 rim sherd; W/R AreaCode.TX 2/1, 1 Colour: Int: 2.5YR5/4 | 10YR3-7/1; Ext: 10YR8/3 | 10YR3-5/1 Diam: 10 χ 22 cm Oval vessel. Interior only faint traces of slip at rim. Exterior matt white paint over black slip, matt to low gloss finish.

fig. 5.38, 3 rim sherd; R/Ws Area Code: -/-, 54 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR7/4 Diam: 12.5 cm

fig. 5.38, 4 rim sherd; R/Ws Area Code: -/-, 40 Colour: Int: 2.5YR4-5/5; Ext: 10R4/4-5 | 10YR7/3 Diam: 12.5 cm

fig. 5.38, 5 rim sherd; R/Ws Area Code: F5/2, 3 Colour: Int: 2.5YR5/5-6; Ext: 2.5YR5/5-6 | 7.5YR5/4 Diam: 16 cm

fig. 5.38, 6 rim sherd; R/Ws Area Code: L4/1, 1 Colour: Int: 10R4/6; Ext: 10YR8/2 | 10R4/6 Diam: 17 cm Red slip interior. White paint over red slip exterior.

fig. 5.38, 7 rim sherd; R/Ws Area Code: LU B1/1-2, 1 Colour: Int: 2.5YR4/4; Ext: 2.5YR5/4 | 10YR7/1 Diam: 2 1 cm Red slip interior, matt or low gloss. Exterior red paint over white slip. Paint thin and evenly applied. Matt to medium gloss finish.

fig. 5.38, 8 rim sherd; R/Ws Area Code: C3/1, 2 Colour: Int: 2.5YR4/5-6; Ext: 2.5YR4/6 | 10YR7/3 Diam: 16 cm?

FIG. 5.38, 9 rim sherd; R/Ws Area Code: A2/3, 2 Colour: Int: 10YR3-4/1-2; Ext: 5YR3-4/1, 5/4 | 10YR6/2-4 Diam: 20 cm Interior unslipped, matt finish. Exterior red paint over light slip, low to medium gloss finish.

fig. 5.38, 10 rim sherd; R/Ws Area Code: O6/ 1, 9

fig. 5.38, 11 rim sherd; R/Ws Area Code.U D1/2, 2 Colour: Int: 2.5YR6/4-6; Ext: 2.5YR5/4 | 10YR7/ 2-3 Diam: 24 cm Thin red slip interior. White slip exterior. Overall medium gloss. Paint thin and evenly applied.

fig. 5.38, 12 rim sherd; R/Ws Area Code: C10/ 2, 2 Colour: Int: 2.5YR4-5/4; Ext: 10YR4/4 | 2.5YR7/5 Diam: 18 cm Red slip interior. White paint over red slip exterior.

fig. 5.38, 13 rim sherd; R/Wp Area Code: E0/1, 3 Colour: Int: 7. 5 YR7/4; Ext: 2.5YR-/6 | 7.5YR7/4? Diam: 1 2 cm Porcelain ware. Light slip interior and exterior. Exterior red paint over light slip. Paint thick and raised.

fig. 5.38, 14 rim sherd; R/Ws Area Code: C4/1, 2 Colour: Int: 2.5YR4/5; Ext: 2.5YR4-5/5 | 10YR7/3 Diam: 16 cm

fig. 5.38, 15 rim sherd; R/Ws Area Code: D 4/1, 3 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5-6 | ioyr8/ 2-3 Diam: 9 cm

fig. 5.38, 16 rim sherd; R/Ws Area Code: G 6/1, 2 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/6 | 10YR8/3-4 Diam: 1 1 cm

142 PARASKEVI YIOUNI

.men ·γί»-λ

fig. 5.39: Slightly closed painted vessels with rim angle from 91-1 14o.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 143

fig. 5.39, 1 rim sherd; R/Ws Area Code: B2/3, 1 Colour: Int: 5YR4/4; Ext: 5YR4/4 | 10YR8/2 Diam: 12 cm

fig. 5.39, 2 rim sherd; R/Ws Area Code: H7/ia, 3 Colour: Int: 2.5YR5/5; Ext: 10R4/5 | 10YR6-7/4 Diam: 16 cm

fig. 5.39, 3 rim sherd; R/Ws Area Code: LXI C5/2, 1 Colour: Int: 2.5YR3-5/2-4; Ext: 10R3 -4/2-3 | 10YR5-6/2 Diam: 16 cm Interior red slip, low gloss finish. Exterior red paint over light slip, medium gloss finish.

fig. 5.39, 4 rim sherd; R/Ws Area Code: O6/ Ί, 5 Colour: Int: 2.5YR5-6/4; Ext: 10R4-5/4 | 10YR7/4 Diam: 18 cm Red slip interior. Red paint on light slip exterior.

fig. 5.39, 5 rim sherd; R/Ws Area Code:Dq/i A, 1 Colour: Int: 2.5YR4-5/5; Ext: 2.5YR4/5 | 10YR7- 8/3 Diam: 15 cm

fig. 5.39, 6 rim sherd; R/Ws Area Code: LXI D5/3, 1 Colour: Int: 2.5YR5/5; Ext: 2.5YR4/5 | 10YR7/2 Diam: 17 cm Interior red slip, matt to low gloss finish. Exterior red paint over white slip, matt to medium gloss finish.

fig. 5.39, 7 rim sherd; W/R Area Code: LXII A5/2, 3 Colour: Int: 10R4/3-4; Ext: ioyr8/i | 10R4/3-4 Diam: 18 cm Red slip interior and exterior, interior low gloss finish. Exterior medium to high gloss finish. White paint remains glossy where not eroded.

FIG. 5.39, 8 rim sherd; R/Ws Area Code: LI Ε 1/3, ' Colour: Int: 7.5YR5-6/2; Ext: 10R5/3 | 7.5YR5/2 Diam: 30-3 1 cm Interior roughly smoothed. Exterior uneven pink paint over body, poorly burnished. Low gloss highlights on matt surface.

fig. 5.39, 9 rim sherd; W/R Area Code: L7/1, 1 Colour: Int: 2.5YR4/6; Ext: 10YR8/3 | 2.5YR4/6 Diam: 13 cm Red slip interior. White paint over red slip exterior. White eroded from overall high gloss. Paint applied over scratched outline.

fig. 5.39, 10 rim sherd; R/Ws Area Code: A2/1, 7 Colour: Int: 2.5YR5/4; Ext: 2.5YR4/4-6 | 10YR8/3 Diam: 14 cm? Red slip interior, matt to low gloss finish. Red paint over white slip exterior. Low to medium gloss finish

fig. 5.39, 11 rim sherd; W/R Area Code: LI C1/1, 2 Colour: Int: 10R4/4-5; Ext: 10R4/4-5 | 10YR8/3 Diam: 16 cm Red slip interior and exterior. Thick matt white paint burnished to medium gloss exterior.

fig. 5.39, 12 rim sherd; R/Ws Area Code: LXI E5/F. A, 7

FIG. 5.39, 13 rim sherd; W/R Area Code: LNT, 3 Colour: Int: 2.5YR4/5-6; Ext: 10YR8/3 | 2.5YR4/ 5-6 Diam: 16 cm Red slip interior. White paint over red slip exterior.

144 PARASKEVI YIOUNI

r m ' /) I i C^ '

0 4

fig. 5.40: Slightly closed painted vessels with rim angle from 91-1 14o.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 145

fig. 5.40, 1 rim sherd; R/Ws Area Code.T Y 7/1, 3 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/4 | 10YR7/2-3 D iam: 10 cm

fig. 5.40, 2 rim sherd; R/Ws Area Code: D6/2, 8 Cohur: Int: 2.5YR4-5/4; Ext: 2.5YR4/4-5 | 10YR7-8/3

fig. 5.40, 3 rim sherd; R/Ws Area Code: D 2/1, 3 Colour: Int: 2.5YR4/4-5; Ext: 2.5YR4/4 | 10YR7/4 Diam: 12 cm

fig. 5.40, 4 rim sherd; R/Ws Area Code: LXI C5/3, 1 Colour: Int: 10R4/4 | 10YR7/4; Ext: 10R4/4 | 10YR7/ 3-4 Diam: 12 cm Interior and exterior red paint over body. Low to medium gloss finish. Surface eroded. Paint and surface now uneven.

fig. 5.40, 5 rim sherd; R/Ws Area Code: -/-, 59 Colour: Int: 5YR3-5/2; Ext: 2.5YR4/2-4 | 7.5YR7/4-5 Diam: 12 cm

fig. 5.40, 6 rim sherd; W/R Area Code.-ΎΧ 9/2, 1 Colour: Int: 5YR5/4; Ext: 10YR7/2-3 | 2.5YR4/5-6 Diam: 13 cm Reddish slip interior. White paint over red slip exterior. Medium gloss finish.

fig. 5.40, 7 rim sherd; W/R Area Code: E00/1, 1 Colour: Int: 10R4/4; Ext: 10YR8/2 | 10R4/4 Diam: 15 cm Red slip interior. White paint over red slip exterior. Both surfaces medium gloss finish.

fig. 5.40, 8 rim sherd; R/Ws Area Code: D 9/ 2, 3 Colour: Int: 5YR3/1-2; Ext: 5YR4/2-4, 3/1 | 10YR7/ 2-3 Diam: 17 cm Burned standard ware. Very high gloss finish both surfaces.

fig. 5.40, 9 rim sherd; W/R Area Code: C2/1, 1 Colour: Int: 10R4/4; Ext: 10YR8/3 | 10R4/4 Diam: 17 cm Red slip interior. White paint over red slip exterior. Both surfaces medium gloss finish.

fig. 5.40, 10 rim sherd; R/Ws Area Code: B6/2, 8 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 75YR7/4-5 Diam: 20 cm?

fig. 5.40, 11 rim sherd; R/Ws Area Code.Lll A2/1, 2 Colour: Int: 10YR8/3-4; Ext: 10R4/4 | 10YR7 -8/3-4 Diam: 22 cm Porcelain ware. White slip interior and exterior. Interior medium gloss. Exterior red paint somewhat smeared, burnished to high gloss.

fig. 5.40, 12 rim sherd; R/Ws Area Code: L6/1 K, 17 Colour: Int: 10YR5-6/2-3; Ext: 2.5YR4/4 | 10YR6-7/3-4 Diam: 8 cm Very crude with only roughly finished surfaces. Interior unslipped. Exterior badly eroded.

fig. 5.40, 13 rim sherd; W/R Area Code.lÄ E4/1, 1 Colour: Int: 5YR3-4/2-3; Ext: 10YR7/2 | 5YR3-4/2- 3, 2.5YR3/2-4 Diam: 1 1 cm Black or red slip interior and exterior with high gloss. Thick white paint from matt to gloss very uneven

fig. 5.40, 14 rim sherd; R/Ws Area Code: E6/2, 3 Colour: Int: 10R4/4-5; E**·' loYR8/3 | 5YR4/4, 10R4/4 Red slip interior, low to medium gloss finish. White matt eroded paint with traces of burnish visible over medium red slip exterior.

fig. 5.40, 15 rim sherd; W/R Area Code: C2/3, 2 Colour: Int: 10R4-5/4; Ext: 10YR8/3 | 10R4-5/4 Diam: 16 cm Red slip interior. White paint over red slip. White now eroded to matt otherwise traces of medium gloss finish.

146 PARASKEVI YIOUNI

12 f ^J '

0 4

fig. 5·41: Slightly closed painted vessels with rim angle from 91-1 14o.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 147

FIG. 5.41, 1 rim sherd; R/Ws Area Code: B3/2, 2 Colour: Int: 2.5YR6/5 | 7.5YR3-4/1-4; Ext: 2.5YR5/ 6 I 10YR8/2-3 Diam: 15 cm Very unevenly slipped interior shows body.

fig. 5.41, 2 rim sherd; R/Ws Area Code:'AT>$/', 1 Colour: Int: 10R4/5; Ext: 10R4/5 | 10YR7/2-3 Diam: 19 cm Red slip interior and fine white slip exterior. Overall medium gloss. Paint somewhat streaked from finishing process.

fig. 5.41, 3 rim sherd; R/Ws Area Code: D 1/2, 3 Colour: Int: 2.5YR4/4-5; Ext: 2.5YR4/4-5 | ioyr 7/3 Diam: 12 cm

fig. 5.41, 4 rim sherd; W/R Area Code: C2/1, 2 Colour: Int: 10R4/4; Ext: 10YR7/3 | 10R3-4/4 Diam: 28 cm? Red slip interior. White paint over red slip exterior. White eroded to matt finish, otherwise both surfaces medium gloss finish.

fig. 5.41, 5 rim sherd; R/Ws Area Code: F5/2, 1 Colour: Int: 2.5YR4-5/4-5; Ext: 2.5YR4/4-5 | 10YR7-8/4 Diam: 2 1 cm

fig. 5.41, 6 rim sherd; R/Ws Area Code.lXll A5/1, 2 Colour: Int: 10R5/3; Ext: 10R4-5/3 | 10YR7/1-2 Diam: 18 cm Red slip interior, medium gloss finish. Red paint over white slip exterior. Eroded to matt finish.

fig. 5.41, 7 rim sherd; R/Ws Area Code .JÏ4/1, 1 Colour: Int: 2.5YR4/6; Ext: 2.5YR4/5-6 | 10YR8/4 Diam: 12-13 cm

fig. 5.41, 8 rim sherd; R/Ws Area Code: -/- Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5-6 | 10YR7/4 Diam: 15 cm

FIG. 5.41, 9 rim sherd; R/Ws Area Code ;LXI D 4/1, 3 Colour: Int: 2.5YR6/4-5; Ext: 2.5YR4/5-6 | ioyr 7/3 Diam: 16 cm? Light red slip interior. Red paint over white slip exterior. Paint smeared. Medium gloss finish.

fig. 5.41, 10 rim sherd; R/Ws Area Code: J6/1, 2

fig. 5.41, 11 rim sherd; W/R Area Code.T X 9/1, 2 Colour: Int: 2.5YR4-5/4; Ext: 10YR7/3 | 2.5YR4- 5/4 Diam: 16-17 cm Red slip interior. White paint over red slip exterior, medium gloss finish.

fig. 5.41, 12 rim sherd; R/Ws Area Code:O>j/i, 3 Colour: Int: 10YR3-6/1; Ext: 5YR2-3/2 | 10YR7-8/ 2-3 Diam: ? Burned porcelain ware. Light slip interior. Red paint over white slip exterior, medium gloss finish.

i48 PARASKEVI YIOUNI

, Ι Μ~Λ " III M K

4

5 Ι Ο '

0 4

fig. 5.42: Slightly closed painted vessels with rim angle from 91-1 14o.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 149

fig. 5.42, 1 rim sherd; W/R Area Code: 111 A2/1, 1 Colour: Int: 10R4/5; Ext: 10YR7-8/3 | 10R4/5 Diam: 14 cm? Red slip interior and exterior. Interior low gloss. Exterior white paint applied and surface burnished to high gloss.

fig. 5.42, 2 rim sherd; R/Ws Area Code:B6/i, 4 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/4-5 | 10YR7/3-4 Diam: 15 cm

fig. 5.42, 3 rim sherd; R/Ws Area Code: -/-, 43 Colour: Int: 2.5YR4/4; Ext: 2.5YR4/4 | 10YR6/3-4 Diam: 18 cm

fig. 5.42, 4 rim sherd, R/Ws Area Code: Β 1/3, ι Colour: Int: 2.5YR5/5; Ext: 2.5YR8/4 | 7.5YR7/5, 10YR7/5 Diam: 18 cm Portion of applied knob projects from wall of vessel. May be basal portion of handle.

fig. 5.42, 5 rim sherd; R/Ws Area Code.TX 3/1, 1 Colour: Int: 2.5YR5/4; Ext: 2.5YR4/4-5 | 10YR7/4 Diam: 19 cm?

fig. 5.42, 6 rim sherd; R/Ws Area Code: Β 1/3, 2 Colour: Int: 2.5YR4/4-5; Ext: 2.5YR4/4-6 | 10YR7/5 Diam: 18 cm

15O PARASKEVI YIOUNI

ι- ̂η ' r~m

, · fan

„Π^η „ι ^^ fig. 5.43: Open painted vessels with rim angle from 85-900.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 151

fig. 5.43, 1 rim sherd: R/Ws Area Code: B6/2, A,B6/X, 4 Colour: Int: 10R5/5, 2.5YR4/5; Ext: 10R4/4, 2.5YR4/ 4 I 7.5YR7/4-5, 10YR6 Diam: 18 χ 14 cm May be oval vessel

fig. 5.43, 2 rim sherd: R/Ws Area Code: D7/2, 1 Colour: Int: 2.5YR4/4; Ext: 2.5YR4/4-5 | 10YR7-8/3 Diam: 15 cm

fig. 5.43, 3 rim sherd; R/Ws Area Code: LI D3/2, 2 Colour: Int: 10R4/4; Ext: 10R4/5 | 10YR7-8/3 Diam: 12 cm Red slip interior. White slip exterior. Interior and exterior medium gloss and paint thin.

fig. 5.43, 4 rim sherd; R/Ws Area Code:B$/i, 2 Colour: Int: 2.5YR5/4-6; Ext: 2.5YR4/4-6 | 10YR8/4 Diam: 17 cm

fig. 5.43, 5 rim sherd; R/Ws Area Code: B4/2, 5 Colour: Int: 2.5YR5/5; Ext: 2.5YR4/5 | 10YR8/4 Diam: 15 cm

fig. 5.43, 6 rim sherd; R/Ws Area Code: LU B 1/3, 2 Colour: Int: 2.5YR4-5/4; Ext: 10R4/3-4 | 10YR7/1-2 Diam: 12 cm Red slip interior, matt to low gloss finish. Red paint over white slip exterior, paint thin and evenly applied. Low to medium gloss surface.

fig. 5.43, 7 rim sherd; R/Ws Area Code: E6/2, 4 Colour: Int: 5YR4-5/4; Ext: 2.5YR4/4 | 7.5YR5-6/2, 6/3 Diam: 15 cm Sparse and thin interior wash. Red paint over burnished body exterior. Low gloss finish. Unusually coarse ware.

fig. 5.43, 8 rim sherd; R/Ws Area Code: C6/2, 5 Colour: Int: 5YR4-5/4; Ext: 5YR4/3-4 | 10YR7/5 Diam: 6 cm? Unusual slip and colour. Thick streaked paint exterior over slip. Streaks parallel to edges of design.

fig. 5.43, 9 rim sherd; R/Ws Area Code: D5/2, 4 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR7-8/3

fig. 5.43, 10 rim sherd; R/Ws Area Code: D 1/3, 1 Colour: Int: 2.5YR4-5/5 | 7.5YR6/5; Ext: 2.5YR4/4, 5YR4-5/3-4 I 10YR7/2 I 10Y Diam: 12 cm Body visible both interior and exterior under slip.

fig. 5.43, 11 rim sherd; R/Ws Area Code:B8-g Baulk/1-2, 2 Colour: Int: 2.5YR4/5; Ext: 10R4/5 | 10YR7/4 Diam: 13 cm

fig. 5.43, 12 rim sherd; R/Ws Area Code: C6/2, 3 Colour: Int: 5YR3-4/2-4; Ext: 2.5YR4/4 | 10YR7/3 Diam: 16 cm

fig. 5.43, 13 rim sherd; R/Ws AreaCode.TX 1/1, 1 Colour: Int: 10R4/4, 2.5YR2-3/2; Ext: 10R4/4 | 7.5YR7/4-5 Diam: 15 cm

fig. 5.43, 14 rim sherd; R/Ws Area Code: LXI B4/2, 3 Colour: Int: 2.5YR4/5-4; Ext: 2.5YR4/4 | 10YR7-8/2 Diam: 18 cm Red slip interior. Red paint over white slip exterior. Both surfaces eroded to matt finish.

152 PARASKEVI YIOUNI

4

fig. 5.44: Open painted vessels with rim angle from 85-900.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 153

fig. 5.44, 1 rim sherd; R/Ws Area Code: A2/2, 3 Colour: Int: 10R4-5/4; Ext: 10R4/4-5 I 10YR7-8/3 Diam: 17 cm? Red slip interior, matt to low gloss finish. Red paint over light slip exterior, low to medium gloss finish.

fig. 5.44, 2 rim sherd; R/Ws Area Code: E8/3 A, 1 Colour: Int: 2.5YR4/5-6; Ext: 2.5YR4/5 | 10YR7/3 Diam: 16 cm

fig. 5.44, 3 rim sherd; R/Ws Area Code: B2/2, 4 Colour: Int: 2.5YR4/5 | 10YR8/4-5; Ε*ί: 2.5YR4/6 | 10YR8/4-5 Diam: 18 cm Red paint over white slip interior and exterior. Much of interior surface removed by erosion.

fig. 5.44, 4 rim sherd; R/Ws Area Code: E8/2, 1 Colour: Int: 2.5YR5/5; E*t: 2.5YR4/5 I 10YR4/4 Diam: 17 cm

fig. 5.44, 5 rim sherd; W/R Area Code:-/-, 16 Colour: Int: 2.5YR4-5/4-5; Ext: 10YR7/2-3 | 2.5YR4-5/4-5 Diam: 16 cm

fig. 5.44, 6 rim sherd; W/R Area Code: L6/K, 23 Colour: Int: 2.5YR4-5/5-6; Ext: 10YR8/2-3 I 2.5YR4-5/4-6 Diam: 16 cm Red slip interior. White paint over red slip exterior. Paint applied over scratched design outline. Medium gloss finish. (White eroded to matt).

FIG. 5.44, 7 rim sherd; R/Ws Area Code: Bi/ 2, 2 Colour: Int: 2.5YR5/5; Ε*1: 2-5**4/5 I 10YR8/4 Diam: 17 cm

fig. 5.44, 8 rim sherd; R/Ws Area Code: -/-, 56 Colour: Int: 10R4/4-5; E*t: 10R4/5 | 10YR7/2 Diam: 16 cm

fig. 5.44, 9 rim sherd; R/Ws Area Code: B2/2, 3 Colour: Int: 2.5YR5/6; Ext: 2.5YR4/6 | 10YR8/4 Diam: 16 cm

fig. 5.44, 10 rim sherd; R/Ws Area Code: -/-, 58 Colour: Int: 2.5YR4/5; Ext: 10R4/4 | 10YR7/3 Diam: 14 cm

fig. 5.44, 11 rim sherd; R/Ws Area Code: B7/ Ί, 10 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR7/4 Diam: 13 cm

fig. 5.44, 12 rim sherd; R/Ws Area Code: C5/2 Pit A, 1; C5/2, 9 Colour: Int: 2.5YR4-5/5; Ext: 2.5YR4/5-6 | 10YR7/ 3-4 Diam: 1 1 cm Paint applied using outline and infill technique.

fig. 5.44, 13 rim sherd; R/Ws Area Code: E7/2 Pit A, 1 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/4-5 | 10YR7/3 Diam: 9 cm

fig. 5.44, 14 rim sherd; R/Ws Area Code: D 9/2, 4 Colour: Int: 2.5YR4/4-5; Ext: 2.5YR4/5-6 | ioyr 7/4 Diam: 13 cm

154 PARASKEVI YIOUNI

3 I Ws ι 4 I qp

. ΓΚ7) . r~^

13 14 fig. 5.45: Open painted vessels with rim angle from 85-900.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY i55

FIG. 5.45, 1 rim sherd; R/Ws Area Code: L6/1, 1 Colour: Int: 2.5YR2-5/2-4; Ext: 2.5YR3-4/4-6 | 10YR6-7/3-4 D iam: 14 cm

FIG. 5.45, 2 rim sherd; R/Ws Area Code: A 1/3, 6 Colour: Int: 2.5YR4/4; Ext: 2.5YR4/4-5 | ioyr6- 7/3 D iam: 13 cm Red slip interior, matt to low gloss finish. Red paint over light slip exterior, low to medium gloss finish.

fig. 5.45, 3 rim sherd; R/Ws Area Code: B6/2, 9 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 7.5YR7/4-5 Diam: 15 cm

fig. 5.45, 4 rim sherd; R/Ws Area Code: C4/1, 1 Colour: Int: 2.5YR5/4 | 7.5YR6/4; Ext: 2.5YR5/5 I 10YR7/3 Diam: 16 cm Exterior paint applied clearly using outline and infill technique. Interior only partially slipped.

fig. 5.45, 5 rim sherd; R/Ws Area Code: ¥6/1 C wall early bid., 1 Colour: Int: 2.5YR4/4; Ext: 2.5YR4/4-5 I 10YR7/ 3-4 Diam: 19 cm Very fine piece with even high gloss finish exterior.

fig. 5.45, 6 rim sherd; R/Ws Area Code: A6/R, 1 Colour: Int: 10R2-4/1-2; Ext: 2-5YR2-3/2-4,4/6 | 10YR5-7/2, 5/1 Diam: 14 cm

fig. 5.45, 7 rim sherd; W/R Area Code: Plough, 9

fig. 5.45, 8 rim sherd; R/Ws Area Code:Bz/i, 1 Colour: Int: 2.5YR5/5; £**-' 2.5YR4/5 | 10YR7-8/4 Diam: 16 cm

fig. 5·45> 9 rim sherd; R/Ws Area Code: C7/1, 2 Colour: Int: 2.5YR4-5/4; Ext: 2.5YR4/5 | ioyr8/ 3-4 Diam: 13 cm

fig. 5.45, 10 rim sherd; R/Ws Area Code: LXII A5/1, 1 Colour: Int: 2.5YR2-3/2; Ext: 2.5YR5/5 I ioyr6/ 3-4 Diam: 14 cm Red slip interior, low to medium gloss surface. Red paint over body exterior. Paint thin and evenly applied, medium gloss finish.

fig. 5.45, 1 1 rim sherd; R/Ws Area Code: -/-, 67 Colour: Int: 2.5YR4-5/5-6; Ext: 2.5YR4/5-6 | 10YR7/4 Diam: 13 cm

fig. 5.45, 12 rim sherd; R/Ws Area Code: A 2/3, 8 Colour: Int: 2.5YR5/4-6; Ext: 10R4/4 | 10YR7/2-3 Diam: 16 cm Red slip interior; red paint over light slip exterior. Both surfaces medium gloss finish.

fig. 5.45» *3 rim sherd; R/Ws Area Code: L5/1 - Plough, 5 Colour: Int: 5YR3-4/2-4; Ext: 2.5YR3-4/5-6 | 7.5YR7/4-5, 10YR7/4-5 Diam: 16.5 cm Interior reddish slip.

fig. 5.45, 14 rim sherd; R/Ws A rea Code: C 1 0/ 2 , 1 Colour: Int: 5YR5/5; Ext: 2.5YR4/4-5 | 10YR7/4 Diam: 18+ cm

156 PARASKEVI YIOUNI

ΠΕ . TO ,rm ΓΤ7Ί

Π^ .ΓΊΤ7

V=™=d4 14

fig. 5.46: Open painted vessels with rim angle from 85-900 (1-7) and from 60-84° (8-14).

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 157

FIG. 5.46, 1 rim sherd; R/Ws Area Code:C*j/i, 4 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR7-8/4 Diam: 8.5 cm?

fig. 5.46, 2 rim sherd; R/Ws Area Code.O^/i, 1 Colour: Int: 2.5YR5/5-6; Ext: 2.5YR4/6 | 10YR8/3 Diam: 9 cm

fig. 5.46, 3 rim sherd; R/Ws Area Code: O6/ la, 2 Colour: Int: 2.5YR5/4; Ext: 7.5YR5-6/4 | 10YR7/3 Diam: 1 1 cm Red slip interior. Red paint over light slip exterior. Low to medium gloss.

fig. 5.46, 4 rim sherd; R/Ws Area Code: C6/2, 4 Colour: Int: 2.5YR3-4/3-4; Ext: 10YR7/3 | 2.5YR2-4/2 Diam: 12 cm Reddish slip interior. White paint over red slip exterior. White paint thick and uneven and shows no signs of gloss now but surface may have eroded.

fig. 5.46, 5 rim sherd; R/Ws Area Code: C 10/1, 4 Colour: Int: 10R4/4; Ext: 10R4/4 I 7-5YR7/5> 10YR7/4 Diam: 14 cm

fig. 5.46, 6 rim sherd; R/Ws Area Code:0>j/iy 7 Colour: Int: 2.5YR3-4/1-3; Ext: 2.5YR4/2-4 | 10YR7/2-3 Diam: 16 cm? Red slip interior. Red paint over light slip exterior.

fig. 5.46, 7 rim sherd; R/Ws Area Code: 06/ 1,4 Colour: Int: 2.5YR5/5; Ext: 2.5YR4/5 | 10YR7/4 Diam: 17 cm Red slip interior. Red paint on light slip exterior.

fig. 5.46, 8 rim sherd; R/Ws Area Code: A2/2, 1 Colour: Int: 7.5YR6/4, 2.5YR5/4; Ext: 2.5YR4/5 | 10YR8/3 Diam: 20 cm

Red slip interior, matt to low gloss. Red paint over white slip exterior, low to medium gloss.

fig. 5.46, 9 rim sherd, R/Ws Area Code: D 1 0/ 1 , 5 Colour: Int: 10R3-4/3-4; Ext: 10R3/2-5 | ioyr6- 7/2-3 Diam: 16-17 cm

fig. 5.46, 10 rim sherd; W/R Area Code: LXI E5/3, 4 Colour: Int: 5YR2-3/1-2; Ext: 10YR7/2 | 5YR3-4/ 1-2 Diam: 22 cm Dark slip interior and exterior. Interior low gloss unevenly streaked finish. Exterior white paint over slip, low to medium gloss finish. Some eroded matt areas on white paint.

fig. 5.46, 11 rim sherd; R/Ws Area Code:¥j/i, 1 Colour: Int: 2.5YR4/5; Ext: 10R4/5 | 10YR7/5 Diam: 16 cm

fig. 5.46, 12 rim sherd; R/Ws Area Code: C 5/1, 1 Colour: Int: 5YR3-4/2; Ext: 5YR4-5/4 | 10YR7/3 Diam: 13.5 cm Only very faint traces of interior slip.

fig. 5.46, 13 rim sherd; R/Ws Area Code: A$/i, 3 Colour: Int: 5YR5/4-5; Ext: 2.5YR4/4-6, 5YR4/4 | 10YR7/4 Diam: 1 1 cm Red slip interior. Matt to low gloss finish. Red paint over light slip exterior, low to medium gloss finish.

FIG. 5.46, 14 rim sherd; R/Ws Area Code:B8-$ Baulk/ 1-2, 1 Colour: Int: 5YR5/ 4 | 7.5YR6/4; Ext: 5YR5/4 | 10YR7/ 2-4 Diam: 15 cm Possibly porcelain ware. Reddish slip or wash over body interior.

158 PARASKEVI YIOUNI

■ ) »1 . ' ψ>1

fig. 5.47: Open painted vessels with rim angle from 60-840.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 159

FIG. 5.47, 1 rim sherd; R/Ws Area Code: Ύ Υ 10/2, 2 Colour: Int: 10R4/4-5; Ext: 10YR8/4 | 10R4/4-5 Diam: 16 cm Red slip interior. White paint over red slip exterior. Medium gloss finish.

fig. 5.47, 2 rim sherd; W/R Area Code: Coo/ Ί, 1 Colour: Int: 10R4/4; Ext: 10YR8/3 | 10R4/4 Diam: 17 cm? Red slip interior; low gloss finish. Exterior, white paint over red slip, medium gloss finish except where white eroded.

fig. 5.47, 3 rim sherd; R/Ws Area Code: C2/2, 3 Colour: Int: 2.5YR2-4/2; Ext: 2.5YR3-4/4 | 10YR4/ 3-4 Diam: 13 cm

fig. 5.47, 4 rim sherd; R/Ws Area Code: D 9/ 2, 1 Colour: Int: 2.5YR4/6; Ext: 10R4/6 | 10YR7/3-4 Diam: 17 cm

fig. 5.47, 5 rim sherd; R/Ws Area Code: B7/ Ί, 2 Colour: Int: 2.5YR5/5-6; Ext: 2.5YR4/6 | ioyr8/ 2-3 Diam: 14 cm

fig. 5.47, 6 rim sherd; R/Ws Area Code:'!>'/', 3 Colour: Int: 2.5YR4/6; Ext: 2.5YR4/6 | 10YR8/4 Diam: 12 cm

FIG. 5.47, 7 rim sherd; R/Ws Area Code: O6/ "i, 1 Colour: Int: 10R2-4/1-3; Ext: 10R3-4/2-3 | ioyr6- 7/2-3 Diam: 18 cm Red slip interior. Red paint on light slip external paint fairly thick.

fig. 5.47, 8 rim sherd; R/Ws Area Code: C5/2 A, 1 Colour: Int: 2.5YR 4-5/5; Ext: 2.5YR4/6 | 10YR7/3 Diam: 15 cm

FIG. 5.47, 9 rim sherd; R/Ws Area Code: UlBi/Y. B, 1 Colour: Int: 2.5YR5/4 j 10YR7/3; Ext: 2.5YR6/6 | 10YR7/3 Diam: 14 cm Red slip over body interior, no design, low gloss. Red paint over light colour slip exterior. Paint thin and even. Burnished lightly to matt or low gloss.

fig. 5.47, 10 rim sherd; R/Ws Area Code: L5/1, 1 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR7-8/3 Diam: 16 cm

fig. 5.47, 11 rim sherd; R/Ws Area Code: B9/2, 4 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR7-8/3 Diam: 16 cm

fig. 5.47, 12 rim sherd; R/Ws Area Code: A1/2, 6 Colour: Int: 10R4/3-4; Ext: 10R4/3 | 10YR7/8-3 Diam: 16 cm Red slip interior and red paint over white slip exterior. Eroded to low gloss finish.

fig. 5.47, 13 rim sherd; R/Ws Area Code: D8/2, 1 Colour: Int: 10YR4-6/1; Ext: 10YR2-3/1 | 10YR7- 8/1 Diam: 18 cm Burned porcelain ware. Light slip interior. Paint over light slip exterior.

i6o PARASKEVI YIOUNI

,ι m ' ,ι ία ^

„I Ö 1 I

ΜΓΛ - ,.rm

fig. 5.48: Open painted vessels with rim angle from 60-840.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 161

FIG. 5.48, 1 rim sherd; R/Ws Area Code: LU A3/1, 1 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR6/3-4 Diam: 21 cm Red slip interior. Red paint over burnished body exterior. Unevenly burnished. Highlights of low gloss on exterior surface.

fig. 5.48, 2 rim sherd; R/Ws Area Code: -/-, 50 Colour: Int: 2.5YR5/4; Ext: 10R4/5 | 10YR6-7/4 Diam: 17 cm

FIG. 5.48, 3 rim sherd; R/Ws Area Code: D$/i, 2 Colour: Int: 2.5YR4/5; Ext: 10R3/6 | 10YR7-8/4 Diam: 15 cm

fig. 5.48, 4 rim sherd; R/Ws Area Code:O>j/i, 1 Colour: Int: 2.5YR4-5/4; Ext: 2.5YR4-5/6 | 10YR8/2-3 Diam: 14 cm

fig. 5.48, 5 rim sherd; W/R Area Code.TY 10/2, 1 Colour: Int: 2.5YR3/2-6; Ext: 10YR8/3 | 2.5YR3/2-6 Diam: 14 cm? Red slip interior. White paint over red slip exterior. Medium gloss finish.

fig. 5.48, 6 rim sherd; R/Ws Area Code.-Oq/i, 1 Colour: Int: 10R4/4 | 7.5YR6/4; Ext: 10R4/5 | 10YR7-8/3 Diam: 18 cm Interior unevenly slipped.

fig. 5.48, 7 rim sherd; R/Ws Area Code: A 1/3, 4 Colour: Int: 2.5YR5/4-5; Ext: 2.5YR4/5 | 10YR7-8/3 Diam: 14 cm Red slip interior; matt to low gloss finish. Red paint over white slip exterior; low to medium gloss finish.

fig. 5.48, 8 rim sherd; R/Ws Area Code: AS/ 1, 2 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/6 | 10YR7-8/2 Diam: 17 cm

fig. 5.48, 9 rim sherd; R/Ws Area Code: D7/ Ί, 2 Colour: Int: 2.5YR5/5; Ext: 10R4/4 | 10YR7-8/4 Diam: 18 cm

fig. 5.48, 10 rim sherd; R/Ws Area Code: C3/2, 1 Colour: Int: 2.5YR4/4-5; Ext: 2.5YR4/5-6 | 10YR7/ 3-4 Diam: 18 cm

fig. 5.48, 11 rim sherd; R/Ws Area Code: LXI E4/2, 3 Colour: Int: 10R5/4-5; Ext: 10R5/4-5 | 10YR7/3 Diam: 17 cm Red slip interior, low gloss finish. Red paint over white slip exterior, low to medium gloss finish.

fig. 5.48, 12 rim sherd; R/Ws Area Code: D 0/2, 2 Colour: Int: 2.5YR5/5 | 7.5YR6/4; Ext: 10R4/5-6, 2.5YR4/5-6 I 10YR7/3-4 Diam: 12 cm

fig. 5.48, 13 rim sherd; R/Ws Area Code: C6/1, 2 Colour: Int: 2.5YR4/5-6; Ext: 2.5YR4/5 | 10YR8/4 Diam: 21 cm?

fig. 5.48, 14 rim sherd; R/Ws Area Code: -/-, 49 Colour: Int: 2.5YR4-5/4, 2-3/2; Ext: 2.5YR2-4/2- 4 I 10YR7/3 Diam: 18-19 cm

fig. 5.48, 15 rim sherd; R/Ws Area Code: O6/ 1, 3 Colour: Int: 7.5YR6-5/4; Ext: 2.5YR4-5/4 | 10YR7/3-4 Diam: 18 cm Interior red slip, low gloss finish. Exterior red paint on light slip, medium gloss finish. Paint smeared.

fig. 5.48, 16 rim sherd; R/Ws Area Code: D 1/2, 3 Colour: Int: 2.5YR5/4-5; Ext: 2.5YR4/4-5 | 10YR8/2-3 Diam: 15 cm Exterior paint applied using outline and infill technique.

i62 PARASKEVI YIOUNI

. ̂ S~7 ' τ - m - 7

.* w - ? ' ΙΓΊ

16 V^rf

fig. 5.49: Open painted vessels with rim angle from 60-840.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 163

FIG. 5.49, 1 rim sherd; R/Ws Area Code: AB6/2 X, 7 Colour: Int: 10R4/4 | 10YR6/3-4; Ext: 10YR6/3-4 Diam: 20 cm Exterior unpainted.

fig. 5.49, 2 rim sherd; R/Ws Area Code: L6/K, 4 Colour: Int: 2.5YR5/6; Ext: 2.5YR4/5 | 10YR8/3 Diam: 16 cm

fig. 5.49, 3 rim sherd; W/R Area Code: -/-, 26 Colour: Int: 10R3-4/4-5; Ext: 10YR8/2-3 | 10R3-4/4-5 Diam: 15 cm

fig. 5.49, 4 rim sherd; R/Ws Area Code: K8/1, 2 Colour: Int: 2.5YR4/6; Ext: 5YR4/6 | 10YR7/4 Diam: 27 cm?

fig. 5.49, 5 rim sherd; R/Ws Area Code: -/-, 77 Colour: Int: 2.5YR4/5 | 7.5YR6-7/4; Ext: 2.5YR4/4 | 10YR8/3 Diam: 20-2 1 cm Interior rim band 2 cm over body.

fig. 5.49, 6 rim sherd; W/R Area Code:-/-, 14 Colour: Int: 10R4/5; Ext: 10YR8/4 | 10R4/5 Diam: 15-16 cm

fig. 5.49, 7 rim sherd; W/R Area Code: LXII A5/2, 2 Colour: Int: 2.5YR5/5 | 10YR6/3; Ext: 2.5YR5/5 | 10YR7/3 Diam: ? Red paint over body interior and exterior. Paint thin and evenly applied. Surface eroded but traces of low gloss finish remain.

fig. 5.49, 8 rim sherd; R/Ws Area Code: LXI D5/2, 3 Colour: Int: 5YR6-7/4 | 7.5YR6-7/2; Ext: 2.5YR6/5 | 10YR6-7/2 Diam: 22 cm Interior very uneven and sparse pink slip over body, matt finish. Exterior: pink over light slip. Paint much eroded. Medium gloss finish.

fig. 5.49, 9 rim sherd; R/Ws Area Code: L6/1 K, 13 Cobur: Int: 10YR7-8/2; Ext: 2.5YR3-4/5 | 10YR8/2 Diam: 17 cm Porcelain ware.

fig. 5.49, 10 rim sherd; R/Ws Area Code: B2/2, 2 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR8/3-4 Diam: 17 cm?

fig. 5.49, 11 rim sherd; R/Ws Area Code: F7/1, 2 Colour: Int: 2.5YR4/4-5; Ext: 2.5YR4/4-5 | 10YR7/4 Diam: 16 cm

fig. 5.49, 12 rim sherd; R/Ws Area Code: B8/1, 1 Colour: Int: 2.5YR4/5; Ext: 2.5YR4/5 | 10YR7-8/4 Diam: 10.5 cm

fig. 5.49, 13 rim sherd; R/Ws Area Code: LI E1/1, 1 Colour: Int: 2.5YR4/5 | 10YR7/3-4; Ext: 2.5YR4/5 | 10YR6-7/3-4 Diam: 22 cm Red paint over white slip interior and exterior. Medium burnish. Paint thin with uneven edges. Applied using outline and infill technique with short strokes.

fig. 5.49, 14 rim sherd; W/R Area Code: H8/ ia, 1 Colour: Int: 5YR4/4; Ext: 2.5YR4/5-6 | 7.5YR7/5 Diam: 15 cm

FIG. 5.49, 15 rim sherd; W/R Area Code: H M Plough, 1 Colour: Int: 2.5YR3-4/6; Ext: 10YR7-8/4 | 2.5YR3-4/6 Diam: 14.5-15 cm Red slip interior. White thick and raised paint over red slip exterior. Both surfaces medium gloss finish. Surface now badly eroded.

fig. 5.49, 16 rim sherd; R/Ws Area Code: C00/2, F. 2, 1 Colour: Int: 10YR4/5 | 10YR7/3; Ext: 10R4/5 | 10YR7/3 Diam: 20 cm Red paint over white slip interior and exterior.

i64 PARASKEVI YIOUNI

> " / Λ WT~7

0 4

fig. 5.50: Open painted vessels with rim angle less than 6o°.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 165

FIG. 5.50, 1 rim sherd; R/Ws Area Code: C2/ 1 y 3 Colour: Int: 2.5YR4/4; Ext: 2.5YR3-4/4 | 10YR7/3 Diam: 12 cm

fig. 5.50, 2 rim sherd; R/Ws Area Code: G7/ia, 3 Colour: Int: 2.5YR4/5-6; Ext: 2.5YR4/6-7 | 7.5YR7/ 4-5 Diam: 17 cm

fig. 5.50, 3 rim sherd; R/Ws Area Code: K6/1, 2 Colour: Int: 2.5YR4/6 | 10YR7/4; Ext: 5YR4/6, 2.5YR4/6 I 10YR7/4 Diam: 16 cm Red paint over light slip interior and exterior.

fig. 5.50, 4 rim sherd; R/Ws Area Code: LI Ε 1/3, ι Colour: Int: 2.5YR4-5/2; Ext: 2.5YR2-3/2 | ioyr 4/1 Diam: 18 cm Red slip interior. Thin paint over thin white slip exterior. Interior low gloss, exterior medium gloss.

fig. 5.50, 5 rim sherd; R/Ws Area Code: A8/2 F .A, 2 Colour: Int: 10R3/5 I10YR7/3-4; Ext: 10R3/5 | 10YR4/3 Diam: 20-21 cm Red paint over light slip interior and exterior.

fig. 5.50, 6 rim sherd; W/R Area Code: -/-, 13

fig. 5.50, 7 rim sherd; W/R Area Code: Tg 3/4, 1 Colour: Int: 10YR8/1-2 | 5YR4/2-3; Ext: ioyr8/i- 2 I 2.5YR4/2-4 Diam: 22 cm? Thick raised white paint over reddish slip. No sign of burnish on white, red medium gloss finish - both surfaces.

fig. 5.50, 8 rim sherd; R/Ws Area Code: A2/3, 3 Colour: Int: 10R4-5/5 | 10YR7/4-5; Ext: 10R4-5/5I 10YR7/4-5 Diam: 18 cm Interior and exterior red paint over white slip. Much smeared, low gloss finish.

fig. 5.50, 9 rim sherd; W/R Area Code: -/-, 79 Colour: Int: 10YR2-3/1; Ext: 10R4/4 | 7.5YR6-5/4 Diam: ? Vessel foot. Interior rough and unslipped. Exterior paint thick and raised, unevenly applied over body (unslipped).

i66 PARASKEVI YIOUNI

( <® / η

7 w ^J^^^J* 0 4

fig. 5.51: Painted body sherds.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 167

fig. 5.51, 1 body sherd; R/Ws Area Code: A1/2, 9 Colour: Int: 2.5YR5/5, 5YR6/3; Ext: 2.5YR4/5 | 10YR7-8/3 Red slip over body interior, eroded to low gloss finish. Red paint over white slip exterior applied using outline and infill technique. Paint thick and slightly raised. Eroded to low gloss.

fig. 5.51, 2 body sherd; R/Wp Area Code: L6/K, Colour: Int: 10YR6-7/4; Ext: 2.5YR4-5/4-5 | 10YR6-7/4 Porcelain ware.

fig. 5.51, 3 body sherd; R/Wp Area Code:TX6/i9 1 Colour: Int: 10YR7/3-4; Ext: 2.5YR4/4-6 | 10YR7/ 3-4 Porcelain ware. Interior badly eroded to matt finish, slipped. Exterior red paint over light slip eroded to matt finish.

fig. 5.51, 4 body sherd; R/Wp Area Code.TY 2/1,5 Colour: Int: 10YR7-5/3; Ext: 5YR4-5/4-6 | 10YR7- 8/3 Porcelain ware. White slip interior, matt to low gloss. Exterior red paint over light slip, high to medium gloss finish.

fig. 5.51, 5 body sherd; R/Wp Area Code.TX 4/1, 5 Colour: Int: 10YR7/3; Ext: 2.5YR4/6, 5YR4-5/4-5 | 10YR7/3 Porcelain ware. White slip interior. Red paint over white slip exterior. Medium to high gloss finish.

fig. 5.51, 6 body sherd; R/Wp Area Code: Τ Υ 4/1, 4 Colour: Int: 10YR7-8/2-3; Ext: 2.5YR4/5, 5YR5/ 4-5 I 10YR8/3-4 Porcelain ware. Light slip interior, low to medium gloss finish. Red paint over white slip exterior. Medium to high gloss finish.

fig. 5.51, 7 body sherd; R/Wp Area Code: D8/2, 5 Colour: Int: 7.5YR6-7/2-4; Ext: 2.5YR4/6 | ioyr8/ 2-3 Probably porcelain ware. Traces of light slip interior. Exterior red paint over white slip. Medium to high gloss finish.

fig. 5.51, 8 body sherd; Pink/Pink Area Code: F 4/1, 8a Colour: Int: 2.5YR6/4; Ext: 2.5YR6/4 | 2.5YR4/2-4 'Pink over pink' exterior. Light overall slip or wash. Design produced by wiping away wet slip to expose red underslip.

i68 PARASKEVI YIOUNI

1

{SV ^ 0 4

fig. 5.52: Painted base sherds.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 169

fig. 5.52, 1 base sherd; R/Ws Area Code: LLV 3.30-3.40 [1] Colour: Int: ior5/4-5,2.5YR4-3/s> | 10YR5/2- 4,7.5YR5/4; Ext: 2.5 YR5/4-5 I 10YR7/2-3 Base Ext: 2.5YR5/4 Diam: 7 cm Medium gloss burnish. Paint slightly raised somewhat smeared. Interior also burnished.

fig. 5.52, 2 base sherd; R/Ws Area Code: C1/3, 3 Colour: Int: 2.5YR4-5/4 | 7.5YR5-6/3; Ext: 2.5YR4/ 5 I loR7/3 Diam: 6 cm Interior slip eroded and now only traces remain over body.

fig. 5.52, 3 base sherd; R/Ws Area Code: B6/1, B6/2, AB6/2 X, 8 Colour: Int: 2.5YR5-6/5; Ext: 2.5YR5-6/5-6 | 10YR8/3-4 Diam: 5.5 cm

fig. 5.52, 4 base sherd; R/Ws Area Code: C5/2, 5 Colour: Int: 5YR5/4; Ext: 5YR4-5/5 | 10YR7-8/4 Diam: 14 cm

fig. 5.52, 5 base sherd; R/Ws Area Code: O7/1, 7/2, 10 Colour: Int: 7.5 YR6/4; Ext: 10R4/3-4 | 10YR8/4 Diam: 5.75 cm Interior: bottom unslipped but top red slip over body, low gloss finish; base roughly smoothed. Exterior red paint over light slip. Applied using clear outline and infill technique. Medium gloss eroded to matt finish.

fig. 5.52, 6 base sherd; R/Ws Area Code: Do/ 2, 5 Colour: Int: 2.5YR5/4 | 7.5YR6/4; Ext: 2.5YR4/4-6 | 10YR7/4, 7.5YR7/4-5 Diam: 6 cm Interior unevenly slipped over body.

17O PARASKEVI YIOUNI

2

4

6

7

FiG. 5.53: Applied decoration: sherds with raised bands.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 171

1 2

6

8

9 '^-^

fig. 5.54: Applied decoration: sherds with raised bands and blobs.

172 PARASKEVI YIOUNI

1 1

^^

Γ] Γ" Γ^ Γ" Ι /í>sv,) 1 (jjX ■ Cftw^y | 6 Μ 5

4

7 8

• 9 ^^^ 10

1 4

fig. ̂.55: Impressed decoration: vessels decorated with finger pinchings (1-4), finger nail impressions (5-8) and instruments (9-11).

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 173

1

fig. 5.56: Bases decorated with sharp pointed instrument.

174 PARASKEVI YIOUNI

fig. 5.57: Plain pottery: correlation of surface treatment and fabric types.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 175

fig. 5.58: Plain pottery: correlation of surface treatment and paste quality.

176 PARASKEVI YIOUNI

fig. 5.59: Relative frequency of fabric types in plain, impressed and applied pottery.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 177

fig. 5.60: Relative frequency of surface coating in plain, impressed and applied pottery.

fig. 5.61: Relative frequency of fine, medium and coarse vessels in plain, painted, impressed and applied pottery.

178 PARASKEVI YIOUNI

fig. 5.6?: Relative frequency of vessel shapes in plain, painted and impressed pottery.

THE EARLY NEOLITHIC POTTERY: TYPOLOGY 179

fig. 5.63: Relative frequency of open pots, closed pots and neck jars in first and second building phases.

fig. 5.64: Relative frequency of different vessel forms from the three building phases.

i8o PARASKEVI YIOUNI

fig. 5.65: Relative frequency of surface coating in the three building phases.

fig. 5.66: Relative frequency of fabric types in the three building phases.

Chapter 6 The Early Neolithic Pottery: Functional Analysis

6.1 INTRODUCTION

In recent years, a considerable amount of archaeological research has been oriented towards the functional analyses of ceramics. Such studies are based on information obtained from two interrelated approaches: ethnoarchaeological research and technological investigation of experimental pottery.

The ethnography of pottery has a greater literature than any other material culture analysis. Studies generally give information on the number of pots in individual households, on their form and stated use.1 In some works, information on vessel life expectancy and recycling of ceramics is included.2

Technological investigation provides information on the contribution of different materials (types of fabric and non-plastic inclusions) and techniques, to vessel suitability for particular functions. Such studies have developed mainly after the ig7o's.s

Functional analyses of Early Neolithic ceramics from the Balkans have been published in two works. The first concerns the material from Early and Middle Neolithic Franchthi (Vitelli 1989). This study concludes that pottery manufacture was very infrequent during the Early Neolithic period, with an average production of 12-13 pots per year. From the estimated number of vessels, Vitelli concludes that the inhabitants of Early Neolithic Franchthi did not have enough vessels for seed or harvest storage. Furthermore, evidence for the use of ceramics for cooking could be found only in the Middle Neolithic material. In the second publication, Gardner, examining the pottery fabrics from Anzabegovo (Anza), Achilleion and Sitagroi, concludes that vessels were not used near the fire (Gardner 1978). Instead, according to Gardner, vessels were used for storage, or as display dishes.

6.2 ESTIMATION OF THE NUMBER OF POTS FROM ΝΕΑ NIKOMEDEIA

A rough sense of the scale of local pottery production is a basic prerequisite for studying the function of vessels. In Nea Nikomedeia, as in every other Neolithic site of the Balkans, an impressive number of pottery sherds has been found.

1 A review of the literature can be found in Kramer 1979 and Rice 1987. * e.g. Foster i960; David and Henning 1972; Stanislawski 1977; DeBoer and Lathrap 1979; Longacre 1985.

5 A review of the research into the mechanical and thermal properties of ceramics can be found in Bronitsky 1986.

181

i82 PARASKEVI YIOUNI

Several methods can be used in order to measure the quantity of the pottery (Millet 1980; Orton 1982). The method which was used for the estimation of the number of pots from Nea Nikomedeia was determined by the available data. During his post-excavation work Dr. Rodden counted the sherds recovered in every square/spit of the main excavation grid (squares A1-M8) and measured their surface area The total surface area of the ceramic material from the main excavation grid was calculated to be equal to 1,343,400 cm2. Subsequently, only the feature sherds were stored in the Museum. In order to use the preliminary work carried out by the excavator it was decided to use as an estimator of the equivalent number of pots, the surface area of the sherds. For the calculation of the average surface area of the Nea Nikomedeia pot, the following method was developed by the present author.

Initially, the surface area of a pot with rim angle from 85-900 was calculated. The measured vessel was chosen so because its ratio of rim diameter/height was equal to the estimated average ratio for this particular vessel form. The average ratio of rim diameter/ height was determined from a number of whole and reconstructible pots. For the estimation of the surface area of the vessel with rim angle from 85-900, the pot was divided into a number of conical segments, of 1 cm height (fig. 6.1). The surface area of each segment was calculated from the following expression:

S = π * (Ra + RJ * h where R , B^ are the radii, and h is the oblique height of the conical segment. The surface area of the base was calculated from the following equation

Sb." *R2

and it was added to the total. Using the data from the histogram of rim diameters for the vessels with rim angle from

85-900 (fig. 5.28), the surface area of each sherd was calculated, and consequently the average surface area for this vessel form was estimated. The basic assumption for this generalization is that the ratio of rim diameter/height is constant. This means that by changing the rim diameter of the pot, all other dimensions {i.e. Ra, Rb, Rc ... , and h) would change proportionally.

The same method was used to calculate the average surface area of all other vessel forms. In table 6. 1 the average surface area and the occurrence of each vessel form are presented. As the surface area is directly related to the size of the pot, vessel forms which include tall pots with large rim diameters have a larger average surface area (from table 6. 1 compare the average surface area of the vessels with rim angle larger than 1 1 50 with that of the vessels with rim angle less than 6o°).

The average surface area of the Nea Nikomedeia pot was calculated to be equal to 1 205 cm2. As has been explained above, for this calculation the average surface area of each vessel form and its occurrence were used. Dividing the total surface area of the recovered ceramic material from the main grid (1,343,400 cm2), by the average surface area of the Nea Nikomedeia pot (1205 cm2), the equivalent number of vessels was found to be equal to 1115.

The underlying assumption of these estimations is that the number of sherds into which a vessel breaks is random, and independent of the size of the pot. In reality, larger vessels will break into a larger number of sherds. This will bias the distribution of rim diameters

THE EARLY NEOLITHIC POTTERY: FUNCTIONAL ANALYSIS 183

Surface Area

*(Ra + Rb)h

R1 = 10.2 cm

d =10.5 cm 65.0 cm Π2 2

r^ =10.5 cm 66.0 cm

R =10.5 cm 66.0 cm é 2

^ =10.5 cm 66.0 cm

F_, =10.5 cm 66.0 cm

_, _10_4 cm J 65·7 cm'

7 2 Ral10.2an / 64.7 cm

/ 2

R_ __9.j_cm / 63.1 cm

Rlo_9_ciii / 66·3 cm

_____Ql1_=_ 8^7 cm ______ _ / 62.2 cm _ / 2

R_2_l7_ic__ / 62.5 cm

y 2 _-.___ _3_6-7cm __ _ yS

68.85cm

yS 2 Rl4=_5SLC}L·^^

73·6 cm

^^~^*^^ 2 2

/ surface of base (nR ) = 78.5 cm

Surface of Vessel = 853 cm

fig. 6. 1 : Calculation of the surface area of a vessel with rim angle 85-900.

Table 6.1: Average surface area (cm2), volume (cm5), occurrence and the diameter/height ratio of the different vessel forms from Nea Nikomedeia.

Vessel Ratio Surface Volume Occurrence

neckjar 0.59 1942 10416 9%

larger than 115o 0.61 4540 28032 3%

θ1"11^ ΐ·θ4 1451 7886 33%

85~9Ο° !·52 8ο8 358° 25%

60-840 2.11 725 2850 23% less than 60o 2.58 580 X9l6 6%

i84 PARASKEVI YIOUNI

towards the larger values. Consequently, the estimated average area of the Nea Nikomedeia pot (1 205 cm2) might be larger than the actual value. In this case, the equivalent number of pots will be larger than 1115.

6.2.1 Recovery Rate The recovery rate of the ceramic material is affected by many factors. Post-depositional factors, such as the collection of sherds for various purposes (as temper in pottery manufacture or as building material) or the effect of natural processes such as rain or wind action, can alter the original amount of the ceramic material deposited in one site. Recycling of damaged vessels or sherds can also affect greatly their quantity and appearance. In Nea Nikomedeia we know that spindle whorls and clay roundels were made from pottery fragments.

An estimation of the recovery rate of the ceramic material from Nea Nikomedeia was obtained from the W/R painted and the impressed vessels. Using the method described earlier, it was calculated that among the 1115 vessels, 6 impressed and 2 W/R pots are included. During the study of impressed pottery, sherds belonging to at least 36 different vessels have been distinguished. That is, the sherds could be separated into 36 groups which differed either in fabric type, or decorative motifs or in vessel form. Similarly the W/R sherds could be attributed to 9 vessels at least. This suggests that the rate of recovery of ceramic material at Nea Nikomedeia is around 20%.

The total surface area of the impressed and the W/R pottery might have been underestimated during Dr. Rodden's post-excavation work, since it is probable that some sherds from the undecorated parts of the vessels were not counted with the decorated pottery. Taking into consideration that only two thirds of the surface of impressed vessels was decorated, a value of 25-30% for the recovery rate of the ceramic material must be more representative. It can be estimated then that the number of vessels that entered the archaeological deposit of the main excavation grid at Nea Nikomedeia, ranges from 3700 to 4500.

6.2.2 Rate of Production In order to obtain a rough sense of the scale of local pottery production per year, an estimate of the time in which these deposits were formed is needed. Taking into consideration that three building phases were distinguished, the site should have been occupied for 50 years at least.

An upper limit for the duration of occupation is more difficult to estimate. The recorded ceramic material used in this study comes from the main excavation grid, where the thickness of deposits ranges from 20-65 cm. An indication of the length of occupation could be obtained by comparing the depth of these deposits with those of the contemporary Thessalian sites.

Although a direct relation between depth of stratigraphy and length of occupation does not exist, an indication for the rate of the accumulation of deposit can be obtained.

As can be seen from table 6.2, the occupation deposit accumulated during the Decorated Phase in Thessaly ranges from 1.60-4.20 m. On the basis of C14 dates, this period lasted for 250-300 radiocarbon years. Taking the lowest value (1.60 m), a period of 100 years would be suggested for the Nea Nikomedeia debris. Consequently, the occupation

THE EARLY NEOLITHIC POTTERY: FUNCTIONAL ANALYSIS 185

Table 6.2: Thickness of occupation deposits of the Thessalian sites with early decorated wares.4

Sites Thickness (mm) Duration (C14 years)

Otzaki 4.20 250 Achilleion 1.60 250 Prodromos 3.15 2 50 Gendiki 3-3° 25°

Magoulitsa 3.40 250

Agia Anna 3.10 250

deposit from the main excavated grid most probably had accumulated over a period of 50 to 150 years.

From the estimated number of years of occupation and the calculated number of pots, we arrive at an estimated annual production of 25-90 pots per year, for all the houses of the main excavation grid.

The annual production of vessels which has been calculated for an area covering o. 1 hectare, suggests that pots were produced frequently in Nea Nikomedeia, in contrast with Franchthi where a number of 1 2-1 3 pots per year was estimated for the whole site (an area of around 13,300 m2, including the cave and the settlement on the shore). The difference in pottery production between the two Early Neolithic sites could indicate a difference in the type of site occupation (ephemeral/seasonal/permanent site), different economic activities or a greater dependence, of the inhabitants of Franchthi, on containers made from perishable materials. The possibility that the recovery rate of the ceramic material was lower in Franchthi than in Nea Nikomedeia should also be taken into consideration.

The fact that ceramic containers played an important role in the activities of the Nea Nikomedeia inhabitants is also reflected by the variation in shape and size of the vessels. Pots range in shape from open dish-like bowls, to hole-mouthed jars. In size there is a continuous range from the miniature vessels, 4-6 cm in height, to the large neck jars which reach up to 60 cm in height. This suggests that vessels had a variety of functions. The functions of ceramic containers can be separated into three broad categories: processing of food, storage and transfer or transport (Rice 1987, 208).

4 Sources: Otzaki: Milojcié 1971; Achilleion: Gimbutas 1989; Prodromos: Chourmouziadis 1971; Gendiki:

Theocharis 1962; Magoulitsa: Papadopolou 1958; Agia Anna: Chourmouziadis 1969.

i86 PARASKEVI YIOUNI

6.3 FUNCTION OF VESSELS

6.3.1 Cooking Vessels Vessels used for cooking must have a good resistance to thermal shock in order to withstand repeated heating and cooling without fracturing (Rye 1981, 26). Two factors are usually controlled by the potters: the shape of vessels, and the types of fabric used for the manufacture of pots. As far as the vessel's shape is concerned, angular pots are usually not used for cooking, because they have many points of stress which might initiate fracture when exposed to sudden changes of temperature. Instead, the majority of cooking vessels have rounded contours (Woods 1986).

The composition of the fabric may also be relevant. The use of different clays and tempers for different vessel forms or function classes of pottery, is widely known from ethnographic research (Rice 1987, 226). This is also attested by analyses of experimental ceramics (Rye 1976, 1981; Bronitsky 1983; Steponaitis 1984). Pottery which contains inclusions in the pot fabric which have low coefficient of thermal expansion or one similar to that of the fired clay, should prove most resistant to the effects of thermal shock. Such inclusions are calcite, feldspars and grog (Rye 1981, 31-36). Furthermore, large amounts of non-plastic inclusions, preferably coarse, will increase the porosity of the fabric. Porous fabrics maximize thermal shock resistance, because they allow the pot to expand and contract during the heating and cooling phases.

The non-plastic inclusions could be either added by the potters, or they could be naturally included in the original clay which has been used for the manufacture of pots. As was discussed in Chapter 4 (section 4.7.4) the Nea Nikomedeia potters were most probably using material from the original clay beds after a refining process.

If the Nea Nikomedeia potters were aware of the effects of the pottery fabric's composition on the performance of vessels used for different tasks, a correlation between fabric types and vessel forms would be expected. In order to investigate this, the relative frequency of every fabric type, used in each vessel form, as well as the proportion of fine, medium and coarse textured paste encountered in each vessel form, have been calculated. They are presented in figs. 6.2 and 6.3.

From fig. 6.2, it can be seen that all fabrics were used for the manufacture of each vessel form. Moreover, the relative frequency of each fabric type is similar in all vessel forms. Only two slight variations are noticed. In neck jars the use of fabric Β is increased at the expense of fabric A. A slight increase in the use of fabric Β is also noticed in the open pots with rim angle less than 6o°. In the case of neck iars, the increase could be attributed to the rise of red-brown slipped pots (figs. 6.4 and 6.5). As was seen in Chapter 5 (sections 5.3. 1 and 5.4.8), there is a correlation between surface treatment and types of fabric.

fig. 6.3 shows that the relative frequency of coarse, medium and fine textured paste is the same in each vessel form.

Therefore, the Nea Nikomedeia pots do not show any correlation between types of fabric or quality of paste, and vessel form. However the above conclusion does not exclude the use of pots for cooking. Fabrics A, B and C, it seems, are well suited for the manufacture of cooking vessels. Fabrics A and C, which have been used for the manufacture of 60% and 12% of the ceramic material, are rich in calcite and limestone inclusions (Chapter 4, sections 4.6.2 and 4.6.5). Fabric Β (used for 18% of the vessels) contains abundant feldspar and volcanic inclusions (sections 4.6.3 and 4.6.4). Moreover, all types of fabric are rather coarse.

THE EARLY NEOLITHIC POTTERY: FUNCTIONAL ANALYSIS 187

fig. 6.2: Correlation of fabric type and vessel form.

i88 PARASKEVI YIOUNI

fig. 6.3: Correlation of quality of texture and vessel form.

The Nea Nikomedeia pots have rounded contours, a design which tends to eliminate points of high stress. Round bases are not a common feature, amounting to 8% of the base sherds. Flat based vessels were more common, amounting to 25% of the sample. Although round bottomed bases might have a better thermal shock resistance, the archaeological record shows that flat based vessels have also been used for cooking (Woods 1986).

Ethnoarchaeological research shows that cooking vessels can have a variety of shapes, depending on the type of cooking (Iinton 1944; Henrickson and McDonald 1983; Ericson et al 1972; Smith 1985). For boiling, slightly open or slightly closed vessels are regularly used. These forms are common in Nea Nikomedeia (fig. 5.62).

Another characteristic which can be used for the identification of cooking pots, is the detection of sooting and burning on the exterior surface of the vessels. The location of soot can indicate how the vessel has been used (Rice 1987, 235). If the soot occurs primarily on the sides of a vessel, from the base up to or near the maximum diameter, the vessel was probably set in the fire. Vessels with soot on the base and sides were most probably suspended over the fire.

THE EARLY NEOLITHIC POTTERY: FUNCTIONAL ANALYSIS 189

fig. 6.4: Correlation of surface treatment and vessel form.

igo PARASKEVI YIOUNI

fig. 6.5: Correlation of surface treatment and vessel form.

Among the whole pots found in Nea Nikomedeia, one askoid vessel has the larger part of the exterior surface covered with soot, apart from a lighter area near the base. This lighter area may be due to the exposure in hot oxidizing flames. Among the base sherds, there are examples with black patches on the exterior surface. Since most of the sherds are rather small, it is difficult to distinguish whether these black patches are the result of repeated exposure to fire, or are due to the original firing of the pots. It is worth mentioning, however, that black patches were noticed in 31% of the beige uncoated bases, in contrast to 12% of the red-brown slipped and 7% of the pink coated bases.

From the above discussion it can be concluded that definite proof of the existence of cooking pots made from a particular fabric type and/or with a specific shape, is not provided by the Nea Nikomedeia material. However such a differentiation might require a long tradition in pottery manufacture and pottery use. The site of Nea Nikomedeia is dated at the early stages of ceramic production in the Balkan area. Moreover, it seems that the majority of fabrics used by the Nea Nikomedeia potters were able to withstand the firing stress, and a whole vessel with sooting marks has been found.

THE EARLY NEOLITHIC POTTERY: FUNCTIONAL ANALYSIS 191

The above discussion is based on the assumption that the inhabitants of Nea Nikomedeia were using clay cooking pots (boiling vessels in particular). From the study of faunal and floral remains it is known that meat and perhaps other animal products, such as milk, were consumed. Cultivated grains and legumes and probably a wide variety of wild plants, fruits and nuts played an important role in their diet. It is not known though, whether boiling, roasting or baking was a preferred form of cooking. An indirect indication for the use of boiling containers, of some sort, is offered by the botanical sample.

Bitter vetch seeds have been found in Nea Nikomedeia and other Neolithic sites, together with human food remains (van Zeist and Bottema 197 1). Thus, although today bitter vetch is used for animal fodder, it seems that in the Neolithic period man himself consumed these seeds. Bitter vetch is poisonous for man and some animals, but the poisonous substance can be removed by boiling the seeds and pouring off the water (van Zeist and Bottema

Wild bitter vetch was also present in Mesolithic sites.5 Whether the Neolithic inhabitants had adapted their Mesolithic cooking traditions to the new technology (ceramic containers), is difficult to say. As Vitelli argues, taste is something that changes very slowly (Vitelli 1989). As far as Nea Nikomedeia is concerned, the ceramic material offers indications of the use of vessels near the fire but it is difficult to conclude how widespread this practice was.

6.3.2 Storage Vessels Ceramic containers, if used for storing foodstuffs, offer better protection against rodents, than do containers made from perishable material (baskets, leather or wooden containers). The safe storage of the next season's seed grain in particular, would have been essential to the early farmers. In Nea Nikomedeia, a number of large vessels have been found, which most probably were used for the long term storage of foodstuffs.

The largest vessels belong to the hole-mouthed and neck jar categories. They are deep vessels with rim diameters reaching up to 32 cm (fig. 5.26). Their height ranges between 40-60 cm. Moreover large vessels belonging to the slightly closed and slightly open bowls as well as the medium sized neck and hole-mouthed jars, could also have been used for storage. Large sized vessels have also been found in other Early Neolithic sites.6

Halstead, reviewing the faunal and floral remains from Greek Neolithic sites, concludes that their inhabitants were largely dependant, for their diet, on grains and pulses (Halstead 1981). He estimates that the combined annual consumption of cereals and pulses may have been something like 200 kg per head. Using NarroPs (1962) formula:

number of persons = total floor space of all houses in m2 / 10 it can be estimated that 30-50 people were living in the houses revealed in the main excavation grid. For the storage of their annual crop production, a storage volume of 7500- 12500 litres would have been needed. A conversion rate of 1.25 litres per 1 kg of pulses and grain seeds has been used. This rate has been estimated by measuring the volume

5 e.g. Franchthi: Hansen and Renfrew 1978.

6 Achilleion: Gimbutas 1989; Franchthi: Vitelli 1989

i92 PARASKEVI YIOUNI

occupied by ι kg of barley, wheat, peas and lentils (see table 6.3). In order to calculate the average volume of the Nea Nikomedeia vessels the same method used for the calculation of the average surface area has been employed. The volume of each conical segment was calculated from the following expression:

V = ^(Ra8+R*Rb+Rbi)*h/3 where Ra, B^ are the radii, and h is the height (1 cm) of the conical segment. The average volume was found to be 6107 cm5, suggesting a total volume of 22600-27500 litres.

Only a part of the calculated volume however, would have been employed for storage. Furthermore the figure of 22600-27500 litres has been estimated from all the vessels used over a period of 50-150 years. This means that only a small fraction of the total volume was available at a time. Thus, it is obvious that the storage capacity of the Nea Nikomedeia pots was not enough for the annual crop production. It seems then that containers made from perishable materials and/or storage pits were also used for storage purposes.

For the storage of seeds kept for sowing the next year, a much smaller volume capacity is needed. According to Halstead, 10% or even less of the annual crop production was kept for sowing (Halstead 1981, 317). This would correspond to a volume of 750-1250 litres. The large vessels of Nea Nikomedeia range in volume from 36-85 litres. Only 15 to 20 such vessels would have been enough for the inhabitants of the main grid to store the grain kept for sowing. According to ethnographic research, large vessels have a long use life, since they are rarely moved (Foster i960; DeBoer 1985; Kramer 1979). Consequently the storage of seeds kept for sowing was within the capacity of the ceramic production at Nea Nikomedeia, which amounts to 25-90 vessels per year.

Table 6.3: Measurement of the weight of 1 litre of cereals and pulses (nb. the sample comprises cereals and pulses bought at a Greek market).

Volume (litres) Weight (grams)

wheat 1 850

barley 1 750

peas 1 710

lentils 1 890

ACKNOWLEDGEMENTS

I would like to thank the excavator of Nea Nikomedeia, Dr. R.J. Rodden, for the opportunity to study the ceramic material from the site. I am also grateful to Mrs. H. C. Ridley, Dr. K. A. Wardle, Dr. K. Gallis, Mr. G. Touphexis, Dr. P. Chrisostomou, Mr. Pan. Chrisostomou, the late Professor M. Gimbutas, Mrs D. Simoska, Mr. V. Sanev, Mr. T. Nacev and Mr. Z. Beldedovski for permitting me to see the ceramic material from a number of Greek and

THE EARLY NEOLITHIC POTTERY: FUNCTIONAL ANALYSIS 193

Yugoslavian sites. I wish to express particular thanks to the former Ephor of Antiquities for Western Macedonia, the late Dr. M. Siganidou, to the archaeologists of Verroia Museum, Mrs. K. Tzanavari and Mrs. M. Apostolou and the staff of the Museum, for their hospitality.

I would like to thank Dr. J. G. Nandris for his helpful critical appraisal whilst preparing this work, Mr. C. Orton for his advice in statistical techniques, Mrs. C. Cartwright for her help in pétrographie analysis and Dr. D. Griffiths for many constructive discussions and for operating the SEM on my behalf. I am grateful to Dr. I. C. Freestone for his helpful suggestions on the analysis of the pétrographie data. Special thanks are deserved by my parents, Barbara and Dimitri Yiouni, and my husband Euripides Glava, for their constant support throughout the years of my studies.

Finally, I wish to acknowledge the financial assistance of the A. S. Onassis Public Benefit Foundation and the Central Research Fund of the University of London.

P. Y.

Appendix A

The Radiocarbon Dating of Nea Nikomedeia

Table Showing the laboratory information, published dates and dates bc of Nea Nikomedeia.

Year Lab. No. Material Published Date Date bc

1962 0:^55 Charcoal 8180 ± 150 bp 6230 Unknown* GX-679 No Information 7780 ± 270 bp 5830 1967 P-1202 Charcoal 7557 ± 91 bp 5^°7 1967 P-1203A Charcoal 7281 ± 740p 5331 1988 OxA-1603 Triticum monococcum 7050 ± 80 bp 5100 1988 OxA-1604 Triticum dicoccum 7340 ± 90 bp 539° 1988 OxA-1605 Hordeum vulgäre 7400 ± 90 bp 545° 1988 OxA-1606 Lens culinaris 7400 ± 100 bp 545° 1993 OxA-3873 Bone (Ovis) 7300 ± 80 bp 5350 1993 OxA-3874 Bone (Capra) 737° i 80 bp 5420 1993 OxA-3875 Bone (Sus) 728o±9obp 533° 1993 OxA-3876 Bone (Bos) 7370 ± 90 bp 5419 1993*· OxA-4280 Triticum monococcum 6920 ± 120 bp 49^9 1993·· OxA-4281 Triticum dicoccum 7100 ± 90 bp 5149 1993*· OxA-4282 Hordeum vulgäre 7400 ± 90 bp 5449 1993·· OxA-4283 Lens culinaris 7260 ± 90 bp 53 10

• No information is available concerning this date. ·· These dates were obtained from humic acid as a check for contamination.

*95

196 APPENDIX A

Graph showing the uncalibrated radiocarbon dates of Nea Nikomedeia.

4800-

4900-

5000-

5100-

5200-

5300-

5400-

55°°-

PQ 5600- <

I 5700-

5800-

59°o-

6000-

6100-

6200-

6300-

0400 ι τ- ι ι ι Ι Ι Τ Ι ι ι ι Τ I I I if)O>e*> oorfifito οα^ΐΟΦ Ô rH et 00 iDt^o^rsO Ο Ο Ο r*r^t^t^-oooocx)oo

<ΟζΟ e* oCOtO^O^OoocXîOOOO e* c< ©# c*

OOOOOOOOOOOO Laboratory Number

Περίληψη

1. ΠΡΟΛΟΓΟΣ

Τα αρχιτεκτονικά κατάλοιπα που αποκαλύφθηκαν κατά τη διάρκεια των ανασκαφών στη Νέα Νικομήδεια χρονολογούνται στην Αρχαιότερη Νεολιθική περίοδο. Δύο παλιές ραδιοχρονολογήσεις της θέσης κυμαίνονται από το 5607-5331 be (Appendix Α). Πρόσφατες αναλύσεις έδωσαν μετρήσεις που κυμαίνονται από το 5450-5100 be. Μία αρκετά νεότερη χρονολόγηση (4969 be) ίσως να ανήκει σε δείγμα που προέρχεται από τα τελευταία στάδια της Αρχαιότερης Νεολιθικής κατοίκησης. Στην Ύστερη Νεολιθική περίοδο χρονολογούνται, με βάση τα ευρήματα που περιέχουν, οι τάφροι που τέμνουν τις επιχώσεις της Αρχαιότερης Νεολιθικής (FIG. 3.3).

Ο παρών τόμος περιλαμβάνει την εισαγωγή του ανασκαφέα της θέσης Dr. R. J. Rodden, o οποίος εκθέτει το ιστορικό, τους στόχους και την τεχνική των ανασκαφών. Στη συνέχεια παρουσιάζονται λεπτομερώς η στρωματογραφία και τα κτίσματα της Αρχαιότερης Νεολιθικής Περιόδου (Gillian Pyke). Ακολουθεί η αναλυτική μελέτη της Αρχαιότερης Νεολιθικής κεραμεικής η οποία περιλαμβάνει την τεχνολογία, το σχηματολόγιο, τη διακόσμηση και τη χρήση των αγγείων (Παρασκευή Γιούνη). Ο δεύτερος τόμος θα περιέχει μελέτες των πελεκητών εργαλείων (Αντίκλεια Μουνδρέα- Αγραφιώτη), των οστέινων εργαλείων (Rosemary Payne) και των ειδωλίων (John Nandris). O πλήρης κατάλογος των υπόλοιπων ευρημάτων θα συνοδεύεται από μελέτη της κατανομής τους στο χώρο (Gillian Pyke). Αναφορά στη συμβολή της Νέας Νικομήδειας στην ανάπτυξη των αγροτικών κοινωνιών της ΝΑ Ευρώπης (Παρασκευή Γιούνη), θα συμπληρώσει τον τόμο.

2. ΕΙΣΑΓΩΓΗ Η ανασκαφική έρευνα στην Νέα Νικομήδεια ξεκίνησε το 1961 και τελείωσε το 1964. Κατά την πρώτη ανασκαφική περίοδο (1961) χρησιμοποιήθηκε κάνναβος με διαστάσεις 2 x 2 μ. Στις επόμενες ανασκαφικές περιόδους (1963, 1964) τα τετράγωνα του καννάβου είχαν διαστάσεις 4 x 4 μ. Η ανασκαφή προχωρούσε με την αφαίρεση οριζόντιων πασών (spits). Η παρατήρηση των ανασκαφικών σκαμμάτων από σκάλα ύψους 6-7 μ, επέτρεπε την ευκρινέστερη αναγνώριση των αρχιτεκτονικών καταλοίπων (PLATE 3).

3. ΣΤΡΩΜΑΤΟΓΡΑΦΙΑ

Στις προκαταρκτικές δημοσιεύσεις ο ανασκαφέας, Dr. Rodden, είχε προτείνει τη διάκριση δύο οικιστικών περιόδων στην Αρχαιότερη Νεολιθική, με δύο φάσεις ανακατασκευών η κάθε μία Η επανεξέταση των δεδομένων έδειξε ότι η άποψη αυτή θα πρέπει να αναθεωρηθεί: η παρούσα εργασία προτείνει τη διάκριση τριών οικιστικών φάσεων. Σε ορισμένα σημεία του οικισμού αποκαλύφθηκαν αρχιτεκτονικά κατάλοιπα τα οποία πιθανότατα να ανήκουν σε μία τέταρτη οικιστική φάση.

197

1 98 ΠΑΡΑΣΚΕΥΗ ΓΙΟΥΝΗ

ΣΤΡΩΜΑΤΟΓΡΑΦΙΚΗ ΔΙΑΔΟΧΗ ΟΙΚΗΜΑΤΩΝ

Η μελέτη των κατόψεων των οικημάτων δείχνει ότι μπορούν να διακριθούν ομάδες οικημάτων (groups) που χτίζονται και ανοικοδομούνται στην ίδια θέση.

Ομάδα 1: Η ομάδα αυτή περιλαμβάνει τρία διαδοχικά οικήματα (FIG. 2.7). Το αρχαιότερο οίκημα (οίκημα 1) εμφανίζεται στον πυθμένα της πάσας 2. Το επόμενο οίκημα, κόβεται από την τάφρο θεμελίωσης του οικήματος 3.

Ομάδα 2: Στις προκαταρκτικές δημοσιεύσεις είχαν διακριθεί μόνον 2 οικήματα: το οίκημα 3 είχε χρονολογηθεί στην Ύστερη Νεολιθική κατοίκηση της θέσης, ενώ τα υπόλοιπα αρχιτεκτονικά λείψανα είχαν αποδοθεί σε ένα οίκημα το οποίο είχε ένα κεντρικό χώρο και δύο μικρότερα ορθογώνια δωμάτια (FIG. 21). Η επανεξέταση των κατόψεων οδήγησε στην διάκριση τριών οικημάτων. Τα οικήματα 1 και 2 έχουν παρόμοιο σχήμα και διαστάσεις. Οι τάφροι θεμελίωσης του βόρειου και νότιου τοίχου του οικήματος 2 περνούν πάνω από το οίκημα 1. Το τελευταίο οίκημα, οίκημα 3, βρέθηκε πιθανότατα στην πάσα 1.

Ομάδα 3: Η ομάδα περιλαμβάνει δύο οικήματα (FIG. 2.9). Η τάφρος θεμελίωσης του οικήματος 1, που καλύπτεται από το οίκημα 2, εμφανίστηκε στην πάσα 3. Πιθανόν ο τοίχος Α που αποκαλύφθηκε στα τετράγωνα C1-2 να αποτελεί τον ανατολικό τοίχο του οικήματος 1 (βρέθηκαν στο ίδιο βάθος). Το πρόβλημα με την ερμηνεία αυτή είναι ότι ο τοίχος έχει πολύ μεγάλες διαστάσεις. Το οίκημα 2 εμφανίζεται κυρίως στην πάσα 2, αν και μέρος του ανατολικού τμήματος βρέθηκε στην πάσα 1.

Ομάδα 4: Το κύριο πρόβλημα στην ομάδα αυτή είναι ότι οι τάφροι θεμελίωσης των οικημάτων 1 και 2 βρέθηκαν στην ίδια πάσα (spit 2) καθιστώντας έτσι δύσκολη τη διάκριση των οικιστικών φάσεων (FIG. 2.10). Το οίκημα 1 είναι ένα μεγάλο τετράγωνο κτίριο που καλύπτεται μερικώς από το οίκημα 2. Το οίκημα 2 χωρίζεται, με εγκάρσιους τοίχους, σε τρεις εσωτερικούς χώρους. Από το οίκημα 3, που καλύπτει και τα δύο άλλα οικήματα, διατηρείται μόνον ένα μικρό τμήμα Στα τετράγωνα D4-5 αποκαλύφθηκε τμήμα κτιρίου το οποίο πιθανόν είναι σύγχρονο με το οίκημα 3 ή αντιπροσωπεύει μία τέταρτη φάση.

Ομάδα 5: Αποτελείται από δύο οικήματα που χωρίζονται σε δύο εσωτερικούς χώρους (FIG. 2.11). Το οίκημα 2 καλύπτει το οίκημα 1.

Ομάδα 6: Αποτελείται από τρία οικήματα (FIG. 2.12). Το οίκημα 2, που κόβει το οίκημα 1, έχει δύο εσωτερικούς χώρους. Το οίκημα 3 καλύπτει και τα δύο άλλα οικήματα Στα τετράγωνα L4-5 και Μ4 διακρίνονται ίχνη που πιθανόν να ανήκουν σε μία τέταρτη κατασκευή (FIG. 24), αλλά τα στοιχεία δεν είναι αρκετά για την εξαγωγή σίγουρων συμπερασμάτων.

Ομάδα 7: Αποτελείται από τρία οικήματα που βρέθηκαν στην ίδια πάσα (spit I). To οίκημα 1, το οποίο έχει δύο εσωτερικούς χώρους, κόβεται από το οίκημα 2 και το οίκημα 3 καλύπτει το οίκημα 2 στο τετράγωνο ΤΥ5 (FIG. 2.13). Μία τάφρος θεμελίωσης που βρέθηκε στα τετράγωνα ΤΥ5-6 και ΤΧ6-7 πιθανόν να ανήκει σε ένα τέταρτο οικοδόμημα

Ομάδα 8: Αποτελείται από τρία οικήματα (FIG. 214). Στο τετράγωνο ΤΥ5 διακρίνεται μία γωνία που πιθανόν να ανήκει σε ένα τέταρτο κτίριο. Από το αρχαιότερο οίκημα 1 έχει αποκαλυφθεί μόνον ένα μικρό τμήμα Το οίκημα 3 θεωρείται το νεότερο της ομάδας 8, γιατί εμφανίστηκε στην πάσα 1.

Ομάδα 9: Το μοναδικό οίκημα της ομάδος έχει δύο εγκάρσιους τοίχους που σχηματίζουν ένα διάδρομο γύρω από τις δύο πλευρές του. Στην ίδια περιοχή έχουν αποκαλυφθεί και τρεις τάφροι της Αρχαιότερης Νεολιθικής, οι οποίοι κόβονται από το οίκημα 1 (FIG. 215).

ΠΕΡΙΛΗΨΗ 199

ΟΡΙΖΟΝΤΙΑ ΑΝΑΠΤΥΞΗ ΤΩΝ ΟΙΚΟΔΟΜΙΚΩΝ ΦΑΣΕΩΝ

Τα περιορισμένα στοιχεία που ήταν διαθέσιμα επέτρεψαν μόνον ένα γενικό συσχετισμό των οικιστικών φάσεων. Τα οικήματα της αρχαιότερης φάσης (FIG. 217) εντοπίστηκαν στην πάσα 3 και στον πυθμένα της πάσας 2. Η δεύτερη φάση (FIG. 2.18) περιλαμβάνει τα οικήματα που εμφανίστηκαν στο άνω τμήμα της πάσας 2 και στον πυθμένα της πάσας 1. Τα οικήματα της τρίτης φάσης και τα ίχνη μίας πιθανόν τέταρτης εμφανίζονται αποκλειστικά στην πάσα 1 (FIG. 219).

Οι πάσες (spits 1 & 2) στις οποίες εντοπίστηκαν οι ομάδες 1 και 2 έχουν ανασκαφτεί περίπου στο ίδιο επίπεδο και έχουν σχεδόν το ίδιο βάθος. Ετσι τα τρία διαδοχικά οικήματα των ομάδων αυτών μπορούν να αποδοθούν στις τρεις διαδοχικές οικιστικές φάσεις. Το οίκημα 1 της ομάδας 3 ανήκει στην πρώτη φάση ενώ το οίκημα 2 στην δεύτερη φάση.

Καθώς δύο οικήματα της ομάδας 4 βρέθηκαν στην ίδια πάσα (spit 2) είναι δύσκολο να βρεθεί η ακριβής αντιστοιχία τους με τα οικήματα των ομάδων 1-3. Παρ' όλα αυτά, βασιζόμενοι στον συσχετισμό του βάθους της πάσας 2 και του βάθους στο οποίο βρέθηκαν τα οικήματα 1 και 2 των άλλων ομάδων, μπορούμε να αποδώσουμε τα οικήματα 1 και 2 στην πρώτη και δεύτερη οικιστική φάση, αντίστοιχα Με τον ίδιο τρόπο το οίκημα 1 που βρέθηκε στην πάσα 1, αποδίδεται στην τρίτη οικιστική φάση.

Τα οικήματα των υπολοίπων ομάδων (εκτός από την ομάδα 9) έχουν αποκαλυφθεί σε περιοχές όπου μόνον μία πάσα έχει ανασκαφτεί. Για τα οικήματα της ομάδας 5 μπορούμε να παρατηρήσουμε ότι η επιφάνεια της πάσας 1 στο τετράγωνο G6-8 αντιστοιχεί με την επιφάνεια της πάσας 2 των προηγούμενων ομάδων, ενώ η επιφάνεια της πάσας 1 στα τετράγωνα Η6-8 αντιστοιχεί με αυτήν της πάσας 1. Φαίνεται λοιπόν πιθανόν ότι τα οικήματα της ομάδας 5 μπορούν να αποδοθούν στην πρώτη και δεύτερη οικιστική φάση.

Οσον αφορά την ομάδα 6 παρά τις περιορισμένες πληροφορίες που υπάρχουν για ορισμένα από τα τετράγωνα της, οι συσχετισμοί του βάθους της πάσας 1 επιτρέπουν να αποδοθούν τα τρία οικήματα της ομάδας αυτής στις ίδιες φάσεις που έχουν αναγνωριστεί για τις ομάδες 1- 4. Με τον ίδιο τρόπο, τα τρία οικήματα των ομάδων 7 και 8 μπορούν να αποδοθούν στις τρεις διαδοχικές φάσεις Η ομάδα 9 είναι δύσκολο να συσχετιστεί με τις τρεις οικοδομικές φάσεις της θέσης. Για το λόγο αυτό η απόδοση, στην τρίτη φάση, του μοναδικού οικήματος της ομάδας αυτής γίνεται με μεγάλη επιφύλαξη.

4. ΟΙΚΙΣΤΙΚΗ ΚΑΙ ΑΡΧΙΤΕΚΤΟΝΙΚΗ

ΚΑΤΑΣΚΕΥΗ ΤΩΝ ΟΙΚΗΜΑΤΩΝ

Συνολικά αποκαλύφθηκαν 24 πασσαλόπηκτα κτίσματα τα οποία εκτός από μία εξαίρεση (ομάδα 8/3, πιθανόν τριγωνικό) έχουν τετράγωνη ή ορθογώνια κάτοψη. Οι τεχνικές κατασκευής των οικημάτων καθώς και η επιλογή των πρώτων υλών παρέμειναν οι ίδιες καθ' όλη τη διάρκεια της κατοίκησης του οικισμού. Η τάφρος θεμελίωσης των τοίχων σκάβεται σε αρκετό βάθος (34-50 εκ.) και συνήθως φέρει επάλειψη με πηλό. Πιο σπάνια, γεμίζουν την τάφρο με χώμα Οι πάσσαλοι, διαμέτρου 8-20 εκ., τοποθετούνται στη σειρά κατά τον άξονα σε απόσταση 1-1,5 μ. ο ένας από τον άλλο. Ορισμένες φορές οι πάσσαλοι τοποθετούνται κατευθείαν στο έδαφος (χωρίς τάφρο θεμελίωσης: ομάδα 2/3). Οι πασσαλότρυπες έφεραν συνήθως επάλειψη με πηλό. Εσωτερικές αντιρήδες χρησιμοποιούνται σπάνια (μακρύς τοίχος ομάδας 3/2). Ανάμεσα στους πασσάλους τοποθετείται πλέγμα κλαδιών και στην συνέχεια ο σκελετός του κτιρίου καλύπτεται

Qoo ΠΑΡΑΣΚΕΥΗ ΓΙΟΥΝΗ

εσωτερικά και εξωτερικά με πηλό ο οποίος κάποιες φορές είναι αναμειγμένος με άχυρα Τα δάπεδα, τμήματα των οποίων διατηρήθηκαν σε λίγα σημεία του οικισμού (βλ. τετράγωνα Α5/1, Α6/2), κατασκευάζονται από πατημένο πηλό. Σε μία περίπτωση υπήρχε και μία κατώτερη στρώση από χαλίκια (τετράγωνο Β4/1).

ΜΕΓΕΘΟΣ ΚΑΙ ΣΧΗΜΑ ΤΩΝ ΟΙΚΗΜΑΤΩΝ Σε όλες τις οικιστικές φάσεις τα οικήματα έχουν κοινό προσανατολισμό από τα ανατολικά στα δυτικά. Οι τρεις οικιστικές φάσεις διαδέχτηκαν πολύ γρήγορα η μία την άλλη. Στο συμπέρασμα αυτό οδηγεί η απουσία νεκρού στρώματος εγκατάλειψης ανάμεσα στα αρχιτεκτονικά λείψανα των διαδοχικών φάσεων. Το γεγονός ότι υπάρχουν ομάδες οικημάτων που χτίζονται και ανοικοδομούνται στον ίδιο χώρο υποδηλώνει ότι υπήρχε κάποια οργάνωση του χώρου.

Πρώτη φάση: Τα οικήματα είναι μονόχωρα εκτός από δύο ορθογώνια κτίρια τα οποία χωρίζονται εσωτερικά σε δύο τμήματα: στην ομάδα 5/1 ένας εγκάρσιος τοίχος στα ανατολικά μίας πιθανής εισόδου χωρίζει το οίκημα σε ένα τετράγωνο χώρο στα ανατολικά και σε ένα ορθογώνιο στα δυτικά Το άλλο οίκημα (ομάδα 7/1) έχει ένα στενό ορθογώνιο δωμάτιο στα νότια και ένα αρκετά μεγαλύτερο στα βόρεια Επτά από τα οκτώ κτίρια της φάσης 1 έχουν παρόμοιες διαστάσεις (TABLE 3.1). Εξαίρεση αποτελεί το οίκημα στην ομάδα 4/1 το οποίο είναι εξαιρετικά μεγάλο. Η διαφορά στο μέγεθος δηλώνει πιθανότατα μια διαφορετική χρήση του κτιρίου

Δεύτερη φάση: Κτίρια με εσωτερικές διαιρέσεις συναντώνται και σε αυτή τη φάση: κτίρια που έχουν δύο ισομεγέθεις χώρους καθώς και κτίρια που διαιρούνται σε ένα μικρό τετράγωνο δωμάτιο και σε ένα άλλο αρκετά μεγαλύτερο. Οι διαστάσεις τους είναι παρόμοιες με αυτές των οικημάτων της φάσης 1 (TABLE 3.2). Σύμφωνα με τον Dr. Rodden το μεγάλο κεντρικό κτίριο (ομάδα 4/2) χωρίζεται σε δύο ανισομεγέθεις χώρους. Τα αρχιτεκτονικά λείψανα όμως μπορούν να ερμηνευθούν και ως κατάλοιπα δύο οικημάτων: ενός μεγάλου τετράγωνου κτιρίου και ενός μεταγενέστερου, μικρότερου, ορθογώνιου οικήματος.

Τρίτη φάση: Μόνον έξι οικήματα αποδίδονται στην φάση αυτή (TABLE 3.3). Τα κτίρια των ομάδων 1, 2 και πιθανότατα της ομάδας 8 παρουσιάζουν την ακόλουθη διαφοροποίηση: οι τρεις τοίχοι των οικοδομημάτων θεμελιώνονται σε μία τάφρο σκαμμένη στο έδαφος, ενώ οι πάσσαλοι του τέταρτου τοίχου τοποθετούνται κατευθείαν στο έδαφος.

ΜΕΓΕΘΟΣ ΚΑΙ ΠΛΗΘΥΣΜΟΣ ΤΟΥ ΟΙΚΙΣΜΟΥ Το τμήμα που έχει αποκαλυφτεί ανασκαφικά καλύπτει περίπου 8% της συνολικής έκτασης της τούμπας της Νέας Νικομήδειας. Εάν υποθέσουμε ότι η διάταξη, το μέγεθος και η χρονολόγηση των κτιρίων στο τμήμα του οικισμού που δεν έχει ανασκαφτεί είναι παρόμοια με αυτήν του ανασκαμμένου, μπορούμε να υπολογίσουμε ότι στην πρώτη και δεύτερη φάση θα υπήρχαν από 100 κτίρια και στην τρίτη φάση, 75. Για το χονδρικό υπολογισμό του αριθμού των κατοίκων του οικισμού υπολογίζεται η συνολική έκταση των κτιρίων του οικισμού η οποία διαιρείται με την έκταση που θεωρείται απαραίτητη για τη διαβίωση ενός κατοίκου ενός οικήματος (10 m^). Με τον τρόπο αυτό υπολογίζεται ότι ο συνολικός πληθυσμός στην πρώτη και δεύτερη φάση θα ήταν 500-700 κάτοικοι ενώ στη τρίτη φάση μειώνεται στους 300- 375 κατοίκους.

ΠΕΡΙΛΗΨΗ 2Ο1

ΛΑΚΚΟΙ

Οι λάκκοι βρίσκονται διασκορπισμένοι σε όλη την έκταση του οικισμού, χωρίς να ακολουθούν κάποια συγκεκριμένη διάταξη. Δεν παρατηρούνται συγκεντρώσεις λάκκων που θα μπορούσαν να συσχετιστούν με κάποιο οίκημα Στην προκαταρκτική μελέτη του υλικού της ανασκαφικής περιόδου 1961 ο Rodden προτείνει ότι οι τρεις μεγαλύτεροι λάκκοι, οι οποίοι είναι σχετικά ρηχοί και έχουν επίπεδο πυθμένα, δημιουργήθηκαν από την εξόρυξη πηλού για το χτίσιμο των οικημάτων και στην συνέχεια χρησιμοποιήθηκαν για αποθήκευση. Οι λάκκοι απορριμμάτων, οι οποίοι έχουν ακανόνιστο σχήμα, αρκετό βάθος και καμπύλο πυθμένα, περιείχαν κόκαλα ζώων, στάχτες και κάρβουνα Αντίθετα, οι αποθηκευτικοί λάκκοι ήταν σχετικά καθαροί Οι περιορισμένες πληροφορίες που υπήρχαν στα ανασκαφικά ημερολόγια δεν επέτρεψαν την αναλυτική μελέτη του υλικού που αποκαλύφθηκε στις επόμενες ανασκαφικές περιόδους (1963 και 1964).

ΦΟΥΡΝΟΙ ΚΑΙ ΕΣΤΙΕΣ

Στην προκαταρκτική μελέτη του υλικού ο Dr. Rodden περιγράφει δύο φούρνους κυλινδρικού σχήματος που ήταν ανοιχτοί στο επάνω τμήμα (FIG. 3.2). Είχαν χτιστεί με πηλό που περιείχε φυτικές προσμείξεις Ο ανασκαφέας περιγράφει επίσης μία εστία που αποκαλύφθηκε κατά την ανασκαφική περίοδο 1963. Ενδεικτικά μπορούμε να αναφέρουμε ότι στο τετράγωνο Β8/1, στην γωνία της ομάδας 2/2, αποκαλύφθηκε μία κατασκευή η οποία πιθανότατα ήταν φούρνος: πρόκειται για ένα λάκκο οποίος έφερε επάλειψη από καμένο καφέ υλικό, πιθανότατα πηλό. Ενας παρόμοιος '"φούρνος" βρέθηκε στο ίδιο τετράγωνο ανάμεσα στις ομάδες 2/2 και 1/2.

5. ΤΕΧΝΟΛΟΓΙΑ ΤΩΝ ΑΓΓΕΙΩΝ ΤΗΣ ΝΕΑΣ ΝΙΚΟΜΗΔΕΙΑΣ

Η μελέτη βασίζεται σε μακροσκοπική εξέταση του υλικού, μικροσκοπική εξέταση με πολωτικό μικροσκόπιο και σαρωτικό μικροσκόπιο (SEM), και μία σειρά από πειραματικά ψησίματα σε ελεγχόμενη ατμόσφαιρα και θερμοκρασία (ref iring test).

ΚΑΤΑΣΚΕΥΗ ΤΩΝ ΑΓΓΕΙΩΝ Η μακροσκοπική μελέτη του υλικού έδειξε ότι για το πλάσιμο των αγγείων χρησιμοποιήθηκαν η "τσιμπητή τεχνική" (pinching: PLATE 8 a) και η "τεχνική της κουλούρας" (coiling). Πολλές φορές γίνεται ένας συνδυασμός των δύο τεχνικών: το άνω τμήμα του αγγείου χτίζεται με την "τεχνική της κουλούρας", ενώ το κάτω τμήμα με την "τσιμπητή τεχνική". Τα περισσότερα αγγεία έχουν δισκοειδείς η δακτυλιοειδείς βάσεις οι οποίες κατασκευάζονται χωριστά και μετά τοποθετούνται στον πυθμένα του αγγείου (PLATE 8 b). Οι λαβές έχουν κολληθεί στην επιφάνεια των αγγείων, εκτός από ορισμένες περιπτώσεις που τα άκρα τους εισχωρούν στα τοιχώματα των αγγείων (PLATE 9 α). Τα αποτυπώματα ψάθας που υπάρχουν σε κάποια από τα αγγεία οφείλονται σε ψάθες που χρησιμοποιήθηκαν είτε για να στεγνώνουν τα αγγεία, είτε για να διευκολύνουν την περιστροφή των αγγείων κατά το χτίσιμο/τελείωμα τους.

2O2 ΠΑΡΑΣΚΕΥΗ ΓΙΟΥΝΗ

ΕΠΕΞΕΡΓΑΣΙΑ ΤΗΣ ΕΠΙΦΑΝΕΙΑΣ

ΚΟΚΚΙΝΟ-ΚΑΦΕ ΕΠΙΧΡΙΣΜΑ Οι διαφορές χρώματος που παρατηρούνται ανάμεσα στα αγγεία (κόκκινο, κόκκινο-καφέ, καφέ: TABLE 4.4) οφείλονται στην διαφορετική ατμόσφαιρα που επικρατεί από ψήσιμο σε ψήσιμο, και όχι στην χρησιμοποίηση επιχρισμάτων με διαφορετική σύσταση (TABLE 4.2). Επιπλέον η ανάλυση των δειγμάτων με το SEM έδειξε ότι το επίχρισμα παρασκευάζονταν από τους ίδιους πηλούς (πολύ πιο λεπτόκοκκους βέβαια) που είχαν χρησιμοποιηθεί για το χτίσιμο των αγγείων (TABLE 4.2), χωρίς την προσθήκη κάποιας χρωστικής ουσίας. Ορισμένες φορές οι αγγειοπλάστες πρόσθεταν στο επίχρισμα σπασμένη μαρμαρυγία, μία πρακτική που έδινε ένα έντονα διακοσμητικό αποτέλεσμα

ΡΟΖ ΕΠΙΣΤΡΩΜΑ Ενώ το κόκκινο-καφέ επίχρισμα είναι ένα αιώρημα μορίων λεπτόκοκκου πηλού σε νερό (PLATE 9 b' το ροζ επίστρωμα αποτελείται από κομμάτια θρυμματισμένης χρωστικής ουσίας και πηλό (PLATES 10a & 6). Ανάλυση δειγμάτων με το SEM (TABLE 4.2) έδειξε ότι σαν χρωστική ουσία χρησιμοποιήθηκε ένας αργιλικός σχιστόλιθος σκοτεινού χρώματος. Σε αντίθεση με τη μεταγενέστερη Αλοιφωτή Κεραμεική (Crusted Ware), το ροζ επίστρωμα έχει εναποτεθεί πριν από το ψήσιμο των αγγείων. Η σταθεροποίηση του στην επιφάνεια του αγγείου επιτυγχάνονταν με την έντονη στίλβωση του κεραμεικού. Είναι η πρώτη φορά που πιστοποιείται η χρήση επιστρώματος (mineral coating) σε κεραμεικό σύνολο της Αρχαιότερης Νεολιθικής. Μακροσκοπική εξέταση οστράκων με ροζ (και κόκκινο) χρώμα υποδεικνύει ότι μία παρόμοια τεχνική χρησιμοποιήθηκε πιθανότατα και σε άλλους οικισμούς της Αρχαιότερης Νεολιθικής που βρίσκονται στην Δυτική Μακεδονία (Τρίλοφο Βέροιας, Ριζάρι Εδεσσας, Σέρβια) και στην ΝΑ. περιοχή της πρώην Γιουγκοσλαβίας (Anzabegovo, Vrsnik, Lakavica).

ΟΠΤΗΣΗ Τα αγγεία της Νέας Νικομήδειας έχουν ψηθεί σε ανοιχτή φωτιά Αρκετά όστρακα παρουσιάζουν χαρακτηριστικά (τεφρό πυρήνα, "νέφος" στην επιφάνεια) που αποκλείουν μία παρατεταμένη έκθεση των αγγείων σε οξειδωτική ατμόσφαιρα Τα χαρακτηριστικά αυτά όμως είναι πιο σπάνια ανάμεσα στα γραπτά αγγεία Η μικροσκοπική εξέταση του υλικού υποδεικνύει ότι τα αγγεία ψήθηκαν σε θερμοκρασίες που δεν ξεπερνούν τους 750- 800° C.

ΥΛΙΚΑ ΚΑΤΑΣΚΕΥΗΣ ΤΩΝ ΑΓΓΕΙΩΝ Τα αγγεία της Νέας Νικομήδειας μπορούν να διακριθούν, με βάση τα είδη των μη- πλαστικών προσμείξεων που περιέχουν, σε εφτά (7) διαφορετικές κατηγορίες υλικών (fabrics: PLATES 11 ö-14 α). Τα εγκλείσματα που συναντώνται στον πηλό των αγγείων είναι "σύμφωνα'' με τη γεωλογία της περιοχής. Συγκεκριμένα, σε μία ακτίνα 7 χμ γύρω από τη Νέα Νικομήδεια συναντώνται τύποι ορυκτών και πετρωμάτων παρόμοιοι με αυτούς που αναγνωρίστηκαν στον πηλό των αγγείων. Εκτός από τις φυτικές προσμείξεις (PLATE 14 b) που έχουν προστεθεί στον πηλό ελαχίστων αγγείων (1-2 % του δείγματος), τα υπόλοιπα μη- πλαστικά υπήρχαν από την αρχή στα υλικά κατασκευής της κεραμεικής. Τα υλικά κατασκευής καθαρίζονταν από τις μεγάλες σε μέγεθος μη-πλαστικές προσμείξεις με τη μέθοδο της επίπλευσης (settling).

ΠΕΡΙΛΗΨΗ 2Ο3

Μία από τις κατηγορίες υλικών (group 2) χρησιμοποιήθηκε αποκλειστικά για την κατασκευή των γραπτών αγγείων με εξαιρετική στίλβωση που έχουν ερυθρά μοτίβα σε λευκό φόντο (Red/White, Porcelain Ware), ενώ μία άλλη κατηγορία (group 3) χρησιμοποιήθηκε για το χτίσιμο των Porcelain Ware και των γραπτών αγγείων με διακόσμηση λευκή σε ερυθρό φόντο (White/Red). Η σπανιότητα των δύο κεραμεικών κατηγοριών (Porcelain Ware και White/Red) και η ομοιότητα της White/Red κεραμεικής με την αντίστοιχη κεραμεική των οικισμών της Ν.Δ περιοχής της πρώην Γιουγκοσλαβίας, Αλβανίας καθώς και του οικισμού των Γιαννιτσών Β στην Δυτική Μακεδονία, θέτει το πρόβλημα της πιθανής εισαγωγής των αγγείων στη Νέα Νικομήδεια Παρ' όλα αυτά, τα στοιχεία που μας δίνει η συγκριτική μελέτη της κεραμεικής της Νέας Νικομήδειας και των οικισμών που αναφέρθηκαν πιο πάνω, είναι αντίθετα με την υπόθεση της εισαγωγής. Είναι πιθανόν ότι η διαφοροποίηση στα υλικά κατασκευής εξυπηρετούσε τεχνολογικές ανάγκες. Η ταυτόχρονη χρήση πολλών διαφορετικών πρώτων υλών είναι χαρακτηριστική των Νεολιθικών αγγειοπλαστών.

6. ΜΟΝΟΧΡΩΜΗ ΚΑΙ ΔΙΑΚΟΣΜΗΜΕΝΗ ΚΕΡΑΜΕΙΚΗ

Η μονόχρωμη κεραμεική της Νέας Νικομήδειας αντιπροσωπεύεται σε ποσοστό 96% του συνόλου του κεραμεικού υλικού. Τα γραπτά αγγεία καλύπτουν το 88% της διακοσμημένης κεραμεικής ενώ η εμπίεστη και η κεραμεική με επίθετη διακόσμηση εμφανίζονται σε μία αναλογία 9% και 3% αντίστοιχα (FIG. 5.2). Και οι τρεις διακοσμητικοί τύποι ήταν ταυτόχρονα σε χρήση.

ΕΠΕΞΕΡΓΑΣΙΑ ΤΗΣ ΕΠΙΦΑΝΕΙΑΣ

ΜΟΝΟΧΡΩΜΗ ΚΕΡΑΜΕΙΚΗ Έχοντας σαν κριτήριο τον τρόπο επεξεργασίας της επιφάνειας των αγγείων η μονόχρωμη κεραμεική μπορεί να διακριθεί στις ακόλουθες κατηγορίες (FIG. 5.2): αγγεία με κόκκινο- καφέ επίχρισμα τα οποία καλύπτουν το 46% του συνόλου της μονόχρωμης κεραμεικής, αγγεία με ροζ επίστρωμα που καλύπτουν το 24% του συνόλου της μονόχρωμης κεραμεικής, μπεζ μονόχρωμη και κόκκινη-καφέ μονόχρωμη κεραμεική. Η επιφάνεια των αγγείων των δύο τελευταίων κατηγοριών που εμφανίζονται σε μία αναλογία 24% και 3% αντίστοιχα, αφήνεται ακάλυπτη. Για την κατασκευή των μονόχρωμων αγγείων χρησιμοποιήθηκαν έξι (6) από τις κατηγορίες υλικών που αναφέρθηκαν πιο πάνω, αλλά με διαφορετική συχνότητα (FIG. 5.57). Οσον αφορά την καθαρότητα του πηλού, μπορούμε να παρατηρήσουμε ότι υπάρχουν περισσότερα παραδείγματα χονδρόκοκκης και μέσης κεραμεικής στα μπεζ και στα κόκκινα- καφέ μονόχρωμα αγγεία (FIG. 5.58).

ΔΙΑΚΟΣΜΗΜΕΝΗ ΚΕΡΑΜΕΙΚΗ

Γραπτή κεραμεική Η γραπτή κεραμεική μπορεί να διακριθεί στην κεραμεική με κόκκινα μοτίβα σε λευκό φόντο (Red/White) και στην κεραμεική με λευκά μοτίβα σε κόκκινο φόντο (White/Red). Στην πρώτη κατηγορία μπορούμε να διαχωρίσουμε ορισμένα αγγεία, που καλύπτουν το 4% της γραπτής κεραμεικής, τα οποία χαρακτηρίζονται από την εξαιρετική στίλβωση της επιφάνειας τους και το λεπτόκκοκο υλικό που χρησιμοποιήθηκε για την κατασκευή τους (Red/White, Porcelain Ware). Τα υπόλοιπα αγγεία (Red/White, Standard Ware) είναι πολύ πιο συχνά

2Ο4 ΠΑΡΑΣΚΕΥΗ ΓΙΟΥΝΗ

(καλύπτουν το 92% της γραπτής κεραμεικής). Για τη διακόσμηση των αγγείων χρησιμοποιούνται συμπαγή (τρίγωνα, τετράγωνα, ρόμβοι, ταινίες) και γραμμικά μοτίβα (ευθείες, τεθλασμένες, κυματιστές και καμπύλες γραμμές). Σε όλες τις περιπτώσεις η διακόσμηση απλώνεται σε ολόκληρη την επιφάνεια του αγγείου. Παρ' όλα αυτά, για κάθε κατηγορία γραπτής κεραμεικής χρησιμοποιούνται διαφορετικοί συνδυασμοί μοτίβων: τα περισσότερα αγγεία με κόκκινο σε λευκό (Standard Ware) διακρίνονται για τον απλό, κάπως χοντροκομμένο χαρακτήρα των διακοσμητικών μοτίβων (FIG. 5.21, 1; FIG. 5.35, 3 & 4; FIG. 5.36, 10). Τα λεπτόκκοκα αγγεία (Porcelain Ware) κοσμούνται με λεπτότερα γραμμικά μοτίβα (FIG. 5.34, 9; FIG. 5.35, 7; FIG. 5.38, 1), ενώ τα αγγεία με λευκό σε κόκκινο παρουσιάζουν μία διαφορετική οργάνωση της διακόσμησης (FIG. 5.50, 7; FIG. 5.39, 7 & 11 & 13; FIG. 5.49, 6).

Κεραμεικήμε επίθετη διακόσμηση Είναι μία σπάνια κεραμεική κατηγορία Τα αγγεία διακοσμούνται με ανθρώπινα προσωπεία ή φιγούρες, μουσούδες ζώων, ή απλά γραμμικά μοτίβα Τα γραμμικά μοτίβα που κοσμούν τα περισσότερα από τα αγγεία με επίθετη διακόσμηση, είναι ευθείες, καμπύλες, ή τεθλασμένες ταινίες (FIG. 5.53). Συχνά παρουσιάζονται σε συνδυασμό με κομβιόσχημα ή ωοειδή εξογκώματα (FIG. 5.54).

Κεραμεικήμε εμπίεστη διακόσμηση Η διακόσμηση έχει γίνει με εμπιέσεις των άκρων των δακτύλων, με εμπιέσεις των νυχιών, τσιμπιές (ο πιο συχνός τρόπος διακόσμησης: FIG. 5.55) καθώς και με την εμπίεση αντικειμένων που χαράζουν ή δημιουργούν κυκλικά, μακρόστενα και τριγωνικά αποτυπώματα (FIG. 5.55). Αξίζει να αναφέρουμε την ύπαρξη δύο βάσεων που στην εξωτερική τους επιφάνεια φέρουν διακόσμηση (φυτικό μοτίβο, ευθείες γραμμές) η οποία έχει γίνει από πολύ μικρές εμπιέσεις (FIG. 5.56). Για την κατασκευή της εμπίεστης και της κεραμεικής με επίθετη διακόσμηση χρησιμοποιήθηκαν τα ίδια υλικά που χρησιμοποιήθηκαν για τη μονόχρωμη κεραμεική, αλλά με διαφορετική συχνότητα (FIG. 5.59). Οσον αφορά την καθαρότητα του πηλού, το ποσοστό της χονδρόκκοκης κεραμεικής είναι αυξημένο στα αγγεία με επίθετη διακόσμηση (FIG. 5.61). Στην εμπίεστη κεραμεική η συχνότητα είναι παρόμοια με αυτή που παρατηρείται στα μπεζ και κόκκινα- καφέ μονόχρωμα αγγεία

ΣΧΗΜΑΤΟΑΟΓΙΟ Το κύριο κριτήριο που χρησιμοποιήθηκε για τη διάκριση των διαφόρων σχημάτων είναι η "γωνία χείλους" (rim angle) η οποία μετράει την κλίση του άνω τοιχώματος του αγγείου. Ορίζεται σαν η γωνία ανάμεσα στην οριζόντια γραμμή του χείλους και στην ευθεία του άνω τοιχώματος του αγγείου. Το άνω τμήμα των αγγείων της Νέας Νικομήδειας έχει συνήθως ευθεία τοιχώματα, υπάρχουν όμως και κάποια παραδείγματα με ελαφρά κοίλα ή κυρτά τοιχώματα Το κάτω τμήμα είναι στρογγυλό ή οβάλ. Η μετάβαση από το ένα τμήμα στο άλλο είναι ομαλή.

Παρουσιάζεται πρώτα το σχηματολόγιο της μονόχρωμης κεραμεικής η οποία αποτελεί το μεγαλύτερο τμήμα του κεραμεικού υλικού. Η συχνότητα εμφάνισης των διαφόρων σχημάτων παρουσιάζεται στην FIG. 5.62 και η κατανομή των διαμέτρων χείλους στις FIGS. 5.25-5.30. Τα σχήματα που διακρίθηκαν είναι τα ακόλουθα: αγγεία με λαιμό (FIGS. 5.4-5.6) που καλύπτουν το 9% του συνόλου, αγγεία με πολύ κλειστά τοιχώματα (γωνία χείλους μεγαλύτερη από 115°: FIGS. 5.7-5.8) που καλύπτουν 3% του συνόλου, ασκοί (έχουν βρεθεί πολύ λίγα παραδείγματα: FIG. 5.9), αγγεία με ελαφρά κλειστά τοιχώματα (γωνία χείλους από 91-114°: FIGS. 5.10-5.13) που

ΠΕΡΙΛΗΨΗ 2Ο5

αποτελούν την πιο συχνή κατηγορία (33%), ανοιχτά αγγεία με κάθετα τοιχώματα (γωνία χείλους από 85-90°: FIGS. 5.14-5.15) που καλύπτουν το 45% του συνόλου, ανοιχτά αγγεία με γωνία χείλους από 60-84° (FIGS. 5.16-5.18) που εμφανίζονται σε μία αναλογία 23% και αγγεία με πολύ ανοιχτά τοιχώματα (γωνία χείλους μικρότερη από 60°: FIGS. 5.19-5.20) που καλύπτουν το 6% του συνόλου της κεραμεικής.

Οι περισσότερες λαβές είναι απλές διάτρητες κομβιόσχημες προεξοχές (string-hole lugs) Υπάρχουν και λίγα παραδείγματα αδιάτρητων λαβών (FIG. 5.24, 8-10). Ορισμένες ιδιόμορφες λαβές παρουσιάζονται στην FIG. 5.24, 11-12. Οι λαβές, οριζόντια ή κάθετα τοποθετημένες, είναι πιο συχνές στα κλειστά αγγεία αλλά εμφανίζονται με την ίδια συχνότητα στην λεπτή, μέση και χονδρόκοκκη κεραμεική. Τα αγγεία της Νέας Νικομήδειας έχουν χαμηλές δακτυλιόσχημες βάσεις (καλύπτουν το 47% του συνόλου, FIG. 5.21) και δισκοειδείς βάσεις (καλύπτουν το 45%, FIG. 5.22). Σπάνια είναι τα αγγεία με καμπύλο πυθμένα (FIG. 5.23). Ελάχιστα αγγεία στηρίζονταν είτε σε ένα ψηλό πόδι (pedestal base, FIG. 5.23, 4-6) είτε σε περισσότερα πόδια, μικρότερων διαστάσεων (FIG. 5.24, 1-7).

Στην FIG. 5.62 δίνεται η κατανομή των σχημάτων της διακοσμημένης κεραμεικής. Μπορούμε να παρατηρήσουμε ότι στην γραπτή κεραμεική τα αγγεία με πολύ κλειστά τοιχώματα (γωνία χείλους μεγαλύτερη από 115°) είναι πιο συχνά, ενώ τα αγγεία με λαιμό είναι εξαιρετικά σπάνια Στην εμπίεστη κεραμεική παρατηρείται μία αύξηση των αγγείων με ελαφρά κλειστά τοιχώματα (γωνία χείλους από 91-114°) και των αγγείων με πολύ ανοιχτά τοιχώματα (γωνία χείλους μικρότερη από 60°). Η επίθετη διακόσμηση συναντάται σχεδόν αποκλειστικά, στα αγγεία με λαιμό και στα αγγεία με γωνία χείλους από 60-84°.

ΕΞΕΛΙΞΗ ΤΗΣ ΚΕΡΑΜΕΙΚΗΣ

Συγκριτική μελέτη της μονόχρωμης κεραμεικής έδειξε ότι δεν υπάρχει καμιά διαφοροποίηση (στα υλικά κατασκευής και στην επεξεργασία της επιφάνειας: FIGS. 5.66 & 5.65) ανάμεσα στην κεραμεική που προέρχεται από τη πρώτη και δεύτερη οικιστική φάση και σ' αυτήν από τη δεύτερη και την τρίτη φάση. Η μοναδική διαφοροποίηση (αύξηση των αγγείων με λαιμό, στις τελευταίες φάσεις κατοίκησης, και αντίστοιχη μείωση των αγγείων με κλειστά τοιχώματα: FIG. 5.64) είναι πολύ μικρή και δε συνοδεύεται από άλλες τεχνολογικές αλλαγές. Παρόμοια ομοιομορφία παρουσιάζει και η διακοσμημένη κεραμεική. Οι πιο πάνω παρατηρήσεις οδηγούν στο συμπέρασμα ότι οι τρεις οικιστικές φάσεις πρέπει να ακολούθησαν πολύ σύντομα η μία την άλλη, και ότι το τμήμα του οικισμού της Νέας Νικομήδειας που αποκαλύφθηκε ανασκαφικά κατοικήθηκε, πιθανότατα, για σύντομο χρονικό διάστημα

7. ΜΕΛΕΤΗ ΤΗΣ ΧΡΗΣΗΣ ΤΩΝ ΑΓΓΕΙΩΝ

ΥΠΟΛΟΓΙΣΜΟΣ ΤΗΣ ΚΕΡΑΜΕΙΚΗΣ ΠΑΡΑΓΩΓΗΣ Θα πρέπει να επισημάνουμε ότι τα στοιχεία και οι υπολογισμοί που παρουσιάζονται πιο κάτω αναφέρονται μόνον στην κεραμεική που βρέθηκε στην κεντρική τομή (ανασκαφικά τετράγωνα Ω6/Ω7-Μ4/Μ8) της Νέας Νικομήδειας

2θ6 ΠΑΡΑΣΚΕΥΗ ΓΙΟΥΝΗ

ΥΠΟΛΟΓΙΣΜΟΣ ΤΟΥ ΕΛΑΧΙΣΤΟΥ ΑΡΙΘΜΟΥ ΤΩΝ ΑΓΓΕΙΩΝ

Η εκτίμηση της κλίμακας της κεραμεικής παραγωγής είναι βασική προϋπόθεση για τη μελέτη της χρήσης των αγγείων. Για τον υπολογισμό του ελάχιστου αριθμού των αγγείων που βρέθηκαν στην Νέα Νικομήδεια αναπτύχθηκε μία μέθοδος η οποία χρησιμοποιεί σαν βασικά στοιχεία υπολογισμού τη συνολική επιφάνεια των αγγείων (surface area) και τη συχνότητα εμφάνισης των σχημάτων των αγγείων. Χρησιμοποιώντας τη μέθοδο αυτή, υπολογίστηκε ότι ο ελάχιστος αριθμός των αγγείων που βρέθηκαν στην κεντρική τομή της Νέας Νικομήδειας είναι ίσος με 1115 αγγεία

ΠΟΣΟΣΤΟ ΑΝΕΥΡΕΣΗΣ Ανάμεσα στα 1115 αγγεία υπολογίστηκε ότι περιλαμβάνονται 6 αγγεία με εμπίεστη διακόσμηση και 2 αγγεία με άσπρα μοτίβα πάνω σε κόκκινο. Παρ' όλα αυτά, κατά τη διάρκεια της μακροσκοπικής μελέτης του κεραμεικού υλικού διακρίθηκαν όστρακα που ανήκουν τουλάχιστον σε 35 διαφορετικά αγγεία με εμπίεστη διακόσμηση και σε 9 αγγεία με λευκά μοτίβα πάνω σε κόκκινο. Η διαφορά αυτή σημαίνει ότι το ποσοστό ανεύρεσης του κεραμεικού υλικού στην Νέα Νικομήδεια είναι πολύ χαμηλό, δηλαδή γύρω στα 30%-25%. Χρησιμοποιώντας αυτά τα στοιχεία μπορούμε να υπολογίσουμε ότι ο ελάχιστος αριθμός των αγγείων που βρέθηκαν στη Νέα Νικομήδεια (στην κεντρική τομή) κυμαίνεται από 3700 μέχρι 4500 αγγεία

ΕΤΗΣΙΑ ΠΑΡΑΓΩΓΗ Για να αποκτήσουμε μία χονδρική εικόνα της ετήσιας κεραμεικής παραγωγής είναι απαραίτητος ο υπολογισμός του χρονικού διαστήματος κατά το οποίο εναποτέθηκαν τα αρχαιολογικά στρώματα Βασιζόμενοι στα διαθέσιμα στοιχεία υπολογίσαμε μία περίοδο 50- 150 χρόνων για τη δημιουργία των αποθέσεων. Μπορούμε να βρούμε λοιπόν ότι για τα οικήματα που αποκαλύφθηκαν στην κεντρική τομή, η ετήσια κεραμεική παραγωγή/απόκτηση κυμαίνονταν από 25 μέχρι 90 αγγεία Η αντίστοιχη ετήσια παραγωγή (12-13 αγγεία) στο Αρχαιότερο Νεολιθικό Φράγχθι (σπήλαιο και οικισμός στην παραλία), είναι πολύ χαμηλή. Η διαφοροποίηση αυτή ανάμεσα στις δύο θέσεις μπορεί να αντανακλά διαφορές στον τύπο εγκατάστασης (μόνιμος/εποχιακός/προσωρινός οικισμός) ή διαφορές στην οικονομική ενασχόληση των κατοίκων. Μπορεί επίσης να οφείλεται στο γεγονός ότι οι κάτοικοι του Φράγχθι χρησιμοποιούσαν κατά κύριο λόγο σκεύη που ήταν φτιαγμένα από οργανικά υλικά, Η μεγάλη ποικιλία που παρουσιάζουν τα αγγεία της Νέας Νικομήδειας, ως προς το σχήμα και το μέγεθος, ενισχύει την άποψη ότι τα κεραμεικά σκεύη έπαιζαν σημαντικό ρόλο στην καθημερινή ζωή των κατοίκων.

ΧΡΗΣΗ ΤΩΝ ΑΓΓΕΙΩΝ

ΜΑΓΕΙΡΙΚΑ ΣΚΕΥΗ Αύο παράγοντες ελέγχονται συνήθως από τους αγγειοπλάστες έτσι ώστε τα μαγειρικά σκεύη να αντέχουν στους επαναλαμβανόμενους κύκλους θέρμανσης και ψύξης στους οποίους εκτίθενται καθημερινά: το σχήμα και το υλικό των αγγείων (είδη πηλού, είδος και μέγεθος προσμείξεων). Αποφασίστηκε λοιπόν να ερευνηθεί εάν υπάρχει κάποια σχέση ανάμεσα στα σχήματα των αγγείων της Νέας Νικομήδειας και στο υλικό που χρησιμοποιήθηκε για την κατασκευή τους. Τα αποτελέσματα της έρευνας αυτής ήταν αρνητικά: όλα τα είδη των υλικών χρησιμοποιούνται με την ίδια περίπου συχνότητα για την κατασκευή των διαφόρων

ΠΕΡΙΛΗΨΗ 2Ο7

σχημάτων (FIG. 6.2). Δεν υπάρχει επίσης κάποια σχέση ανάμεσα στην καθαρότητα του πηλού και στα σχήματα των αγγείων (FIG. 6.3).

Η έλλειψη εξειδίκευσης δεν αποκλείει αυτόματα τη χρήση κεραμεικών μαγειρικών σκευών. Τα υλικά Α, Β και C περιέχουν μη πλαστικές προσμείξεις που συνήθως συναντώνται στα μαγειρικά σκεύη άλλων θέσεων. Επίσης τα αγγεία της Νέας Νικομήδειας έχουν καμπύλα τοιχώματα, ένα χαρακτηριστικό που βοηθάει την ανθεκτικότητα τους στην θέρμανση και ψύξη. Έμμεση απόδειξη για την ύπαρξη σκευών που χρησιμοποιούνταν για βράσιμο, προσφέρει το αρχαιοβοτανικό δείγμα που περιέχει σπόρους βίκου ( Vicia ervilia). To δηλητήριο που περιέχει ο βίκος μπορεί να απομακρυνθεί όταν οι σπόροι βράσουν και ξεπλυθούν. Άγριος βίκος έχει βρεθεί και σε Μεσολιθικές θέσεις. Σύμφωνα με τα στοιχεία που έχουμε μέχρι τώρα είναι δύσκολο να διακρίνουμε κατά πόσο οι κάτοικοι της Νέας Νικομήδειας χρησιμοποιούσαν τα νέα, κεραμεικά, σκεύη για την απομάκρυνση του δηλητηρίου ή εφάρμοζαν πρακτικές παρόμοιες με αυτές των Μεσολιθικών κατοίκων.

ΑΠΟΘΗΚΕΥΤΙΚΑ ΣΚΕΥΗ Η ασφαλής αποθήκευση της ετήσιας γεωργικής παραγωγής θα πρέπει να ήταν ένα βασικό μέλημα των κατοίκων. Με βάση την πρόταση του Halstead για την ετήσια κατανάλωση δημητριακών και οσπρίων (200 kg για κάθε άτομο) και τον αριθμό των κατοίκων που ζούσαν στα κτίρια που αποκαλύφθηκαν στην κεντρική τομή (30-50 άτομα), μπορούμε να υπολογίσουμε ότι χρειάζονταν αγγεία με χωρητικότητα 7500-12500 λίτρων. Με μαθηματικούς τύπους υπολογίστηκε η χωρητικότητα των αγγείων (22600-27500 λίτρα) που εναποτέθηκαν στην Νέα Νικομήδεια, σε μία περίοδο 50-150 χρόνων. Υπολογίζοντας ότι μόνον ένα τμήμα από τον όγκο αυτόν είχε χρησιμοποιηθεί για αποθήκευση, είναι φανερό ότι τα αγγεία που βρέθηκαν στην Νέα Νικομήδεια δεν ήταν αρκετά για τη φύλαξη της ετήσιας γεωργικής παραγωγής. Για την ασφαλή αποθήκευση των σπόρων που χρησιμοποιούνται για την σπορά της επόμενης χρονιάς (κρατείται 10% της παραγωγής) απαιτείται πολύ λιγότερος όγκος (750-1250 λίτρα). Καθώς τα μεγάλα αγγεία της Νέας Νικομήδειας έχουν όγκο 36-85 λίτρα μπορούμε να δούμε ότι μόνον 15-20 αγγεία χρειάζονται για τον σκοπό αυτόν. Ο αριθμός αυτός των αγγείων είναι σαφώς μέσα στα όρια της κεραμεικής παραγωγής της Νέας Νικομήδειας.

Bibliography

Adams A. E., W. S. Mackenzie and C. Guilford, 1984, Adas of Sedimentary Rocks under die Microscope, London: Longmans.

Bankoff H. and F. Winter, 1979, Ά House Burning in Serbia. What do Burned Remains tell an Archaeologist?', Archaeology 32, 8-15.

BiernoffD., 1969, An Analysis of die Earliest Painted Design Motifs on Pottery from West Anatolia and Greece. Unpublished Ph.D. Thesis, Institute of Archaeology, London.

Biegen C. W., 1975, 'Neolithic Remains at Nemea', Hesperia 44, 227-51. Bronitsky G., 1983, 'Materials science approaches in the study of Virginia ceramics: overview and

initial results', Anthropology 7 (2), 13-20. , 1986, The Use of Materials Science Techniques in the Study of Pottery Construction and

Use', in Schiffer M. B. (Ed.), Advances in Archaeological Method and Theory, 9, 209-76, New York: Academic Press.

Chourmouziadis G. Ch., 1969, 'Πήλινο Ομοίωμα Νεολιθικού Οικίσκου', ΑΑΑ II (iq6q), 36. , 1971» Ή Διακοσμημένη Κεραμεική της Αρχαιότερης Νεολιθικής εις ΘεσσαλΙαν', ΑΕ 1971* 165-87. , 1979» Ενα Ειδικευμένο Εργαστήριο Κεραμεικής στο Νεολιθικό Αιμινι, ΑΑΑ ίο, 207-26.

Chrisostomou P. and Pan. Chrisostomou, i993> 'Νεολιθικές Έρευνες στα Γιαννιτσά και στην Περιοχή τους', Αρχαιολογικό Έργο στη Μακεδονία και Θράκη 4 (ι 99°)» 169-77·

Chrisostomou Pan., 1 994* 'Οι Νεολιθικές Έρευνες στην Πόλη Και την Επαρχία Γιαννιτσών κατά το 1991', Αρχαιολογικό Έργο αχη Μακεδονία και Θράκη *> (iqqi), 111-21.

Clark J. G. D., 1952, Prehistoric Europe, The Economic Basis, 129-51, London: Methuen & Co. Ltd.

Courtois L. G., 1981, 'Etude Physico-chimique de la céramique', in Lambert Ν., La Grotte Préhistorique de Kitsos (Attique) I, S7Q-QO, Paris: ADPF.

and E. Dimou, 1981, 'Annex I: Apercu technologique sur les céramiques Néolithiques de Γ Andre Corycien', in Touchais G., Le matériel néolithique, Γ Andre Corycien I, 173-82 and 248-51. (BCHsuppl. 7).

David N. and H. Henning, 1972, The Ethnography of Pottery: A Fulani Case Seen in Archaeological Perspective, McCaleb Modules Anthroool. 21. Reading Mass: Addison-Weslev.

DeBoer W. R, 1985, The last pottery show: System and sense in ceramic studies', in Pritchard A. C. and Leeuw van der S. (Eds.), The Many Dimensions of Pottery, 529-68, Amsterdam: University of Amsterdam.

, and D. Lathrap, 1979, The making and breaking of Shipibo-Conibo ceramics', in Kramer C. (Ed), Ethnoarchaeology: Implications of Ethnography for Archaeology, 102-38, New York: Columbia University Press.

Ellis L, 1984, The Cucuteni-Tripolye Culture: A study in Technology and the Origins of Complex society, BAR International Series, 217. Oxford.

EricsonJ. E., D. Read and C. Burke, 1972, Research design: the relationships between the primary functions and the physical properties of ceramic vessels and their implications for ceramic distributions on an archaeological site', Anthropology UCLA * (2), 8a-qk.

Fortier A. C, 1981, Chronological and Regional Variation within the Early Neolithic Karanovo I-Kremikovtsi Complex of Bulgaria, Ph.D. Thesis, University of Wisconsin-Madison, Michigan, Ann Arbor, University Microfilms (1981).

209

2 ίο BIBLIOGRAPHY

Foster G. M., i960, 'Life-expectancy of utilitarian pottery in Tzintzuntzan, Michoacan, Mexico', Am. Antiq. 2*, 6o6-q.

French D. H., 1967, Index of Prehistoric sites in Central Macedonia, Athens (privately circulated). , 1970, Tottery Distributions and the Geographical Regions of Macedonia', Zbornik Museum, VI, 5-19·

Gallis C. I., 1082, Καύσβις Νβκρών από tu Νεολιθική Εποχή στη Θεσσαλία, Athens. Garasanin M. V., 1971» 'Anzabegoyo-Barutnica-site néolithique à plusieurs couches. Fouilles de

1969-1970', in Novak G. (Ed), Époque préhistorique et protohistorique en Yougoslavie. Reserches et résultats, 1*6-8, Beograd.

, 1982, 'The Stone Age in the Central Balkan Area', in BoardmanJ., Edwards Ι. Ε., Hammond Ν. G. and Soliberger Ε. (Eds.), Cambridge Ancient History, Second Edition, Volume III, Part I, 75-135, Cambridge: University Press.

Gardner E.J., 1976, The technology of ceramics', in Gimbutas M. (Ed.), Neolithic Macedonia· As Reflected by Excavations at Anza, Southeast Yugoslavia, 159-76, Los Angeles: University of California Press.

, 1978, The pottery technology of the Neolithic Period in Southeastern Europe, Ph.D. Thesis, University of California, Michigan, Ann Arbor, University Microfilms (1980).

Gimbutas M., 1976, Neolithic Macedonia. As reflected by Excavations at Anza, Southeast Yugoslavia, Los Angeles: University of California Press.

, iq8q, Neolithic Achilleion, Los Angeles: University of California Press. Grbi<5 M., P. MaCkic, S. J. Nadj, D. Simoska and B. Stalio, i960, Porodin, kasno-neolitsko naselje

na tumbi kod Bitolja, Bitolj. Halstead P., 1981, 'Counting sheep in neolithic and bronze age Greece', in Hodder, I., Isaac, G.

and Hammond, N. (Eds.), Pattern of the Past: Studies in Memory of David Clarke, Cambridge: University Press.

Hansen J. andj. Renfrew, 1978, 'Palaeolithic-Neolithic Seed Remains at Franchthi Cave, Greece', Nature 271, %aq-w.

Henrickson Ë. and M. McDonald, 1983, iCeramic form and function: an ethnographic search and an archaeological application', Am. Anthr. 8κ (s), i-&.

Hodges H. W. M., 1969, 'Pétrographie examination of the Painted pottery from Nea Nikomedeia', in Biernoff, D., An Analysis of the Earliest Painted Design Motifs on Pottery from West Anatolia and Greece, Appendix A, Unpublished Ph.D. Thesis, Institute of Archaeology, London.

Jones R. E., 1986, Greek and Cypriot Pottery: A review of scientific studies, Athens. 87-107, British School at Athens.

Korkuti M., 1082, 'Vashtëmia-nië vendbanim i neolitit të hershëm', Iliria 1082 (2), Qi-146. Kotsakis K., 1981, 'Excavations at Sesklo 1979: A preliminary report', Anthropologica 2, 87-107. Kramer C, 1979, 'Ceramic Ethnoarchaeology', in Kramer C. (Ed), Ethnoarchaeology:

Implications of Ethnography for Archaeology, 77-102, New York: Columbia University Press.

Lera P., 1983, 'Dy vendbanime prehistorike ne Podgorie', Iliria 1983 (1), 51-77. Linton R, 1944, 'North American Cooking Pots', Am. Antiq. 9, 369-80. Longacre W. Α., 1985, Tottery use-life among the Kalinga, northern Luzon, the Philippines', in

Nelson B. A. (Ed.), Decoding Prehistoric Ceramics, South Illinois: University Press. Middleton A. P, 1987, Technological investigation of the coatings on some "haematite-coated"

pottery from Southern England', Archaeometry 20, 250-61. Millet M., 1980, Ήο'ν much pottery?', in Millet M. (Ed.), Pottery and the Archaeologist, 77-81,

Institute of Archaeology, London: Occasional Publication No. 4. Orton C, 1982, 'Computer simulation experiments to assess the performance of measures of

quantity of pottery', WA, 14 (1), 1-20.

BIBLIOGRAPHY 211

Milojtic-v. ZumbuschJ. and V. Milojdc, 1971, Die Deutschen Ausgrabungen auf der Otzaki- Magula in Thessalien I, Das frühe Neolithikum, Β AM 10/1 1, Bonn.

My lonas G. E., iq2Q, Excavations at Olynthus. I. The Neolithic settlement, Baltimore. Nandris J. G., 1968, The Prehistoric Archaeology of South-East Europe, Unpublished Ph.D.

Thesis, University of Cambridge. , 1971, 'Bos primigenius and the bone spoon', ΒΙΑ ίο, 63-82. , 1977, The Perspective of Long-term Change in south-east Europe', in Carter F. W. (Ed.), An Historical Geography of the Balkans, 2*-*7, London: Academic Press.

Naroll K, 1962, 'Floor Area and Settlement Population', Am. Antiq. 27, 587-9. Nica M., 1977, 'Nouvelles donnés sur le Neolithic Ancien d' Olténie', Dacia XXI, 13-53. , 1978, 'Cuptoare de olariw din epoca neolitic descoperite în Oltenia', Drobeta (Turnu Severin)

2, 18-20. Overweel C. J., 1981, 'Pétrographie thin section and X-Ray Diffraction analysis of pottery from

Sesklo and Achilleion', in Wijnen, M. H., The Early Neolithic I settlement at Sesklo, Appendix I, 105-11, Leiden: Analecta Praehistorica Leidensia XIV.

Papadopoulou M., 1958, 'Μαγουλίτσα. Νεολιθικός συνοικισμός παρά την Καρδίτσαν', θεσσαλικά ι (iQS8), 30.

Phelps W. W., 1975, The Neolithic pottery sequence in Southern Greece, Unpublished Ph.D. Thesis, Institute of Archaeology, London.

Renfrew C, 1973, Trade and craft specialisation', in Theocharis D. R, Neolithic Greece, 179- 01, Athens.

Rice P. M., 1987, Pottery Analysis: A Sourcebook, Chicago and London: The University of Chicago Press.

Rigby V., A. P. Middleton and I. C. Freestone, 1989, The Prunay workshop: technical examination of la Tène bichrome painted pottery from Champagne', WA 21. 1-16.

Rodden Κ J., 1962, 'Excavations at the Early Neolithic site at Nea Nikomedeia, Greek Macedonia (iq6i season)', PPS 28, 267-88.

, 1964 a, 'A European Link with Chatal Huyuk: Uncovering a 7th Millennium Settlement in Macedonia. Part I - Site and Pottery', The Illustrated London News, April 11, ̂64-7.

, 1964 b, Ά European Link with Chatal Huyuk: The 7th Millennium Settlement of Nea Nikomedeia in Macedonia. Part II - Burials and Shrine', The Illustrated London News, April 18, 604-7 , 1964 c, 'Recent Discoveries from Prehistoric Macedonia', Balkan Studies 5, 110-24. , 1965, 'An Early Neolithic Village in Greece', Scientific American 212 (4), 83-gi.

Rye O. S., 1976, 'Keeping Your Temper under Control: Materials and the Manufacture of Papuan Pottery', Archaeology and Physical Anthropology in Oceania 11, 106-37. , 1981, Pottery Technology: Principles and reconstruction, Washington.

Shepard A. O., 10*6, Ceramics for the Archaeologist Washington. Sikalidis K., M. Kesisoglou, E. Mirtsou and K. Alexiadis, 1983, 'Διερεύνηση της Τεχνολογίας

Κεραμεικών Αρχαιολογικών Ευρημάτων της Νεολιθικής Εποχής', Επιστημονική ΕπβτηρΙα της Πολυτεχνικής Σχολής θεσσαλονίκης Ο, ιο7-2^.

Simoska D. and V. Sanev, 1076, Prehistory in Central Pelagonia, Bitola. Smith M. F, 1985, 'Toward an Economic Interpretation of Ceramics: Relating Vessel Size and

Shape to Use', in Nelson B. A. (Ed.), Decoding Prehistoric Ceramics, 255-309, South Illinois: University Press.

Stanislawski Μ. Β., 1977, 'Ethnoarchaeology of Hopi and Hopi-Tewa pottery making: Styles of learning', in Ingersoll D., Yellen J. and Macdonald W. (Eds.), Experimental Archaeology, 378-408, New York: Columbia University Press.

Steponaitis V., 1984, Technological studies of prehistoric pottery from Alabama: physical properties and vessel function', in Pritchard A. and Leeuw S. van der. (Eds.), Tlie many dimensions of pottery, 79-121, Amsterdam: University of Amsterdam.

212 BIBLIOGRAPHY

Theocharis D. R, 1962, 'Από τή Νεολιθνκή Θεσσαλία, Γ, θεσσαλικά 4 (1962), 63. , 1967» Η Αυγή της θεσσαλικής Προϊστορίας: Αρχή και Πρώιμη Εξέλιξη της Νεολιθικής, Volos.

, 1973» Neolithic Greece, Athens. Torrence K, 1978, 'Chipping away at some misconceptions about sampling lithic assemblages', in

Cherry J. F., Gamble C. and Shennan S. (Eds), Sampling in Contemporary British Archaeology, BAR British Series, 50, Oxford.

van Zeist W. and S. Bottema, 1971, 'Plant husbandry in Early Neolithic Nea Nikomedeia', Greece, Acta Botânica Neerlandica 20 (κ), κ24-^8.

Vetnic S., 1972, 'Po eci rada na ispitivanju kultura prvih neolitskih zemljoradnika u srednjem Pomoravlju', Materijali X, 123-68.

Vitelli K. D., 1974, The Greek Neolithic Patterned Urfirnis Ware from the Franchthi Cave and Lerna, Ph.D. dissertation, University of Pennsylvania, Michigan, Ann Arbor, University Microfilms (1980). , 1984, 'Greek Neolithic Pottery by Experiment', in Rice P. M. (Ed), Pots and Potters: Current Approaches in Ceramic Archaeology, 113-31, Los Angeles: University of California Press.

, iq8q, 'Were pots first made for foods? Doubts from Franchthi', WA 21 (1), 17-20. Washburn D. K., 1983, 'Symmetry analysis of ceramic design: two tests of the method on Neolithic

material from Greece and the Aegean', in Washburn, D. K. (Ed.), Structure and Cognition in Art, 138-64, Cambridge: University Press.

, 1984, Ά Study of the Red on Cream and Cream on Red Designs on Early Neolithic Ceramics from Nea Nikomedeia', ATA 88, 305-24.

Weinberg S. S., 1962, 'Excavations at Prehistoric Elateia 1959', Hesperia 31, 158-209. Whitelaw T., 1981, 'The Settlement at Phournou Koriphi, Myrtos, and Aspects of Early Minoan

Social Organisation', in Krzyszkowska O. and Nixon L. (Eds.), Minoan Society, Proceedings of the Cambridge Colloquium 1981, Bristol: Bristol Classical Press.

Wijnen M. H., 1979, 'The Early Neolithic Servia V, in Ridley C. and Wardle Κ. Α., 'Rescue Excavations at Servia 1071-1073: A preliminary report', BSA 74, 180-04.

,1981, The Early Neolithic I settlement at Sesklo. Leiden: Ánalecta Praehistorica Leidensia XIV.

Woods A. J., 1982 a, 'Smoke gets in your eyes: patterns, variables and temperature measurement in open firings', BEFG 1, 11-25. , 1982 b, Thin sections and ceramic technology: an introduction', in Corfield M. and Foley K. (Eds.) Microscopy in Archaeological Conservation, 13-18, United Kingdom Archaeological Conservation Occasional Paper No. 2.

, 1984-85, 'An introductory note on the use of tangential thin sections for distinguishing between wheel-thrown and coil/ring-built vessels', BEFG *, 100-14.

, 1986, 'Form, Fabric, and Function: Some Observations on the Cooking Pot in Antiquity', in Kingery W. D. (Ed.), Technology and Style, 157-72, Columbus, Ohio: American Ceramic Society.

Yiouni P., 1991, The Neolithic Pottery of Nea Nikomedeia in its Balkan Context, Ph.D. Thesis, Institute of Archaeology, London.

, 1994, Technological Analysis of the Neolithic Pottery from Makri (in press BCH 1994).

Plate ι

(a) The Plain of Macedonia showing the low mound of Nea Nikomedeia and the surrounding foothills; (b) Foreman YiannisPapadopoulos (left) with other members of the Greek workforce.

Plate 2

(a) At work clearing the site in 1963; (b) Excavation of the same area.

Plate 3

{a) & (b) Aerial views of the eastern half of the settlement mound, taken from the north west, 1963. Structures of Groups 1, 2, 5 and 6 are visible on the left; Group 4 is the excavated structure

in the centre. (Photographs courtesy of the Greek Army Air Corps.)

Plate 4

(fl) Group 2 structure 3, squares A6-A8 and B6-B8; {b) Close-up of Group 2, structure 3, squares A6-A8 and B6-B8.

Plate 5

(β) Group 4, from south east; (*) Group 4, from north east.

Plate 6

(a) Group 4 structures 1 and 2, from east; (b) Group 5, showing one wall of structure 1 (right), the corner of structure 2 (bottom left) and Late Neolithic Trench (across centre).

Plate 7

(ö) Group 6 structure 2; {b) Group 9 in squares TB1-2 and Tgi-s with Late Neolithic Trench in foreground.

Plate 8

(a) Miniature vessel with traces of pinching; {b) Sherds showing the method of base attachment.

Plate 9

{a) Thrust Lugs'; {b) Red Brown Slip (crossed polars 170 x).

Plate ίο

(a) Pink coating (crossed polars 170X); (b) Pink coating (crossed polars 420X).

Plate ι ι

(a) Fabric A (crossed polars 170X); (b) Fabric A (crossed polars 170X).

Plate 12

{a) Fabric B-i (crossed polars 170X); (b) Fabric B-2 (crossed polars 170X).

Plate 13

(a) Fabric C (crossed polars 170X); [b) Fabric D (crossed polars 170X).

Plate 14

(a) Fabric Ε (crossed polars 170X); (b) Charred plant inclusions (crossed polars 170X).

Plate 15

(a) White on Red Painted Ware; (b) Sherds impressed with finger tips.

Plate 1 6

(a) Sherds impressed with finger pinching and finger nails; [b) Sherds impressed with finger pinching.

Plate 17

(a) Sherds impressed with finger pinching; (b) Sherds impressed with finger nails.