three approaches to task-based syllabus design long crookes

TESOL QUARTERLY Vol 26 No 1 Spring 1992

Three Approaches to Task-BasedSyllabus Design

MICHAEL H LONG and GRAHAM CROOKESUniversity of Hawaii at Manoa

Choice of the unit of analysis in syllabus design is crucial for allaspects of a language teaching program A variety of unitsincluding word structure notion function topic and situationcontinue to be employed in synthetic Type A syllabuses Whileeach is relevant for analyses of the target language and its usenativelike linguistic elements find little support as meaningfulacquisition units from a language learnerrsquos perspective Task hasmore recently appeared as the unit of analysis in three analytic(primarily) Type B alternatives procedural process and tasksyllabuses Each of these has certain limitations too but when thetask syllabus is combined with a focus on form in task-basedlanguage teaching the task receives more support in secondlanguage acquisition (SLA) research as a viable unit around whichto organize language teaching and learning opportunities

Three new task-based syllabus types appeared in the 1980s (a)the procedural syllabus (b) the process syllabus and (c) the tasksyllabus They are distinguishable from most earlier syllabus typesby the fact that part of their rationale derives from what is knownabout human learning in general andor second language learning inparticular rather than as is the case with lexical structural notionalfunctional and relational syllabuses primarily from an analysis oflanguage or language use In addition while differing from oneanother in important ways all three reject linguistic elements (suchas word structure notion or function) as the unit of analysis andopt instead for some conception of task Despite their considerablepotential they are not yet well known outside specialist circles andperhaps for that reason have not received the testing andinvestigation that they (along with older second language syllabustypes) deserve In this paper we present and contrast these threeapproaches to task-based syllabus design and argue that the thirdapproach the task syllabus employed in task-based languageteaching (TBLT) in particular holds special promise

27

MACRO-OPTIONS IN SYLlABUS DESIGN

Syllabus types can be divided into two superordinate classessynthetic and analytic (Wilkins 1974 1976) although it may bemore accurate to view synthetic and analytic as two points on acontinuum rather than as a strict dichotomy (Wilkins 1976)Synthetic syllabuses segment the target language into discretelinguistic items for presentation one at a time

Different parts of language are taught separately and step by step so thatacquisition is a process of gradual accumulation of parts until the wholestructure of language has been built up At any one time the learneris being exposed to a deliberately limited sample of language (Wilkins1976 p 2)

Synthetic that is refers to the learnerrsquos role

The learnerrsquos task is to re-synthesize the language that has been brokendown into a large number of small pieces with the aim of making his[sic] learning task easier (Wilkins 1976 p 2)

The synthetic syllabus relies on learnersrsquo assumed ability to learn alanguage in parts (eg structures and functions) which are indepen-dent of one another and also to integrate or synthesize the pieceswhen the time comes to use them for communicative purposesLexical structural notional and functional syllabuses are syntheticAlthough they need not be so also are most so-called topical andsituational syllabuses for examination of teaching materials showsthat topics and situations have traditionally been used as vehicles forstructural syllabuses (Long amp Crookes in press) mdasha tendency whichhas also begun to occur with some commercially published materi-als that purport to be task-based but are not

Analytic syllabuses offer the learner target language sampleswhich while they may have been modified in other ways have notbeen controlled for structure or lexis in the traditional mannerUsers maintain that

prior analysis of the total language system into a set of discrete pieces oflanguage that is a necessary precondition for the adoption of a syntheticapproach is largely superfluous Analytic approaches areorganised in terms of the purposes for which people are learninglanguage and the kinds of language performance that are necessary tomeet those purposes (Wilkins 1976 p 13)

Analytic that is again refers not to what the syllabus designer doesbut to the operations required of the learner Wilkins (1976) writes

since we are inviting the learner directly or indirectly to recognize thelinguistic components of the language behavior he [sic] is acquiring we

28 TESOL QUARTERLY

are in effect basing our approach on the learnerrsquos analytic capabilities(p 14)

Updating Wilkinsrsquo definition a little analytic syllabuses are thosewhich present the target language whole chunks at a time withoutlinguistic interference or control They rely on (a) the learnersrsquoassumed ability to perceive regularities in the input and to inducerules (or to form new neural networks underlying what looks likerule-governed behavior) andor (b) the continued availability tolearners of innate knowledge of linguistic universals and the wayslanguage can vary knowledge which can be reactivated byexposure to natural samples of the SL2 Procedural process and tasksyllabuses are all examples of the analytic syllabus type Wilkins(1976) classifies situational notional and functional syllabuses asanalytic Notions and functions are clearly linguistic units howeverisolation of which in practice always results in a synthetic syllabussuch that exercises practicing requests or apologies replace exerciseson relative clauses or the present perfect

The analyticsynthetic distinction is partially reflected in a sec-ond classification R V Whitersquos (1988) Type A and Type B sylla-buses However whereas Wilkinsrsquo categories turn on differences inthe way input and learner interact Whitersquos conceptualization isbroader capturing differences in two general approaches to coursedesign instruction language learning and evaluation

Type A syllabuses focus on what is to be learned the L2 They areinterventionist Someone preselects and predigests the language tobe taught dividing it up into small pieces and determining learningobjectives in advance of any consideration of who the learners maybe or of how languages are learned Type A syllabuses White pointsout are thus external to the learner other-directed determined byauthority set the teacher as decision maker treat the subject matterof instruction as important and assess success and failure in terms ofachievement or mastery

Type B syllabuses on the other hand focus on how the languageis to be learned They are noninterventionist They involve noartificial preelection or arrangement of items and allow objectivesto be determined by a process of negotiation between teacher andlearners after they meet as a course evolves They are thus internalto the learner negotiated between learners and teacher as jointdecision makers emphasize the process of learning rather than thesubject matter and assess accomplishment in relationship tolearnersrsquo criteria for success

As will become clear in addition to being analytic all three task-based syllabus types focused on in this paper are primarily Type B

APPROACHES TO TASK-BASED SYLLABUS DESIGN 29

in nature Each allows both language and task to be negotiated inthe classroom Procedural and task syllabuses do have one Type Acharacteristic however for (via different procedures) each makesan initial specification in substantive terms of the kinds of taskslearners will work on before teachers and students ever meet Thatis to say they specify the target tasks learners ultimately need to beable to handle and then allow the tasks teachers and learners workon in the classroom that is the pedagogic tasks to be negotiatedProcess syllabuses conversely are Type B thoroughbreds theyallow negotiation of language and task and in theory at least placeno constraints on the tasks chosen

UNITS OF ANALYSIS THE CASE FOR TASK

Every syllabus needs some unit around which to organise lessonsand teaching materials A case for task as the unit of analysis may bemade on the basis of the problems with potential alternatives andor on the merits of task itself In this section we will briefly considerthe problems with word structure notion function topic andsituation Since the rationale for task as well as its definition variesamong advocates of procedural process and task syllabuses wewill postpone consideration of the merits (and problems) until weexamine the three task-based approaches themselves

Syllabus designers who choose a linguistic elementmdashwordstructure notion or functionmdashas the organizational unit simultane-ously commit to a synthetic Type A syllabus They sometimesattempt to disguise the underlying focus on isolated linguistic formsby avoiding overt drills in the teaching materials that embody thesyllabus and instead while ostensibly dealing with a topic situationor most recently task seed dialogues and texts with the linguisticitem of the day This approach is notorious however for producingstilted samples of the target languagemdashartificial because they arewritten to conform to a set of linguistic specifications (eg a 600-word vocabulary and two verb tenses) supposedly defining ldquolevelsof proficiencyrdquo and so do not reflect how people speak or write(much less learn) the language concerned (see Long amp Crookes inpress) Variants of this position include advocacy of tasks ascarriers or classroom practice devices for traditional syllabus items(Nunan 1989) and the use of pedagogic tasks that are either likelyor guaranteed to elicit particular structures (Loschky amp Bley-Vroman 1990)

Beyond the lack of authenticity synthetic Type A syllabuses areflawed because they assume a model of language acquisitionunsupported by research findings on language learning in or out of

30 TESOL QUARTERLY

classrooms Where morphosyntax is concerned research shows thatpeople do not learn isolated items in the L2 one at a time inadditive linear fashion but as parts of complex mappings of groupsof form-function relationships Nor in principle could languagesbe learned in that way given that many items share a symbioticrelationship Learning English negation for example entailsknowing something about word order auxiliaries and how to markverbs for time person and number Progress in one area dependson progress in the others

Synthetic syllabuses not only present linguistic forms separatelybut also attempt to elicit immediate targetlike mastery of thoseforms Where syntax is concerned research has demonstrated thatlearners rarely if ever move from zero to targetlike mastery of newitems in one step Both naturalistic and classroom learners passthrough fixed developmental sequences in word order negationquestions relative clauses and so onmdashsequences which have toinclude often quite lengthy stages of nontargetlike use of forms aswell as use of nontargetlike forms (See eg Huebner 1983Johnston 1985 Meisel Clahsen amp Pienemann 1981 Sato 1990 andfor review Ellis 1985 Hatch 1983 Larsen-Freeman amp Long 1991Long 1990) As indicated these developmental sequences seem tobe impervious to instruction presumably because linguistic itemshave to be comprehensible and processable before they arelearnable and hence teachable (Pienemann 1984 1987)

Morphological development reveals similar patterns When plurals articles third-person singular s and other morphemes firstappear they tend to do so variably and on certain words or wordclasses first (eg plural s on measure words such as dollars a n ddays) they are not suddenly supplied correctly across allappropriate nouns and verbsmdashagain despite teachersrsquo and textbookwritersrsquo best instructional efforts (Lightbown 1983 Pica 1983Young 1988) Progress is often not even unidirectional Secondlanguage acquisit ion (SLA) frequently involves temporaryldquodeteriorationrdquo in learner performance (so-called backsliding)giving rise to U-shaped and zigzag developmental curves (Seeeg Huebner 1983 Kellerman 1985 Sato 1990)

All synthetic syllabuses not just structurally based ones areflawed in these ways Studies of interlanguage developmentprovide no more support for the idea that learners acquire onenotion or function at a time than for the idea that they master oneword or structure at a time As Prabhu (1984) noted

There are methodological consequencesmdashresulting at least in adifference of emphasismdashto adopting a structural or a functional


syllabus but both kinds of syllabus have the fundamental similarity thatthey look on language acquisition as a planned process of input-assimilation They both rely on the validity of the equation what istaught = what is (or ought to be) learnt (p 273)

The point is moot however since it is the linguistic exponents ofnotions and functions that is structures lexical items intonationpatterns and so on that the learner actually encounters in the inputnot the notions and functions themselves The sequencing of thoseitems may differ from that in a structural syllabus due to forms nowbeing grouped according to communicative function rather thanlinguistic relationships or (supposed) learning difficulty Thelinguistic input to and output demands on the learner howeverstill consist of isolated nativelike structures (e g Would youmind + gerund please as a polite request) mdashstructures which areno more plausible as acquisition units for having their potentialcommunicative function made more salient

If any targetlike linguistic items are learnable separately andcompletely at one time words or collocations may be the mostlikely candidates It seems more reasonable to suppose a learner canconnect items like car and book put on and take off with theirreferents accurately and invariably from Time 1 and do so ondemand not when dictated by some internal syllabus especially ifthe lexical item marks a one-to-one form-meaning relationshipThis belief coupled with advances in text corpus analytictechniques has stimulated renewed interest in the viability of wordsand collocations as units of analysis in syllabus design (Kennedy1987 1990 in press Sinclair 1987 Sinclair amp Renouf 1988) Wheresyllabus design is concerned however problems of authenticityand learnability once again limit the potential of this effort

The authenticity problem arises from the fact that lexicalcollocational or structural frequency counts provide usefulinformation on the relative frequency of occurrence of items inlarge corpora (often of several million running words) but not onthe occurrence of those items in individual texts Therefore ifwriters incorporate authentic examples from the data-based surveyof native-speaker use underlying the linguistic description teachingmaterials based on a lexical syllabus may be expected to improve onprevious work in the way the use of particular vocabulary items andcollocations is illustrated (see eg Willis amp Willis 1988) AS withstructural and notional-functional syllabuses however thematerials are also likely to expose the learner to nonauthenticsamples of the target language overall if whole dialogues or

32 TESOL QUARTERLY

passages are written to conform to word frequency data given thatwhile people demonstrably use (say) 600 words and collocationsmore frequently than others it is unlikely that any single stretch ofauthentic discourse will happen to be lexically graded in this wayThe benefits of the data-based computational work can bepreserved and the problem avoided if the data on use are accessedto guide the presentation of individual items when a brief focus onform is judged appropriate but frequency data ignored in writingtexts That means however that the word is abandoned as the unitof analysis and an alternative is required

The learnability problems for lexical syllabuses are the same asthose for any syllabus using linguistic elements and targetlikemodels as the organizational units While some instantaneousvocabulary acquisition probably does occur normal developmentalprocesses operate here as elsewhere (Blum amp Levenston 1973Laufer 1990) especially when first and second language form-meaning relationships differ in a semantic domain or when non-concrete referents are involved Vocabulary and collocation errorsabound often persisting in advanced learners long after mostgrammatical problems have been cleared up (Hyltenstam 1988Patkowski 1990) As with so many grammatical forms learnersalternate correct use of words with nonnativelike use for longperiods That is they can quickly learn new lexical forms but needtime to understand their precise meaning(s) and selectionalrestrictions ie their use (For data and review see Gass 1989Kellerman 1984)

The last two ostensive units of analysis in synthetic syllabuses aretopic and situation While each is frequently highlighted as a sourceof chapter headings in teaching materials examination shows thatboth units tend to be vague examples often overlap and both haveto date served merely as carriers of linguistic items typically lexicaland structural respectively (for details see Long amp Crookes inpress) The arguments against them as synthetic units therefore arethe same as those against overtly linguistically based syllabuses andmaterials and need not be repeated

In sum whatever the unit of analysismdashstructure notion functionword topic or situationmdashsynthetic syllabuses suffer from somegeneric problems most obviously their static target languageproduct orientation Syllabus content is ultimately based on ananalysis of the language to be learned whether this be overt as inthe case of word structure notion and function or covert as withsituation and topic Further the analysis is conducted on anidealized native-speaker version of that language SLA researchoffers no evidence to suggest that nativelike exemplars of any of


these synthetic units are meaningful acquisition units that they are(or even can be) acquired separately singly in linear fashion orthat they can be learned prior to and separate from language useThe same literature provides overwhelming evidence against all ofthose assumptions in fact

SLA is sufficiently difficult that most learnersrsquo attempts end in atleast partial failure Whatever the relative merits of one unitcompared to another therefore the psychological processesinvolved in learning would seem to have priority over argumentsconcerning alternative ways of analysing the ideal but rarelyattained product While it also involves the acquisition of social andcultural knowledge language learning is a psycholinguistic processnot a linguistic one yet synthetic syllabuses consistently leave thelearner out of the equation

TASK AND THREE TYPES OF TASK-BASED SYLLABUSES

Precursors to Task-Based Syllabuses

Early proposals concerning analytic Type B syllabuses (Macna-mara 1973 Newmark 1964 1966 Newmark amp Reibel 1968Reibel 1969) had little institutional backing and no accompanyingteaching materials distributed by large commercial publishers bothfactors which inhibit the spread of ideas in language teaching goodor bad (Richards 1984) Not surprisingly therefore classroom im-plementation was initially small scale and the result of individualeffort and imagination (Allwright 1976 Dakin 1973 Newmark1971) with one larger institutionalised (ldquocommunicativerdquo ratherthan truly task-based) project the Malaysian Language Syllabus(Kementarian Pelajaran Malaysia 1975 see Long amp Crookes inpress Rodgers 1984 Samah 1984) It is only recently that somemore substantial attempts to use analytic syllabuses have appearedeach using task as the unit of analysis

Procedural Syllabuses

The procedural syllabus is associated with the work in India from1979-1984 of Prabhu Ramani and others on the BangaloreMadrasCommunicational Teaching Project (Prabhu 1980 1984 1987)Early influences were similar to those of the Malaysian communica-tive syllabus but were quickly abandoned

Communicative teaching in most Western thinking has been training forcommunication which I claim involves one in some way or other inpreelection it is a kind of matching of notion and form Whereas the

34 TESOL QUARTERLY

Bangalore Project is teaching through communication and therefore thevery notion of communication is different (Prabhu 1980 p 164)

Prabhu (1987) denies the sufficiency of comprehensible input(Krashen 1982) but he supports the idea that students need plentyof opportunity to develop their comprehension abilities before anyproduction is demanded of them He recognises that acquisition ofa linguistic structure is not an instant one-step procedure andclaims with Krashen that language form is acquired subconsciouslythrough ldquothe operation of some internal system of abstract rules andprinciplesrdquo (Prabhu 1987 p 70) when the learnerrsquos attention isfocused on meaning ie task-completion not language This placeshim firmly in the analytic camp

any attempt to guide [learning] more directly (and whether or notexplicitly) is rejected as being unprofitable and probably harmful Thereis therefore no syllabus in terms of vocabulary or structure nopreelection of language items for any given lesson or activity and nostage in the lesson when language items are practised or sentenceproduction as such is demanded The basis of each lesson is a problemor a task (Prabhu 1984 pp 275-276)

Prabhursquos definition of task for the purposes of the Bangaloreproject was fairly abstract and oriented towards cognition processand (teacher-fronted) pedagogy

An activity which required learners to arrive at an outcome from giveninformation through some process of thought and which allowedteachers to control and regulate that process was regarded as a lsquotaskrsquo(Prabhu 1987 p 24)

In practice two related tasks or two versions of the same task weretypically paired The first or ldquopre-taskrdquo was used by the teacher ina whole-class format perhaps with one or more pupils Its purposewas to present and demonstrate the task to assess its difficulty forthe class (if necessary to modify it accordingly) and perhaps mostcrucial of all what Prabhu (1984) describes vaguely as ldquoto let thelanguage relevant to it come into playrdquo (p 276) The second thetask proper was for the pupils to work on usually individuallyThere followed feedback from the teacher on task accomplishment

Tasks in a procedural syllabus should be intellectually challengingenough to maintain studentsrsquo interest for that is what will sustainlearnersrsquo efforts at task completion focus them on meaning and aspart of that process engage them in confronting the tasks linguisticdemands (Prabhu 1987) Opinion-gap and later information-gapand (especially) reasoning-gap activities were favored in theBangalore project (for discussion see Prabhu 1987) It is important


that learners perceive a task as presenting a reasonable challengethat is as difficult but feasible Difficulty is initially a matter of trialand error and

a rough measure of reasonable challenge for us is that at least half theclass should be successful with at least half the task (Prabhu 1984p 277)

The examples of tasks Prabhu provides are of the kind familiar inthe many variants of so-called communicative language teaching(CLT) which is not task-based in the analytic sense They includecalculating distances and planning itineraries using maps and chartsassessing applicants for a job on the basis of biographical sketchescompleting ldquowhodunitrdquo stories and answering comprehensionquestions about dialogues These are not necessarily activitiesstudents will ever need to do or do in English outside the classroom(although they may be useful for language learning) Similarlyactivities in a procedural syllabus are preset pedagogic tasks notrelated to a set of target tasks determined by an analysis of aparticular group of learnersrsquo future needs

In theory at least the radical departure from CLT the Bangaloreproject represented lay then not in the tasks themselves (seeGreenwood 1985 for a brief critique) but in the accompanyingpedagogic focus on task completion instead of on the language usedin the process (for discussion see Beretta 1989 Prabhu 1990) Twoof the more salient innovations concerned the kind of input to whichpupils were exposed and the absence of overt feedback on errorWith respect to input teacher speech accompanying use of aprocedural syllabus is not preselected or structurally graded butroughly tuned as a natural by-product of the spontaneousadjustments made to communicate with less proficient speakersinside or outside classrooms (Prabhu 1987) Where errors areconcerned ungrammatical learner utterances are accepted for theircontent although they may be reformulated by the teacher (whatPrabhu 1987 p 61 calls ldquoincidentalrdquo as opposed to ldquosystematicrdquocorrection) in the same way that a caretaker reacts to the truth valueof a childs speech and provides off-record corrective feedback inthe process In these and other areas Prabhursquos pedagogic proposalsare strikingly similar to those of the Natural Approach (Krashen ampTerrell 1983)

Despite being an interesting innovative program and all themore praiseworthy for having been carried out under difficultteaching conditions the Bangalore project has been criticised on avariety of grounds one of the chief complaints being its failure tobuild an evaluation component into the design (a criticism rarely

36 TESOL QUARTERLY

made of programs using synthetic syllabuses) More important thanany shortcomings in the way this particular program wasimplemented however is whether or not procedural syllabuses asadvocated by Prabhu are in principle well motivated

There appear to us to be at least three problems with the pro-cedural syllabus as currently conceived

1 In the absence of a task-based (or indeed any) needs identifica-tion no rationale exists for the content of such a syllabus that isfor task selection It is impossible for anyone to verify the appro-priacy of particular pedagogic tasks for a given group of learn-ers without objective evaluation criteria one of which must sure-ly be relevance to learner needs

2 Grading task difficulty and sequencing tasks both appear to bearbitrary processes left partly to real-time impressionistic judg-ments by the classroom teacher Use of a ldquoat least half the taskrdquoby ldquoat least half the classrdquo (or any such ad hoc) criterion forassessing difficulty is not a satisfactory solution for it makes taskachievement a norm-referenced issue reveals nothing aboutwhat made one task ldquoeasierrdquo than another and thereby precludesany generalizations to new materials Moreover if the presenceof a (pedagogic) task in a syllabus is justified (nonarbitrary) atall as we assume it should be then a criterion-referencedapproach is called for The passing grade might vary somewhatbut if a task is a necessary part of the syllabus it is presumablynecessary for all students Seventy percent is accepted as asatisfactory minimum passing grade on many criterion-referenced language tests but higher cut-off points favorincreased decision dependability for such tests (see J D Brown1989a 1990)

3 There are logical arguments having to do with the need fornegative evidence and incomprehensible input in SLA (see egBley-Vroman 1986 L White 1987) and empirical findings oninstructed interlanguage development (Long 1988) whichsupport the need for a focus on form in language teaching yetthis is proscribed in Prabhursquos (as in Krashenrsquos) work

Process Syllabuses

A second task-based approach to course design is the processsyllabus (Breen 1984 1987 Breen amp Candlin 1980 Candlin 19841987 Candlin amp Murphy 1987) The early rationale for processsyllabuses was educational and philosophical not primarilypsycholinguistic with curriculum design proposals for other subject


areas (e g Freire 1970 Stenhouse 1975) constituting an importantinfluence Type A syllabuses were rejected for their interventionistauthoritarian nature

targets for language learning are all too frequently set up externally tolearners with little reference to the value of such targets in the generaleducational development of the learner (Candlin 1987 pp 16-17)

A social and problem-solving orientation with explicit provision forthe expression of individual learning styles and preferences isfavored over a view of teaching as the transmission of preselectedand predigested knowledge This outlook is reflected in Candlinrsquosrather formidable definition of task as

one of a set of differentiated sequenceable problem-posing activitiesinvolving learners and teachers in some joint selection from a range ofvaried cognitive and communicative procedures applied to existing andnew knowledge in the collective exploration and pursuance of foreseenor emergent goals within a social milieu (Candlin 1987 p 10)

Breen and Candlinrsquos focus was and is the learner and learningprocesses and preferences not the language or language learningprocesses They argue that any syllabus preset or not is constantlysubject to negotiation and reinterpretation by teachers and learnersin the classroom Candlin (1984) suggests that what a syllabusconsists of can only be discerned after a course is over by observingnot what was planned but what took place Both Breen andCandlin claim that learning should be and can only be the productof negotiation which in turn drives learning

A Process Syllabus addresses the overall question lsquoWho does what withwhom on what subject-matter with what resources when how and forwhat learning purpose(s)rsquo (Breen 1984 p 56)

Breen (1984 see also Widdowson 1985) advocates replacementof the traditional conception of the syllabus as a list of items makingup a repertoire of communication by one which promotes alearnerrsquos capacity for communication He advocates incorporating acontent syllabus within a process syllabus as an external check onwhat students are supposed to know but he is clear that proceduralknowledge is to replace declarative knowledge as the primaryelement in syllabus content and process is to replace product

Conventional syllabus design has oriented toward language as primarysubject matter An alternative orientation would be towards thesubject-matter of learning a language This alternative provides a changeof focus from content for learning towards the process of learning in theclassroom situation (p 52)

38 TESOL QUARTERLY

The process syllabus is a plan for incorporating the negotiationprocess and thereby learning processes into syllabus design Breen(1984) proposes a hierarchical model with sets of options at fourlevels final selection among which at each level is left for users todecide on Course design consists of providing the resources andmaterials needed for (a) making general decisions about classroomlanguage learning (which students need to learn what how theyprefer to learn it when with whom and so on) (b) alternativeprocedures for making those decisions (the basis for an eventualworking contract between teacher and learners) (c) alternativeactivities such as teacher-led instruction group work andlaboratory use (Breen Candlin amp Waters 1979) and (d) alternativetasks that is a bank of pedagogic tasks students may select from torealise the activities

It is at the level of tasks that the actual working process of the classroomgroup is realized in terms of what is overtly done from moment tomoment within the classroom (Examples at task level would includesuch things as agreeing [sic] a definition of a problem organizing datadeducing a particular rule or pattern discussing reactions etc) (p 56)

Finally procedures are provided for formative evaluation of thee f fec t iveness o f opt ions chosen a t Leve ls b c and d inaccomplishing the goals agreed upon at Level a Breen defines taskas

any structural language learning endeavor which has a particularobjective appropriate content a specified working procedure and arange of outcomes for those who undertake the task lsquoTaskrsquo is thereforeassumed to refer to a range of workplans which have the overall purposeof facilitating language learningmdashfrom the simple and brief exercisetype to more complex and lengthy activities such as group problem-solving or simulations and decision-making (Breen 1987 p 23)

Published criticisms of the process syllabus (see eg Kouraogo1987 R V White 1988) claim that it lacks a formal field evaluationassumes an unrealistically high level of competence in both teachersand learners and implies a redefinition of role relationships and aredistribution of power and authority in the classroom that wouldbe too radical andor culturally unacceptable in some societies Theneed it creates for a wide range of materials and learning resourcesis also noted to be difficult to meet and to pose a threat to traditionalreliance however undesirable on a single textbook which is thesyllabus for most teachers learners and examiners

While understandable these are concerns about the logisticalfeasibility of implementing process syllabuses in certain contextsnot flaws in the process syllabus itself As such they are not


especially pertinent After all one would hardly fault radiation as atreatment for cancer because it is unusable without medicalexpertise consenting patients and radioactive materials Moreoverskepticism about peoplersquos desire and ability to take control of theirown learning is to ignore the success of educational programs of allsorts where learners from different cultural backgrounds have doneexactly that often under the most adverse circumstances (see egArnove 1986 Freire 1970 1972 Hirshon 1983 MacDonald 1985Vilas 1986) as well as 200 years of successful libertarian education(see eg Avrich 1980 Holt 1972 Illich 1971 Spring 1975 andissues of Libertarian Education)

More problematic in our view are some of the same weaknesseswhich we claimed were likely to limit the effectiveness of theprocedural syllabus and which we think are inherent in processsyllabuses

1

40

Like procedural syllabuses process syllabuses deal in pedagogictasks whose availability (in the task ldquobankrdquo) is not based on anyprior needs identification which raises problems for selection Intheir work Breen and Candlin (eg Breen 1987 Candlin 1987)advocate making the range criteria and parameters of choiceknown to teachers and learners but are keen to preserveflexibility to allow for learners and circumstances changing Werecognise that prespecification of syllabus content is preciselywhat Breen and Candlin seek to avoid and accept thatprespecification in most syllabuses and the commerciallypublished materials that embody them suffer from all theweaknesses they allege (in addition to their lack of psycholinguis-tic credibility) We think however that arbitrary selection is dueto the lack of a needs identification not to prespecification perse Moreover while some learners (and teachers) might inpractice recognise which tasks were relevant to their futureneeds (assuming such tasks happened to have been included inthe task bank) and choose to work on them we believe coursedesigners should be better judges of whether and have aresponsibility to ensure that use of class time is as efficient andas relevant as possible and that a (task-based) needs identifica-tion can help achieve this Preselecting pedagogic tasks on thebasis of preidentified target tasks need not mean that learnerchoices in other areas are curtailed although it does admittedlymean limiting the choice of tasks available Nor need it restrictoptions provided at other levels in Breenrsquos (1984) model To usea medical analogy we would like to have patients able to choosefrom among a range of alternative treatments but expect the

TESOL QUARTERLY

physician to limit their choices to remedies for what ails themWhile we recognise that learners are one important source ofknowledge about their needs we believe that a properlyconducted needs identification makes course designers better atdiagnosing those needs (as opposed to wants) than learnersalone We also recognize however following Brindley (1989)that learnersrsquo needs are broad and can change during a course

2 Grading task difficulty and sequencing tasks are discussed byCandlin (1987) where a variety of possible criteria are putforward without any resolution This is a valid reflection of thestate of the art (see Crookes 1986 Nunan 1989 for usefuldiscussion of these issues) but a problem for the process syllabus(and all task-based syllabuses) nonetheless

3 While not ruled out and presumably an option with task designfor the process syllabus no explicit provision is made for a focuson language form For the reasons indicated above in ourcritique of procedural syllabuses we think this is an error

4 It is not clear to what (if any) theory or research in SLA theprocess syllabus is to be held accountable There is relativelylittle reference to the language-learning literature in the writingon process syllabuses This may be a reaction to the tendency forSLA theorists to ignore general education literature when makingproposals for language education However given the strongevidence for at least some uniqueness for language knowledgeand acquisition and given the range of theories developed toaccount for it it is difficult fully to evaluate proposals which arenot obviously and explicitly psycholinguistically motivated

Task-Basked Language Teaching

A third approach to course design which takes task as the unit ofanalysis is task-based language teaching (Crookes 1986 Crookes ampLong 1987a 1987b Long 1985 1989 in press Long amp Crookes1987 in press) TBLT bases arguments for an analytic chiefly TypeB syllabus on what is known about the processes involved in secondlanguage learning (see eg Ellis 1985 Hatch 1983 Larsen-Freeman amp Long 1991 Spolsky 1989) on the findings of secondlanguage classroom research (see eg Chaudron 1988) and onprinciples of course design made explicit in the 1970s chiefly inEFL contexts for the teaching of languages for specific purposes(eg Mackay amp Mountford 1978 Selinker Tarone amp Hanzeli1981 Swales 1985 1990 Tickoo 1988 Widdowson 1979)


The basic rationale for TBLT derives from SLA researchparticularly descriptive and experimental studies comparingtutored and naturalistic learning Results suggest that formalinstruction (a) has no effect on developmental sequences (b) has apositive effect on the use of some learning strategies as indicatedby the relative frequencies of certain error types in tutored anduntutored learners (c) clearly improves rate of learning and (d)probably improves the ultimate level of SL attainment (Doughty1991 Long 1988) These advantages for instruction cannot beexplained as the result of classroom learners having received moreor better comprehensible input which is necessary but insufficient(cf Krashen 1985) for major aspects of SLA Rather while mostcurrent treatment of language as object is undoubtedly wasted forbeing unusable by learners at the time it occurs awareness ofcertain classes of linguistic items in the input is necessary forlearning to occur and drawing learnersrsquo attention to those itemsfacilitates development when certain conditions are met (Schmidt1990a 1990b in press)

To illustrate the following are five examples of how a focus onform can help SLA (a) Work on marked or more marked L2 formscan transfer to implied unmarked or less marked items (EckmanBell amp Nelson 1988 Zobl 1985) (b) Giving increased salience tononsalient or semantically opaque grammatical features maydecrease the time needed for learners to notice them in the inputwhich appears to be necessary if input is to become intake(Schmidt in press Schmidt amp Frota 1986) (c) Increased planningcan promote use of more complex language and possibly ofdevelopmentally more advanced interlingual forms (Crookes1989) (d) Instruction targeted at an appropriate level speeds uppassage through a developmental sequence and extends the scopeof application of a new rule (Pienemann amp Johnson 1987) (e) Twokinds of negative evidence overt feedback on error targeted at anappropriate level and incomprehensible input may help destabi-lize an incorrect rule and can even be essential for this to happen asin cases where the L2 is more restrictive in a given linguistic domainFor example a learnerrsquos L1 may allow two options in adverb place-ment subject-verb agreement after collective nouns or subject pro-noun suppliance in discoursally marked and unmarked contextsand the L2 allow only one of those options While only one of therules is correct when transferred to the L2 however either may becommunicatively successful with L2 speakers with the result thatthe untutored learner may not receive negative input (because theerror never causes a breakdown in communication) and so neverrealise that the form is ungrammatical (L White 1989)

42 TESOL QUARTERLY

The evidence of positive effects for instruction does not supporta return to a focus on forms (plural) in language teaching that is tothe use of some kind of synthetic syllabus andor a linguisticallyisolating teaching ldquomethodrdquo such as audiolingualism the SilentWay or Total Physical Response A focus on forms is ruled out forall the arguments offered earlier against synthetic Type Asyllabuses notably the evidence from SLA research of the need torespect ldquolearner syllabusesrdquo and the related evidence against fullnative-speaker target-code forms as viable acquisition units at thevery least where beginners are concerned

On the other hand the evidence does motivate a focus on form(Long 1991) that is use of pedagogic tasks and other methodolog-ical options which draw studentsrsquo attention to aspects of the targetlanguage code Learner production both grammatical andungrammatical is one source of cues for teachers as to when thiswill be (unproductive interlanguage-sensitive diagnostic testing(eg Pienemann Johnston amp Brindley 1988) is another Whichaspects of the language when how and for which learners all needto be precisely specified (for details see Long in press)

Against this background Long and Crookes (eg Crookes 1986Long 1985) adopt task as the unit of analysis in an attempt toprovide an integrated internally coherent approach to all six phasesof program design and one which is compatible with current SLAtheory There is no suggestion that learners acquire a new languageone task at a time any more than they do (say) one structure at atime It is claimed rather that (pedagogic) tasks provide a vehiclefor the presentation of appropriate target language samples tolearnersmdashinput which they will inevitably reshape via applicationof general cognitive processing capacitiesmdashand for the delivery ofcomprehension and production opportunities of negotiabledifficulty New form-function relationships are perceived by thelearner as a result The strengthening of the subset of those that arenot destabilized by negative input their increased accessibility andincorporation in more complex associations within long-termmemory adds to the complexity of the grammar and constitutes SLdevelopment

The definitions of (both target and pedagogic) task and task typeused by Long and Crookes always focus on something that is donenot something that is said Long (1985) defines (target) task using itseveryday nontechnical meaning

a piece of work undertaken for oneself or for others freely or for somereward Thus examples of tasks include painting a fence dressing achild filling out a form buying a pair of shoes making an airline


reservation borrowing a library book taking a driving test typing aletter weighing a patient sorting letters taking a hotel reservationwriting a check finding a street destination and helping someone acrossa road In other words by lsquotaskrsquo is meant the hundred and one thingspeople do in everyday life at work at play and in between Tasks arethe things people will tell you they do if you ask them and they are notapplied linguists (p 89)

Similarly Crookes (1986) regards it as

a piece of work or an activity usually with a specified objectiveundertaken as part of an educational course or at work (p 1)

Task-based syllabuses utilizing such conceptions of task require aneeds identification to be conducted in terms of the real-worldtarget tasks learners are preparing to undertakemdashbuying a trainticket renting an apartment reading a technical manual solving amath problem reporting a chemistry experiment taking lecturenotes and so forth Valuable expertise in procedures for conductingsuch needs analyses was accumulated by English for specialpurposes (ESP) specialists in the 1970s and 1980s (see eg Berwick1989 Brindley 1989 Candlin Bruton amp Leather 1976 Jupp ampHodlin 1975 Mackay 1978 Selinker 1979) and can still be drawnupon even though most early ESP program designers wereworking within a notional-functional framework Bell (1981)describes a task-based needs identification for a canteen assistant(based on Boydell 1970) as well as the way the resultinginformation can be used for diagnostic and (in Bellrsquos case notional-functional) syllabus design purposes Swales (1990) offers examplesand insightful discussion from the design of a university English foracademic purposes program Yalden (1987) reports on theidentification of the ldquotask typesrdquo relevant for a group of Canadiangovernment officials who would be handling trade and commercein embassies abroad

Once target tasks have been identified via the needs analysis thenext step is to classify them into (target) task types For example ina course for trainee flight attendants the serving of breakfast lunchdinner and snacks and refreshments might be classified as servingfood and beverages Pedagogic tasks are then derived from the tasktypes and sequenced to form the task-based syllabus (for a rationaleand details of these procedures see Long 1985 in press) It is thepedagogic tasks that teachers and students actually work on in theclassroom They will be increasingly complex approximations to thetarget tasks which motivated their inclusion Simplicity andcomplexity will not result from application of traditional linguisticgrading criteria however but reside in some aspects of the tasks

44 TESOL QUARTERLY

themselves The number of steps involved the number of solutionsto a problem the number of parties involved and the saliency oftheir distinguishing features the location (or not) of the task indisplaced time and space the amount and kind of languagerequired the number of sources competing for attention and otheraspects of the intellectual challenge a pedagogic task poses are justa few of the potential grading and sequencing criteria that havebeen proposed (for discussion see G Brown 1989 Brown andYule 1983 Crookes 1986 Long 1985 in press Robinson 1990)

The grading and sequencing of pedagogic tasks is also partly afunction of which various pedagogic options are selected toaccompany their use It is here that some of the negotiation oflearning process urged by Breen and Candlin in their work can bebuilt into TBLT and here too that the findings of a number of linesof SL classroom research over the past 15 years are most helpfulUseful information is available from that work on several relevantissues including but not only the effects on student comprehensionof elaboratively or interactionally modified spoken and writtendiscourse (Parker amp Chaudron 1987 Ross Long amp Yano 1991) theeffects on student production of certain types of teacher questions(eg Brock 1986 Tollefson 1988) the quality and quantity oflanguage use in whole-class and small-group formats (eg Bygate1988 Doughty amp Pica 1986 Longamp Porter 1985) and relationshipsbetween different pedagogic task types (one-way and two-wayplanned and unplanned open and closed here-and-now and there-and-then) on the one hand and negotiation work and interlanguagedestabilization on the other (Berwick 1988 Crookes amp Rulon1988 Pica 1987a Pica Holliday Lewis amp Morgenthaler 1989Robinson 1990 Varonis amp Gass 1985 and for review Crookes1986 Long 1989 Pica 1987b)

Such task-based syllabuses would usually although not exclu-sively imply assessment of student learning by way of task-basedcriterion-referenced tests whose focus is whether or not studentscan perform some task to criterion as established by experts in thefield not their ability to complete discrete-point grammar itemsWhile beyond the scope of this paper it suffices to say that devel-opments in criterion-referenced language testing in the past 15 years(see eg Brindley 1989 J D Brown 1989a 1989b) hold greatpromise for language teaching in general and for TBLT inparticular

TBLT is distinguished by its compatibility with research findingson language learning a principled approach to content selectionand an attempt to incorporate findings from classroom-centeredresearch when making decisions concerning the design of materials


and methodology However it is not without problems of its ownof which the following are some of which we are aware There areno doubt others

1

2

3

4

5

46

We have outlined what we hope is a coherent rationale howeversketchy for TBLT Its research base is as yet limited and someof the second language acquisition and classroom researchfindings referred to may bear alternative interpretations giventhe recency small scale and questionable methodology of someof the studies involvedGiven an adequate needs analysis selection of tasks is relativelystraightforward Assessing task difficulty and s equenc ingpedagogic tasks are more problematic Little empirical supportis yet available for the various proposed parameters of taskclassification and difficulty nor has much of an effort beenmade to define some of them in operational terms (but seeRobinson 1991) Identification of valid user-friendly sequenc-ing criteria remains one of the oldest unsolved problems inlanguage teaching of all kinds (for useful discussion seeSchinnerer-Erben 1981 Widdowson 1968)There is also the problem of finiteness which afflicts all units wehave discussed How many tasks and task types are there Wheredoes one task end and the next begin How many levels ofanalysis are needed What hierarchical relationships existbetween one level and another For example just as wecriticised topic and situation for their vagueness and for thetendency for examples of each to overlap so it must berecognised that task sometimes has the same problem Sometasks for example doing the shopping either could or willinvolve others for example catching a bus paying the farechoosing purchases paying for purchases and so on and someof those ldquosubtasksrdquo could easily be broken down still further forexample paying for purchases divided into counting money andchecking changeTBLT is relatively structured in the sense of being preplannedand guided While we have argued for this in terms of efficiencyor relevance to studentsrsquo needs others could equally well objectto the lesser degree of learner autonomy that the structuringadmittedly produces They could claim that general learningprocesses need more protection than task relevance and that ifthis is done language learning will take care of itselfA few programs have been reported that reflect some principlesof TBLT although often within a different content-based ESL

TESOL QUARTERLY

framework (eg Early Mohan amp Hooper 1989 Yalden 1987)and classroom studies have been conducted of several issues insuch programs (eg Chaudry 1990 Long in press Rankin1990) but no complete program has been implemented andsubjected to the kind of rigorous controlled evaluation we thinkessential

SUMMARY AND CONCLUSIONS

Advocates of process syllabuses procedural syllabuses and TBLTdiffer in the rationale for their proposals in the ways they definetask in whether they conduct a formal needs analysis to determinesyllabus content in how tasks are selected and sequenced and inthe methodological options such as group work and a focus onform that they prescribe and proscribe Their proposals may welldiffer in other areas too but full comparable statements are notavailable for all three proposals on several issues including testingand evaluation

All three proposals have some areas of agreement however mostfundamentally their rejection of synthetic Type A syllabuses andthe units of analysis on which they are based and their adoption oftask as an alternative Consequently all share certain problems Aserious one is the difficulty of differentiating tasks especially tasksand subtasks nested within them which in turn raises questions as tothe finiteness of tasks (or task types) or their generative capacityAnother problem is the issue of task difficulty that is ofdetermining the relevant grading and sequencing criteria These areproblems never resolved for synthetic syllabuses either of coursedespite periodic discussion of such criteria as frequency valencyand (undefined and so unhelpful) ldquodifficultyrdquo but that does notabsolve users of tasks from doing better Finally none of theproposals has yet been subjected to a rigorous field evaluation asituation which will be difficult to resolve at least in the U S wherefunding continues to be allocated to personnel ldquotrainingrdquo but not toresearch on foreign and second language acquisition and languageteaching


ACKNOWLEDGMENTS

We thank Chris Candlin Kevin Gregg Peter Robinson Charlie Sato DickSchmidt and two anonymous TESOL Quarterly reviewers for detailed oftenhighly critical comments on an earlier version of this paper We have incorporatedthose of their suggestions that would not have involved abandoning the wholeenterprise Finally we thank TESOL Quarterly Editor Sandra Silberstein for theseemingly infinite patience and care with which she edited the manuscript Errorsthat remain are very much our responsibility

THE AUTHORS

Michael H Long is Professor of ESL and Chair of the PhD Program in SecondLanguage Acquisition at the University of Hawaii at Manoa He is a member of theEditorial Board of Studies in Second Language Acquisition and coeditor of theCambridge Applied Linguistics Series

Graham Crookes is Assistant Professor of ESL at the University of Hawaii atManoa where he teaches courses in second language learning and languageteaching methodology and supervises the teaching practicum He is a member ofthe Editorial Advisory Board of the TESOL Quarterly and has published in theTESOL Quarterly Studies in Second Language Acquisition Language Learningand Applied Linguistics

REFERENCES

Allwright R (1976) Language learning through communication practiceELT Documents 76 (3) 2-14

Arnove R F (1986) Education and revolution in Nicaragua New YorkPraeger

Avrich P (1980) T h e m o d e r n s c h o o l m o v e m e n t A n a r c h i s m a n deducation in the United States Princeton NJ Princeton UniversityPress

Bell R T (1981) An introduction to applied linguistics Approaches andmethods in language teaching London Batsford

Beretta A (1989) Attention to form or meaning Error treatment in theBangalore Project TESOL Quarterly 23 (2) 283-303

Berwick R F (1988) The effect of task variation in teacher-led groups onrepair o f Engl ish as a fore ign language U n p u b l i s h e d d o c t o r a ldissertation Vancouver Canada University of British Columbia

Berwick R (1989) Needs assessment in language programming Fromtheory to practice In R K Johnson (Ed) The second languagecurriculum (pp 48-62) Cambridge Cambridge University Press

Bley-Vroman R (1986) Hypothesis testing in second language acquisitiontheory Language Learning 36 (3) 353-376

Blum S amp Levenston E (1973) Universals of lexical simplificationLanguage Learning 28 (2) 399-415

Boydell T H (1970) A guide to job analysis London Bacie

48 TESOL QUARTERLY

Breen M P (1984) Process syllabuses for the language classroom InC J Brumfit (Ed) General English syllabus design (ELT DocumentsNo 118 pp 47-60) London Pergamon Press amp The British Council

Breen M P (1987) Learner contr ibut ions to task design InC N Candlin amp D Murphy (Eds) Lancaster Practical Papers inE n g l i s h L a n g u a g e E d u c a t i o n V o l 7 L a n g u a g e l e a r n i n g t a s k s

(pp 23-46) Englewood Cliffs NJ Prentice HallBreen M P amp Candlin C (1980) The essentials of a communicative

curriculum in language teaching Applied Linguistics 1 (2) 89-112Breen M P Candlin C N amp Waters A (1979) Communicative

materials design Some basic principles RELC Journal 10 1-13Brindley G (1989) Assess ing achievement in the learner-centred

curriculum Sydney Australia Macquarie University National Centrefor English Language Teaching and Research

Brock C A (1986) The effects of referential questions on ESL classroomdiscourse TESOL Quarterly 20 (1) 47-59

Brown G (1989) Making sense The interaction of linguistic expressionand contextual information Applied Linguistics 10 (1) 98-108

Brown G amp Yule G (1983) Teaching the spoken language CambridgeCambridge University Press

Brown J D (1989a) Criterion-referenced test reliability University ofHawairsquoi Working Papers in ESL 8 (l) 79-113

Brown J D (1989b) Language testing A practical guide to proficiencyp l a c e m e n t d i a g n o s t i c and achievement tes t ing U n p u b l i s h e dmanuscript University of Hawaii at Manoa Department of ESLHonolulu

Brown J D (1990) Short-cut estimators of criterion-referenced test con-sistency Language Testing 7 (1) 77-97

Bygate M (1988) Units of oral expression and language learning in smallgroup interaction Applied Linguistics 9 (l) 59-82

Candl in C N (1984) Syllabus design as a critical process ( E L TDocuments No 118 pp 29-46) London Pergamon Press amp The BritishCouncil

Candlin C N (1987) Towards task-based language learning InC N Candlin amp D Murphy (Eds) Lancaster Practical Papers i nEnglish Language Education Vol 7 Language learning tasks (pp 5-22)Englewood Cliffs NJ Prentice Hall

Candlin C N Bruton J amp Leather J M (1976) Doctors in casualtySpecialist course design from a database International Review ofApplied Linguistics 14 245-272

Candlin C N amp Murphy D (Eds) (1987) Lancaster Practical Papersin Engl i sh Language Educat ion Vol 7 Language learn ing tasks Englewood Cliffs NJ Prentice Hall

Chaudron C (1988) Second language classrooms Research on teachingand learning New York Cambridge University Press

Chaudry L (1990) TBLT vs ldquoregularrdquo language teaching A comparativeanalysis of classroom language Unpublished manuscript University ofHawaii at Manoa Department of ESL Honolulu


Crookes G (1986) Task classification A cross-disciplinay review (TechRep No 4) Honolulu University of Hawaii at Manoa Social ScienceResearch Institute Center for Second Language Classroom Research

Crookes G (1989) Planning and interlanguage variation Studies inSecond Language Acquisition 11 (4) 367-383

Crookes G amp Long M H (1987a) Task-based second languageteaching A brief report (Pt 1) Modern English Teacher 24 (5) 26-28

Crookes G amp Long M H (1987b) Task-based second languageteaching A brief report (Pt 2) Modern English Teacher 24 (6) 20-23

Crookes G amp Rulon K A (1988) Topic and feedback in nativespeakernonnative speaker conversation TESOL Quarterly 22 (4)675-681

Dakin J (1973) The language laboratory and modern language teachingLondon Longman

Doughty C (1991) Second language instruction does make a differenceEvidence from an empirical study of second language relativizationStudies in Second Language Acquisition 13 (4) 431-469

Doughty C amp Pica T (1986) ldquoInformation gaprdquo tasks Do they facilitatesecond language acquisition TESOL Quarterly 20 (2) 305-326

Early M Mohan B A amp Hooper H R (1989) The Vancouver SchoolBoard Language and Content Project In J H Esling (Ed) Multicultu-ral education and policy ESL in the 1990s A tribute to Mary Ashworth(pp 107-122) Toronto Canada Ontario Institute for Studies inEducation

Eckman F Bell L amp Nelson D (1988) On the generalization ofrelative clause instruction in the acquisition of English as a secondlanguage Applied Linguistics 9 (l) 1-20

Ellis R (1985) Understanding second language acquisition O x f o r d Oxford University Press

Freire P (1970) Pedagogy of the oppressed Harmondsworth EnglandPenguin

Freire P (1972) Cultural action for freedom Harmondsworth EnglandPenguin

Gass S M (1989) Second language vocabulary acquisition A n n u a lReview of Applied Linguistics 1988 9 92-106

Greenwood J (1985) Bangalore revisited A reluctant complaint E L TJournal 39 (4) 268-273

Hatch E (1983) Psychol inguist ics A second language perspect ive Rowley MA Newbury House

Hirshon S (with J Butler) (1983) And also teach them to read WestportCT Lawrence Hill

Holt J (1972) How children fail Harmondsworth England PenguinHuebner T (1983) Linguistic systems and linguistic change in an

interlanguage Studies in Second Language Acquisition 6 (1) 33-53Hyltenstam K (1988) Lexical characteristics of near-native second

language learners of Swedish ]ournal of Multilingual and MulticulturalDevelopment 9 (1 amp 2) 67-84

50 TESOL QUARTERLY

Illich I (1971) Deschooling society Harmondsworth England PenguinJohnston M (1985) Syntactic and morphological progressions in learner

English Canberra Australia Department of Immigration and EthnicAffairs

Jupp T C amp Hodlin S (1975) Industrial English London HeinemannKellerman E (1984) The empirical evidence for the influence of the L1 in

interlanguage In A Davies C Criper amp A P R Howatt (Eds)Interlanguage (pp 98-122) Edinburgh Edinburgh University Press

Kellerman E (1985) If at first you do succeed In S M Gass ampC G M a d d e n ( E d s ) I n p u t i n s e c o n d l a n g u a g e a c q u i s i t i o n(pp 345-353) Rowley MA Newbury House

Kennedy G D (1987) Quantification and the use of English A case studyof one aspect of the learnerrsquos task Applied Linguistics 8 (3) 264-286

Kennedy G D (1990) Collocations Where grammar and vocabularyteaching meet In S Anivan (Ed) Language teaching methodology forthe nineties (pp 215-229) Singapore SEAMEO Regional LanguageCentre

Kennedy G D (in press) BETWEEN and THROUGH The companythey keep and the functions they serve In K Aijmer amp B Altenberg(Eds) English corpus linguistics Studies in honor of ]an SvartvikLondon Longman

Kementerian Pelajaran Malaysia [Malaysian Ministry of Education](1975) Engl ish Language Syl labus in Malays ian Schools Tingatan[Level] IV-V Kuala Lumpur Kementerian Pelajaran

Kouraogo P (1987) EFL curriculum renewal and INSET in difficultcircumstances ELT ]ournal 41 (3) 171-178

Krashen S D (1982) Princip les and pract ice in second languageacquisition Oxford Pergamon Press

Krashen S D (1985) The input hypothesis London LongmanKrashen S D amp Terrell T D (1983) The natural approach Language

acquisition in the classroom San Francisco CA Alemany PressLarsen-Freeman D amp Long M H (1991) An introduction to second

language acquisition research London LongmanLaufer B (1990) lsquoSequencersquo and lsquoorderrsquo in the development of L2 lexis

Some evidence from lexical confusions Applied Linguistics 11 (3) 281-296

Lightbown P M (1983) Exploring relationships between developmentaland instructional sequences in L2 acquisition In H W Seliger amp M HL o n g ( E d s ) C l a s s r o o m o r i e n t e d r e s e a r c h i n s e c o n d l a n g u a g eacquisition (pp 217-243) Rowley MA Newbury House

Long M H (1985) A role for instruction in second language acquisitionTask-based language teaching In K Hyltenstam amp M Pienemann(Eds) Modelling and assessing second language acquisition (pp 77-99)Clevedon England Multilingual Matters

Long M H (1988) Instructed inter language development InL M Beebe (Ed) Issues in second language acquisition Multipleperspectives (pp 115-141) New York Harper amp Row


Long M H (1989) Task group and task-group interactions Universityof Hawairsquoi Working Papers in ESL 8 (2) 1-26 (Reprinted in S Anivan(Ed) Language teaching methodology for the nineties pp 31-50 1990Singapore SEAMEO Regional Language Center)

Long M H (1990) The least a second language acquisition theory needsto explain TESOL Quarterly 24 (4) 649-666

Long M H (1991) Focus on form A design feature in language teachingmethodology In K de Bot D Coste R Ginsberg amp C Kramsch( E d s ) F o r e i g n l a n g u a g e r e s e a r c h i n c r o s s - c u l t u r a l p e r s p e c t i v e(pp 39-52) Amsterdam John Benjamins

Long M H (in press) Task-based language teaching Oxford Basi lBlackwell

Long M H amp Crookes G (1987) Intervention points in secondlanguage classroom processes In B K Das (Ed) Patterns in classroominteraction in Southeast Asia (pp 177-203) S ingapore S ingaporeUniversity PressRELC

Long M H amp Crookes G (in press) Units of analysis in syllabus designIn G Crookes amp S M Gass (Eds) Tasks in language l earningClevedon England Multilingual Matters

Long M H amp Porter P A (1985) Group work interlanguage talk andsecond language acquisition TESOL Quarterly 19 (2) 207-227

Loschky L amp Bley-Vroman R (1990) Creating structure-basedcommunication tasks for second language development University ofHawairsquoi Working Papers in ESL 9 (l) 161-212

Mackay R (1978) Identifying the nature of the learnerrsquos needs InR Mackay amp A Mountford (Eds ) English for specific purposes(pp 21-42) London Longman

Mackay R amp Mountford A (Eds ) (1978) English for speci f icpurposes London Longman

MacDonald T (1985) Making a new people Education in revolutionaryCuba Vancouver Canada New Star

Macnamara J (1973) Nurseries streets and classrooms Some compari-sons and deductions Modern Language Journal 57 250-254

Meisel H Clahsen H amp Pienemann M (1981) On determiningdevelopmental stages in second language acquisition Studies in SecondLanguage Acquisition 3 (2) 109-135

Newmark L (1964) Grammatical theory and the teaching of English as aforeign language In D P Harris (Ed) The 1963 conference papers ofthe English language section of The National Association for ForeignStudent Affairs (pp 5-8) New York National Association for ForeignStudent Affairs

Newmark L (1966) How not to interfere with language learningInternational Journal of American Linguistics 32 (l) 77-83

Newmark L (1971) A minimal language teaching program InP Pimsleur amp T Quinn (Eds) The psychology of second languagelearning (pp 11-18) Cambridge Cambridge University Press

52 TESOL QUARTERLY

Newmark L amp Reibel D A (1968) Necessity and sufficiency inlanguage learning International Review of Applied Linguistics 6145-164

Nunan D (1989) Designing tasks for the communicative classroomCambridge Cambridge University Press

Parker K amp Chaudron C (1987) The effects of linguistic simplificationand elaborative modifications in L2 comprehension University ofHawairsquoi Working Papers in ESL 6 (2) 107-133

Patkowski M (1990) Age and accent in a second language A reply toJames Emil Flege Applied Linguistics 11 (l) 73-89

Pica T (1983) Adult acquisition of English as a second language underdifferent conditions of exposure Language Learning 33 (4) 465-497

Pica T (1987a) Interlanguage adjustments as an outcome of NS-NNSnegotiated interaction Language Learning 37 (4) 563-593

Pica T (1987b) Second language acquisition social interaction and theclassroom Applied Linguistics 8 (l) 1-25

Pica T Holliday L Lewis N amp Morgenthaler L (1989) Compre-hensible output as an outcome of linguistic demands on the learnerStudies in Second Language Acquisition 11 (1) 63-90

Pienemann M (1984) Psychological constraints on the teachability andlearnability of languages Studies in Second Language Acquisition 6186-214

Pienemann M (1987) Psychological constraints on the teachability oflanguages In C Pfaff (Ed) First and second language acquisitionprocesses (pp 143-168) Rowley MA Newbury House

Pienemann M amp Johnston M (1987) Factors influencing the develop-ment of language proficiency In D Nunan (Ed) Applying secondlanguage acquisition research (pp 45-141) Adelaide Australia NationalCurriculum Resource Centre

Pienemann M Johnston M amp Brindley G (1988) Constructing anacquisition-based procedure for second language assessment Studies inSecond Language Acquisition 10 (2) 217-243

Prabhu N S (1980) Reactions and predictions [Special issue] Bulletin4 (l) Bangalore Regional Institute of English South India

Prabhu N S (1984) Procedural syllabuses In T E Read (Ed) Trendsin language syllabus design (pp 272-280) S ingapore S ingaporeUniversity PressRELC

Prabhu N S (1987) Second language pedagogy O x f o r d O x f o r dUniversity Press

Prabhu N S (1990) Comments on Alan Berettarsquos ldquoAttention to form ormeaning Error treatment in the Bangalore projectrdquo TESOL Quarterly24 (l) 112-115

Rankin J (1990) A case for close-mindedness Complexity accuracy andattention in closed and open tasks Unpublished manuscript Universityof Hawaii at Manoa

Reibel D A (1969) Language learning analysis lnternational Review ofApplied Linguistics 7 (4) 283-294


Richards J C (1984) The secret life of methods TESOL Quarter ly 18 (1) 7-23

Robinson P (1990) Task complexity and second language narrativediscourse Unpublished manuscript University of Hawaii at ManoaDepartment of ESL Honolulu

Robinson P (1991) Motivating theories of task complexity Unpublishedmanuscript University of Hawaii at Manoa Department of ESLHonolulu

Rodgers T S (1964) Communicative syllabus design and implementa-tion Reflections on a decade of experience In J A S Read (Ed)Trends in syllabus design (pp 28-51) Singapore Singapore UniversityPress

Ross S Long M H amp Yano Y (1991) Simplification or elaborationThe effects of two approaches to text modification on foreign languagereading comprehension Unpublished manuscript University of Hawaiiat Manoa Department of ESL Honolulu

Samah A A (1964) The English language (communicational) curriculumfor upper secondary schools in Malaysia Rationale design andimplementation In J A S Read (Ed) Trends in language syllabusdesign (pp 193-214) Singapore Singapore University Press

Sato C J (1990) The syntax of conversation in inter language develop-ment Tubingen Germany Gunter Narr

Schinnerer-Erben J (1981) Sequencing redefined (Practical Papersrsquo inEnglish Language Education No 4 pp 1-29) Lancaster EnglandUniversity of Lancaster

Schmidt R W (1990a) The role of consciousness in second languagelearning Applied Linguistics 11 (2) 17-46

Schmidt R W (1990b July-August) Input interaction attention andawareness The case for consc iousness-ra is ing in second languageteaching Paper presented at the 10th Encuentro Nacional de ProfesoresUniversitarios de Lingua Inglesa Rio de Janeiro Brazil

Schmidt R W (in press) Consciousness learning and interlanguagepragmatics In G Kasper amp S Blum-Kulka (Eds) I n t e r l a n g u a g epragmatics Oxford Oxford University Press

Schmidt R W amp Frota S N (1986) Developing basic conversationalability in a second language A case study of an adult learner ofPortuguese In R R Day (Ed) Talking to learn Conversation insecond language acquisition (pp 237-326) Rowley MA NewburyHouse

Selinker L (1979) The use of specialist informants in discourse analysisInternational Review of Applied Linguistics 17 (2) 189-215

Selinker L Tarone E amp Hanzeli V (1981) English for academic andtechnical purposes Rowley MA Newbury House

Sinclair J McH (1987) Collocation A progress report In R Steele ampT Threadgold (Eds) Language topics Essays in honor of MichaelHalliday (pp 319-331) Philadelphia PA John Benjamins

54 TESOL QUARTERLY

Sinclair J McH amp Renouf A (1988) A lexical syllabus for languagelearning In R Carter amp M McCarthy (Eds) Vocabulary and languageteaching (pp 140-158) New York Longman

Spolsky B (1989) Conditions for second language learning O x f o r d Oxford University Press

Spring J (1975) A primer of libertarian education Montreal CanadaBlack Rose Books

Stenhouse M (1975) An introduct ion to curr iculum research anddevelopment London Heinemann

Swales J (1985) Episodes in ESP Hemel Hempstead England PrenticeHall

Swales J (1990) English in academic and research settings Cambridge Cambridge University Press

Tickoo M (Ed) (1988) ESP State of the art Singapore S ingaporeUniversity PressRELC

Tollefson J W (1988) Measuring communication in ESLEFL classesCross Currents 15 (1) 37-46

Varonis E M amp Gass S M (1985) Non-nativenon-native conversa-tions A model for negotiation of meaning Applied Linguistics 6 (l)71-90

Vilas C M (1986) The Sandinista revolution National liberation andsocial transformation in Central America New York Monthly ReviewPress

White L (1987) Against comprehensible input The input hypothesis andthe development of L2 competence Applied Linguistics 8 (l) 95-110

White L (1989) The adjacency condition on case assignment Do learnersobserve the subset principle In S M Gass amp J Schacter (Eds)Linguistic perspectives on second language acquisition (pp 134-158)Cambridge Cambridge University Press

White R V (1988) The ELT curr iculum Design innovat ion andmanagement Oxford Basil Blackwell

Widdowson H G (1968) The teaching of English through science InDakin J Tuffen B amp Widdowson H G Language in educationThe problem in commonwealth Afr ica and the lndo-Pakistan sub-continent (pp 115-175) London Oxford University Press

Widdowson H G (1978) Notional-functional syllabuses 1978 (Pt 4) InC H Blatchford amp J Schachter (Eds) On TESOL rsquo78 EFL policiesprograms practices (pp 33-35) Washington DC TESOL

Widdowson H G (1979) Explorations in applied linguistics O x f o r d Oxford University Press

Widdowson H G (1985) Learning purpose and language use OxfordOxford University Press

Wilkins D A (1974) Notional syllabuses and the concept of a minimumadequate grammar In S P Corder amp E Roulet (Eds) Linguist icinsights in applied linguistics (pp 119-128) Brussels AIMAV ParisDidier

Wilkins D A (1976) Notional syllabuses Oxford Oxford UniversityPress


Willis D amp Willis J (1988) Collins COBUILD English Course LondonCollins

Yalden J (1987) Principles of course design for language teachingCambridge Cambridge University Press

Young R (1988) Variation and the interlanguage hypothesis Studies inSecond Language Acquisition 10 (3) 281-302

Zobl H (1985) Grammars in search of input and intake In S M Gass ampC G M a d d e n ( E d s ) I n p u t i n s e c o n d l a n g u a g e a c q u i s i t i o n(pp 329-344) Rowley MA Newbury House

56 TESOL QUARTERLY

MACRO-OPTIONS IN SYLlABUS DESIGN

Syllabus types can be divided into two superordinate classessynthetic and analytic (Wilkins 1974 1976) although it may bemore accurate to view synthetic and analytic as two points on acontinuum rather than as a strict dichotomy (Wilkins 1976)Synthetic syllabuses segment the target language into discretelinguistic items for presentation one at a time

Different parts of language are taught separately and step by step so thatacquisition is a process of gradual accumulation of parts until the wholestructure of language has been built up At any one time the learneris being exposed to a deliberately limited sample of language (Wilkins1976 p 2)

Synthetic that is refers to the learnerrsquos role

The learnerrsquos task is to re-synthesize the language that has been brokendown into a large number of small pieces with the aim of making his[sic] learning task easier (Wilkins 1976 p 2)

The synthetic syllabus relies on learnersrsquo assumed ability to learn alanguage in parts (eg structures and functions) which are indepen-dent of one another and also to integrate or synthesize the pieceswhen the time comes to use them for communicative purposesLexical structural notional and functional syllabuses are syntheticAlthough they need not be so also are most so-called topical andsituational syllabuses for examination of teaching materials showsthat topics and situations have traditionally been used as vehicles forstructural syllabuses (Long amp Crookes in press) mdasha tendency whichhas also begun to occur with some commercially published materi-als that purport to be task-based but are not

Analytic syllabuses offer the learner target language sampleswhich while they may have been modified in other ways have notbeen controlled for structure or lexis in the traditional mannerUsers maintain that

prior analysis of the total language system into a set of discrete pieces oflanguage that is a necessary precondition for the adoption of a syntheticapproach is largely superfluous Analytic approaches areorganised in terms of the purposes for which people are learninglanguage and the kinds of language performance that are necessary tomeet those purposes (Wilkins 1976 p 13)

Analytic that is again refers not to what the syllabus designer doesbut to the operations required of the learner Wilkins (1976) writes

since we are inviting the learner directly or indirectly to recognize thelinguistic components of the language behavior he [sic] is acquiring we

28 TESOL QUARTERLY













30 TESOL QUARTERLY











32 TESOL QUARTERLY














34 TESOL QUARTERLY














36 TESOL QUARTERLY






Process Syllabuses











38 TESOL QUARTERLY










1

40


TESOL QUARTERLY










42 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY













30 TESOL QUARTERLY











32 TESOL QUARTERLY














34 TESOL QUARTERLY














36 TESOL QUARTERLY






Process Syllabuses











38 TESOL QUARTERLY










1

40


TESOL QUARTERLY










42 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY











32 TESOL QUARTERLY














34 TESOL QUARTERLY














36 TESOL QUARTERLY






Process Syllabuses











38 TESOL QUARTERLY










1

40


TESOL QUARTERLY










42 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY














34 TESOL QUARTERLY














36 TESOL QUARTERLY






Process Syllabuses











38 TESOL QUARTERLY










1

40


TESOL QUARTERLY










42 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY














36 TESOL QUARTERLY






Process Syllabuses











38 TESOL QUARTERLY










1

40


TESOL QUARTERLY










42 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY






Process Syllabuses











38 TESOL QUARTERLY










1

40


TESOL QUARTERLY










42 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY









38 TESOL QUARTERLY










1

40


TESOL QUARTERLY










42 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY










1

40


TESOL QUARTERLY










42 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY










42 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY












44 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY







1

2

3

4

5

46


TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY






ACKNOWLEDGMENTS


THE AUTHORS



REFERENCES











48 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY









































50 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY




































52 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY



































54 TESOL QUARTERLY

























56 TESOL QUARTERLY

























56 TESOL QUARTERLY