Approaches to Language Syllabus Design
Chapter One from D. A. Wilkins, Notional Syllabuses. Oxford University Press, 1976.

1.0 Introduction: synthesis and analysis

One of the major decisions that has to be taken in the teaching of foreign languages is on what basis we will select the language to which the learner will be exposed and which we will expect him to acquire. If we look at existing text-books, at existing syllabuses and at the discussions that have been conducted in journals and books devoted to the teaching of languages, we will see that a variety of approaches have been proposed or adopted. In the case of older textbooks decisions appear to have been taken on a more or less subjective basis, whereas in more recent years the criteria employed have been made more and more explicit. To a considerable extent the different ways of structuring courses reflect different ways of looking at the objectives of language learning and teaching. If a close analysis of objectives has been made, the most obvious pedagogic strategy to adopt in planning to meet those objectives is to follow the components of the analysis step- by-step. Since the learning of a language is most commonly identified with acquiring mastery of its grammatical system, it is not surprising that most courses have a grammatical (or 'structural') pedagogic organisation. Of course there is enormous variety in the ways in which language may be presented in grammatically structured teaching materials themselves, but there are also quite other ways of defining the content of language courses. There are courses based on the systematic introduction of vocabulary and others which take language situations as the starting-point. There are those that adopt a functional approach that resembles parts of the notional syllabus that is to be proposed here. The attempt has also been made to give an operational definition to the objectives of language learning and to plan courses accordingly.

While admitting that in practice these approaches are not necessarily mutually exclusive, regarding them from the linguistic point of view, I would wish to argue that they can be grouped into two conceptually distinct types of approach which could be labelled synthetic and analytic. Any actual course or syllabus could be placed somewhere on the continuum between the wholly synthetic and the wholly analytic, but the actual decision procedures that have been followed in the process of selection will show that it tends towards one pole or the other.

A synthetic language teaching strategy is one in which the different parts of language are taught separately and step-by- step so that acquisition is a process of gradual accumulation of the parts until the whole structure of the language has been built up. In planning the syllabus for such teaching the global language has been broken down probably into an inventory of grammatical structures and into a limited list of lexical items. These are ordered according to criteria which are discussed in the next section. At any one time the learner is being exposed to a deliberately limited sample of language. The language that is mastered in one unit of learning is added to that which has been acquired in the preceding units. The learner's task is to re-synthesize the language that has been broken down into a large number of smaller pieces with the aim of making his learning easier. It is only in the final stages of learning that the global language is reestablished in all its structural diversity.

In analytic approaches there is no attempt at this careful linguistic control of the learning environment. Components of language are not seen as building blocks which have to be progressively accumulated. Much greater variety of linguistic structure is permitted from the beginning and the learner's task is to approximate his own linguistic behaviour more and more closely to the global language. Significant linguistic forms can be isolated from the structurally heterogeneous context in which they occur, so that learning can be focussed on important aspects of the language structure. It is this process which is referred to as analytic. In general, however, structural considerations are secondary when decisions are being taken about the way in which the language to which the learner will be exposed is to be selected and organized. The situational, notional and functional syllabuses described below (pp 15-20 ) are analytic in this sense, as are approaches based on operational definitions.

2.0 Synthetic approaches

The majority of language courses and syllabuses are and probably always have been constructed on synthetic lines. Language learning is a complex task. However, a complex task can usually be broken down into a series of simpler tasks. In recent years and particularly under the influence of advances in the psychology of learning the identification of the smaller learning tasks has been carried out with increasing linguistic sophistication. The tasks are identified with items derived from the description of the language. In those courses which are commonly labelled 'traditional' the control of new linguistic items introduced in any one text-book lesson or unit was not particularly strict. Whole paradigms were presented at a time and often quite distinct linguistic structures would be treated in the same lesson. in the last twenty years or so the use made by structural linguists of the technique of minimal contrast as a criterion for identifying distinct linguistic structures has encouraged text-book writers and syllabus constructors to simplify the learning task still further by reducing to a minimum the quantity of new language in any learning unit. As a result the same learning content is spread over more units and a longer period of time. However, although there is now much more explicit recognition of the criteria that are involved in this process of selecting and ordering language, the learning principle that underlies both types of text remains the same. You facilitate learning if you present the learner with pieces of language that have been pre-digested according to the categories found in a description of the language.

I should add too that matters of method and the exact form in which the new language is presented are not in question here. There are, of course, some very real differences. A new linguistic structure may be presented in the form of an explicit rule; it may be presented as a paradigm; it may be embedded in a dialogue; it may occur in a series of analogous sentences intended to promote inductive learning. None of these differences is relevant to the discussion here. If the content of teaching is in the first place a limitation and an ordering of the forms of the linguistic system, the approach is synthetic.

As methods of teaching have changed, so have the processes by which language is selected and graded. In the case of older text-books decisions appear to have been taken on a more or less subjective basis. At least there is very little discussion of the criteria that were employed. In contrast, the language teaching literature of the past thirty years or so is full of discussions of the various factors to be taken into consideration in deciding which forms of language were to be taught and in which order. This is not the place for a critical and detailed review of the literature on this topic, but a brief discussion of the criteria that have been proposed will be useful since it will be necessary to mention some of them later.

Although in most modern courses control of vocabulary and of grammatical structure go hand-in-hand, the attention of methodologists was first directed to vocabulary. This was presumably because the vocabulary that is needed for predictable day-to-day use of language was markedly different from the somewhat literary and arbitrary vocabulary that learners actually met in their predominantly readingbased courses. It was felt that ways should be found of ensuring that the vocabulary learned should be less haphazardly distributed, more in keeping with the likely needs of the learners and not so large as to constitute a severe learning burden. The aim was to see that the vocabulary content of courses consisted of, in short, the most useful words.

The criteria that have been used in establishing the relative usefulness of words are frequency (1), range (2), availability (3), familiarity (4), and coverage (5). The notion of frequency is self-evident. Range relates to the distribution of a lexical item over a number of different types of text. Availability (disponibilité) accounts for lexical items which may not be particularly frequent but which are readily available to the speaker when he needs them. As with familiarity it is measured by means of speakers' responses rather than by the statistical analysis of texts. In establishing the availability of lexical items subjects are asked to to list the words which they would find most useful in certain defined areas of interest. The degree of familiarity of an item is assessed by asking the subjects to rank words in a given list on a familiarity scale. The coverage of a lexical item is rated high if it expresses a range of meanings or is capable of replacing other items of more specific meaning in particular contexts.

Pedagogic considerations are not ignored in the process of selection. Some items will be promoted because they are particularly useful in the classroom situation. Others will be assessed for their teachability in the light of the techniques for teaching meaning that the teacher wishes to employ. Insistence on the use of ostensive procedures, for example, will make it very difficult to teach some items in the early stages. Again, an item might be deferred where comparison with the mother-tongue suggests that it might present an exceptional degree of difficulty.

As Reibel observes, what is happening here is that we are taking the language behaviour and the language knowledge that we aim to produce in our learners, we are analysing the linguistic components of the desired performance and isolating its units. We are then teaching the units piece by piece so as to get back to the very position from which we started (6). The process of synthesis that is required of the learner is itself based on the results of a prior analysis on the part of the course-book writer. This is true not only of the lexical but also of the grammatical content of language learning. Historically, intensive discussion of grammatical selection and grading is a more recent development, but most writers and methodologists would agree that the grammatical component is central in foreign language learning and in synthetic approaches it is the organization of the grammatical content that provides the essential structure for courses and syllabuses. Each unit of learning usually focusses on some particular aspect of grammatical structure and, whether or not this is made explicit, teachers themselves will usually identify an individual unit by means of some grammatical label.

The vocabulary that is chosen for inclusion in a general language course is only a small proportion of the total lexicon of the language. The process of selection therefore is no less important than that of ordering. With the grammar the position is rather different. The ultimate goal of a general course will be to teach virtually the whole of the grammatical system. Whereas limitation will be necessary in the case of courses of short duration or of those having as a goal some kind of restricted language competence, the problems faced in determining the grammatical content of general courses are more those of staging and sequencing (7). By what criteria does one decide which grammatical structures will have to be taught at certain stages and how they will have been sequenced in relation to one another within each stage?

The linguistic criteria that are most often cited in relation to the grammatical content of teaching are simplicity, regularity, frequency and contrastive difficulty (8). There is no particular difficulty in understanding any of these concepts. It is suggested that more simple language should be taught before more complex on the not unreasonable assumption that simplicity of structure implies ease of acquisition. Judgements of simplicity are still made on a largely intuitive basis, since linguistics has not yet provided us with a means of measuring complexity which has proven psycholinguistic validity. The criterion of regularity requires that the most productive linguistic structures should be taught before those of low productivity. The reason why the content of the early stages of so many courses is similar is that they deal with those linguistic forms that have the greatest generalizability and whatever type of linguistic description has been used to derive the language content, the same, basic facts are likely to emerge. Some grammatical forms are so necessary to any use of the language that they can only be avoided in the early stages of a course at the cost of the greatest artificiality. The criterion of frequency is rarely used at all rigidly. It is more often simply a matter of deferring to a later stage the learning of forms that are evidently obscure or rarely used. A great deal has been written on the subject of constrastive difficulty. Most of it, however, remains at the level of description and there is very little discussion of how our understanding of particular contrastive problems influences the detail of course and syllabus design. In general it is suggested that the early stages of learning should be devoted to language forms which present the fewest contrastive difficulties.

Other criteria once again involve the interaction with pedagogic considerations. If it is intended that new language forms should be presented in a context of day-to-day language use, forms which have special social utility or probability of occurrence are likely to be promoted. As with vocabulary, grammatical forms will have higher or lower priority according to their degree of pedagogic utility, their appropriateness to the classroom context and their teachability in the light of the methods and techniques that the teacher wishes to adopt. Most important of all is the fact that the whole of this strategy of teaching is based on the principle of working from the familiar to the unfamiliar and of using the familiar to teach the unfamiliar. The efficient teaching of one item will presuppose the prior acquisition of certain other items. The factors involved will be partly linguistic and partly pedagogic and they will result in preferred orderings of grammatical forms -- what have been called relations of recommended precedence (9).

One of the problems faced in selecting and grading language is that the various lexical and grammatical criteria conflict with one another as often as they complement one another and there is no way in which weightings can be given to them. A highly desirable lexical item may cause grammatical difficulties. Productive forms may nonetheless be complex. As a result, a good deal of the decision-making remains subjective. The individual teacher, writer, syllabus-constructor will, in any given instance, have to decide for himself to which criterion he will attach the greatest importance.

The syllabus that results from the application of these criteria will be a grammatical syllabus. The use of a grammatical syllabus can be regarded as the conventional approach to language teaching since the majority of syllabuses and published courses have as their core an ordered list of grammatical structures. The vocabulary content is secondary in importance and certainly rarely provides the basic structure of a course. The view is widely held that until the major part of the grammatical system has been learned, the vocabulary learning, load should be held down to what is pedagogically necessary and to what is desirable for the sake of ensuring adequate variety in the content of learning. From this point on, the grammatical syllabus will be regarded as the archetype of a synthetic approach to syllabus design.

3.0 Reservations about synthetic approaches

In recent years a number of arguments of varying degrees of importance and validity have been put forward for questioning the adequacy of a grammatical syllabus. It is not generally denied that what is learned through a grammatical syllabus is of value to the learner. It is rather suggested that this is not the necessary or the most effective way of designing language courses and that, in any case, language learning is not complete when the content of a grammatical syllabus has been mastered.

Reibel, in the article already referred to above, argues that the elaborate procedures of analysis leading to re-synthesis are superfluous since they aim to recreate the very language behaviour that was the starting-point of the analysis. In that case, he says, why not base language learning directly on the language corpus from which the analysis was derived? There are ways of exploiting the language found in a corpus that could lead to effective learning, However the analysis that underlies a grammatical syllabus is not often in practice based on an identifiable corpus. It is more likely to be based on existing descriptions of the language and on what, by common consent, course producers have actually included in language learning materials and syllabuses. If one shares Reibel's view, therefore, how does one choose the corpus of authentic language material on which the learning is to be based? The time available for learning is short and it does not seem reasonable to suggest that a random exposure to language will suffice. The issue of selection will still have to be faced and if one does not want to use criteria that stem from grammatical descriptions of the language, others will have to be used in their place.

One danger in basing a course on a systematic presentation of the elements of linguistic structure is that forms will tend to be taught because they are there, rather than for the value which they will have for the learner. Sometimes irregular verbs are introduced for the sake of completeness even where they are likely to be of little use to the learner. However this is a criticism of actual syllabuses not of grammatical syllabuses in principle, since the proper operation of the criteria listed above should prevent this kind of thing occurring. The danger is greatest where learners require the language for some restricted purpose. If the content is planned with general linguistic considerations in mind and with inadequate attention paid to the grammatical (and lexical) characteristics of the language performance implicit in the learners' objectives, much time may be spent in the acquisition of language that is at best marginally relevant and too little time on forms that are of particular value to this group of learners. In a word, learning will be inefficient.

One characteristic of grammatical syllabuses, a characteristic that is also found in some kinds of teaching material, is that what has to be learned is identified as a form and rarely as a set of meanings. Most syllabuses are in fact an inventory of grammatical forms. It is very rare for grammatical meanings also to be specified. The assumption seems to be that form and meaning are in a one-to-one relation, so that the meaning to be learned in association with a particular grammatical form would be self-evident. In practice, language is not like that. A single grammatical form may be semantically quite complex. The learning of grammatical meaning needs to be planned no less than the learning of grammatical forms. If this is not done, it will tend to be assumed that learning is complete when there is mastery of the formal devices or when a partial semantic interpretation can be put upon a form. In materials themselves learning of form is sometimes adequately provided for, but the learning of meaning is neglected. This kind of criticism can be met without abandoning the framework of a grammatical syllabus.

A greater difficulty and one to which there is not an obvious answer lies in the fact that the syllabus is an ordered list of structures. If the content is expressed by use of grammatical terminology, units will be identified by such labels as the definite article, the past tense, transitive sentences, adverbs of frequency, the order of adjectives, the comparative and so on. Alternatively the content might be expressed through examples, or, most likely of all, through both. The items that are identified in this way are only rarely syntactic structures like transitive sentences or the order of adjectives above. More often they are items which contrast paradigmatically with other items in the syllabus and which may well be morphologically distinct. By this I mean that the definite article contrasts with the indefinite article and the past tense with the present tense. Each of these is a term in a grammatical system and the total number of terms is limited. It is possible to learn all the terms of a system and the exact relationship between them. Success in learning the grammar of a foreign language is usually measured in terms of the degree of mastery over paradigmatic systems of this sort.

Although these systems are listed exhaustively in a syllabus, the syntactic structures in which they occur in the language are not. Of course the fundamental facts of syntax are almost inevitably taught, but there remains a good deal that is not. Let us take an example. In the unit labelled comparative the learner will learn such facts as that older is in contrast with old and oldest. He will learn that this is a typical comparative formation and that certain other adjectives form their comparatives differently. He will also learn that a comparative adjective co-occurs with than and he will probably practise the comparative through syntactic structures like John is older than Peter. What the syllabus or the course will never do, either at this point or at a later stage, is make it clear that the comparative occurs in sentences like the ruins were older to a considerable degree than had originally been thought or older than the discovery of electricity was the invention of the steam-engine. A multiplicity of other comparative sentences are also possible, all syntactically distinct from one another. The fact is, therefore, that the inventories found in grammatical syllabuses give insufficient attention to syntax. There is a good reason why this should be so. The actual number of structures (sentences) that are possible in any language is infinite. This is because, as linguists have recently once again emphasized, a finite set of grammatical rules is capable of producing an infinity of sentences. Any syllabus which is an itemized list of points or structures to be learned is inevitably incomplete. It is impossible therefore, for a grammatical syllabus to cover the grammatical facts exhaustively. On the other hand they could (and should) be modified to take greater account of the importance of the acquisition of syntax. It would not be difficult even within a grammatical syllabus to show that sentence structures can be extremely varied, so that the learner is not encouraged to believe that what he is being taught constitutes the structures of the language. One of the major reasons for questioning-the adequacy of grammatical syllabuses lies in the fact that even when we have described the grammatical (and lexical) meaning of a sentence we have not accounted for the way in which it is used as an utterance. It is this apparent paradox that has led philosophers to try to define meaning as use. The fact is that sentences are not confined in use to the functions suggested by the grammatical labels that we give to them, nor does one use of language require the selection of one particular grammatical form. In an interesting discussion of the pedagogic significance of the discrepancy between what is signified grammatically and what is actually communicated Widdowson makes the following observation: (10)

One might imagine, for example, that the imperative mood is an unequivocal indicator of the act of commanding. But consider these instances of the imperative: 'Bake the pie in a slow oven', 'Come for dinner tomorrow', 'Take up his offer', 'Forgive us our trespasses'. An instruction, an invitation, advice and prayer are all different acts, yet the imperative serves them all; -- and need serve none of them: 'You must bake the pie in a slow oven', 'Why don't you come to dinner tomorrow?' 'I should take up his offer', 'We pray for forgiveness of our trespasses'. But one might suppose, nevertheless, that though there are several different kinds of act that can be performed by the imperative, when an order is to be given, it is always the imperative which is used. But this, of course, is not the case either. Just as one linguistic form may fulfil a variety of rhetorical functions, so one rhetorical function may be fulfilled by a variety of linguistic forms.
Widdowson then quotes a number of examples from Labov of the different way in which a teacher may phrase a command." (11)
This should be done again.
You'll have to do this again.
You can do better than this.
It's my job to get you to do better than this.
Given the teacher's relationship to his pupils, these will all be intended as commands and, indeed, will be interpreted as such, although, of course, these sentences are not synonymous. The most significant thing about this example is that the teacher will normally be understood as lie intends to be understood. If this was not so, communication could scarcely take place. We are able to convey grammatical meaning in any situation where speaker and hearer are familiar with the grammatical system. The hearer knows the grammatical rules that the speaker is using. Since those things that are not conveyed by the grammar are also understood, they too must be governed by 'rules' which are known to both speaker and hearer. People who speak the same language share not so much a grammatical competence as a communicative competence. Looked at in foreign language learning terms, this means that the learner has to learn rules of communication as well as rules of grammar. The conventions that relate the linguistic form of an utterance to its actual communicative effect are not universal. What is permissible in the use of one language may not be permissible in another. Since there will be similarities and differences between languages, the learning of the communicative conventions no less than the learning of the grammatical conventions has to be planned for. A grammatical syllabus, however, provides for the acquisition of a grammatical competence and embodies the assumption either that grammatical function and communicative function are the same thing, or that the learner himself can readily acquire knowledge of the communicative aspects of language during or after his acquisition of the grammatical system. The aim of the present study is to find a better way of taking account of the communicative aspects of language than is possible within the framework of a grammatical syllabus.

For learners, probably the most striking way in which the knowledge of language developed through a grammatical syllabus fails to measure up to their communicational needs is in its lack of situational relevance. They may have learned through oral, active methods and, indeed, have a considerable practical command of grammatical structures, but the language that they rehearse in the classroom will be inadequately related to what is needed in the situations in which they may actually want to use the language. Even the relatively uninformed learner is aware that there are ways of using language that are appropriate to the situation in which and for which it is required. What one may want to say in one situation differs from what is needed in another. Some specialised uses of language are quite distinctive and there is enough variety in language for one to be cautious about regarding language as a monolithic whole. The language that occurs in any one situation can be regarded as a sub-variety of the whole in the particular way in which it exploits the grammatical core of the language. The grammatical syllabus focusses learning on the core and not on the distribution of that core in particular uses. As a result, even the learner who knows the core may not be able to communicate adequately when he finds himself in a situation requiring language.

I should add that there is language teaching, based on a grammatical. syllabus, which is sometimes called situational. The label is most commonly applied to a method of teaching in which language is always taught in association with some physical characteristic of the classroom. Objects, pictures and activities are used to illustrate and give meaning to grammatical and lexical forms. Tenses, for example, are often presented in association with physical activities on the part of teacher or pupils. It is clear, however, that the situation referred to here is a pedagogic, classroom situation, not a situation of natural language use. It cannot, therefore, meet situational language needs, however effective it might be as a pedagogic device. A grammatical syllabus can also be 'situationalized' by presenting language in the form of dialogues. The dialogues are written to illustrate a grammatical point and apart from the use of situationally restricted formulae, they rarely resemble natural language use, nor do they enlighten the learner's understanding of the appropriateness of form to context and purpose. An interesting point about virtually all modern courses derived from a grammatical syllabus is that the intensive practice materials involve the repeated production of sentences having like structures. Such sentences are not related to one another thematically as would be the case in natural language use. Equally, in natural language interaction sentences of identical structure scarcely ever co-occur. Neither of these approaches represent any radical revision of the grammatical syllabus to take account of situational needs.

The contrast between language as it is experienced in the classroom and language as it is known to be used in society often makes it difficult for the learner to appreciate the value of what he is learning. The motivation of learners is hard to sustain when success is measured in terms of the proportion of the grammatical system known. Although some learners are able to see that an investment in learning effort now should produce practical benefits in the future, many are looking for a much more immediate return for the effort expended. Their motivation will be less likely to fade if they are continually aware that this is not an unapplied, and from their point of view perhaps unapplicable system, but a genuine means of communication. The argument is the stronger if the learners are already in a situation where they can or need to use the language they are learning. On grounds of motivation, therefore, as well as on linguistic grounds, there are reasons for looking for an alternative to the grammatical syllabus as a strategy for structuring the learner's experience of language.

4.0 Analytic approaches

The prior analysis of the total language system into a set of discrete pieces of language that is a necessary precondition for the adoption of a synthetic approach is largely superfluous in an analytic approach. As we shall see, this is not to say that we make no use of the structural facts of language in making decisions, merely that they are not the starting-point. Analytic approaches are behavioural (though not behaviourist). They are organized in terms of the purposes for which .people are learning language and the kinds of language performance that are necessary to meet those purposes. The problem of putting an analytic approach into practice is largely one of finding a way to express what it is that people do with language, so that the unavoidable process of limitation or selection can take place. The units in any language teaching material based on an analytic approach are not primarily labelled in grammatical terms. They are identified according to whatever behavioural metalanguage we have eventually decided upon. Of course, since it is language behaviour we are concerned with, it is possible, indeed desirable, that the linguistic content of any unit should also be stated, but it is a content that is derived from the initial behavioural analysis. It cannot be established independently of it.

Actual language behaviour is structurally very varied and since the organization of teaching in analytic approaches is in terms of types of actual language behaviour, structural diversity in any analytic syllabus or teaching materials is inevitable. In the actual teaching process pieces of this heterogeneous structural content will be extracted and isolated so that learning can be focussed upon them. The need for the learner to benefit from significant linguistic generalizations cannot be ignored, but the units of language treated in this way will not necessarily be minimally distinct from one another as is usually the case in synthetic approaches. Whether or not the aspects of language structure involved are brought explicity to the learner's attention is a methodological matter and does not concern us directly here. However, since we are inviting the learner, directly or indirectly, to recognize the linguistic components of the language behaviour he is acquiring, we are in effect basing our approach on the learner's analytic capacities. This approach is therefore in contrast with those approaches that rely more upon his capacity to synthesize.

4.1 Operational definitions

An operational definition is a way of stating terminal or intermediate objectives rather than a way of constructing actual teaching sequences. The statement of objectives would not normally include a specification of linguistic structures, that is to say, there would be no inventory of words or structures such as we have seen above. As the components of the language behaviour are not specified, the learner's task cannot be one of re-synthesis. Operational definitions are, therefore, here categorized as analytical. The potential of operational definitions in language teaching is largely unexplored, so that there is only brief discussion of the issue here.

As the name suggests, the essence of an operational definition is that it defines the operations that the individual is capable of performing. In the case of language behaviour the operations are language operations. The operations themselves must be identified in such a way as to be measurable in strictly quantitative terms. The quantities themselves should reflect identifiable aspects of behaviour. The approach is perhaps best understood through some examples.

We could set as objectives for a science student studying through the medium of a foreign language that he should be able to read student text-books in his field of specialization at a rate of x words per minute (w.p.m.) and with y degree of comprehension (as measured by some standard form of comprehension measurement). An intermediate objective for the same student might be that he could read the same type of text with the same degree of comprehension but at a rate of x/2 w.p.m. Alternatively the rate could be kept stable but the degree of comprehension varied. Objectives for a secretary could be that she should be able to take foreign language dictation given at x w.p.m. with y degree of accuracy. In setting intermediate goals, x and y could be varied, although, presumably, the ultimate value of y would have to be 100%. This would also apply to an interpreter whose capacity in simultaneous interpreting can be measured in terms of rate and accuracy.

These are two interesting things about these examples. In an cases the form of the linguistic input to the learner is held constant. There is no attempt to define such notions as elementary, intermediate and advanced in terms of the degree of complexity of the language structure, as would be the case in a synthetic approach. The language is the language of the terminal behaviour and levels of proficiency are assessed in terms of the degree of capacity to perform the terminal behaviour itself. Levels of proficiency indicated in these terms are more easily relatable to actual language performance than statements of numbers of linguistic structures or lexical items known.

The second point is that in each of these examples the language performance takes the form of a response to a clearly identifiable external linguistic stimulus. It is unlikely that other kinds of language performance could be identified and measured in such simple terms. Much of our production of language is not relatable to external stimuli in this way. It is true that the situational and semantic dimensions that are discussed below are capable of handling productive aspects of language and that, at the same time, they can be thought of as contributing to operational definitions. In that case, however, it is clear that measurement of language proficiency is not going to be quite as straightforward as the above examples suggest. The full operationalization of language teaching objectives still seems to be something for the future.

4.2 Situational syllabuses

I have suggested that the framework for most foreign language teaching is provided by a grammatical syllabus and that dissatisfaction with this shows itself most readily in concern that the language acquired in this way is not adequate for situational needs. It is hardly surprising, therefore, that the most commonly proposed alternative is to take situational needs as the starting-point and thereby to construct a situational syllabus to replace the grammatical syllabus. It is the only other kind of syllabus that is at all widely used as a basis for the construction of teaching materials.

The argument for the situational syllabus is fairly straightforward. Although languages are usually described as general systems, language is always used in a social context and cannot be fully understood without reference to that context. Our choice of linguistic forms may be restricted according to certain features of the social situation and, in any case, we need the language so that we can use it in the situations that we encounter. Therefore, rather than orientate learning to the subject and its content, we should take account of the learner and his needs. We should predict the situations in which the learner is likely to need the language and then teach the language that is necessary to perform linguistically in those situations. It will be a more efficient process because it will include only what is relevant to the learner. It will be more motivating because it is learner- rather than subject-centred. The distinction between language for learning and language for use will disappear. Units in the syllabus will have situational instead of grammatical labels.

In order to carry out the behavioural analysis that underlies the situational syllabus, we must have a set of parameters for describing the significant features of situations. These features include the physical context in which the language event occurs, the channel (spoken or written) of communication, whether the language activity is productive or receptive, the number and the character of the participants, the relationships between them and the field of activity within which the language event is taking place. Obviously, different syllabuses will result for different types of learner. The exact contents of a syllabus will be the result of a careful behavioural prediction and will consist of an inventory of language situations and a description of the linguistic content of each of these situations.

Situational courses do exist. They consist of learning units with labels like 'At the post office', 'Buying a theatre ticket', 'Asking the way' and so on. In all probability they are successful in what they set out to do, but there are reasons for doubting whether they can be taken as a model for the general organization of language teaching. The difficulty centres on just what is meant by 'situation'. With examples like the ones above there is no great difficulty. They are situations with fairly evident, objectively describable physical characteristics. The language interactions that are taking place are closely related to the situation itself. There will be grammatical and lexical forms that have a high probability of occurrence in these kinds of language event.

However it would be naive to think that the speaker is somehow linguistically at the mercy of the physical situation in which he finds himself. What the individual says is what he has chosen to say. It is a matter of his intentions and purposes. The fact that there are some situations in which certain intentions are regularly expressed, certain linguistic transactions regularly carried out, does not mean that this is typical of our language use. Even in the restricted physical situations that have been mentioned so far language does not have to be related to the situation. I may have gone into the post office, not to buy stamps, but to complain about the non-arrival of a parcel, to change some money so that I can make a telephone call or to ask a friend of mine who works behind the counter whether he wants to come to a football match on Saturday afternoon. Making complaints is not (or should not be!) what one typically goes to a post office for. The making of requests, the seeking of information, the expression of agreement and disagreement can take place in almost any situation. There are probably no situations where we typically express possibility, probability, certainty, doubt or conviction and yet the need to do so is demonstrated by the frequency with which they are expressed in our speech.

One way in which this problem might be overcome is by extending the notion of situation to include uses of language like those just mentioned which are the product of internal processes and not of the influence of situational features. Once we do this, however, we move into the realms of the unpredictable. The content of an utterance is determined by the state of mind of the speaker. That in turn is the product of his life's experience. We could predict his language behaviour only if we had complete knowledge of the universe. By broadening the concept of situation in this way we have rendered it virtually inoperable since we are no longer able to describe the features of a situation in objective terms. At the same time we have lost the benefit of the insight into language that is provided by our awareness that relationships between language and situation do exist.

It seems best, therefore, to retain the term situation for the sum of the observable and independently describable features of the context in which a language event occurs. Language use is then seen as a continuum. At one end of the scale the form and content of utterances is fairly predictable from a description of the situational context. At the other end the situational context of utterance is almost wholly irrelevant and prediction would be possible only if one knew what, in practice, one cannot know -- the learned and inherited characteristics of the participants.

Examples of language use under the control of observable stimuli are, if anything, atypical. A situational syllabus will be valuable insofar as a learner's need is to be able to handle language situations of this sort. The limited aims of a tourist, a waiter or a telephone switchboard operator might be provided for adequately in this way. However, they would, by definition, be unprepared for anything 'out of the ordinary'. If we were to attempt to use a situational syllabus for any learner whose needs could not be identified in these situational terms, including the general language learner, we would fail to provide him with the means to handle significant language needs. Useful as a situational syllabus may be in certain circumstances, therefore, it does not offer a general solution to problems of syllabus design (12).

4.3 Notional syllabuses

The discussion so far has suggested that there are limits to what can be achieved through grammatical and situational syllabuses. The grammatical syllabus seeks to teach the language by taking the learner progressively through the forms of the target language. The situational syllabus does so by recreating the situations in which native speakers use the language. While in neither case would it be denied that languages are learned for the purposes of communication, both leave the learner short of adequate communicative capacity. We have now, in effect, dealt with the existing situation in syllabus design and in doing so have provided a context against which the proposals for a notional syllabus, which are the major concern of the present book, can best be understood and judged.

The notional syllabus is in contrast with the other two because it takes the desired communicative capacity as the starting-point. In drawing up a notional syllabus, instead of asking how speakers of the language express themselves or when and where they use the language, we ask what it is they communicate through language. We are then able to organize language teaching in terms of the content rather than the form of the language. For this reason the resulting syllabus is called a notional syllabus (13).

The advantage of the notional syllabus is that it takes the communicative facts of language into account from the beginning without losing sight of grammatical and situational factors. It is potentially superior to the grammatical syllabus because it will produce a communicative competence and because its evident concern with the use of language will sustain the motivation of the learners. It is superior to the situational syllabus because it can ensure that the most important grammatical forms are included and because it can cover all kinds of language functions, not only those that typically occur in certain situations.

The process of deciding what to teach is based on consideration of what the learners should most usefully be able to communicate in the foreign language. When this is established, we can decide what are the most appropriate forms for each type of communication. The labelling for the learning units is now primarily semantic, although there is no reason why the structural realization should not also be indicated. A general language course will concern itself with those concepts and functions (14) that are likely to be of widest value. In the same way, in the provision of a course for a more specialized language learner, the limitation is on the types of content that he needs to express and not on the number of structures he needs to know or the situations in which he will find himself. In short, the linguistic content is planned according to the semantic demands of the learner.

Although, as we shall see in the final chapter, the criteria that are used in establishing a grammatical syllabus need not be wholly irrelevant in the creation of a notional syllabus, there is no reason to expect that what we identify as being semantically necessary to the learner will coincide with what is grammatically the most simple, the most regular or the easiest to learn. To put it another way, the forms that are needed to express the semantic needs will be extremely varied. Even if one could identify a 'simple' need, it is unlikely that there would be a 'simple' form that met it. The learning material derived from a. notional syllabus will, therefore, almost inevitably be linguistically heterogeneous. Although we will probably choose to isolate particular forms from this rich linguistic enviornment to ensure adequate learning of the grammatical system, there will be no ordered exposure to the grammar of the language. A notional syllabus is, therefore, to be classified as an example of an analytic approach to language teaching.

We can now see the position of the notional syllabus in relation to other forms of syllabus design. We have also seen what considerations motivate the proposal that we should approach syllabus construction from semantics. The remainder of this study is devoted to a more detailed examination of the concept of a notional syllabus, but, first, there is one major difficulty that has to be overcome. Whereas in the case of the grammatical syllabus a framework is readily available in the form of one of the many descriptions of the language and in the case of the situational syllabus much progress has been made in recent years in identifying the relevant situational features, when it comes to attempting to produce a notional syllabus, there is no available semantic (notional) framework in terms of which it can be prepared. The next chapter, which constitutes the major part of this study, is devoted to the exposition and exemplification of a framework which could be used in the setting- up of a notional syllabus. In the final chapter there is discussion of the way in which the framework might be used in producing a notional syllabus and of the language teaching situations in which a notional syllabus might most valuably be adopted. There is also consideration of some of the implications of a notional approach for other aspects of language teaching.
(1) The following works deal in some detail with vocabulary control in general and frequency in particular:

H.Bongers: The History and Principles of Vocabulary Control. Wocopi. Woerden. 1947.

Committee on Vocabulary Selection: Interim Report on Vocabulary Selection. London. King. 1936.

L.K.Engels: The fallacy of word-counts. IRAL 6/3 1968

C.C.Fries and A.A.Traver: English Word Lists. Ann Arbor. University of Michigan. 1950.

M.A.K.Halliday, A.Mclntosh and P.D.Strevens: The Linguistic Sciences and Language Teaching. London. Longman. 1964. Chapter 7.

A.S.Hornby: Vocabulary control -- history and principles. ELT 8/1 1953.

W.R.Lee: Grading. ELT 17/3, 17/4, 18/2. 1962/63.

W.F.Mackey: Language Teaching Analysis. London. Longman 1965.

(2) Committee on Vocabulary Selection: ibid.
(3) Institut Pédagogique National: Le Français Fondamental. (1er degré) Paris.

R.Michéa: Mots fréquents et mots disponibles. Les Langues Modernes. 47.

(4) J.C.Richards: A psycholinguistic measure of vocabulary selection. IRAL 8/2 1970.

(5) W.F.Mackey and J-.G. Savard: The indices of coverage: a new dimension in lexicometrics. IRAL 5/2 & 3. 1967.

I.A.Richards: Basic English and its Uses. London. Kegan Paul. 1943.

(6) D.A.Reibel: Language learning analysis. IRAL 7/4 1969.

(7) For a lengthier discussion of these terms see: M.A.K.Halliday, A.Mclntosh and P.D.Strevens: Op. Cit. pp. 207-212.

(8) The works by Lee, Mackey and the Institut National Pédagogique cited above also include discussion of criteria of grammatical selection and grading. Most books on methodology also contain some discussion, e.g.

J.A. Bright and G.P.McGregor: Teaching English as a Second Language. London. Longman. 1970.

C.C.Fries: Teaching and Learning English as a Foreign Language. Ann Arbor. University of Michigan. 1947.

R. Lado: Language Teaching. New York. McGraw-Hill. 1964. See also:

L.Dusková and V. Urbanová: A frequency count of English tenses with application to teaching English as a foreign language, in Prague Studies in Mathematical Linguistics. 2. Munich. Hueber. Prague. Academia. 1967.

ILK George: A verb-form frequency count. ELT 18/1 1963.

(9) See K.Bung: The Specification of Objectives in a Language Learning System for Adults. Strasbourg. Council of Europe. 197 3.

(10) H.G.Widdowson: The teaching of rhetoric to students of science and technology. In Science and Technology in a Second Language. London. Centre for Information on Language Teaching and Research. 1971. pp. 38-39.
(11) W.Labov: The Study of Non-standard English. National Council of Teachers of English. (U.S.A.). 1969. pp. 54-56.

(12) I should, perhaps, point out that the conceptual distinction between, say, a grammatical and a situational syllabus need not always be so clearly drawn in practice. A grammatical syllabus could be situationalized to the extent that a situational context (e.g. in the form of a dialogue) could be created to illustrate the grammatical structure being presented. A situational syllabus may similarly be grammaticalized by the deliberate exclusion or inclusion of grammatical structures in otherwise 'natural' materials. Neither of these processes can obscure the fundamental contrast in the underlying strategies.

(13) The term notional is borrowed from linguistics where grammars based on semantic criteria are commonly called notional grammars (cf. formal grammars where the criteria used in analysis are formal).

(14) The exact sense in which these terms are used here is explained on pp 23-24



Topic 2 and 3 of 12
GO BACK TO: Topic 2: Synthetic Syllabuses
GO BACK TO: Topic 3: Analytic Syllabuses
GO BACK TO: Topic 9: Selecting, Grading and Sequencing Content

[Home] [Topics Index] [WebBoard]
Copyright by University of Illinois at Urbana-Champaign, 1998