Contents 

1. Introduction 

2. Analysis of the Syllabus

i) Selection of Structures and Functions 


ii) Sequencing of Structures and Functions 


iii) Integration of Structures and Functions


iv) Topics, Situations and Lexis

3. Conclusion: Evaluation of the Suitability of this Syllabus for its Target Constituency

4.References 

5.Appendices 

i) Contents of G 
   Contents of D 

ii) Grammatical Summary of G
    Grammatical Summary of D

iii) Wordlist for Each Lesson of G
     Wordlist for Each Lesson of D

iv) Alphabetical Wordlist for G
     Alphabetical Wordlist for D

1.  Introduction: Description of the learners and the course objectives. 

The situation which we set out to analyse here is that of young children, typically about eleven years old, beginning a course in general English. While for these learners, as for any others, there are many factors in the domains of affect and cognitive style which should be taken into account in the lanning and implementation of a syllabus, these factors are mainly subjective needs, or wants, rather than objective needs. So, although we can discover useful information about the kinds of topics or tasks that are likely to be more motivating for each particular learner, perhaps through interviews or questionnaires, an objective needs analysis is largely irrelevant. 

Consequently, the task of syllabus design is often replaced, at least to a certain extent, by that of coursebook selection. Despite the administration of tests designed to monitor variously progress, attainment or proficiency, detailed objectives may not be specified. In such a situation there is no great pressure for a 'high surrender value', no a proiri reason to favour either competence or performance ( Widdowson 990 : 132 ). Unconcerned by such dilemmas as whether our students will need to open a bank account before they register with a doctor or vice versa, we are free to design the syllabus or coursebook according to purely linguistic criteria.

Despite interesting work in recent years on the lexical syllabus, the majority opinion would probably still agree with Johnson 
( 1982 : 94 ) that in the early phases of learning the organisation of the syllabus should be structural. Dubin & Olshtain (1986 : 38) expand on this idea of different principles of syllabus design at different levels: 'Thus it may be necessary to use a structural/situational syllabus for the first years of the course of study, moving to a functional plan of organisation, followed by a notional/skill combination, leading finally to a fully communicative design for the final phases of the course.'

However, there is a danger that an overemphasis on structure at the expense of the development of language use may not allow the learners to develop their communicative competence. This problem is addressed in the design of Generation 2000 1 in three ways. Firstly, while retaining the structural syllabus as the fundamental principle of organisation, this is integrated with a functional syllabus. Secondly, in additional to the general lessons which conform to that plan, there are other lessons, such as the skills lessons, which are outside the structural/functional syllabus ( see note on following page ). Finally, the use of communicative tasks rather than structural exercises shows that a structural syllabus may be implemented through a communicative methodology.

2. Analysis of the Syllabus

i) Selection of structures and functions 

It has been claimed (e.g. Dubin & Olshtain 1986 : 47) that the selection and ordering of linguistic items in coursebooks for beginners has no theoretical justification. To consider the possibility of founding our choices on theoretical principle we will turn later to the findings of SLA research, but in practical terms sufficient justification may be found in the fact that the patterns introduced first are not only very common in use, but also highly generative.

While analyses of corpora and concordancers have produced some interesting data in recent years in the field of lexis, there is 
no reason to doubt to any great extent the observations and intuitions of coursebook writers as to what are the most common 
structures and functions in the language. In fact the great similarity in the choice of structures to include in books at this level argues for the correctness of the choice. As a first benchmark for comparative analysis with G, we may examine the list of structural elements covered in CE, following Nunan's (1988: 144) suggestion. 

CE is much shorter than G,and so it is natural that G should go on to cover much more material, but that aside, the similarities 
between the two syllabuses are much more salient than the differences. Particularly noteworthy is the fact that all the items covered in units one and two of CE are covered in the first -4-three lessons of G, with the exception of two prepositions which are in lesson six. This correspondence continues substantially throughout the course : the great majority of structures covered are covered in both books, and may be considered to comprise a common core. 

One of the more significant omissions in G as compared with CE is that the stative verbs are not covered. The structure 'Be+N+Adj' ( CE3 ) and 'What...like?' ( CE4 ) are not in G. Nor is 'How much? ( price )' ( CE8 ), but 'How much? ( quantity ) and 'How many?' ( CE9 ) are in G11, and 'How often?', also CE9, is in G37. Other items cited in the contents for CE and not included in G are : 'so', 'then', 'each', 'another', 'other', and some prepositions.

On the other hand, G is generally much richer than CE. For example, both courses introduce simple questions of polar and wh- 
type in lesson one, but G also has 'How old' and 'What colour', which are treated in unit three of CE. Thus we can note two points about the syllabus of G. Firstly, there is an overwhelming correspondence between the choice of structures in G and those in CE, suggesting that there are good practical and intuitive, if not theoretical reasons for these choices. Of all the ninety-seven items listed in the contents of CE, only seventeen are not also covered in G, half of which are prepositions. Secondly, the major point of difference lies not in the choice of what to include, but how much.

In our examination of coverage of functions, we used for comparison the syllabus of Discoveries 1, which, like Generation 2000 1, is based on a structural/functional syllabus, and is designed for young beginners. As in the comparison with CE for structures, so in the comparison with D for structures and functions, we note that G covers a lot more ground but also has a greater amount of input even if we restrict the comparison to the section of G that the other courses overlap, approximately the first third. So, while several major functions and structures present in G are not covered in D, because, for example it mentions neither past nor future, of the functions introduced up to lesson seventeen of G, virtually all are present also in D, give or take some quibbling about the precise definitions of certain functions. Those which come later in G are covered in subsequent levels of Discoveries. 

In the first few lessons, both books introduce useful functions concerning greeting, identifying, introducing, asking for and giving personal information, and talking about relationships. Other than that, there is no great similarity between the ordering of functions throughout the syllabus, an aspect of the syllabus design we will return to in the following section. As we noted earlier, there is no need for such a course to have a high surrender value, and so no reason why any particular functions should be given priority beyond these first few which are of obvious utility. From there on the principle on which the lessons are ordered is that of the structural sequence. 

One useful result of the comparison of functions in G with those in D is that it clarifies the idea that a distinction may be made between core functions such as 'identifying', 'warning', 'making suggestions', which would be associated with one or more specific 
exponents taught as structures, and less essential ones such as 'asking and saying how old people are', which are very limited 
both in the range of possible linguistic realisations and so also in the potential for generating other functions or meanings from 
similar structures.

Of those functions covered in D and not G, only a few might be considered serious omissions:

Apologise
Accept apologies
Give opinions
Express approval
Complain
Criticise.

Many other items listed in the 'language use' section of the contents for D (see appendix) are not really functions at all, but situational or topic-based categories of language use, or notions such as 'time'. Taking this into account, it is clear that G covers a much larger number of functions than D. This is not due simply to delicacy of category descriptions but reflects the much richer linguistic input. Conversely, it is characteristic of D that certain basic functions such as 'identifying', 'describing', 'talking about likes' are often repeated. This repetition does not constitute recycling in the sense of a spiral syllabus design as the same functions are 
repeated in consecutive lessons. For example, 'location' is the subject of lessons 22, 23 and 24. In terms of structure, three lessons are required to cover 'Is there?', 'Yes,there is', 'No,there isn't', 'Are there?', 'There aren't any'. Such repetition, designed to create a slower pace and so aid consolidation for young learners, inevitably has the effect of restricting the amount and variety of input that they receive, which must be one of our main concerns.

In contrast, of the forty functions covered in G:

30 occur in 1 lesson each
6 occur in 2 lessons each 
2 occur in 3 lessons each
2 occur in 4 lessons each. 

The functions repeated four times are the most general, describing' and 'talking about the past'.

The first lesson on 'describing' (7) is the lesson where adjectives are first focused on, and the structure 'There is' is used. In the second (11) 'There are' is used, and 'There is' is now used for uncountables'. The third (22) introduces 'There was', and the fourth (41) also uses 'There is'. 

'Talking about the past' is realised in lessons :

23 : was
24 : simple past - other verbs
26 : both of the above 
43 : both of the above 

The functions treated three times are 'talking about location' and 'making requests'.

'Talking about location' is realised in lessons : 

1 : Where is in, at
6 : Where is in, on, under, near, in front of, behind
12 : Where is

'Making requests' is realised in lessons :

3 : Imperative 
17 : Can I...?, Can you...?
34 : Can I...?, Can you...?

The functions which occur twice are :

Giving instructions
Talking about ability
Talking about likes
Talking about necessity or obligation
Talking about purpose
Talking about reasons

While it may be noted that for most of these functions, the exponents used to realise them are virtually the same each time, there are important exceptions to this, where the linguistic range is extended. An example of this is the function 'describing'. We should also note the fact that those functions that are selected for repetition are recycled at intervals rather than being drawn out over consecutive lessons. This point is partially contradicted by the treatment of 'talking about the past', but it is natural that such a major structural item as the past tense should require more than one lesson, and note how it is recycled in lesson 43. More characteristic is the treatment here of 'location', which may be contrasted with that noted above for D. Another important aspect of the syllabus design is the ordering of the structures and functions, and it is to this that we turn now. 

ii) Sequencing of structures and functions

In the previous section we noted that the structures and functions covered in D are roughly the same as those in the first third of G. For the main structural items, the sequencing is also similar:

But while this degree of correspondence seems to imply some principle of sequencing, it is not clear if any principle is n fact being used. Within lessons in G, there are certain obvious functional links. For example, lesson 21 has 'Making suggestions', and also 'Agreeing and disagreeing with suggestions'. So functional requirements override the structural, in order to include the two parts of the adjacency pair. But such functional links are not developed between lessons. In order to exploit sequencing potential, we might have 'Making suggestions' in a cluster of lessons which would also include 'Discussing future plans', 'Inviting', and 'arranging to 
meet' (cf. Johnson 1982: 72), but this potential is not realised in G. 

Important insights into sequencing may also come from SLA research, although there are radical differences between naturalistic acquisition and classroom based learning which are likely to limit the applicability of SLA theory to the practice of language teaching and learning (White 1988 : 82). A particular problem for a content syllabus is that, as Prabhu points out, 'it assumes that the development of the internal system is a discrete item, additive process - an assumption which goes counter to the highly plausible perception in interlanguage studies that the process is a holistic one, consisting of a sequence of transitional systems'(1987: 73). This difficulty may be overcome by a process-based approach to syllabus design, which recognises 
that as the concept of level is non-linear, so too should the syllabus be non-linear (cf. Lewis 1993: 43-48, Nunan 1988: 34). 

Alternatively, at least at the lower levels, interlanguage development might be accommodated by recycling structures and 
functions in a spiral. The need for recycling is summarised by McLaughlin (1987: 143): 

'Lightbown (1985) pointed out that second-language acquisition is not simply linear and cumulative, but is characterised by 
backsliding and loss of forms that seemingly were mastered. She attributed this decline in performance to a process whereby 
learners have mastered some form and then encounter new ones that cause a restructuring of the whole system.'

While it has been shown that formal instruction has only a marginal effect on acquisition order for language in use (Ellis 1985: 218- 224, Larsen-Freeman & Long 1991: 299 -312) positive effects are noted for the rate of acquisition and possibly the ultimate level of attainment. Furthermore, even if structures are taught at a time when a learner is not ready to incorporate them into his IL, they will have an immediate effect on performance, and are likely to have a subsequent effect on competence (Ellis 1990: 170).

So while SLA research may eventually provide valuable guidance for syllabus designers, studies of IL do not provide the basis 
for syllabus sequencing. As Brown (1994: 69) cautions, 'The fallacies of many statements [comparing first and second language acquisition] should be more clearly perceived, with carefully integrated considerations of the several types and domains of comparison and some of the issues and implications of first language acquisition.' 

While we agree with Corder (1979; cited in Widdowson 1990: 139) that 'efficient language teaching must work with, rather than 
against natural processes', fifteen years on it is still not clear what is the best way to achieve that. While it would seem to implicate a structural sequencing of syllabus content to conform to the natural order of acquisition, even if this were possible in view of our limited knowledge of SLA, and in the light of Ellis's distinction between order and sequence (Widdowson 1990 : 141), such a view is not uncontroversial.

'There are the general arguments against grammatical grading of content (whether this grading is based on traditional criteria or 
more recent criteria stemming from SLA research) on the grounds that grammatical grading distorts the language available to the 
learner. It could well interfere with language acquisition which is more a global than a linear process, different aspects of grammar developing simultaneously rather than one structure being mastered at a time.' (Nunan 1988: 34).

This objection may be overcome if we can implement a structural/ functional syllabus through a communicative methodology, which would not artificially restrict input and undermine authenticity. Appropriate methodology would also overcome another of Nunan's objections (op. cit.: 28), to the structural syllabus, that grammatical complexity does not necessarily equate with learning 
difficulty. In a structural syllabus, just as in a lexical one, grading for difficulty may be achieved 'through the complexity of the linguistic task, rather than through some arbitrary restriction on the lexical and syntactic content' (Renouf 1987: 178).

Willis (1990: 5) notes that 'There is a tension, perhaps a basic contradiction between a grammatical or structural syllabus and a 
communicative methodology. A grammatical syllabus demands a methodology which focuses on the correct production of target forms. It is form-focused. A communicative methodology, if it involves real communication, demands that learners use whatever language best achieves the desired outcome of the communicative activity.'

However, the essence of this problem lies in the assumption that a grammatical syllabus must be implemented through a 'present and practice' methodology. Although the underlying principle of organisation in G is structural, that is not to say that the structures are the main focus in the lessons. Even in the general lessons, communicative tasks such as information gap and problem solving activities are more prevalent than exercises which simply practise usage, while the other lessons are outside the framework of the structural syllabus. Thus, as Willis himself notes (op. cit.: 6-7), a possible solution is to take an eclectic approach. 

iii) Integration of structures and functions

While the grammar in each lesson is generally graded according to what is considered to be simpler, the notion of simplicity does not always apply; for example, there is no reason to think that the modal 'can' is any more or less simple than any other modal. In such cases we would apply other criteria, one of which would be the utility of the functions associated with particular structures. However, just as there are many structures which are equally simple, there are many functions which may make competing claims for priority. As Nunan (1988: 85) suggests, while the lack of any direct one-to-one relationship between linguistic functions, notions, and grammatical forms 'leads to a certain amount of arbitrary decision making about which forms to introduce with which functions, some form/function relationships naturally suggest themselves, particularly at lower proficiency levels'.

While it may seem unnecessarily restrictive, one-to-one mapping of form on function is justified by SLA research: '...initially, at least, learners attempt to maintain a relationship between one invariant surface linguistic form and a single function, Andersen's (1984c) "one-to-one principle". They are motivated to do so to keep their IL system internally consistent.' (Larsen-Freeman & Long 1991: 68) So even if, as in the case of the modals, we choose a particular structure according to criteria of utility as well as simplicity, we may still be faced with the dilemma of choosing which of its functions to prioritise.

Thus, in CE, 'can' is introduced in unit 3 to express permission, a function that does not occur in G. 'Can' expressing ability is resented in unit 6 of CE, lesson 4 of G, while 'can' for possibility is in lesson 21 of G. So the fact that there is an overlap between the structures taught does not necessarily imply that the same functions are taught.

As we do not meet the two functions of 'can' in a single lesson, neither do we find together various ways of expressing ability or 
possibility. At higher levels this principle might be considered to be unduly restrictive and we may note that it is note maintained in the higher levels of Generation 2000. For example, lesson 23 of book two has three different exponents for the function 'making suggestions'. 

Irrespective of the choice of function to illustrate use of a structure, or the choice of structure given to realise a function, 
'A more constructive approach to describing language in structural or functional terms is to see the two as complementary, with each supporting and enriching the other. The relationship between the two can best be expressed in the form of this simple equation : 

Structure + Context = Function

(Hutchinson & Waters 1987 : 32).

The function should thus be regarded as a pragmatic phenomenon (Lewis 1993: 4) in which the context is essential to the meaning 
of the utterance, in contrast to the traditional structural syllabus, where context was secondary or, in grammar-translation methods, even non-existent. The corollary of context-dependency is that the context can help learners to derive meaning from utterances which would be incomprehensible if presented as decontextualised sentences.

Contextualisation makes it easier to decode functions, allowing the possibility of introducing useful functions at an early stage 
even if their exponents are grammatically complex (cf. White 1988: 79). Although there is no pressure in G to introduce early 
functions with a high surrender value, the value of not excluding whatever fails to meet criteria of structural simplicity lies in the naturalness of the language. The fewer criteria that work to restrict the selection of linguistic items, the more potential to foster natural language use. So along with the structures and functions in the syllabus, we must also consider contextualisation; the topics and situations in which the structures and functions are met and the lexis that arises within them.

iv) Topics, situations and lexis

Although the functions which are matched to each structure determine the orientation of each of the general lessons, there must still be contextualisation in terms of topics and situations (White 1988 : 73). Johnson (1982 : 92) regards as a major advance in syllabus design the flexibility to combine various kinds of lessons in a multi-dimensional syllabus. Although a multi-dimensional syllabus may lead to rigidity (Prabhu 1987: 92), G overcomes this potential problem. On the one hand, the general lessons are contextualised in the traditional way, by selecting topics and situations which are appropriate to the functions being used in those lessons, and also appropriate for the kind of learners the course is designed for (Dubin & Olshtain 1986: 45). On the other hand, apart from the general lessons, there are also the skills lessons, reading lessons, and project lessons, which are outside the structural/functional organisation of the course.

One lesson in every ten is devoted to skills development, practising reading, writing, listening and speaking through connected activities designed to integrate the four skills. Additionally, another lesson in each block of ten is devoted to reading, reflecting the emphasis on receptive skills in the communicative approach. This is a particularly rich source of input as the reading texts (rather oddly described by the publishers as 'semi-authentic') are much longer than the texts which would be used in presentations in traditional lessons. They are not selected to present any particular structures or functions, but for comprehension, which is aided by activities to encourage prediction and active comprehension.

Although lexis is not a basic principle in the design of the syllabus for this course, its importance is clearly recognised. Willis (1990 :9) points out that '...it is the reception of lexis which will form the most important activity in the earliest stages...This seems self-
evident, but it was not many years ago that the vocabulary content of courses was deliberately restricted until students had developed a certain mastery of basic structures.' 

As the methodology implied by G does not implicate such a rigid sequence of presentations of structures, there is no such constraint on the amount of lexis which students encounter. As each function occurs in the context of a particular topic or 
situation, the lexis associated with these will naturally arise. At the stage of topic selection, Dubin & Olshtain (1986: 11) say 
'...lexical decisions must override other considerations, giving lexis the proper emphasis and suitable focus it richly deserves within the thematic unit.'

However, 'focus' implies that certain words should be singled out for special attention, which in general terms is not the case in 
G. While G does satisfy the need for rich lexical input, with a total of some 1500 words, this is achieved at the expense of a clear indication of what is to be learnt. For the fact is that not all words are equally important. As Willis points out (1990:vi), the 700 most frequent words of English account for around 70% of all English text, and so it should be possible to identify the most important words and make sure that they are learnt. As not all input will become intake (Lewis 1993 : 27), this entails recycling key lexical items, which is not done in G, the great majority of words occurring just once.

The problem is that, while there is a natural correlation between certain functions and the structures exemplified in their exponents, the extent of such a correlation with lexical words is very limited. Despite Willis's point (1990: vi, 38) that the commonest words in the language occur with the commonest patterns, Renouf, commenting on the COBUILD project (1987: 177), notes that 'In the course of our analysis, we became aware of facts about he language which had not been available to us intuitively. One of these was the fact that words thought to be lexical were less likely to be so the more frequent they were in the text.' 

So while G gives plenty of opportunities for the learning of the most common structural words, it is not certain that the rich lexical input will actually lead to uptake, and it might be preferable to have a reduced number of items with more recycling.

3.Conclusion : Evaluation of the suitability of this syllabus for its target constituency.

Generation 2000 1 is well suited to the typical Greek beginner. The use of a structural syllabus reassures children who want the learning goal of each lesson to be specified and while Johnson (1982 : 63) says that a semantic syllabus may be more attractive than a structural one, many learners appreciate the specification of discreet grammatical items to be mastered. It may be argued that this fosters a false impression of the process of language acquisition, but that is compensated for by the lessons which fall outside the structural/functional framework of the course, and by the large quantity of roughly tuned input.

The principle behind the design of the course is that the classroom must be a language-rich environment, and the effect of this works in two opposing senses. On the one hand, it means that certain things do not need to be presented. If, for the sake of naturalness and authenticity, a lesson contains the names of certain days but not others, it may reasonably be assumed that 
the others will arise in the course of communicative interaction during the lesson. As the course implicates a communicative 
methodology, it would be perverse to expect that such items should be 'presented', and to criticise 'omissions'.

On the other hand, there is an enormous amount of lexis in each lesson, and in contrast to what we said above about indulging the 
learners' desire to master whatever they meet, it is not possible for them to learn in one lesson a hundred words, most of which 
are not recycled. This is not a defect in the course design as such, but it is the point at which we experience a clash between 
the communicative principle of the course design, and the conservative Greek learner.


References 

Abbs, B. & Freebairn, I. 1986. Discoveries Student's Book 1.

Harlow: Longman.


Brown, H. D. 1987,1994. Principles of Language Learning and Teaching.

Englewood Cliffs, N.J.: Prentice Hall.


Dubin, F.& Olshtain, E. 1986.Course Design.

Cambridge: Cambridge University Press.


Ellis, R. 1990. Instructed Second Language Acquisition. 

Oxford: Basil Blackwell.


Granger, C.& Beaumont, D. 1994. Generation 2000 Student's Book 1.

Oxford: Heinemann.


Hutchinson, T.& Waters, A. 1987.English for Specific Purposes.

Cambridge: Cambridge University Press.


Johnson, K. 1982.Communicative Syllabus Design and Methodology.

Oxford: Pergamon Press.


Larsen-Freeman, D.& Long, M. 1991. An Introduction to Second Language Acquisition Research. 

Harlow: Longman.


Lewis, M. 1993. The Lexical Approach.

Hove: Language Teaching Publications.


McLaughlin, B. 1987. Theories of Second Language Learning.

London: Arnold.

Nunan, D. 1988. Syllabus Design.

Oxford: Oxford University Press.


Prabhu, N.S. 1987. Second Language Pedagogy.

Oxford: Oxford University Press.


Renouf, A. Moving On. In Sinclair, J. M. (ed) 1987. Looking Up.

London: Collins.


Rossner, R., P.Shaw, J.Shepherd, J.Taylor, and P.Davies. 1979

Contemporary English, Pupil's Book 1.

London: Macmillan.


White, R.V. 1988. The ELT Curriculum.

Oxford: Blackwell.


Widdowson, H. G. 1990.Aspects of Language Teaching.

Oxford: Oxford University Press.


Willis, D. 1990. The Lexical Syllabus.

London: Collins.