IH Journal of Education and Development

IH Journal Issues

open all | close all


open all | close all

IH Journal Issues:

Published in:

The Language Syllabus: Why Not Start With Lexis? by Dave Willis

This is the first of two articles which will look at the content of language teaching – what is to be taught – and how that content should be organised. In this article I will argue for the importance of lexis as a starting point, and will emphasise the importance of the most frequent words and phrases in the language.

Some years ago I asked a class of secondary school teachers on an in-service course to list the ten most frequent mistakes their learners made. Without letting them all know what was happening I asked some groups to list mistakes made in the first year and the other groups to list mistakes in the fourth year. When they had agreed on consolidated lists for each of the two years the teachers were dismayed to see that seven of the errors appeared in both lists: things like marking questions lexically but not structurally (Where you live?/ Where do you live?); omission of articles (My mother is teacher); failure to mark past tense (Yesterday I go to cinema). They all agreed that they had taught these items conscientiously, but that many learners simply failed to take them on board when using the language.

The teachers were, however, rightly and understandably indignant when I suggested, tongue in cheek, that students had learned very little in four years. They said that their fourth year students had a much wider vocabulary, and that they could speak with a greater degree of fluency. They also pointed out that fourth year learners were more likely to produce these forms accurately when they had time to think about things – when doing a grammar test for example. The problems occurred when learners were trying to put the language to use. Problems like this are all too common. I am sure many teachers all over the world have had similar experiences.

We do not know how learners develop the target language. But we do know that it is not a straightforward process. It is not a matter simply of learning the grammar and then applying that knowledge. As teachers we are very familiar, for example, of the situation in which learners appear to know something and yet not know it. Given time to work things out they can produce do-questions, for example, but when they are using the language spontaneously they fail to apply this knowledge and produce questions like Where you live? rather than Where do you live?


But most classroom practice is based on the assumption that learners acquire grammatical insights of an abstract kind and then go on to apply those insights. Lessons focus on grammatical abstractions like past and present tenses, simple and continuous aspect, the definite and indefinite articles. Learners apply this abstract knowledge to create ‘structures’, which are then filled out with appropriate vocabulary. There are two problems with this approach. The first is that some of these concepts (simple/continuous; definite/indefinite) are too complex and abstract for learners to assimilate in the early stages, unless they have close counterparts in their first language. Secondly, all of these grammatical features are, initially, redundant.  Learners can often get their meanings across without needing to use the newly taught structure. This does not mean that the right grammar is not important in the long run: simply that learners will often ignore it in the short term when they are using the target language to express what they want to mean.

This means that an approach which focuses primarily on grammatical abstractions is unlikely to produce positive results in the short term, or to sustain motivation. Learners find that they are unable to apply what they have ‘learned’, which is discouraging for them. The more precisely and exclusively teaching focuses on these abstractions, the greater the sense of failure.

Let us consider another possibility for the way people learn a new language. Let us assume that learners begin by stringing together lexical items and that they gradually develop a more complex grammar as they attempt to realise more complex meanings and to realise meanings more precisely. This certainly seems to reflect what happens in many learning situations. As soon as learners are expected to use the language they resort to making themselves understood by stringing words together with little in the way of grammatical marking – even though the grammatical framework has been very much the focus of teaching.

So from the point of view of learning it would seem useful to focus on building vocabulary and allowing time for the grammar to develop. But there are problems from the teaching point of view. Grammar is a relatively closed system. We can itemise the tenses and determiners in English. We can list categories like singular and plural and countable and uncountable nouns. So in this way we can put together a reasonably definitive grammatical syllabus. Vocabulary, on the other hand, is an open system – fluent speakers of a language have a vocabulary of around 50,000 words.  How are we to decide what to teach – where to begin?

Let us look at some figures. Although fluent users have thousands of words, some words are very much more important than others. In fact the 700 most frequent words in the language account for around 70% of text – of all the language we hear or read and speak or write. Clearly these 700 words are of central importance. Learners need to be introduced to them as soon as possible. The next 800 words on the frequency list account for another 10% of text, so a vocabulary of 1500 words brings us to 80% coverage. 2500 words bring 86% coverage, so the next 1000 words cover only 6%. There is a clear pattern here. It emphasises the importance of the most frequent words in the language. We can readily identify the most frequent words in the language. If we focus learning on these items we can provide a syllabus which will give learners a solid foundation.

But words on their own are not enough. Widdowson (1989) has this to say:

…communicative competence is not a matter of knowing rules for the composition of sentences…  It is much more a matter of knowing a stock of partially pre-assembled patterns, formulaic frameworks, and a kit of rules, so to speak, and being able to apply the rules to make whatever adjustments are necessary according to contextual demands.  Communicative competence in this view is essentially a matter of adaptation, and rules are not generative but regulative and subservient. (Widdowson 1989: 135)

In other words communicative competence – the ability to use a language – depends on having a large number of phrases at our disposal. We do not simply string words together, we string together words and prefabricated phrases. Think, for example, of the phrase as a matter of fact. We carry this in our heads as a single item. We do not have to compose it from its constituent words any more than we compose the word unhappiness from the constituents un, happy and ness. If we look closely at Widdowson’s words they provide a perfect illustration of what he is saying:

…(communicative competence)  (is not a matter of) knowing (rules for) (the composition of) sentences…  (It is much more a matter of) knowing (a stock of) partially pre-assembled patterns, formulaic frameworks, and (a kit of) rules, (so to speak), and (being able to) (apply the rules) to make (whatever adjustments are necessary) (according to) contextual demands. (Communicative competence) (in this view)(is essentially a matter of )adaptation, and rules are not generative but regulative and subservient.

Here I have attempted to show the fixed phrases using italics and brackets. Well over half of the paragraph is made up of fixed phrases.

The two most frequent words in the language are the and of. Of course these words are not much use on their own. They are frequent because they feature in a large number of very frequent phrases: the beginning of; the end of; the middle of; the front of; the back of … Here we have a pattern, the … of, which combines with words denoting a part (beginning, middle, end etc.) to produce phrases. The same applies to the words a and of, which offer the pattern a … of. This can be completed by, for example, words denoting quantities or containers: a lot of; a few of; a load of; a box of; a tin of etc.


So the most frequent few words in the language are valuable not for themselves, but for the patterns and phrases in which they occur. We recognise this by referring to words like the and of as structure words, as opposed to content words like middle; end; box; load and so on.

Let us go on to take a look at one of the most frequent content words in the language, the noun way. Why is this word one of the most frequent nouns in the language? First it is polysemous – it has a number of different meanings. In the case of way these meanings are central to everyday concerns:

    1. Method, means: It’s a good way to meet people.
    2. Manner, style: He smiled in a funny way.
    3. Direction: Is this the way to St. Pancras?
    4. Distance, extent: It’s a long way to Tipperary.
    5. Time: Christmas is a long way off.

Again we should note that way features in a large number of patterns and syntactic frames such as:

  1. Noun Phrase + to + infinitive: The cheapest way is to hire a van.
  2. Noun Phrase + of + –ing: The different ways of cooking fish.
  3. Noun Phrase + relative clause: The only way a telegram can be delivered


We can move from here to the notion of pattern grammar (Hunston and Francis 1999). The pattern Noun Phrase + of + -ing can be completed with noun phrases like the best way; one way; another way and so on. But it can also be completed by a number of other nouns:

METHOD: way; method; means; process

LIKING/DISLIKING: love; hope; hatred; fear; horror

POSSIBILITY: likelihood; chance; possibility; risk; danger

OUTCOME: result; effect; aim; intention; advantage; disadvantage

The important insight of pattern grammar is that patternings are not random. They relate to groups of words which are identifiable under a number of semantic headings as shown above.

So the word way offers input to a language syllabus in three ways:

  • as a word which expresses a number of central concepts
  • as a component in a large number of common phrases (on the way; in the way; in a way; the .. way to; one way of… )
  • as an exemplar of valuable syntactic frames

Some words are frequent because they collocate with a large number of other words. Look at the word strong for example. In English we talk about: a strong personality; a strong leader; strong wind/current; a strong supporter; strong coffee; strong language; a strong case ; strong evidence; a strong possibility; a strong team; a strong currency; a strong taste.

Some words have a large number of metaphorical meanings. Verbs of motion provide a good example. In English life is a journey: We go through life, reach the age of …, reach adolescence; get to seventy… And discourse is a journey: I’ll come to that soon; We have already covered that; Let’s go back over that… So words like come and go are used not only with their literal meaning of physical motion, they also have a range of metaphorical uses.

An awareness of these features of language and language use puts us in a position to think about a syllabus which is based primarily on lexis and on patterns and fixed phrases. We can argue that a language syllabus should be based on the most frequent words in the language and on their various meanings and patterns of use (see Sinclair and Renouf 1989; Willis 1990). We can also argue, as I have done above, that the accumulation of lexical items, including fixed phrases, is more consistent with natural language development than a syllabus based on grammatical abstractions. But there are, of course, problems with the notion of a syllabus based on lexis. The most frequent words in the language may be readily identifiable, but they are also highly complex. They are involved in a huge number of patterns and phrases. And it is all very well to talk about the 70% of text that is covered by the most frequent 700 words, but that leaves another 30%. It has been estimated that 3000 words are needed to provide a basic competence in the language. How do we move towards that figure of 3000? And what about basic grammar? How do we provide coverage of, for example, verb tenses and the articles?

Learners need to encounter frequent words and phrases in contexts which help understanding and recall. And in working hard to cover the most frequent words of the language we must make sure that we help learners to go beyond that to provide them with a wide range of topic vocabulary. In the next issue I will look at lesson planning and at the notion of a pedagogic corpus to seek answers to these questions.

Author’s Bio:
Dave Willis has been in ELT for more than forty years. Apart from the UK he has worked in Ghana, Cyprus, Iran and Singapore, including twenty years as a British Council officer. His last job was at the Centre for English Language Studies at Birmingham University, where he worked until 2000, mainly on MA TEFL/TESOL and Applied Linguistics programmes. He is now an Honorary Senior Research Fellow.
His last book, Doing Task-based Teaching was written with his wife, Jane, and published by OUP in 2007. His main area of interest is language description and the place of language study within a task-based approach – the subject of his book, Rules, Pattern and Words (CUP 2003). For more information please visit: www.willis-elt.co.uk

Similar Articles: