The Language Teacher

This article is not against contexts. Obviously, suitable contexts can help people to acquire new words, and there are good reasons for using contextualised learning a lot of the time. The idea I want to discuss here is that learning vocabulary from lists might not be such a bad thing either.

The main argument against the use of word lists is that they are an unnatural way of acquiring vocabulary items. This, of course, is true. Word lists ARE unnatural, but so are many of the other things that we do to teach foreign languages, and it seems unfair to single out word lists in this way.

My own view is that word lists have an important role to play in the acquisition of a new language, and that this role is particularly important at the beginning stages of learning a new language. The reason for this is quite simple. When you first start to learn a new language, the biggest problem you face is that you can't recognise any of the words. Nothing that you see or hear in the new language makes any sense at this stage, because all the words are unfamiliar.

There are two solutions to this problem. The first solution is for teachers to try and protect students from the consequences of their ignorance. Typically, we do this by teaching the students a very small number of words, and then restricting the sorts of texts that the students meet: graded readers, language text-books, and so on. Some people have actually put forward the view that this is the only real way to teach a new language, and that students should not be allowed to acquire a large vocabulary until they have mastered the basic grammatical system with a very small restricted vocabulary. Although this view is less common now than it was twenty years ago, in practice, most beginners courses restrict themselves to a basic vocabulary of two or three hundred words, and make a virtue out of this fact. The result is that students are able to operate reasonably well within a classroom context, where the lexical environment is very limited and very predictable. Outside this protected environment, however, they are often unable to cope.

It isn't obvious to me that establishing a protected environment is actually the best way of getting learners to operate effectively in the wider world. Instead, I believe there might be a case for teaching people very large vocabularies when they first start to acquire a language. There are two main reasons why I think that this kind of approach might be an appropriate one: one is a linguistic reason, the other a psychological one.

The linguistic justification is that a severely restricted target vocabulary, say 500 words or so for a beginners' course, doesn't really make much sense in terms of what we know about the lexical structure of languages. In all languages, a small number of words account for a very large proportion of the material which we see or hear on a regular basis. In general any corpus of language has a small number of words that occur many times, and a larger number of words that appear only once or twice. You can see this characteristic very easily if you take a short piece of writing, and count up the number of times each word appears in the text. Then, construct a graph by taking the most frequent word, and plotting the number of times it occurs in your text. Next, add to this total the number of times the next most frequent word occurs, and plot this new total on your graph. Then, do the same with the third most frequent word, and so on. The result should be a graph which looks something like Figure 1. This graph is a cumulative frequency count of the words in your text. What the figure shows is that a very small number of words account for a surprisingly large proportion of the text.

Instead of working with a small text, you can do similar counts using very large corpora, and if you do this, you still get the type of curve that appears in Figure 1. The actual figures for English suggest that a basic vocabulary of about 2,000 words accounts for about 80% of what we see or hear (the points marked by the arrows in Figure 1). Other languages produce slightly different shapes of curve, (for instance, some languages don't have articles, or auxiliary verbs, and this affects the number of high frequency words), but all languages produce the characteristic curve with a very steep slope followed by a much slower rise that is shown in Figure 1. Of course, the graph does not show that a person with a vocabulary of 2,000 words will understand 80% of what they see or hear in English! In most texts, the really important meanings are carried by the words that the learner is not likely to know. What the graph does show is that a person with a vocabulary of 2000 words is going to be able to recognise at least some of the words s/he hears. In contrast, the graph suggests that a person whose vocabulary is limited to only 500 words or so will meet a very large number of unfamiliar words in almost any common context. These unfamiliar words will be enough to prevent most learners at this level from understanding very much, except in very unusual circumstances. The obvious conclusion, from a linguistic point of view, is that a vocabulary of 500 words is pretty useless, while a vocabulary of 2,000 words goes a considerable way towards a realistic level of competence. The linguistic evidence, then, suggests that it might be sensible to teach beginners a very large vocabulary very quickly, and not restrict their lexical development to small vocabularies acquired over an extended period of time.

The second reason why it might make sense to teach people large vocabularies very early on in their learning career is a psychological one. Most learners have a rather naive understanding of what learning a language involves, but most people are sure that learning a language means learning lots of new words. So, in a sense, most learners expect to have to learn vocabulary, and it therefore makes a lot of sense to capitalise on these expectations. And unlike many aspects of language learning, vocabulary acquisition is a skill which it is easy for non-specialists to understand and easy for them to evaluate their own performance in. A score of 75% on a well constructed vocabulary test is very straightforward to interpret: (a) It means that you know only three quarters of the vocabulary that you ought to know, and (b) it is obvious what you have to do about it. A score of 75% on a grammar test, or a translation test is much more difficult for an untrained person to make sense of.

Most methodologies don't make anything of this interest in words. Instead, they set extremely low targets for vocabulary knowledge, and tend to play down the importance of learning words. Typically learners are not set vocabulary targets at all, and when they are, the number of words they are expected to pick up is very small. This seems to me to be a mistake, both on linguistic and on psychological grounds.

What is the alternative then? One possible alternative is to deliberately make the early stages of learning a language focus on the acquisition of vocabulary. Instead of allowing the basic vocabulary to be acquired slowly over a period of many years, it might be possible to teach it all, deliberately, and relatively quickly. This might sound like a radical suggestion, but there are several reasons why it is worth considering.

Firstly, young children learn their first language by acquiring single words in the first instance. They eventually get round to putting these words together into phrases and sentences, but this development takes a long time. Children don't start using two word utterances until they have a basic vocabulary of about 100 words. Longer utterances only come with a much bigger vocabulary, and even then, the syntax of these utterances is pretty peculiar. It takes several years before children start to use a sentence structure that clearly resembles that of a normal adult. If a single word stage is important for children learning their L1, then quite possibly a similar stage might be natural for L2 learners as well.

The second reason for building a vocabulary quickly is that a large vocabulary does in fact allow you to communicate with other people over a wide range of unpredictable situations. The communication that results might not be perfect, but at least the words are being put to some real use -- much like the imperfect sentences of children in fact. If you know the words, it's easy to learn how to use them properly. Corrective feedback from other speakers will eventually teach you how to formulate correctly what you want to say; if you don't know the words you need to use, then this natural process cannot even get started.

The third reason for emphasising vocabulary rather than grammar is that grammar is mainly about patterns. Patterns are much easier to recognise if you have a lot of data to work with, particularly if the patterns are statistical ones, rather than absolute regularities. It is very difficult to recognise a pattern if you only have a few instances to work with. If you only know 100 words, it's very difficult to discover the regularities that they show when you meet them in texts, especially when these texts have been specially constructed by a well-meaning text-book writer! When you know 2,000 words, it is very much easier to see the patterns in the way these words behave. Human brains seem to be particularly good at recognising patterns in complicated data. Restricting language input to a relatively small vocabulary hides the patterns, and makes it difficult for the brain to carry out its natural learning function.

So far, then, I have suggested that it might be possible to construct a language teaching methodology that focused initially on vocabulary acquisition, and aimed to build up a very large vocabulary very quickly. I've indicated some of the reasons why I think this might be an interesting approach. The question that arises now is what would an approach of this sort look like in practice?

The best bet so far seems to be a memory system called the key-word method, which is mentioned in the Nation interview and the Ellis article in this issue. Research using this method suggests that it is possible for adults to learn very large numbers of words in a relatively short space of time -- fifty words in an hour is not uncommon. At that rate, a learner would be able to learn the 2000 word target vocabulary in as little as forty hours. Of course, it would be very hard work, but most learners seem to expect to work hard in this way, and it ought to be possible to develop a reward schedule that made the learning fun as well. The main argument against the use of methods like the key-word method is that they encourage learners to think that words in the L1 are directly linked to words in the L2 in a one-to-one fashion. People who take this view think that this is misleading, and that it encourages the learners to remain over-dependent on translation. I think this is wrong. For me, the biggest problem in learning a foreign language vocabulary is not learning the exact meanings of the words. It's much more difficult to learn the actual shape of the words, to learn to recognise the words as physical objects, and to learn to distinguish them from similar forms reliably. Traditional vocabulary teaching has tended to ignore this problem completely. My own view is that if learners can recognise the word forms reliably, then they can be encouraged to work with authentic materials which use these forms frequently. Once you start to meet the familiar word forms in natural contexts, then it will soon become obvious how they relate to each other, and how they behave in sentences.

To me, this suggests that two types of learning activities ought to accompany this initial vocabulary learning stage. In the first activity, students might be provided with authentic texts, and simply be asked to mark any word forms that they recognise. These texts could be either written or spoken texts. The object of the exercise is not for the students to understand the text, but simply for them to recognise any of the words they already know. Given time and an increasing vocabulary, comprehension should become automatic. In the meantime, exercises of this sort should train the students ears and eyes to recognise unfamiliar forms quickly and reliably. The second type of activity involves word games of different sorts, specifically designed to get the students using their new vocabulary. Word games do not provide the naturalistic, communicative contexts that language teachers usually think of when they are trying to provide contexts for using an L2. But, in fact, artificial contexts of this sort provide a very good environment for using words. Again, it's significant that word puzzles are incredibly popular with L1 speakers, and it is surprising that language teachers have not exploited this popularity more.

Conclusion

The idea I have explored in this paper is that there might be a case for teaching languages starting from vocabulary. I have suggested that it might be possible to teach beginners much bigger vocabularies than we usually aim for, and that it might be possible to achieve these vocabulary targets very quickly if we concentrated on words. I have explained the linguistic and the psychological reasons why I think this kind of approach might be worth trying. One word of caution seems in order, however. I use this vocabulary-based approach in my own language learning whenever I have to learn something of a new language for a trip to a foreign country, and as far as I can see, it is very successful. I have never used this word-based method in class with other learners, though. This is partly because school syllabuses have made it impossible to experiment in this way, and partly because it has just seemed too radical. Needless to say, if any reader of The Language Teacher feels brave enough to give it a try, I would be very happy to hear from you.

Paul Meara is Director of the Wales Applied Language Research Unit at Swansea University of Wales, and Chair of the British Association for Applied Linguistics. His three bibliographical volumes on second language vocabulary and lexis have become the standard reference works in this field. He can be contacted at Centre for Applied Linguistics, University College, Swansea SA2 8PP, United Kindom.