Getting The Balance Right

Paul Meara and Nick Ham

In 1975, Hunter Diack published a book of vocabulary tests called Test Your Own Word Power. The book was aimed at native speakers of English and claimed to be able to assess roughly how many words the reader knew. It did this through a series of simple tests, one of which should you care to try it appears at the end of this article. Diack's tests implied that the maximum number of words a native speaker was likely to know was about 30,000 words. He reckoned that most people who have completed a university degree course would have a vocabulary size of about 29,000 words. At the times Diack's estimates were considered suspicious low, but recent, more careful studies have confirmed his work.

However, Diack's tests do not seem to work for non native speakers of English. For example, Izawa (1993) found that he scored a total of about 13,000 words on Diack's test, but felt that the test overestimated his real knowledge, Which was closer to 7,500 words. Diack's own analysis suggests that a score as low as 7,500 is typical of a not very bright 12 year old with little interest in reading. It is difficult to imagine that this could be an accurate description of a senior, experienced English language teacher. So what is going on?

There seem to be two reasons why the tests don't work properly for learners of English. The first is that Diack isn't really interested in low levels of proficiency and this is reflected in the way the tests were constructed. This explanation clearly doesn't apply to Izawa however. The second reason is more interesting. Diack's tests are based loosely on a frequency ordering of the words in the test. The words at level one are "easier" than the words at level two or three because they are more frequent, and that means that the average native speaker is likely to have come across them. The words at level four, on the other hand, are relatively rare, and not often used in speech. If you read a lot you might know some of these words, but you are very unlikely to have met them otherwise. The level five and six words are very unusual and many of them only occur in specialist areas such as medicine. Diack's tests work because he is able to use his native speaker intuitions to make an informed guess about how likely you are to have come across a particular word in the course of your everyday life. Words that children would know are "easy" words, while words only known to doctors, lawyers and so on are "hard" words.

The problem for non native speakers is that this equivalence doesn't always work. Just because a word is frequent, it doesn't mean that a learner will have met it and learned it. Conversely, just because a word is infrequent it doesn't follow that a learner won't know it. For example, a second language learner . who is training as a doctor may have a moderate level of ordinary English, and yet have a much greater knowledge of terminology than a well-read native speaker.

The real difficulty, then, seems to be that non native speakers of English often have unbalanced vocabularies. Perhaps what we really need is a negative vocabulary test, rather than a positive one not so much a test that identifies the hard, unusual words You know, as one that identifies the easy, ordinary words you don't know. A diagnostic test that located these gaps might be a very useful thing to have.

 WORD FREQUENCY INFORMATION

Another way to help learners redress their own 'vocabulary imbalance' is simply to provide them with word frequency information. This enables them to focus on high frequency vocabulary, consolidating their basic knowledge. The new edition of the Longman Dictionary of Contemporary English (LDOCE 3) takes this approach. Indicators are given for all words in the most frequent 3000 words in spoken or written English on the BNC / Longman Lancaster Corpus. An example of the marking system is shown below:

The indicators tell us that intend comes within the 2000 most frequent words in spoken English and the 1000 most frequent words in written English. In this way, int- words such as intelligent, intend, intention and interest are indicated whereas intemperate, intent, intercede and interchange are not. Thus the learner can distinguish core productive vocabulary frosts more peripheral vocabulary at a glance. The distinction between spoken and written English is also clearly useful to learners, many of whom are less aware of and less competent in spoken English. For many words, the differences in spoken and written frequency are quite dramatic. In LDOCE 3, a number of such spoken/written comparisons are presented in graph form:

Frequencies of the verb mean in spoken and written English

Based on the British National Corpus and the Longman Lancaster Corpus

This graph shows that the verb mean is much more common in spoken English than in written English. This is because it is used in a lot of common spoken phrases

 I SEE WHAT YOU MEAN!

What lies behind these figures? An analysis of a 1000 line lemmatized concordance for the verb mean from the BNC spoken corpus shows that over 73% of all occurrences are of the phrase I mean. By contrast, the sense 'have the meaning...' accounts for only 3% of spoken occurrences. In fact, in spoken English, mean is very much a vehicle for a wide variety of common spoken phrases, such as I see what you mean, how do you mean?, that means... and so on. Many of these phrases appear superficially transparent unlike the kind of stock 'idioms' that some learners know so well yet they are often quite subtle in meaning and context, and are unfamiliar to many learners. All these phrases are therefore included in LDOCE 3, with definitions and examples.

Another aspect which can be helpful to learners is the relative frequency of synonyms. Many learners tend to use more formal, infrequent words than native speakers. This may be for a variety of reasons culture, Ll interference, teaching methodology/materials and so on. These examples of the verb permit are taken form the Longman Learner's Corpus:

I hoped to meet another man who would permit me to hold his hand...

Moreover, the landlady doesn't permit me to go downstairs in the sitting room...

By showing the relative frequency of synonyms such as permit, let and allow, and at the same time teasing out the differences between them, learners can be made more aware of important word choice factors:

Frequencies of the verb let, allow and permit in spoken and written English Based on the British National Corpus and the Longman Lancaster Corpus This graph shows that the verb let is much more common in spoken English than allow and permit. Allow is more common in written English. Permit is a formal word meaning to officially let someone do something.

Graphs of this type can be used to draw the learner's attention to regional differences. In particular the distinction between British and American English is a crucial one, since many learners are taught and examined exclusively in one variety. LDOCE 3 contains a number of British / American frequency comparisons:

Frequencies of the nouns shop and store in British and American English Based on the British National Corpus and the Longman Lancaster Corpus This graph shows that when talking about a building where things are sold, shop is the usual word in British English and store is the is the usual word in American English. Americans use shop to mean a small shop where one particular type of thing is sold. In British English store is used to mean a very large shop that sells many different types of things, and is usually in the expression department store.

Having chosen the right word, the next step is to use it idiomatically. Here again, frequency information can be of help. Graphs can be used to give the learner a quick visual profile of tile grammatical behavior of a particular word, or to point out a range of frequent collocations:

This graph shows how common the different grammar patterns of the verb decide are Based on the British National Corpus and the Longman Lancaster Corpus

Frequency information therefore, is one of the practical ways in which learners can be helped to balance and consolidate their vocabularies, and at the same time become more aware of the key issues affecting both word choice and word use.

Click here to go to Diack's Test No. 19

Back to top