These and similar questions came up during a fascinating series of lectures given this week by Paul Meara (“the world’s leading researcher in modelling vocabulary knowledge” according to Paul Nation), at the Pompeu Fabra University here in Barcelona.Traditionally, estimates of vocabulary size have been based on the number of words that subjects could define on a list taken at random from a dictionary: if the list represented 10% of the total words in the dictionary, the number of known words would then be multiplied by ten to give the total. But the method is fraught with problems, not least ‘the big dictionary’ effect: “The bigger the dictionary used, the more words people are found to know” (Aitchison 1987, p.6).
More sophisticated, and more sensitive, tests have since been designed, including Paul Nation’s widely used and very reliable Vocabulary Levels Test (described in Nation 1990), which targets five levels of word frequency (including a university word list) and involves matching words with simple definitions.
Meara himself has devised a number of vocabulary size tests, including the EVST (originally commissioned as a placement test by Eurocentres). Elegantly simple and very easy to administer, this checklist-type test requires takers simply to say which words they recognise in a sequence of frequency-based lists. But, as a way of controlling for wild guessing – or shameless lying! – the lists also include ‘pseudo words’, such as obsolation and mudge.
All the above tests are tests of receptive vocabulary knowledge. Testing a user’s productive vocabulary is more problematic. One approach is the aptly-named ‘spew test’, where test-takers are asked to produce as many words they can that share a common feature, e.g. that start with the letter B. Taking a somewhat different tack, Meara reported on some intriguing research he has done, matching frequency profiles of learner texts with statistical models of different vocabulary sizes. A student writes a text and a profile is generated in terms of the relative frequency of its words; the program then searches for a best match (a bit like the way that fingerprints are matched up), which in turn yields a fairly exact estimate of the learner’s vocabulary size. Magic! (You can check the program out for yourself at Paul’s _lognostics website. It’s called V-size).
But what does vocabulary size mean? And does size matter? Certainly, it seems that having a big vocabulary is a prerequisite for reading (and presumably listening) ability. As Bhatia Laufer (1997) puts it, “By far the greatest lexical obstacle to good reading is insufficient number of words in the learner’s lexicon. [In research studies] lexis was found to be the best predictor of success in reading, better than syntax or general reading ability” (p. 31).
More than that, vocabulary size may be a reliable predictor, not just of reading success, but of overall linguistic competence. Certainly, in first language acquisition, the processes of vocabulary development and grammar development are closely intertwined, with the former possibly driving the latter. Tomasello (2003), for example, cites research that shows that “only after children have vocabularies of several hundred words [do] they begin to produce in earnest grammatical speech”, which suggests to Tomasello “that learning words and learning grammatical constructions are both part of the same overall process” (p. 93).
If this is the case in first language acquisition, does it not also suggest that – for second language learning – the learner needs to assemble as big a lexicon as possible, and as soon as possible – even if this means putting other areas of language learning ‘on hold’?
Aitchison, J. 1987. Words in the Mind: An introduction to the mental lexicon. Oxfrod: Blackwell.
Laufer, B. 1997. ‘The lexical plight in second language reading” in Coady, J. and Huckin, T. (eds.) Second Language Vocabulary Acquisition: A Rationale for Pedagogy. Cambridge: Cambridge University Press.
Nation, I.S.P. 1990. Teaching and Learning Vocabulary. Boston, MA: Heinle and Heinle.
Tomasello, M. 2003. Constructing a Language. Cambridge, MA: Harvard University Press.