View Content #18515

John Read is Associate Professor of Applied Language Studies at the University of Auckland, New Zealand, specialising in second language vocabulary assessment and the testing of English for academic and professional purposes. He is the author of Assessing Vocabulary (Cambridge, 2000) and Assessing English Proficiency for University Study (Palgrave Macmillan, in press). He has written numerous vocabulary tests for various purposes and is perhaps best known for his work on depth of vocabulary knowledge, particularly the development of the word associates format.

The first question to ask about this topic is which vocabulary should be assessed. The obvious answer for classroom teachers is to focus on the words presented in the coursebook or specified in the curriculum. However, English teaching materials do not always give much explicit attention to vocabulary and it may be necessary to consult appropriate vocabulary lists. The best lists are based on a careful analysis of how frequently words occur and in what range of contexts. Modern computer corpus analysis has made an enormous contribution in helping us to identify the most useful words, both in the language as a whole and in particular varieties or genres of English.

The next question is which aspects of word knowledge should be assessed. The traditional approach is to focus on the form-meaning link: can the learners demonstrate that they know the meaning of a given L2 word and, conversely, can they produce the L2 word which corresponds to a given meaning (often expressed as the L1 equivalent)?  However, we now recognise the vocabulary knowledge involves much more than that. Word meaning changes according to context, and high-frequency words are particularly likely to have a range of senses, both literal and figurative. Words also cluster into “word families”, and thus learners need to know that origin, original, originate and originally share a core sense, but they have different grammatical functions and shades of meaning.

Another significant aspect of vocabulary knowledge is the way that words go together in combinations. Although idioms, phrasal verbs and other fixed expressions are familiar as sources of difficulty for even advanced learners, corpus analysis helps to reveal how much of normal language use is made up of multi-word lexical units of various kinds.  Collocations have received a great deal of attention: does the learner know that we say heavy rain rather than strong rain, and we take (not do or make) a vacation?  Multi-word units also have pragmatic functions: do learners have the vocabulary resources to be able to make a polite request, apologise appropriately or express sympathy to a bereaved person?

This brings us to the question of how we assess these different aspects of word knowledge. A number of innovative formats have been devised for vocabulary research, but in operational tests the basic item types remain largely the same:
•    multiple-choice items of various kinds
•    matching of words and definitions
•    filling gaps in words or sentences
•    translation between L1 and L2
•    self-assessment tasks, using checklists or rating scales: (How well) do you know this word?
These formats lend themselves well to computer-based testing, which is one reason for their enduring appeal. What is changing is the use of these test items to assess aspects of vocabulary beyond the form-meaning link for individual L2 words and consequently more focus on testing vocabulary in context.