The focus of this timeline article is on the development and validation of measures to assess the vocabulary knowledge and ability of second language (L2) learners. There is a parallel and somewhat longer tradition of work on L1 vocabulary assessment, particularly by reading researchers in the United States. Some key references from the L1 research prior to the 1980s have been included in the present timeline, mainly on the basis that they had a strong influence on the thinking of L2 researchers as they began to work on their own measures of vocabulary knowledge. Since that time, L1 and L2 research have continued along largely parallel tracks, with regrettably little interaction between the two (for an account of the state of the art in L1 vocabulary assessment, see Pearson et al. Reference Pearson, Hiebert and Kamil2007).
Obviously the inclusion in examinations of questions and test items to assess learners' knowledge of foreign language vocabulary goes back much further than the 1980s. But the argument here is that it was during that decade that researchers began to publish reports of particular measures and procedures designed to assess defined aspects of L2 vocabulary knowledge or ability, rather than simply to test whether learners knew sets of words that they were supposed to have studied.
The decade of the 1980s is significant in that during that period (as has been frequently observed) a small band of dedicated researchers was working in comparative isolation from the mainstream of two disciplines in which one might have expected them to have a more central role: second language acquisition (SLA) research and language testing. SLA was then, and is still, predominantly concerned with the acquisition by learners of syntactic and morphological features of the target language. At the same time, language testing was under the influence of the communicative approach to language teaching, and the focus of attention in the literature shifted from knowledge of the lexical and grammatical elements of the language to the performance of language use tasks. Thus, in both cases, work on vocabulary was not seen as a priority.
At the conceptual level, the result of this isolation from wider influences was a relative lack of theory to account for the nature of L2 vocabulary knowledge and the processes of lexical development in L2. From a testing perspective, it meant that there was only a limited basis for defining the constructs being measured by particular language tests. Distinctions such as receptive vs. productive knowledge, or breadth and depth of knowledge, were made as if they were unproblematic and could be clearly understood without much further elaboration. Paul Meara has stood out among scholars working on L2 vocabulary development as one who has long pursued a more theoretical approach informed by psycholinguistics, employing computer simulations and modelling along with a range of innovative measures to operationalise the theoretical concepts (Meara Reference Meara2002, Reference Meara2009).
Apart from the work of Meara and his associates, the timeline shows how a number of authors in the last fifteen years have sought to identify various components or dimensions of vocabulary knowledge and to apply different approaches to the definition of the construct.
At the operational level, the separation from language testing meant that early papers in the 1980s and 1990s presented rather limited evidence for the quality of the test instruments. They relied typically on classical reliability statistics and correlations with concurrent measures, as well as t-tests or simple analyses of variance where appropriate. There was quite a lag before the principles of modern validation theory – which had been widely disseminated in the field of language testing through Bachman's (Reference Bachman1990) landmark volume – were reflected in research articles on L2 vocabulary tests. In the last ten years more sophisticated procedures, notably Rasch analysis, have been applied to the evaluation of vocabulary instruments, along with more complex and well-formulated validity arguments in some recent cases. In addition, there have been some critical re-evaluations of widely known vocabulary measures, to take a fresh look at their conceptual and empirical adequacy. It is symptomatic of the return of vocabulary assessment to the mainstream that there has been a noticeable increase in the number of articles on the topic accepted for publication in the major research journal Language Testing.
The selection of publications here is based on the author's extensive knowledge of the field, cross-checked against books found in the timeline itself – Read (2000), Milton (2009), Schmitt (2010), Nation & Webb (2011) – and a search of recent volumes of the major journals in applied linguistics.
The items in the timeline represent the following themes, which are labelled with the appropriate letter(s) in the last column of the table:
A Vocabulary size (breadth)
B Depth of vocabulary knowledge
C Knowledge of spoken vocabulary
D Productive use of vocabulary in speech and writing
E Role of context in vocabulary assessment
F Speed of lexical access
G Definition of the vocabulary construct
H Further validation or critical analysis of an existing vocabulary measure
John Read is an Associate Professor of Applied Language Studies at the University of Auckland, where he conducts research and teaches courses in language assessment, and vocabulary teaching and learning. He is the author of Assessing vocabulary (Cambridge, 2000), was co-editor of Language Testing (2002–2006), and has been President of the International Language Testing Association (2011–2012).
