Although it is a core component of a language user's overall linguistic competence, vocabulary has traditionally not received the kind of attention it deserved, being usually lumped with other kinds of competence till 1980s and 1990s when researchers started focusing on it with great interest. The present paper discusses what the notion of lexical competence entails, and also attempts to look at how vocabulary researchers have conceptualized its dimensions, especially the breadth and depth of vocabulary knowledge. The paper also discusses the different ways in which these dimensions of lexical competence can possibly be measured, drawing data from an ongoing cross-sectional study on lexical competence in Assam University.
It is an established fact that vocabulary teaching and learning has traditionally received little attention in second and foreign language programmes despite vocabulary being, as Richards and Renandya (2002, p. 255) say, “a core component of language proficiency”, providing “much of the basis for how well learners speak, listen, read, and write.” The Cinderella of language pedagogy, vocabulary, continued playing a secondary role even when Communicative Language Teaching (CLT) emerged on the horizon in the 1970s although “CLT gradually brought in its wake a major re-think on the role of vocabulary as people started recognizing the meaning-making potential of words and, therefore, their importance for the second and foreign language learners” (Syam Choudhury, 2010, p. 308). The turnaround happened when in the late 1980s ground-breaking developments in lexicography started taking place, involving the “extensive corpora of spoken and written language and the creation of sophisticated computer-based access tools for such corpora” (Carter, 2001, p. 43). Later, vocabulary teaching got an impetus when Lewis (1993) put forward his 'lexical approach', focusing on the development of learners' lexical competence. Lewis not only broadened the notion of vocabulary to include 'lexical chunks', i.e., groups of words commonly occurring together or collocations, but also proposed that since language was basically 'grammaticalised lexis' rather than 'lexicalised grammar', lexis and lexical competence should be at the centre of language pedagogy.
To understand what lexical competence is, it is essential to try to understand what it means to know a word. Richards (1976) was one of the first applied linguists to propose the concept of 'knowing a word', which, in his opinion, included knowing the degree of probability of encountering a word in speech or print, the limitations on the use of the word according to variations of function and situation, the syntactic behavior associated with the word, the underlying form of the word and the derivations that can be made from it, the network of associations between the word and other words in the language, the semantic value of the word, and the different meanings associated with the word. Largely retaining this lexical knowledge framework of Richards, Nation (1990) added pronunciation as an important component to make the framework more inclusive. In addition, Nation (1990) made explicit the distinction between the receptive and productive knowledge of vocabulary, pointing out that production involved a higher level of knowledge of vocabulary than reception did. Later, Nation (2001) revised his early framework to point out that the knowledge of a word or, in other words, lexical competence, includes three kinds of knowledge:
(I) knowledge of form (spoken form, written form and word parts),
(ii) knowledge of meaning (form and meaning, concept and referents, and associations), and
(iii) knowledge of use (grammatical functions, collocations and constraints on use).
Even before Richards and Nation, Dale (1965, cited in Read 2000) developed the following four-stage scale to represent the varying degrees of lexical competence:
Stage 1: 'I never saw it before.’
Stage 2: 'I have heard of it, but I don't know what it means.’
Stage 3: 'I recognize it in context…it has something to do with…’
Stage 4: 'I know it.’
It is important to point out here that Dale developed this scale for first language users. For second language learners, Paribakht and Wesche (1993) produced a Vocabulary Knowledge Scale, quite similar to the one developed by Dale but having one additional stage: 'I can use this word in a sentence.’
Thus the author finds that lexical competence has been conceptualized differently by different researchers, depending on their view of what constitutes vocabulary knowledge. What is common to all the viewpoints is an understanding that lexical competence is multidimensional and learning a word is a complex and gradual process.
A common feature in vocabulary studies is to look at lexical competence in terms of a number of easily measurable dimensions. One of the most commonly accepted views of vocabulary acquisition is that the acquisition of word knowledge occurs along a continuum of development. The fundamental idea is that the word knowledge develops in some kind of a hierarchical order. In line with the continuum perspective, Henriksen (1999) proposed a three-dimensional model of lexical competence:
(I) partial-to-precise knowledge,
(ii) depth of knowledge, and
(iii) receptive-to-productive dimension.
The first dimension, partial-to-precise knowledge, deals basically with the breadth or size of vocabulary knowledge, and is conceptualized as a kind of journey of the learner from simple word recognition moving through several stages of partial knowledge to a precise comprehension level resulting from a widening of the knowledge base of the learner. However, vocabulary size cannot be the only dimension by which the author can come to an understanding of the lexical competence of a language user. In this context, the second and third dimensions of Henrikson's model gain importance. The second dimension, depth of knowledge, pertains to the relationship of a word to other words in the lexicon. The relationship could be paradigmatic (antonymy, synonymy, hyponymy, etc.) or syntagmatic (collocational restrictions). The third dimension, the receptive-to-productive dimension, pertains to the level of mastery of vocabulary knowledge reflected in the learner's comprehension and production abilities. Receptive vocabulary is obviously bigger in size since it pertains to the ability to comprehend a lexical item only. On the other hand, productive vocabulary entails the ability to use a lexical item in production. In comparison with Henrikson's multi-dimensional model, Meara (1996) proposed a framework with only two dimensions: vocabulary size, and organization, i.e., the ways in which the words are related to one another. These two dimensions, Meara (1996, p. 15) pointed out, had the advantage that they were relatively “independent” of the items that contributed to them, and did not require “a detailed understanding of the way individual lexical items function.” A more comprehensive framework is the one designed by Chappele (1998), who put forward a quadridimensional framework of lexical competence having the following: vocabulary size (i.e., the total number of words that a person knows), knowledge of word characteristics (i.e., the knowledge of each word from vague to precise), lexicon organization (i.e., the manner in which words are stored in the mind of the leaner), and processes of lexical access which enable a language user to access his or her vocabulary knowledge while writing or speaking. In a more recent framework developed by Qian (2002), the researcher drew on the earlier models of lexical competence and proposed another quadri-dimensional framework having vocabulary size, depth of vocabulary knowledge, lexical organization, and automaticity of receptive-productive knowledge, a dimension stressed by Meara (1996) as well since it is believed that the hidden lexical competence of automaticity helps in the development of both receptive and productive vocabulary.
In all the models of lexical competence discussed above, two dimensions which have always featured are vocabulary size and the depth of vocabulary knowledge. In the light of this, it is believed that a brief discussion on these would be fruitful.
Vocabulary size is often characterized as the breadth of vocabulary knowledge or the number of words a person knows, and is one of the most basic dimensions of lexical competence. The breadth of vocabulary knowledge is closely associated with three interesting questions:
(i) how many words are there in the language under consideration?
(ii) how many words does the native speaker of the language know? and
(iii) how much vocabulary does a second language learner need?.
Goulden, Nation and Read (1990) counted the number of word families in Webster's Third New International Dictionary (1964) , one of the largest dictionaries of English. After excluding entries such as proper names and alternative spellings, Goulden, Nation and Read (1990) found that the dictionary contained about 54,000 word families. But this learning goal is far beyond the reaches of second language learners and even for the native speakers. Recent studies (Nation & Waring, 1997, p. 7, for instance) suggested that an average university-educated English speaker knows around 20,000 word families. Even though there are large variations among the individuals, a figure like this, which excludes proper names, compound words, and abbreviations, etc., is generally accepted. Liu and Nation (1985, cited in Hunt A. & Belgar D, 2002) and Nation (1990) found that a second language reader needs a requisite knowledge of a minimum of 3,000 words to achieve 95% coverage of a general text. However, most second language researchers nowadays recommend a basic vocabulary of at least 3,000 word families, and for more specialized needs, a working vocabulary of over 5,000 word families (Nation, 1990).
A different way of describing lexical competence is by specifying how well a particular word is known. There has been little agreement among second language researchers about what depth or quality entails. Meara (1996) viewed the depth of knowledge as the interaction between individual words and envisaged depth as the organization of words in the mental network. According to him, depth of vocabulary knowledge, which he calls “organization”, refers to the relation that a word might have with other words in the language (p. 9). Based on the dimension proposed by Meara (1996), Read (1998) identified three types of relations to the word, i.e., paradigmatic (synonymy, antonymy, hyponym, etc.), syntagmatic (collocation) and analytic (one aspect of the meaning). It is assumed that learners with high vocabulary proficiency have dense and more organized networks than less proficient ones.
The question that arises now is how these dimensions are to be measured to make a useful distinction between learners at different levels of second language proficiency. The following paragraphs discuss some of the ways in which the lexical competence of second language learners has usually been measured. In this context, the author would draw examples from a research on lexical competence carried out under his supervision at Assam University to substantiate his viewpoints.
Traditionally, researchers researching on lexical competence have focused on measuring the breadth of vocabulary knowledge. A variety of measures have been developed to assess it. Initially, vocabulary used to be measured using two sharply contrasting methods:
In the dictionary-sampling method, the words were sampled from the dictionary and then, a person's vocabulary size was estimated by multiplying the number of sample words known by the ratio that the sample of words bears to the total number of words in the dictionary. Nation (1990, p. 76) generated a sample formula to demonstrate the vocabulary size using the dictionarysampling method:
For example, for a dictionary of 20,000 words, from which a sample of 100 words is selected, if a person knows 20 of the sample words, his vocabulary size would be 4,000 words (20 x 20, 000/100). But this method was not successful because the estimated vocabulary depended heavily on the size of the dictionary and the dictionary's definition of a word (Anderson & Freebody, 1981; Lorge & Chall, 1963). As a result, lexical research has tended to use the frequencybased sampling method, as an alternative one for selecting test items for measuring vocabulary size. Anderson and Freebody (1981, p. 23) claimed that “frequency is a characteristic of a word which probably is very strongly related to the chances that the word will be known”. That is, the words which occur more frequently will be learned relatively earlier so that the learner's knowledge of words at a given frequency level displays their overall vocabulary size.
Another crucial thing to consider is the nature of the test format to be used for the measurement of lexical competence. One of the formats is the one used by Checklist Tests. This format allows a large number of words to be tested in a short space of time. Target words are presented in a list with one non-word item for every two real words. The learners are merely required to check if they know them. If some of these non-words are checked, the indication is that the learner is over estimating his or her vocabulary knowledge. Meara and his colleagues are closely associated with this format, having developed a book of pencil-and-paper Checklist Tests called the EFL Vocabulary Tests (Meara, 1992) and a commercial computerized version called the Eurocentres Vocabulary Size Test (EVST) (Meara and Jones, 1990).
Another test for measuring the breadth of vocabulary knowledge is the Vocabulary Levels Test developed by Paul Nation (1990). Rather than giving a single estimate of the total vocabulary size, it measures the knowledge of words at the following five frequency levels:
(i) 2,000 words, related to basic everyday oral language,
(ii) 3,000 words, related to words required to begin reading authentic texts in a second language
(iii) 5,000 words, related to words necessary for reading authentic texts,
(iv) 10,000 words, related to words necessary for University study, and
(v) a special level called University Word Level dealing with academic words in pedagogical contexts
Both Checklist Tests and the Vocabulary Levels Tests have been accepted by a number of researchers (Laufer and Paribakht, 1998; Qian, 1999) as appropriate for measuring vocabulary size in terms of validity and reliability. However, for measuring the vocabulary size (the receptive breadth) of students at the post-intermediate and advanced levels in a cross-sectional study carried out in Assam University across five academic departments (Social Work, Visual Arts, Computer Science, English and Linguistics), the Vocabulary Levels Test (VLT) framework, proposed initially by Nation (1990) and updated later by Schmitt, Schmitt and Calpham (2001) was used. This VLT framework has the following five levels: 2000-word level, 3000-word level, 5000-word level, 10,000-word level and the level of academic vocabulary. According to Nation (1990), the words in the 2,000- and 3,000-word levels include only high-frequency words in English; the 5,000-word level is the boundary between the high frequency and low frequency words, and the 10,000-word level includes low frequency words. The academic vocabulary level contains specialized vocabulary items required for University studies. The words contained in the 2,000 to 10,000 word levels are based on the word frequency list of Thorndike and Lorge (1944), cross referenced with the frequency data from Kucera and Francis (1967) and the General Service List (West, 1953). The updated format of the VLT, used for the Assam University study, consisted of ten clusters of words from each frequency level, each cluster consisting of six words and three definitions. An example of the cluster is given below:
Thus, the task of each of the participants in the study was to select three items from the group of six words to match the corresponding definitions. At each level, the test-takers were required to match sixty words against thirty definitions. Therefore, three hundred items were there in the VLT as against one hundred and fifty short definitions. The maximum possible score was one hundred and fifty points (three definitions in each cluster X 10 clusters across 5 frequency levels). The VLT was administered to ninety-eight students (fifty-seven post-intermediate students of the Social Work, Visual Arts and Computer Science departments, and forty-one advanced ones of the English and Linguistics departments).
Table 1 presents the average percentage of the correct responses of the VLT at the two levels:
Table 1. Vocabulary Levels Test (VLT) Results Across Each Word-Frequency Level (for Measuring the Receptive Breadth of Lexical Competence)
As can be seen in Table 1, the post-intermediate level students achieved 93.10% at the 2,000-word level, 84.56% at the 3,000-word level, 71.81% at the 5,000-word level, 82.69% at the Academic Word Level (AWL) and 33.80% at the 10,000-word level. On the other hand, the advanced level students scored 98.05% at the 2,000-word level, 92.76% at the 3,000-word level, 82.93% at the 5,000-word level, 89.70% at the Academic Word Level and 44.31% at the 10,000-word level. The students at the advanced level of communicative competence scored a higher percentage (81.40 %) in the VLT than those at the postintermediate level (73.02%). In other words, there was an increase in the correct number of responses in the VLT (for measuring the receptive breadth) with an increase in the level of communicative competence. The same was the case with the Productive Vocabulary Levels Test (PVLT), the format for which was designed by Laufer and Nation (1999), used for measuring the productive breadth of lexical competence of the participants. The structure of the test was a further development of the VLT, with each level consisting of eighteen items at each of the 2,000-, 3,000-, 5000-, University Word Level (UWL) and 10,000-word levels. A meaningful sentence context was presented for each item and the missing target word was to be supplied. The initial letters (ranging from one to five) of the target word were provided to prevent the test-takers from filling in another word. An example is provided below:
Table 2 presents the average percentage of the correct responses of the PVLT at the two levels:
As can be seen in Table 2, the post-intermediate level students attained 71.25% at the 2,000-word level, 37.13% at the 3,000-word level, 34.70% at the 5,000-word level, 44.13% at the Academic Word Level and 16.47% at the 10,000-word level. On the other hand, the advanced level students achieved 77.37% at the 2,000-word level, 53.66% at the 3,000-word level, 44.85% at the 5,000-word level, 59.20% at the Academic Word Level and 24.12% at the 10,000-word level. The students at the advanced level of communicative competence scored a higher percentage (51.84%) in the PVLT than those at the postintermediate level (40.90%). In other words, there was an increase in the productive breadth of lexical competence with an increase in the level of communicative competence.
Table 2. Productive Vocabulary Levels Test (PVLT) Results Across Each Word-frequency Level (for Measuring the Productive Breadth of Lexical Competence)
The number of ways for measuring the depth of vocabulary knowledge in lexical competence literature is relatively small when compared with those for measuring vocabulary size. The fact that there are so many components involved in the depth of vocabulary knowledge makes the measuring of depth a complex task. Schmitt (1998) pointed out two approaches in measuring the depth of vocabulary knowledge: the developmental approach and the dimensions approach. The first attempts to measure how vocabulary develops over time and the other measures which word knowledge types are known.
One measure in the developmental approach which is increasingly gaining some significance is Paribakht and Wesche's (1993) Vocabulary Knowledge Scale (VKS), which has already mentioned earlier. It uses five scales to capture certain stages in the initial development of the core knowledge of given words, combining a self-report with some verifiable evidence of word knowledge in the form of a synonym, first language translation, etc. The learners use the scale to report how well they know each of the target words. However, VKS is essentially concerned with describing the very basic stages through which a word might pass and no attempt is made to account for more detailed knowledge about a word that develops over time.
The dimensions approach, on the other hand, focuses on word associations. The Word Associates Test (WAT) developed by Read (1993, 1998) was one of the first attempts to measure the associative and collocational word knowledge, in addition to conceptual knowledge. This test is based on the concept of word association by creating items that consist of a target word and eight other words. The task is to identify the related words, or associates, for each target word. Read (1998) concludes that the test composed of these items gives a good overall view of the test taker's vocabulary knowledge.
For the present study, the WAT developed by Read (1998) was used. The WAT consisted of 40 adjectives as target words, and two boxes, each containing four words. The box on the left contained some words that had a paradigmatic relation with the target word, representing one aspect of its meaning (synonymy or polysemy). The box on the right contained certain words having a syntagmatic relation with the target word, i.e., its possible collocations. An example is given below:
The participants in the study had to select the words from the boxes with regard to their being synonyms and collocations of the given words. Each item always had four correct choices. However, these choices were unevenly distributed in the two boxes. There were three kinds of possibilities: (1) one answer from the left and three words from the right; (2) two answers from the left and two from the right; (3) three answers from the left and one from the right. This arrangement was designed to reduce the chances of guessing (Read, 1998, p. 46)
Table 3 presents the average percentage of the correct responses of the WAT at the two levels:
As can be seen in Table 3, the advanced-level students (73.87%) scored a higher percentage of correct responses in the WAT than the post-intermediate-level students (67.53%). These results indicated that the advanced-level students possessed a higher degree of depth of vocabulary knowledge than the post-intermediate-level students.
Table 3. Word Associates Test (WAT) Results (for measuring the depth of lexical competence)
The findings of the study discussed above revealed a significant difference in the lexical competence of the students at the post-intermediate and advanced levels of communicative competence. It could be assumed that these differences were largely due to their different learning environments. The advanced-level students were from the departments of Linguistics and English and, therefore, got more exposure to the second language than the postintermediate level students. The advanced-level students in the present study received an average of sixteen hours of instruction in the second language, and most of what they did related to the English language and literature. On the other hand, the post-intermediate-level students received only an average of five hours of second language instruction per week. Therefore, students at the advanced level of communicative competence had a greater opportunity to learn and use the second language than the post-intermediate students, and this surely had a bearing on the results of the tests conducted. However, one has to admit that since vocabulary acquisition is an ongoing process, cross-sectional studies like the one from which some data for this paper were drawn can provide only a snapshot of the lexical competence of the learners at a particular point in time. A more profitable endeavour would perhaps be to conduct longitudinal studies, which would be able to provide a more comprehensive view of what is meant by knowing a word.
This paper has attempted to describe and analyse the notion of lexical competence and its dimensions. It has also tried to look at the different ways in which the two most important dimensions of lexical competence, breadth and depth of vocabulary knowledge are usually measured in lexical competence research, drawing data and examples from an ongoing research in Assam University. Overall, an idea which has hopefully come through in the discussion is that lexical competence is, far from being a simple sum of all the lexical items that a person knows, a complex phenomenon which, however, may be defined in terms of some measurable dimensions.