Lexical Processing in Second Language Learners
eBook - ePub

Lexical Processing in Second Language Learners

Papers and Perspectives in Honour of Paul Meara

Tess Fitzpatrick, Andy Barfield, Tess Fitzpatrick, Andy Barfield

Share book
  1. 176 pages
  2. English
  3. ePUB (mobile friendly)
  4. Available on iOS & Android
eBook - ePub

Lexical Processing in Second Language Learners

Papers and Perspectives in Honour of Paul Meara

Tess Fitzpatrick, Andy Barfield, Tess Fitzpatrick, Andy Barfield

Book details
Book preview
Table of contents
Citations

About This Book

This book presents studies from authors at the cutting edge of second language vocabulary research, whose output represents much of the current focus and direction of work in this area. The authors address various aspects of L2 lexical processing and explore different models of acquisition, processing and storage. The studies are linked by the fact that the authors have all belonged to the same dynamic and influential vocabulary acquisition research group led by Paul Meara. Alison Wray provides an overview of how Meara has led this group's research activities in an innovative PhD programme, and John Read and Paul Nation contribute a critical evaluation of Meara's wide-ranging contributions to the field of vocabulary acquisition research. The research studies presented here are relevant and replicable, offering researchers and teachers many valuable and critical insights into lexical processing in second language learners.

Frequently asked questions

How do I cancel my subscription?
Simply head over to the account section in settings and click on “Cancel Subscription” - it’s as simple as that. After you cancel, your membership will stay active for the remainder of the time you’ve paid for. Learn more here.
Can/how do I download books?
At the moment all of our mobile-responsive ePub books are available to download via the app. Most of our PDFs are also available to download and we're working on making the final remaining ones downloadable now. Learn more here.
What is the difference between the pricing plans?
Both plans give you full access to the library and all of Perlego’s features. The only differences are the price and subscription period: With the annual plan you’ll save around 30% compared to 12 months on the monthly plan.
What is Perlego?
We are an online textbook subscription service, where you can get access to an entire online library for less than the price of a single book per month. With over 1 million books across 1000+ topics, we’ve got you covered! Learn more here.
Do you support text-to-speech?
Look out for the read-aloud symbol on your next book to see if you can listen to it. The read-aloud tool reads text aloud for you, highlighting the text as it is being read. You can pause it, speed it up and slow it down. Learn more here.
Is Lexical Processing in Second Language Learners an online PDF/ePUB?
Yes, you can access Lexical Processing in Second Language Learners by Tess Fitzpatrick, Andy Barfield, Tess Fitzpatrick, Andy Barfield in PDF and/or ePUB format, as well as other popular books in Languages & Linguistics & Psycolinguistics. We have over one million books available in our catalogue for you to explore.

Information

Year
2009
ISBN
9781847699855

Chapter 1

Introduction: Meara’s Contribution to Research in L2 Lexical Processing

JOHN READ and PAUL NATION

Introduction

We are delighted to have been invited to contribute to this volume in honour of Paul Meara because we have long admired his work, as a researcher, a teacher and presenter, a writer, a bibliographer and a supervisor of doctoral theses in the field of second language vocabulary studies. Our perspective is that of outsiders, in the sense that we are the only authors in the book who have not worked closely with Paul as colleagues or research students at Swansea. This puts us in a good position to reflect on the wider impact of his scholarship in relation to our own work and the field generally. We will begin with some personal observations and then go on to review some major strands of his research in a more formal way.
Both of us met Paul for the first time at the AILA World Congress of Applied Linguistics in Greece in 1990, but we had been aware of his work for some time before that. Paul Nation remembers how pleased he was to read Paul’s review article ‘Vocabulary acquisition: A neglected aspect of language learning’, which was published in Language Teaching and Linguistics in 1980. At that time, there truly were only a small number of scholars with an interest in L2 vocabulary learning and it was heartening to realise that here was another researcher who was so obviously committed to work in this area. This led to correspondence and a regular exchange of papers, articles and ideas, which has continued to this day. However, for reasons that we will discuss further in a moment, his research and ours have proceeded mostly along separate tracks; Meara and Nation have collaborated only once, on a co-authored article (Nation & Meara, 2002) for Norbert Schmitt’s introductory survey of applied linguistics.
After the initial meeting in Greece in 1990, John proceeded straight to London to spend three months on sabbatical, working with Paul at Birkbeck College. This turned out to be a key transition point in Paul’s career because he was in his last few months at Birkbeck, having already accepted an intriguing new opportunity to establish a specialised research unit on vocabulary acquisition in what was then the Centre for Applied Language Studies at University College, Swansea. Subsequently, we have both visited him in Swansea and have met him regularly at international conferences. However, we have never been able to entice him to New Zealand. He travels widely in Europe and has an impressive range of collaborations with scholars in his own continent. We know that he also makes trips to Canada, the USA and Japan, but it seems that our part of the world is a hemisphere too far for him. Still, we hope that some day he may visit us and give our students the chance to meet him and hear firsthand one of his stimulating presentations on his research.
It is interesting to note in this regard that the two universities, Swansea and Victoria, which are perhaps the foremost centres in the world for doctoral research in second language vocabulary, are located at opposite ends of the earth. It is stretching the metaphor too far to say that they are poles apart in other respects, but there are certainly differences. The Swansea PhD programme (as described by Alison Wray in the Preface to this volume) is distinctive, if not unique, in its emphasis on distance study and the structured way in which it builds up the candidates’ expertise in research methodology, while still giving them the freedom to follow their own interests. The doctoral candidates at Victoria follow a more conventional programme of study, tailored to their individual needs and interests, alongside students working on a variety of other topic areas.
More importantly, there are differences in research orientation. Our work at Victoria grew out of very practical concerns to promote more effective vocabulary learning by students of English as a second or foreign language. In this, we built on the achievements of our mentors at the English Language Institute in Wellington, H.V. George and Helen Barnard, who in turn linked us to the British tradition of vocabulary study represented by giants in the field like Harold Palmer and Michael West. As a result, a great deal of the research effort at Victoria has gone into the development of word lists, vocabulary tests and computer programs to analyse the lexical content of texts – all intended ultimately to be of practical value to teachers and learners. By contrast, the work of Paul and his doctoral students at Swansea is motivated by more theoretical questions about the nature of vocabulary acquisition and the state of the learner’s mental lexicon at various stages of development. There is a stronger influence from psychology, seen not only in the research questions, but also in the types of measures and analytical procedures they have used in their experiments. Paul’s ongoing interest in theoretical models and computer simulations (of which we will say more later) is further evidence of the impact of psychology on his thinking.
The contrast in approaches has been brought home to Paul Nation on more than one occasion when he has been giving talks in various parts of the world. Members of the audience sometimes ask questions that he struggles to find an answer to, like ‘What model of vocabulary storage and learning is this based on?’ It usually turns out that the questioner has been in the PhD programme at Swansea. This indicates the distinctive orientation of Paul Meara’s work. It also highlights the fact that, after more than 15 years of operation, the Swansea programme has an impressive roll of graduates who are not only challenging inquisitors of other people’s research, but also, in many cases, productive researchers in their own right, as can be seen in the contents of this volume.
Paul Meara has noted in the past the lack of continuity in L2 vocabulary research, with many one-off studies where someone had researched an idea and written an article on it, but never followed that up with further research so that the individual studies could be set in the framework of a larger continuing investigation. Paul even produces bar graphs to show the piecemeal nature of the research, as a way of encouraging researchers to pursue vocabulary studies more persistently.
To some degree, it seemed in the 1980s and early 1990s that Paul was something of a dilettante himself within vocabulary research, in that he published a number of short papers in which he explored an interesting concept without apparently taking it further (see Meara, 1990a, 1992d). However, in reviewing his work for this chapter, we have been struck by the fact that there are definite lines of enquiry that he has pursued systematically over many years on his own account and through the work of his doctoral students. We will now move to a more formal analysis of some of the major strands of his work, to identify those continuities and point out connections with the work of others, including our own.

The Yes/No Format

We begin with the aspect of Meara’s work that is perhaps closest to our own interests. He brought a significant innovation into the field of second language testing by introducing the Yes/No format, in which learners are presented with a large set of words and simply asked to indicate whether they know each one or not. This type of test had previously been used by L1 reading researchers, going back at least to Sims (1929), under the name checklist. A modern version of the format was introduced to L1 research initially by Zimmerman et al. (1977) and then by Anderson and Freebody (1983), with one crucial addition to the basic concept: a certain proportion of the words were not real words in the target language, in order to allow the scores to be adjusted for guessing, confusion, overoptimism or whatever else might lead learners to claim knowledge of words that did not exist. Meara was one of the first to propose that this checklist test would be a useful vocabulary measure for second language learners.
Meara and Buxton (1987) introduced the format to readers of Language Testing as an alternative to multiple-choice items as measures of word knowledge. They raised various questions about the reliability of these items to assess individual target words, but also argued that a multiple-choice test could not adequately sample the number of words required to make a good estimate of vocabulary size, which the authors saw as the primary purpose of a vocabulary test. Elsewhere, Meara (1996) has expressed the view that up to a level of about 5000 words, vocabulary size is the single best indicator of the state of the learner’s lexical development.
It is easy to see how the Yes/No format would appeal to Meara from a variety of perspectives, apart from its effectiveness as a measure of vocabulary size.
• As the format lends itself well to electronic delivery, the design of computer-based Yes/No tests was for Meara an early venture into programming, which has subsequently led to a range of other computerised measures, some of which are now available as freeware under the Lognostics brand (www.lognostics.co.uk).
• Although the test-takers’ task is a very simple one, the scoring of a Yes/No test poses a number of challenges and, as discussed further below, has engaged Meara’s attention for many years.
• Similarly, the issue of what form the nonwords should take and how they influence the nature of the task continues to intrigue him. Early on, Meara and Buxton (1987) recognised that plausible nonwords in English could in fact turn out to be actual words in the learners’ L1, especially if they were speakers of cognate languages like French. This was the focus of a later study in Canada (Meara et al., 1994).
• Apart from its value as a research tool, the format has practical applications in language teaching for placement testing and as a means of monitoring the expansion of learners’ vocabulary development over time. Thus, Meara has sought ways to make Yes/No tests widely available to language schools and programmes in various forms, as a product on disk distributed under the auspices of Eurocentres (Meara & Jones, 1990), as a bound set of pen-and-paper tests (Meara, 1992a), and more recently both as a published CD-ROM (Meara & Milton, 2005) and as downloadable files from the Lognostics website (e.g. Meara & Miralpeix, 2006).
The scoring of Yes/No tests has been the major concern. The original basis for scoring was the Signal Detection Theory, as developed to account for the performance of sonar operators in locating enemy submarines, hence the names of the four possible responses: hit, miss, false alarm and correct rejection. Meara has written about Yes/No scoring in several papers (Huibregtse et al., 2002; Meara, 1992e, 1994), identifying the main issue as being the need to find a method of scoring that satisfactorily models the test-takers’ response behaviour without, for instance, overpenalising those who respond Yes to a small number of the nonwords. Other researchers have also taken up the question (e.g. Beeckmans et al., 2001; Eyckmans, 2004; Eyckmans et al., 2007; Mochida & Harrington, 2006). An emerging conclusion appears to be that, for most practical purposes, it is sufficient to use a relatively simple scoring formula, such as the proportion of the hits minus the proportion of false alarms (Mochida & Harrington, 2006).
One important application of the Yes/No format has been in DIALANG, the web-based diagnostic assessment system for adult learners of 14 European languages (www.dialang.org). Meara was commissioned to develop for the system a vocabulary size test consisting of 50 actual words (all verbs, to facilitate comparability across the languages involved) plus 25 nonwords. The basic function of the test within DIALANG is to determine the learners’ general level of language competence so that, in subsequent tests of various skills, they can respond to texts and items that are broadly at their level of ability. Alderson’s (2005) validation research on the English version of the DIALANG measures shows that the vocabulary test is remarkably good as a general measure of competence, as reflected in very substantial correlations (0.61 to 0.72) with the skills tests.
People associated with the Swansea research programme have extended Meara’s contributions in various ways. For his doctoral research, Shillaw (1999) investigated the possibility of dispensing with nonwords in Yes/No tests for English learners in Japan and relying instead on the misfit statistics provided by Rasch analysis to identify errant responses. And Milton and Hopkins (2006) have developed a version of the test called Aural_Lex, which presents the test items orally, rather than in written form on the screen. They have used Aural_Lex to investigate whether native speakers of Arabic produce underestimates of their English vocabulary knowledge in written tests because of their well-attested tendency to transfer L1 reading strategies to the decoding of English words (Milton & Hopkins, 2006; Milton & O’Riordan, 2007).
In addition to his research on the validity of DIALANG, Alderson (2005) also re-examined the general concept of diagnosis in language assessment and advocated a greater role for diagnostic tests in the field. If this call is taken up, it will give fresh impetus to the development of vocabulary measures such as Yes/No tests, and Meara’s pioneering efforts will be even more widely recognised. There is evidence that this is already happening, as reported in yet unpublished research into computer-based diagnostic testing.

Word Associations

Some of Meara’s earliest work on vocabulary acquisition investigated the potential of word association data to give insights into the state of the second language learner’s mental lexicon. In fact, his very first published article on vocabulary had the title ‘Learners’ word associations in French’ (Meara, 1978). The basic word association task – presenting subjects with a list of preselected words one by one and asking them to give the first word that came to mind in response – was a well-established research tool with native speakers, which had yielded two findings in particular that were of interest. One was that native speakers produced a remarkably stable set of responses to many common stimulus words, with relatively few ‘deviant’ items. The other finding was a developmental sequence whereby younger children gave a preponderance of responses that were syntagmatically related to the stimulus word (as in blackbird), but there was a clear shift as they grew older to more paradigmatic responses (blackwhite), which predominate in the output of adult subjects (Entwisle, 1966; Postman & Keppel, 1970).
These findings suggested that word association responses would provide interesting data for L2 vocabulary studies. Meara and his students at Birkbeck undertook a number of studies in the 1980s (Meara, 1982b, 1984), which showed that learner responses were much more diverse and less stable than those of native speakers. In addition, learners had a strong tendency to produce ‘clang’ associations: responses that sounded similar to the stimulus words but were not semantically related at all. Thus, the word association task did not seem such a productive avenue for investigating how learners organised their knowledge of L2 words. Subsequently, Söderman (1993b) found that clang associations were likely to result from a lack of familiarity with the stimulus words, rather than any phonological basis for the organisation of learner lexicons. Singleton and Little (1991) made a similar point, using data from a rather different elicitation task, the C-Test.
In the meantime, Meara moved on to a variety of other ways of exploring the basic concept of word association. One line of development began during the three-month period Read spent at Birkbeck in 1990. He was looking for a practical means of assessing depth of word knowledge and, working with Meara, devised the word associates format (Read, 1993, 1998), which essentially required learners to select appropriate associations rather than supplying them. The items consisted of a target word plus eight other words, four of which were associates (with varied semantic relationships to the target) and four distractors. Although Read has been cautious in his claims for the validity of the format as a measure of deep word knowledge, others have taken it up for their own research purposes and modified it in various ways (e.g. Greidanus & Nienhuis, 2001; Qian, 2002; Schoonen & Verhallen, 2008). Meara himself stripped the format down to its essence by creating a set of French tests (Meara, 1992b) with items comprising pairs of words, some related to each other (vache ‘cow’ and lait ‘milk’) and others not (avion ‘plane’ and écrire ‘to write’). Bogaards (2000) conducted trials of several of these tests to compare the performance of native speakers of French with that of Dutch native speakers who had advanced levels of proficiency in French, with somewhat mixed results. In both versions of the format (Read’s original one and the paired version), native speakers do not always identify the intended associations and there are issues related to the role of the nonassociated elements, so that the format remains an interesting research tool rather than a test in widespread practical use. Meara has largely moved away from seeing depth of knowledge as a useful concept. He believes that it does not help us understand how the lexicon works, although it may be useful in assessing how well individuals know particular words (Meara, 1992d: 69).
A second line of enquiry adopted by Meara has been to use chains of associations to investigate the overall structure of a person’s vocabulary, building on the widely used metaphor of the lexical network or web. The original elicitation task involved presenting participants with pairs of words such as cold…desire and instructing them to produce a series of words to link the two stimulus ones, as in coldhotpassiondesire. The resulting response chains were analysed using the mathematical principles of Graph Theory, which provides both measures and a visual representation of the density of associations among the words in a person’s lexicon. The initial exploratory work at Birkbeck (Meara, 1992c; Welsh, 1988, cited in Wilks & Meara, 2002) led to further research by Wilks (1999, cited in Wilks & Meara, 2002). Despite Wilks’ efforts to refine the basic model, these studies were unsuccessful in supporting the hypothesis that the associative chains of L1 speakers would be shorter and denser than those of L2 speakers, partly because participants tended to produce idiosyncratic and unpredictable responses.
Wilks and Meara (2002) took the work a stage further by developing a new elicitation procedure. As with the word associates format, it entailed a switch from supplying to selecting associations – except that in this case, associations were accepted at face value rather than being judged correct or incorrect. The participants received a questionnaire with 40 sets of five randomly selected high-frequency words in French and were asked to circle two of the w...

Table of contents