Vocabulary and English for Specific Purposes Research - Averil Coxhead 2018
Pre-university, undergraduate and postgraduate vocabulary
Martínez et al. (2009) carried out a corpus-based study of specialised vocabulary in research articles in agricultural sciences (Agro Corpus) to find out more about the AWL (Coxhead, 2000) and items outside that list. The corpus contained 826,416 running words from 218 research articles written by academics based in English-speaking universities and published between 2000 and 2003. The corpus analysis included a comparison across sections of the articles: introduction, method, results and discussion. Eight AWL types covered 0.4% of the agriculture corpus: data, analysis, significant, similar, significantly, area, response and concentration. The highest proportion of AWL items was in the discussion sections. Martínez et al. (2009) suggest that discussion sections of research articles tend to be more argumentative (c.f. Martínez, 2003) which could be a reason for the higher use of AWL vocabulary. The discussion sections tend to be longer than method or results sections, meaning that these lexical items have more opportunity to occur. The distribution of lexis across the other sections varied, with less variation and fewer AWL items in results sections and more variation in lexis in methods sections. Table 6.3 shows examples of AWL items across the sections of the AgroCorpus.
The comparison of these findings from Agriculture with Chen and Ge’s (2007) medical study (see the following), Coxhead’s (2000) general academic vocabulary research and Hyland and Tse’s (2007) analysis of lexis across disciplines and in student and professional writing, revealed that some, but not very much, vocabulary is shared between the corpus-based findings of all four studies. In part, this finding is not surprising, because of the different research goals and corpora involved in each study. Martínez et al. (2009, p. 196) conclude that their study shows ’that the more specific the corpus, the greater the specificity of items, and consequently, the lower the variability.’
Table 6.3 Most frequent academic word families in the sections of the AgroCorpus (Martínez et al., 2009, p. 190)
A further step in the study by Martínez et al. (2009) was to select several items from the corpus and carry out a qualitative analysis using a technical dictionary and examples of the lexical items in context and in clusters from the AgroCorpus. A good example is culture, which relates to plant cultivation, as in these examples: blueberry cell cultures, cultures were grown, cultures were maintained and cultures were incubated. Martínez et al. (2009) also point out that high frequency words from West’s GSL (1953) occur with specialised meanings in Agriculture in their corpus, for example, effect and control, and that learners may not know these words before starting their studies. This point provides a counter argument to assuming that learners would know the first 2,000 words of English before commencing study of an academic nature (Coxhead, 2000). More analysis of target words through consultations with dictionaries and experts, as Martínez et al. (2009) have done in their work enriches quantitative findings and provides more evidence of language in use in specialised subjects.