An investigation of the vocabulary representative of postgraduate International Law texts

The reading requirements for postgraduate (LLM) International Law courses are challenging, particularly for second language students. As vocabulary is a key component in reading ability, this research aimed to help these students by identifying a discipline-specific vocabulary core of words essential for LLM reading. To this end, the 1.98-million word DSVC-IL corpus was compiled from 12 domains of International Law, comprised of 401 5000-word samples from 21 genres across three communicative functions (prescriptive, descriptive and hybrid). The DSVC-IL corpus is demonstrated to adequately represent the target domain of LLM reading, and this is confirmed by Law experts. A Single Word (SW) List was extracted on the basis of frequency, range, keyness and dispersion, with the BNC written academic component as a reference corpus, and using Kilgarriff’s (2009) Simple Math statistic for keyness, and Savický and Hlaváčová’s (2002) Average Reduced Frequency (ARF) to measure evenness of distribution. This study demonstrates that the SW List of 1026 flemmas is linguistically representative of the target domain. Moreover, a comparison of the ranked ARF scores of list items in two randomly-split halves showed items are evenly distributed (rs=.98, p<.000). The SW List provides 26.37% coverage of the DSVC-IL, and 23.87% coverage of other texts sampled from the domain. The study also demonstrates that while the list contains words classed elsewhere as general English or general academic, many are discipline-specific in their meaning, usage and/or lexical patterning. For instance, 15% of the items are shown to have a law-specific intra- or intertextual function, the knowledge of which would greatly benefit student readers. This research makes an important contribution to our understanding of International Law vocabulary, providing a sound foundation for the design of EALP teaching materials. Furthermore, it offers a new method for establishing and reporting representativeness, through the use of an innovative Representativeness Argument adapted from Toulmin (1958/2003).