The Louvain EAP dictionary: From corpus to dictionary and back

Abstract

In this presentation, we will introduce the Louvain English for Academic Purposes dictionary (LEAD), viz. a customisable web-based dictionary-cum-writing-aid tool. The LEAD aims to meet the growing needs of non-native speakers, be they students of English or researchers, who have to write academic texts that conform to the established conventions of the genre (more particularly its phraseology).

The dictionary contains a rich corpus-based description of c. 900 academic words from the Academic Keyword List (Paquot, 2010) that express key functions in academic discourse. The list contains nouns (e.g. issue, contrast, parallel), verbs (argue, discuss, emerge), adjectives (differing, opposite), adverbs (namely, notably, however), prepositions (despite, such as) and conjunctions (while, albeit). To extract the phraseology (collocations and recurrent phrases) of these words, we made use of large corpora of academic texts, in particular the academic component of the British National Corpus which we supplemented with a number of home-made discipline-specific corpora. We also made use of the International Corpus of Learner English (Granger et al., 2009) to identify EFL learners’ difficulties in using these particular words.

The main originality of the LEAD is its customisability: the content is automatically adapted to users’ needs in terms of discipline and mother tongue background. Before using the dictionary, users have to select a domain (currently business, medicine, linguistics) and specify their L1 background (currently French and Dutch). Domain selection makes it possible to customise the output and illustrate the phraseological environment of a search word by means of example sentences automatically extracted from a corpus of either business, medicine or linguistics texts. One of the purposes of L1-background identification is to give feedback on errors and problems that a specific L1 population typically encounters.

The dictionary is not only corpus-informed but can also be described as a dictionary-cum-corpus as users have direct access to discipline-specific corpora. We make use of a new open source web-based corpus analysis system, viz. CQPweb, developed by A. Hardie (Hardie, 2009) for two main purposes: - give access to concordances of academic words and their collocations to provide users with more examples and make it possible for them to check whether a collocation or phrase that is not in the LEAD is correct or not; - query words that are not in the LEAD so that users can check how to use a word even though it is not in the dictionary. As we are focusing on a very specific and quite limited vocabulary, we want to be able to provide another kind of feedback than the very frustrating “No match found!” when the search word is not in the dictionary.

The automatic customisation of the dictionary to users’ discipline and L1 coupled with direct corpus access makes it a particularly dynamic tool. The inclusion of error warnings and targeted exercises gives it the status of a hybrid tool, i.e. both a dictionary and a learning resource. The LEAD is also a highly flexible tool, which could easily be customised to other L1 background populations, other disciplines, and other languages.

References

Granger, S., Dagneaux, E., Meunier, F. & Paquot, M. (2009). The International Corpus of Learner English. Handbook and CD-ROM. Version 2. Louvain-la-Neuve: Presses universitaires de Louvain. Hardie, A. (2009). ‘CQPweb – combining power, flexibility and usability in a corpus analysis tool’. Paper presented at the 30th ICAME conference, Lancaster, 27-31 May 2009. Paquot, M. (2010). Academic Vocabulary in Learner Writing: From Extraction to Analysis. London & New-York: Continuum