Przemysław Kaszubski

School of English

Adam Mickiewicz University

Poznań

Contact: http://main.amu.edu.pl/~przemka/przemek.html

A CASE FOR CORPUS-BASED EFL PEDAGOGY

ABSTRACT

A corpus of English essays written by Polish students of the language (3rd year and above) is currently being compiled in the Poznań School of English, in association with the International Corpus of Learner English Project, directed by Prof. Sylviane Granger from the Catholic University of Louvain, Belgium. The aim of the ICLE Project is, among others, to enable and facilitate multi-faceted computational contrastive studies of learner-derived material -- using compatible sub-corpora of English writing gathered from many different first language backgrounds (i.e. French, German, Dutch, Chinese, Czech, Polish, as well as native English). Other objectives of ICLE are those of stimulating research into the advanced EFL students’ interlanguage, and of contributing new -- massive and L1-relevant -- data to the development of English language pedagogy.

In accordance with ICLE and its corpus-based possibilities, this paper attempts to show how various statistical data (particularly instances of over- and under-use) obtained from comparing learner corpora of written English (specifically the Polish sub-corpus PICLE) and a corpus of native English writing might enhance the EFL teaching/learning process. Special attention will be paid to ways of adapting written English coursebooks to suit the posited needs of Polish students of English.

1. ABOUT THE INTERNATIONAL CORPUS OF LEARNER ENGLISH

The International Corpus of Learner English, ICLE for short, is a computerised corpus of English essays written by advanced adult foreign learners of the language. The constituent parts of ICLE include the controlling native English sub-corpus named LOCNESS, as well as a number of learner sub-corpora, among them Chinese, Czech, Dutch, Finnish, French, German, Japanese, Spanish, Swedish and Polish. The corpus is constantly growing in size, as new material is being added and new national teams join in. First comparative studies have already been carried out on the corpus; as is the case here, their authors mostly concentrate on lexical features characterising a respective national variety of student interlanguage.

Essays collected for ICLE have to meet the following sampling criteria:

a) argumentative or literature topics;

b) length: 500-1,000 words;

c) adult writers (university students);

d) advanced level of English (at least 3rd year of English studies);

e) EFL (not ESL);

f) max. 1,000 words per author;

g) each author has to fill in a questionnaire called the ‘Learner Profile’, which contains relevant biographical information (a copy of the Profile used with PICLE is in the Appendix).

The reader interested in the ICLE Project may refer to Granger (1993) and (1994); some information is also provided in Kaszubski (1996). Up-to-date information on the Project can be obtained from the Web homepage of the Centre for English Corpus Linguistics at the Catholic University of Louvain: http://www.fltr.ucl.ac.be/FLTR/GERM/ETAN/CECL/cecl.html

2. THE POLISH SUB-CORPUS OF ICLE (PICLE)

The Polish sub-corpus of ICLE, called PICLE, contains currently 150 essays (95,588 words of electronically available running text), a larger part of which (62,601 words; 100 essays) has been tagged grammatically. Another estimated 25-30 thousand words of new material is currently being added, so the data available at the moment amount to about 120-125,000 words. The word limit to be achieved is 200,000 words.

The compilation of PICLE thus far has been facilitated thanks to co-operation links with outside institutions (mainly English Departments in Poland) and individuals, who have contributed material to the corpus. At the same time, I can assure all those interested in joining the Project now that there is always room for more. Apart from shortening the time of PICLE compilation, each portion of ‘external’ texts also helps to make the corpus larger and less Poznań-specific, thus more representative nationally, a question which cannot be overstated in the light of possible research based on it.

More current information on PICLE and corpus linguistics matters can be obtained from my WWW homepage at http://hum.amu.edu.pl/~przemka.

Back to Unpublished articles....