An Evaluation of Intermediate Students' Approaches to Corpus Investigation
Kennedy, Claire, Miceli, Tiziana, Language, Learning & Technology
This paper reports on our experience in using a corpus of our own compilation, Contemporary Written Italian Corpus (CWIC), in teaching intermediate students at Griffith University in Australia. After an overview of the corpus design and the training approach adopted, we focus on our initial evaluation of the effectiveness of the students' investigations.
Much has been written on what can be done with corpora in language learning: what kinds of discoveries can be made with different types of corpora. There is relatively little on how learners actually go about investigations. Since we intend for our students to progress from classroom use to independent work as a result of using a Web-based version of CWIC, we have been seeking to understand how successful they are at extracting information from this corpus in the absence of a teacher. Our initial study highlighted the complexity of the process and the specialized skills required. We found that lack of rigor in observation and reasoning contributed greatly to the problems that arose, as did ignorance of common pitfalls and techniques for avoiding them. We, therefore, conclude the paper with an outline of proposed changes to our apprenticeship program, aimed at better equipping the students as "corpus researchers."
Much of the literature on the use of corpora in language teaching relates to courses for advanced and highly motivated students of English for Specific/Academic Purposes (e.g., Johns, 1988, 1991a, 1991b; Levy, 1992; Mparutsa, Love, & Morrison, 1991; Stevens, 1991; Tribble, 1991) or translation (e.g., Aston, Gavioli, & Zanettin, 1998; Bernardini, 1998; Gavioli, 1996). So, when contemplating the introduction of work with corpora into the undergraduate Italian program at Griffith University in Australia, we were aware of the need to tailor the experience to quite a different target group, for whom Italian is usually a foreign rather than a second language and whose intentions for its use are less ambitious. In other contexts, our students might be regarded as reaching intermediate or higher-intermediate levels.
Our aim was to provide these students with a corpus to use primarily as a reference resource while writing. In view of their proficiency level and the types of writing tasks in which they engage, we sought a corpus that would supply models of personal writing on everyday topics. At the time, the only corpus of contemporary written Italian available to us was a collection of newspaper material, so we first resolved to create our own corpus, which we have named CWIC, or Contemporary Written Italian Corpus. Secondly, we also decided to initiate the students into corpus use in a gradual and guided manner and thirdly to attempt to evaluate the effectiveness of their work with CWIC as soon as possible. This paper discusses the implementation of those three decisions, with the focus on an initial evaluation exercise and its implications for our approach to training students.
The CWIC Project: A Corpus for our Teaching and Learning Context
Most of our students begin their university experience with no prior knowledge of Italian and can attend a maximum of 400 contact hours during their three years in our program. Additionally, they are not usually able to spend time in Italy during their studies nor are there many local opportunities for immersion. We estimate that, on the average, they graduate with "basic vocational proficiency" in reading and listening, on the scale used in Australia (Wylie & Ingrain, 1999), while their ratings in speaking and writing are lower, somewhere between "basic social proficiency" and "basic vocational proficiency."
In their second year, the students begin intensive writing practice with letters and diaries, some creative writing and informative pieces based on their own experience. In their third year, the work is more academic in the sense that they bring their analytical skills to bear on the topics. …