Some background
Why study Phraseology and Discourse?

SOME BACKGROUND

While the field of Foreign Language Learning (FLL) has been researched for over 50 years, many questions remain unanswered. In addition, FLL constitutes (perhaps more than ever) a major societal challenge in our modern world where mobility and knowledge of foreign languages can no longer be considered as optional but have become an essential asset of European citizenship and social integration.

Learning a foreign language involves a combination of extra-linguistic (age, motivation, socio-cultural environment, etc.) and purely linguistic factors. The influence of linguistic factors has often been studied from two radically different perspectives: cross-linguistic influence (or transfer) and developmental issues (see Cook 1993, Ellis 1994 or Gass 1996). While proponents of the cross-linguistic trend support the existence of positive and/or negative transfer from the learner’s mother tongue, other theoreticians suggest that all foreign language learners, irrespective of their mother tongue, are faced with similar developmental issues.

Although Second Language Acquisition (SLA) studies have gained wide recognition, some major methodological issues or weaknesses should be addressed:
- the term Second Language Acquisition (SLA) has often been used as a general term encompassing all aspects relating to the ‘mastering of a language which is different from the mother tongue’, hence disregarding or blurring important differences between second language and foreign language or between acquisition and learning
- the transfer and developmental perspectives have often been studied separately, using different research methodologies, hence losing the benefit of cross-fertilization (Oshita 2000, Granger 2002)
- the amount of solid empirical data used to support the theories is often very scarce
- more attention has been devoted to phonology, lexis or syntax, while other domains such as phraseology or discourse have been largely neglected
- more attention has been devoted to the early stages of learning
- the influence of classroom practice has too often been neglected, disregarding the fact that studying FLL inevitably involves considerations not only of learners’ performances per se but also of the teaching and learning methods which have helped the learners attain their proficiency level in foreign languages

The present project aims to address the above-mentioned issues and has three main objectives:

• THEORETICAL: studying
  - both developmental and cross-linguistic influence on the learning of foreign languages
  - in two (so far) largely neglected areas, i.e. phraseology and discourse
  - at more advanced proficiency levels
• METHODOLOGICAL:
  - demonstrating the importance of corpus data and methodology to complement experimental data. We will make use of The International Corpus of Learner English (ICLE) which currently contains over 2 million words of argumentative essay writing from university students of English from 14 different mother tongue backgrounds. ICLE is used as a research tool for analysing features of written interlanguage grammar, lexis and discourse. For a bibliography of publications relating to ICLE click here.
  - evaluating automatic language analysis tools
• APPLIED: reinforcing the link between theory and practice in FLL research by
  - taking classroom practice into account in FLL research
  - integrating research findings in teaching practice (adapting the curriculum, improving teaching methods and offering tailor-made tools for advanced learners)

WHY STUDY PHRASEOLOGY AND DISCOURSE?

Two levels of linguistic analysis will be investigated in this project:

• Phraseology, which can be defined as the study of cooccurring lexical units, covers a wide range of expressions with diverging degrees of fixedness and opacity.
• At the discursive level we will concentrate on a number of linguistic devices indicating the global structure of a text: metadiscourse markers and segmentation markers.

We propose to study these two linguistic levels for the following reasons, which apply equally well to both levels:

• These two research domains are fairly young. Linguistic as well as psycholinguistic research on the discourse level did not start until the early 70s. The specific domain of discourse markers is even younger, since the first significant monographs on the topic only appeared in the late 80s (e.g., Schiffrin, 1987). Nevertheless, this domain of research is nowadays in full expansion (Fraser, 1999; Redeker, 2000). The emergence of phraseology as a field in its own right is even more recent : it was only in the 80s and 90s that the field established itself firmly in theoretical and applied linguistics (Pawley & Syder 1983; Peters 1983; Sinclair 1991; Cowie 1998). Before the 1980s « it was still possible to dismiss phraseology as a linguistic activity of only minority interest and with poor prospects of recognition as a level of language or of linguistic description » (Cowie 1998: 18).
  
  
• Both levels are important for the mastery of a language. As stressed by Odlin (1989), errors at the discourse level, in particular at the level of coherence very often have far- reaching consequences because they might lead to incoherent text production. A number of authors indicate that knowledge and understanding of discourse markers (in particular, connectives) are indicative of the degree of proficiency of a foreign language speaker (Geva, 1986; Goldman & Murray, 1992). Phraseological errors are universally recognized as those that most clearly distinguish native from non-native language, even at an advanced proficiency level. This aspect is also included in the very recent development of work aiming at the integration of these dimensions in foreign language teaching (Lee, 2002; Victori, 1999; Conrad, 1999; Nattinger & DeCarrico 1992).
  
  
• Even if it is not totally true that these domains have been left out of the scope of attention of FLL researchers, it is undeniable that these studies have tackled in the first place the levels of phonology, lexis and syntax (Benson, 2002; Odlin, 1989:23; Trillo, 2002). There is, however, a rapid development in this area (Flowerdew, 1998). Both the phraseological and the discourse level are the focus of renewed interest in contrastive studies (Evans, 1998). The contrastive analysis of Discourse Markers (DM) seems indeed to be the next point on the research agenda:
  
  Finally, how do DMs compare across languages? Some preliminary data [...] suggest that there is a general correspondence between the markers, but certainly not an exact mapping. If so, to what extent are they similar, to what extent different [...]? In the acquisition of a second language, which DMs are learned first, and is this influenced by the native language? (Fraser, 1999:950)
  
  Solid empirical studies of phraseology also figure high on the agenda of second language acquisition and foreign language teaching specialists. One of the great challenges for future research will be to determine the respective roles played by phraseological competence and creative skills in SL and FL learning and adapt teaching practices accordingly (Granger 1998).

• These two levels also make it possible to contrast the role played by a whole range of factors that are supposed to favour or hinder the adequate use of these linguistic devices. Some of these factors support the cross-linguistic hypothesis (real or perceived proximity between languages), others support the developmental hypothesis (linguistic proficiency at other levels of L2), others still are cognitive factors (cognitive load, reader-oriented writing, writing metaknowledge and strategies).

In addition to these reasons justifying the choice for the study of phraseology as well as of global coherence markers, a number of reasons motivate the study of these two levels within the same project.

• In the first place their joint study will make it possible to contrast the impact of cross-linguistic and developmental factors at a local level (phraseology) and at a global level (discourse structure markers). This is especially important as the discourse level is claimed by some authors to be especially transfer-prone. If we only studied the discourse level, the results could be biased. On the other hand, both the phraseological level and the discourse level will allow us to take into account an aspect of the learners' interlanguage which has hitherto been largely neglected: the role of the production context, be it immediate (the co-text, corresponding to the immediate syntagmatic environment of the studied words, which is crucial for phraseological studies) or larger (the entire text as advocated by the discursive approach).
  
  
• The second justification is methodological. The study of these two linguistic levels requires the recourse to methods involving different types of corpus analyses. This type of approach is extremely time-consuming if all the necessary corpora have to be collected from scratch, but time-saving if, as is the case in the current project, the research team already has many of the required corpora at its disposal and is ready to share them and collaborate, if necessary, to collect more data.