The Use of Conventional Expressions ofThanking, Apologizing, and Refusing

Abstract

Different strands of pragmatics research have emphasized the importance of conventional expressions in the realization of speech acts. Cross-cultural comparisons of speech act realization by native speakers have caught our attention with titles derived from such culturally-based formulaic expressions as Poison to your soul (Coulmas, 1981) and May God increase your bounty (Bodman & Eisenstein, 1988). Even mundane expressions may be conventional, however, and Coulmas and others brought to our attention the role of conventional expressions as part of the social contract of communication. Coulmas (1981) describes conventional expressions as “tacit agreements, which the members of a community presume to be shared by every reasonable co-member. In embodying societal knowledge they are essential in the handling of day-to-day situations” (p. 4). The construct of “reasonable co-member” of a community is a natural link to second language pragmatics, as learners of the L2 may see themselves or be seen by others as joining a new speech community through language acquisition. Second language researchers are also interested in conventional expressions, although as often from the perspective of general L2 development as from L2 pragmatics (BardoviHarlig, 2006). In fact, Granger (1998) even attributes the research in pragmatics as one impetus to study formulas, observing that “the formulaic nature of many pragmalinguistic rules has necessarily contributed to bringing the study of prefabs to the fore” (p. 145). Although the use of conventional expressions has garnered significant attention from sociolinguisticallyand comparatively-oriented investigations in L2 pragmatics, there has been relatively less attention paid to conventional expressions from acquisitionally-focused research in L2 pragmatics. Early studies include Scarcella (1979) which examined the production of conventional expressions on a written DCT; more recent studies include Kecskes (2000, 2003) which investigated the interpretation of transparent and idiomatic formulas, Roever (2005) which examined the identification of formulas in appropriate settings, Bardovi-Harlig (in press) which explored the recognition of conventional expressions as pragmalinguistic resources, and Bardovi-Harlig et al (in press) which reported on the influence of the use of the expressions on length of learners’ turns. The strong cultural associations of conventional expressions as well as their highly consistent linguistic composition makes conventional expressions a natural area of exploration for L1 influence. This paper explores the influence of first language and level of development on the use of conventional expressions in the realization of three speech acts, namely, expressions of gratitude, apologies, and refusals. Using an experimental approach, we have sought to reproduce the conditions for the use of conventional expressions. We employed a computer-delivered aural discourse completion task (DCT) with timed presentation of scenarios and a recorded interlocutor to simulate turn taking, and through piloting have identified scenarios that promoted high use of conventional expressions by native speakers. With the exception of a few studies, the design of most investigations of first language influence compares the production of a single L1 group to the production of native speakers of the target language. The four main language groups in this study—Arabic, Chinese, Japanese, and Korean— permits the investigation of the influence of L1s across languages, following the research design for studies of L1 influence proposed by Jarvis (2000). Only a few interlanguage pragmatics studies have multiple L1s (Bodman & Eisenstein, 1988; Eisenstein & Bodman, 1986; Hinkel, 1994; Park & Nakano, 1999); without such studies, it is difficult to tell which characteristics of L2 pragmatic development are unique to specific languages or cultures, and which are shared by all learners.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 16 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.