Open Collections

Everest, Terri Josephine

University of British Columbia

Words are particular about the company they keep, so educators have long endeavored to crack the collocational code, accelerating SLA. This endeavor is supported by powerful digital tools enabling corpus analysis. Many academic genres have thus been analyzed, phrases linked to moves (functions), and writing roadmaps developed. However, fewer professional genres, notably the job cover letter, have garnered attention. Studies in this “occluded genre” (Swales, 1996, p. 46) usually feature small corpora by EFL writers sharing one L1. Instead, this cross-sectional study compares three English cover letter corpora in an ESL environment: a balanced corpus with samples from books; another NES one; and a third by Canadian immigrants with diverse L1s. After a pilot study, three questions were posed. One categorized move fulfillment by genre-specific verb-noun combinations (VNCs); another identified learners’ lexico-grammatical VNC errors; and a third quantified fulfillment by VNCs versus other lexical means, also identifying obligatory versus optional moves. CLAWS4 POS-tagged corpora, analyzed in Excel (and secondarily AntConc) using Sinclair’s node-collocate approach (1991) with a 4-4 span for VNCs. Lemmatized VNCs were extracted semi-automatically; MI and t- scores determined significance (Hunston, 2002), with t- scores ranking VNCs. Using Swalesian and Bhatian ESP genre analysis, VNCs were categorized into eight moves, adapting Upton and Connor’s (2001) framework. Error analysis addressed verbs, nouns, and other infelicities, paralleling others (e.g., Nesselhauf, 2005). VNCs with (borderline) significance, whether genre-specific or another category (e.g., field-specific) semi-automatically populated Excel sheets in a case (letter) by variable (move) format. Moves unfulfilled by genre-specific VNCs were examined using “intuitive notion of topic” (Brown & Yule, 1983, p. 69) to determine which other means, including structural elements, were instrumental. Regarding outcomes, learners employed more colloquial VNCs in moves. Determiner, verb tense/aspect, and preposition errors predominated. While over half the learners’ genre- specific VNCs were correct, many were fixed phrases. VNCs were integral in move fulfillment; other means still complemented or replaced them. More moves appeared obligatory than optional and some evolving. Learners rarely fulfilled some, perhaps lacking genre awareness. This study concludes with pedagogical recommendations. Future avenues are addressed and facilitated by my corpora, archived at cIRcle.

Text

2022-07-20

Attribution-NonCommercial-ShareAlike 4.0 International

http://hdl.handle.net/2429/82175

Interdisciplinary Studies

University of British Columbia

UBCO

http://creativecommons.org/licenses/by-nc-sa/4.0/