Wim Gombert

CHAPTER 6. Chunks in writing 87 INTRODUCTION In the previous chapter on writing skills, the SB and DUB groups did not di er much in writing skills a er 6 years of instruction. However, the DUB students wrote longer sentences and longer texts, which could be related to uency. ese ndings were in line with Piggott et al. (2020) who found that the learners with implicit instruction and more L2 exposure wrote longer sentences and texts. She also examined the number of chunks used in her writers’ text and she found that implicitly taught students excelled in outcome measures such as length and formulaic sequences (chunks). Also, in chapter 4 (on reading and listening skills), it was argued that the mastery of chunks can facilitate language processing speed, as chunks are readily available as a result of repeated exposure (Gustafson & Verspoor, 2017), and once automatized, chunks can form shortcuts in processing (Pawley & Syder, 1983; Wray & Perkins, 2000). e aim of the present paper is to explore the use of chunks in the texts of learners a er six years of an SB or DUB program. Taking a usage-based perspective (Ellis & Cadierno, 2009), we expect that the learners’ use of chunks will re ect the amount and kind of input they are receiving. We expect the DUB learners to use more chunks overall and the SB students to use a more limited range of chunks. We will also explore the relationship between the use of chunks and syntactic complexity to see if indeed the use of chunks facilitates longer, more complex sentences. LITERATURE As Sinclair (1991) has pointed out a substantial part of a native speaker’s language consists of a wide range of conventionalized expressions (Wray, 2002; Ellis, 2008b) which we will refer to as chunks. Because chunks are so pervasive in a native-like repertoire, they are also a crucial aspect of L2 development. Chunks contribute to uency and authenticity of L2 use and may also speed up L2 development (Gustafsson & Verspoor, 2017). From a usage-based perspective, L2 development is primarily driven by frequency and salience of structures in the surrounding input (Ellis & Cadierno, 2009), which we will refer to as L2 exposure. Chunks have been labelled di erently throughout the literature, including, but not limited to, “chunks”, “formulaic sequences”, “conventionalized ways of saying things”, and “multiword expressions” (Forsberg, 2010; Myles, 2012; Verspoor et al., 2012; Wray, 2000), all considered form-use-meaning mappings (FUMMs) by Verspoor (2017). According to Wray (2000), chunks are preferred ways of expressing certain concepts within speech communities, resulting in extensive use and they may therefore contribute to uency and authenticity of L2 use. e following characteristics are repeatedly cited