Gap-Filling in English as L2 as a Form of Text Construction Using Contextual Cues

Abstract

The purpose of the study is to analyse the processes involved in gap-filling in English as L2, which may be considered to be a form of text construction. It is based, on the one hand, on models of the multilingual lexicon which assume that both syntactic and semantic properties are stored within lexical entries (Herwig, 2001), or more precisely, lemmas (e.g., Jiang, 2000), and, on the other hand, on models of reading and writing which involve both bottom-up and top-down processing (de Bot, Paribakht, & Wesche, 1997; Schoonen, Snellings, Stevenson, & Van Gelderen, 2009). The study, carried out with advanced (MA-level) students of English philology, consisted of two parts, a multiple-choice test, and a “free production” gap-filling task, in which the students could provide any words that, in their view, met the semantic, syntactic and collocational requirements of the text. The tasks were followed by a short retrospective questionnaire, aiming to investigate the reasons for the students’ lexical choices. As the results show, gap-filling is a very complex process, where excessive reliance on bottom-up information, disregarding the context, often results in errors, although errors can also result from paying insufficient attention to such local cues as syntactic and collocational restrictions. They also prove the importance of maintaining a coherent mental model (Johnson-Laird, 1983) of the text during comprehension, and of acquiring accurate lemma information from L2 input, instead of copying lemma information from L1 lexical entries (cf. Jiang, 2000).