Cognitive linguistics inspired mediation on L2 grammar instruction: The English preposition tutor

Abstract

English prepositions have always posed great challenges on second language (L2) learners due to their polysemous nature. Also, lexical semanticists (Quirk et al., 1985) often present such linguistic items as arbitrary, causing learners to heavily rely on memorisation. The present study explored the efficacy of Cognitive Linguistics (CL) inspired mediation on English preposition learning by adopting a computerized tutorial system called The English Preposition Tutor. In this context, mediation, as the main instructional intervention, was prefabricated and was explicitly delivered. Moreover, mediation was provided in the forms of animated schematic diagram. Applied cognitive linguists argued that schematic diagrams, with variation among studies, enabled L2 learners to visualize the conceptual representations of prepositions and to formulate a more accurate form-function mappings (Tyler, 2012; Boers & Lindstromberg, 2006). 77 year-one English majors were randomly assigned to one of the four groups. The three treatment groups differed on the type of mediation materials received: (a) schematic diagrams of the target and distractor prepositions (b) metalinguistic explanations and sentence examples of the target and distractor prepositions, and (c) no mediation yet with an indication of correctness after selection. The control group did training on English articles instead. All groups went through the same procedure of data collection: a pretest, training, a posttest, and a delayed posttest. Three prepositions (in, at, over) were trained, making up a total of twelve prepositional senses. Measurements included two tests: an untimed grammaticality judgment test and a picture elicitation test. The dependent variables are accuracies. ANOVAs showed that the cognitive group outperformed the traditional groups at the production test only. Furthermore, spatial cues were acquired substantially better than non-spatial cues at the production test whereas comparable gains were found at the processing test.