The acquisition of the perfective/imperfective aspectual distinction in French : output-based instruction vs. processing instruction

Abstract

The effect of grammar instruction on second language acquisition continues to be a source of debate in SLA research. Previous studies have shown that input-based instruction such as Processing Instruction (PI) is more effective than traditional grammar instruction (TI) for the acquisition of grammatical structures such as object pronouns in Spanish, the Spanish preterite, the simple present vs. the present progressive in English, and the ser/estar contrast in Spanish (VanPatten and Cadierno, 1993; Cadierno, 1995; Buck, 2000; Cheng, 2004). This quasi-experimental, classroom-based study examines the effects of output-based instruction (OB) and PI on the acquisition of the perfective/imperfective aspectual distinction in French, shown to be difficult to master by English-speaking learners due to its linguistic and pragmatic complexity. Specifically, the research design investigates whether OB instruction and PI have significant effects on the learners' performance involving the interpretation and production of the passé composé and the imparfait in narration. Two second semester university level French classes at the University of Texas at Austin were randomly assigned to one of the two groups: an output-based instruction group (n=18) and a processing instruction group (n=17). A distinct instructional treatment was developed for both groups, and a pretest/posttest procedure was used to assess the effect of instruction. The two posttests were administered one day and one month after instruction. All tests except the pretest included a written interpretation task, a controlled written production task, and a written composition. The findings show that both the OB and the PI groups improved their performance significantly on the assessment tasks and that there was no statistical difference between the groups on any of the tasks. These results differ from those of previous PI research and suggest that either type of instruction, output-based or processing, may have had a significant impact on the learners' developing system. The type of output-based instruction used in this study differs from TI in that it does not include a mechanical component. The results of the present study are consistent with Farley (2004b) in showing that approaches to grammar instruction that are meaning-oriented may bring about significant effects on SLA.