This study compares L2 Spanish learning after inductive, consciousness-raising tasks vs. deductive explanations in three U.S. high schools. Both groups improved significantly on picture description and grammaticality judgment assessments, but the deductive group made modestly stronger judgment gains. Implications for the accessibility of useful explicit L2 knowledge will be discussed.
This study compares L2 Spanish learning following inductive consciousness-raising tasks vs.deductive teacher explanations in three U.S. high schools. Participants came from eight intact, third-year classes, with two in each school experiencing three 90-minute lessons on the pronoun se containing either deductive teacher explanations (n = 48) or consciousness-raising tasks (n = 50). Two other classes in one school did unrelated work as a control group (n = 30). Deductive instruction provided ready-made rules for se, followed by sentence-level examples. Learners then identified tokens of se in narrative texts. Following Adair-Hauck and Donato (2016), the consciousness-raising tasks first had learners interpret the meaning of each text before having their attention drawn to the target structure. Learners then proposed rules for se in small groups before being guided toward a consensus through whole-class, dialogic interaction. Identical communicative tasks followed in both groups for the remaining 45 minutes. At issue is the impact of explicit, inductive problem-solving over grammar on L2 learning, as advocated in both social and cognitive L2 theories. Quantitative assessments included written picture description and grammaticality judgment tasks administered as pre-, post-, and delayed posttests, while qualitative data was gathered by recording all whole-class interactions and 3-4 sets of volunteers during small-group work. The results indicate that although both groups improved significantly in uses of se after instruction, the deductive group made modestly stronger gains on their grammaticality judgments. Nonetheless, both groups overgeneralized se to contexts that shared some of its semantic properties, while also exhibiting L1 transfer errors. Excerpts from the qualitative data suggest that, in all three schools, the deductive format may have more efficiently resolved learners’ uncertainties over se within the available time. Implications for the accessibility of explicit grammatical knowledge for language use and the implementation of deductive and guided inductive instruction will be discussed.
