
Full text loading...
This multifactorial corpus-based study focuses on verb-complementation constructions (Marcus started to draw a picture vs. Marcus started drawing a picture) and contrasts 3,119 occurrences of gerundial and to-infinitival constructions across native and non-native (ESL) English varieties. Using logistic regression modeling, I analyze how grammatical contexts constrain the syntactic choices of American and Hong Kong English speakers. The regression model reveals a level of complexity in the uses of gerundial/to-infinitival complements that had so far remained unnoticed. Specifically, speakers make syntactic decisions based on comprehensive grammatical contexts rather than single isolated semantic parameters (as previously reported). Further, for the two types of speakers different grammatical features play an influential role in the association of a given predicate with a particular complement type. This suggests that the two speaker populations do not share the same abstract knowledge of the semantic and morpho-syntactic constraints associated with each of the investigated type of complementation. Ultimately, this study shows that combining cognitively oriented theoretical frameworks with rigorous empirical corpus approaches helps distinguish what motivates native and ESL speakers’ syntactic choices.