Proceedings of the Annual Meeting of the Cognitive Science Society

The present study examines how natural language aptitude subskills predict individual differences in learning Python and Java. Past work has demonstrated that overall performance on the Modern Language Aptitude Test (MLAT), a standardized measure of language aptitude, is a strong predictor of both the speed and accuracy with which individuals learn Python. However, language aptitude is a broad multidimensional construct made up of individual subskills. In the present study, we examine how two of these subskills - sensitivity to form and meaning mapping - relate to programming outcomes in both Python and Java. Results indicate that both sensitivity to form (MLAT IV) and meaning mapping (MLAT V) are related to programming acquisition in both languages - this relationship remains even after controlling for fluid intelligence. We also examined how programming skills tied to semantics and syntax related between Python and Java in a subset of learners who learned both languages. These results demonstrated that proficiency in Python predicted individual differences in both syntactic and semantic knowledge in Java. Taken together, these results further elucidate the role of natural language aptitude in programming learning and suggest that semantic and syntactic content may transfer across programming languages.