Abstract
Since the rise of generative grammar, one of the central aims of linguistic theory has been to identify the constraints that define the space of possible grammars. This talk focuses on one such class of constraints: islands, encapsulated syntactic domains that block movement and yield ungrammaticality when extraction is attempted. Far from being marginal anomalies, island effects reveal the boundaries of the language faculty’s computational capacity. Drawing on findings from approximately 40 experimental studies in English and Italian, I present findings that broaden the empirical landscape of island effects and challenge traditional assumptions. I conclude by exploring the theoretical implications of these results for the architecture of grammar.
