Locality at the Interface

There are many approaches to locality, which differ in how they answer a central question: where do locality constraints originate? All available theories successfully explain specific subsets of the data, yet none captures the full range of locality phenomena. This suggests a “division of labour” among locality accounts, whereby variation in locality effects (within and across languages) is partitioned into distinct explanatory domains. This observation raises two possibilities: either “locality effects” comprise ontologically distinct phenomena stemming from heterogeneous sources, or there is a common denominator linking phenomena such as strong and weak islands, superiority effects, and subject–object asymmetries in A'-dependencies. If the latter is correct, a fundamental question arises: why has no unified account emerged? If these effects share an underlying basis, one might expect a unified theoretical framework to capture them. The persistence of this gap remains largely unaddressed in the literature. This thesis addresses this issue by developing a framework that seeks to explain why locality effects are multifaceted and resistant to a single overarching explanation. The core proposal is that locality effects arise from the interaction between purely linguistic factors and cognitive dynamics that extend beyond the language faculty. Most (if not all) locality-related phenomena are argued to result from the interplay between cognitive mechanisms responsible for language processing and the grammatical representations underlying them. This interaction is located at the Interface, understood as the point of interaction between grammar (conceived as a set of mental representations and their combinatory rules) and the domain-general principles and mechanisms through which language emerges and is put to use (Chomsky et al. 2023). In pursuing these questions, the dissertation offers novel insights into specific cases, including argument–adjunct asymmetries in extraction from English weak and strong islands and intervention effects in Spanish object relative clauses, both investigated experimentally. In the final chapter, the long-standing issue posed by the (subject part of the) CED is addressed theoretically.