The Person Case Constraint: Unconditional Interfaces and Faultless Derivations

This thesis advances a theoretical move toward a grammatical model devoid of interface conditions by proposing a novel feature-based structure-building mechanism. In the standardly assumed architecture of grammar, ungrammaticality is often explained in terms of a violation of some condition on an output of the syntactic module. However, some recent research in linguistics has attempted to move away from such an approach to ungrammaticality and proposed to reinterpret ungrammaticality as non-generability. In this approach, ill-formed structures are construed not as defective but as impossible to generate with the available syntactic operations. In order to advocate for the latter approach to ungrammaticality, this thesis examines an interface condition called the Person Licensing Condition (PLC), which was proposed to account for a linguistic phenomenon known as the Person Case Constraint (PCC). It is shown in the thesis that the PLC fails to capture cross-linguistic variation in the PCC patterns and in the way illicit structures are remedied. It is further argued that previous, Agree-based accounts of the PCC variation, also reliant on an interface condition, cannot fully explain all the patterns and that they obscure the source of the variation. This thesis proposes an alternative account of the PCC, which involves a version of Merge that is constrained by feature valuation, and what is known as articulated person features. The proposed syntactic system allows arguments to engage with each other by valuing the same unvalued feature, and this explains person restriction effects seen in the PCC patterns. Illicit argument combinations are underivable in the system, and an interface condition is not needed to account for ungrammaticality induced by the PCC. This novel account of the PCC also captures the variation fairly well, and because the variation is attributed to different ways syntactic elements are featurally specified in different languages, the reason why we obtain such variation is clear. In defending the person features used in the proposed syntactic mechanism, this thesis also speaks to the typology of grammatical persons and shows that the same features explain not only syntactic restrictions like the PCC but also a variety of morphological neutralization phenomena.Ph.D