Rethinking Locative Inversion and Pseudo-locative Inversion: EPP Satisfaction as the Key to Polarity Focus and VP Ellipsis

Abstract

This paper investigates a fundamental asymmetry between Locative Inversion (LI) and Pseudo-locative Inversion (PI) constructions in English: only PI constructions are compatible with VP ellipsis, despite their structural similarities. Previous accounts have struggled to explain why PI allows polarity focus and VP ellipsis, whereas LI does not. This study offers a novel explanation by demonstrating that the key distinction lies in how the Extended Projection Principle (EPP) is satisfied, which shapes both the syntactic structure and information structure of these constructions. Specifically, it argues that while PI constructions fulfill the EPP through the insertion of the expletive there in Spec-TP, allowing TP to function as a complete predicative domain and enabling polarity focus, LI constructions rely on Edge Feature (EF)-driven movement. In LI, the locative phrase temporarily occupies Spec-TP before moving to the left periphery, leaving the predicative domain incomplete and blocking polarity focus. Empirical evidence from various inversion constructions, including Negative Inversion, So-Inversion, and Comparative Inversion, supports this analysis by demonstrating that filling Spec-TP is essential for establishing a subject-predicate relation, enabling polarity focus, and licensing VP ellipsis. The findings contribute to the broader theoretical understanding of how syntactic derivations influence information structure, challenging discourse-driven explanations. This research underscores the syntactic basis of polarity focus licensing and offers a unified account of LI and PI constructions within the Minimalist framework.