Cross-Linguistic Variations in Word-Final Position: The Parametric Hierarchies, Connections and Networks

Abstract

Word-final position is widely recognized as a structurally weak and restricted domain, yet languages differ strikingly in how they regulate segments and clusters at the right edge. While some systems categorically prohibit final consonants, others allow only a subset of segments, and still others impose process-based adjustments such as final devoicing. This study proposes that such typological variation can be systematically captured through parametric hierarchies and networks. Drawing on a broad cross-linguistic sample-such as Italian, Japanese, Malayalam, Spanish, English, French, Polish, Dutch, Vietnamese, Turkish, among others-we argue that three macro-parameters-Final Onset, Branching Rhyme, and Devoicing-provide the structural backbone of word-final phonotactics. These macro-parameters project meso- and micro-settings that regulate segmental classes and positional distributions. Crucially, we introduce floating sub-parameters, which connect macro-domains and generate implicational patterns, explaining why certain restrictions (e.g., final devoicing) arise only under specific configurations. By treating phonology not as a set of isolated parameters but as an interconnected network, the model captures both robust cross-linguistic generalizations and language-specific outcomes. In this way, the study extends parametric theory beyond linear hierarchies and offers a unified framework for understanding the organization of phonological variation at word edges.