Polarity Markers in Hindko: Licensing Conditions and Corpus Evidence

Abstract

This article explores the distribution and licensing conditions of Negative Polarity Items (NPIs) in Hindko, an Indo-Aryan language spoken primarily in the Hazara region of Khyber Pakhtunkhwa (KPK) and parts of Punjab, Pakistan. Drawing on corpus data collected from naturalistic speech, written narratives, and structured elicitation tasks (N = 1,240 tokens), this study documents the full inventory of Hindko NPIs, analyzes the syntactic and semantic contexts that license their occurrence, and assesses the degree to which cross-linguistically established licensing conditions hold for this under-described language. The findings reveal that Hindko NPIs operate within a hierarchically organized licensing space involving sentential negation, questions, conditionals, and downward-entailing operators. Statistical analysis of corpus frequencies demonstrates that negation accounts for 68.4% of all NPI licensing contexts, with interrogative and conditional environments contributing 16.2% and 9.7%, respectively. Notably, Hindko exhibits several typologically unusual features, including a class of 'free-choice' items that appear sensitive to modal and generic contexts. The paper advances current theoretical accounts of NPI licensing by integrating evidence from a language peripheral to the mainstream South Asian linguistic literature