A Sociolinguistic Corpus Study of Allophonic Variation in Punjabi Language

Abstract

This study examines allophonic variation in the Punjabi language using a purpose-built sociolinguistic speech corpus. The corpus consists of 1008 transcribed and annotated tokens of allophony from 72 respondents from beauty parlors representing high, middle and lower class income in the capital city of Pakistani Punjab. To reveal social stratification (e.g., class, gender and age) in the use of prevocalic and postvocalic, 14 Punjabi lexical items were finalized. A quantitative approach was applied to address the research question and chi-square test was applied to investigate the association between social stratification and the use of allophones. The respondents were found using alternative pronunciations of the same lexical items, with no change in meaning (except [raaz]/[raaj], which are distinct words). Variationist sociolinguistics has documented class, gender and age stratification for English; however, there is scarce comparable evidence for Punjabi, a major South Asian language. Social class has also not been operationalized through consumption in an actual service setting. This study bridges these two gaps by creating a speaker-annotated corpus, and pairing it with a beauty-service spending index of people of varying social classes. The findings suggest that there are social meanings carried by allophones of Punjabi and that spending on beauty services may be a useful proxy for class in this community.