Socially Driven Variation in Noun Insertions: Pluralised English-Origin Nouns in Spoken and Written Hindi

Abstract

The status of insertions in the context of language mixing has been debated for some decades. This paper contributes new data by investigating the language of plural marking on English-origin nouns in a Hindi matrix in written and spoken Hindi–English in India. The central question was whether insertions should be regarded as loanwords or code-switches. Google search results were tallied out for Hindi- and English-plural-marked English-origin nouns in Devanagari script. A spoken corpus of YouTube interviews of Bollywood personalities was also analysed quantitatively. Over 60 common English-origin words were analysed in detail through Google searches. The Bollywood corpus consisted of interviews with 28 male and female speakers, and contained over 140,000 words. Basic statistical analyses in the form of chi-square tests were carried out to compare the distributions of tokens of interest. The Google searches showed that English-origin words fell into three categories, based on the relative preference for Hindi or English plural marking. The category for which Hindi plurals were dominant clearly contained established loanwords, but the other two categories showed characteristics of both loanwords and code-switches. The Bollywood interviewees showed an overwhelming preference for English plural marking on English-origin nouns in a Hindi matrix, and hence for code-switching as the primary strategy for single-word insertions. This is the first attempt to disambiguate loanwords and code-switches in Hindi–English by combining online usage patterns with conversational data from a spoken corpus. Generalisations along the lines of ‘English-origin word x in a Hindi matrix is a loanword/code-switch’ or ‘Hindi-English bilinguals incorporate English insertions into a Hindi matrix as loanwords/code-switches’ are unhelpful, and likely to be grossly imprecise given the large and internally variable bilingual population of India. Future research should take into account key sociolinguistic variables and the nature of the speech situation.