Current research issues

The phonology of InSAfE is still open research territory. I shall concentrate on the possibilities offered by the study of aspiration. P, T, K have aspiration patterns that are different from the prototypical English patterns of aspiration in all initial positions. Detailed research has still to be undertaken, and a preliminary analysis suggests the following in vernacular mesolectal speech:

P is always unaspirated before /ɑ:/ , /ɔ:/ , /ʊ/ , /ɒ/ , /eɪ/ , /oʊ/ and /εə/. Thus park, pork, put, pot, pay, poke, pair all have unaspirated initial P. Likewise P is always unaspirated before /r/ and /l/, e.g. in pray and play. This means that /r/ and /l/ are voiced in InSAfE in contrast to many varieties of English in which the aspiration on initial consonants causes /r/ and /l/ to become voiceless. In all other contexts whether P is aspirated or not, depends on the particular word. Taking P before /e/ as an example, the following words always have aspiration – pen, pebble, pet; whereas penny, pepper, petal, peck are always unaspirated. It has still to be researched whether there is intra-speaker variability (i.e. pronouncing the same word differently) or variation across speakers. Speakers who produce aspiration invariantly with initial P, T, K would be judged as putting on a ‘Speech and Drama’ accent. The dialect has minimal pairs like p^hea and pee; p^hiece and piss (pronounced [pi:s]). It also has near-minimal pairs like p^het and petal, p^hen and pencil.

Similar principles apply to T and K. The reason for this unusual system is two-fold. Firstly it represents a shift from languages with differential patterns of aspiration towards the general English norm. The Indic languages have phonemic distinction between aspirated and unaspirated P, T, K. Speakers appear to be comfortable with the categorical absence of aspiration in some words and its categorical presence in others. On the other hand, as the Dravidian language, Tamil, does not have aspiration, its speakers have to adopt this feature afresh in their English. The InSAfE mesolect seems a happy compromise between the two systems: no aspiration before certain back vowels, certain diphthongs and both liquids; and in all other contexts aspiration is word-dependent. The actual minimal pairs are marginal: both pee and piss cited above are, in fact, taboo words, and therefore do not occur in the same register as p^hea and p^hiece. The second reason for this unusual system is that it is probably a stage in the language acquisition-cum-lexical diffusion process. It is not hard to envisage a gradual shift to a system with aspiration for all initial P, T, K.