Non-Vowel Phonology

Abstract

That there is probably more comment upon pronunciation alternants involving ti and du in items like action and duke in the eighteenth century than in the two preceding, would seem to suggest that an active and salient phonological change is taking place. Indeed, some early-eighteenth-century observers not only see sociolinguistic significance in innovative pronunciations in this area, but see them as much more advanced in the lexicon than do others. The changes under discussion here typically involve voiced and voiceless dental obstruents in syllable-initial position, preceding a vowel or semi-vowel segment with a high F₂ (palatal) configuration — [i]/[j]. In such palatal environments the change itself seems to have a diachronic sequence something like [ti] → [tj] → [t∫] → [∫], alongside the corresponding [dj] → [dʒ] → [ʒ]. It would appear that a two-element unit ultimately becomes perceived as a simplex. The progress of the changes is, however, phonologically quite transparent. We see a movement away from a segment characterized by a relatively high degree of obstruency to one where a more vowellike configuration — in terms, at any rate of formant structure prominence — is produced. The linear combination of obstruent and palatal [i] vowel comes to be perceived and interpreted as a single segment, whose internal structure is composed of precisely these two units. The segment so produced is more ‘vowel like’ (i.e. the palatal vowel element is perceived as the more prominent in the ‘mix’), resulting in a segment higher up on the sonority scale — i.e. some kind of fricative.