Plenary Speaker

Jonathan Harrington

Institute of Phonetics and Speech Processing, Ludwig-Maximilians University of Munich, Germany

Abstract

The dramatic effect of sound change on phonological categories is immediately apparent in considering both the often opaque relationship between orthography and pronunciation and, more recently and closer to the venue of the present conference, the clockwise re- organization of the front lax vowels in New Zealand English in the last fifty years. There is now also considerable evidence that phonological change arises from synchronic variability that exists at a number of different levels between a speaker and hearer, but the mechanisms by which the continuous fine-grained variation in speech communication give rise to categorical sound changes are not entirely understood. A related issue is that on the one hand sound change seems to take place incrementally at a such a slow rate that much of it is quite possibly not noticeable: this would seem to be one of the reasons why sound change is unlikely to be teleological or planned. But at the same time a lack of teleology seems to be at variance with the evidence that the outcome of so much sound change, such as the diachronic chain shifting of New Zealand English vowels, is typically quite systematic. I will discuss, partly with reference to recent sound changes in Standard English, the extent to which this dichotomy is resolved firstly in John Ohala's model, in which listeners misinterpret coarticulation, and secondly in recent exemplar approaches to speech that account for sound change as incremental shifts to categorical generalizations across exemplars stored in the lexicon. In the final part of my talk, I will discuss the relationship between incremental sound change and some recent experiments demonstrating not only that speakers imitate each other often unwittingly, but also that these unintended imitations may well be regulated by social factors, even though imitation seems to be below the level of the speaker's awareness. My general conclusion is that we may begin to resolve some of these paradoxes within self-organizational models in which both phonological stability and categorical change can be shown to be properties that emerge out of the fine-grained variability of speaker-hearer interactions.