Josef Fruehwald (2012)
In this talk, I will argue that it is possible to determine whether the influence of one segment on another is phonetic or phonological by examining how they interact during sound change. I will begin with an introduction to the Philadelphia Neighborhood Corpus, which has been developed at Penn. It represents both the cumulative effort of fieldwork since the 1970's and the most recent developments in automated acoustic analysis (Labov & Rosenfender, 2011). I will then briefly present some of the initial results from the corpus, specifically the reversal of some sound changes which were originally identified as new and vigorous in the 1970's (Labov, Rosenfelder & Fruehwald, forthcoming). The remaining presentation will focus on the interrelation of phonology and phonetics, and how this is affected by sound change. Specifically, some sound changes are conditioned, or favored in certain linguistic contexts. Whether this conditioning is phonetic or phonological is an open question, and is another in instantiation of the general question of whether the influence of one segment on another is phonetic or phonological (e.g. Cohn 1993; 2007). I argue that a comparison of the rates of sound change across contexts is an important diagnostic for determining whether and when conditioning of a sound change is phonetic or phonological. This methodology is actually borrowed from historical syntax (Kroch 1989), but requires modification in statistical practice for the cases of sound change examined here. Most interestingly, I will argue that it is possible to answer the question 'is it phonetic or phonological' with diachronic data that would have been impossible otherwise.