Time to pull out the stops: Spirantization in Pacific Northwestern English.

Abstract

Anecdotal evidence suggests that English speakers in North America are spirantizing stops. Spirantization is a consonant lenition process active in many languages, but has traditionally been assumed not to be active in English, except as a lexical or affixal process. This study is an investigation into the spirantization of stops using recordings of eight age/sex matched dyads (n=16) from a small corpus of Pacific Northwestern English speech. Representative monologues were extracted for each speaker and analyzed acoustically. Alveolars and voiceless aspirated stops were excluded from analysis as alveolar stops have a large number of allophones, and aspiration may block spirantization. The resulting tokens were rated for the absence of a stop burst, the presence of frication noise, and the presence of formant structure. Scores for each factor were then used to create a coarse‐grained spirantization score for each token, with lack of a release burst, and presence of frication or formants indicating incomplete closure. Preliminary results from four speakers (88 tokens total) show that spirantization is indeed occurring: 44% of the tokens were spirantized. Additionally, it appears that velars are much more likely to spirantize than bilabials: 77% of velars spirantized compared to 26% of bilabials.