Questioning questions: The illusion of variation in African American English polar question intonation

Polar questions (PQs) in African American English (AAE) have often been described as having three distinct final contours-rising, falling and level. In this paper, I test the claims that variation in PQ intonation for AAE speakers from the Mississippi Delta can be predicted by i) syntactic form: presence or absence of auxiliary inversion or ii) truncation effects related to post tonic syllable number. Semi-spontaneous questions were elicited from AAE speakers in the Mississippi Delta that varied in inversion status as well as the segmental material needed to evaluate truncation effects. Ultimately, the presence/absence of inversion did not constrain question intonation, while post nuclear segmental material was a statistical predictor of final contour. Thus, Similar to Armstrong (to appear), the study demonstrates that the illusion of variation in question contours in AAE may be most attributable to truncation effects. Namely, I will propose that PQs in AAE have a consistent final low boundary tone (L%), with a realization that varies based on number of post tonic syllables. Therefore, the greater the distance from pitch accent to final target for a given question, the more faithful an approximation of the boundary tone will be realized.