Tonal alignment is contrastive in falling contours in Dinka

This study investigates a contrast in tonal alignment, involving falling contours in Dinka. Such a contrast, first reported in Andersen (1987), calls into question the assumption that tonal alignment cannot distinguish contour tone patterns of the same shape within the syllable domain (Odden 1995:450, Silverman 1997:479-480, Yip 2002:29). A qualitative description of the Dinka phenomena is followed by a detailed production study, in which the realization of the contrast is investigated under manipulations of time pressure (cf. Caspers & van Heuven 1993). The primary correlate of the contrast is indeed tonal alignment: the early-aligned fall sets in during the onset or early in the vowel; the late-aligned peak sets in well into the vowel. The time pressure study also suggests that it is unlikely for more than two patterns of alignment to be accurately produced and perceived, given various phonetic limitations (Xu & Sun 2002; ’t Hart, Collier & Cohen 1990; House 2004). On this basis, it is argued that a representation with reference to mora edges (Prieto, D’Imperio & Gili Fivela 2005; Morén & Zsiga 2006) is less attractive: this representation is not sufficiently restrictive. Instead, the contrast is represented using a binary feature [±late-aligned]. This representation is adequate in an explanatory sense, in that the category boundary is in line with the quantal threshold hypothesized in the model of pitch perception of House (1990). The results also strengthen the case for three-level vowel length in Dinka (Andersen 1987, Remijsen & Gilley 2008).