Re - representing phonology : consequences of Q Theory ∗

Since the development of Optimality Theory (OT; Prince & Smolensky (2004)) and Harmonic Grammar (HG; Legendre et al. (1990)) in the 1990’s, phonological theory has focused on the grammatical constraints that conspire to produce output generalizations. In the 1970s and 1980s, however, the focus was on representations, both above and below the level of the segment. Traditional feature matrices gave way to the autonomous features of Autosegmental Phonology (Leben (1973, 1978, 1980); Goldsmith (1976), McCarthy (1981)). The ability of features to link independently to timing units permitted, among other things, the representation of contour segments: