Dialect variation in East Norwegian tone

In recent analyses of Scandinavian tonal accent, two competing hypotheses can be identified, the privativity hypothesis and the timing hypothesis. The more widespread is the privativity hypothesis, which assumes that the difference between the accents consists in accent 2 having an initial, lexical tone that is absent from accent 1. Otherwise the melodies are identical. Due to greater, structural complexity, accent 2 is furthermore regarded as the marked member of the contrast. The timing hypothesis on the other hand assumes different timing of identical melodies to be the essence of the contrast. In this chapter, I argue that when East Norwegian dialect data are taken into consideration, the timing hypothesis appears as the only one that allows for a unified analysis of the differences between the three dialects discussed. At the input level, the three dialects are assumed to have the same melody. The well established privative pattern of Urban East Norwegian (UEN) emerges as a surface phenomenon that is the result of a constraint that bans low tones from metrical heads. The surface timing patterns of the two other dialects, North Gudbrandsdal and Oppdal can be accounted for by minimal adjustments of the UEN grammar. A supplement to this article, containing OT-analyses of the three dialects discussed in the paper, is available in this BORA collectio