The prosody of Swedish underived nouns: No lexical tones required

Abstract

This paper provides a detailed representational analysis of the morpho-prosodic system of underived nouns in a dialect of Swedish. It shows that the morphology, stress and tonal patterns are not as complex as they first appear once the data are looked at in sufficient detail. Further, it shows that the renowned Swedish "lexical pitch accent" is not the result of lexical tones/tonemes. Rather, Swedish is like all other languages and uses tones to mark the edges of prosodic constituents on the surface. "Accent 2" occurs when tones mark the edge of a structural uneven trochee (i.e. recursive foot) and "accent 1" occurs elsewhere. This analysis is counter all other treatments of North Germanic tones and denies the almost unquestioned assumption that there is an underlying tone specification on roots and/or affixes in many North Germanic varieties. At the same time, it unifies the intuitions behind the three previous approaches found in the literature.