|
The Prosodic Morphology of Swedish| old_uid | 9727 |
|---|
| title | The Prosodic Morphology of Swedish |
|---|
| start_date | 2011/02/28 |
|---|
| schedule | 10h-12h |
|---|
| online | no |
|---|
| location_info | D 143 |
|---|
| summary | Swedish unmarked prosody spectacularly comes to the surface in hypocoristic word formation, whereby long (Kata'rina) and short ('Bo) names are turned into disyllabic forms ('Kattis, 'Bosse) with stress on the initial syllable. In this presentation I look at the less spectacular but much broader effects of prosodic wellformedness in the lexicon, relating to both the stress system and the distribution of tone accent.
The stress systems of the Germanic languages in general tend to be analyzed as phonological, e.g. based on a phonological routine for stress assignment (an algorithm), but with a certain morphological influence expressed as properties of suffixes (cohering/non-cohering) or final syllables (extrametrical/or not). Despite the generalizations one can express via a stress algorithm (e.g. predominantly penultimate stress), there is always a residue left, which requires diacritic marking.
I’ll argue instead that much of the stress system, at least for Swedish, is specified directly in morphemes. Morphemes are tonic, pretonic or posttonic, which amounts to metrical specification, either directly (many roots, suffixes and prefixes are tonic), or as subcategorizations (pretonic prefixes, posttonic suffixes). In addition, there are prosodically unspecified morphemes (all kinds), which get stress by a phonological routine (stress rightmost). This allows us to better understand how morphemes (etymologically native or not) combine or don’t combine in the derivational
lexicon. This approach makes Swedish stress look much more like the accentual systems of Greek, Japanese and Basque, which exhibit dominance relations between morphological specifications of roots and suffixes.
Swedish also has a marginal lexical tone accent distinction, which is closely tied to the stress system, historically and synchronically. Here, too, the traditional analysis is to treat so-called ‘accent 2’ as due to a phonological generalization (polysyllabicity). I will show instead that accent 2 is, and must be, lexical (in simplex forms) and that it fits right into the lexical generalizations found for the stress system. |
|---|
| responsibles | Copley |
|---|
| |
|