Modelling Intonational Variation in Swedish: Two reports from the SIMULEKT project

Our object of study in the research project SIMULEKT (Simulating Intonational Varieties of Swedish) (Bruce et al., 2007) is the prosodic variation characteristic of different regions of the Swedish-speaking area, shown in Figure 1. The seven regions correspond to our present dialect classification scheme. In our work, the Swedish prosody model (Bruce & Garding, 1978; Bruce & Granstrom, 1993; Bruce, 2007) and various forms of speech synthesis play prominent roles. Our main sources for analysis are the two Swedish speech databases SpeechDat (Elenius, 1999) and SweDia 2000 (Engstrand et al., 1997). SpeechDat contains telephone-transmitted speech from 5000 speakers, registered by age, gender, current location and self-labelled dialect type, according to Elert’s (1994) suggested Swedish dialect groups, which is a more fine-grained classification with 18 regions in Sweden. The dialect project SweDia 2000 collected a word list, an elicited prosody material, and extensive spontaneous monologues from 12 speakers (younger and elderly men and women) each from more than 100 different places in Sweden and Swedish-speaking parts of Finland.