Frisian accents in Dutch:Exploring perceptual salience and attitudes towards prosodic traits

This study explores the perception and indexicality of prosodic features in spoken Dutch. Specifically, it investigates whether intonational, timing and voice quality features that occur in the regional variety spoken in the province of Fryslân index regional belonging to listeners from the area. Linguistic attitudinal research assumes that existing attitudes towards a specific social group can be evoked through language. Irvine (1996) argues that linguistic features can hold iconicity, i.e. that attitudes towards a group can be transferred to features themselves, subsequently allowing these features to evoke attitudes without group identification. Similarly, features that have previously held no social meaning may become meaningful through a change in indexical order (Silverstein, 2003). This study investigates whether prosodic features identified by language professionals as ‘Frisian’ sounding are sufficiently perceptually salient to invoke a local social meaning. This is accomplished through an investigation of folk perception towards Dutch spoken by a Frisian native speaker. Frisian is a minority language in the Netherlands spoken by approximately 450.000 (bilingual Dutch) speakers in the province of Fryslân. Whilst a salient Frisian accent in the Dutch variety spoken in the area is often commented upon, it is a previously unexplored topic of variationist sociolinguistic research. The prosodic features nasalisation, stress placement (reciprocal pronouns and genitive compounds), and sentence intonation (in declarative, interrogative and imperative constructions) are included in this study, which contains two experimental tasks. The first task consists of recorded Dutch sentences containing these Frisian features, presented to participants in a pair-wise forced choice experiment answering the question “which sentence sounds more Frisian?”. The outcome of the 45 forced choices is a hierarchy of perceptual salience. The second task is an adaption of the matched-guise task and provides validation of the features’ perceptual salience alongside attitudinal information. In this task, three recordings of each feature are manipulated to match the Standard Dutch pronunciation, resulting in two stimuli sets: one with Frisian features and one without. The resulting 36 stimuli are presented individually to the participants. For each stimulus, the participants answers the question “was this sentence spoken by a native speaker of Frisian?” on a 6-point Likert-scale. Additionally, participants provide judgments of the speaker’s intelligence, friendliness, attractiveness, strangeness, wealth and modernity for each stimulus. A stronger difference between attitudes than between ‘Frisianness’ in both stimuli sets is indicative of iconicity of the prosodic features. Although the study is currently ongoing, preliminary results suggest that participant background strongly impacts both perceptual salience and attitudes towards Frisian and its speakers. By taking into account the social background of listeners, the results of this study contribute to sociolinguistic theory on the development of indexicality of linguistic features. It also yields interesting insights into the differences in salience between voice quality features, intonational features and timing features within one variety of speech.