Nanna Haug Hilton

Syllable reduction and articulation rates in Danish, Norwegian and Swedish

This investigation compares articulation rates of phonological and phonetic syllables in Norwegian, Swedish and Danish to investigate differences in degrees of syllable deletion (reduction) among these three languages. For the investigation two sets of data are used: one consisting of recorded speech from radio news and another consisting of sentences read aloud. The results of the comparative investigation show that in both data sets Danish exhibits a much larger degree of syllable reduction in speech than Norwegian and Swedish. The finding that certain syllable deletion processes take place in Danish but not in Norwegian and Swedish is viewed as typological. The results indicate that Danish words are shorter than their Norwegian and Swedish counterparts. This could be a contributing factor to problems that arise in inter-Scandinavian communication.

Swedish is beautiful, Danish is ugly? Investigating the link between language attitudes and intelligibility

Abstract This paper investigates the hypothesis that attitudes towards a linguistic variety and intelligibility of that variety are linked. This is done by eliciting language attitudes and word recognition scores in 154 Danish and Swedish schoolchildren and adolescents between 7 and 16 years. Language attitudes towards the neighboring language are elicited by means of a matched-guise experiment while word recognition is tested by auditorily presenting the participants with 50 spoken stimuli in their neighboring language (Danish for Swedish children and vice versa) in a picture-pointing task. Results revealed that while Danish children held more positive attitudes towards Swedish than vice versa and their word recognition scores were generally higher than those of their Swedish peers, the correlation between these two variables is very low, indicating that the two variables are only loosely linked.

University students' context-dependent conscious attitudes towards the official South African languages

Abstract This paper considers the results of an empirical investigation of overt language attitudes held by students attending North-West University, South Africa. Attitudes elicited from 325 students with mainly Setswana, Sesotho, Afrikaans and English as home languages are analysed comparatively. The study explores the relationship between language preferences in an educational environment and language preferences in other domains of students’ daily lives. The study thus indicates how different languages fulfil different roles in a multilingual university environment. The results presented are compared with findings reported in earlier attitudinal investigations from the Eastern Cape. In addition, a comparative analysis of attitudes by Afrikaans speakers is conducted considering students attending classes at two different campuses of the North-West University. At the traditional Afrikaans university campus, Potchefstroom, students show a higher regard for Afrikaans than students who attend classes at the Vaal Triangle. These findings are used in a discussion of context dependency of overt language preferences.

Why is Danish so difficult to understand for fellow Scandinavians

We investigate the relative contribution of duration and speech reduction on intelligibility.We manipulate spoken Danish sentences with respect to duration and pronunciation accuracy.For semantically unpredictable sentences in spoken Danish, the effect of duration is stronger than the effect of speech reduction. It has consistently been shown that among the three mainland Scandinavian languages, Danish is most difficult to understand for fellow Scandinavians. Recent research suggests that Danish is spoken significantly faster than Norwegian and Swedish. This finding might partly explain the asymmetric intelligibility among Scandinavian languages. However, since fast speech goes hand in hand with a high amount of speech reduction, the question arises whether the high speech rate as such impairs intelligibility, or the high amount of reduction. In this paper we tear apart these two factors by auditorily presenting 168 Norwegian- and Swedish-speaking participants with 50 monotonised nonsense sentences in four conditions (quick and unclear, slow and clear, quick and clear, slow and unclear) in a translation task. Our results suggest that speech rate has a larger impact on the intelligibility of monotonised speech than naturally occurring reduction.

Using Social Media to Measure Language Use.

ABSTRACT Investigations of language use in multilingual regions are traditionally done through the usage of retrospective questionnaires, either distributed on paper, or digitally. The aim of the present study was to test a new method of measuring language choice and use: by using social media to contact individuals several times a day and ask them what language(s) they are using, how, and with whom. It was investigated whether this method could provide real-time data on an individuals language use. One hundred participants living in the Netherlands were given a language background questionnaire and were asked to provide information through Facebook®. It was found that the social media data generally reflect the questionnaire data. More importantly, the social media data include more exact information on cases of codemixing and allow to compile a typology of language contact situations. It was concluded that social media inquiries could prove useful in future language use studies. Methodological issues, in particular concerning privacy and problems related to crowd sourcing of language data through social media, are discussed.

Word and phrasal stress disentangled: Pitch peak alignment in Frisian and Dutch declarative structures

This paper investigates intonational pitch variations and pitch peak alignment in declarative sentences and is part of a larger study of declarative, interrogative and imperative grammatical constructions in the Frisian-Dutch contact situation. Frisian is a minority language spoken in the province of Fryslân in the Netherlands. Following Jun [19], we devised a reading task in which phrasal intonation could be analysed while cancelling out focus effects. The reading task contains nine sentences per language, each with three trisyllabic words (SVO): three with focus on the first word, three on the second, and three on the last. For each set of three sentences, lexical stress is equally divided across the syllables of the focused word. A subset of 20 of the study’s 40 bilingual Frisian-Dutch native speakers performed the task in Frisian and Dutch. Pitch measurements were conducted according to the Melodic Analysis of Speech method and adapted MAS+ method, allowing for fine-grained phonetic analysis. Results suggest both Frisian and the local Dutch variety show delays in pitch peak alignment when compared to previous research, with Frisian showing a stronger delay in focus realisation. Additionally, an age effect in Frisian pitch production suggests a possible change in the language.