Focusing foreign-sounding Swedish – a study on sociolinguistic awareness and language attitudes in Sweden, in collaboration with the Language Council of Sweden

Focusing foreign-sounding Swedish
– a study on sociolinguistic awareness and language attitudes in Sweden

Project description (background, aims, data and methods)

Understanding how foreign accents, in all their diversity, are perceived and responded to in majority societies has become increasingly important in an era of rising international migration. Much of the research on attitudes towards foreign accents has, however, focused either on non-native speech in general or on a single specific accent. Relatively few studies have explored attitudes towards the broader range of foreign accents present within a single nation. Yet, such an approach is crucial for capturing the complex dynamics of inclusion and social cohesion in society. It enables the exploration of key questions, such as: Are accents associated with different first languages evaluated differently? How do potential status differences affect the opportunities for social inclusion of speakers with a foreign accent?

In the present Flexit-project, perceptions and attitudes towards a range of “foreign-sounding” accents in contemporary multilingual Sweden are explored. Different methods have been used to collect both quantitative and qualitative data. Two large quantitative studies have been carried out in the project. Study 1 consists of a questionnaire that follows up on a number of questions from a language attitude survey conducted 25 years ago (Wingstedt 1998). Its aim is to examine changes in attitudes towards L2 Swedish over the past decades. Study 2 (= the main study of the project) is a speaker evaluation experiment, focusing on perceptions and attitudes towards a range of specific accents. Over 2,000 participants took part in the experiment. The project also includes interview and group discussion data from a smaller number of participants – both L1 and L2 speakers of Swedish.

The project is guided by the following research questions:
• How is foreign-accented Swedish perceived and assessed in Sweden today?
• How proficient are people at identifying various accents?
• Do different accents trigger different listener reactions?
• Is there a hierarchy of preference among participants regarding different forms of accented Swedish?
• What is the impact on listeners’ perception of a speaker when they fail to identify the speaker’s accent, particularly if the accent is misattributed to another language?

Main results

The results from the first study (N=1,330) indicate a substantial increase in the number of people today who hold a positive perception of L2 Swedish, as compared to the situation 25 years ago. However, closer examination in Study 2 reveals uneven acceptance across different accents. This study exposed participants to accents originating from Arabic, Bosnian, English, Finnish, German, Persian, Polish, Somali, Spanish, and Turkish respectively, as well as suburban Swedish accent. The answers to the accent identification task show that only English, Finnish and German accents are recognised by a majority of participants, whereas suburban accent is correctly identified by just over 40 %. The other accents were difficult to identify (less than 20 % correct guesses) and were often mistakenly assumed to originate from Arabic.

Listener reactions uncovered numerous instances of both linguistic stereotyping and reverse linguistic stereotyping. Accents from North-Western European languages generally elicited positive listener reactions. In contrast, accents from other regions were rated more negatively. Suburban-sounding Swedish (i.e., a native accent of Swedish) is the accent that received the most negative listener reactions. The findings highlight the existence of a pronounced accent hierarchy within Swedish society. These outcomes are reinforced by data from the interviews and group discussions. The analysis of misidentifications reveals that a presumed Arabic accent often results in more negative judgements about the speaker. Collectively, these findings underscore that attitudes towards languages largely mirror attitudes towards the speaker group or the country and culture associated with the language, rather than towards the language itself.
 
An analysis of the attitude data across various subgroups of listener background variables shows that two variables in particular – gender and political preference – have a significantly strong impact on language attitudes. Male listeners consistently rate all speakers less favourably compared to female listeners. Furthermore, respondents who align with right-wing parties tend to assign significantly lower scores to the majority of accented speakers. These patterns align with an increasingly pronounced trend in Swedish society in recent years: a gender gap, wherein women tend to shift towards the left and men towards the right on various political issues.

Methodological issues

Several important methodological issues in language attitude research are highlighted by the present study:
1. By combining multiple data collection methods (surveys, speaker evaluation experiments, group discussions, interviews), a more nuanced understanding and deeper insights into prevailing attitudes towards linguistic variation can be achieved.
2. Many studies using speaker evaluation experiments lack an identification task, such as: ”What is the linguistic background of this speaker?” or ”What is this speaker’s first language?”. However, such a task is essential to determine which accents participants believe they are evaluating. The identification task is central to interpreting the collected data. In the present project, the existence of an accent hierarchy could not have been revealed without it.
3. The study also demonstrates that analysing misidentifications of speakers’ linguistic backgrounds offers significant insights into people’s attitudes towards various forms of linguistic variation. Thus, the project makes a forceful call for the increased inclusion of this often-overlooked type of data in future research on language attitudes.

Further research

Regarding the consequences of prevailing language attitudes for speakers of foreign-sounding Swedish, it is important to emphasize that the empirical data collected in this project come from surveys, listening experiments, and interviews rather than real-life situations. Consequently, the results cannot be used to make definitive statements about the possible existence of discrimination in concrete, real-world contexts. Instead, the project maps perceptions of and attitudes towards different types of migration-related language variation in Swedish society today – knowledge that is important, not least for further research. As Lindemann (2005: 210) suggests: “While further research is needed into how expectations translate into evaluations of actual speakers, knowledge about the expectations themselves is an important step towards addressing language discrimination in several ways”.

Thus, studies of language attitudes in real-life situations that are central to social inclusion (e.g., in recruitment processes or interviews with housing agencies) are called for. An application for such a project is currently being prepared by Bijvoet and Senter.

Collaboration and dissemination of research

The project has been planned and implemented in collaboration with the Swedish Language Council, which operates under the Institute for Language and Folklore (Isof). Isof is a government authority mandated “to build, collect, and disseminate knowledge about Sweden’s language and culture”. A key responsibility of the Language Council is to raise awareness of perceptions and attitudes towards language and language variation in Swedish society today – including migration-related language variation. It is with this objective that the Language Council took on the present project.

The research outputs have been disseminated in various forms: in scientific and popular science publications, in talks at national and international scientific conferences and workshops, but also to a wider audience outside of academia (e.g., at Isof’s website and at the Gothenburg Book Fair) in order to maximise the benefit of the research. Several of the above-mentioned publications are now used as course literature at university departments across Sweden. The dissemination of knowledge generated by the project has also taken place through mass media participation (including radio and television interviews). Additionally, the project’s findings have been reported in several newspaper articles and news broadcasts from Swedish Radio (both in Swedish and other languages).

The project has contributed to increased collaboration between the Department of Scandinavian Languages (Uppsala University) and Isof, not least through staff exchanges. Several colleagues and master students from the department have undertaken both short- and long-term assignments at the Language Council. Additionally, colleagues from the Language Council are regularly invited to give presentations at seminars and relevant university courses. This ongoing exchange of theoretical and practical knowledge benefits both workplaces.

References

Isof: https://www.isof.se/other-languages/english/about-the-institute

Lindemann, S. (2005): Who speaks “broken English”? US undergraduates’ perceptions of non-native English. International Journal of Applied Linguistics 15(2). 187–212.

Wingstedt, M. (1998): Language ideologies and minority language policies in Sweden. Historical and contemporary perspectives. Stockholm: Stockholm University dissertation.