Transcript analysis: sociolinguistics and Foucault
This essay considers the use of interactional sociolinguistics in the analysis of a short transcript (reproduced as appendix one for ease of reference). A brief introduction to the tradition is followed by the analysis itself. This analysis is then subjected to a critique from the Foucauldian perspective, to highlight potential strengths and weaknesses in the use of interactional sociolinguistics for discourse analysis.
The tradition of interactional sociolinguistics
Taylor (in Wetherell et al., 2001a) introduces sociolinguistics as her third generic approach to discourse analysis, claiming a focus on language use within specific groups, for example, families (see Fitch 2001). Hymes (1974) identifies the ‘speech community’ as the entity with which one starts; in ethnographies of communication, ‘one starts with a social group and considers the entire organisation of linguistic means within it’ (page 47). This is the context within which Fitch (2001, page 60) presents three units of analysis identified by Hymes for use in a sociolinguistic study:
- speech situations
- speech events
- speech acts
Hymes (1974) identifies further units of analysis as components of speech, rules of speaking and functions of speech (pages 53–64), although these could be more relevant to conversation analysis (Moerman [1988, in Fitch 2001 pages 61-62] would approve of these being included in sociolinguistics as supporting his approach to data collection and analysis). This also demonstrates the closeness and overlap of discourse analytic approaches. Hymes (1984) sees the purpose of sociolinguistic analysis as ‘understanding and explaining the interaction of language and social life’ (page 66).
Fitch (2001) states that sociolinguistics focus on ways of speaking, speech communities (see above) and native terms for talk (page 57). These are illustrated by Gumperz (2001) who refers to ‘differences in linguistic and socio-cultural knowledge’ (page 139) in interethnic interactions. Gumperz shows two individuals from different sociocultural backgrounds having different ways of speaking. If these individuals were interacting with other individuals from the same sociocultural background, the degree of misunderstanding would be minimal. In the case studies used in his article, Gumperz highlights the group-based nature of sociolinguistics – the interaction between two individuals from two different social groups is fraught with miscommunication and misunderstanding. A study of each group by the other would clarify and reduce misunderstandings in what Gumperz calls ‘a strategy for self-diagnosis of communication difficulties’ (page 139).
The article also shows how different native terms for talk can sound the same but have completely different meanings, something Gumperz calls ‘context-bound interpretive preferences’ (page 139). The study of terms in each group can assist the understanding of other groups and facilitate communication between them. This is expanded by Tannen (2001, page 150) who used sociolinguistics to study the extent to which language is affected by cultural background, discovering that the use of language extends to sub-cultural aspects of society, including ethnic and regional style and gender. The impact of gender is not considered in Gumperz’s study referred to above, but it is possible that Participant A, coming from a patriarchal society, may feel that his situation is belittled by being referred to as female Participant B (Gumperz, 2001, page 139), and might be a reason he sees the recording of the interview as signifying importance (ibid, page 142) (a counter to the possible indifference conveyed by the interviewer’s gender).
Tannen (2001) also identifies key focus points for her study that deal with relations of power and solidarity, stating that in the context of gender relations, it cannot be assumed that males always dominate and females are always submissive and that all behaviour and language supports and endorses this view. Only by studying group interactions can such aspects be identified, demonstrating the relevance of sociolinguistic study.
From this brief exploration of sociolinguistics, the key features are the study of language use within groups, that social life and the use of language interact in many ways, and that studying this can facilitate understanding of such things as power relations and different cultural approaches to communication. The overall purpose of sociolinguistics is to explicate differences and facilitate or change approaches to communication to avoid misunderstandings as recounted by Gumperz (2001).
There are limitations to this analysis. The only information provided is the transcript – the fact that no context is given is not usual for a sociolinguistic study. Using Hymes’ (1974) units of analysis, the study is concerned with a group of individuals socially classified as being intellectually disabled. It is assumed that these individuals share a way of communicating within their group, both verbal and nonverbal. The native terms for talk appear to be shared by the participants in both exchanges as the language used appears to be the same.
There is no information about facial expressions, tone of voice or bodily movements of either the interviewer or Michelle, so there can be no analysis of ‘interactive synchrony’ (Erickson and Schulz, 1982, cited in Gumperz, 2001). It would also have been helpful to indicate whether the participants had the same conceptions of what the interview was about. The questions at line 6 and 17 might indicate surprise at being asked such things; for example, Michelle might have understood the interview to be about how she could develop her computing skills, and being asked about typing might cause her to reframe the interview according to that belief. Further, the patterns of stress and intonation within the language – the prosody – are missing. If present, line 17 could register indignation at being asked what may have been perceived by Michelle to be a stupid question.
The interviewer seems to use more ‘female’ than ‘male’ language and the relationship between interviewer and interviewee tends to be like a parent to child. The question in line 8 is unlikely to have been delivered without an encouraging, supportive tone, which would be inappropriate with an adult. It may reflect the nature of Michelle’s intellectual disability and explain the slightly patronising tone of the interviewer’s utterances. This assertion clearly reflects my own beliefs regarding ‘male’ and ‘female’ language, something that Silverman (2001) identifies as ‘membership categorisation devices’ and ‘category-bound activities’ (pages 122–123). A female interviewer might have made Michelle feel more comfortable. Would she have spoken to a male interviewer in this way?
Line 7 is not a direct answer to the question in line 6 and seems to be Michelle’s way of demonstrating her abilities, perhaps ‘showing off’ or seeking to impress the interviewer of gain his approval. This might have been a reaction to the interviewer’s tone that could have been perceived as disparaging of typists
The extract appears to demonstrate Gumperz’s (2001, page 138) point that in sociolinguistics, grammatical knowledge may be shared; however the communicative style is different.
The extract appears to have three distinct phases. The first considers Michelle’s wish to change jobs and the typing. The non-answer to the question in line 6 by Michelle signifies a move to the second phase, where computer training is the subject of the interview. The third phase begins at line 16 when the motivation behind the training is explored. This ties in with Gumperz (2001) and his analysis of the interview exchange he studied.
There appears to be good speaker-listener co-ordination apart from the one major jump made by Michelle at line 6. There are no interruptions in this exchange and the answers follow the questions asked, demonstrating an understanding of the requirements of the question. From this it can be assumed that Michelle and the interviewer come from the same cultural background. The verbal strategies used appear to be direct rather than indirect, which may be a consequence of the interview situation and perhaps the perception of the interviewee held by the interviewer and/or researchers.
This extract shares many commonalities with extract 1; therefore the focus will be on differences.
In this interview, the language used by the interviewer appears to be more ‘male’ than ‘female’: it is more direct and does not have the patronising tone observed in the previous extract. An indication of some of the body language is provided, demonstrating both synchrony and speaker-listener co-ordination.
The interviewer interrupts Shane when he is describing serving motel guests at line 3, and Shane interrupts the interviewer at line 8. This could be a male power play and an attempt to dominate the conversation (genuine interruptions), or an attempt at solidarity, demonstrating support for the relevant speaker (overlaps) (Tannen, 2001, page 157).
The interviewer also uses ‘back-channel signs of interest’ (Gumperz, 2001, page 143) when he encourages further discussion in line 5 with the word ‘yes’ and again in line 7 when he says ‘Excellent! That’s good’. Line 7 also marks a change in topic and therefore a change of phase in the interview.
A Foucauldian Critique
Approaching the analysis of the transcript using a Foucauldian approach highlights some areas that the sociolinguistic approach ignores or understates.
Identity of the Researcher
Sociolinguistics considers the identity of the researcher from the point of view of the interaction that takes place between the researcher and the researched. However, what is not considered is whether the researchers have their own agenda or are working to a political agenda – perhaps to change the quality of help for the subjects of the study. Foucauldian analysis would view any political involvement as caught in the interplay between knowledge and power (Hall, 2001, page 76). Thus the interviewer is wielding some form of power, whether to shame the government of the day into changing policy, or by working for the government to produce research that fulfils a specific political agenda.
The Research Context
The lack of contextual information relative to the study and to the interviews is noted by sociolinguistics, but can be accommodated by using a more conversation-analytic approach to data collection and analysis, as advocated by Moerman (1988, in Fitch 2001). However, to ignore or avoid the context is to remove significant influences on the research. Political influence has already been mentioned; others include social, historical and cultural influences (Wetherell, 2001, page 388-389).
An example would be the construction of the idea of ‘intellectually disabled’. This ties in with Foucault’s view that the concept of mental illness was created by how it was described by society in general, politicians framing legislation and discursive practices. This constitutive, productive nature of discourse produces a specific discursive formation (Hall, 2001, page 73) that allows the idea of mental illness to take shape and be part of a particular culture. Subjects of the discourse would be created by the discursive formation, producing someone who is ‘intellectually disabled’. The idea of intellectual disability is created by the discourse, along with ‘mentally disabled’ people’s place in society, whether that place is in a secure mental health wing of a hospital or in the family home being cared for by family members. Thus Foucauldian analysis would consider a wider range of influences on the research area and its implantation than sociolinguistics allow for.
Links with Social Practices
Following on from the research context, Foucauldian analysis would consider how the discursive practices interact with other social practices – again, this would be ignored by the sociolinguists. The idea of discursive formation also includes how society’s institutions deal with the idea of intellectually disabled people in these extracts. This is demonstrated by the additional training provided to help those perceived to be intellectually disabled become fully functioning economic units for the state, rather than a burden to be supported at the state’s expense. This links with one of the current debates in discourse analytic research; namely whether discourse is only relevant to the cultural aspects of society or whether it also encompasses what Wetherell (2001, page 390) refers to as ‘material practices’. The sociolinguistic approach would appear to deal with both material and cultural issues, but at a superficial level. Foucauldian analysis sees discourse as affecting and being part of both at a far more fundamental level, and therefore requiring the inclusion of institutions that are part of the discursive formation surrounding the subject of the research.
Power and Knowledge within the Research
Knowledge is constructed by discourse. The idea of intellectual disability has been considered to be a social construct. Another aspect of knowledge that would have a bearing on the research is the knowledge of the researcher and how they use that knowledge in the research. The analysis indicated that some of the questions from the interviewer in extract 1 could be perceived as patronising. This might derive from the intellectually superior position of the interviewer as an academic or a professional interviewer. Society constructs what is considered valuable in the intellectual arena, and thus is another focus of Foucauldian analysis that sociolinguistics would ignore.
The interview situation is a situation where the interviewer, in addition to having power conferred by their apparent intellectual superiority, will have the power of being the one who is asking the questions, thereby determining what sort of answers will be given and the general topic of conversation. While sociolinguistics completely ignore the power relations between interviewers and participants, Foucauldian analysists would automatically consider this to be part of the context of the research and in each interview.
This analysis has demonstrated that different traditions of discourse analysis produce different ideas as to what is happening through the use of language and the interaction between language and social context. Sociolinguistics offers a narrower, group-centric view of this interaction than Foucauldian analysis which takes a much broader view and considers anything that has meaning in the context of the area being studied.
Fitch, K. (2001) “Reading Five – The Ethnography of Speaking: Sapir/Whorf, Hymes and Moerman” in Wetherell, M., Taylor, S. and Yates, S. J. (eds.) Discourse Theory and Practice: A Reader, London, Sage/The Open University
Gumperz, J. J. (2001) “Reading Eleven – Interethnic Communication” in Wetherell, M., Taylor, S. and Yates, S. J. (eds.) Discourse Theory and Practice: A Reader, London, Sage/The Open University
Hall, S. (2001) “Reading Seven – Foucault: Power, Knowledge and Discourse” in Wetherell, M., Taylor, S. and Yates, S. J. (eds.) Discourse Theory and Practice: A Reader, London, Sage/The Open University
Heritage, J. (2001) “Reading Four – Goffman , Garfinkel and Conversation Analysis” in Wetherell, M., Taylor, S. and Yates, S. J. (eds.) Discourse Theory and Practice: A Reader, London, Sage/The Open University
Hymes, D. (1974) Foundations in Sociolinguistics: An Ethnographic Approach, Cinnaminson NJ, University of Pennsylvania Press Inc.
Silverman, D. (2001) “Reading Ten – The Construction of ‘Delicate’ Objects in Counselling” in Wetherell, M., Taylor, S. and Yates, S. J. (eds.) Discourse Theory and Practice: A Reader, London, Sage/The Open University
Tannen, D. (2001) “Reading Twelve – The Relativity of Linguistic Strategies: Rethinking Power and Solidarity in Gender and Dominance” in Wetherell, M., Taylor, S. and Yates, S. J. (eds.) Discourse Theory and Practice: A Reader, London, Sage/The Open University
Wetherell, M. (2001) “Reading 27 – Debates in Discourse Research” in Wetherell, M., Taylor, S. and Yates, S. J. (eds.) Discourse Theory and Practice: A Reader, London, Sage/The Open University
Wetherell, M., Taylor, S. and Yates, S. J. (2001a) Discourse as Data: A Guide for Analysis, London, Sage/The Open University
Wetherell, M., Taylor, S. and Yates, S. J. (2001b) Discourse Theory and Practice: A Reader, London, Sage/The Open University
Appendix One: Transcript Two from the Course Assignment Booklet