a Foucauldian approach to discourse analysis

A Foucauldian notion of discourse (1) holds that:

  • discourse is a culturally constructed representation of reality, not an exact copy
  • discourse constructs knowledge and thus governs, through the production of categories of knowledge and assemblages of texts, what it is possible to talk about and what is not (the taken for granted rules of inclusion/exclusion). As such, it re/produces both power and knowledge simultaneously
  • discourse defines subjects framing and positioning who it is possible to be and what it is possible to do
  • power circulates throughout society and, while hierarchised, is not simply a top-down phenomenon
  • it is possible to examine regimes of power through the historicised deconstruction of systems or regimes of meaning-making constructed in and as discourse, that is to see how and why some categories of thinking and lines of argument have come to be generally taken as truths while other ways of thinking/being/doing are marginalised.

There are of course a range of critiques of this social theory – how much it denies material reality, whether it disallows agency, whether anything precedes discourse and so on (2) .

Turning this way of understanding discourse into method to apply to textual analysis means asking of the text or texts questions such as:

  1. What is being represented here as a truth or as a norm?
  2. How is this constructed? What ‘evidence’ is used?  What is left out? What is foregrounded and backgrounded? What is made problematic and what is not? What alternative meanings/explanations are ignored? What is kept apart and what is joined together?
  3. What interests are being mobilised and served by this and what are not?
  4. How has this come to be?
  5. What identities, actions, practices are made possible and /or desirable and/or required by this way of thinking/talking/understanding? What are disallowed? What is normalised and what is pathologised?

(1) M. Foucault, The Archeology of Knowledge (London: Routledge, 1972/1995 ed, trans R. Sheridan); see also M Foucault, ‘Politics and The Study of Discourse’ in The Foucault Effect: Studies in Governmentality ed. Graham Burchell, Colin Gordon and Peter Miller (Chicago: University of Chicago Press, 1991), 53-72.

(2) See for example Nancy Hartsock ‘Foucault on Power: A Theory for Women?’ in Feminism/Postmodernism ed. Linda Nicholson (London: Routledge, 1990), 157-175; Paul Patton ‘Foucault’s Subject of Power’, Political Theory Newsletter 6. no. 1 (1994): 60-71; Jana Sawicki ‘Feminism, Foucault, and the “Subjects” of Power and Freedom’ in Feminist Interpretations of Foucault ed Susan Hekman (Pennsylvania: University of Pennsylvania Press, 1996); David Hoy, “Foucault and Critical Theory’ in The Later Foucault ed. Jeremy Moss (London, Sage, 1998), 18-32.

About pat thomson

Pat Thomson is Professor of Education in the School of Education, The University of Nottingham, UK
This entry was posted in discourse, research methods. Bookmark the permalink.

33 Responses to a Foucauldian approach to discourse analysis

  1. Pingback: policy as problematisation | patter

  2. Marguerite says:

    Hi Pat,
    I appreciated this blog post. Foucault is definitely one of the great social constructionists. I can see threads in this that are similar to some of Ian Hacking’s work (The Social Construction of What?)

    Are you going to be using critical DA for your research?

    Marguerite

    Like

    • pat thomson says:

      hi marguerite. thanks for the comment, and yse there are connections with hackins approach to the way that categories of knowledge are produced. I do use this approach in some of my work. Im putting this blog together for the doc researchers I work with and anyone else who finds it useful.

      Like

  3. Nolu says:

    I believe my understanding of Foucault is deepening as I am in midst of locating my work within his framework. I am shaping my doc proposal it is a struggle though.

    Like

  4. Sue Bartow says:

    Hello. I would like to ask you a question. I have been using Foucauldian Discourse Analysis to look at the discursive possibilities in teaching with social media – to see what positions and subjectivities teachers occupy and activate. My data consist of interview transcripts and lengthy observations. Typically researchers consider discourse to be written or spoken text. I would like to argue that the classroom is a text also, that my notes are records (albeit subjective) of discursive productions, productions that can be read as text. Are you aware of any support for including that data, of anything written by or about Foucault’s concept of discourse or any other research conducted similarly? Thank you for your thoughts.

    Like

    • pat thomson says:

      discourse is understood in at least two ways – text and conversation as per CDA – and as actions, the material environment, social relations, texts of all kinds .. this is Foucault. you can certainly justify the classroom as discursive and as a text to be read for inclusions exclusions etc. theres a lot of Foucauldian work that takes this approach. check out valerie harwoods book on disorderly children, richard neisches book on ed leadership, stephen balls writings on foucault, ian hunter on the school … their references will give you other leads too. ive got a paper on academia edu on deconstructing leadership that takes this line too [free].

      Like

      • Sue Bartow says:

        Hello again, Pat. I have been holding onto your recommendations like gold and have spent the day digging into them. I am still struggling though. I am still searching or waiting for two of the Foucault references in the paper you recommend and my library can’t find its copy of the Harwood book. The more I read though the more concerned I am becoming that discourse, even if it is situated in and interacting with social practice, is really just written, spoken, or visual text (like pics or movies). I can’t find anything that says the arrangement of chairs in a classroom is discourse. It’s obvious to me that “classroom” is a construction, part of a discourse of education in the last few centuries but I can’t find anyone that says I can use my notes of what people are doing as well as what they are saying in a classroom as data for a Foucauldian discourse analysis. It seems like I am either using the wrong model or concept or methodology to analyze my data or I need to find something else – like ritual or something – to talk about what happens in a classroom. Sorry to bug you again – but am I making needless circles or do I have a problem? When I read Foucault (Discipline & Punish, The History of Sexuality) it seems to fit but I can’t find something clear in The Archaeology of Knowledge or The Order of Discourse. Thanks so much for your help.

        Like

      • pat thomson says:

        Discourse produces expert systems of knowledge, exclusions inclusions etc ,, which are instantiated in classroom systems including their organisation rules etc. these are discursive, part of a discursive assemblage… See Ball article on policy as text and discourse and Hall http://edst.educ.ubc.ca/sites/edst.educ.ubc.ca/files/courses/Hall-Foucault-power%20knowledge%20and%20discourse.pdf.
        Hope that helps.

        Like

      • Sue Bartow says:

        Thank you so much! I am getting right on it and am so grateful for your help.

        Like

    • Thorunn says:

      Hi Sue. I’m currently reading this book and I remembered your comment on here when came across this:
      “… Even the design of buildings such as prisons reveals a social logic that specifies ways of interpreting persons and the physical and social landscapes they occupy” (Foucault, 1979).
      This is from page 490 in The SAGE handbook of qualitative research (3. ed), edited by N.K. Denzin and Y.S. Lincoln.

      Like

  5. Wow, thank you so much for this. 🙂 And also for the sources

    Like

  6. SheriO says:

    Thanks for this. I went to Academia.edu to look up your paper about deconstructing leadership and was taken into a pay wall. Is the paper available outside a paywall?
    I’m interested in deconstructing leadership as my action research doctoral project involves me, in my role as Academic VP for the graduate students’ union, pressing for greater transparency in assessment practices in online doctoral program (Ed. D) and also providing skills workshops or forums on aspects of doctoral education neglected in our program like ‘what is originality?’ “problem finding,’ or “research questions.” I’m also interested in exploring the process by which doctoral programs change, let’s say to permit networked or collaborative research projects, for example.

    Like

  7. C says:

    Hi…I’m using Foucault discourse analysis for my thesis and feel quiet stuck with analysis do u provide pvt supervision

    Like

    • pat thomson says:

      No I’m sorry I don’t. I just blog.

      Like

    • suebartow says:

      A resource I found really helpful is Willig’s 2008 book: Introducing Qualitative Research in Psychology (2nd ed.) published by McGraw Hill Open University Press. She develops guidelines I modified slightly and used in my dissertation – doing Foucauldian discourse analysis of classroom teachers using social media in their teaching.

      Like

  8. Christine says:

    Any advice on reading reviewing discourse analysis on television fan sites. How television viewers make meaning of what they watched and go online to talk about it…specifcally how they made meaning of what they just wanted reflecteed on blog, social media, or fan forums?

    Like

    • SheriO says:

      I wonder if ‘fandom’ and ‘Foucault’ in a Google Scholar search might bring up some hits. Coincidentally, my neighbour’s daughter compted an MA in female fandom,

      Like

  9. Randi says:

    Is there a “real” methodology for FDA? It seems that most academics do not clearly mention their approach. I do understand that Foucault himself rejected the idea of “spelling out” a methodology. As a doctoral student working in a science/engineering background, I find it difficult to make the scholars in my field understand this. Is there any systematic methodology developed for this ?

    Like

    • maelorin says:

      Oh gods. Foucault in sci/eng. How’s that going? In my understanding of his work, Foucault (and many other social researchers) resists formulaic statements of methods or methodologies because they are not examining concrete phenomena, rather they are making sense of social phenomena – and the processes of sense-making influence the phenomena you are trying to make sense of (even if in no other way than that making-sense-of alters your understanding of the phenomenon, and thus of the phenomenon itself).

      Much of the social/sociological research approaches that I have come across are not only interpretive (and critical) of the phenomenon of interest, but also interpretive and critical of themselves. This, it seems to me, is an important feature of any research, and of the framework used from which to accumulate and to interpret an understanding of the research subject. In social research, the research process itself is as much a subject of the research as the phenomenon under investigation.

      One might explain, or justify, all of this as being about an evaluation of the research approach itself as a meaningful way to learn meaningful things about your phenomenon of interest. (I think a reason why some of the language around social research is so difficult to understand, so difficult to interpret, for those of us new to it is because the subject of the research is *simultaneously* the social phenomenon *and* the research of that phenomenon.

      Perhaps be very explicit when discussing what the subject of your discussion/analysis actually is at every stage in your text(s). ‘Science’ and ‘engineering’ generally assume that the subject of research is a material, physical phenomenon. The methods used should explicitly isolate that phenomenon as much as possible.

      Social research, in contrast, assumes that the phenomenon under investigation *cannot* be isolated from other phenomena – indeed, the interactions between phenomena are an *integral part* of the phenomenon of interest. Therefore, they cannot be isolated from other factors – including the research methods employed to investigate them. Thus, the researcher must examine the possible influence of their chosen methods and methodology upon the phenomenon, an on their interpretation of what they find (and on what they don’t find).

      I moved out of an IT/IS school into business because my research integrates tech and socio-legal perspectives regarding privacy and identification technologies … Foucault in ‘hard’ science or engineering must bewilder your academic peers. (“How do you measure *that*?). I have found that every time I answer questions about my research, I learn something new about it. So write, and write more. Explain, and question, everything you can think of…

      Like

    • Sue Bartow says:

      This discussion thread has provided some of the most helpful questions, comments, and suggestions on Foucault and qualitative research I have found. Maelorin, your comments are tremendous! From my past posts in this discussion you can tell I wrestled with FDA for my dissertation. My committee members generally accept FDA or at least my very qualitative approach. However, when I tried to submit an article based on my research to a journal, I got some serious push back from two of the reviewers. Fortunately for me, the editors were super helpful. One of them shared this resource that I ultimately used to justify/connect Foucault’s work to a research methodology trajectory – Keller’s Analysing Discourse: An Approach from the Sociology of Knowledge. It’s easy to find by Googling the title as it is an online forum article but here is the URL: http://www.qualitative-research.net/index.php/fqs/article/view/19. It’s Reiner Keller. I doubt it will convince your science/engineering folk. Maybe some of the work in physics on how the observer empirically distorts or impacts a phenomenon will help them recognize that the measurer impacts the physical world as well as the idea that science is a contextual effort. We influence our own understandings by what we choose to ask, even in the physical world. Of course that discussion gets complicated.

      Like

  10. Dear Pat
    Thanks for the insights on FDA. I am using this method for my research by attempting to construct an archive for analysis. Could you recommend any source that actually talks about constructing an ‘archive’ to conduct FDA.

    Like

  11. Pingback: happy birthday to patter… a reflection on the last two years | patter

  12. Pingback: patter’s 2013 top ten | patter

  13. Pingback: do I have to transcribe the interviews I do? | patter

  14. Pingback: patter’s year | patter

  15. Josh says:

    Hi! I’m actually working on determining and defining the “digital self” of the youths today using this method. Still working on it.

    Like

  16. Justine says:

    I have the impression a lot of people get stuck with Foucauldian discourse analysis because they read too many other people’s versions of Foucault. When I read ‘The Archaeology of Knowledge’, it puts me into a certain state of mind, makes things appear a particular way. It becomes a critical lens (to use the standard metaphor), a key part of which is sensitivity to powerful social institutions. I can’t understand Foucauldian DA without institutions being absolutely front and centre.
    My understanding of ‘Archaeology’ is that Foucault was prompted to write it in response to critics and students who asked questions about the methodology used for ‘Discipline & Punish’. Its actually the methods chapter of that book, which alerts us to the centrality of powerful social institutions, both abstract and concrete.

    Like

  17. Pingback: The Language and Discourse of Teaching Job Adverts | sputniksteve

  18. Pingback: Lecture notes 1.30.18 | Hum 105 – Hist 105 Honors

  19. Pingback: My Four-Year-Old Dances Like a Stripper – Undecidedly Undiagnosed

Leave a comment