![]()
Part I
Discourses
Discourse analysis already has a history in psychology, and I set the scene for an analysis of discourse dynamics in this first part of the book. In the first chapter I work through the consequences of a definition of discourses as sets of statements which constitute an object. I elaborate a set of criteria for defining what a discourse is. I run through seven necessary criteria and then three additional points to do with institutions, power and ideology. Through these criteria I use post-structuralist work on discourse to show how the study of language must attend to tensions and contradictions which express political matters. Without these three supplementary points, an analysis of discourse could become just another method, just an academic exercise, and then just as pointless as other frameworks psychologists use to describe action and experience.
I cannot pretend that a focus on institutions, power and ideology is not a moral/political matter (and perhaps it is only that), and the following chapters in this book rest on the assumption that amoral/apolitical psychology is worse than useless. Each theoretical resource used by discourse analysts in psychology (whether it be sociology of scientific knowledge, microsociology, analytic philosophy, rhetoric, semiology or post-structuralism) can all too easily draw us into versions of linguistic relativism, idealism. A focus on the structures of language at the expense of an analysis of material coercion is one that discourse analysts feel differently, and ambivalently, about. Chapter 2 takes this issue seriously, and outlines a realist view of language and the material resources which make discourse possible. The study of the dynamics which structure texts has to be located in an account of the ways discourses reproduce and transform the material world.
![]()
Discovering discourses, tackling texts
A means of indicating the transition from a sign to a sign is offered by quotation marks. Thus âCaliforniaâ is a sign that denotes California; âCaliforniaâ is written with ten letters, whereas California grows oranges. The transition may be further repeated. Thus âCaliforniaâ is the name of a sign, namely of âCaliforniaâ, but is not the name of California. In writing quotes we have to watch that the sign combination occurring in our sentence is always one level higher than the object to which it refers. Thus âCaliforniaâ is written with one pair of quotes, and âCaliforniaâ is written without quotes. It would be difficult to add quotes to California; we then would have to construct huge quotes and put those of the left end into the Pacific Ocean, and those of the right end into Nevada.
(Reichenbach 1947: 68)
What is a âdiscourseâ? This chapter is concerned with the task of defining discourses. I will draw attention to some of the descriptions of discourse outside psychology, and then set out some criteria for identifying discourses. My main focus will be on the practical problems which confront a researcher attempting to carry out a discourse analysis. However, each practical problem raises broader issues about the nature of language, discourse and texts. I will also argue, towards the end of the chapter, that discourse analytic research should go beyond seven necessary criteria for the identification of discourses, and consider the role of institutions, power and ideology.
Discourse research strikes a critical distance from language, and one useful aspect of the approach is the reflexivity urged upon a researcher, and reader. When discourse analysts read texts they are continually putting what they read into quotation marks: âWhy was this said, and not that? Why these words, and where do the connotations of the words fit with different ways of talking about the world?â I want to argue, however, that this reflexivity needs to be grounded if it is to have progressive effects, and that work in the post-structuralist tradition can ground discourse and reflection historically in a useful way. In addition, the study of discourses carried out by Foucault and his followers has implications for how we describe the emergence of academic psychology and the âpsycomplexâ in Western culture, and for how we understand the discipline and its objects today (Rose, 1985, 1989).
Foucault (1971), for example, described how a discourse which was about âmadnessâ as a medical category came into being, and the ways in which a medical discourse emerged alongside related ways of speaking about individual âpathologyâ which involved the categorisation of a section of the population. Debates over rationality and responsibility in the nineteenth century were informed by such discourses. In another study Foucault (1975) and co-workers collected legal papers and accounts given of a murder at that time, and showed how discourses of individual reason and âmadnessâ framed the possible explanations that could be given of the event. Foucault (1977, 1981) also connected the development of discourses which describe and prescribe forms of rationality, responsibility and pathology with discipline, surveillance and power. These discourses informed legal practice, and they helped constitute contemporary psychology. Discourses about the person that we employ today, then, have a history.
It is also possible for discourses about the nature of mental processes, and to whom one attributes them, to fall into disuse. At the beginning of the century, for example, at the very moment when âcross-culturalâ psychology was busily demarcating certain human âracesâ as not fully mentally developed, there was an area of research devoted to plant psychology (Crellin, 1989). The attribution of mental states was framed by discourses pertaining to plants as almost sentient beings. In contrast, the dominant psychology we have today is informed by particular conceptions of rationality, discourses in which one attributes to individual human beings internal mental states which, we suppose, direct behaviour (Costall and Still, 1987).
A number of issues arise from the history of discourse. Discourses do not simply describe the social world, but categorise it, they bring phenomena into sight. A strong form of the argument would be that discourses allow us to see things that are not âreallyâ there, and that once an object has been elaborated in a discourse it is difficult not to refer to it as if it were real. Discourses provide frameworks for debating the value of one way of talking about reality over other ways. Types of person are also being referred to as the objects of the discourses. When we look at discourses in their historical context, it becomes clear that they are quite coherent, and that as they are elaborated by academics and in everyday life they become more carefully systematised. Discourse analysis deliberately systematises different ways of talking so we can understand them better. A study of discourse dynamics takes off from this to look at the tensions within discourses and the way they reproduce and transform the world.
A good working definition of a discourse should be that it is a system of statements which constructs an object. However, this definition needs to be supported by a number of conditions. In the main section of this chapter, then, I will set out seven criteria, the system of statements that should be used to identify our object, to enable us to engage with, and in, discourse analysis. Potter and Wetherell's (1987) account of the âmethodâ, their âten stages in the analysis of discourseâ, is useful, but sometimes bewilders new researchers as it dawns on them that each step rests on a bedrock of âintuitionâ and âpresentationâ. At points the reader is told, quite rightly, that discourse analysis is like riding a bike, is warned that the stages are not sequential, and advised that âthere is no analytic methodâ (Potter and Wetherell, 1987: 169). Billig et al.'s (1988) suggestions that a researcher needs to look for âimplicit themesâ is also right, but can also lead to worries on the part of someone new to the area that they may not be picking up the themes that matter. Similarly, when Hollway (1989) draws on her own intuitive feel for what is going on in discourse, how the accounts of her interviewees resonate with her own experience, she produces fascinating and plausible analyses. But how could we do this too? I do not want to suggest that the criteria presented here constitute a method, that they should necessarily be employed sequentially, but that they will help to clear up some of the confusions that have followed the incorporation of discourse ideas into psychology.
SEVEN CRITERIA FOR DISTINGUISHING DISCOURSES
These seven criteria deal with different levels of discourse analysis. There is a degree of conceptual work that needs to go into the analysis before the material is touched, and then, as the analysis proceeds, it is necessary to step back a number of times to make sense of the statements that have been picked out. Each criterion raises questions about the theoretical framework the researcher is using. Along the way I will mark some âstepsâ in an analysis of discourse dynamics (and you will have to imagine quotation marks around the word âstepsâ from now on).
1) A discourse is realised in texts
First, though, where do we find discourses? It would be misleading to say that we ever find discourses as such. We actually find pieces of discourse. I want to open up the field of meanings to which discourse analysis could be applied beyond spoken interaction and written forms by saying that we find discourses at work in texts. Texts are delimited tissues of meaning reproduced in any form that can be given an interpretative gloss.
Take two examples: (i) I was given a small Liquid Crystal Display electronic game for Christmas. The buttons on the left and right move a male figure at the bottom of the screen from side to side. The figure is waving a crucifix at the ghosts descending from the top of the screen to their graves. As each ghost is prevented from landing and is despatched to the flames at the right-hand side I get awarded ten points (and the penalty for letting each spirit through is a lost life). This is a text. A Christian discourse inhabits this text, and it is the translation of this text into a written and spoken form that renders that discourse âvisibleâ or, more accurately, in which the category âdiscourseâ becomes appropriate; (ii) as I work through this chapter a second time a melody comes wafting up the stairs. The unmarried couple I share a house with are playing host to her parents who have come to cosset the new baby. Grand-dad has got his hands on the electric organ and is quietly finding his way round the keyboard. The tune is uncertain snatches of âHere comes the brideâ. A little narrative of heterosexual bonding within familial discourse inhabits this melody as the text.
It is useful, as a first step, to consider all tissues of meaning as texts and to specify which texts will be studied. All of the world, when it has become a world understood by us and so given meaning by us, can be described as being textual. Once the process of interpretation and reflection has been started, we can adopt the post-structuralist maxim â[t]here is nothing outside of the textâ (Derrida, 1976:158). This does not necessarily commit us to a particular position on the nature of reality, textual or otherwise. I deal with that issue in the next chapter. I am merely drawing attention to the effects of describing, for research purposes, the world in this way. Speech, writing, non-verbal behaviour, Braille, Morse code, semaphore, runes, advertisements, fashion systems, stained glass, architecture, tarot cards and bus tickets are all forms of text. In some cases we could imagine an âauthorâ lying behind the text as source and arbiter of a true meaning. But the lessons to draw from this list are, first, that, as Barthes (1977) argued, there need not be an author, and, second, that once we start to describe what texts mean we are elaborating meanings that go beyond individual intentions, discourses that are transindividual. The second step in a discourse analysis, then, should be a process of exploring the connotations, allusions and implications which the texts evoke. A helpful guide to this exercise in cultural anthropology is Barthes' (1973) work on modern âmythâ.
Sometimes different discourses are available to different audiences. The distinction between the inside and outside of psychology is a good case example. On the one hand, the Ď sign gives a text a meaning for those of us inside psychology. The discourses which inhabit a text containing that sign will often be discourses coherent to psychologists. On the other hand, an image of Freud's face gives a text a meaning for those outside the discipline. The discourses which give that sign meaning, and it often means âpsychologyâ for outsiders, would not be accepted by many psychologists. It is right, then, to adopt the formulation that discourses are âlinguistic sets of a higher order than the sentence (while often reducible to a sentence) and carried out or actualized in or by means of textsâ (Marin, 1983 162). Discourse analysis, then, involves two preliminary steps:
1 Treating our objects of study as texts which are described, put into words; and
2 Exploring connotations through some sort of free association, which is best done with other people.
2) A discourse is about objects
âAnalysisâ necessarily entails some degree of objedification, and in studies of discourse there are at least two layers of objectification. The first is the layer of ârealityâ that the discourse refers to. It is a commonplace in the sociology of knowledge (e.g., Berger and Luckmann, 1971) that language brings into being phenomena, and that the reference to something, the simple use of a noun, comes to give that object a reality. Discourses are the sets of meanings which constitute objects, and a discourse, then, is indeed a ârepresentational practiceâ (Woolgar, 1988a: 93). The representation of the object occurs as previous uses of the discourse and other related discourses are alluded to, and the object as defined in the discourses is referred to. Some local councils have had to close off sewer entrances to stop young children from going down to look for ninja turtles. The turtle discourse constitutes these beings as objects for children, and when the children refer to turtles they are referring to the objects of the discourse. They think, as most of us do when we talk about things, that they are talking about real objects in the world. Discourses are, according to one post-structuralist writer, âpractices that systematically form the objects of which they speakâ (Foucault, 1972: 49).
Discourse constructs ârepresentationsâ of the world which have a reality almost as coercive as gravity, and, like gravity, we know of the objects through their effects. Take, for example, descriptions of medi-eval Anglo-Saxon sorcery in which the world is full of spirits and physical illness is attributed to the shots fired by elves (Bates, 1983). What we now can describe as âdiscoursesâ created and reproduced spirits and elves. Then, they were real in the way that atoms and electrons are real today. Many of the objects that discourse refers to do not exist in a realm outside discourse. There are fuzzy borders between the set of things we know exist outside discourse and the things which may have a reality only within it. The first layer of reality, then, is the reality of the objects of the discourse, the things the discourse refers to.
The second layer of reality, of objectification that a discourse sometimes refers to is that of the discourse itself. One example is a badge given away at the Commonwealth Institute in London in 1988 with âDialogue on Diarrhoeaâ printed around the top. It says âinternational newsletterâ around the bottom, and these phrases frame a picture of a woman feeding an infant with a spoon. There were also huge posters around the cafeteria with the same message blazoned across them. At the first level of meaning, we have the object âdiarrhoeaâ, and the badge is a text which reproduces the object in particular ways: (i) we know that âdiarrhoeaâ is, among other things, a medical description, and so we can identify a medical discourse; (ii) we assume that the woman feeding the infant is the mother, and so a familialist discourse also touches the text; and (iii) we understand the image and message as located in an appeal, located in a discourse of charity. The second layer of reality, then, is that of the âdialogueâ, and here there is a reflection in the text on a discourse, and the text says that there is another âobjectâ which is the set of statements about diarrhoea. A discourse is about objects, and discourse analysis is about discourses as objects. This criterion, then, takes us into a third and a fourth step of analysis:
3 Asking what objects are referred to, and describing them (turtles, diseases, ghosts etc.); and
4 Talking about the talk as if it were an object, a discourse.
3) A discourse contains subjects
The object that a discourse refers to may have an independent reality outside discourse, but is given another reality by discourse. An example of such an object is the subject who speaks, writes, hears or reads the texts discourses inhabit. I will...
