A genealogy of postmodern subjects: Discourse analysis and late capitalism

Potter and Wetherell’s discourse analysis

Of course, Potter and Wetherell did not invent the idea, or the practice, of DA whole cloth. In the early 1980s, previous sociological studies were criticized for ignoring the variability of discourse and a strong case was made for DA as the way to proceed in sociology (e.g., Mulkay, Potter, & Yearley, 1983). Potter and Wetherell (1987) also built on prior critiques of “essentialist” methods and “dominant” paradigms in social psychology (e.g., Gergen, 1978, 1985; Harré, 1979; Rosier, 1974). They criticized survey questionnaires and experiments as making rash assumptions about causality, generalizability, and the correspondence of meaning between the researcher and the researched. Drawing heavily on Gilbert and Mulkay’s (1984) discourse analytic study of biomedical scientists, Potter and Wetherell (1987) proposed a DA which focuses on language function, construction, and variation: “The principal tenet of discourse analysis is that function involves construction of versions, and is demonstrated by language variation” (p. 33). That the same “object” is constituted variably in different scenarios demonstrated for the authors how talk or discourse constructs reality to achieve particular functions: “Different versions and forms of talk are the analysable trace of the way language is used to bring about different ends” (p. 122). Discourse (and language and speech) was not to be treated as an “indicator of some other state of affairs” but to be analysed in terms of “how discourse or accounts of these things is manufactured” (p. 35, drawing on Gilbert & Mulkay, 1984; Potter, Stringer, & Wetherell, 1984). Their critique of social psychological work emphasized how “other socio-psychological methodologies have ignored or covered over the constructive, active use of language in everyday life” (Potter & Wetherell, 1987, p. 32). This version of DA has since been situated within a more general “social constructionist” criticism in/of psychology (e.g., Gergen, 1985; Potter, 1996a).

Another influence on Potter and Wetherell’s (1987) DA was the methodology of conversation analysis (e.g., Sacks, Schegloff, & Jefferson, 1974). Potter and Wetherell described the central focus of conversation analysis (CA) as an examination of sequences and the organization of “accounts” “in context” to discover how different types of actions and explanations are “produced and managed” (1987, p. 80). Conversation analysis “emphasises the vital nature of understanding each conversational turn in terms of the sequences in which it is embedded” (p. 94). These core principles of context and sequence were carried forward into the new discourse analytic method. It follows from this conceptualization of language and reality that personhood and “self” are also “relativized” because they are constructed in situated discourse to serve particular functions. As Potter and Wetherell (1987) put it “the question is not what is the true nature of the self, but how is the self talked about, how is it theorized in discourse?” (p. 102). In discourse, there is no psychological essence to a person, but it is possible to analyse function, construction, and variation in discourses about the self. Thereby it becomes “meaningless” to ask which social psychological theories of the self are “correct,” as they are each merely “ways of making sense of the self with their own context, relative advantages and disadvantages” (p. 102). How might such an approach to the self be located in time?

From Fordism to flexible accumulation

In The Condition of Postmodernity David Harvey (1990) argued that that there was a “necessary relation between the rise of postmodernist cultural forms, the more flexible modes of capital accumulation,” and “new dominant ways in which we experience space and time … within the organization of capitalism” (p. vii). In so doing, he drew upon “regulation school” theory (e.g., Aglietta, 1979/2001; Lipietz, 1986), which extended Marxist economic analysis by describing the different modes of capitalistic production and economic regulation which exist within older dynamics. Capitalism’s mode of production, its “regime of accumulation” (Aglietta, 1979/2001; Lipietz, 1986) was described as being enacted through a coherent network of individuals who configure a correspondence between workers, conditions of production, capitalists, financiers, and political and economic agents, which enable the regime to function for a period of time. Each regime needs to be “internalized” through consistent individual behaviours and social processes, and through norms, laws, habits, and regulating networks.

One such regime is Fordism. Henry Ford expanded prior models of corporate business organization (e.g., Taylor, 1911/1997), by explicitly connecting business and production with a whole new kind of society and human being—linking mass production with mass consumption and the democratic society, state actions, and questions of morality, discipline, and compliance (Harvey, 1990, pp. 125–140). The success of Fordism is said to have been a result of a relative balance of power between organized labour, corporate capital, and the nation state (see Polanyi’s 1944/2001 seminal work on the rise of the “market society”). Fordism was the dominant “regime of accumulation,” which lasted, intact, until around 1973.

By the mid-1960s the fixed exchange rates of the post-war period were floating, no longer linked to a fixed rate for gold (the Nixon shock) and were therefore more volatile. Harvey (1990)—drawing also on Huyssens (1984)—proposed that a shift in the capitalist mode of production and regulation occurred around 1972, from Fordism to “flexible accumulation.” The breakdown of Fordism was accompanied by a move away from Keynesian economic policies toward the empowerment of financial capital to determine social conditions, a move supported by aggressive neo-conservatism typified by the Reagan Administration in the US and Thatcher’s Government in the UK. London’s “big bang” reforms in 1986 signalled the implementation of Thatcher’s doctrines of unfettered competition and accelerated the geographical mobility of funds. In this regime of flexible accumulation, it became much more possible to make profit from shifting currency values and interest rates without producing any goods or services at all. “Paper entrepreneurialism,” “creative accounting,” and the production of debt or “fictitious” capital all occurred in this period (Harvey, 1990, p. 163). Autonomy from “real” production has been understood as the genesis of the contemporary globalized economy based on an “informational mode of production” (see Castells, 1996). Extending capitalism’s necessary condition of relentless expansion, the informational economy has been described as creating a global space which “is distinguished by flexible production, flexible political identities, porous borders, and an accompanying general fragmentation or fluidness (an ongoing process of disaggregation or ‘unbundling’) of modernity’s foundational concepts” (Perry & Maurer, 2003, p. ix).

As laid out by regulation school theory, Harvey spoke of the necessary integration between societies’ individual’s norms, attitudes, and practices; the mode of flexible accumulation; and neo-conservative political ideology. The emergence of the era of flexible accumulation coincided with a parallel shift in cultural conditions from the “solids of Fordism” (Harvey, 1990, p. 171) to an emphasis on the new, fleeting, ephemeral, fragmented, and contingent (see also Jameson, 1984 for investigation of a similar thesis). The internalization of competitive individualism was central to an entrepreneurial culture, and individualism was a necessary condition for the success of the capitalist mode of flexible accumulation. Whilst relationships between Fordism and psychology became critiqued in the 1980s (e.g., Bramel & Friend, 1981; Sampson, 1983), critical psychologists have paid less attention to the kinds of selves required by capitalism in an era of flexible accumulation. However, anthropologist of finance Bill Maurer asserts that “Subjectivities were constituted through the practices and requirements of industrial capitalism itself” (2002, p. 120), but “modern” subjectivities became too “static, limiting, predictable and linear” to explain “the kind of economic world we currently inhabit” (p. 121). Is it possible to understand the seemingly revolutionary project of Potter and Wetherell’s (1987) as a fit to the regime of flexible accumulation, which focused only on the “surface” and accepted and valorized the postmodern tropes of ephemerality, fragmentation, contradiction “and the chaotic currents of change as if that is all there is” (Harvey, 1990, p. 44)?

A genealogy of postmodern subjects

To develop such an analysis we draw on Michel Foucault’s notion of “genealogy” as a form of historical investigation which looks for the interplay of relations across seemingly unrelated fields, and diachronic discontinuities within seemingly unified disciplines (Foucault, 1972, 1977, 1966/2009). Critical of “continuous histories” which present the “spirit of an age,” “totalities,” and synchronies which gloss over differences, Foucault aimed to engender distrust in narratives of progress and evolution that purport to explain the present. Forms of psychology that at once present themselves as revolutionary and as developments of prior disciplines would appear to be logical sites for genealogical analysis.

We are not the first to extend Foucault’s ideas about genealogy to psychology. In Inventing Ourselves, Nicholas Rose (1996) undertook a genealogy of subjectification which “focuses directly on the practices that locate human beings in particular ‘regimes of the person’” (p. 25): “its domain of investigation is that of practices and techniques, of thought as it seeks to make itself technical” (p. 23). Rose’s (1996) genealogy is inspired by Foucault’s concept of governmentality (cf. Gordon, 1991), and Rose understood government as a “perspective from which one might make intelligible the diversity of attempts by authorities of different sorts to act upon the actions of others,” including strategies which “operate through ‘technologies’ of the self” (1996, p. 29). This investigation involves analysis along the following “linked pathways”: problematizations, technologies, authorities, teleologies, and strategies (pp. 25–29).

Our genealogy aims to show the political contingency of DA by first examining the treatment and interpretation of the scholars that Potter and Wetherell cited as having “laid the foundations for our current understanding of the operation of social texts” (1987, p. 8), “the foundations of discourse analysis” (pp. 9–31), upon which they “hope to indicate how a new style of socio-psychological research can be erected” (p. 32). These are linguistics (Chomsky, 1966), speech act theory (Austin, 1962), ethnomethodology (Garfinkel, 1967), and semiotics (Saussure, 1916/1966). Although Ferdinand De Saussure’s influential 1911 lectures (collected in the book Course in General Linguistics, 1916/1966) loomed large in the development of DA, our genealogy is focused on tracing the transition from Fordism to flexible accumulation. Consequently, we narrowed our discussion to the sense that Potter and Wetherell made of Chomsky, Austin, and Garfinkel, all three of whom published in the late 1960s as Fordism was entering a period of crisis. We aim to outline the elements, notions, and ideas that Potter and Wetherell retained from these foundations, how they extended and transformed them, as well as those elements they discarded. By so doing, we aim to show the contingency of Potter and Wetherell’s ontological conclusions from their observations about some very general features of language use that other scholars had previously considered.

Chomsky: From deep to surface structure

Noam Chomsky is well known for his critical writing on US foreign policy; international and American power relations; and news media, global politics, capitalism, and conflicts. However, Potter and Wetherell refer to Chomsky’s earlier work in linguistics (e.g., Chomsky, 1965, 1966). In 1955 Chomsky joined the Massachusetts Institute of Technology (MIT), Harvard University and has taught there ever since. Chomsky’s book Cartesian Linguistics (1966) aimed to “discuss the history of linguistics in the modern period” (p. 1), in which questions studied previously in the 17th, 18th, and early 19th centuries appeared to re-emerge. Such questions pertained to peoples’ ability to sustain coherence in response to novel situations, and the discussion of innovative and creative aspects of language use.

Chomsky’s theories—SYNTAX and “generative grammar” (Chomsky 1965, 1966) posited a base system (rules that generate underlying grammatical relations) and a transformational system (rules of deletion, rearrangement, etc. which form questions or imperatives, etc. in speech). These concepts assumed an ontology of “mind”—language was understood as reflecting the nature of inner thought, and as being an instrument of expression. The “deep structure,” “constitutes an underlying mental reality … a mental accompaniment to the utterance” (Chomsky, 1966, p. 34), to which the “surface structure” may or may not correspond. This dichotomy between the deep and surface structure of language was perhaps Chomsky’s most lasting legacy in cognitive science and linguistics.

Potter and Wetherell (1987, pp. 172–182) acknowledged Chomsky’s contribution to linguistic theory whilst they rejected the possibility of investigating or discovering anything about “the world out there,” “the world under the skull,” or anything “lying beyond” language:

Because discourse analysts do not accord different status to the “inner” and the “outer,” or to the “mental” and the “non-mental,” the question of how, precisely, a person’s description of their mental state represents or matches that mental state becomes irrelevant. The descriptive accuracy of discourse and its adequacy as a map or chart of private, subjective, mental experience is the non-issue from our perspective. (p. 179)

This assertion was underpinned for Potter and Wetherell (1987) by their interpretation of the evidence of “variation” in language use. Variation in language was used as evidence to support a very different theory of language and of the self than Chomsky’s. It underpinned a broad critique of the ontology of “inner” events. For example, they re-cast the notion that the same people espouse different “attitudes” or constitute the same object differently at different times, as evidence that language is variable and constitutive of social opinions (pp. 43–55). As in Foucault’s early work, “discourse” became a term that Potter and Wetherell used to describe text and language, but without any implication of underlying (or “deep”) structures to human communication. Quotations from interviews in DA papers were not “illustrations” of deeper structures; “they are the topic itself, not a resource from which the topic is built” (p. 172).

They asserted that variation in language is used to deal with “an ever changing kaleidoscope of situations” (Potter & Wetherell, 1987, p. 156). Of course, Chomsky had also considered “variation” in language use, but for him language’s “boundless” potential to deal with an “unbound and unpredictable set of contingencies” (Chomsky, 1966, p. 20) indicated the presence of innovation and creativity in the transformational system to adapt and transform the base structure. Chomsky also suggested that the “boundless” creative potential of language emphasized by Descartes and his followers, “leads to a search for coherence, unity in experience” (1966, p. 18). David Harvey (1990) asserted that the “categorical fixities” of the Enlightenment project were increasingly challenged across many fields by the mid 1900s. Simultaneously, a new project was created—the myth of “heroic modernism”—which aspired to redeem us from “the formless universe of contingency” (p. 31), manifest in material practices and projects to re-build war-torn economies. Across many academic disciplines scholars employed “relativism and multiple perspectivism” to access a “unified, though complex, underlying reality” (p. 30). A quote from the modernist architect Le Corbusier: “By order bring about freedom” (Harvey, 1990, p. 32), recalls the sentiment in Chomsky’s work that language has a fixed underlying structure, but is limited in its surface possibilities only by the limits of creativity and innovation of expression.

In contrast to this humanist enlightenment sentiment, Potter and Wetherell endow only the discourse analyst—and not all human language users—with coherence in language. According to them, coherence needed to be built during analysis of a discourse’s functions, from the “detail of passages of discourse, however fragmentary or contradictory” (1987, p. 168). Coherence allowed the reader to assess the “validity” of a discourse analytic study: “a set of analytic claims should give coherence to a body of discourse”; “how a discourse fits together and how discursive structure produces effects and functions” (p. 170). Coherence was, for Potter and Wetherell, not given in the nature of the mind, but was an ideal to be achieved through the practice of discourse analysis.

In sum, variation and contradiction in language performance were taken by Chomsky and by Potter and Wetherell to evidence very different ontologies of the person. Chomsky’s conceptualizations in linguistics appear in line with the heroic modernist search for coherence, whilst Potter and Wetherell discarded the potential for language to indicate “realities,” “selves,” or “states” outside of language in line with the postmodern emphasis upon the “surface.” If we engender the distrust of narratives of progress and evolution as suggested by Foucault’s genealogical principle of discontinuity, then Potter and Wetherell’s departure from Chomsky’s ontological position seems less revolutionary than as co-occurring with the shift from Fordism to flexible accumulation, toward the internalization of a different kind of personhood. In other words, Potter and Wetherell’s use of Chomsky’s linguistics can be understood as a “problematization”—one of Rose’s “linked pathways” in the genealogical analysis of subjectification that demonstrates the “psychologised understanding of what it is to be human as the site of a historical problem” (Rose, 1996, p. 23). Potter and Wetherell’s use of Austin’s work in the philosophy of language from the same period demonstrates another slant on this shift.

Austin, indexicality, and “truth”

John Langshaw Austin was a British philosopher of language who served the UK government’s foreign intelligence service during the Second World War, before becoming White’s Professor of Moral Philosophy at the University of Oxford’s Centre for Ethics and Philosophy of Law. In the mid-1950s Austin visited Harvard University, where he met Noam Chomsky and gave the lectures from which the material was later collected into the book How to do Things with Words (1962), and the essay “A Plea for Excuses” (1956/1968). He is widely understood as a pioneer in ordinary language philosophy—which pays close analytic attention to the features of “ordinary” everyday language use. Austin reacted against the logical positivist philosophy of language, and has since been described as “a thinker whose whole work is permeated by a desire to make his audience aware of and lose confidence in the habitual imposition of preconceived categories” (Gustafsson & Sorli, 2011, p. 2). We could understand Austin’s philosophy as reacting against the “formless universe of contingency” (Harvey, 1990, p. 31), not by searching for “coherence, unity in experience” in the way Chomsky did, but by re-conceptualizing contingency in a way that attempts to formalize the contingent relationships between language, material conditions, and situational circumstances.

In How to do Things with Words, Austin (1962) cited Emmanuel Kant’s notion of the “descriptive fallacy” (p. 3)—the mistaken notion in the philosophy of language of the existence of “statements” which directly report reality. Kant believed that such utterances “masquerade” as descriptions of facts which report on truth or falsity, yet “statements” should be “read” as containing indications about the context within which they were made, and as to how they should be heard and perceived. Austin argued that within the speech act, “stating” and “describing” have no “unique position” in relation to “facts” (1962, p. 149), and that “the traditional statement is an abstraction, an ideal, and so is its traditional truth or falsity” (p. 148). Potter and Wetherell (1987) shared Austin’s focus upon the importance of “context”: “the total speech-act in the total speech situation is the only actual phenomenon … we are engaged in elucidating” (Austin, 1962, p. 148). For Potter and Wetherell (1987) “accounts are examined as they occur, in context, embedded in conversational sequences” (p. 81). However, Potter and Wetherell diluted Austin’s notion of “necessary conditions” upon which utterances depend to achieve effects or functions. Austin’s doctrine of infelicities insisted that the “happiness” of an utterance requires, for example, “an accepted conventional procedure having certain conventional effects”; “the appropriate persons and circumstances”; that persons must have the “thoughts and feelings invoked by the design of the procedure and intend to conduct themselves [accordingly]” (Austin, 1962, pp. 14–15).

Indexical language was not taken as evidence by Austin of an ontology which views the subject, the self (the world “under the skull”), and reality (the world “out there”) as constructed by language. Rather, for Austin, the surface appearance of language depends upon internal and external conditions; for example the appropriate circumstances, thoughts, feelings, and intentions. Potter and Wetherell (1987) elevated and extended this part of Austin’s argument: “Given the essentially performative and indexical nature of language use how can researchers constitute it as a neutral record … of cognitive mental states?” (p. 145). The indexicality of language use was taken as further evidence that language is constructive, thereby supporting the methodological focus on discourse as “a topic in its own right.”

The functionality or “action-orientation” of talk for Potter and Wetherell (1987) was made evident by variations within or across talk: “Different versions and forms of talk are the analysable trace of the way language is used to bring about different ends” (p. 122). So, although Potter and Wetherell insist on attention to context, they moved away from Austin when they asserted that there are only different versions of reality which are “manufactured” through “accounts,” and that discourse is to be studied as the topic in its own right: “our focus is exclusively of discourse itself, how it is constructed, its functions, and the consequences which arise from different discursive organisations” (Potter & Wetherell, 1987, p. 178). In effect, only discursive context and consequences can be studied. Brewster Smith (1994) quoted Gergen (1991) drawing on the philosophy of Richard Rorty in his critique of postmodern theories of selfhood, “to eradicate the distinction between world and mind, object and subject” meant that “world” and “mind” become entries in the discursive practices of the culture (Brewster Smith, 1994, p. 408).

As with Potter and Wetherell’s (1987) interpretation of Chomsky, their interpretation and extension of Austin’s work seems consistent with the shift from Fordism to flexible accumulation. Since the publication of Discourse and Social Psychology, some critical psychologists have argued that the focus on discourse alone undermines the potential for discourse analysis to be used as the powerful socio-political critique of social contexts that it aspired to become (e.g., Nightingale & Cromby, 2002).

David Harvey (1990) asserted that neo-conservatism and postmodern cultural forms and philosophies were necessarily related to the late capitalist mode of flexible accumulation, which is capable of operating “in ways that seem almost oblivious of the constraints of time and space that normally pin down material activities” (p. 164), in other words, in ways that are oblivious to material context.

We were reminded that such dematerialization of context can serve quite conservative ends by a conversation recorded by journalist Ron Suskind (2004) with a senior neo-conservative advisor to George W. Bush, which demonstrates neo-conservative governance strategy as a kind of “war on reality”:

We’re an empire now, and when we act, we create our own reality. And while you’re studying that reality—judiciously, as you will—we’ll act again, creating other new realities, which you can study too, and that’s how things will sort out. We’re history’s actors … and you, all of you, will be left to just study what we do.

A psychology which rejects the possibility of discerning through language anything “fixed” about the self, reality, or any contextual conditions; a theory where we create ourselves anew each time we speak idealizes such perpetual, endless studying. This quote demonstrates the underpinning genealogical principle of approaching analysis “laterally” to identify “fragmentary, dissociable discursive acts working in conjunction” perhaps with “mutual unawareness” (Foucault, 1981; see also Hook, 2001).

Garfinkel: Practice and the achievement of reflexivity

Harold Garfinkel served in the Second World War before joining the new Department of Social Relations at Harvard University, a period when army and post-war corporate and government demands were made for studies into the sociology of “common sense” (Australasian Institute for Ethnomethodology and Conversational Analysis, 2011). John Heritage has argued that Garfinkel “spoke directly to the mood of the moment,” politically and within the discipline of sociology, by placing “renewed stress on the role of human agency in social life”; reacting against prior sociological methodologies which saw “social actors as simply the passive bearers of sociological and psychological attributes” (Heritage, 1984, p. 2). Garfinkel’s early work is best represented by Studies in Ethnomethodology (1967), an investigation into the “facts” of society by analysing the settings of everyday activities, and societies’ members’ ways of making these settings “accountable.” We will not attempt to do justice to the complexities and subtleties of Garfinkel’s works here, but will focus on a key difference between the scope of Garfinkel’s project and Potter and Wetherell’s.

Garfinkel spent his academic career studying “the sociology of everyday life,” including investigations into the nature of inter-subjectivity and the social constitution of knowledge. He insisted that the topic of study for sociologists should be the “methods of common understandings” in moment-by-moment social interaction which formed the “operational structure” of “interpretative work” by members of a society. He described these phenomena and practices as “an endless, on-going, contingent accomplishment,” requiring and constituting “knowledge, skills, competence and entitlement” (1967, p. 2). The multitude of these methods was indicated by the “endless list” of ways in which people speak. People demonstrate their “bona fide” status in society by making evident the “rational” (known in common) character of their activities and sense-making, via the elaboration of “institutionalized knowledge of the real world” (p. 53). Common sense knowledge was described by Garfinkel as the “seen but unnoticed,” “expected background features,” which entitle claims and conduct (e.g., pp. 36, 118). All features and properties of activities are “to be treated as a contingent accomplishment of socially organised common practices” (p. 34). Organizations of activities sustain, resist, or facilitate those methods “for making their affairs accountable-for-all-practical-purposes” (p. 34).

Garfinkel discussed “variation” and “function” long before Potter and Wetherell did. He asserted that variation, in “the endless list of ways that people speak,” indicates the multitude of methods used to accomplish “accountability.” These methods included: “marking, labelling, symbolizing, emblemizing … analogies … miniaturizing, imitating … used in recognizing, using and producing the orderly ways of cultural settings from within those settings” (1967, p. 31). Potter and Wetherell (1987) described how Garfinkel showed “the constructed nature of even apparently obvious natural categories like gender” (p. 156), and that ethnomethodology and DA share an interest in “how categories are constituted in everyday discourse and the various functions they satisfy” (p. 116).

However, Garfinkel never suggested that discourse was to be studied as “the topic in its own right” (Potter & Wetherell, 1987, p. 35), with regard to gender or any other topic. For Garfinkel (1967), the “stability” of a person’s actions varied depending on the “real conditions” that guarantees their “motivated compliance” with a background texture of relevances: a perceived legitimate order of beliefs about life in society as seen from “within” the society (e.g., p. 54). Like Austin, Garfinkel included talk and texts as contingent parts of a whole system. Conduct, events, and speech were seen as continual processes of accomplishing “normative accountability” (to use Garfinkel’s term) via the elaboration of rules and context. Garfinkel attempted to identify and assess “seen but unnoticed” background features which are accomplished in an interaction, similarly to the way Austin identified the conditions required for speech-acts to achieve “happiness” or be felicitous. Potter and Wetherell’s reference to Garfinkel’s study of “apparently obvious natural categories like gender” (1987, p. 156) is a reference to his study of Agnes (1967, pp. 116–185), who managed her continuous interpersonal exchanges of “passing” as a woman by using certain “devices” including avoiding particular situations, navigating physical examinations and learning skills that accord with a particular identity.

Potter and Wetherell (1987) took variation as evidence that discourse is constructed to serve certain functions in particular contexts. Although this emphasis on variation recalled Garfinkel, Potter and Wetherell tended to conflate variation with contradiction in a manner that Garfinkel did not. Variation (contradiction and fragmentation) across passages of talk or text (discourse) demonstrated to them that language constructs reality and the self. However, for Garfinkel, variation indicated methods used by members of a society to achieve accountability within that society (accountability is the function); Potter and Wetherell extended the notion of variation to support their ontological perspective, and to detach their “speaking subjects” (1987, p. 95) from materiality. It is one thing to insist, with Garfinkel, that gender identities are a contingent ongoing accomplishment, and quite another to insist that they are constructed with, by, or in words alone. This shift may be particularly problematic for analyses of gender. Transgender scholars have been vocal critics of the theory that their lives and experiences ought to be figured to argue the seemingly revolutionary argument that gender is “socially constructed” (Namaste, 2000). Indeed, attention only to words as sites of gender construction can erase the contexts and lived experiences of people whose bodies and identities do not fit the “two-gender system” that Garfinkel identified as normative (Lombardi, Wilchins, Priesing, & Malouf, 2001).

Potter and Wetherell (1987) acknowledged that their own book was “subject to the same processes and the same constructed action oriented aspect” and said that the work of social scientists should also be seen as “versions of versions” or “accounts of accounts” (p. 4). Garfinkel similarly (1967) applied his ethnomethodology concepts to social scientists that were also described as partaking in “motivated compliance” to be seen as bona fide members of society, by demonstrating rationality and reflexivity in their writings. Reflexivity for Garfinkel was conceptualized as a “device,” used to accomplish demonstrable rationality based on speaking or writing and conducting oneself in line with “seen but unnoticed” background expectancies; drawing on the assumption that other members share common sense knowledge of the social world and its particular organized or institutionalized practices. In discourse analysis the concept of reflexivity is elevated into a methodological ideal, a “research strategy” (Potter & Wetherell 1987, p. 183) to be applied to any data/discourse. In the past quarter century, “reflexivity” has increasingly become taught to students as a technical achievement of qualitative psychology; “part of a process of making research more accountable, more transparent and easier to evaluate” (Lyons & Coyle, 2007, p. 114). Reflexivity is supposed to increase accountability by acknowledging the researcher’s influence, opening the text up to alternative interpretations. The linguistically constructed (therefore variable, contradictory) reflections of the researcher’s “self” is “accounted for” whilst they “construct a purposeful account of their texts” (p. 114).

As this difference between Garfinkel, Potter and Wetherell, and more recent qualitative psychologies makes clear, the moral weight placed upon “doing” reflexivity can vary considerably. In The Order of Things, Foucault (1966/2009) rejected the notion that “reflexivity can be grounded by ontology” because “man” is the object and subject of knowledge (Smith, 2005, p. 18). Smith (2005) highlighted that Foucault asserted the importance of historicizing reflexivity as a “condition of knowledge”:

Reflexivity is the form of thought of a particular regime of truth at a particular period in history … if reflexivity does now mark out a field of disciplinary practices, it does so in consequence of the particular practices of governance which mark the modern world and not as a result of the ontology of being human. (p. 19)

If we think again of reflexivity in Garfinkel’s terms as functional, motivated, and achieved, then Potter and Wetherell and later authors could be understood not as “more reflexive” than other psychologists, but as leaning more on reflexivity to demonstrate their rationality.

Conclusion: From genealogical shifts to constantly shifting accountability

A transformation took place away from subjects with deep mental realities, material contexts, and practices in Potter and Wetherell’s (1987) theory of the person, a shift that was not given by these authors’ quite valid observations about the variability, indexicality, and social nature of language. If genealogical analysis asks “what forms of life are the aims, ideals or exemplars for these different practices for working on persons … ?” (Rose, 1996, p. 27), then the three examples above suggest a shift from theories of the modern subject that fit with Fordism to theories that fit with the flexible mode of accumulation. Extending Rose’s genealogies of subjectification into the late 20th century, as we do here, suggests that Potter and Wetherell’s DA can be seen as one of those technologies which takes “modes of being human” as its object and which “produce[s] and enframe[s] humans as certain kinds of being … within a technological field” (Rose, 1996, pp. 26–27). Rose (1989b) has demonstrated the analysis of teleologies by drawing on Max Weber’s (1922/1978) analysis of capitalism as a system enmeshed within a programme of governance, operating upon the “psycho-physical apparatus” of humans:

Economics, in the form of a model of economic rationality and rational choice, and psychology, in the form of a model of the psychological individual, have provided the basis for similar attempts at the unification of life conduct around a single model of appropriate subjectivity. (p. 28)

Potter and Wetherell can be read as dismantling the “psychological” framework that Weber critiqued, within a different era in which selves were obliged to find freedom in “ferment, instability and the fleeting qualities of the postmodern” (Harvey, 1990, p. 156).

This new ontology of personhood was certainly “revolutionary” in the sense of being novel, but perhaps less “revolutionary” in the sense of being more liberating than the psychologies that went before it. Before drawing our argument to a close, it is worth considering how locating Potter and Wetherell’s (1987) text as the psychology that fits the mode of flexible accumulation might reframe their critique of modernist psychology.

Potter and Wetherell’s (1987) book consists of a series of demonstrations of the ways the discourse analysis method can show us that “consistency in accounts is often overstated by the various aggregating techniques commonly used by psychologists” (p. 34). They described “orthodox social researchers” as aiming “to obtain or measure consistency,” and that attempts to highlight consistency in previous social psychological work covers over the “function of participants’ constructions” (p. 163). Instead, they insisted upon the existence of variability, as noted above. Interviewers should “try to generate interpretative contexts … in such a way that the connections between the interviewee’s accounting practices and variations in functional context become clear,” and “techniques which allow diversity rather than those which eliminate it are emphasised” (p. 164). The presence of variability is visibly normative within this ontology, as Potter and Wetherell (1987) expend effort to explain away consistency: “orderliness in discourse will be viewed as a product of the orderly functions to which a discourse is put” (p. 34), “consistency may be the product of accounts sharing the same function” (p. 35). Such passages suggest the validity of considering whether variability, rather than consistency, could be an “achievement of the researcher rather than a feature of the discourse” (Potter & Wetherell, 1987, p. 103).

We would argue that the traditions critiqued by Potter and Wetherell (1987) were not as invariably opposed to contradiction as this account would suggest. As Morawski (2005) has shown, “reflexive” critiques of the modernist science model of 20th century experimental psychology has often co-existed with the practice of that science, from William James’ “psychologist’s fallacy,” through Horace Mann Bond’s experimental manipulations of the conditions for IQ testing, to Susan Rosenzweig’s assertion that “the experimental situation in psychology is itself a psychological problem” (Morawski, 2005, p. 89). Morawski conceptualized these critical works as using forms of reflexivity to uncover the implicit assumptions contained in epistemologies, methodologies, and ontologies. Potter and Wetherell’s criticisms were presented without considering psychology’s complex intellectual history, or its relationship with other fields, in a way that seems as “general, afunctional, decontextual” (1987, p. 54) as the constructs they bemoaned in the workings of experimental and survey methods in social psychology.

Following Morawski (2005), perhaps our genealogy can show “the cultural and political contingency of the methods that psychologists [in this case Potter and Wetherell, 1987] purportedly derived from scientific epistemology” (p. 95). The author of a discourse analysis is engaged in a contradictory project that is committed to an ontology of language as only properly analysed when its endless contradictions are brought forth and when the analysis itself heroically escapes the ontological world in which all language is riddled with contradiction. Whilst the emphasis on reflexivity appears to make the author of a discourse analysis accountable for interpretation at first glance, discourse analysis also moved away from Garfinkel’s critique of “reflexivity” toward rationalizing reflexivity as technique.

Of course we are not the first to describe how discourse analysts might gerrymander ontologies with regard to their own and their participants’ discourses (Hammersley, 2003). However, by locating Discourse and Social Psychology (Potter & Wetherell, 1987) in historical time we hope to have provided an alternative to the textbook consensus that discourse analysis is a radical epistemology that exemplifies how qualitative methods differ from the methodological and epistemological “mainstream.” Discourse and Social Psychology might not be a radical break with psychology, as its fluid movement between critique of psychology, and epistemology of the psychological, suggests. It is not our intention to suggest “improved” alternative directions for psychology in this article, or to spawn new methodologies from our exploration of the political contingency of Potter and Wetherell’s discourse analysis. We have hoped to demonstrate through genealogy that we should be sceptical of knowingly post-Kuhnian texts which make claims of radicality and revolution within a multi-paradigmatic discipline. As Erin Driver-Linn (2003) argues, psychology papers often reference Kuhn’s work, and such references provide psychologists’ claims with rhetorical leverage, but Kuhn’s (1962/1996) model of historical change in science is often cited superficially and uncritically. We should be suspicious of those who claim to portend the next in a series of paradigm shifts.

Maurer (2002) described finance in the era of flexible accumulation as “distributed [in] networks of human and technological agents … [that] seem to work all by themselves and rework the world” (p. 18). Within the capitalist mode of flexible accumulation and the accompanying thesis of postmodernism, “anything which could be recognised as a ‘subject’ would disperse into new networks of power with no originary point, cybernetic structures of infinite feedback and nonoriginary unfolding” (p. 125). Maurer asks:

What kind of politics are possible in a world where there is no easily identifiable subject on whom to place blame, no one whom we can take to task for a now nonlocatable structure of domination that is itself a part of us? (p. 141)

Similarly, the ontology of ephemerality, fragmentation, and contradiction in discourse analysis creates a position for the author that only appears accountable for the psychological interpretation of others. Michael Billig (1995) argued that the postmodern thesis overlooked the “banality of nationalism”—that the national and the international have always been intermixed, and this is “especially marked in the case of a nation [the United States] bidding for global hegemony” (p. 128). He can remind us perhaps that we should be aware of the reality behind the “unreality politics” mentioned above.

Billig asserted that a large part of the postmodern thesis, shared among scholars, was that “nationhood is unravelling”: it’s “as-if, at this moment around the world, vast armies are not practising their battle manoeuvres beneath national colours” (1995, p.139). He theorized that “philosophy performs an ideological task: it formulates abstract, universal principles which, in fact, express, and mask, the particular interests of ruling groups” (p. 154). In an era when governments, corporations, and individuals are inherently fractured by contradiction and no longer treated as whole, stable entities as they arguably were during Fordism, discourse analysis mandates flexible subjects who float free of material context. The judicious study of new and constantly shifting realities forbids the discussion of worlds that are both inner and outer to our words.