One important objection to the notion of ‘internal conversation’ rests on a broader trend within contemporary social theory that is concerned with the possibility that theoretical claims about agency lead proponents to make claims about agents which are empirically inadequate. So too that these ensuing claims might find themselves implicated, knowingly or otherwise, in broader political contestations within wider society (though the suspicion persists that those most vocal about the purportedly intrinsically political nature of social science and social theory systematically, indeed hubristically, overestimate the influence of social scientists and theorists outside the academy). On such a view, affirmation of a  choosing, reflective and deliberative self is unavoidably embroiled within a broader project of neoliberal governmentality (Rose 1998, 1999, 2007). Nonetheless, it seems profoundly mistaken to therefore expunge conceptual talk about subjects from the admissible range of contemporary social theory. Recognising this as a mistake is entirely compatible with the methodological move Miller and Rose (2008) make in their genealogical investigations into ‘technologies of the self’: “what understandings of the people to be acted upon – whether explicit or implict – underpinned these endeavours, and how did they shape or reshape the ways in which these individuals understood and acted on themselves?” (Miller and Rose 2008: 7).

While it might be objected from a realist standpoint that such an investigation presupposes a theory of the subject, it is important that we nonetheless understand the ambiguous role such an account occupies within a broader inquiry of this form. As the author’s themselves observe, “that question could only be answered on the basis of some explicit or implicit assumptions about human mental processes. Yet for us, the historical forms taken by those presuppositions were exactly what we were studying” (Miller and Rose 2008, 7). It is perfectly possible to accept the validity of such an injunction, construing it as an exercise in bracketing to facilitate a specific form of inquiry, while rejecting the more radical implication that all claims about the underlying properties and powers of human subjects are, in actuality, claims about the cultural resources and reflexive technologies which are distributed within a given social context at any given time. Such an absolute injunction would involve a failure to distinguish between what Bhaskar (2011: 21) terms the transitive and intransitive objects of scientific inquiry: the “changing cognitive objects that are produced within science as a function of scientific practice” and “the unchanging real objects that exist outside the scientific process” respectively.

Bracketing the real properties and powers of human subjects may be a useful move for investigation into the empirical variety of reflexive technologies over time, insofar as that it minimises the role that our prior (transitive) commitments play in the empirical investigation. In doing so it facilitates a largely descriptive, though nonetheless valuable, mode of investigation which traces out the socio-cultural factors involved in empirically observable changes in how human beings “recode variations in moods, emotions, desires, and thoughts” (Rose 2006, 223). In doing so it can lead to empirically rich accounts of how, for instance, the rise of a “psychological language of self-description: the language of anxiety, depression, trauma, extroversion, and introversion” was connected, inter alia, to the use of psychological tests of intelligence and personality from vocational guidance to military promotion” and the rise of “psy technologies for marketing commodities” or the “proliferation of psychotherapies” (Rose 2007, 187-188). Within the framework of this thesis, the objects of such investigation are understood as cultural resources (ideational objects) and reflexive technologies (the socio-cultural application of these objects). It is necessary to understand the properties and powers of such cultural objects in order to explain socio-cultural variation in modes of “seeing, judging, and acting upon human normality and abnormality” and how we “our desires, moods, and discontents” are mapped onto differing images of the human (Rose 2006, 187-193).

Through investigation into such objects, whether or not it takes a genealogical form, it becomes possible to empirically flesh out the sense in which, as Archer puts it, “our reliance upon the public domain for thinking can be upheld, without this determining what we do with it – that is the contents of our mental activities” (Archer 2003, 69). The difficulty with Rose’s work, as well as the broader corpus of sociological thinking of which it is an outstanding exemplar, lies in its inability to make sense of how such cultural objects (which it has mapped in a highly detailed and sophisticated manner) are mediated at the level of an individual subject. Such subjects are perpetually implied within Rose’s work, with continual references to reflexivity in Archer’s sense (i.e. the relationship of a self to a self) implicit in the substantive claims made about shifting constellations of technologies of selfhood, yet remain curiously absent. Unfortunately this absence precludes the possibility of gaining concrete explanatory purchase on how particular cultural objects are mediated by particular subjects – Rose’s account is laudable in its detail at the level of the former yet conspicuous in its generality at the level of the latter.

Rose actually does offer something analogous to a theory of the subject, though unsurprisingly it is framed in terms which deny this. He writes that his engagement with the question of subjectivity is offered “not in terms of the effects of ‘culture’ upon ‘the person’, or in terms of a ‘theory of the subject’, but by seeking to characterize the mode of action, as it were, of the diverse psy technologies of subjectification that I have discussed” (1998, 170). As well as the tacit admission that the lack of engagement with subjectivity was the glaring omission in his otherwise accomplished body of work, his explicit framing of the ensuing account in terms of the mode of operation of ‘psych technologies’ is telling, in that it leaves the subject as little more than an explanatory lever, invoked merely to flesh out the absent subjective moment of his broader account. What explains this continual hostility to abstract models of the subject?

Archer (2000) addresses the same question in an exploration of a body of work which is undergirded by a similar cultural politics. As she observes of Richard Rorty’s anti-humanism, his injunction against substantial conceptions of the human has a normative component to it. Given Rorty’s desire to nonetheless make normative claims pertaining to human beings, it is inevitable that the human resurfaces and, with it, so too does the problem of structure and agency (Archer 2000, 40-43). So too with Rose and his Deleuzian account of the subject, as well as its concomitant insistence that “the ‘question of agency’ as it has come to be termed, poses a false problem” (Rose 1998, 186). On the one hand, it is denied that the human is “an actor essentially possessed of agency” and on the other that they are a “passive product or puppet of cultural forces”. The reintroduction of agency into Rose’s ontology leads him to make a move surprisingly reminiscent of structurationist theory, transcending the dichotomy of structure and agency through central conflation (Archer 1995). His attempt to avoid an affirmation of agency leads instead to a particular understand of the bridge between structure and agency, such that former is understood to continually shape the latter through an ongoing process of ‘enfolding’, only to be reproduced and sometimes transformed by the latter, as a consequence of the radical contingency and therefore underdetermination which characterises the process.

However the point here is not to critique Rose but to elucidate the difficulties inherent in the treatment of subjectivity within contemporary social theory. I have suggested that there is a methodological objection to abstract treatments of subjectivity, which can be relativised to a particular mode of inquiry and dismissed when claimed to apply more broadly. There is also a normative objection, a broad discomfort with are assumed to be unavoidably normative implications of talk about ‘humans’ and ‘subjects’, as well as the belief that academic thought and talk about such matters entrenches the hegemony of the ‘liberal self’. But there is also a concern about the empirical difficulties which are seen to be contained within theoretical accounts of human properties and powers. Such difficulties become more pronounced when the issue is framed within a particular substantive area of investigation. For instance Heaphy (2012) adroitly identifies the implications which the widespread uptake of Giddens et al within sexuality studies has had on the empirical portrayal of the lives of LGBT individuals within contemporary Britain.

Heaphy takes issues with a pervasive tendency to hold up LGBT lives as exemplars of reflexivity in the first sense, identifying a range of strands in the sexualities literature of which this is true (Heaphy 2012: 17). He argues that, as a whole, these represent a “powerful story” about LGBT lives as “reflexively achieved forms of existence that are the exemplars of the life politics of self-fashioning” (Heaphy 2012: 19). Furthermore he suggests that the appeal of such accounts stems from the affirmation of LGBT agency implied by them, in contrast to the previously dominant Foucauldian vision of sexualities which tended to stress disciplinary subjection. Arguments about LGBT reflexivity, as perhaps did Foucault’s account in an earlier political era, have an intuitive plausibility because of the wider social circumstances in which they are articulated. As Heaphy observes, “it seems clear, after all, that lesbian and gay sexualities hare more ‘empowered’ and visible in the culture than ever before, and recent legislation in Britain and elsewhere (such as the Civil partnership and other Acts) seems to promote and defend the legitimacy of same-sex relationships” (Heaphy 2012: 19).

However Heaphy raises a number of problems with such accounts. He suggests that these prevailing narratives of LGBT reflexivity have been characterised by a “blurring of arguments about theoretical possibilities and empirical actualities” i.e. a theoretical affirmation of agency leads proponents to make claims about agents which are empirically inaccurate. In doing so the realities of difference are occluded, such that “exclusive and well-resourced lesbian and gay experience is valorized while other experiences are made invisible”. This, he argues, is a consequence of insufficient attention to power, particularly in an indifference to the “relationship between power and sociological narration” (Heaphy 2012: 20). He goes on to argue that in order to take the “differences that are shaped through the intersections of class, race and ethnicity, generation, geographical location and like” seriously we must acknowledge “that there is no one lesbian and gay experience or forms of existence, and that lesbian and gay living should be studied in their diversity of forms”. In doing so, we might come to ask “how significant resources (economic, social, cultural and corporeal) are in shaping different possibilities for lesbian and gay living, and how their embodiment gives rise to different possibilities for identification, relating and life political practice” (Heaphy 2012: 21). Heaphy argues that a move towards reflexive sociology within sexuality studies, as part of a Bourdieusian turn which moves the study of LGBT lives away from Giddens and Beck, would help rectify this worrying tendency to homogenise the lived experience of LGBT individuals and treat their lives as if difference didn’t matter.

While applauding Heaphy’s broader aims and accepting elements of his critique, this direction of travel is nonetheless revealing of profound conceptual confusions relating to what reflexivity is and how it operates. The broader shift he identifies from Foucauldian conceptions of sexuality (excessively structural) to voluntaristic accounts influenced by Giddens (excessively agential) reveal an inability within sexuality studies, as well as social theory more broadly, to come to terms with the problem of structure and agency. One approach elucidates the role of structure while obliterating agency. The other elucidates the role of agency while obliterating structure. The two approaches each contain an element of truth but, in their inability to proceed beyond their own theoretical terms of reference, neither is able to do justice to the ambivalence of human experience.

Both freedom and constrain co-exist in our daily experience. We choose and yet we are denied choice. We shape our circumstances and yet our circumstances shape us. We make our way through the world and yet the maps we use and the paths we choose from forever elude our full understanding, let alone our control. We are subjects and we are subjected. In fairness to Giddens, attempting to reconcile this duality is at the heart of his theoretical project. Yet the empirical inadequacies which so often result from attempts to adopt his approach as an explanatory framework are indicative of the conceptual error at its heart. Unless we conceptualise reflexivity in a properly mediatory manner, as being the human power which allows the pursuit of courses of actions by (fallibly) taking stock of objective circumstances and our subjective concerns, the problems Heaphy correctly identifies will inevitably ensue. But if we do understand reflexivity in such a way, these problems do not occur. The issue here is not reflexivity as such. The issue is conceiving of reflexivity in a way which detaches it from the constraints and enablements an individual is contingently subject to at any given moment. If we conceive reflexivity in a manner which is fundamentally relational, such that our degree of freedom or constraint is an empirical matter emergent from our circumstances at a particular moment in time and the biographical pathway which led us to such circumstances, then these contrasting images of human life (LGBT or otherwise) as either overly-free or overly-constrained simply do not emerge.