Copyright Sociological Research Online, 2000


Roger Slack (2000) 'Reflexivity or Sociological Practice: A Reply to May'
Sociological Research Online, vol. 5, no. 1, <>

To cite articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 17/1/2000      Accepted: 7/4/2000      Published: 31/5/2000


The paper constitutes a response to May's concept of reflexivity, and argues that debates on reflexivity have missed the need to ground their claims in the life world of society members - thus promoting the very ironic stance they seek to address. A re-articulation of claims to reflexivity is made in the distinction between 'essential' and 'stipulative' reflexivities wherein the former is grounded in members' observable-reportable natural language practical actions, while the latter remains the province of the analyst and subjects members' versions to sociological remedy. The paper suggests a return to the work of Garfinkel (1967) as a means of respecifying the grounds of the reflexivity debate.

Ethnomethodology; Reflexivity; Sociological Description; Sociological Research

Reflexivity or Sociological Practice

In this brief paper, I want to examine some of May's (1999) comments about the scope and status of reflexivity in sociology. In so doing I will argue that he has continued with the very epistemological moves that he seeks to critique. This is, for me, an indication of the continued confusion and conceptual conflation that exists around the notion of reflexivity in sociological discourse.

In order to offer some way out of this morass, I will advocate two modes of reflexivity - the stipulative and the essential. I will argue that the former stands as a sociological achievement with which society's members need not be concerned, while the latter stands as an achievement of society members with which sociologists by and large have not been concerned. Taking an ethnomethodological standpoint I shall argue that we need to attend to members' reflexivity and to take it seriously in our explications of social life. In sum, the only way out of the postmodernist, structurational and textual maze is to attend to the practical essential reflexivity of society members[1].

At the outset I should say that May's paper is an interesting examination of the current variants of reflexivity to be found in sociological circles. However, the central point of my argument here is that, in fabricating a new set of dichotomies between endogenous and referential reflexivity, May repeats the fundamental mistake at the heart of sociological writings on reflexivity - that is, he regards reflexivity as a sociological concern to be addressed and remedied by sociologists as opposed to society members. What counts as reflexivity is an achievement of the sociologist for sociology. Just as those who advocate the textual turn (Clifford and Marcus, 1996) have moved us away from the world into the text, May moves us into a concern with reflexivity for sociological practice as undertaken by sociologists. My problem with this is that social life is undertaken by society members, who have no need of (or interest in) sociological descriptions in the formulation of accounts of their practical actions for all practical purposes (for a discussion of formulations as conversational objects see Heritage and Watson, 1979). Another stratum of sociological reflexivity perpetuates the belief that there is a sociological language which can describe the actions of members in some 'better' manner than they can describe these actions themselves in ordinary language. The move from the world into the text, rightly criticised by May, is mirrored by the move into a concern with sociological description that is methodologically ironic upon members' own accounts.

To put this another way, what May is doing is competing with members' descriptions by ironicising them[2]. Members' accounts are in no way remediable by sociological reflexivity - there is no need for a sociological meta-language to produce accounts since society members have done this adequately and for all practical purposes in and through their natural language (this is the crux of Sacks' (1963) paper on 'sociological description'). Indeed we might say that the realisation of sociological accounts is just as much a topic for study as the practical activities (including accounting) of scientists, pilots, traffic wardens and others. This is the focus of the recommendation from Sacks (1974) that May quotes (¶2.4).

It is good to see that he appreciates the input of ethnomethodology in the formulation of reflexivity, yet it is here that he appears to stumble, arguing that:

'A reflexive concern with the mediation between professional and lay concepts is sidelined in favour of meticulous descriptions of mundane activities within everyday life' (¶2.5)

This is problematic in that it presents a false dichotomy - what could the notion of professional and lay concepts mean to an analysis that treats these as the same? Garfinkel was attending to the use of the terms lay and professional in conventional sociological discourse. In this form there is a disjuncture of the type May suggests in that sociological accounts seek to compete with those of members. Ethnomethodology respecifies this and moves us towards a methodologically non-ironic sociology that has no interest in professional-lay dichotomies but which seeks to explicate the world of society members as it is apparent to them in and through the accounts they use. The distinction between lay and professional accounts vanishes. This is, I contend, in line with the arguments advanced by Schütz (1954) and does not constitute a break when developed by Garfinkel. Indexicality is a property of the lifeworld, a point to which Schütz may be seen to allude in his brief discussion of 'pre-meditation' and the like (p.64). However, it is not my intention to argue with May about phenomenology, as there are more important issues at stake here.

The point is that May follows the methodologically ironic path set out by those whom he seeks to criticise and does not appear to realise that he perpetuates the selfsame relations of sociological re-description and competition with members' accounts. Further, May grounds his analysis in a correspondence theory of knowledge in which he cannot but perpetuate these relations. As Graham Watson rightly points out, those who seek to be reflexive ground their claims to reflexivity in a correspondence epistemology - arguing that through the schema they advocate we can see the world aright. What such versions fail to recognise is that, in a coherence epistemology, members have already rendered the world accountable for all practical purposes, and there is no need for a sociological re-description. As Watson points out, the attempts to 'manage' reflexivity fail to recognise the fact that reflexivity is 'an essential and inevitable property of all discourse' (Watson, 1983, p. 29) and therefore not amenable to management.

Now this does not mean that we can, in the words of Ashmore, 'relax and stop worrying' (Ashmore, 1989, p. 97). Indeed, the modes of reflexivity management advanced by Ashmore together with those developed in May's work attest to the problematic and pervasive nature of reflexivity in sociology. Sociologists have sought to manage reflexivity as opposed to recognising that it is an essential feature of members' accounts. It is past time that sociology abandoned the articulations of reflexivity that are grounded in correspondence epistemology and moved on back to the world of society members.

May (par. 3.1), following Latour, calls for a return from the text to the world - this is commendable but we are forced to ask whose world is May advocating that we return to? Certainly it does not seem to be that of members since their accounts of the world are subject to reflexive remedies. This raises an important question for the future of sociology, indeed it calls into question the relevance and future of sociology per se. What should sociology have as its subject matter? From May's account it does not seem that the representations he offers are those in which members could see themselves and their practical activities in anything but the most tangential sense. I would argue that May's concern with representation is, in fact, a concern with re-presentation in sociologists' accounts that renders the life world in some form of what I want to call sociological strangeness[3].

One is reminded of Ryle's (1960, pp. 75-79) description of the discussion between the college accountant and the student wherein the accountant's ledger provides an account of the life of the college that is very different to that which the student has. When looking at the ledger's systematic and detailed description, the student is led to ask if their own description of college life is in some way deficient. Ryle argues that an assumption that the student's account is deficient is incorrect in that systematic descriptions of the accountant's ledger type leave room for accounts of the type provided by the student. As Bogen points out, an assumption that the accountant's record, by virtue of its comprehensive nature, is the record is only correct 'if we accede to an epistemic amnesia concerning the manifold nature of our own experiences and the diversity of our ordinary linguistic practices' (Bogen, 1999, pp. 33-34). While Bogen goes on to say that there is 'no essential rivalry' between the two descriptions, I would argue that in the case at hand, this is not so. To be sure, May is offering us a potentially ledger like description that does seek to compete with members' accounts. The call to reflexivity and the concomitant sociological re-description does require 'epistemic amnesia' on the part of members if it is to be recognised in the manner that May advocates. Indeed we might say that the sociology he advocates has this as a necessary condition. I should like to move on to outline some elements of this sociology and its epistemological commitments before offering an alternative model.

In his call to reconstruction May does recognise that deconstruction is not the answer - for which we should be thankful. However, he does pose an interesting dichotomy in that he suggests 'philosopher kings' remain uncrowned while maintaining a commitment to the correspondence epistemology that constitutes their realm. This leads to his central question - 'how far should we go'? From an ethnomethodological standpoint, this question makes little sense; the 'we' obviously refers to professional sociology, and of course that is not the province of ethnomethodology at all. For a would-be reflexive republican May's adherence to the ironic tenets established by monarchs is most surprising. Steering a middle course between 'belonging' and 'positioning' within 'endogenous' and 'referential' reflexivities May attends to a spectrum of 'reflexivity within actions to reflexivity upon actions' (¶3.12). My question is why not have a reflexivity that recognises accounts as a constituent feature of the circumstances they describe? We might call this a reflexivity of practical actions. Of course, this exists in the respecification of ethnomethodology (which saw the problems of formal, methodologically ironic sociology well before [a decade or more] they became modish concerns).

May's oscillation between the two poles he posits must be rather confusing since it cannot satisfy either pole and must perforce result in partial connections to a traditional sociological distinction between individual and social conditions. There is a telling adherence to alternatives - either real or imagined - within this mode of reflexivity. Things could be different, to be sure, but that they were as they are is surely at least as important? A focus on the ways that things could be otherwise elides a concern for the concerted actions that have made them as they are - these are the achievements of society members and surely sociology must have some commitment to them? Claims to knowledge cannot be vouchsafed by May's oscillator reflexivity - they are already and immediately a product of the world of society members in the natural attitude. If I might return to the comments above regarding sociological practice as a topic of study, we can say that May's reflexivity presumes more of an impact for social science than has hitherto been seen. This is not a call for interpreters over legislators, to use Bauman's (1987) terms. Rather than engaging with difference as a universal problem, we should explicate the manner in which difference, discrimination, justice and change come about as a part of the actions of society members in the natural attitude. This does not mean to imply that we should be uninterested in these processes, but that qua sociologists we should seek to explicate as opposed to change[4].

It is incumbent on those who seek to criticise to propose something better, and it is to this that I turn in the final part of this paper. We have seen that May's reflexivity is a sociological achievement, and that it is an example of what I have called (Slack, 1996) stipulative reflexivity. In stipulative reflexivities, what will count as reflexive is the achievement of the person espousing the particular variety of reflexivity. That is, reflexivity is decided upon at the outset as a sociological property which can be achieved by following sociological stipulations as opposed to a reflexivity that grounds itself in the accounts members provide as a part of the settings they describe. Reflexivity in the stipulative mode relies on the sociologist as opposed to the society member. As noted above, the stipulative mode seeks to compete with members' own accounts even while it advocates emancipation and difference. It is almost as if society members cannot be trusted to provide accounts of their own actions and that accounts must in some sense be 'repaired' by sociological intervention. I can do no better in describing the mode of reflexivity that I advocate than by quoting Garfinkel's classic description of accountability:

'In exactly the ways that a setting is organised, it consists of members' methods for making evident that settings' ways as clear, coherent, planful, consistent, chosen, knowable, uniform, reproducible connections, - i.e., rational connections. In exactly the way that persons are members to organised affairs, they are engaged in serious and practical work of detecting, demonstrating, persuading through displays in the ordinary occasions of their interactions the appearances of consistent, clear, chosen, planful arrangements. In exactly the ways in which a setting is organised, it consists of methods whereby its members are provided with accounts of the setting as countable, storyable, proverbial, comparable, picturable, representable - i.e. accountable events' (Garfinkel, 1967, p. 34. Italics in original)

I would argue that the central problematics of the reflexivity debate have been laid out in the continuing program of respecification developed by ethnomethodology, and that the concerns expressed by May and others are sociological problems not problems for members. In that this is the case, we might say that the point is to explicate members' accounts and accounting practices and not to regard them as in some way deficient or in need of remedy. Reflexivity is a recurrent trouble in sociological analysis precisely because this crucial move is not made. Ethnomethodology made such a move one third of a century ago[5], and its ongoing program (see, for example, Lynch 1993) continues to advocate and exemplify such a move - perhaps it is time that others engaged in the business of sociology took this move seriously and began to understand the implications for their own sociological practice?


1Notes 1By 'essential' I do not mean to refer to some residual essentialism or phenomenological eidos, but to the ways in which 'members accounts, of every sort, in all their logical modes, with all their uses, and for every method for their assembly are constituent features of the settings they make observable' (Garfinkel, 1967 p. 8).

2It is a continual source of fascination to me that sociological accounts of reflexivity move us further away from the natural attitude and ever closer to the text, the profession and, consequently, the ironic. A first class account of various modes of irony is to be found in Anderson and Sharrock (1983).

3This coinage has the direct opposite meaning to Garfinkel's notion of 'anthropological strangeness' (Garfinkel, 1967, p. 9).

4Obviously if one is wedded to the legislative model of sociology then this makes little sense, but one might ask how much impact sociology has had and how much a reflexive sociology of the type described by May would have.

5I am grateful to one anonymous reviewer for pointing out the implications of this to me. I would hope that other sociologists recognise that this move has been made, and above all take it into account in discussing reflexivity.


I would like to thank Dr. Andrew Carlin of University College Dublin and Dr. Rod Watson for their comments on this paper - of course the usual caveats apply.


ANDERSON, D. and SHARROCK, W. (1983) 'Irony as a Methodological Theory: A sketch of four sociological variations', Poetics Today, Vol. 4, No. 3. pp. 565-579.

ASHMORE, M. (1989) The Reflexive Thesis. Chicago: University of Chicago Press.

BAUMAN, Z. (1987) Legislators and Interpreters: on modernity, postmodernity and intellectuals. Cambridge, UK: Polity.

BOGEN, D. (1999) Order without Rules. Albany: SUNY Press.

CLIFFORD, J. and MARCUS, G. E. (Eds.) (1996) Writing Culture: the Poetics and Politics of Anthropology. Berkeley: University of California Press.

GARFINKEL, H. (1967) Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.

HERITAGE, J.. and WATSON, R (1979) 'Formulations as Conversational Objects', in Psathas, G. (Ed.) Everyday Language: Studies in Ethnomethodology. New York: Irvington.

LYNCH, M. E. (1993) Scientific Practice and Ordinary Action. Cambridge, UK: Cambridge University Press.

MAY, T. (1999) 'Reflexivity and Sociological Practice', Sociological Research Online, Vol. 4, No. 3, <>.

RYLE, G. (1960) Dilemmas. Cambridge: Cambridge University Press.

SACKS, H. (1963) 'Sociological Description', Berkeley Journal of Sociology. Vol. 8, pp. 1-16.

SACKS, H. (1974) 'On the Analysability of Stories by Children' in Turner, R. (Ed.) Ethnomethodology Harmondsworth: Penguin.

SCHÜTZ, A. (1954) 'Concept and Theory Formation in the Social Sciences', The Journal of Philosophy, Vol. LI, No. 9, pp. 257 - 273.

SLACK, R. S. (1996) Varieties of Sociological Reflexivity. Unpublished PhD Thesis: Department of Sociology, University of Manchester.

WATSON, G. (1983) 'Make Me Reflexive - But Not Yet: strategies for managing essential reflexivity in ethnographic discourse', Journal of Anthropological Research, Vol. 43, pp. 29-41.

Copyright Sociological Research Online, 2000