Copyright Sociological Research Online, 1999


Martyn Hammersley (1999) 'Sociology, What's It For? A Critique of Gouldner'
Sociological Research Online, vol. 4, no. 3, <>

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

Received: 19/7/1999      Accepted: 28/9/1999      Published: 30/9/1999


In this paper I argue against the influential 'mission statement' offered by Alvin Gouldner in his book For Sociology. This treats the discipline as supplying a reflexive perspective on social life that will lead political action towards the realisation of Enlightenment ideals. I point out that Gouldner is inconsistently reflexive, preserving his own position from the corrosive effects of the kind of sociological analysis he applies in criticising others. I contrast this with the reflexivity advocated by Steve Woolgar, deriving from the sociology of scientific knowledge. Woolgar suggests that reflexivity is always selective, and that (contrary to Gouldner) there is no sociological essence to be realised. Instead, the task for sociologists is to construct and reconstruct both sociology and its context, so as to bring off a practically successful mode of knowledge production. I argue that while Woolgar's position points to some genuine problems, it too is unsatisfactory. This is because he shares with Gouldner the assumption that a sociological perspective can be the basis for action in the world; what might be called the 'grand conception' of its role. I conclude that a more modest approach is required, similar to the position taken by Max Weber. This treats sociology as no more than a source of specialised factual knowledge about the world. Its practical value is considerable, but nevertheless limited. Above all, it cannot offer a self-sufficient answer to questions about 'what's wrong?' or 'what is to be done?'.

Constructionism; Reflexivity; Sociological Theory; Sociology's Role; Value Neutrality


Alvin Gouldner gave the title 'For Sociology' to one of his books (Gouldner 1973). However, it seems to me that before we can be for sociology, we need to know what sociology is for. And this is not a question that is straightforward to answer. Indeed, it is one about which there seems to be considerable disagreement among sociologists.

One reason for the difficulties in answering this question, I suggest, is that how we set about answering it may be implicated in the view of sociology that we have. This would not have been news to Gouldner, of course; that point was central to his notion of a reflexive sociology (Gouldner 1970). However, my argument will run in a very different direction from his.

A first point to be made is that we can interpret the question of what sociology is for in at least two different ways. On the one hand, we can ask what it is for in the sense of what purpose it actually serves in the world; in other words, we can ask about its function in sociological terms. Alternatively, we can ask about what it stands for: to what value or values is it committed? That is, what purpose should it serve? Gouldner is one of the few sociologists to have addressed both these questions in some detail. However, he does not distinguish explicitly between them and this leads to fundamental problems with his argument, or so I will claim.[1]

Gouldner is probably best known for his application of sociology to itself. Paradigmatic here is the article 'Anti-Minotaur: the myth of a value-free sociology', where he analyses the doctrine of value freedom as an occupational myth or ideology of American sociologists (Gouldner 1962). Of course, Gouldner was not the first to extend sociological analysis to the explanation of beliefs and knowledge. However, he was among the first to extend it to sociologists' own beliefs.

What is striking about his sociology of sociology from my point of view, though, is its selective application of sociological analysis. As already mentioned, he applies this to the notion of value neutrality; and in other places he extends it to what he sees as Howard Becker's commitment to researcher partisanship, to the normative functionalism of Talcott Parsons, to the work of Erving Goffman, and to both orthodox and critical forms of Marxism (Gouldner 1970, 1973, and 1980). What he does not do, however, is to apply it to the sociological approach that he himself puts forward. While he treats other approaches as serving various personal interests and social functions, he presents his own as motivated by a commitment to universal values. In other words, at this point, he switches from talking about what functions sociology serves to talking about what it 'stands for' - what it should be committed to (Gouldner 1968). In effect, he moves out of sociological explanatory mode into what I will call rational talk about the mission of sociology.

Moral gerrymandering or the full reflexivity?

This kind of selective application of sociological argument is not uncommon. Woolgar and Pawluch criticised it some years ago in the field of social constructionist analyses of social problems, labelling it as 'ontological gerrymandering' (Woolgar and Pawluch 1985). They point out how constructionists present social problems as socially constituted, rather than as reflecting actual conditions, yet in doing so themselves claim objective knowledge of those conditions. They draw a line between social constructions and objective facts in such a way that the definers of social problems are presented as addressing illusions whereas the researcher deals with reality.

Moreover, while sociologists apply constructionism to some social problems, implying that these are the invention of moral entrepreneurs, other problems are treated as genuine. In the case of these latter problems, constructionism only enters to explain why it took so long for a problem to be recognised and to be put on the political agenda. For example, there are few radical constructionist accounts of the problems of rape or child sexual abuse. Their genuine status as social problems is usually taken for granted; and, we might say, rightly so.[2] Yet this selective application of constructionism raises important questions about its validity; signalled by the evaluative character of the term 'ontological gerrymandering'.

It is also worth pointing out that what is involved here is as much moral as ontological gerrymandering. A line is being drawn separating what are legitimate and illegitimate problem claims. This is the direct parallel with Gouldner's sociology of sociology. As I noted, when he discusses views with which he disagrees he employs concepts like 'myth' and 'ideology' to point to the ulterior motives and unsavoury social functions that they allegedly serve; yet, when he discusses his own views, he presents them as embodying universal human values, and therefore as not in need of sociological explanation (see Hammersley 1999:ch4). Moral character work is being done through this selective treatment. But even more important, for my purposes, is the underlying assumption on which this selectivity relies. In effect, Gouldner treats sociology as only appropriate for explaining what is morally or politically unacceptable - taking it to be unnecessary when dealing with what he believes to be sound in moral or political terms. We might say that his analysis has a Kantian structure. Thus, he presents the advocates of value freedom as caught in the realm of necessity, so that their behaviour is open to explanation in causal terms: on the basis of interests and social functions. By contrast Gouldner presents himself as operating in the realm of freedom, acting in accordance with universal principles.

So, moral gerrymandering involves moving the line between the realms of necessity and freedom in such a way as to put oneself and those one agrees with on one side of the line, and those with whom one disagrees on the other side of it. Underlying this is a conception of humanity as lying between the animals and the gods. Like animals, human beings are in the realm of necessity; but, unlike animals, through commitment to universal values and great effort they can raise themselves into the realm of freedom. And, in terms of this framework, those whose behaviour remains caused can be blamed for lack of commitment or effort. This is the reason why Gouldner's analyses of other sociological positions are simultaneously causal and negatively evaluative in moral terms.

It is important to note that Gouldner does not supply any supporting argument for why the line between those whose views are caused and those whose views are a product of freedom should be drawn in the place where he draws it. For instance, he provides no evidence that American advocates of value freedom were in fact motivated in the way he claims, or that their advocacy of this principle had the effects he cites. Nor does he show that his own commitment to 'objective partisanship' is free from the influence of interests and does not serve an undesirable social function.[3]

A more fundamental point, though, is that it is possible to produce speculative (and perhaps even accurate) explanations for any set of beliefs in terms of interests or social functions, including those of Gouldner himself.[4] Given this, the obvious conclusion to draw might be that sociological analysis ought to be applied comprehensively across the board: everyone should be treated as operating in the realm of necessity, including the analyst him or herself. In fact, it could be argued that this is the logical implication of Gouldner's reflexive application of sociology to itself; in other words, that he simply failed to be reflexive enough.

A social constructionist version of this comprehensive reflexivity seems to be the solution advocated by Woolgar and Pawluch, in the article mentioned earlier. What is involved here is not simply a temporary suspension of cognitive and moral realism, but an anti-realism or anti-essentialism. These authors outline three ways of reading their argument about ontological gerrymandering. First, as providing instructions for how to do it effectively. Secondly, as a critique instructing social problems researchers to be more careful in their work so as to avoid ontological gerrymandering. Or, thirdly, as indicating that such gerrymandering is unavoidable; that, to use my terms rather than theirs, it is a fact of life. To the extent that they recommend a solution to the problem, Woolgar and Pawluch follow the third of these options, suggesting that more research must focus on the rhetorical devices used by sociologists to establish their claims about the world.

This is very much in line with Woolgar's work in the sociology of scientific knowledge. Central to this sub-discipline has been its treatment of science as a thoroughly social activity, and its rejection of the idea that scientific conclusions are to be explained as reflecting reality. Rather, the findings of science are shown to be constituted through the activities of scientists. Woolgar interprets this argument in a radically constructionist fashion, and has argued that sociologists should set out to undermine the hegemonic moral order underlying the realist discourse of science (Woolgar 1988a). Moreover, he has extended this argument to the sociology of scientific knowledge itself, challenging sociologists' own referential claims. Like Gouldner, he requires sociologists to be reflexive; though for him what this means is that they should be explicit about the ways in which their accounts rhetorically constitute the social world. Thus, Woolgar has argued in favour of forms of sociological writing that subvert their own representational claims (Woolgar 1988b).

Where Gouldner identifies the essence of sociology with a commitment to universal values (identified as truth, freedom, unity, peace and justice - see Gouldner 1973:ch2), Woolgar denies that it has any essence at all; not even - or perhaps, more accurately, especially not - commitment to truth as a transcendent value. A recent illustration of this is that, in response to the problem of funding sociological research, he has argued that sociologists should be in the business of reconstructing their work so as to configure a population of users, thereby attracting the funds from private companies and government agencies that it needs (Woolgar 1999). Thus, rather than seeing the task of sociology as being to meet the needs of potential users, he argues that what is required is a proactive approach in which sociologists play a direct role in constructing those 'needs'. In these terms, neither sociology nor the needs it can serve are simply given: both are open to construction and reconstruction. The analogy here is with technology, which configures its users, and with commercial firms, who seek to create their own markets through advertising and 'public relations'.

In the terms I introduced at the beginning of this paper, according to Woolgar, we can ask about what sociology is for in the sense of how it functions in the world, and of how it could function in the world; but there is no point in asking what it stands for. For him, to ask that question is to attribute to sociology some sort of moral essence; and this is illegitimate. This is the point at which the contrast between his position and that of Gouldner is at its sharpest. So, Woolgar's answer to the question of what sociology is for would be that it is for whatever we can and want to make it for.

I suspect that some sociologists would be inclined to respond to this more or less in the terms of Gouldner's jibe about the difference between living off and living for sociology (see Gouldner 1970:15). Woolgar's position seems to leave us with no option but to live off the discipline; though I am sure that he would claim to have undercut the very distinction between 'living off' and 'living for'. In my judgement, though, the problem is more deep-rooted: it is that his argument leaves us with no basis for action at all. It is not possible to live by doing whatever we find that we can do, since there are many things that we could do that are incompatible. There are also some things which, if not done, result in the extinction of life.[5] Furthermore, in Woolgar's terms, any appeal to what we want to do is surely open to reflexive deconstruction, in exactly the same way as are the 'needs' of sociology's 'users'.

What is wrong with Woolgar's position, I suggest, is that it is the application of a third person perspective, which may be suitable for reflection on the world, to a situation where the sociologist must think of him or herself as an actor in the world. And, for action, only a first person perspective will do; even though we can employ knowledge derived from third person perspectives - about others and even about ourselves - in formulating our plans.

Moreover, this point extends beyond constructionism. The causal-functional form of analysis that Gouldner deploys to explain other views is also a third person perspective, and is similarly disabling when applied to situations demanding action. Thus, in determining what sociology is for, we cannot avoid relying on some kind of rational or intentional model of ourselves, because deliberation and decision are required. This is why Gouldner lapses into moral gerrymandering: if he were fully reflexive he could not have put forward an alternative to the approaches he criticised.

Woolgar and Pawluch propose that ontological gerrymandering may be unavoidable. And, indeed, Woolgar engages in it himself in his critique of essentialism in science and sociology. That critique relies on a claim that these enterprises have no essence; which is just as much a philosophical (as opposed to a sociological) claim as the argument that they do have one. Woolgar is faced with two options. He could try to justify his acount of the sociality of science as capturing reality, or he could treat it simply as one construction among many, and as no more true than any of its competitors (in terms beyond its own). Indeed, even the idea that his account is superior because it is more reflexive assumes that constructionism is true, since only then would reflexivity of the kind he recommends add value to an account.

There is a genuine and difficult problem here, it seems to me. On the one hand, Woolgar's anti-essentialism is not sustainable: it gives us no basis for action in the world. However, he is surely correct that there is a contradiction in the way that sociological analysis is often deployed selectively. This is a contradiction which comes home to roost when Gouldner applies sociology to itself; there, it generates special pleading for his own notion of what sociology stands for.

The grand conception of Sociology's Role

In my view, the source of the problem lies in an assumption which both Gouldner and Woolgar share. This is the idea that sociology supplies a perspective which is comprehensive and self-sufficient, in the sense of providing the appropriate basis for action in the world. I will refer to this as the grand conception of sociology's role.

An early source of this can be found in Auguste Comte's initial formulation of the sociological enterprise. He saw historical development as involving the replacement of religious and metaphysical ideas by scientific ones. Sociology was the last of the sciences to be completed, and for Comte it provided the framework through which all the others were to be interpreted in order to organise the new social order. A similar commitment to the idea of social science providing a comprehensive and self-sufficient basis on which social and political practice can and should be based is to be found in Marx; and this has been inherited by many versions of critical research. Moreover, aside from the influence of these sources, the late arrival of sociology as a discipline - and the resulting competition with politics, economics, and psychology for space in the academy and in the public sphere - has resulted in a tendency towards imperialism: to rejection of the ways of explaining human behaviour provided by other disciplines (see Strong 1979). In the case of sociology, this has been encouraged by, and has also shaped, its association with egalitarian and other political views on the Left.[6]

Gouldner's work is very obviously within this grand tradition. For him, sociology plays much the same role as it did for Comte, as Reason did for some of the philosophes, and as materialist philosophy and political economy did for Marx. The central message of The Coming Crisis of Western Sociology is that sociologists must not separate their work from the rest of their lives; the implication is that they must 'live for' sociology, and equally important that sociology can provide the basis for right living. Thus, Gouldner proposes the establishment of communities of sociologists who develop theory which they can use reflexively to reorganise themselves, thereby in turn creating the emancipatory conditions that will facilitate further theoretical breakthroughs (Gouldner 1973:ch4). What is envisaged here is a continuing dialectic between sociological theory and practice. Equally important, he sees this as forming part of a wider political process in which the structure of the whole society is transformed, facilitated by a strong relationship between radical sociologists and political activists; albeit with some autonomy preserved on both sides.

By contrast, Woolgar is not obviously part of this grand tradition. He detaches himself from most of the sociological domain assumptions on which Gouldner relies. Nonetheless, it is clear that he sees constructionist sociology not simply as a useful tool to be adopted for analytic purposes, but as providing a basis for how sociologists should operate in the world. Reflecting this, there is a striking parallel between his analysis of science and of older forms of the sociology of science, and Gouldner's critique of value neutrality. Both employ the concepts of 'ideology' and 'myth' (Woolgar 1988a:pp14, 48, 84, 101; Gouldner 1962). That Gouldner does so is not at all surprising: it is compatible with the functionalism and Marxism on which he draws. However, use of these concepts is very surprising, and puzzling, in Woolgar's case. After all, with what could 'ideology' and 'myth' contrast in his account? But the most important point for my argument here is that, like Gouldner, Woolgar applies sociology reflexively in order to show sociologists what they ought to (or at least can) do; hence his recommendation that they should subvert 'the ideology of representation' and configure their own audience. Moreover, underlying his argument here seems to be a radically pragmatist notion of truth, and indeed of virtue, whereby what is true or virtuous is simply what comes off as true or virtuous in the course of social interaction. On this basis, Woolgar would probably deny the very distinction between theory and practice (see Woolgar 1986).

So, despite the sharp differences between them, both Gouldner and Woolgar adopt the grand conception of sociology's role. Against this, I will suggest that the idea that sociology can supply a comprehensive and self-sufficient perspective on the world is false, whether interpreted in Gouldnerian or Woolgarian terms.

Central to the grand conception is the idea that it can replace the basis on which people currently operate within the world. This is sometimes formulated as 'scientific sociology replacing folk sociology'. While this terminology is no longer common, and certainly would not be accepted by Woolgar, it is useful because it highlights the assumption, central to the grand conception, that social action routinely relies on a knowledge base which is similar in character to the disciplinary knowledge of sociology, albeit one that is inferior to it.[7]

A common response to the grand conception these days is to deny the superiority of scientific sociology and to insist on the value and validity of lay perspectives. But it is worth pointing out that this position itself retains the assumption that those lay perspectives are analogous in character to sociology. My argument, by contrast, is that human action is not simply a product of knowledge, and especially not of the kind of propositional knowledge that sociology supplies. Rather, social action involves contexted processes of interpretation whereby we make sense of what is going on, in a way that is designed to serve our practical purposes and concerns. And these processes of interpretation rely on diverse forms of knowledge, some deriving from experience, as well as on judgements of various kinds; rather than being the 'application' of a body of general knowledge or even of a method. Furthermore, practice involves evaluations of what is good or bad, better or worse, acceptable or unacceptable, worth pursuing or not worth pursuing etc. And these value judgements are not derivable solely from sociological analysis.

Indeed, they are not derivable even from sociological analysis of the kind that Gouldner recommends, where universal values other than truth are already built in. For example, while most people are probably committed to social justice, they often disagree considerably about what is and is not just. And this disagreement does not stem entirely from different interests or from discrepant views about the facts of the case concerned. Disagreement also arises from the differential priority people give to various possible interpretations of the principles of justice, both in general and in the particular case being considered. Take the example of selective secondary schooling. On some interpretations of both retributive justice (the concern with just deserts) and of distributive justice (concerned with equitable shares) selective entry of pupils into schools on the basis of ability is justifiable. On other interpretations of distributive justice (notably those which emphasise parity) it would not be.[8]. And while there are all sorts of reasons and causes why people adopt one interpretation of a value rather than another on particular occasions, all of us probably use all of these interpretations on some occasions. The result of this is that there is scope for reasonable disagreement about their application to particular cases. Given this, while social scientists may legitimately be able to claim that their work produces propositional knowledge about the social world that is more likely to be sound than accounts from other sources, it is unclear why their judgements about the priority among and interpretation of different principles of social justice (or of any other value) should be treated as superior to anybody else's.[9]


For these reasons, it seems to me that we must adopt a more modest view of the social role that sociology can play in the world than what I have called the grand conception. Sociology cannot tell us what is wrong and what should be done. And this conclusion carries two relevant implications for my topic. The first of these is that sociology alone cannot tell us what it is itself for, in the sense of what it should stand for or aim at. In other words, even as sociologists we cannot live by sociology alone. We cannot do sociological work as a practical activity by relying exclusively on sociological knowledge. We have to draw on other resources as well. Thus, reflexivity, in the sense of the application of sociology to itself, cannot provide the basis for specifying the mission or the method of sociology.[10]

The second implication of my argument, however, is that Gouldner's attempt to commit sociology to the practical realisation of a comprehensive set of universal values was a mistake. As noted earlier, he sees sociology as committed not just to truth but also to justice, peace, and human unity. Now, one does not have to be a postmodernist to find this totalising perspective unacceptable. Arguments against it have long been available from within liberalism, for example in the work of Isaiah Berlin, who stresses the plurality of values and the conflicts among them (see Gray 1995). Once that plurality is recognised, it becomes clear that there will be occasions when one value will have to be prioritised over others. In the context of Gouldner's sociology, this may mean that political considerations arising from the pursuit of unity must override the concern with truth. More insidiously, such prioritising of other values over truth may happen without the sociologist even being aware of it. In short, Gouldner's position opens the way for bias (Hammersley and Gomm 1997).

Ironically, against this background, it seems that the most cogent answer to the question of what sociology is for is precisely the one which is implicit in what Gouldner dismissed as an occupational ideology: Weber's doctrine of value neutrality. According to this, what sociology stands for is truth, and its proper function is to produce value-relevant knowledge about the social world. This view is often regarded as a product of positivism. Yet, in fact, it derives from post-Enlightenment critiques of Hegelianism (and by implication of nineteenth century positivism) for attempting to read a harmonious set of values into History.[11] In this respect, Gouldner is closer to positivism than is Weber; indeed, his attachment to the model of Saint-Simon is explicit (Gouldner 1973:69).

For me, the conclusions to be drawn from Gouldner's reflexive analysis of value freedom as an occupational ideology are two-fold. First, he is wrong to use what he claims to be sociological knowledge about the functions of this doctrine as an evaluative standard by which to judge it. This involves moral gerrymandering. The second conclusion I draw is that what his argument suggests is not the fallacious character of the principle of value freedom but the failure of many American sociologists in the 1950s to live up to it. Like Gouldner, they often treated sociological method and knowledge as a standard by which to judge the rationality of the people being studied. And it must be said that this tendency remains widespread; as evidenced by continued use of concepts like 'myth' and 'ideology' in much the same way that Gouldner uses them.

In some ways, the principle of value neutrality is analogous to an influential view in the sociology of scientific knowledge that preceded Woolgar's: what was referred to as the strong programme (Barnes 1974; Bloor 1976). The central plank of that programme was that we should set about the task of explaining why people hold particular beliefs, and how these function socially, in a manner that is indifferent to whether we take those beliefs to be true or false. This was dubbed the symmetry principle. And, in parallel terms, it seems to me that as sociologists we ought to adopt the principle of moral symmetry. In other words, in sociological analysis we should focus entirely on describing what happens and explaining why, and do this in a way that takes no account of whether we believe what we are studying is good or bad. In other words, we should try to avoid moral gerrymandering.[12] Moreover, the only way to do this is to abandon the grand conception of sociology's role, and to adopt a more realistic and modest position.

Perhaps I need to underline that this does not mean that as individuals we should become indifferent to matters of right and wrong, in ethical or political terms. It is simply that in doing sociological analysis we should suspend our commitments to moral and political goals other than truth, in order to maximise our chances of achieving a sound factual and theoretical understanding of the world. In my view, producing that kind of understanding is precisely, and entirely, what sociology is for.


1The failure to distinguish between these two issues is endemic in the main traditions on which Gouldner drew: functionalism and Marxism.

2Nelson (1984) provides a constructionist account of child abuse as a social problem; I am indebted to David Silverman for this reference. It is striking, though, that she nowhere suggests that child abuse is merely a construction, but focuses rather on how it came to be placed on the public policy agenda. Closer to a radical constructionist position is Davies' work on the discursive construction of childhood sexual abuse; but she explicitly brackets the question of its existence in order to focus on the way it is formulated in Sylvia Fraser's autobiographical account (see Davies 1995).

3There are several places where Gouldner implies that sociological analysis should be applied reflexively in a fully consistent manner. One of these is at the start of his article 'The sociologist as partisan', where he writes: 'Sociology begins by disenchanting the world, and it proceeds by disenchanting itself. Having insisted upon the non-rationality of those whom it studies, sociology comes, at length, to confess its own captivity' (Gouldner 1973:27). However, he does not follow through the implications of this.

In the final chapter of The Coming Crisis, Gouldner addresses the question of whether his own views should themselves be subjected to the kind of sociological critique he has applied to the work of others (Gouldner 1970:481-2). However, he argues that he is not the person to do this, and goes on to point out that the fact that his views may be a product of his biography does not invalidate them. This is true enough, but his criticism of others' work does rely on the assumption that validity is implicated in social origin and functioning (see Hammersley 1999:ch4). Indeed, this seems to be central to his whole notion of reflexivity.

The final section of the final chapter of The Coming Crisis is entitled 'Reflexive sociology looks at itself', but the concern is with why this approach emerged at the time that it did, not with how its content is socially constituted. The explanation given is that 'as [sociologists'] most immediate work environment - the universities themselves - become drawn into the coalescing military-industrial-welfare complex, it becomes unblinkingly evident that sociology has become dangerously dependent upon the very world it has pledged to study objectively' (Gouldner 1970:511). Thus, reflexive sociology is presented as a response to the reality it documents, rather than as serving some latent social function. This is analogous to constructionists' treatment of what they take to be 'genuine' social problems.

4An interesting example of this kind of table-turning occurred in the debate over postmodernist anthropology: see, for example, Roth (1989).

5As this indicates, I see no problem with appeals to 'death and furniture' in defence of realism, contrary to Edwards et al (1995). These have the same status, it seems to me, as Wittgenstein's appeal to those things which there is no point in questioning: see Wittgenstein (1969). For a defence of realism, see Hammersley (1998).

6The grounds for this association are spurious, in my view, since the two main forms of sociological explanation - appealing to aspects of actors' social backgrounds and to features of the situations they face - inevitably involve causal analysis and therefore (at best) set limits to freedom and equality. Furthermore, denial both of the existence of a universal human nature and of normativity - which is central to sociological imperialism - undercuts the basis for critique. We might describe this as the unintended lesson of poststructuralism: see Dews (1987, 1995).

7Thus, in Woolgar's case the realist folk methodology of natural scientists is taken to be similar in structure to the knowledge produced by the sociology of science, and can therefore be replaced by it. In this respect, while Woolgar draws on ethnomethodology, he has abandoned the ethnomethodological principle of indifference. On the relationship between constructionism and ethnomethodology, see Button and Sharrock (1993). See also Button (1991).

8On different interpretations of educational inequality, see Hammersley (1997).

9Some sociologists today even reject the idea that sociology can supply a superior perspective in terms of propositional knowledge about facts. Yet it is difficult to see how the discipline could survive without relying on this claim to superiority. If it does not, on average, provide superior knowledge in some respect - however fallible and qualified - what could be its justification? Even those who argue that the task of sociology is to give voice to the marginalised presumably rely on sociological analysis to identify the victims of marginalisation; as evidenced by the fact that they are selective in whom they seek to give voice to. Going back to an earlier example, few sociologists would wish to give voice to those perpetrating child sexual abuse; and yet there is a clear sense in which such people are socially marginalised. Presumably, this selectivity is based on theoretical analysis; as well as being open to the charge of moral gerrymandering.

10See also Holmwood's (1999) argument for the need for a lack of reflexivity in doing sociological work.

11For detailed discussion of Weber's conception of value neutrality, see Bruun (1972); Scott (1995). See also Shapiro (1978) and Eden (1987), who document the influence on him of Nietzsche's critique of Enlightenment thinking. Elsewhere, I have provided an account of the implications of this principle for how the research communities must be organised: see Hammersley (1995) and Foster et al (1996).

12We should also try to avoid ontological gerrymandering. However, contrary to Woolgar, this does not require us to abandon all claims to representational knowledge, simply to avoid treating the validity or invalidity of beliefs as relevant to their description and explanation.


BARNES, B. 1974. Scientific Knowledge and Sociological Theory. London: Routledge.

BLOOR, D. 1976. Knowledge and Social Imagery. London: Routledge.

BRUUN, H.H. 1972. Science, Values and Politics in Max Weber's Methodology. Copenhagen: Munksgaard.

BUTTON, G. (ed) 1991. Ethnomethodology and the Human Sciences. Cambridge: Cambridge University Press.

BUTTON, G. and SHARROCK, W.W. 1993. 'A disagreement over agreement and consensus in constructionist sociology'. Journal for a Theory of Social Behaviour. 23, 1: pp1-25.

DAVIES, M. 1995. Childhood Sexual Abuse and the Construction of Identity: Healing Sylvia. London: Taylor and Francis.

DEWS, P. 1987. Logics of Disintegration. London: Verso.

DEWS, P. 1995. Introduction to The Limits of Disenchantment. London: Verso.

EDEN, R. 1987. 'Weber and Nietzsche: questioning the liberation of social science from historicism' in Mommsen, W.J and Osterhammel, J. (eds.) Max Weber and his Contemporaries. London: Allen and Unwin.

EDWARDS, D., ASHMORE, M., and POTTER, J. 1995. 'Death and furniture: the rhetoric, politics and theology of bottom line arguments against relativism'. History of the Human Sciences. 8, 2: pp25-49.

FOSTER, P., GOMM, R. and HAMMERSLEY, M. 1996. Constructing Educational Inequality. London: Falmer.

GOULDNER, A.W. 1962. 'Anti-Minotaur: the myth of a value-free sociology'. Social Problems. 9: pp199-213.

GOULDNER, A.W. 1968. 'The sociologist as partisan: sociology and the welfare state'. The American Sociologist 3: pp103-116.

GOULDNER, A.W. 1970. The Coming Crisis of Western Sociology. New York: Basic Books.

GOULDNER, A.W. 1973. For Sociology: renewal and critique in sociology today. Harmondsworth: Penguin.

GOULDNER, A.W. 1980. The Two Marxisms. London: Macmillan.

GRAY, J. 1995. Berlin. London: Fontana.

HAMMERSLEY, M. 1995. The Politics of Social Research. London: Sage.

HAMMERSLEY, M. 1997. 'Educational inequalities' in Block 5, U. (ed.) EU208 Exploring Educational Issues. Milton Keynes: The Open University.

HAMMERSLEY, M. 1998. 'Get real! A defence of realism', in P. Hodkinson (ed) The Nature of Educational Research: Realism, relativism, and postmodernism. Manchester: Manchester Metropolitan University.

HAMMERSLEY, M. 1999. Taking Sides in Research: essays on bias and partisanship. London: Routledge.

HAMMERSLEY, M. and GOMM, R. 1997. 'Bias in social research', Sociological Research Online, 2:1 <>

HOLMWOOD, J. 1999. 'Sociology and its audience(s): sociological argument, "validity" and "values"', paper presented at the annual conference of the British Sociological Association, 6-9th April, University of Glasgow.

NELSON, B. 1984. Making an Issue of Child Abuse: political agenda setting for social problems. Chicago: University of Chicago Press.

ROTH P. A. (1989) 'Ethnography without tears', Current Anthropology, 30, 5, pp555-69.

SCOTT, A. 1995. 'Value freedom and intellectual autonomy'. History of the Human Sciences 8: pp69-88.

SHAPIRO, W. 1978. The Nietzschean Roots of Max Weber's Social Science, unpub PhD thesis, Cornell University.

STRONG, P.M. 1979. 'Sociological imperialism and the profession of medicine'. Social Science and Medicine 13A: pp199-215.

WITTGENSTEIN, L. 1969. On Certainty. Oxford: Blackwell.

WOOLGAR, S. 1986. 'On the alleged distinction between discourse and practice'. Social Studies of Science 16, 2: pp309-17.

WOOLGAR, S. 1988a. Science: the very idea. London: Routledge.

WOOLGAR, S. (ed.) 1988b. Knowledge and Reflexivity: new frontiers in the sociology of knowledge. London: Sage.

WOOLGAR, S. 1999. 'The social basis of interactice social science' Paper given at:Interactive Social Science: a conference on the production and consumption of social science research. ALSIS, Brighton, England.

WOOLGAR, S. and PAWLUCH, D. 1985. 'Ontological gerrymandering: the anatomy of social problems explanations'. Social Problems 32: pp314-27.

Copyright Sociological Research Online, 1999