Notes on: Hammersley, M. (1992). On Feminist Methodology. Sociology 26 (2): 187 – 206. DOI: 10.1177/0038038592026002002 Dave Harris Feminists claim that they have a distinctive approach, a commitment to qualitative, for example, or particular 'methodological and epistemological assumptions'(187). There are different positions, but four underlying themes: 1. Gender is seen to be central to human social relations, and a key reason for differences in power. It's crucial and must be included in any analysis. Conventional social science is primarily about the experience of men 'presented as if it were human experience'. Women's experience must be studied, conventional social science must be assessed for male bias. 2. Personal experience is emphasised instead of scientific method, sometimes with empiricist undertones. This can lead to the argument that personal experience cannot be invalidated, that women have access to truths and reality that are not available to men. They are oppressed and know about this at a bodily level, while men lack empathy and therefore sociological imagination. Sometimes, women have double consciousness, knowing both the dominant culture and their own deviant perceptions. Sometimes this leads to the argument that only women can do feminist research, to unstructured qualitative data and analysis, the denial of objective rules and objectivity in research which is seen as a masculine perspective: objectivity itself can be '"an excuse for a power relationship"' [citing Stanley and Wise] (189). 3. There should be no hierarchies between researcher and researched, researchers should share personal experiences and authentic relations. Sometimes this means equal participation of people studied, so that participants can see what is significant to them. Sometimes it needs to an interest in conscientisation. The researcher is also relevant in research, and has a duty to explain underlying reasoning procedures and their own experience. Hierarchy is rejected on ethical grounds, with non-hierarchy particularly relevant to women; on methodological grounds, only authentic relations will deliver the truth; a practical interest in consciousness-raising and emancipation where the researched must necessarily be involved. 4. Research is aimed at emancipation rather than the production of valid knowledge. Truth can only be discovered in struggle against oppression and should be related to how it achieves emancipation. This is conscientisation. The opposite also follows — that mere description is conservative, and that struggle itself is necessary to produce adequate social science research. There are clearly some underlying ideas about method: 1. Gender had been given relatively little attention, not much research had focused on gender differences and the experiences of women, and the importance of gender is an extraneous variable had been 'underplayed' (191), and research from boys and men generalised instead. Gender difference on the research process itself had been ignored. All these are valuable, but is gender of unique importance? Other variables have also been neglected — some black feminists have mentioned race, some poststructuralist feminists have talked about a 'wider plurality of differences'. Establishing 'essentializing definitions of womanhood' has led to problems, and gender has been unduly prioritised – 'while gender is very important, in my view it should not be given any preestablished priority over other variables' (192) [you're dead mate]. 2. Claiming the superiority of direct experience over method is fairly common in qualitative methodology, but this privileges experience as a form of 'direct access to the truth' (192). We can never get outside our own physiology and culture. Experience is important, but it cannot replace method. It risks 'false cultural assumptions being embedded in the data', it still contains error and may prevent 'efficient data collection'. Instead, experience should be corrected by method, and vice versa. Few philosophers stress exclusive reliance on method or procedure, and few would reject method completely. We need to retain the idea of making the means of doing research public, and 'therefore open to collective assessment and improvement', explicit, and more instrumental about choosing appropriate strategies. We should apply methods 'as a trial, in a tentative manner', and we may have to change them. Feminist challenges might help counter rigid methodological ideas, but they also encourage an idea that experience is 'beyond question' [193). We need instead 'communal questioning of assumptions… Those judged to be open to reasonable doubt', we want to involve the research community in suspending assumptions that are not shared, and attempting to establish validity within shared assumptions. Feminists can suggest that direct experience prevails over method and gives uniquely valid insights — 'feminist standpoint epistemology'. Other groups also claim privileged insights, including Marxists, but also Nazi scientists. There can be diversity arising from different social locations, and this can be fruitful, but 'we must beware of claims that one group or category of people necessarily has more valid insights than another'. Experiences are construction, with the capacity for error as well as truth. All information is processed by factors including 'cultural conditioning, social context'. Perhaps women have unique insight into the motives of their oppressors — but the reverse could also be argued for effective oppression. Women can appeal to both dominant culture and also their own direct experience, as can some members of both dominant and subordinate cultures — 'there is probably some truth in it' (194), but so can other categories of persons experiencing more than one culture. This can even be seen as 'the norm in large, complex societies'. There is a parallel in the Marxist view that the proletariat have uniquely valid insights, but this usually necessitated separating a true vision from actual beliefs: the latter often diverse and inconsistent. Same problems arise with the category of woman, and there is an implicit elitism here too — that insight belongs to 'feminists involved in the struggle'. Of course feminist themselves displayed diverse and inconsistent views, leading to a problem about which ones we should select. Stanley and Wise suggest that all experiences and views are valid in themselves, but this involves relativism and 'provides no basis for the rational resolution of disagreements'. It also minimises the real conditions of women's oppression independently of their experiences. This diversity can actually be an advantage in helping us minimise bias, as Merton once said — using both insider and outsider perspectives will enlarge our chance of understanding. This is preferable to any attempt to privilege some category or declare all points of view to be equally valid. One consequence on the emphasis on experiences is an undue emphasis on unstructured methods — these are valuable but may not be appropriate 'to all purposes' (195) as many feminists recognise. 3. Ethnographers have often discussed the variations in the role of participant and observer. There are both observer effects and ethical concerns. The consensus these days seems to be that no one should try to act as a neutral or uninvolved observer, even if they are not actually committed to a particular group. It is common to argue that the researchers' own values should be explicit [and the authority for this goes back to Weber]. Collaborative research has emerged in a number of other fields, including educational research [Stenhouse]. So these are not distinctive feminist ideas. There are ethical and political issues raised by feminists, but problems with them. For example, there are 'ethical, methodological and tactical considerations' which lead to an argument against hierarchy in the research relationship. It is not clear if the elimination of hierarchy just involves research on women. It would in any case be difficult to study women alone because their lives are 'so closely interrelated' with those of men. If we focused on unique women's experiences, we might have more trouble researching 'the world that produced these experiences' (196). If men are included as sources of data, they may insist on introducing their own hierarchy, requiring special control by feminist researchers — such as choosing 'between being dominant or being dominated'(196) [so when researching men, hierarchical research techniques might be needed]. Women in powerful positions might raise the same dilemma, and researching them means 'feminists may also be forced into hierarchical relationships, one way or the other'. How desirable are non-hierarchical relationships and equality of control over the research process? Feminist discussions often use a one-dimensional notion of hierarchy, implying all-pervasive control usually exercised by a group or person at the expense of others. Such 'power is rarely all-encompassing' in pluralist societies, so it is unlikely that researchers will dominate the lives of their subjects. They do have limited control over the project, defining the topic, deciding how data is to be connected, how the analysis is to be done and so on, but this is not even 'a claim to total control over the research process itself', because they can be denied access to places or information. Research is usually marginal to the research subjects. It is not the same as controlling employers, clients or students [which are probably worse]. Can all inequalities be rejected anyway? Can there be no hierarchy as in representative democracy? Is democracy always the best way to make decisions? Even if researchers do dominate the research process, this still might not be against the interests of others. Having a say is not always an overriding issue. This is to deny the expertise of the researcher and reduce her opinion to that of any other woman. This has led to accusations that publication involves domination. Research does involve a claim to intellectual authority, which is circumscribed and based on the idea that the findings of research are 'on average, less likely to be in error than information from other sources' (197), and this is so because there is a whole research community which can scrutinise research. Everyone is entitled to their own opinions, but they need not be treated as equally likely to be true. We do not even do this in everyday life. Of course it is not only research findings that are superior — we also trust information based on first-hand experience. It is rational nevertheless to give them more attention than mere opinions. Some feminists would reject this claim to intellectual authority because it conflicts with a commitment to equality, the right for individual groups to define reality for themselves. But this still assumes a link between understanding a situation and then being able to change it by activities such as redefinitions. Successful action equally depends on accurate information and therefore expertise. It would be 'an anti-realism' to deny such expertise. We see the problems with Stanley and Wise who pursue the anti-hierarchical argument to a logical 'and contradictory conclusion' [they don't want to be seen to be telling other feminists anything should be, because there is no one feminist research; the personal is the main principle for research anyway — '"this may be a contradiction at the heart of what we believe "' (198), but their book necessarily makes a claim to valid knowledge]. Researchers always do if they publish their findings, and this makes them obliged to check on the reliability of information themselves. There is a methodological case against hierarchy in that hierarchy will distort the context and thus the data. This assumes that you get data only from unconstrained accounts by informants. Informants have more interests and commitments than those of the research team, however. It also assumes that the only data is provided in the accounts of informants — but what of the wider social forces that influence behaviour, what of the threats to reliable observation, do people always know their own motives? Informers' accounts should always be examined. Much data is observational anyway. The researchers are necessarily more active, collating information, and there are always threats to validity. There is no 'single ideal authentic situation in which data are necessarily error free' (199). The final argument is that if the goal of research is to bring about female emancipation, then women need to be involved in it. This assumes that research should aim at female emancipation. Overall, the specifics of the research investigation need to be investigated. There is no 'authentic related (personhood)' as in Stacey, and the rest of life shows all the usual dilemmas and inequalities as research does, so it is naive to insist that that one relationship is nonhierarchical. It is argued that scientific enquiry should attempt to change the world rather than just describe it, in the classic work of Marx, and others argue that truth should be judged in terms of practical efficacy, including US pragmatists. Again this claim is not unique to feminism, and it has a number of problems. Changing the world should not be the immediate goal of research, nor should it be judged against that goal. Emancipation seems to be too simplistic, implying a neat division into oppressors and oppressed, with no proper analytical content for those concepts anyway. Oppressions of different kinds cannot be reduced to sexual oppression, so many people will be both oppressors and oppressed — we know for example that black feminists criticise white feminists for neglecting racism, and even black women in Western societies might be regarded as oppressors on an international scale. For that matter, oppression itself is problematic, assuming some state where there is never a frustration of preferences, where only real needs are identified. This will depend usually on a '"naturalist conception of the human agent"' (200) who can identify all their needs, or that experts can identify needs. Human agency is too fluid. It is not easy to identify needs, and we tend to reconstruct them on the basis of their own feelings and our interpretations of what others say. There may be inconsistency in disagreement. Interests or needs sascribed to a group might not match their actual beliefs and this contradiction can be a problem. There is a particular problem with distinguishing genuine desires and those which are actually against the group's interests. Finally, there might be several alternative reconstructions of what the needs are and much genuine disagreement. If oppression is difficult, so is emancipation. If oppression is not simple and reducible to a single dimension, nor can emancipation be. Marxists claim that their view will be validated by history which will finally reconcile different views in the triumph of the proletariat. Just as his views were limited, so will be the argument that women's emancipation will abolish all the other forms of oppression — 'indeed, it is possible that it could worsen them' (201). Without emancipation, however, 'negative critique can be little more than negativism for the sake of it'. There are inequalities between sexes and feminist politics need not wait until these issues have been resolved, but researchers do have a responsibility to attempt conceptualisation. One thing needs to be reformed is the instrumental view of the relationship between theory and practice: practice is complex and it cannot just be driven by research findings. There is a pragmatic criterion for truth in the argument for emancipation, and practice again suggests greater complexity. Successful action can be based on false assumptions, and vice versa. Truth and effectiveness are different values with no necessary connection. We cannot judge the value of research 'entirely in terms of practical success'(202) and research anyway requires 'some autonomy from practical concerns', rejecting the idea that research that is not immediately useful in emancipation is illegitimate. Political movements must reflect on goals and strategies and gather information, but with them, practice shapes the process of enquiry, and this is both good and bad — the bad bits involve the pressures of practice and the commitments of the audience restricting topics of enquiry and limiting testing. Raising wider questions may be counter-productive from the point of view of immediate practice, however. This makes a case for some form of 'institutionalised inquiry that is not geared to the immediate requirements of any single political or social practice', but whose findings might be relevant to a wide range of practices. It should not be directed to pragmatic goals or judged in pragmatic terms, but judged against a wider view of public relevance. This is the traditional basis for scholarship and science, and we need to defend it against those who would 'wish to tie inquiry to their own practical goals, and this includes feminists'. In conclusion, feminism is a major contributor to the social sciences, but feminist methodology is not 'a coherent and cogent alternative to non-feminist research'. Ideas are shared with non-feminist literature. Many ideas are unconvincing. Separating off a specifically feminist methodology 'homogenises other research and other views' (203). It suggest a strong link between political and philosophical assumptions and research decisions, but research is 'a more pragmatic enterprise'. Methodological distinctions can lead to Balkanisation [apparently Merton's term]. It is counter-productive in that there would be no obligation to adopt feminist methodology if you are a non-feminist, or to respect their findings. Others should engage in debates with feminists. 'This article is offered in the spirit of such a debate' (203) [There are some interesting notes, including note 12, 204, that says that collaborative research with other women is not unproblematic and might actually involve a form of exploitation]. Replies: Gelsthorpe,L. (1992) Response to Martin Hammersley's Paper 'on Feminist Methodology'. Sociology, 26 (2): pp 213 – 18. DOI: 10.1177/0038038592026002004 This addresses the first three headings [since Ramazanoglou has addressed the 4th]. Gender is privileged by feminist but not any uniform way, as some of the feminists quoted indicate — there are intersections. Nor should emphasis on gender just mean an emphasis on women. Early feminist work did privilege gender for rhetorical reasons, but later material looks at masculinity and femininity. Gender is a crucial mediating variable when discussing racial or class oppression, for example. Female emancipation should not be emphasised if feminism is seen as 'a transitory stage to an ideal human state which would accommodate the diversity of men and women' (213), but it is acceptable 'as a necessary corrective to male domination' (214). The focus on experience is indeed misplaced if it assumes that anything goes, and critical awareness is abandoned. However much recent feminist work talks about 'the role of experience in method rather than versus method'. Quantitative methods are commonly dismissed as inconsistent with feminist values, mostly because they offer only a shallow understanding and can even ignore sex and gender differences. It is intensive quantification in particular that needs to be tackled. The emphasis on experience does recognise personal reflexivity in the sense that researchers are also human beings and that they use normal human methods to understand. This is part of the 'quest to be open and honest' rather than an attack on objectivity. Scientistic objectivity does need to be questioned. None of this implies not being 'critical, rigorous or accurate'. Instead, researchers should make ' interpretive schemes explicit'. Hammersley is right to express a concern of the view that accounts of women's experiences offer better knowledge of situations — this is his version of standpoint theory. Nevertheless researching standpoints can be valuable, especially if we remember that people do not just occupy one standpoint. This is a safeguard against excluding those who do not seem to adopt that standpoint: 'we cannot assume that black/white, young/old, and so on, experience life in the same way' (215). Knowledge can emerge from the struggle to overcome domination, raising awareness of 'multi dimensional experiences of reality' and revealing how everyday worlds are shaped. It is another thing to claim that this would provide better knowledge. We should focus instead on knowledge production, how the process of knowing can be traced to particular standpoints [sociology of knowledge then? Any shared criteria or is this relativism?] . There is a need to distinguish good from less good knowledge [how exactly, though?] , but the knowledge claims of science 'can be viewed as deeply suspicious' [citing Feyerabend and Popper — neither of them feminist as far as I know]. Cain argues for 'successor science'[so that's where it comes from in Denzin's stuff on indigenous people]. Personal reflexivity is linked with theoretical reflexivity, and we can pursue the second by drawing upon a number of 'guidelines developed from realist ontology and epistemology' to provide a foundation [not a feminist foundation then?] — Including 'hermeneutic techniques, theoretical pragmatism' (216). She still argues that knowledge produce from this standpoint 'must "work" for those from whose standpoint it was produced' [cf plugging in]. But an adequate theory must explain not only ourselves but also 'those whom we investigate'[does she mean women and their male oppressors, or researchers and researched?]. Feminists are already aware of the difficulties of managing hierarchical relationships. They should be abandoned in research because otherwise the relationship involves 'treating people as mere objects', with a link between sex objects and research objects for Stanley and Wise. [BIzarre universal humanism which invovles, to take Schutz's example, a full intimate knowledge of the postal worker's world as she delivers the mail] It goes against the traditional advice to be entirely neutral and to refuse all questions about the researchers opinion. However there is a difference between dismantling power differentials between women and women and between female researchers and men. The first will value interactive methodology, but not necessarily the second, for example where male offenders are being researched. There are problems with democratising the research process discussed by feminists themselves, and Barker has even argued about the dangers of a false equality where researchers somehow have to negate their knowledge and skills in a gesture of equality. So while Hammersley's comments are useful, he does tend to miss out the debate and internal critique in feminism. Feminists have been struggling with these issues and there is no single feminist line. So his discussion ought to really begin with his concluding points on the complexity of the issues. There is no single feminist position on all these issues. Finally, feminists have not particularly argued for a domination of feminist methodology — there is no consensus on methodology although there are methodological preferences.  'Hammersley has thus demolished a case that never really was'(217). Ramazonoglou, C. (1992) On Feminist Methodology: Male Reason Versus Female Empowerment. Sociology 26 (2): 207 – 12 Hammersley has a 'fundamental opposition to the central tenets of feminist thinking' (207) which can be assembled from the piecemeal points he makes. Contradictions include acknowledging the positive contribution feminism is made while finding feminist methodologies unconvincing. He wants to privilege reason 'in some sort of established scientific community'. For him, knowledge reflects 'a conceptual split between the goals of science and rationality and the goals of politics and personal commitment'[old weberian fact/value stuff]. Any paradigms trying to link these two threatens an obstacle to open debate and dogmatism. The transformation of gender relations cannot be the point of feminist research, therefore. Feminism along with Marxism can be dismissed as political and dogmatic and therefore not scientific or rational. To give feminists any credit involves accepting empirical findings is valid but not accepting the grounds on which this knowledge has been produced. Hammersley does not 'seem to see his own place in the research process as a gendered one' and assumes that his stances value free and gender free. The same goes with his demands that public relevance is the issue, but this ignores how the public domain itself is gendered [and dominated by class]. He can only rely on some tradition of scholarship and science, which again is taken for granted: the rational validator is do not seem to have political positions or make key assumptions or have practical goals. If reason and emotion are linked, feminists might be forgiven an emotional reaction at this point, referring to 'unthinking sexism and implicit racism… Widespread in sociology'. (208) Hammersley might at least have looked at 'the feminist critique of the Enlightenment dualisms'[but are these valid critiques, just to close the circle]. Hammersley's position has been analysed and rejected by feminists. Value neutrality is not possible, objectivity is not separable from subjectivity, any more than mind is from body reason from emotion. Feminists have attacked these conceptions of rationality and 'male dominated and male defined relevance in the public domain'. Western scholarship in science depend on views that are 'blatantly sexist and racist, and privilege middle-class males'. Notions of reason and science are 'masculinist' and used to exclude and belittle women and other subordinates. Feminist scholars themselves have experienced sexism and subordination in the so-called neutral Academy. All schools of sociology 'systematically privilege male knowledge, experience and interests but without acknowledging that this is what they do' (209). There are intersections with other forms of domination. Feminist research does not just choose gender as an additional variable but 'offers new theories of power' and how it works in social divisions, including those between women. Feminists are aware of the difficulties of validating experience and there is disagreement. This should not detract from the point that 'women's experiences of gendered power relations' as a source of knowledge has been hitherto ignored by 'masculinist sociology'. Even theorising has not been possible, hence 'new ways of knowing', feminist methodology. If we study domestic violence, we cannot study it 'apolitically' [but should we try to?], Especially if we see '"normal" domestic behaviour as violent' [so we have politicised it first by definition]. Knowledge gained of domestic violence from initially political aims has led to new knowledge and identified the 'social relationships, ideas and institutions which allow such violence to develop'. This sort of effort both empowers women and problematises notions of knowledge and rationality. Feminist methodologies address issues of ways of knowing and criteria for valid knowledge. Hammersley has ignored this because he lacks 'a theory of power' and has a 'supposedly ungendered stance' (209). He is ready to accept inequalities of class, race and gender while ignoring the effects they have on power, on ways of knowing, and on experience. Feminist methodologies are not claiming privilege. There are varied explorations of how we validate knowledge. These raise difficult problems about truth which had been evaded by sociology before. We now know about 'power struggles throughout the research process' (210) and can challenge the way they have been silenced 'as leading to gendered knowledge which privileges male interests'. There are differences between feminists, including methodological and epistemological differences. Some are empiricist, others realist. Sometimes standpoint epistemology helps share the experiences of the researched, but in other cases, there are power relations dividing subjects from researchers, and subjects do not 'theorise themselves'in the same ways in which they are 'theorised by feminists' [false consciousness]. Feminist methodologies are about seeking knowledge and truth, but also forms of political commitment. This is a necessary characteristic, an 'immensely ambitious project in seeking ways of knowing which avoids subordination'. There is no alternative to political commitment or 'any other ways of knowing'. Validation is always a problem. Ways of knowing which Hammersley likes also have political goals, and his support for them 'will by default empower men by maintaining the status quo'. Weaknesses identified in feminism in fact applied to sociology in general: feminism has just made them explicit. Threats to validity are matters of concern for funding bodies colleagues and students, so the 'problems of explanation raised by feminism' (211) are 'necessary and urgent'. Feminism has exposed the role of power relationships. It has exposed unreasonable assumptions about 'reason'. It might have no general solution, but we cannot dismiss it in the name of 'the assumed superiority of a rational scientific community'. Researchers have political commitments and elements of subjectivity. It is more logical to accept them rather than 'to assume that some of us can rise above them'. In general, feminist work has been open, creative and productive, but Hammersley's position threatens to close off further discussion by seeing gender as 'an extraneous variable within a male appropriation of knowledge'. This would disempower women and 'should be resisted'. back to Hammersley page