«GENDER, METHODOLOGY AND PEOPLE’S WAYS OF KNOWING: SOME PROBLEMS WITH FEMINISM AND THE PARADIGM DEBATE IN SOCIAL SCIENCE ANN OAKLEY Abstract This ...»
SOCIOLOGY Vol. 32 No. 4 November 1998
GENDER, METHODOLOGY AND PEOPLE’S WAYS OF
KNOWING: SOME PROBLEMS WITH FEMINISM AND
THE PARADIGM DEBATE IN SOCIAL SCIENCE
This paper examines the character of the debate about ‘quantitative’ and ‘qualitative’ methods in feminist social science. The ‘paradigm argument’ has been central to feminist social science methodology; the feminist case against ‘malestream’ methods and in favour of qualitative methods has paralleled other methodological arguments within social science against the unthinking adoption by social science of a natural science model of inquiry. The paper argues in favour of rehabilitating quantitative methods and integrating a range of methods in the task of creating an emancipatory social science. It draws on the history of social and natural science, suggesting that a social and historical understanding of ways of knowing gives us the problem not of gender and methodology, but of the gendering of methodology as itself a social construction.
Key words: Feminism, gender, methodology, research.
Since the 1960s, feminist1 social scientists2 have had a great deal to say about methodology. The case that has been mounted against mainstream/ ‘malestream’ social research has been an important part of the project of women’s studies; and within this, the dualism of ‘quantitative’ and ‘qualitative’ methods has played a central role. The use of ‘qualitative’ research methods has been aligned with a feminist perspective, while ‘quantitative’ methods have been seen as implicitly or explicitly defensive of the (masculinist) status quo. The feminist critique has joined other modern approaches to knowledge which dispute the appropriateness of a ‘natural science model’ to the social sciences (Bryman 1988; Hammersley 1989;
This paper takes a critical look at what has been called a ‘paradigm dialog’ (Guba 1990) – although the character of the discussion is often more oppositional than this phrase would suggest. The paper argues that the debate about research methods is more than a dialogue/argument concerning the best research technique to use in which circumstances; it offers a narrative which is about the relations between the social and scientiﬁc division of labour, the cultural production of masculinities and femininities, and the processes used to establish an understanding of the social and material world. Seen from this viewpoint, methodology is itself gendered; and one of the chief functions of the quantitative/qualitative dichotomy is as an ideological representation. The paper 708 ANN OAKLEY suggests that, in order to understand the relationship between methodology and gender, we need to consider the intellectual and social origins of both social and natural science, and the history of the relationship between them.
This history indicates that the ‘feminist methodological case’ has been made largely in ignorance of the way in which different approaches to knowledge have historically been sited within social and natural science and have been used by social reformers, including feminists. One explanation of the feminist dispute with the value of quantitative methods and the claim to own qualitative ones is that this methodological position contributes an important part of the ‘professionalisation project’ of feminist social science. The paper ends by suggesting that maintaining the division between ‘quantitative’ and ‘qualitative’ methods and the feminist case against quantiﬁcation is ultimately unhelpful to the goal of an emancipatory social science.
The ﬁrst section of the paper examines in more detail the nature of the feminist criticism of quantitative methods.3 Secondly, some issues to do with truth-claims in qualitative work are brieﬂy considered. This is followed by a historical outline of how the paradigm argument is located in the historical identities of both natural and social science. Lastly, an argument is put for dissolving the dualism of quantitative and qualitative methods and adopting a feminist empiricist approach which is more likely to promote policy-relevant research ‘for’ women. The focus of the paper is on feminist arguments about methodology as the main ‘problematic’, but the argument is relevant to methodological debates more generally, because embedded in the feminist ‘case’ against quantitative methodologies are other epistemological positions contending the superiority of ‘qualitative’ methods.
Hearing the Silent
When academia was ﬁrst challenged by feminism in the late 1960s and early 1970s, and the biases of ‘masculine’ knowledge and women’s invisibility were revealed, the argument quickly developed that ‘positivist, quantitative research methodology’ (Mies 1983:120) cannot be used uncritically to further the political goals of academic women’s studies, because the voices of women as an oppressed social group are unlikely to be heard using such an approach.
Accordingly, the early feminist methodology texts all celebrated qualitative methods as best suited to the project of hearing women’s accounts of their experiences (see, for example, Bowles and Duelli-Klein 1983; Roberts 1981;
Stanley and Wise 1983).
‘Qualitative’ methods include participant observation, unstructured/semistructured interviewing, (some) life history methods and focus groups. These came to be seen as epistemologically distinct from the ‘quantitative’ methods of surveys, experiments, statistical records, structured observations and
GENDER, METHODOLOGY AND PEOPLE’S WAYS OF KNOWINGcontent analysis, although, in practice, the feminist critique mainly equated qualitative methods with indepth face-to-face interviews (Stanley and Wise 1993:3), and quantitative methods with enumeration in some form or other, and with the epistemological/philosophical position underlying the use of statistical techniques (Mies 1991:67). As Bryman (1988) has noted of the methodological literature generally, quantitative and qualitative methods tend to be portrayed as mutually antagonistic ideal types, and even as representing two different ‘paradigms’ of social science itself. The contemporary opposition dates from the 1960s, when Glaser and Strauss’s The Discovery of Grounded Theory (1967) was followed by a spate of books on qualitative methods (see, for example, Fletcher 1974; Loﬂand 1971; Schatzman and Strauss 1973).
While the ‘paradigm argument’ pre-dated the feminist critique of methodology, the arrival of feminist scholarship introduced new themes. It became
clear that the dualism of quantitative and qualitative is paralleled by others:
hard/soft; masculine/feminine; public/private; rational/intuitive; intellect/feeling;
scientiﬁc/artistic; social/natural; control/understanding; experiment/observation;
objective/subjective; separation/fusion; repression/expression; autonomy/ dependence; voice/silence (Belenky et al. 1986; Gilligan 1982; Millman and Kanter 1987; Reinharz 1984). The dualism of ‘quantitative’ and ‘qualitative’ methods became inextricably bound up with the central contentions of women’s studies – that traditional social science ignores or marginalises women, that all the major social theories explain the public world of labour but not the private world of work and the home (Elshtain 1981; Stacey 1981), and that the areas of social life which have particularly concerned women – caring, bodies, emotions (Rose 1994; Martin 1987; Williams and Bendelow 1996) – have hardly been part of the sociological landscape at all.
The feminist critique contests quantitative research on a number of grounds: that the choice of topics often implicitly supports sexist values;
female subjects are excluded or marginalised; relations between researcher and researched are intrinsically exploitative; the resulting data are superﬁcial and overgeneralised; and quantitative research is generally not used to overcome social problems (Jayaratne and Stewart 1991:86; Jayaratne 1983:145–6).
Signiﬁcantly, such criticisms of quantitative research overlap with general feminist critiques of mainstream/‘malestream’ social research (see, for example, Eichler 1988). ‘Normal’ social science research has been equated by
feminist critics with ‘rape’: ‘Research is frequently conducted on rape model:
the researchers take, hit and run’ (Reinharz 1984:95).4 The Three Ps Underlying the various arguments in the feminist case against quantitative methods are three fundamental objections: the case against positivism; the case 710 ANN OAKLEY against power; and the case against p values, or against the use of statistical techniques as a means of establishing the validity of research ﬁndings.
(a) The reality of positivism Positivism is an approach to knowledge which sees material and social worlds as equivalent, and which limits knowledge to ‘facts’ knowable through human experience (Kolakowski 1972; see Bryant 1985). A positivist social science is primarily a search for social ‘facts’ and for social laws which will predict behaviour. The adequacy of a scientiﬁc theory within positivism is guaranteed by its ‘objectivity’ or lack of bias; following rules of valid inference means that knowledge claims should be veriﬁable by anyone. Removal from the research process of researchers’ own values and experiences is essential to this requirement of veriﬁability (Jaggar 1983). A basic precept is the ‘subject/object’ dichotomy: what is studied is an ‘object’, which the knower/researcher can look at in a value-free and neutral way. All these constructions are problematic within feminist and other modern critiques of knowledge. For this reason, positivism has effectively become a ‘term of abuse’ (Giddens 1978:237). In the most extreme versions of feminist anti-positivism, the difference between ‘fact’ and ‘ﬁction’ disappears completely, and ‘truth’ and ‘objectivity’ are regarded as synonymous with ‘lies’ and ‘subjectivity’ (Stanley and Wise 1993:171). ‘Objectivity’ is reframed as ‘male subjectivity’ (Caplan 1988).
Supporting the rationale for the rejection of positivism is the association between it and ‘the scientiﬁc method’ which is condemned on political grounds. Science props up ‘the fantasy of a calculable universe where everything operates according to quantiﬁable laws’, a fantasy which is ‘central to the calculating project of bourgeois democracy and the strategies of global domination’ (Walkerdine 1989:208).
(b) The power of knowing When social science imitates natural science, the resulting delusion is that the barriers between researcher and researched keep researchers ‘safe from involvement or risk’ (Reinharz 1984:368). The idea of a social world to be known about implies a knower; the knower is the expert, and the known are the objects of someone else’s knowledge, not, most importantly, of their own.
But feminist knowers must reject ‘any mode of explanation which requires or sanctions the imposition upon the female subject of the theorist’s own views as to who she is, what she wants, and what she should have’. This is because, ‘One cannot survey the human race from a great and learned distance, proclaiming loudly that one has found the truth for the good of all the nameless, faceless abstractions one has never really bothered to take seriously’ (Elshtain 1981:303). The hierarchical situation – the position of the researcher as expert knower – invalidates any data that come out of the
GENDER, METHODOLOGY AND PEOPLE’S WAYS OF KNOWINGresearch process (Mies 1983:123). Just as hierarchy produces data that are by deﬁnition invalid, so it is contended that where non-hierarchical relationships between researcher and researched exist, the resulting data are intrinsically more valid (Acker, Barry and Esseveld 1991:146).
There may be practical reasons why hierarchical research methods do not work, for example because research participants treated in this way respond to their objectiﬁcation by not trusting researchers, and therefore by lying to them (Edwards 1990). But this is not the essential objection. The essential objection is that the unequal power relationship between the knower and the known conﬂicts with the moral obligation at the heart of feminism to treat other women as you would yourself wish to be treated, and in this sense is seen to be at odds with feminism’s emancipatory ideal. The notion of ‘objectivity’ calls up the ideological screen of ‘objectiﬁcation’ and the adjective ‘objectionable’. It has been argued that objectivity in research methodology is nothing more or less than an excuse for the same sort of obscene power relationship which leads women to be sexually assaulted and murdered (Stanley and Wise 1983:169).