Martyn Hammersley - The Limits of Social Science: Causal Explanation and Value Relevance

SAGE has been part of the global academic community since 1965, supporting high quality research and learning that transforms society and our understanding of individuals, groups, and cultures. SAGE is the independent, innovative, natural home for authors, editors and societies who share our commitment and passion for the social sciences.

Find out more at: www.sagepublications.com

Causal explanation and Value relevance

.

is a revised version of Why critical realism fails to justify critical social research, Methodological Innovations Online, 4(2), 2009, pp. 111. It incorporates some elements from Research as emancipatory: the case of Bhaskars realism, Journal of Critical Realism, 1, 1, 2002, pp. 3348.

An earlier version of was presented at a symposium on Methodological Issues in Research on Social Class, Education and Social Mobility, at the British Educational Research Associations annual conference in Manchester, September 2009, and at the British Sociological Associations annual conference at the London School of Economics, April 2011.

is based on a talk given at the conference Collisions, Coalitions and Riotous Subjects: The Riots One Year On, at South Bank University, September 2012, and at the British Sociological Associations annual conference at the Grand Connaught Rooms, London, April 2013.

I would like to thank several good friends and colleagues for their help and support, in particular Barry Cooper, Jeff Evans, Judith Glaesser, Roger Gomm, and Anna Traianou. Also, thanks to all those who responded with questions or comments at the talks on which these chapters are based.

At the start of his book Forms of Explanation, Alan Garfinkel (1990: vii) asks: If social science is the answer, what is the question? This is not an idle quip, his whole book is about the importance of being clear in social inquiry about exactly what questions we are addressing since this has implications for what would count as adequate answers. However, it should be noted that the format implicit in Garfinkels question here is perhaps not so much questionanswer as problemsolution: he is asking what problem social science is designed to solve in short, what the point of it is.1

In the spirit of Garfinkels own book, we might consider whether this question is an appropriate one to ask: is social science the kind of thing that is, or ought to be, designed to solve a problem? Garfinkels question can be read as implying a very particular view about the character, purpose, and justification of social science to the effect that it should be designed to provide a service to policymakers, practitioners, and/or relevant publics; and that social scientists ought to be accountable in terms of how well they have done this. While this sort of view is increasingly influential particularly on the part of funders of research, governments, interest groups, and others who use research findings it is open to serious question, and certainly should not be accepted at face value; not least because it may have damaging practical consequences (Hammersley 2011a; Holmwood 2011a).

Of course, demands for social science to produce usable knowledge, and to have an impact on policy and practice, do not come solely from outside the research community. Many social scientists also see this as their task, and perhaps even regard their work as only justifiable in these terms. This is true, most obviously, of those approaches to social inquiry that are explicitly tied to particular political and social movements, such as feminism, anti-racism, LGBT (lesbian, gay, bisexual, and transgender) rights, or disability activism. Here, the demand is frequently that research must make a difference in terms of its effects, rather than simply be aimed at producing knowledge. And much the same is true of calls for public sociology (Burawoy 2005), or for public forms of social science more generally (Cannella and Lincoln 2004), as well (even more obviously) as arguments in favour of policy science, action research, participatory inquiry and practitioner research (Lerner and Lasswell 1951; Lasswell 1971; Mies 1983; Carr and Kemmis 1986; Reason and Bradbury 2001). Indeed, this view is now to be found quite widely across social science, albeit usually in largely taken-for-granted forms.