'Science' in the Social Sciences

3. Use and Interpretation of Statistical Data

Perhaps the leading figure in the philosophical dispute about the use and interpretation of statistical data in the social sciences was Peter Winch. In his book, The Idea of a Social Science and Its Relation to Philosophy (published in 1958 and still in press), Winch exploited Wittgenstein’s revolutionary re-thinking of the concept of ‘grammar’ and sought to apply it to topics central to social-scientific thought and research – action, reason, explanation, causation, rule and others. For the later Wittgenstein, ‘grammar’ encompasses far more than it does in traditional linguistics: it comprises the rules of use of words and expressions in the language of everyday life, and not just the ‘syntax’ of such expressions. Applying Wittgenstein’s method of ‘grammatical elucidation’, Winch argued that the concept of a ‘cause’ is not logically (grammatically) appropriate for the explanation of ‘human action’ in the sense in which the word ‘cause’ is used in the natural sciences.
*Image Courtesy of Francis & Taylor, Inc.

For Winch, “all behavior which is meaningful (therefore all specifically human behavior) is ipso facto rule-governed” (Winch, 1988 ed.,:52). Rules are not determinants, since it is a central feature of the concept of a ‘rule’ that rules can be broken, and this is in sharp contrast to the causal concept of a ‘law’ (in the natural-scientific sense of this word). Thus, for example, Durkheim’s effort to explain acts of intentional self-destruction (suicides) in causal terms would constitute a logically inappropriate program, and not simply one that failed to work out empirically as rigorously as he had hoped.

At first blush, Winch’s critique appeared devastating to the entire enterprise of constructing a social science. The promise of Yule’s arguments about using statistical data in order to discover etiological (causal) connections in the domain of the social world, and, along with it, the promise of being able to predict outcomes and from there to control them by means of informed social policies, appeared misconceived at best. However, it is crucial to bear in mind that Winch’s argument pertained solely to human actions or activities. Social states of affairs such as rates of inflation or levels of (un)employment cannot be exempted from causal reasoning, and because such states of affairs lend themselves to quantification and measurement the project of causal modeling could proceed, with whatever limitations that might arise in their construction being practical rather than logical.

Whether or not causal propositions in economic theory are strictly lawful or ‘nomological’ is still a debated issue, but this need not concern us here. See Hausman, 1994, Parts II and III for some discussion.

*Winch, P. (2007). The Idea of a Social Science and Its Relation to Philosophy. London and New York: Routledge Classics.
Many Taylor & Francis and Routledge books are now available as eBooks. For more information visit www.tandf.co.uk and eBookstore www.tandfebooks.com.