Hostname: page-component-cd9895bd7-hc48f Total loading time: 0 Render date: 2024-12-28T11:23:04.152Z Has data issue: false hasContentIssue false
  • English
  • Français

Explanation in the Social Sciences: Singular Explanation and the Social Sciences*

Published online by Cambridge University Press:  08 January 2010

David-Hillel Ruben
Get access Rights & Permissions [Opens in a new window]

Extract

Are explanations in the social sciences fundamentally (logically or structurally) different from explanations in the natural sciences? Many philosophers think that they are, and I call such philosophers ‘difference theorists’. Many difference theorists locate that difference in the alleged fact that only in the natural sciences does explanation essentially include laws.

Type
Papers
Information
Royal Institute of Philosophy Supplements , Volume 27 , March 1990 , pp. 95 - 117
Copyright
Copyright © The Royal Institute of Philosophy and the contributors 1990

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 Winch, Peter, The Idea of a Social Science (London: Routledge, 1967). Quote from p. 86.Google Scholar

2 Taylor, Daniel, Explanation and Meaning (Cambridge University Press, 1973). Quote from p. 75.Google Scholar

3 Lesnoff, Michael, The Structure of Social Science (London: Allen and Unwin, 1974). Quote from p. 106.Google Scholar

4 Louch, A. R., Explanation and Human Action (Oxford: Blackwell, 1966), 233.Google Scholar

5 Michael Scriven wrote extensively on explanation, but I will confine my remarks to his ‘Truisms as the Grounds for Historical Explanations’, in Theories of History, Gardiner, Patrick (ed.) (New York: The Free Press, 1959), 443–75.Google Scholar

6 I do not think that it is always easy to see when an explanation is an explanation of a singular event, and when it is an explanation of a class of similar events, or a regularity. Grammatical form by itself may be a poor guide. Of course, there is a sense in which a token is always explained as a representative token of some specific kind. In a rough-and-ready way, though, I think we can see the intention behind the distinction between singular explanation and the explanation of a regularity.

Why did Vesuvius erupt in 79 C.E.? Why is there this unexpected reading in the light emission from such-and-such star? Why did Protestantism arise in northern Europe in the sixteenth century? These are requests for singular explanation. On the other hand, when a student explains the results of mixing two chemicals in a test tube in the laboratory, he is unlikely to be engaged in giving a singular explanation of the particular mixing at hand. He is in fact explaining why mixings of those types of chemicals have that sort of result. The claims of this paper are relevant only to genuinely singular explanations, not to ones which might have the grammatical form of a singular explanation but are in truth the explanation of a regularity.

7 He sometimes counts Deductive-Statistical as a third, but I ignore this here.

8 This assumption is made by David Lewis, since his conditions for something's being a full explanation are set so high: ‘Every question has a maximal true answer: the whole truth about the subject matter on which information is requested, to which nothing could be added without irrelevance or error. In some cases it is feasible to provide these maximal answers’. Lewis, David, Philosophical Papers II (Oxford University Press, 1986), 229.Google Scholar

9 If, as Lewis asserts, a complete explanation is a maximally true answer, it is not easy to see how it could ever be feasible to provide one. Lewis' theory is bound to drive one into an extreme form of explanatory idealism. The causal history of each thing to be explained stretches into the indefinite past; surely I can never provide all of that. On my view, the whole truth about something will typically contain many full explanations for the same thing, so that one can proffer a full explanation about a thing without having to state the whole relevant truth about it. David Lewis, ibid.

10 Hempel, C. G., Aspects of Scientific Explanation and Other Essays in the Philosophy of Science (New York: Free Press, 1965), 362.Google Scholar

11 Winch's criticisms of the covering law account of explanation were certainly not confined to the law and argument theses. He also argued against the view that having a reason is or can be the cause of the action it helps to rationalize. I consider the battle against this to be lost, principally due to the work of Donald Davidson. Once issues about causation, explanation, and law are disentangled, I can see no convincing argument against the claim that the having of reasons can be a type of cause.

12 No orthodox theorist would consider this condition by itself sufficient. Accidental generalizations have this form, too. Further, universally quantified material conditionals are true when their antecedent terms are true of nothing. So, if this condition were by itself sufficient for lawlikeness, and if nothing in the universe was an F, then both of the following would be laws of nature: (x) (Fx⊃Gx) and (x) (Fx⊃−Gx).

There are various proposals for adding further conditions to the one above. Some are proposals for strengthening the generalization by adding a necessity-operator: laws of nature are nomically necessary universally quantified generalizations. See for example: Kneale, William, ‘Natural Laws and Contraryto-Fact Conditionals’, Analysis 10 (1950), 121–25CrossRefGoogle Scholar; Popper, Karl, The Logic of Scientific Discovery, (London: Hutchinson, 1972)Google Scholar, Appendix *10, 420–41; Fisk, Milton, ‘Are There Necessary Connections in Nature?’, Philosophy of Science 37 (1970), 385–404.CrossRefGoogle Scholar Others ascribe to the universally quantified generalization an additional special epistemic status, or a special place in science, or impose further syntactic requirements. See Braithwaite, Richard, Scientific Explanation (Cambridge University Press, 1964), Ch. IX, 293–318Google Scholar; Nagel, Ernest, The Structure of Science (New York: Harcourt, Brace & World, 1961), Ch. 4, 47–78Google Scholar; Mellor, D. H., ‘Necessities and Universals in Natural Laws’, Science, Belief, and Behaviour, Mellor, D. H. (ed.) (Cambridge University Press, 1980), 105–19.Google Scholar Hempel himself is unclear about how to complete the list of sufficient conditions for something's being a law: ‘Though the preceding discussion has not led to a fully satisfactory general characterization of lawlike sentences and thus of laws….’, (Aspects, 343)Google Scholar.

Suppose, though, that the orthodox view does not provide even a necessary condition, let alone a sufficient one, for a sentence's stating a law of nature. If so, my argument would have to proceed somewhat differently. I am sympathetic to some of these non-orthodox views, but I do not deal with any of them here, nor with how their acceptance would alter my argument. See Dretske, Fred, ‘Laws of Nature’, Philosophy of Science 44 (1977), 248–68CrossRefGoogle Scholar; for a reply to Dretske, see Niiniluoto, Ilkka, Philosophy of Science, 45 (1978), 431–9CrossRefGoogle Scholar; Armstrong, David, What Is A Law of Nature? (Cambridge University Press, 1987).Google Scholar

13 Hempel, , Aspects, 336.Google Scholar

14 Ibid., 337.

15 In any event, Hempel himself, in ‘Aspects’, more or less abandons the view that every prediction is a potential explanation, in his discussion of Koplik spots and measles. Another case worth considering is this: suppose c and d both occur. Both are potential causes of e, but, since c occurs and actually causes e, d is merely a potential but pre-empted cause of e. One can predict but not explain e's occurrence from the occurrence of d, its pre-empted cause; e's explanation must be in terms of c, its actual cause, and not in terms of its preempted cause.

16 I am convinced by the argument of McCarthy, Tim, ‘On An Aristotelian Model of Scientific Explanation’, Philosophy of Science 44 (1977), 159–66.CrossRefGoogle Scholar See my Explaining Explanation (London: Routledge, 1990)Google Scholar, Ch. 6, where I develop the point at considerable length.

17 Scriven, , 444.Google Scholar

18 Ibid., 446.

19 See Nickles, Thomas, ‘Davidson on Explanation’, Philosophical Studies 31 (1977), 141–5CrossRefGoogle Scholar, where the idea that ‘strict’ covering laws may be ‘non-explanatory’ is developed.

20 Scriven, , 445.Google Scholar

21 See Davidson, , ‘Mental Events’, reprinted in his Essays on Actions and Events (Oxford University Press, 1980), 207–27.Google Scholar Also in Experience and Theory, Foster, and Swanson, (eds) (London: Duckworth, 1970), 79–101.Google Scholar

22 Scriven, , 458.Google Scholar

23 Scriven, , 464.Google Scholar

24 Scriven, , 465.Google Scholar

25 Hempel, , ‘Aspects’, 421–3.Google Scholar

26 I do not say that nothing but properties matter; names sometimes seem to matter to explanation too. Naming in one way can be explanatory, naming in a different way can fail to be so. If I do not know that Cicero=Tully, I can explain why Cicero's speeches stopped in 43 B.C.E., by the fact that Cicero died in that year, but not by the fact that Tully died in that year. But in any case, properties matter to explanation, even if names do too.

27 Hume, David, A Treatise of Human Nature, Selby-Bigge, L. A. (ed.) (Oxford University Press, 1965), 88.Google Scholar I think that much of the motivation for the inclusion of a generalization in every full explanation stems from the Humeian analysis of causation.

28 Aristotle gives us a definition of the incidental or accidental in Metaphysics V, 30, 20–5Google Scholar: ‘“Accident” means (1) that which attaches to something and can be truly asserted, but neither of necessity nor usually … for neither does the one come of necessity from the other or after the other, nor … usually … And a musical man might be pale; but since this does not happen of necessity nor usually, we call it an accident. Therefore, since there are attributes and they attach to subjects, and some of them attach only in a particular place and at a particular time, whatever attaches to a subject, but not because it was this subject, or the time this time, or the place this place, will be an accident.’ For a defence of my interpretation of Aristotle, see my Explaining Explanation.

29 See Friedman, Michael, ‘Theoretical Explanation’, in Reduction, Time and Reality, Healey, Richard (ed.) (Cambridge University Press, 1981), 1–16.Google Scholar See also his ‘Explanation and Scientific Understanding’, Journal of Philosophy, 71 (1974), 5–19Google Scholar, and the reply by Kitcher, Philip, ‘Explanation, Conjunction, and Unification’, Journal of Philosophy, 73 (1976), 207–12.CrossRefGoogle Scholar

30 Scriven, , 448.Google Scholar

31 The analogy was suggested to me by Peter Lipton.

32 This Hempelian notion of full explanation is to be contrasted with an even stronger sense of full explanation that one finds in the work of David Lewis. See notes 8 and 9 above.

33 This is the Hempelian ideal of full explanation. I describe in notes 8 and 9 above David Lewis' even more demanding ideal of full explanation.

34 Ayer, A. J., ‘Man as a Subject for Science’, Philosophy, Politics and Society, third series (Oxford: Blackwell, 1967), 6–24.Google Scholar

35 Davidson, Donald, ‘Mental Events’, Experience and Theory, 79–101.Google Scholar

36 I have discussed this briefly in ‘Marx, Necessity, and Science’, in Marx and Marxisms, Parkinson, H. R. (ed.) (Cambridge University Press, 1982), 39–56, see 54–6.Google Scholar