Hume and the Problem of Causation by Tom L. Beauchamp and Alexander Rosenberg (review)

BOOK REVIEWS 119 cern, I would submit that a faith with richer doctrinal content could be still more fruitful for Christian existence, and actually has been in the lives of many saints. Wiles's conclusions cannot be effectively challenged on his own methodology . Although he wishes to stand " recognizably within historical Christian tradition," he considers it appropriate " to test, to review and where necessary to revise both the traditional affirmations of faith and its contemporary insights" (p. 128). Accordingly he takes Scripture and traditicn as providing something less than authoritative witness to divine revelation. He treats them as offering materials to be interpreted ar,cording to the norms of contemporary experience and common sense. This method guarantees in advance that the novelty of the gospel will be blunted. An alternative approach would have to reckon with the possibility that God might do and reveal things beyond all that the human imagination might be inclined to conceive. The idea of Christianity as a radical reversal of all human expectations is almost totally absent from the present volume. Yet if this idea is correct, Christianity cannot help but be far more distinctive than Wiles would have it. AVERY DULLES, S.J. The Catholic University of America Washington, D.O. Hume and the Problem of Causation. By ToM L. BEAUCHAMP and ALEXANDER ROSENBERG. New York: Oxford University Press, 1981. Pp. xxv + 340. $23.50. Although of interest primarily to philosophers of an empiricist persuasion , familiar with the recent literature on causality, this clear, careful, and precise volume can be easily summarized. Beauchamp and Rosenberg do very careful textual analysis of David Hume's works, but they are just as interested in constructing "a unified and defensible Humean theory of causation" (p. 32; italics added). Their principal target throughout is J. L. Mackie's The Cement of the Universe (1974), and they accuse Mackie of "offer[ing] epistemological answers for metaphysical problems" (p. xxiii, in a very handy analytical table of contents, referring to chapter 7, which ends with the claim that Mackie " seems to confuse ontological and epistemological accounts of causation " (p. 282]). The account of (metaphysical) causality Beauchamp and Rosenberg offer is a modification-involving a criterion they call "Nominal Extensionality '' (p. 269)-of a proposal due to Jaegwon Kim (1973), inter- 120 BOO:K'. REVIEWS preting Hume's account as requiring ''the time-like, the space-like, and ... abstract items " to be brought " together under the category of a structured event" (p. Z51). (This will be referred to again later.) Beauchamp and Rosenberg provide their own summary: "[J. A.] Robinson and most all recent writers on causation believe that Hume holds a pure regularity theory of causation.... J. L. Mackie, who allies himself with Robinson, ... [says Hume means] that statements of causal connection are nothing but statements of de facto constant conjunction" (pp. 31-32). This view, Beauchamp and Rosenberg then say, is indefensible because it fails to "distinguish causal laws from statements of de facto regularity" (p. 32). This they counter, saying, "After all, there is the second definition of ' cause ' . . . [plus] Hume's repeated assertion that, ' According to my definitions, necessity makes an essential part of causation ' (Treatise, 407)" (ibid.). And what they then propose is to provide their defensible reconstruction of Hume's theory of causal events. Aside from Mackie, Beauchamp and Rosenberg take on G. E. M. Anscombe (chapter 3), William Kneale (chapter 4), and Bertrand Russell and Richard Taylor (chapter 5)-among others. In chapter 8, Beauchamp and Rosenberg take up Hume's views on causal explanation, where their opponents are C. J. Ducasse (Causation and the Types of Necessity, 1924), R. G. Collingwood (An Essay on Metaphysics, 1940), H. L. A. Hart and A. M. Honore (Causation in the Law, 1959), in addition to Mackie. Throughout, they maintain they are not talking (at least primarily) about epistemological skepticism-except with respect to the claims of rationalism . According to the authors, Hume is not at all skeptical about scientific rationality; indeed, that is what (they say) he is appealing to against the rationalists ! Undoubtedly, Beauchamp and Rosenberg will encounter opposition from within the empiricist camp-especially from people like Mackie who favor an epistemological (and skeptical...