Understanding Human Action: Social Explanation of the Vision of Social Science by Michael A. Simon (review)

BOOK REVIEWS of course, object to the implicit nominalism, to the anti-realism of the entire Beauchamp/Rosenberg/Hume venture; if they have a cavil about the book, it will be that they are never given so much as a mention, much less a hearing.) Granted all these objections from people who would have wanted Beauchamp and Rosenberg to have written something different, their book is an impressive accomplishment. They would seem to have successfully extricated Hume from the entanglements of those, friends and foes alike, who would make him out to be a me;e regularity theorist and skeptic. They seem clearly to be right in claiming that Hume was skeptical primarily about rationalist claims with respect to causality. Whether Beauchamp and Rosenberg are as successful in defending their " event" view of causality (derived from Jaegwon Kim) is a matter likely to be debated for some time in empiricist circles-while process philosophers grumble from the sidelines that they come perilously close to giving up the rational empiricism they work so mightily to defend. Philosophy Department University of Delaware Newark, Delaware p AUL T. DURBIN Understanding Human Action: Social Explanation of the Vision of Social Science. By MICHAEL A. SIMON. State University of New York Press, 1981. Pp. 226. Cloth, $29.50; paper, $9.95. " This book," the author writes, " is an attempt to see what it is about human social life that makes a social science based on the model of natural science impossible" (p. viii). The "principal argument of the book " is outlined in the introduction as follows : The social studies have essentially to do with human actions; actions are free in the sense that they are not subject to explanation and prediction on the basis of strict causal laws; what empirical science is capable of investigating successfully cannot be free in that sense; hence the social studies cannot be empirical sciences (p. 2). Simon's first three chapters are devoted to establishing the first premise of this argument. In Chapter 1, he attempts to show that action "must be conceived as a logical primitive," in the sense that it is "irreducible to any concepts that do not themselves presuppose the notion of action or agency" (p. 7). The thesis of his second chapter is that "social relations as we know them are not possible without a concept of action which BOOK REVIEWS they presuppose, and that this concept has no essential role in language other than to serve this mediating function" (p. 27). And in Chapter 3, he locates the subject matter of social science in human action and its products (p. 41). In his first two chapters, and again in Chapters 4 and 7, Simon makes use of the work of some prominent philosophers of action. These chapters exhibit a less than adequate familiarity with the literature in the field. In Chapter 1, for example, Simon employs an objection of Ryle's to volitional theories of action (p. 18) and Melden's version of the argument that there can be no causal connection where there is a logical connection (p. 19), without considering recent, and even not-so-recent, meritorious replies, e.g., by D. M. Armstrong, A. C. Danto, Donald Davidson, and Alvin Goldman. And in Chapter 4 he advances a version of contextualism -" Reasons explain ... by representing the action as appropriate under the circumstances" (p. 62)-without even mentioning the weighty objections raised by such influential figures in the field as Davidson , Shaffer, and Davis. Simon's discussions of action are sometimes imprecise or seriously incomplete at important junctures. The following are representative instances of this problem. (1) Simon reports that "What is essential" to his " understanding of human actions is that they be conceived neither as random nor as determined in accordance with causal laws" (p. 131). Thus it is crucial for him to get clear on what determinism is. Yet he elaims-what is demonstrably false-that "If human beings are a proper subject matter for deterministic science, their behavior must in principle be totally predictable" (p. 170). Simon also contends, in apparent ignorance of Goldman's well-known attack on the position, that a person cannot predict his own actions...