The Thomist: A Speculative Quarterly Review

640 BOOK REVIEWS not ever choosing to harm these values. If this view is irrational, it is not simply because it is not consequentialist. Harris goes on in the same vein to suggest that the double effect doctrine might assume that the states of mind which will be sacrificed are all good, and likely to worsen. Thus, the double effect doctrine is as rational as the action of those South American Indians who killed their newly baptized babies so that they would die in the state of grace. Actually, Harris's view should be that the double effect doctrine is less rational than the Indians' killing of the babies since it involves the plainly false assumption that the states of mind to be sacrificed are good. This is hardly serious philosophical criticism. In short, this book raises serious questions but does not present compelling answers to them. More than anything else, it reveals the necessity for serious analytical work in the theory of human action and in related areas of moral psychology. The Center for Thomistic Studies University of St. Thomas Houston, Texas. JOSEPH M. BOYLE, JR. Identity and Essence. By BARUCH A. BRODY. Princeton, N.J.: Princeton University Press, 1980. Eminently readable and carefully argued, Identity and Essence successfully combines a purely formal definition of identity, a version of the Leibnizian doctrine of the identity of indiscernibles, with an Aristotelian theory of essences . Brody develops this position through a sustained, often critical, consideration of the prominent alternatives presented by his contemporaries. The reasoning is compact, never superfluous, and always insightful, even if not always convincing in every respect. Brody begins by defending the adequacy of a general theory of identity. He challenges the assumption that " the truth conditions of claims concerning identity may vary as the type of entity in question varies," e.g., physical object, person, or property (4). For any entities to be identical, it is necessary and sufficient that they have all of their properties in common. By definition, x = y = def. [ (F) (Fx s:= Fy) ]. Brody takes a property to be anything that is had by an object, even only a single object. To establish the relatively more controversial point that having all properties m common is sufficient for identity, Brody offers the following proof: " (i ) Suppose that a and b have all of their properties in common (ii ) a certainly has the property of-being-identical-with-a ·BOOK REVIEWS (iii) so, by supposition, does b (iv) then b = a" (9). 641 With its construal of identity itself as a property, this proof and thus the definition it is supposed to help establish appear circular. Most of chapter one is devoted to refuting these and other objections to the general theory. Taking a cue from Carnap, Brody argues that his definition of identity is not circular, but impredicative. ' Identity ' does not occur in the definiens (' the relation which holds between x and y just in case they have all of their properties in common ') , Brody insists, nor must the relation of identity be understood to understand the definiens. Brody's definition is satisfactory from the perspective of coming to know that two objects are identical. Less convincing, however, is the claim that his general theory is non-circular. I leave aside the problem of relating his definition and proof to one another as well as the question of the legitimacy of construing identity as both a relation and a property. The crucial issue, on Brody's own terms, is whether' sharing a property' can be understood without presupposing the relation of identity to be defined by that phrase (in conjunction with other phrases) . Brody is right in insisting that the definition, like its application, requires knowledge of neither all of the objects ' properties nor their identity-properties. However, even the meaning of' sharing a property' presupposes the relation of identity. Brody's theory is meant to apply unrestrictively to objects and properties. Suppose, then, that x and y have property F in common or, in other words, Fx and Fy. To assert that x and y have F in common can only mean that there is an identity between the F that is predicated...