Robert L. Frazier ( )

Facing Evil
Princeton University Press, 1990. xii + 250 pp. $29.95

Facing evil, according to Professor Kekes, is difficult because of the unpleasant emotions evil evokes. Facing his book on this subject poses no such problems. Kekes's language is vivid, his presentation is lively and the book is well organized. In addition, he discusses an intriguing set of topics. The book contains an illuminating discussion of the tragic view of life and the pervasiveness of evil, an instructive discussion of responses to the tragic view, proposals for reducing the amount of evil caused by humans, and, intertwined throughout, a clearly presented virtue based theory of morality. And these are merely subsidiary discussions.

The heart of the book is a vigorous argument that theories focusing on the evaluation of acts (what Kekes calls ``choice-morality'') cannot propose an adequate response to evil's occurrence, while virtue based theories (character-morality) can. The argument has three main parts. First Kekes defines important concepts and argues that much of the evil done by persons is unchosen (Chs. 1-4). Then he argues that persons can be morally culpable for unchosen evil and that this is inconsistent with the acceptance of choice-morality (Chs. 5-9). Finally he proposes what he takes to be an adequate moral response to moral evil (Chs. 9-12). I will concentrate on the first two parts.

Evil is the central notion in Kekes's argument. He takes it to be ``undeserved harm inflicted on human beings'' (p. 4). And he takes moral evil to be, roughly, evil for whose occurrence a person may be culpable. Persons are harmed when something interferes with their living their versions of the good life. Harm is undeserved when there is no ``morally acceptable justification for causing it'' (p. 65). The only acceptable justification is that the harm is necessary to promote human welfare (pp. 56-57).

This characterization of evil appears to be in tension with another of Kekes's views: that what a person receives should be a function of that person's character (pp. 58-59). Some harmful acts that promote human welfare are bound to do so irrespective of individuals' characters. Thus I do not see how Kekes can avoid the unwanted result that harming a person might be justified because it promotes human welfare, regardless of that person's character.

Choice also plays a central role in Kekes's arguments. Roughly, choice requires ``the availability of alternatives and the possession and exercise of capacities to make decisions regarding them'' (p. 67).

Given Kekes's characterizations of evil and choice, it should be obvious that not all evil is the result of persons' choices. Earthquakes, for example, can result in unchosen evil. However, Kekes is far more interested in unchosen evil caused by persons' having deficient characters and in whether such evil is moral evil. One of his particularly persuasive examples of a vice that may lead to unchosen evil is dogmatism. A dogmatic person may have mistaken moral principles inculcated from childhood, along with an aversion to questioning them. Such a person, according to Kekes, lacks the cognitive capacity required for choice. Yet that person may cause great evil. Is this evil moral evil?

Here is where Kekes thinks choice-morality goes wrong. Supposedly, it has as a consequence that if ``agents did not choose to do the evil they have done, they should not be held morally accountable'' (p. 86). Thus the evil done by dogmatists would not be moral evil and they would not be culpable for it. Nonculpability, Kekes says, would preclude certain responses to evildoers. (Kekes mentions such things as shunning their company and creating institutions to protect us from them (p. 103). It is hard to see why these responses are not open to someone who does not believe the evil to be moral evil. We shun mad dogs and have institutions to protect us from them.) Kekes argues that this view about evil, what he calls ``the soft reaction to evil'', should be rejected and, consequently, that choice-morality should be rejected. He proposes that we accept the hard reaction to evil, which holds that ``the morally significant fact is that the agents of unchosen evil habitually cause undeserved harm, and this should affect our moral evaluation of them'' (p. 86). At times he even seems to hold the stronger view that all evil caused by humans is moral evil, whether or not it is habitual (p. 104).

Kekes supplies a number of arguments against the soft reaction to evil. They include an argument from preferences (that we prefer persons to be virtuous shows our commitment to the censure of persons who perform unchosen vicious acts (p. 90)), an argument from symmetry (if we are willing to praise persons for unchosen virtuous acts, we should be willing to censure persons for unchosen vicious acts (p. 101)) and, as his main argument, an argument from provenance (we should reject the soft reaction to evil because its source is an unacceptable view). This last argument can be outlined as follows.

Underlying the soft reaction to evil is the Kantian principle that ought-implies-can (p. 89). In turn, this principle is justified by the view that moral agency requires choice (pp. 88). This view about moral agency presumes Kant's conception of morality as involving free, rational beings willing law on to themselves (p. 88). Kant's notion of freedom is unacceptable because it requires the will to be independent of natural causes (p. 139). Consequently, we should reject the Kantian conception of morality and, since it provides the soft reaction's foundation, ``we should do likewise with choice-morality and the soft reaction to unchosen evil'' (p. 142).

Kekes also rejects two other theses associated with this conception of morality: that moral agency entails fundamental moral worth and that since persons are fundamentally good, causing moral evil involves a corruption of that good.

Although Kekes's discussions are interesting and much of what he says is right, there appears to be a gap between his arguments and what they are supposed to establish. He seems to present the available options as two: virtue based moral theories and act based moral theories that accept the Kantian notion of freedom. These are not the only options. The ``can'' of the Kantian principle ``ought-implies-can'' has been viewed as expressing a number of different notions of possibility, ranging from the sort of freedom against which Kekes argues to mere physical possibility, against which he has no argument. Indeed, the Kantian principle has even been denied by some proponents of act based theories who hold that there can be real conflicts of obligation. If Kekes's arguments are intended to justify dismissing all act based moral theories, they fail. If they are intended to refute theories accepting the Kantian notion of freedom, then, even if sound, they do not show that we should prefer the approach to morality taken by virtue based theories.

In any case, for someone who is puzzling over whether the evaluation of acts or the evaluation of agents is fundamental, Kekes's book is valuable. He formulates a credible version of a virtue morality and actually argues that it can account for a moral phenomenon (culpability for unchosen evil) that act based morality cannot. Even if, as I believe, his main argument is flawed, it helps the reader to see more clearly what is at issue between the two approaches to morality.


Robert L. Frazier