Eudaimonism and Moral Absolutes

I agree with so much of Matthew O’Brien’s brilliant essay on why God matters in ethical theory that I hesitate to pursue a disagreement on a relatively minor issue. But because the disagreement illuminates some important questions, I cheerily forge ahead.

First, the areas of agreement: O’Brien and I agree that human beings have a definite nature, that this nature implies a certain final end for human beings, and that human actions are morally right or wrong depending on whether they are ordered as means to that end. We further agree that human nature was created by God, who ordains human nature to its final end and commands us to attain that end. Thus, by acting well, we perfect our nature and obey and please God, but by acting badly, we act contrary to our nature and disobey and displease him. A complete moral theory therefore involves God in various important ways, including in giving a full description of the final end (which involves contemplating God) and in explaining the nature of moral obligation (which involves obeying Him).

Moreover, O’Brien and I also agree that, notwithstanding the foregoing, in treating many ethical issues we can avoid referring to God. I believe that this is possible because we can work from a description of the final end that theists and atheists can both accept as being true as far as it goes (e.g., that the end is rational activity in a community of goodwill) and then consider the relationship of actions to that end.

O’Brien and I disagree about whether, having restricted our premises in this way, we can show that there are some actions that are always wrong, regardless of the circumstances. I think this can be done; O’Brien thinks it cannot.

Before going into why we disagree, it’s worth recalling why, in a eudaimonistic system, some actions are always wrong. The reason is that, for at least some ends, there are actions that are never ordered as means to those ends, regardless of the circumstances in which the action is chosen. For instance, if the end is winning a baseball game, intentionally walking in the winning run in the bottom of the ninth is never ordered to the end. Hence, given a description of the final end, if there are some actions that are never ordered to that end, these actions will be always wrong. In a eudaimonistic system, therefore, claims that certain actions are always wrong are based on claims about the causal structure of the world—claims that certain causes will not have certain effects, regardless of the circumstances.

Now, in an earlier article, O’Brien had argued that, in order to demonstrate that certain actions ought never be done, “you need to be able to appeal to God’s legislation of the moral law.” This is not so, and the reason is that God’s legislation does not affect the causal structure of the world, does not change which actions are ordered to which ends. Hence, unless we are adopting some kind of divine command theory of morality, an absolute divine prohibition on a certain kind of action will not help explain why such actions are always wrong. Given a kind of action, either such actions are never ordered to the final end, or else sometimes they are. If such actions are never ordered to the final end, then for exactly that reason such actions are always wrong, and there is no need to rely on a divine prohibition to ground their wrongness. On the other hand, if such actions are sometimes ordered to the final end, then God would not absolutely prohibit them, for it would be senseless to prohibit such actions in those circumstances in which they are actually ordered to the final end, and God does nothing senselessly. Hence, I stand by the conclusion in my earlier article that bringing in divine prohibitions does not help us explain why some actions are always wrong.

But as O’Brien ably shows in his latest article, God can figure in moral theory in various ways, not just as a legislator. Thus, in this article O’Brien has a quite different argument about why we need God to show that certain actions are always wrong. He writes that, if we bracket the existence of God, “it is difficult to see how the final end could be specified determinately enough to demonstrate that certain kinds of injustice, such as intentionally killing the innocent, could never be chosen in order to advance the final end.”

Now, this seems to be an attractive line of argument, for it is true that the less definitely the final end is specified, the more difficult it becomes to show that certain actions are wrong. The assumption, however, is that complete ethical theories that mention God give very definite descriptions of the final end (thus supporting arguments that some actions are always wrong), but that these descriptions are weakened and made vague when we re-describe the final end more generally to avoid mentioning God (with that result that such arguments no longer go through).

This, however, is simply not the case. In his full-blown and expressly theological ethical theory, Aquinas describes the final end as “the vision of the divine essence” (Summa Theologiae Ia-IIae.3.8), which is surely a much grander thing than my “rational activity in a community of good will,” but it hardly provides a more determinate basis for explaining why certain actions are always wrong. Indeed, in explaining why intentionally killing the innocent is always wrong, Aquinas has but the briefest and simplest of arguments: he merely says that the lives of the innocent conserve and promote the common good (and so the attainment of the final end), and that therefore killing them is always wrong (Summa Theologiae IIa-IIae.64.6). This argument makes no use whatsoever of the theological character of the final end, and so such an argument is quite as open to me with my definition of the final end as it was to Aquinas with his. If it suffices for him, it suffices for me as well.

Of course, O’Brien will say that it does not suffice, that such arguments prove not that killing the innocent is always wrong, but only that it is generally so. We still have to worry about the extreme cases, like the bombing of Hiroshima and Nagasaki, in which killing the innocent seems to advance the final end. What are we to say about such cases?

As Aquinas’s argument about killing the innocent suggests, the description of the final end employed—whether it be theological or not, and regardless of its specificity—is largely irrelevant. For, no matter how the end is specified, we can always invent sufficiently bizarre circumstances such that, in those circumstances, performing the action reputed to be always wrong will in fact advance the end.

For instance, take Aquinas’s ethics, in which the final end is the vision of the divine essence, and then assume that, if I but kill this one innocent man, untold millions will receive the grace of repentance and so come to share the beatific vision in heaven, thus greatly glorifying God. (Indeed, if he had had more insight into the Scriptures, Caiaphas might have seen himself as being in precisely this situation.) In such a case, killing the innocent seems to advance this fully theological final end. Or again, I said above that walking in the winning run in the bottom of the ninth is never the way to win a baseball game. But what if I know that the opposing team has been cheating, that the man at bat has a bad conscience about it, and that, if I walk him in, he will be overcome with remorse and will confess the whole scheme, with the result that the umpires will declare that the cheating team has forfeited the game and my team wins after all? In such circumstances, walking in the winning run actually wins the game for my team.

The problem with such examples, just as with the Hiroshima and Nagasaki problem, is that they trace the consequences of our actions too far. Only philosophers think that I am responsible for all the foreseeable consequences of my actions, and thus that all of these consequences must be taken into account in determining whether my action advances the final end. Normal people realize that my responsibility is more circumscribed. Which consequences of my action are properly attributed to me—and so count in determining whether my action advances the final end—and which consequences are too remote is an extremely difficult question. The answer in any particular case will depend heavily on the particular facts, but saying that all foreseeable consequences count goes too far.

There is a clear parallel in the law, for there is an immense body of legal doctrine concerning what lawyers call proximate causation, the whole purpose of which is to determine for which of the consequences causally following from his actions a defendant may be held liable. (See Prosser & Keaton on Torts, §§ 41-45.) Although, in general, defendants are responsible for the reasonably foreseeable consequences of their actions, as when a man who builds a fire on his own property on a windy day is held liable when the wind carries the fire to his neighbor’s house, this is not always the case. One clear exception is that defendants are not generally liable for the intentional wrongdoing of others, even when such wrongdoing is a reasonably foreseeable consequence of the defendant’s action, the idea being that the subsequent wrongdoer is the responsible party.

This principle, incidentally, suffices to dispatch the Hiroshima and Nagasaki problem: President Truman’s dropping the bomb should be evaluated taking into account its natural and inevitable consequences—such as the death of thousands of innocents—but not taking into account other consequences that are properly chargeable to others, such as the deaths of even more thousands of innocents who would have been killed in an American invasion of the Japanese home islands. Those deaths, had they occurred, would have been the responsibility not of Truman but of the Japanese authorities, who wrongfully resisted American forces waging a just war. It is a mistake, therefore, to count them in considering whether Truman’s action advanced the final end.

As I said above, determining for which consequences of his actions an agent is responsible (and thus which consequences need be counted in determining whether the agent’s action advances the final end) is a vexed question about which it is extremely difficult to generalize. But merely pointing out the existence of this question suffices, in a general way, to solve the puzzle of how, in a eudaimonistic system, certain actions can be said to be always wrong. To wit, to show that an action is always wrong, we need not show that such actions never advance the final end, including in the most bizarre circumstances; we need show only that such actions do not advance that end in normal and usual circumstances—that is, in all cases except those in which the advancement of that end would be beyond the responsibility of the agent. Under that standard, it becomes fairly easy to justify the proposition that intentionally killing the innocent—that is to say, murder—is always wrong.