Warranted excuses and free will

Those who hold the Exculpating View are likely to find this analysis ques­tion-begging. If it is extraordinarily difficult to avoid fallacious unconscious reasoning based on excessive self-regard, as the two-thirds Milgram com­pliance rate suggests, giving in to it should not reflect badly on us.

The point of the objection is that we should be held responsible only for choices that are ours to make, and if we cannot help reasoning as we do - it is, remember, unconscious reasoning - it follows that the choice is not really ours. Let’s use the term “moral self” to describe those aspects of a person that engage in moral choice. Unconscious reasoning that we can’t easily avoid seems to come from outside the moral self, and for that reason it does not reflect badly on the moral self.

Take an extreme illustration. Suppose that a Milgram subject believes he is morally infallible, but he believes it only because a brain tumor has given him delusions of grandeur. And suppose that because of this belief he becomes a Milgram compiler in just the way that the corruption-of-judgment theory suggests. He is, in other words, a typical Milgram complier, with the one difference that excessive self-regard (the sin of pride) has become part of his make-up only because of the misfortune of the brain tumor. Surely, we should hold him blameless, because his judgment has been corrupted by something foreign to his moral self.

If that is right, however, we must consider the possibility that even in less extreme cases - everyday cases where we can’t point to an obvious cause like a brain tumor - susceptibility to excessive self-regard also derives from causal factors foreign to the moral self (brain chemistry, psychological laws, upbringing).

Clearly, we are here treading in the vicinity of the general question whether moral responsibility and determinism are compatible - an aspect of the Problem of Free Will, which one writer has aptly described as the most difficult problem in philosophy.^[480] I have no reason to believe myself equipped to solve that problem. A distinguished philosopher once warned that “it is impossible to say anything significant about this ancient problem that has not been said before.”^[481] He wrote these words in 1964; if they were true then, they obviously remain true now. Instead, I will simply lay out, with a minimum of argument, the views about free will and compatibilism that underlie the argument of this chapter. More importantly, though, I will show how these views respond to the objection I have just rehearsed.

Melvin Lerner’s deterministic line of argument suggests a blanket dis­claimer of responsibility for all bad acts, and - as legal theorist Michael Moore rightly argues - this implication amounts to a reductio ad absurdum of the theory that caused action is blameless action.^[482] Not that everyone would regard the implication as a reductio. Lerner believes “that (a) the way people act is determined by their past experience and their biological inheritance, and (b) this perspective neutralizes the condemning or blaming reaction to what people do.”^[483]

Yet Lerner finds that he himself blames members of his family for actions of which he disapproves. His explanation: “I want to, must, believe that people have ‘effective’ control over important things that happen, and I will hang on to this belief, even when it requires that I resort to rather primitive, magical thinking.”^[484] A few moments’ reflection will reveal that to abandon this “primitive, magical thinking” is to abandon all the reactive attitudes such as gratitude, resentment, forgiveness, and indignation - that is, to abandon the cement of the social universe.^[485] That is one reason Moore calls the argument a reductio ad absurdum.^[486] Before accepting its drastic conclusion, we should explore the possibility that praising and blaming do not require primitive, magical thinking, even in a deterministic universe.

Moore’s preferred alternative is to insist that we are morally responsible for our choices, whether or not determinism is true. To avoid the counter­intuitive argument that his position blames people for acting even though they could not do otherwise, he adopts G.E. Moore’s analysis of the phrase “could have done otherwise”: it means “could have if the actor had chosen to do otherwise.”^[487] According to this analysis, even an actor whose behavior is determined could have done otherwise, as long as the causal laws that link choosing with doing remain valid. For then, the actor could have done otherwise if only he had chosen to.

I cannot accept this alternative, however, because it falls prey to the well- known objection that “He could have done otherwise if he had chosen to do otherwise” can be true even of someone who could not have chosen to do otherwise. As Susan Wolf illustrates the objection, “the fact that a person attacked on the street would have screamed if she had chosen cannot possibly support a positive evaluation of her responsibility in the case if she was too paralyzed by fear to consider, much less choose, whether to scream.”^[488] Indeed, “she could have screamed if she had chosen to” may be true (in a hypothetical sort of way) even if the victim had fainted - hardly a condition under which it is reasonable to insist that she could have screamed!

Wolf suggests a better characterization of the ability to do something, namely that one possesses the necessary skills, talents, and knowledge to do it, and nothing interferes with their exercise.^[489] And indeed, this may come close to another of Michael Moore’s ideas, namely that one can do something if one has the capability and opportunity to do it.^[490] If this characterization of freedom is right, atom-by-atom physical determinism seems pretty much beside the point.⁵⁷

A worry nevertheless arises about whether Wolf’s alternative will help us understand the Milgram experiment or similar cases where psychological forces distort our judgment at the unconscious level. Moore rightly maintains that “the freedom essential to responsibility is the freedom to reason practically without the kind of disturbances true [psychological] compulsions represent”;^[491] and Wolf likewise insists that “agents not be psychologically determined to make the particular choices or perform the particular actions they do.”⁵⁹ But what if we aren’t free in that way? In that case, even Wolf’s definition of ability will lead to the conclusion that Milgram’s compliers were unable to act differently.

However, the numbers in the Milgram experiments suggest that such distortion does not rise to the level of determination. If every last one of Milgram’s thousand subjects had complied with the experimenter, we would undoubtedly conclude that some powerful psychological force, as irresistible as the brain tumor in our earlier example, compels our obedient behavior and excuses otherwise wrongful compliance. Our only remaining puzzles would be isolating and identifying the force, and explaining why naive observers don’t predict the result. Furthermore, if only one or two of the thousand subjects complied, we would likewise suspect that pathology had something to do with it, precisely because the experimenter’s orders prove so easy for normal people to resist.

In the actual experiment, the numbers fall in between. The two-thirds compliance rate provides strong evidence that some previously unsuspected psychological force distorts the judgment of otherwise normal people. But, because a third of the subjects did not comply, the evidence hardly supports the hypothesis of an irresistible compulsion.

The corruption-of-judgment theory I have defended here grounds the urge to comply in cognitive dissonance, a dynamic that all people share. But it links the subjects’ susceptibility to the urge to excessive self-regard, which two-thirds of us (apparently) have despite our conscious beliefs to the con­trary, and the rest do not. This difference, no doubt, results from differences in how we are put together and brought up.

What makes the warranted-excuse theory distinctive is its insistence that such differences are not morally neutral brute facts about us that excuse bad judgment. When distorted moral judgment arises from bad values like excessive self-regard, it seems wrongheaded to release the actor from blame for his actions.

Notice that on this approach, we blame people only for their chosen actions, not for their characters; the warranted-excuses approach should not be confused with the theory that actions are wrong only because they man­ifest bad character. Michael Moore criticizes this “character theory” because it implies that people of bad character deserve to be punished even if they do nothing blameworthy.^[492] It is important to understand why his objection does not apply to the warranted-excuses approach. On our approach, the ground for criticizing Milgram compliers is not that they have bad characters. It is that they knowingly administered lethal electrical shocks to an innocent person pleading with them to stop. The character trait that renders them susceptible to authoritarian pressure is a moral fault, but that fact functions only to rob them of an excuse, not to explain why shocking the learner is wrong.

The warranted-excuses approach does share features with the character theory. First, as Moore observes, it recognizes two very different sorts of moral judgments we make about persons - judgments that they are blame­worthy because of their wrongful actions, and judgments that their characters are bad.⁶¹ Second, it accepts Moore’s point that the latter judgments are “a kind of aesthetic morality”⁶² which in effect judges people by how well formed their souls are. But, unlike Moore, it rejects any implication that “aesthetic morality” is illegitimate. What should we judge people (as dis­tinguished from their actions) by except the content of their characters?

Moore confuses matters when he marks the distinction between the two sorts of moral judgments by describing them as judgments holding people responsible for their chosen actions and judgments holding them responsible “for being the sort of people that they are.”⁶³ To be sure, this description makes character-based moral judgment sound irrational, because holding people responsible for being who they are sounds irrational.

Both kinds of moral judgments are legitimate, and the warranted-excuse approach utilizes both. It assigns blame by judging actions, and accepts or rejects excuses by judging character. This procedure is fair, because it grounds blameworthiness solely in what we do, and withholds deterministic excuses only when the bad acts result from judgment corrupted by bad character. Deterministic excuses remain available whenever our judgment is corrupted by forces beyond our moral selves - forces outside of us in the way that bad character in not outside of us. But doubts surely remain, unless there is something we can do to guard against corrupted judgment and wrongful obedience.