The Compatibilist's test to determine responsibility

[reposted December 12, 1998]
Part II: desires vs. disabilities

Using the Compatibilist’s subjunctive
test sentence to determine responsibility

Part I: Intentional acts and acts of negligence

S
ome philosophers seek to argue that Ralph’s having freely done X (or, Ralph’s being morally or legally responsible for having done X) is compatible with Ralph’s doing of X having been caused by factors over which Ralph had no control. For, to say that Ralph did X freely (or, that Ralph was responsible for his performing X) is to say that Ralph could have done otherwise (than X). And to say that Ralph could have done otherwise (than X) means that   The Compatibilist supposes that when a person acts freely, his choice will “determine” his actions; for had he chosen differently, he would have acted differently. Because of this, the person’s actions follow upon his choice, not upon someone else’s (and against his choice), nor yet independently of his choice.

  On the other hand, the Compatibilist supposes that had the external causes been different, the person would not have chosen as he did. Thus, the external causes determine the individual’s action, but do so in such a way that the individual’s choice, unconstrained by coercion or by ignorance, is involved in the mechanism by which the external causes bring about the action.

  So, when the Compatibilist uses his test for freely acting, “Had he chosen differently, he would have acted differently,” he relies on the fact that the sentence will often be true even though the individual did not choose differently; the test sentence is “counterfactual.” The test sentence will be false on those occasions when the act was not a function of the individual’s own choice but was a function of something else. Where an individual cannot accurately anticipate the consequences of an action, those consequences will not be a function of his choice. We might think of one’s choices as something like the movements of a joy-stick: while its motions are normally followed by predictable consequences, yet, if circumstances change the rules (the complicated algorithm) by which the movements of the joy-stick affect the consequent happenings, and the actor has no warning about what changes have taken place, then the subsequent happenings cannot be attributed to the movements he makes with the joy-stick.

  Thus, if Ralph chooses to flip the light-switch in order to illuminate the room, but—as he could not have anticipated—flipping the switch instead sends a lethal dose of electricity through Sam, then the result cannot be attributed to Ralph’s choice; his choice was not to electrocute Sam, but to illuminate the room. It is, of course, true that he did choose to flip the switch, supposing reasonably that this would lead to the illumination of a room, not to the death of a person. But in a world in which deadly functions are arbitrarily assigned to innocent activities such as flipping a light switch, the actor has no way to avoid causing harm, since he has no way of knowing whether his moving the chair or opening the blinds or his failing to move at all within a twenty-second period of time might cause someone to be electrocuted.

  In using the Compatibilist’s test sentence, we need to be careful how we fill-in for “X,” since the actor can be said to have “done” a host of actions. It is clear that when he flips the light-switch in order to illuminate the room, he does freely flip the switch; this fits with the Compatibilist’s use of the test sentence, and it is not counter-intuitive, since we will grant that Ralph was responsible for flipping the switch. The question of responsibility and blame would center on Ralph’s causing the death of Sam by flipping the switch; so we want to put this into the slot occupied by “X.” The proper substitution yields: “had Ralph wished (or willed, or chosen) to do otherwise (than electrocute Sam), Ralph would have refrained from performing the act which electrocuted Sam.” But notice, since Ralph did not choose to cause the death of Sam, the antecedent is true; but because he nevertheless did cause Sam’s death by electrocution, the consequent is false. So the test sentence is false, and therefore, given the Compatibilist’s supposition that Ralph was responsible if and only if the test sentence is true, the test does yield the result that Ralph was not responsible for causing Sam’s death by electrocution.

  Consider the example where Wilbur is driving in his convertible and is attacked by bees. In order to see how the Compatibilist’s hypothetical sentence applies, we must pick a candidate for “X;” for what action are we inclined to excuse him? Perhaps “X” is failing to stop and thereby causing an accident. Yet here, Wilbur did not choose to fail to stop; because the painful attack of the bees monopolized his attention, his choosing to stop would not have resulting in his stopping. (I can imagine someone’s arguing that he made a choice to fend off the bees—at any rate, he was trying to fend them off; and so we might continue, his choosing to fend off the bees resulted in his failure to notice the stop sign and his failure to stop and so forth. So we will try this: is it true that had he chosen not to fend off the bees, he would have avoided the accident? It seems artificial to speak of this as a choice, but perhaps not much more artificial than a clerk’s choice to surrender money to a gunman. In any case, the test would seem to turn out in the same way: no matter what he had chosen to do, the same result would have ensued since he would have been unable to concentrate sufficiently to carry out either a choice not to fend off the bees or a choice to attend fully to the car.)

  We can, by the way, imagine a person with the training, or perhaps with a powerful motive, to resist the bees’ urgent claim on his attention. Imagine that a soldier is trying to escape through the lines of an enemy which will surely butcher him if it finds him; he constantly reminds himself of the absolute necessity of his remaining silent and still no matter what happens. In this state of mind, he steps into a nest of bees which attack him furiously. Here, it seems, the soldier has a choice and exercises it; he is provided with a choice by having prepared himself for just such an eventuality; here, the outcome does depend upon his choice—if he has the ability to suppress his instinctive reactions. (You might recall that Martin Luther King’s passive resisters went through extensive training to make them capable of suppressing their normal reactions to the provocations which were anticipated as a result of their planned non-violent “sit-ins.” In effect, they were trained to suppress their normal reactions to the attacks of stinging bees.) But with no such warning and mental preparation, Wilbur lacked the ability to ignore the bees’ attack. I take it that Wilbur might have tried to concentrate on slowing down and stopping the car, but would have been unsuccessful in this effort. If he tried, then I suppose we can say that he chose; but the events occurred despite his choice, so the hypothetical is false in the case of Wilbur and the bees.

  Compare the example of Wilbur and the bees to a case of negligence. Suppose that Elmer is driving past a nudist beach and is too intent upon the scenery to notice the stop sign. Here Elmer does not, of course, choose to run the sign. But he does choose to pay inordinate attention to the scenery, its deleterious consequences upon his attending to his driving being foreseeable. Again, when Boardman negligently pushes the bale of accumulated memos (or, in the standard legal example, a barrel of flour) out of his third-floor window without having roped off the area below or even having looked below and sounded an alarm, he does not choose to injure the passerby. But he does choose to eject the heavy object in circumstances where doing so foreseeably creates a substantial risk of injury to a passerby. In these cases of negligence, had either Elmer or Boardman chosen to reduce the risk to a negligible amount, he would have done so; thus, on the Compatibilist’s test, each is responsible for creating a substantial risk of injury to other persons.

  I have used some examples with which I think the Compatibilist can plausibly deal. It is worth remembering that examples of a kleptomaniac or a person with an elevator phobia might be harder for the Compatibilist. It seems that the person with a phobia can’t really choose to get into the elevator (though he can choose to try—forcing himself to try to take some preliminary steps towards the elevator). Here it seems plausible to say that if he had chosen to get into the elevator, he would have performed whatever duty required the elevator ride (saving his young child, e.g.), but that he was unable to make this choice. But might it perhaps be like the case of the bees—that given his phobia, no choice would have led to rescuing his child by means of an elevator ride? If that works, the example could be accommodated by the Compatibilist’s test sentence. In any case, the issue is whether the individual’s choices were deprived of their typical effects or whether he was unable to choose.

  Ayer and others use constraint as their paradigm of lack of freedom. So, how does coercion work? It gets you to make a choice which you would not have made in normal circumstances. Suppose a friend needs aid, but someone threatens you with violence or penury if you lend help. Imagine a writer with leftist beliefs who is “blacklisted” in the McCarthy era; a director wants to employ him, but is threatened with being blacklisted himself if he does so. No doubt the director will be a hero if he ignores the threats; but suppose he succumbs—as many in the McCarthy era did. And later his friends and self-appointed judges charge him with disloyalty to a friend. He points to the coercion as his excuse. But how is the excuse supposed to function? Perhaps like this: given the external threat, it would have been unreasonable for the director to aid his friend. But this seems to imply that had he chosen to help his friend, he would have done so; he does not seem to have been deprived of the ability to alter the course of events—to help his friend (as was the case in the bizarre situation where the light switch is wired in a nonstandard way). It looks as though we should say that the director did make a rational choice: his defense would seem to be that, under the circumstances, he did nothing wrong, nothing blameworthy. Loyalty to a friend cannot be expected to trump one’s duties to one’s family or one’s need to earn a living in the only way one knows.

  (Of course, there are cases and there are cases. We might imagine a lesser threat; in that case, the excuse might be only partial. It is hard though not completely unreasonable to expect a person to do his duty in the face of large though not devastating personal sacrifice: in such an event, he is responsible, though with mitigating factors. The mitigation comes from our recognizing that although his choosing differently would indeed have resulted in a different outcome, the choice would have been hard, linked as it was to large personal loss. We want people to be selfless to some degree, but since few are, we think it harsh (not fair) to hold them to standards which most would not measure up to. Here, then, the mitigation comes in at the reasonableness of the standards, not in the ability of the person to affect changes through his choices.)

  On the other hand, under some threats (for many of us, confrontation with a gunman would suffice), one may simply go to pieces. In such a case, it is not that one’s choices, unfortunate as they are, are rational. Rather, one is here deprived of rational choice; perhaps this is comparable to the example of the bees. No matter what one chooses in this situation, he will not have the ability to resist the threat and carry out his choice on account of his paralyzing fear.

  A problem which philosophers (certainly including me) and perhaps also lawyers have with many of these cases, e.g., kleptomania, is that it isn't obvious how they should be described: very different alternative descriptions suggest themselves. Yet, an accurate, cannonical description would seem essential for us to understand how an example should be treated in our theories. In the example of the driver attacked by bees, perhaps we should say that the person is deprived of the ability to attend to his driving. But then he would be understood to be in a situation something like one where the person does not know that the time for doing something (e.g., for breaking to slow down to a stop) has arrived. Maybe it’s like an excuse of ignorance (the light-switch case), his choices not having their ordinary consequences. In a number of the examples, e.g., that of a phobia, the worry would seem to be that the individual is not able to choose: even if he would have done what we think he ought to have done had he chosen otherwise than he did, if he did not have the ability to choose, our attending what he would have done had he possessed that ability seems moot—and perhaps perverse. (This concern is pursued in the sequel.)

Part II: Desires vs. Disabilities

F
irst suppose that we have revised the Compatibilist’s conditional in the way Davidson thinks we should, by removing any intentional verb of action from the if-clause; otherwise we descend into a vicious regress.1 Imagine, then, that the Compatibilist claims that a person is responsible, that he acts freely, that he could have done otherwise, when this conditional is true:

Now consider a new example: When a child drowns within sight of Rebecca, she is said not to be responsible because she had previously been handcuffed to the cabaña; we will suppose that Rebecca wanted to help the child, but was prevented from doing so by the handcuffs. Continuing the example, because the child was also within sight of Boardman, Boardman is said to be responsible since he had not been handcuffed to the martini which he was enjoying too much to be willing to bother with a mere child. According to Davidson, the individual’s failure to rescue the drowning child is in each case caused by prior events; how, then, can it make sense to hold Boardman responsible for the drowning but not to hold Rebecca responsible? Why shouldn’t the excuse, “I couldn’t help it,” be equally available to both?

  We might further elaborate on the example: imagine that, throughout his life, Boardman has been inveterately and unwaveringly selfish—that he resembles the “before” portrait of Scrooge. Now the Compatibilist says that Boardman cannot intentionally change his own set of desires nor their relative strengths; indeed, Davidson agrees in Ch. 4 that one’s pro-attitudes are not intentional actions, and that this is so is an important part of his argument. Thus, if Boardman is to act otherwise, some external cause (e.g., a visitation of Christmas ghosts as in The Christmas Carol, by Chas. Dickens) will be required. Given that no such external cause has by now changed Boardman’s pattern of desires, how can it make sense to hold him responsible (rather than, e.g., his parents who, by spoiling him at an early age, caused him to desire his own personal gain so absolutely)? (Of course, you see where that will go. Boardman’s parents aren’t stupid: they will avail themselves of the same excuse, and so will their parents, and so on back—to the monkeys or, if you prefer, to Adam and Eve.)

  The comedian, Flip Wilson, used to invoke the excuse, “The Devil made me do it.” For our example, suppose that although the Devil did not make Boardman continue to relish his martini, the Devil did cause Boardman to want to relish his martini; perhaps the Devil described martini drinking in such seductively attractive terms that Boardman was caused to want to engage in such conduct more than he wanted to do anything else. Since a person’s desires are not subject to his command, how is the excuse Boardman offers, “The Devil caused me to want to do it,” less of an excuse than Wilson’s “The Devil made me do it.”? If we reject the satanic excuse which Boardman offers, are we not holding Boardman responsible on account of a fact over which he has no control—namely, that he happened to have selfish, martini-loving desires rather than altruistic ones?

  Of course, had Boardman previously been worried about this character trait of favoring his selfish desires, he might have signed up to serve with some altruistic outfit such as Doctors-without-borders, hoping that the experience would change him in fundamental ways. But, first, he would have to have desired that his characteristic desires become different (which, of course, would raise the problem all over again; he must already have had the desire to change his character, and where did that come from?); and, secondly, he cannot intentionally have created or changed those desires, but can at most can have put himself into a situation where desire-changing causes might be hoped to do their work. But, anyway, let us suppose that he has tried to appreciate the “superior pleasures of helping others,” but that he finds, perhaps to his chagrin, that he just hasn’t come really to appreciate them; self-centered activity, such as drinking martinis, simply remains more attractive to him. So, although he has tried in the only (indirect) way available to change himself, he now finds himself sipping a martini while a child is drowning (and while Rebecca is imploring him to help, for God’s sake), and finds himself wanting to continue sipping rather than to go for a nasty swim. How does it make sense to hold him responsible, to blame him for his failure to save the child? In our story, Boardman seems a plaything of his desires and therefore of the prior events which caused those desires to exist. Must we simply accept it as a primitive axiom of morality that a person is blameworthy when bad things are caused by his not having possessed stronger social desires? These are, of course, his desires, but so too is his instinctive desire to draw breath; yet if his drawing breath in some set of circumstances led to foreseeably tragic consequences, would we hold him responsible for the consequences of being motivated by his own desire?

  A little earlier, when I said that “a person’s desires are not subject to his command,” I meant that a person can not intentionally make it the case that he desires something or doesn’t desire it. But we do speak of one’s controlling his desires; on the sort of account offered by the Compatibilist, when we intentionally control our desire to do one thing, we exercise that control as the result of our having some other, stronger desire to which the first is opposed. So, if we upbraid Boardman for not having thwarted his desire to sip the martini, he will reply that unfortunately he did not happen to have a stronger desire which was opposed to it, more’s the pity.

  Since one cannot make it to be the case that he does or does not desire something, then why should it be important to morality whether Boardman was caused not to save the child by his having been chained to a cabaña or by his having desired to do something else in preference to saving the child? What is the sense behind treating him differently when his action is caused by a desire which he has been caused to have than when his action is caused by a disability which he has been caused to have? Are we not really succumbing to latent superstition when we treat his desire differently from his disability? For in either case, he is simply acting in accordance with the causal elements which have been created in him from the outside. Although the set of desires existing in a person might be regretted as tragic and as potentially dangerous to others, how can it make sense to hold the person responsible for the effects of causes which, given their existence, he could not have altered?

  It is tempting to answer, and some Compatibilists who have tried to work out the implications of their position sufficiently far (as Schlick does) have answered, following in Bentham’s utilitarian footsteps: blame and punishment are simply society’s tools for shaping the behavior of persons. Hart, you will remember, has some pretty devastating things to say about Bentham’s views in his essays in Punishment and Responsibility: punishment and blame2 might well be efficient tools for shaping behavior when applied to cases where a person could not have avoided doing what he did; even in the state of Wisconsin, crimes of strict liability—for example, “felony murder”—are thought to be useful tools to change the behavior of potential miscreants. The typical Anglo-American legal requirement of mens rea seems to be justified solely on the grounds of fairness, not efficiency.

  The difficulties of the Compatibilist’s view seems to stem from its Humean theory of the soul in which a person does not brook one desire unless he is motivated to do so by a second desire. But now avert your gaze from this Humean view of the soul to a Rationalist view—the sort of framework evident in Chisholm’s article and perhaps in Malcolm’s: the person who is responsible for his action is now seen as having the power to curb his desires; and an exercise of this power need not be caused by some other, stronger desire. Rather, it is supposed that the person might simply recognize that he ought not to succumb to some urgent desire in these circumstances, and as a result of so recognizing, not succumb. The Rationalist view of the soul is not itself free of difficulties, of course: explaining the mechanism by which someone comes to exercise such a power is a substantial problem, as it is explaining why a person might choose to exercise the power in a given set of circumstances. But because a person’s allowing himself to act upon a given desire—e.g., the desire to remain on shore sipping his martini—is viewed as subject to the person’s executive control, it makes sense, on the Rationalist view, to hold a person such as Boardman responsible for failing to save a drowning child. For on this view, the claim that he could (and should) have done otherwise is not seen as a matter of what he would have done had his desires been otherwise than they were. It is seen as a matter of whether he possessed control over his desires as they existed on that occasion.

Footnotes:

Footnote 1  Suppose that it is true of some individual in some set of circumstances that if he had chosen to swim out to rescue the drowning child, he would have rescued the drowning child; then, by the usual version of the conditional (which uses as its verb, “chooses,” or “tries”), he is responsible for letting the child drown if he fails to swim out to save the child. Nevertheless, it is possible that because of some earlier traumatic episode involving his swimming (when he was young, a practical joker put piranhas into the backyard swimming pool, leading to horrific results), the individual could no longer bring himself to choose to swim, no matter what horrible consequences seemed to be in the offing; then although the conditional analysis would seem to say that the individual could have rescued the child, we surely want to say that he was not in fact free to do so, because he was not free to choose or try. The problem here is that when the if-clause uses an intentional verb, one can always go on to ask whether the individual was free to perform that; and if he was not, then he was not free after all to do otherwise. Thus, if we allow intentional verbs in the if-clause, then each time we say he was free to perform that action, the analysis of this will itself use an intentional verb, and the same question will arise regarding the performance reported by that verb: hence, a vicious regress ensues. The way out is to employ a verb which is not an intentional verb: that forestalls the regress. This is what recommends the verb, “desires;” for it does not seem to be intelligible to say that a person either could or could not have desired to do such-and-such a thing.
  Notice the presumed result of applying the conditional analysis: his upbringing and other past experiences have caused Boardman to desire not to save the drowning child, since he believed that doing so would interfere with his enjoyment of the martini he was sipping at the time; nonetheless, it is true that if Boardman had desired to save the child (i.e., if he had been motivated by a different set of desires than the set which actually did motivate him), he would have done so (being an expert swimmer and still sober). Thus, the Compatibilist is able to argue that Boardman acted freely when he refrained from saving the child, and so, can be held responsible, even though he did what he did as a result of external causes (those leading to his desiring not to bother with the child).

Footnote 2  Note that how one understands “punishment” and “blame” must also alter if Bentham’s view be accepted. Recall that Lady Wooton prefers to speak of “accountability” rather than of “responsibility.”