A Metaskeptical Analysis of Libertarianism and Compatibilism
A Metaskeptical Analysis of Libertarianism and Compatibilism
Abstract and Keywords
This chapter outlines the author's arguments for versions of libertarianism and compatibilism that are consistent with the broader metaskeptical thesis. It considers the “all-things-considered” cases for both compatibilism and libertarianism for members of his “intuition group”—people with starting intuitions and attitudes about responsibility that resemble his own. He argues that libertarianism is an implausible alternative because its epistemic costs outweigh its alleged pragmatic and moral benefits. The more promising case for compatibilism relies on questionable assumptions. He continues to challenge those assumptions in the final chapter as he develops the case for eliminativism about moral responsibility.
From the first time I encountered the problem of free will in college, it struck me that a clear-eyed view of free will and moral responsibility demanded some form of nihilism. Libertarianism seemed delusional, and compatibilism seemed in bad faith. Hence I threw my lot in with philosophers like Paul d’Holbach, Galen Strawson, and Derk Pereboom who conclude that no one is truly morally responsible. But after two decades of self-identifying as a nihilist, it occurred to me that I had continued to treat my friends, colleagues, and acquaintances as morally responsible. Hardly ever did I call on my philosophical views to excuse people’s actions.
Like Shaun Nichols, I have always found the arguments for nihilism or skepticism compelling. The TNR principle, which is at the heart of these arguments, seems to me to reflect a basic principle of fairness. Yet like Nichols I hold myself morally responsible all the time. I feel that I deserve blame for my bad actions and praise for good actions and accomplishments. I (p.134) also commonly hold other people morally responsible—my family, friends, criminals, and (perhaps especially) athletes who are connected in any way to Boston sports.1 At the gut level, then, I find the belief that people can be morally responsible (PMR henceforth) to be intuitively compelling as well. Of course, the TNR and PMR beliefs are not inconsistent on their own. The inconsistency arises only for those who agree with Nichols that libertarianism is “delusional” or at any rate false. Indeed, the appeal of libertarianism is precisely that it allows us to retain two intuitively compelling beliefs, one about the conditions of moral responsibility (TNR) and the other that people are capable of being morally responsible (PMR).
So metaskeptics have at least three options when trying to determine the most reasonable all-things-considered position about responsibility. Compatibilism allows us to regard ourselves and others as morally responsible agents, though at the cost of rejecting the highly intuitive TNR principle. Nihilism allows us to retain TNR, but at the cost of denying the highly intuitive belief that people can be morally responsible. Finally, there is libertarianism, which allows us to accept both TNR and PMR, but at the price of some thorny empirical and theoretical problems that will be discussed below. In this chapter I will evaluate the cases for libertarianism and compatibilism for people who share my core starting intuitions about the conditions of moral responsibility. In the next chapter, I will consider the case for nihilism or eliminativism. The goal of these final two chapters is not to arrive at a decisive conclusion (a good thing because I do not) but rather to lay out and apply the method metaskeptics can employ for arriving at all-things-considered responsibility judgments. Readers may apply the same method and arrive at a different verdict. Indeed, the conclusions generated by this method will almost certainly vary from culture to (p.135) culture and possibly from person to person. Nevertheless, the method itself is something that can be adopted universally as a means of identifying the most reasonable position for people in each “intuition group” to embrace.
The Case for Libertarianism
The criticisms that have been directed at libertarianism are well known and compelling. The event-causal varieties (e.g., Kane 1996 or Ekstrom 2000), have little if anything in the way of empirical support (Vargas 2007). Agent-causal varieties rely on metaphysical views that are incoherent at worst, empirically dubious at best. More damaging to the view is that even when one grants libertarians their metaphysics, it remains unclear how their accounts can provide agents with the right kind of control for moral responsibility. For these reasons, like Nichols, I have never found libertarianism to be a serious contender in the free will and moral responsibility debate. Daniel Speak, however, has identified a controversial assumption that lurks behind these criticisms. He writes: “Opponents of libertarianism make their arguments on the assumption that moral and pragmatic considerations about the value of libertarian agency are strictly irrelevant to its philosophical assessment” (Speak 2004, 353). And he goes on to provide compelling reasons to doubt this assumption, concluding that “the case for libertarianism may very well stand or fall with an overall assessment of its importance to human life (moral and otherwise)” (353). Pointing to the recent “pragmatic encroachment” (Jonathan Kvanvig’s term) into issues of epistemic justification in general, Speak argues for an axiological approach to free will and moral responsibility. His methodological conclusions are similar to my own as described in chapter 5: We should try to arrive at “all-things-considered” (p.136) judgments about the condition for free will and moral responsibility, judgments that incorporate values as well as facts. The pragmatic or moral reasons for accepting libertarianism may be significant enough to outweigh evidentiary or theoretical factors that count against the position.
I agree with Speak that this outcome cannot be dismissed out of hand.2 With this in mind, let us examine some considerations that seem to favor libertarian theories of free will and moral responsibility.
The Phenomenological Fit
One virtue of libertarianism is that it seems to describe the phenomenology of agency and responsibility more effectively than competing positions. As Galen Strawson (a skeptic) writes:
We are, in the most ordinary situations of choice, unable not to think that we will be truly or absolutely responsible for our choice, whatever we choose. Our natural thought may be expressed as follows: even if my character is indeed just something given (a product of heredity and environment or whatever), I am still able to choose (and hence act) completely freely and truly responsibly, given how I now am and what I now know; this is so whatever else is the case—determinism or no determinism.
(Strawson 1986, 116)
Strawson gives a vivid description of this phenomenology in the following example: Imagine, he writes, that it is Christmas Eve and you are going to buy a bottle of scotch with your last twenty-dollar bill. Right outside the liquor store, there is a beggar who is clearly in need. You must decide now whether to spend your money on the scotch, or give it to the beggar.
(p.137) You stop, and it seems quite clear to you — it surely is quite clear to you — that it is entirely up to you what you do next — in such a way that you will have deep moral responsibility for what you do, whatever you do. The situation is in fact utterly clear: you can put the money into the beggar’s tin or you can go in and buy the scotch. You’re not only completely, radically free to choose in this situation. You’re not free not to choose. That’s how it feels.
(1986, x [preface])
Strawson is right, I believe, that libertarianism best captures the subjective experience of moral deliberation, a point that must count in its favor. Some libertarians have made the much stronger claim that belief in libertarian free will is required in order for us to believe that we engage in rational deliberation. According to this view, when we deliberate we must assume (a) that there are two possible outcomes, and (b) that we have the power to bring about either one of them. As Van Inwagen (1983, 155) writes:
If someone deliberates about whether to do A or to do B, it follows that his behavior manifests a belief that it is possible for him to do A—that he can do A, that he has it within his power to do A—and a belief that it is possible for him to do B.
If this view is correct, the cost of denying libertarianism would indeed be enormous. In order to consistently deny libertarian free will I would have to deny, in addition, that I deliberate—and the belief that I deliberate is close to non-negotiable. But Van Inwagen is mistaken on this point. For while it does feel as though there are two metaphysically open alternatives when I deliberate, it does not feel impossible that I am determined to (p.138) deliberate in a particular way and to arrive at a particular conclusion. Indeed, this does not even seem unlikely. Furthermore, the possibility that my deliberation is determined does not make it seem pointless, since the deliberation would still play a causal role in determining the outcome. To think otherwise confuses determinism with fatalism.
So what precisely do we need to believe in order to engage in rational deliberation? Pereboom (2008) lays out two conditions both of which are consistent with the truth of determinism and thus with the denial of libertarian free will. First, there must be epistemic openness: that is, we must not know the outcome of our deliberation in advance. Second, we must believe that the deliberation is, under normal conditions, efficacious—that the final choice will be causally influenced by the deliberation process. Pereboom’s argument for the sufficiency of these conditions is in my view decisive. Here is an example from my own past (and readers may think back to some episode in their own lives that required deliberation): About ten years ago, I faced a decision about whether to attend a graduate program in philosophy or to continue my so-called career as a freelance writer. At the time, I had no idea which choice I would end up making. I also believed (correctly) that my deliberations would have causal influence over the final verdict, whatever that turned out to be. So the act of deliberating made perfect sense. Yet even during the deliberation process, it did not seem implausible that the outcome was inevitable. And after the decision was made, as I was reflecting back on the process, it seemed to me rather likely that the deliberation occurred in a deterministic fashion. There is a feeling of metaphysical openness, but it is mostly confined to the present moment. At no point in the process did the possibility of determinism or the lack of libertarian free will undermine (p.139) my belief that the deliberation was rational, genuine, and effective.
Another factor to be considered is the connection between libertarianism and what Kane (1996) calls “objective worth.” If objective worth is important to us, and it is incompatible with non-libertarian accounts of free will and responsibility, then we would have a powerful pragmatic reason to accept libertarianism, all things considered. So in this section I try to shed light on three questions:
1. What is objective worth?
2. Is objective worth so defined valuable or important?
3. Is objective worth only compatible with libertarian conceptions of free will and responsibility?
(1) What is Objective Worth?
Kane illustrates the notion of objective worth with the story of Alan the artist. Alan is depressed because no one appreciates his paintings. To cheer him up, a rich friend of Alan’s arranges to have his paintings purchased at an art gallery for $10,000 apiece. Alan assumes that his work is finally receiving the critical and popular recognition he feels it deserves, and so he feels much better. Kane then asks us to image two worlds. In the first, Alan is mistaken about the merit of his paintings—his rich friend has tricked him into thinking he is a great artist and that he is being recognized as one. In the second, Alan’s beliefs about his talent are correct. The rich friend has not deceived him to lift his spirits, he merely provided the necessary impetus for others to appreciate him. In both worlds, “Alan dies happily believing he is (p.140) a great artist, though only in the second was his belief correct” (Kane 1996, 97).3
The key question, then, is whether it would make any difference to Alan which of these worlds he belonged to. According to Kane, if we agree that there is an important difference in value for Alan in the two worlds, then we must endorse a notion of objective worth.
Since Alan’s subjective experiences would be identical in both worlds, any difference in their value must consist in something over and above the feelings of happiness he experiences from the belief that he is a great artist. There is some value, then, in Alan’s belief being true and his feeling of happiness being warranted.
(2) Is Objective Worth Valuable?
Compatibilists and skeptics about responsibility would likely agree that this type of objective worth is valuable. As Kane writes, there is something demeaning about being praised for works that are in reality worthless—even if we never know that we were being deceived. The controversy, then, focuses on the answer to the final question:
(3) Is Objective Worth Only Compatible with Libertarian Conceptions of Free Will and Mr?
According to Kane, libertarian free will, the capacity to be ultimately responsible for our characters, is “of a piece” with this notion of objective worth:
If, like Alan, we think that the objective worth of our acts or accomplishments is something valuable over and above the felt satisfaction the acts have or bring, then I suggest that we will be inclined to think that a freedom requiring (p.141) ultimate responsibility is valuable over and above compatibilist freedoms from coercion, compulsion, and oppression. … Such freedoms would be enough, if we did not care in addition, like Alan, about more than what pleases us—namely, if we did not care in addition about our “worthiness” or “deservingness” to be pleased. Are our acts or deeds, our accomplishments and we ourselves, objectively worthy of recognition, respect, love, or praise, as Alan wished for his paintings or as he might wish for himself? It is this concern for the objective desert of our deeds and characters that leads naturally to a concern about whether those deeds and characters have their ultimate sources in us rather than in someone or something else (fate or God, genes or environment, social conditioning or behavioral controllers). Do they ultimately redound to us or not? The answer to this question does not have to do merely with subjective happiness, but with the objective worth of our selves and our lives. (97–98)
Here Kane attempts to draw an analogy between Alan’s paintings and our actions and character. Just as Alan wants his paintings to be worthy, we want our actions and characters, our selves, to be truly worthy of respect, recognition, and praise.
Even when we grant Kane our desire for worthiness, however, there still seems to be a gap in how that is connected with libertarian free will. Mele (1999) points out correctly that compatibilists may have an equally strong preference to live in an objectively worthy world. About Kane’s example of Alan, he writes:
A compatibilist can reasonably prefer actually being a great artist and having the associated feelings and experiences of success to having the same feelings and experiences (p.142) while falsely believing himself to be a great artist. He can prefer that these qualitative states are grounded in his excellent work and that he not be radically deceived about himself. (103)
Indeed, in Kane’s example of the objectively worthy world, nothing required that Alan be the ultimate source of his talent and abilities as a painter. This world only required that Alan’s paintings actually be of high quality. Kane’s remark that concern for objective worth leads us “naturally” to a concern about “whether those deeds and characters have their ultimate sources in us rather than in someone or something else” seems unmotivated by his own example. Alan does not indicate any concern about whether his talents and abilities have their ultimate source in him. Why should this be such a concern for us?
Mele’s objection can be taken further. Objective worth does not even seem dependent on a compatibilist notion of responsibility. Consider the case of someone who is concerned with the objective worth of his ability (like Alan), but who does not presume himself to be responsible for that ability in any way:
Joe the naturally funny guy. Joe has the talent to make people around him laugh heartily. He is able to find the humor in a conversation or situation as it’s happening, and the right words come into his head quickly enough for him to articulate them. Irreverence, wit, and originality come naturally to Joe. He takes pride in this aspect of his character—it is an appealing quality, part of the reason he is well liked by his friends and family. Nevertheless, it is not an aspect of Joe’s character for which he considers himself responsible in any way. He does not know what combination of causes produced his sense of humor—family, genes, upbringing—but he is certain that the source is not him in the libertarian sense. What is more, Joe has not made a (p.143) higher-level endorsement of a decision to be funny, he has not responded to reasons in favor of making jokes. He has not studied or worked hard to cultivate his wit. It requires little or no effort whatsoever. Joe, in other words, does not consider himself responsible for this aspect of his character in either the compatibilist or the libertarian sense. He sees it as a natural trait that he is lucky to have, like being physically attractive or having a photographic memory.
Now consider a world in which Joe has all the same beliefs about his wit, the same pride in being funny, the same sense of satisfaction every time one of his remarks provokes a round of hearty laughter. But in this world, Joe’s wealthy father has paid the people around him to laugh at his jokes. In fact, Joe’s remarks are not funny at all. They fail to capture the irony of a situation, they are often deeply offensive or too politically correct, sometimes both at once. But the actors around Joe never let on. Joe goes to his grave believing himself to be as funny as Peter Sellers in his prime.
In this case, as in Alan’s, we can agree that there is value to being in the first world, the one in which Joe is genuinely funny. There is objective worth in having a good sense of humor over and above the subjective feelings of happiness and satisfaction that come with it. But the objective worth of this quality is disconnected from the need to be responsible for the quality in any way at all, let alone the need for it to have its “ultimate source in us rather than in someone or something else.” Joe is certain that he is not responsible for the gift. But this does not affect the value he places on the quality’s authenticity. The only thing that could undermine the authenticity of his ability would be the discovery that he is in fact not funny. Objective worth, then, does not appear to be “of a piece” with ultimate responsibility or indeed responsibility of any kind.
Though not a libertarian himself, Smilansky’s (2000) “illusionism” provides the axiological or pragmatic case for libertarianism with additional ammunition. Smilansky argues that we should retain the illusion of libertarian free will due to its moral importance. Two of his worries are pragmatic in nature. First, he argues that knowing in advance that we will not be ultimately responsible for future actions would undermine our motivation to refrain from wrongdoing. Smilansky calls this the “present danger of the future retrospective excuse.” The second and related “danger of worthlessness” posits that our motivation to perform morally worthy actions will be harmed by the knowledge that our responsibility will be of the “significantly shallower compatibilist variety” (Smilansky 2000, 53).
Smilansky’s claims are based on some controversial assumptions about the nature of moral motivation, and, until recently, there was no empirical evidence to support his pessimistic claims. This changed with a study conducted by Vohs and Schooler (2008). Subjects in the study were divided into two groups. In the first condition, subjects read a passage from Francis Crick’s The Astonishing Hypothesis stating that scientists have denounced the notion of free will. In the second condition, subjects read a different excerpt from Crick’s book that did not refer to free will. The authors then gave the subjects cognitive tasks that featured opportunities to cheat, including one where they could pocket some extra money for doing so. The authors predicted that the subjects who had read the denunciation of free will would be more likely to cheat on their tasks, and the results supported the prediction. Vohs and Schooler conclude their paper by asking: “Does the belief that forces outside the self determine (p.145) behavior drain the motivation to resist the temptation to cheat, inducing a “why bother” mentality? … Or perhaps denying free will simply provide the ultimate excuse to behave as one likes” (54).4
Vohs and Schooler do not make the distinction between compatibilist and libertarian free will; rather, they assume that free will must be of the libertarian variety and their scenarios are designed to challenge the existence of libertarian free will in the minds of their subjects. Thus, their studies provide an important challenge both to “happy hard determinists” (Smilansky’s term) and to compatibilists who argue that denying libertarianism would have either beneficial or neutral implications for human beings as a whole. But is the challenge successful? I do not believe so. For one thing, our behavior just after hearing that a cherished belief is false may have little or no bearing on how we would act after further reflection.5 The authors cite no studies that document such a correlation. A religious person may experience some moral lapses just after having his faith challenged, but soon after come to realize that his values did not depend on his belief in God after all. The debate over the moral implications of denying libertarian free will focuses on long term effects. Vohs and Schooler may merely have shown that people should not lose their belief in libertarian free will fifteen minutes before they submit their tax return.
The relevant experiment would be one that tested the motivations of people who have denied libertarian free will and moral responsibility for long periods. Unfortunately, the pool of subjects for such an experiment is small, so Smilansky is limited to hypothetical examples. He raises a concern, for example, about the moral motivations of heroes facing situations of grave risk. He writes:
(p.146) Think about someone on the eve of a mission which he knows is more likely than not to cost him his life: say, being parachuted into al-Qaida territory in Afghanistan, or being sent to infiltrate the Mafia. Do we not think that such a person is affected by what he thinks others will think and feel about his or her actions? That he will be honored by his colleagues, or respected by the public, or even forever hero-worshipped by his young nephew, in a real, robust way? A robust way assumes that it was, in fact, in a strong sense up to him not to do this, and that hence he is deserving of our appreciation (not only because to do so makes good utilitarian sense), and has through his choice acquired high moral value.
(Smilansky 2010, 197)
According to Smilansky, it is crucial that these courageous heroes believe that they have done something that most people would not have done and that they themselves could have chosen not to do. And the heroes must also believe that other people believe this about them as well. (That is what will make them heroes in the eyes of the others.) If their heroism is just an “unfolding of the given,” and they and other people know this, they will lack sufficient motivation for undergoing serious risk.
In a speculative case like this, argument can take us only so far. Not surprisingly, I see no connection between the heroes’ motivations and their belief in libertarian free will. The notion that determinism could undermine their heroism strikes me as bizarre. A hero is a hero. They do things that most people cannot do. We do not find beautiful men and women any less attractive because we know that their appearance is not in a deep sense up to them. The desire to achieve a moral end is all the motivation that a moral hero needs. Consider people like Martin Luther King, Jr., or Mahatma Gandhi. Do we imagine that (p.147) they performed heroic actions so that they could deserve praise and appreciation? The most obvious motivation for battling injustice is to end injustice. There is of course room for cynicism about the real causes of moral heroism, but we can at least say this: There is no evidence whatsoever that a belief in libertarian free will and moral responsibility played even the slightest role in motivating people like King, Gandhi, or other, less celebrated heroes who risk their lives to help others. As Franco and Zimbardo (2006, 33) note, “The first response of many people who are called heroes is to deny their own uniqueness with statements such as, ‘I am not a hero; anyone in the same situation would have done what I did,’ or, ‘I just did what needed to be done.’” This could of course be false modesty on their part. But there is no reason to think that it is, nor is there any reason to think that their heroic behavior was motivated even in part by the knowledge that it would render them praiseworthy in the ultimate libertarian sense.
In addition to his pragmatic concerns about motivation, Smilansky, like Kane, argues that the real moral value of moral heroism is essentially connected to the ultimate responsibility of the hero. He writes:
True appreciation, deeply attributing matters to someone in a sense that will make him worthy, is impossible if we regard him and his efforts as merely determined products. … From the ultimate perspective, all people, whatever their efforts and sacrifices are morally equal: i.e., there cannot be any means of generating “real” moral value. … there is a sense in which our notion of moral self-respect, which is intimately connected with our view of choices, actions, and achievements, withers when we accept the ultimate perspective. From the latter any sense (p.148) of moral achievement disappears, as even the actions of the “moral hero” are simply an unfolding of what he happens to be. No matter how devoted he has been, how much effort he has put in, how many tears he has shed, how many sacrifices he has willingly suffered.
(Smilansky 2000, 163, his italics)
These claims strikes me as implausible even in the abstract—a kind of Kantianism gone wild—and when applied to concrete cases, they become even more implausible. Consider a holocaust survivor who owes her life to the courage of a family that risked being captured by the SS in order to shelter her. Imagine that she later became convinced that determinism was true. Would she no longer find real moral value in the family’s actions? Would she no longer truly appreciate them? As a matter of psychological fact, the answer is almost certainly no. And the woman’s psychology is indeed the relevant factor. Appreciation, true appreciation, is an affective response. If, after due reflection, the determinist holocaust survivor still feels appreciation, if she still finds “real value” in the actions of the family that sheltered her, then the actions have value—at least, in the context of the present discussion, which is after all focused on pragmatic or axiological concerns. The same goes for the self-respect of the members of the family. If they sense deep appreciation on the part of people they helped, if they believe that their actions served a worthy purpose, they will likely feel a deep sense of self-respect, in spite of not being the ultimate source of their actions.
The Case against Libertarianism
In the preceding discussion we examined several possible axiological or pragmatic benefits of retaining a belief in libertarian free will and moral responsibility. I argued that in each case the (p.149) benefits were not all they were cracked up to be. In this section I examine the costs of a adopting a belief in libertarianism and argue that the costs, when taken together, weigh against an all-things-considered endorsement of libertarianism.
The Difficulty of Articulation
Much of the debate in the free will literature about libertarianism concerns the intelligibility or coherence of the position. Nagel (1979), Strawson (1986), and Wolf (1990), among many others, argue that the type of free will and ultimate responsibility endorsed by the libertarian is logically impossible, while Pereboom (2001), Mele (1995), Kane (1996), and O’Connor (2000) have defended the position’s coherence. What is uncontroversial is the difficulty of articulating precisely how the falsity of determinism provides us with the kind of ultimate responsibility the libertarian treasures. This difficulty has led to some diminished expectations on the part of many libertarians. Wiggins, for example, writes:
One of the many reasons I believe why philosophy falls short of a satisfying solution to the problem of freedom is that we still cannot refer to an unflawed statement of libertarianism. Perhaps libertarianism is, in the last analysis, untenable. But if we are to salvage its insights we need to know what is the least unreasonable statement the position can be given. (1973, 33, my italics)
Kane (1996, 106) cites this passage as in line with his own project. “Can we show that such [libertarian] freedom is coherent,” he asks, “and give a clearer account of it than libertarians have offered in the past, an account that might show how such a freedom could exist in the natural order?” Libertarian ambition reaches its nadir in Van Inwagen’s (1983, 2000) “mind argument” (p.150) against indeterministic free will. Since the consequence argument is supposed to rule out the possibility of deterministic free will, one might expect Van Inwagen to become a skeptic. Instead, he concludes that free will undeniably exists but “remains a mystery.”
If we are to embrace libertarianism, it appears that the most we can hope for is an account that could exist in the natural order, and offers the “least unreasonable” explanation for how indeterminism adds anything to compatibilist accounts of responsible agency. I cannot think of another theory, in any area, that we would even consider accepting on such a flimsy basis. The difficulty of articulation, then, counts as a significant cost to accepting libertarian free will.
Of all the theories of moral responsibility, libertarianism is the most empirically demanding. As Vargas (2007) notes, it is not enough that the world contain indeterministic events. The burden of libertarianism “is that it must hold that the indeterminism shows up at particular times and places” (142). The evidence that this occurs is not just lacking, it is non-existent:
The main problem concerning the empirical plausibility of this view is that there are no accepted scientific models of indeterministic brain events. Indeed, virtually all brain science proceeds on the supposition of thoroughly deterministic explanations of the operations of the mind, and what evidence there is weighs against indeterministic interpretations of brain phenomena.
(Vargas 2007, 144)
Of course, brain science could one day vindicate an account like Kane’s. But the bare possibility that libertarianism could (p.151) conceivably enjoy empirical support in the future is not enough. In other areas of our epistemic lives, we resist accepting beliefs that we have no positive reason to believe are true. This is a core epistemic principle, at least for naturalists—and for non-naturalists, too, in most areas of their lives. Rejecting this principle in order to retain our belief in libertarian free will would fracture our epistemology. The pragmatic considerations in favor of the belief would have to be overwhelming to offset this cost.
These epistemic costs raise troubling moral issues as well. If we accept libertarianism, we can then hold people ultimately responsible for their behavior. We believe they can deserve blame and punishment, independent of consequentialist considerations. Given that we lack epistemic justification for this belief, we are open to the charge of what Double (2002) terms “moral hardness.” He notes that libertarians use the assignment of moral responsibility to turn otherwise immoral treatment into “just-deserts goods.” Thus, even if there were pragmatic value in accepting libertarianism from our own perspective, we would still have to consider the moral cost to others. Vargas (2007, 140) makes a similar point:
Imagine a criminal asking you why he should be made to serve for a longer period of time, when the only answer appeals to a conception of agency which has no evidence in its favor… . It is the wrong kind of faith to suppose that the moral acceptability of our denying clearly valuable things to these agents is justified or justifiable. It matters whether there is evidence that suggests we are libertarian agents, and so we must do better than to believe it on faith.
(p.152) To Vargas’s and Double’s points, I would add that not only is there no evidence for libertarian free will and responsibility, but there is no clear way to articulate what it would be if it did exist or how it might justify the criminal’s blameworthiness and punishment. Imagine a libertarian who, following Van Inwagen, must explain to the criminal that the type of agency that justifies his punishment is incoherent and therefore “a mystery.” The acceptance of libertarianism on such flimsy grounds is therefore inconsistent with a moral principle that requires a great deal of epistemic justification before permitting the mistreatment or harm of others.
Conclusion: An All-Things-Considered Rejection of Libertarianism
When we employ the above considerations in generating all-things-considered judgments, the case for libertarianism turns out to be pretty weak. On the one hand, we can reject libertarianism and still believe in the effectiveness of deliberation and in the existence of objective and moral worth; on the other, accepting a belief in libertarian free will creates inconsistencies with core epistemic and moral principles that we apply in all other areas of our lives. In short, the axiological and pragmatic benefits of accepting libertarianism do not seem to offset the costs. Libertarianism is therefore not a viable all-things-considered conclusion—at least for people whose epistemic and moral commitments resemble my own.
The Case for Compatibilism
Compatibilism about moral responsibility is both celebrated and mocked for its “best of both worlds” qualities. Compatibilists (p.153) can be empirically sound naturalists and at the same time claim all of the practical and moral benefits cherished by the libertarian (with the exception of a metaphysically open future.6) The primary cost of compatibilism is that one has to reject the TNR principle. Compatibilists and nihilists tend to agree about relevant metaphysical and non-moral issues. Both accept that our actions are the result of factors that trace back beyond our control. Their disagreement is over whether it is fair or appropriate to hold agents morally responsible for such actions.
The debate between compatibilists and nihilists about moral responsibility, then, boils down to which of two intuitive beliefs they choose to reject: the TNR principle or the belief that people can be morally responsible (PMR). If the TNR principle has intuitive force (as it does for me), embracing compatibilism will require that we revise our concept of moral responsibility so that so that it can accommodate our lack of ultimate responsibility. The all-things-considered case for compatibilism then becomes equivalent to what Vargas (2007) calls “moderate revisionism” about moral responsibility. In order to highlight the contrast, I will henceforth refer to skeptical or hard incompatibilist or nihilist accounts as “eliminativism” about moral responsibility. (The difficulty in coming up with an acceptable name for this position is a significant pragmatic cost of adopting it.) In this section, I will offer a critical examination of the case for revisionist compatibilism, one that, as always, is restricted to members of my intuition group.
The Strength of Our Commitment to Moral Responsibility
Following Strawson (2003 ), many contemporary compatibilists call attention to the strength of our commitment to moral responsibility and, in particular, to the reactive attitudes (p.154) that are bound up with this commitment.7 Strawson claims that the real question in the debate is whether the truth of determinism could cause us to reject our responsibility related attitudes. Strawson answers no: “We cannot, as it were, seriously envisage ourselves adopting a thoroughgoing objectivity of attitude to others as a result of the theoretical conviction of the truth of determinism” (Strawson 2003, 82). Strawson argues further that a commitment to the reactive attitudes entails a commitment to moral responsibility. If he is right, eliminativism is not option. It would require adopting “the objective attitude” on an exclusive basis, something that we are not psychologically capable of doing. And it cannot be rational to do what we are not capable of doing. Nichols (2007) refers to this strategy as the “insulationist move”: our attitudes and practices related to moral responsibility are insulated from theoretical threats posed by incompatibilist arguments.
It is certainly true that we are deeply committed to seeing others and, especially, ourselves as morally responsible agents, and as appropriate candidates of attitudes like resentment. Indeed, as Strawson argues, it is hard even to imagine what it would be like to give up this belief in its entirety. But is the rejection of moral responsibility and its accompanying attitudes and practices impossible? I do not believe so. Both Bennett (1980) and Russell (1995) contrast the Strawsonian position with that of the Spinozist. The Spinozist does indeed try to take the objective attitude on an exclusive basis, and he does so precisely because of a general theoretical belief in determinism. True, we often think we are morally responsible, but this is only because we are ignorant of the causes of our behavior. Once we are aware of these causes, the Spinozist argues, our responsibility-presupposing attitudes fade away. As Stuart Hampshire writes in his (p.155) study of Spinoza, once the causes of a certain type of behavior are understood,
… we do in fact cease to apply purely moral epithets to them as responsible agents. When the behavior now causally explained is what was formerly regarded as morally wicked, we come to regard it as the symptom of a disease, curable, if at all, by the removal of its causes; expressions of moral disapproval come to seem useless and irrelevant.
(Hampshire 1951, 158)
Now if determinism is true, the Spinozist goes on, then every action has causes for which the agent is not morally responsible. The only question is whether we can identify the causes of the action, and this is a purely epistemic question that should have no bearing on whether the agent actually deserves blame or praise for that action. How many practicing Spinozists there are in this world is an open question, but Spinoza himself may serve as a counterexample to Strawson’s insulationist claims. By most accounts Spinoza was said to live pretty much in accordance with his beliefs, showing little traces of resentment or bitterness over the considerable injustices done to him. Many Buddhist masters, too, are said to have banished emotions like resentment and hatred as a result of an understanding that the behavior of others is caused, and that the source of bad behavior is ignorance. The most appropriate reaction, they feel, is compassion, just as one might feel compassion for a sick person. And Charles Darwin expresses a similar view in his notebooks:
This view [the denial of free will] should teach one profound humility, one deserves no credit for anything. (yet one takes it for beauty and good temper), nor ought one to blame others … we must view a wicked man like a sickly (p.156) one. It would be more proper to pity than to hate & be disgusted.”8
Darwin, Spinoza, and a Buddhist monk might not appear to have much in common, but one thing they do seem to share is a belief that claims about moral desert are inappropriate if our actions have causes for which we are not responsible. It seems, then, that theoretical concerns can have an effect on our reactive attitudes and responsibility practices. As Galen Strawson writes, “The roots of the incompatibilist intuition lie deep in the very reactive attitudes that are invoked to undercut it. The reactive attitudes enshrine the incompatibilist intuition” (1986, 89).
One might object that all known cultures have believed in the concept and importance of moral responsibility (even if there is variation concerning the conditions for its application). Indeed, my analysis in Part One supports this claim. None of my examples presented a culture that denied the existence of responsibility altogether. The universality of the commitment to moral responsibility seems to indicate that the concept plays a valuable and perhaps indispensible role in our social lives.
This is a serious objection to which I offer two tentative replies.9 First, the importance of the responsibility concept in regulating the moral behavior of groups in the past does not entail its indispensability. It may be that present and future societies can flourish without the concept—especially if appropriate analogues are provided in its place. (See Pereboom (2001), Sommers (2007a), and the following chapter for a description of these analogues.) At one time, virtually all societies believed in the importance and truth of religion. Religious beliefs were intimately connected with moral norms and practices, and many predicted disaster if society at large were to lose its faith. But the doomsday scenarios have not come to pass in countries such as (p.157) the Czech Republic where a large proportion of the citizenry have become atheists and agnostics. The same may be true for societies that give up their belief in desert-entailing moral responsibility. We will not know for sure until we have a concrete example.
My second reply concerns the scope of my ambitions in these two chapters. My goal is not to speculate about what would happen if an entire society retained or rejected their belief in moral responsibility. Rather, it is to examine the implications of the various positions for individuals. It is arguable that no society (especially in the West) will ever fully embrace the notion that we are not free and morally responsible. Powerful psychological and cultural forces push us to believe in the reality of these concepts, and most people lack the resources and the inclination to resist those forces. Particular individuals, on the other hand, can become committed nihilists at any time and endeavor to diminish the influence of responsibility-presupposing attitudes on their beliefs and behavior. This is why I focus on examples like Darwin and Spinoza—individuals whose theories caused them to deny free will and moral responsibility but who lived in societies where these concepts were, and still are, pervasive. Indeed, there is a real sense in which it would be inappropriate for me, as a metaskeptic, to be a missionary for a particular first-order position. Many people in my society may not belong to my intuition group—they may not share my core starting intuitions, and they could, therefore, arrive just as reasonably at a different final verdict.
In sum, I agree with Strawson and his followers that our intuitions and sentiments pull many of us in the direction of holding a person responsible when he meets appropriate compatibilist conditions. But another set of intuitions pulls us towards mitigation or exoneration once we see that factors beyond the (p.158) agent’s control are the ultimate cause of his or her action. A slew of recent studies in experimental philosophy suggest that we are deeply conflicted when it comes to our beliefs about moral responsibility. Factors such as emotional salience, concreteness, mood, psychological distance, and even the messiness of our surroundings can incline us in one direction or the other.10 It seems, then, that we cannot make the case for compatibilism solely on the basis of our belief in moral responsibility or our commitment to responsibility-presupposing attitudes.
Can the TNR Intuition Be Explained Away?
The compatibilist at this point may argue that incompatibilist principles only seem plausible because of some mistaken assumptions about what they involve. Incompatibilist principles like TNR will lose their intuitive appeal once they are properly understood. Nahmias, Coates, and Kvaran (2007) and Nahmias and Murray (2010) have recently developed an “error theory” for incompatibilist judgments. According to these authors, many people who appear to have incompatibilist intuitions are interpreting determinism to involve the bypassing (or causal impotence) of desires and conscious deliberation. But this is a mistake, since deterministic or naturalistic accounts of human behavior do not entail that our desires and deliberation are causally impotent. The authors conclude that to the extent that we have them, incompatibilist intuitions can be traced to this mistake of confusing determinism with fatalism.
To support this interpretation, the authors developed two kinds of deterministic scenarios—one in which conscious deliberation is bypassed, and one in which it is not. The authors operate under the assumption that when the decision-making process is described mechanistically as neural processes and (p.159) chemical reactions in the agent’s brain, subjects tend to think that beliefs, desires, and reasoning are causally impotent. This assumption provides the key manipulation for the deterministic scenarios. In one condition, the agents’ decision-making is described “in terms of neuroscientific, mechanistic processes (neuro scenarios)”; in the other, decision-making is described “in terms of psychological, intentional processes (psych scenarios).” The authors made three central predictions:
1. Most people will judge that determinism is not threatening to free will and moral responsibility if determinism is described in nonmechanistic (psychological) terms.
2. Significantly more people will judge determinism to be threatening if determinism is described in mechanistic (neuroscientific) terms.
3. People will significantly increase their attributions of free will and moral responsibility [in both conditions] in response to descriptions of specific agents who perform bad acts in comparison to agents and actions described in abstract ways. (Nahmias et al. 2007, 221)
The results appear to offer significant support for their predictions. Subjects tended to give compatibilist responses in the abstract/psych condition but incompatibilist responses in the abstract/neuro condition. Neuroscientific descriptions of decision-making seem to be more of a threat to responsibility than psychological ones.
The authors’ target is a study by Nichols and Knobe (2007) that offers support for the view that the “folk” have incompatibilist commitments. They object to Nichols and Knobe’s description of determinism, which includes the claim that “given the past, each decision has to happen the way that it does.” The (p.160) “had to happen” language, according to Nahmias et al., leads subjects to think that the actions they perform will occur no matter what their desires, intentions, and reasons may incline them to do.
However, if Nahmias and colleagues have grounds to object to the “had to happen” language, incompatibilists can complain in turn that the authors’ own description of the “psychological” condition does not make determinism sufficiently salient. Consider the final paragraph of the description:
So, if these psychologists are right, then once specific earlier events have occurred in a person’s life, these events will definitely cause specific later events to occur. For instance, once specific thoughts, desires, and plans occur in the person’s mind, they will definitely cause the person to make the specific decision he or she makes.
(Nahmias et al. 2007, 224)
The authors focus on perhaps the least threatening aspect of determinism (that our thoughts and desires determine our actions); the description would seem to gloss over the historical aspect of determinism, the notion that these thoughts, desires, and plans are determined by events that occurred before the agents were born.
Nahmias and Murray (2010) have recently conducted a study that more straightforwardly examines the possibility that subjects conflate determinism with fatalism in the Nichols and Knobe study. After giving subjects Nichols and Knobe’s description of determinism, they asked them whether the agent’s actions could have been different if his or her desires, intentions, and thoughts had been different. The results:
(i) the vast majority of participants who express apparent incompatibilist intuitions interpret determinism to involve (p.161) bypassing, while those who express prima facie compatibilist intuitions tend not to, and … (ii) there is a dramatic correlation between the degree to which participants take a scenario to involve bypassing and the degree to which they attribute MR and FW to agents in that scenario, but the results also suggest that (iii) the difference in attributions of MR and FW between the abstract conditions is caused by people’s bypassing interpretations. Subjects may be conflating fatalism with determinism, assuming that agents would perform their actions whether or not they had the intention/desire to do so (207).
I think the general strategy of trying to explain away incompatibilist intuitions is an excellent one. If compatibilists can show that the intuitive force of the TNR principle is based on an error, then we will have good reason to view the intuition as unreliable. Since our commitment to TNR is the primary reason to resist compatibilist accounts of moral responsibility, the case for compatibilism will be all but made. But there is a significant problem with the studies Nahmias and his colleagues appeal to, one that applies to most contemporary experimental work on free will and moral responsibility: Rather than probe for intuitions about the relevant principles philosophers employ in their arguments, the studies probe for intuitions on the compatibility question itself.11
The roots of this problem can be seen Nahmias et al.’s seminal article “Is Incompatibilism Intuitive” (2006), the first systematic empirical investigation into our intuitions about free will and determinism. The study probes for intuitions on the compatibility question and finds that the folk give largely compatibilist responses. To explain why such a study is philosophically important, the authors claim (correctly) that arguments for incompatibilism always appeal to intuitions about key premises (p.162) and cases. Next, they argue (again correctly) that we have no reason to accept incompatibilism unless we find it intuitive upon reflection. The problem arises in the next step of their argument. Nahmias et al. offer a series of quotations from incompatibilists who claim that the folk are natural pretheoretic incompatibilists. One of these incompatibilists, Robert Kane, writes:
In my experience, most ordinary persons start out as natural incompatibilists. They believe there is some kind of conflict between freedom and determinism; and the idea that freedom and responsibility might be compatible with determinism looks to them at first like a “quagmire of evasion” (William James) or “a wretched subterfuge” (Immanuel Kant). Ordinary persons have to be talked out of this natural incompatibilism by the clever arguments of philosophers.
The authors cite similar passages from Galen Strawson, Thomas Pink, and Laura Ekstrom. The motivation for the studies, then, seems clear. Kane and his fellow incompatibilists are making empirical claims about folk intuitions that can, and should, be tested. Subsequent articles in the experimental literature have followed Nahmias and framed the issue in like manner, selecting from a now-standard menu of passages about the intuitiveness of incompatibilism and then providing data to support or undermine these claims. What is odd from a dialectical standpoint, however, is that the cited passages almost always come from introductory sections of incompatibilist articles and books. The role of such passages is, as one might expect, largely rhetorical, to get readers into an incompatibilist mood. There is no incompatibilist argument featuring a premise like, say, “free will and moral (p.163) responsibility are incompatible with determinism” (and for good reason since this would be an obvious case of question-begging). Nor are there arguments of the form (1) we intuitively find free will and moral responsibility to be incompatible with determinism, (2) therefore free will and moral responsibility are incompatible with determinism. So while Nahmias et al. (and the subsequent work that tested intuitions about the compatibility question) are correct in stating that incompatibilist arguments appeal to intuitions, the appeal is for incompatibilist principles, not for the compatibility question itself.
What, then, are the implications of the compatibilist results of Nahmias and colleagues? The authors write:
Either determinism obviously precludes free will or those who maintain that it does should offer an explanation as to why it does. The philosophical conception of determinism—i.e., that the laws of nature and state of the universe at one time entail the state of the universe at later times—has no obvious conceptual or logical bearing on human freedom and responsibility. So, by claiming that determinism necessarily precludes the existence of free will, incompatibilists thereby assume the argumentative burden.
(Nahmias et al. 2006, 30)
It seems, however, that incompatibilists can accept that they bear this “argumentative burden” but at the same time claim that they have discharged it—with arguments! After all, Van Inwagen’s “consequence argument,” Strawson’s “basic argument,” and Pereboom’s “four-case argument,” to name only a few, are designed precisely to lead the reader to the conclusion that determinism precludes free will and moral responsibility. Again, the arguments appeal to intuitions, but to intuitions about cases and narrower principles (such as TNR), and not to intuitions (p.164) about the question posed in the experimental studies. In order to truly test the plausibility of the incompatibilist position, then, we would need to examine the intuitions that support the premises and principles of their arguments and not their introductory rhetoric.
Nahmias and colleagues anticipate an objection of this kind, but their response is problematic. They acknowledge that they have not tested something like the consequence argument directly, but state that their results nonetheless provide indirect evidence against its intuitive plausibility, because the scenarios “present conditions in the past that, along with the laws of nature, are sufficient conditions for the agent’s action.” (2006, 45) They assume, in other words, that philosophically unsophisticated subjects should have somehow internalized or implicitly grasped the consequence argument before ever hearing it! Van Inwagen has received credit for reviving incompatibilism in the past century in large part because he found a non-obvious way of revealing the threat of determinism to free will. Indeed, the whole point of developing an incompatibilist argument (rather than just asserting the conclusion) is to draw out the implications of determinism that might otherwise go unappreciated. It is not clear, then, that the results of these studies give much in the way of indirect evidence against incompatibilist arguments either.
When we apply this criticism to Nahmias et al. (2007) and Nahmias and Murray (2010), we see that the results do not explain away our intuitions about incompatibilist principles such as TNR. Rather, at best, they provide an error theory for our intuitions about the compatibility question. Incompatibilists do not, however, rely on those intuitions, so they would not need to be too worried if they could be explained away. The independent intuitive plausibility of principles like TNR (e.g., (p.165) Nagel 1979, Van Inwagen 1983, Strawson 1986) leads me to believe that they cannot be accounted for in the manner that Nahmias and colleagues suspect. Furthermore, the best manipulation and design cases often emphasize explicitly that agents’ reasoning capacities are part of the causal process leading to behavior—that they are not bypassed—which suggests that the determinism/fatalism confusion cannot account for our intuitions in these cases.
Finally, it is worth noting that even if the folk are guilty of this confusion, I am not guilty of it. The results might show that leaping from determinism to fatalism is tempting and common, so that we should be especially wary of it in our own deliberations about the conditions for moral responsibility. But if we are fully aware of this danger and still find the TNR principle to be intuitive (as I do), then what reason do we have to think that our intuitions are still being corrupted by the confusion?
The Ethical and Pragmatic Benefits of Compatibilism
Thus far, the case for all-things-considered (revisionist) compatibilism is inconclusive. While it is true that we have a strong natural commitment to responsibility-related attitudes and practices, many of us seem also to have a natural commitment to a condition for moral responsibility that is incompatible with naturalistic accounts of human agency. There is, moreover, no compelling evidence that the intuitions supporting this condition (and the TNR principle) can be explained away. There may still be ethical and pragmatic considerations, however, that make it reasonable to accept compatibilism over the eliminativist alternative. Indeed, a crucial feature of P. F. Strawson’s defense of compatibilism is based on an “assessment of the gains and losses to human life” that the reactive attitudes bring with (p.166) them. As Bennett puts it: “The greatest single achievement of ‘Freedom and Resentment’ … is its showing how the question ‘Ought we to retain praise, blame, etc.?’ could be a fundamentally practical one rather than having a strict dependence upon a perpetually troublesome theoretical question” (1980, 30).
So what are the practical benefits of retaining our moral-responsibility practices and attitudes? Strawson and contemporary compatibilists in this tradition (such as Wolf, Watson, and McKenna) focus on the centrality of the reactive attitudes to interpersonal relationships. To regard human beings exclusively with “the objective attitude”—as the eliminativist must—would lead to a cold, bleak, tragic world of human isolation. Embracing compatibilism allows us to retain the natural range of interpersonal attitudes.
In Sommers (2007a), I argue that philosophers have taken an unduly pessimistic view of the objective attitude. They misconstrue what it would mean to deny moral responsibility and renounce the attitudes that presuppose it. I focus on Wolf’s (1981) description as it is the clearest and most eloquent expression of the prevailing pessimistic view. Imagining a world in which we regarded each other solely with the objective attitude, Wolf writes:
We would still imprison murderers and thieves, presumably, and we would still sing praises for acts of courage and charity. We would applaud and criticize, say “thank you” and “for shame” according to whether our neighbors’ behavior was or was not to our liking. But these actions and words would have a different, shallower meaning than they have for us now. Our praises would not be expressions of admiration or esteem; our criticisms would not be expressions of indignation or resentment. Rather, they would be bits of positive and negative reinforcement (p.167) meted out in the hopes of altering the character of others in ways best suited to our needs. (1981, 391)
The eliminativist can agree with most of the claims in this passage (e.g., “our criticisms would not be expressions of indignation and resentment”) but would question the gratuitous use of words like “shallower.” Wolf continues:
An act of heroism or of saintly virtue would not inspire us to aim for higher or nobler ideals, nor would it evoke in us a reverence or even admiration for its agent. At best we would think it is a piece of good fortune that people occasionally do perform acts like this. … We would not recoil from acts of injustice or cruelty as insults to human dignity, nor be moved by such acts to reflect with sorrow or puzzlement on the tide of events that can bring persons to stoop so low. Rather, we would recognize that the human tendency to perform acts like this is undesirable, a problem to be dealt with, like any other, as scientifically and efficiently as possible. (391)
Here I believe Wolf’s worries are mostly groundless. Why, when we take the objective attitude, should an act of heroism not inspire us to aim for this ideal? As noted above in the discussion of Smilansky’s illusionism, we do not perform heroic acts merely to deserve praise for them; we perform them because we think the act will serve a worthy purpose. Consider the firemen who entered the World Trade Center on September 11, 2001. Did they risk serious injury and death in order to deserve praise for their bravery? Surely their primary motivation was to save the people trapped inside the building from being burned alive. Such actions inspire admiration and reverence whether or not we attribute any kind of desert-entailing responsibility to the people who performed them. Wolf’s second claim is similarly wrong. (p.168) Why should we not recoil from acts of cruelty? If Wolf’s emphasis here is on cruelty as an insult to human Dignity with a Kantian capital “D,” then she is perhaps right. But this does not stop us from recoiling at the sight of cruelty, not as an insult to Dignity, but as a cause of intense physical and psychological suffering. We all recoil at the sight of a man being eaten by a tiger, or burned alive, without blaming or resenting the animal or the fire. And of course we would be moved to reflect with sorrow at how an event of human cruelty has come to pass. Not blaming the criminal in no way diminishes our sorrow at the suffering of the victim.
Wolf concludes her indictment of the objective attitude with the following observations:
The most gruesome difference between this world and ours would be reflected in our closest human relationships—the relations between siblings, parents and children, and especially spouses and companions. We would still be able to form some sorts of association that could be described as relationships of friendship and love. One person could find another amusing or useful. One could notice that the presence of a certain person was, like the sound of a favorite song, particularly soothing or invigorating. We could choose friends as we now choose clothing or home furnishing or hobbies, according to whether they offer, to a sufficient degree, the proper combination of pleasure and practicality. Attachments of considerable strength can develop on such limited bases. People do, after all, form strong attachments to their cars, their pianos, not to mention their pets. Nonetheless I hope it is obvious why the words “friendship” and “love” applied to relationships in which admiration, respect, and gratitude have no part, might be said to take on a hollow ring. (391)
(p.169) Once again, the pessimism here seems unmotivated. When we take the objective attitude towards other human beings, we do nothing more than see them as natural things. A human being is still a human being—the most exciting, infuriating, unpredictable, lovable, loathsome natural thing in the world. So we would not merely find them useful or amusing. We would not choose our friends like we would choose home furnishings, hobbies, songs, pianos, or pets. We choose friends like we choose human friends—that’s all. Nothing in the objective attitude prevents us from recognizing, appreciating, cherishing the rich and wonderful qualities of another person. It remains the choice that brings the greatest rewards and the deepest disappointments in all of human existence. Rejecting our responsibility-related attitudes and practices would not change this at all.
The disagreement over these questions is perhaps best seen as an example of how individual differences in temperament can influence our philosophical beliefs. As may be apparent, I have a hard time understanding people who despair at the prospect that we lack free will and moral responsibility. Yet I know that they exist. I know that their feelings are genuine and that they have an equally hard time understanding me. Smilansky (forthcoming) claims to find the optimistic hard determinist perspective “literally incredible.” Indeed, he independently uses the example of the 9/11 firemen to argue for the opposite conclusion:
Think about the firefighters who rushed into the collapsing, burning buildings in 9/11. Envisage a situation in which a few of them who survived, after losing most of their mates, and perhaps being physically harmed themselves, are sitting around and talking about things. Assume that a hard determinist philosopher comes in and tells them that fundamentally they are not better than any common thief or rapist, explaining that ultimately everyone (p.170) is what he is, as an inevitable outcome of forces beyond his control. Do you think that they would welcome this philosopher? Or that, if one or two of them were to listen to him, and begin to understand and internalize what he was saying, that those people would not feel that something very unpleasant and deeply threatening was being said?
Not only do we both use the same example, but we both litter our analyses with rhetorical questions—indicating that we think the answer is obvious! Ultimately, I believe this disagreement boils down to differences in personal psychologies or temperaments. Rational argument can only take us so far. I would note, however, that Smilansky is wrong to attribute to hard determinists the view that “fundamentally [the firefighters] are not better than any common thief or rapist.” True, hard determinists recognize that we are the inevitable outcome of forces beyond our control. But these forces still shape us into better or worse people. Rapists and thieves are bad people. It may not be their fault that they are bad people, but they are bad people nonetheless. Similarly, the 9/11 firemen were good people, courageous people, whether or not they deserve credit for their courage. Only those who are committed to a radically Kantian understanding of moral worth can claim otherwise.
Nichols’s Empirical Defense of Moral Responsibility
Even if one does not despair at the thought of hard determinism, there may be other benefits to holding on to a belief in moral responsibility. Nichols (2007) offers a pragmatic defense of responsibility-presupposing attitudes, appealing to evolutionary models and research in experimental economics. (See (p.171) chapter 2 for a more detailed description of these theories and models.) Following Frank (1988) and Trivers (1971), Nichols argues that attitudes like moral outrage and guilt help to secure norms that motivate human cooperation. He appeals to the work of Fehr and Gachter (2002) and Boyd et al. (2003), who discuss the role of retributive attitudes in motivating altruistic punishment. Cooperation might deteriorate, Nichols worries, if we were to eliminate the attitudes that are associated with responsibility and desert.
How much support does this research lend to the all-things-considered case for compatibilism? Not very much in my view. Nichols’s argument works like this:
1. Reactive attitudes motivate individuals to punish free-riders at a cost.
2. In economic games, when punishment options are removed, cooperation deteriorates and free-riding behavior increases.
3. Therefore the reactive attitudes are necessary for fostering cooperation.
This argument is not valid, because there may be other ways of raising the expectation of punishment. Indeed, in our society, the law and the criminal justice system exist for this very purpose. The concepts of legal responsibility and legal punishment can be separated from beliefs about moral responsibility, as utilitarians since Bentham have repeatedly stressed. The experimental evidence only shows that cooperation suffers when punishment is absent. One might argue that laws are designed to reflect the sentiments and intuitions of society, that without retributive attitudes our laws might not be sufficiently punitive. To this, I reply again that we are considering eliminativism at the level of the individual. As a purely descriptive matter, it is (p.172) unlikely that most of society will reject responsibility-related beliefs and attitudes. The function of moral anger only requires that potential offenders expect to find it in others as a response to wrongdoing. Any particular individual, then, can reject attitudes like moral anger without undermining its function—as long as he or she stays in the relative minority.
Indeed, Nichols himself notes that the best strategy for an individual in public goods games is never to punish, as long as there are enough punishers in the population to sustain cooperation. But he accuses anti-retributivists of “second-order free-riding” since they get the benefits of retributivism without paying the costs (Nichols 2007, 426, n. 11). In one sense he is correct. Presumably, however, eliminativists who engage in this kind of free-riding do so not for personal gain, but rather because of their commitment to a principle of fairness (the TNR principle). Surely the pragmatic gains of a position adopted for moral reasons should not count against the position.
The goal of this chapter has been to outline arguments for versions of libertarianism and compatibilism that are consistent with the broader metaskeptical thesis. I considered the “all-things-considered” cases for both compatibilism and libertarianism for members of my “intuition group”—people with starting intuitions and attitudes about responsibility that resemble my own. I argued that libertarianism is an implausible alternative because its epistemic costs outweigh its alleged pragmatic and moral benefits. The more promising case for compatibilism, I argued, relies on questionable assumptions. I will continue to challenge those assumptions in the final chapter as I develop the case for eliminativism about moral responsibility.
(1.) In Sommers (2005), I observed that denying moral responsibility took most of the fun out of watching sports—and that in this arena, I was willing to suspend disbelief. During my dissertation defense, Galen Strawson noted that “it almost seemed as if I was serious about sports being the most difficult place to rid myself of a belief in moral responsibility.” The reason it almost seemed that way was because I was serious.
(2.) Another option might be Smilansky’s “illusionism,” in which one finds libertarianism to be false but embraces the illusion that we have libertarian free will because of the moral benefits. It is unclear to me, however, to what extent this would count as an illusion—since the axiological case for “unillusioned” libertarianism relies in large part on pragmatic and moral considerations.
(4.) Baumeister et al. (2009, 267) have expanded upon these results in a study which suggests that inducing disbelief in free will leads to an increase in aggression and a reduction in willingness to help. They also tested for “chronic disbelief” in free will and observed a correlation between high scores on this scale and a lack of helpfulness.
(5.) Having one’s entire worldview shaken up can certainly have some short-term negative effects on behavior. Fifteen minutes after Grady Little inexplicably allowed Pedro Martinez to give up the lead to the Yankees in game 7 of the 2003 ALCS, I might have committed (p.210) acts that are far more immoral than cheating on a psychology quiz.
(6.) I do not discuss the value of metaphysically open futures since they are, in my view, an utterly bizarre thing to want—a bit like wanting to be buried in outer space. Clearly, the weight given to the benefits and costs of each position is a subjective matter.
(7.) I use the vague “bound up” language to avoid taking a stand on whether the reactive attitudes constitute or presuppose the existence of moral responsibility. Either interpretation can apply in the following discussion.
(9.) Thanks to an anonymous referee for throwing this monkey wrench into my argument.
(10.) See e.g., Nahmias et al. (2006), Nichols and Knobe (2007), and Feltz and Cokely (2009). Sommers (2010) provides a review and critique of this literature. For reasons discussed below, I do not appeal to the studies in this literature as a means of supporting one side or the other.