J Ethics (2016) 20:229–246 DOI 10.1007/s10892-016-9234-9
Moral Responsibility, Reactive Attitudes and Freedom of Will Robert Kane1
Received: 28 May 2016 / Accepted: 1 June 2016 / Published online: 13 June 2016 Springer Science+Business Media Dordrecht 2016
Abstract In his influential paper, ‘‘Freedom and Resentment,’’ P. F. Strawson argued that our ordinary practices of holding persons morally responsible and related reactive attitudes (such as blame, resentment, indignation, and moral approval) were wholly ‘‘internal’’ to the practices themselves and could be insulated from traditional philosophical and metaphysical concerns, including concerns about free will and determinism. This ‘‘insulation thesis’’ is a controversial feature of Strawson’s influential paper; and it has had numerous critics. The first purpose of this paper is to explain my own reasons for thinking that our practices of holding responsible cannot be entirely insulated from incompatibilist concerns about freedom and determinism. The second purpose is to argue that these incompatibilist concerns are in fact legitimate concerns: There are sound reasons to believe that our ordinary practices of holding persons morally responsible do require at least sometimes in our lives that we must be capable of acting freely in a manner that is not determined. I defend this thesis by spelling out why I believe various compatibilist strategies attempting to show that moral responsibility is compatible with determinism fail to show this. In the course of this critique, a general theme will emerge: In order to do full justice to our ordinary practices of holding persons responsible and the freedoms thus involved, one must distinguish between different types of freedom, and in particular, between freedom of action and freedom of will. Keywords Determinism Free will Incompatibilism Moral responsibility Reactive attitudes
& Robert Kane
[email protected] 1
Austin, TX, USA
123
230
R. Kane
1 Introduction Can our ordinary practices of holding persons morally responsible, and the reactive attitudes associated with those practices, be entirely ‘‘insulated’’ from incompatibilist concerns about free will and determinism? P. F. Strawson thought so. In his influential paper, ‘‘Freedom and Resentment,’’ he argued that the application and justification of these ordinary practices of holding responsible and related reactive attitudes (such as blame, resentment, indignation and moral approval) were wholly ‘‘internal’’ to the practices themselves and could be insulated from traditional philosophical and metaphysical worries, including worries about free will and determinism. To believe these ordinary practices of holding morally responsible would have to be modified in some ways, mitigated or even possibly abandoned, if we found that our actions were determined by prior causes, he argued, was to ‘‘overintellectualize’’ the issues. This ‘‘insulation thesis’’ is one of the most controversial features of Strawson’s paper; and it has had numerous critics.1 Interestingly enough, one of the most prominent of these critics was Strawson’s son, Galen Strawson, who in his book, Freedom and Belief, took issue with his father’s contention in ‘‘Freedom and Resentment’’ that ordinary practices of blaming and other reactive attitudes could be entirely insulated from metaphysical worries about determinism. Against this contention, Galen Strawson argued that ‘‘the roots of the incompatibilist intuition lie deep in the very reactive attitudes that are invoked in order to undercut it. The reactive attitudes enshrine the incompatibilist intuition’’ rather than being insulated from it. (Strawson 1986: 89) I agree with Galen Strawson on this issue, though my reasons are not all the same as his. The first of several purposes of this paper is to explain my reasons for thinking that our practices of holding morally responsible cannot be entirely insulated from incompatibilist concerns about freedom and determinism (Sect. 2). The second purpose is to argue that these incompatibilist concerns about freedom and determinism are in fact legitimate concerns: There are sound reasons to believe that our ordinary practices of holding persons morally responsible do require at least sometimes in our lives that we must be capable of acting freely in a manner that is not determined. I defend this thesis by spelling out (in Sects. 3, 4 and 5) why I believe that various compatibilist strategies attempting to show that moral responsibility is compatible with determinism fail to show this. In the course of this critique of compatibilist strategies, a general theme will emerge that I have argued for in writings over the past four decades: In order to do full justice to our ordinary practices of holding persons morally responsible and the freedoms thus involved, one must distinguish between different types of freedom, and in particular, between freedom of action and freedom of will. Both types of freedom and notions of responsibility associated with them, I will argue, are required to do full justice to our ordinary practices of holding persons morally responsible.
1
See e.g., Nagel (1986: 124–126), Watson (1987), Russell (2011), Fischer (2014).
123
Moral Responsibility, Reactive Attitudes and Freedom of…
231
2 Ordinary Practices: Excuses and Exemptions There are different ways to highlight problems with the thesis that our ordinary practices of holding responsible, and related reactive attitudes, can be entirely insulated from incompatibilist concerns about free will and determinism. I discuss several ways in this section that are particularly important for my purposes. The first and most obvious of these ways is by considering certain features of ordinary practices of holding responsible involving common excuses and exemptions. Consider, for example, Peter Strawson’s own account in ‘‘Freedom and Resentment’’ of various excusing and exempting circumstances under which ordinary practices of holding responsible would have to be modified in some ways, mitigated or even abandoned, and in which it would be appropriate for reactive attitudes, such as blame and resentment, to be ‘‘mollified’’ or withheld. Among the legitimate excusing conditions recognized in ordinary practices, Peter Strawson argues, belong all those circumstances which might give occasion for the employment of such expressions as ‘‘He didn’t mean to,’’ ‘‘He hadn’t realized’’… and also all those which might give occasion for the use of the phrase ‘‘He couldn’t help it’’ when this is supported by such phrases as ‘‘He was pushed,’’ ‘‘He had to do it,’’ ‘‘It was the only way,’’ ‘‘They left him no alternate’’ etc. (Strawson 1962: 23) As critics of the insulation thesis (including Galen Strawson) have pointed out, these ordinary excusing conditions (‘‘he couldn’t help it,’’ ‘‘he had to do it,’’ ‘‘they left him no alternative’’) strongly suggest that when an agent is not free to do otherwise, he or she is not legitimately deemed morally responsible for the behavior in question. And, as critics often further point out, this seems to land us squarely in traditional metaphysical debates that have concerned incompatibilists about whether causal determinism rules out the freedom to do otherwise and whether the freedom to do otherwise is required for moral responsibility. In other words, given ordinary conditions in which it would be appropriate to modify or withhold ascriptions of moral responsibility, on Peter Strawson’s own account, one cannot entirely insulate questions of moral responsibility from metaphysical questions about freedom and determinism. One might, of course, still argue in various ways that moral responsibility is compatible with not being free to do otherwise and hence with determinism or to argue that moral responsibility does not require the freedom to do otherwise at all. But this would be to engage squarely with the traditional metaphysical debates (as we will do in Sects. 3–5) rather than arguing that ordinary practices of holding responsible can be insulated from these debates. A second important way of highlighting problems with the insulation thesis is by focusing on practices of ascribing responsibility, culpability and blame in ordinary courts of law. A widely cited condition among legal theorists for such ascriptions was stated by the influential legal theorist Hart (1970). Hart argued that a necessary condition for ascribing responsibility and culpability to agents in legal contexts was that the agents must have had a ‘‘fair opportunity to avoid wrongdoing,’’ or more
123
232
R. Kane
generally, a ‘‘fair opportunity to have done otherwise’’ than they have done. In an important recent article, Brink and Nelkin (2013) argue persuasively that Hart’s criterion is not only crucial for understanding legal and criminal responsibility, but for understanding moral responsibility in general, in an accountability sense of moral responsibility that would justify blame, sanction and punishment. They argue, as a consequence, that Hart’s condition of fair opportunity to avoid wrongdoing or to avoid doing otherwise is a crucial part of the ‘‘architecture’’ of ordinary practices of ascribing moral as well as legal responsibility. (Brink and Nelkin 2013: 284) Hart does not offer a precise definition of what counts as a ‘‘fair’’ opportunity to avoid wrongdoing. He rather appeals to common conditions cited in ordinary practices in the law for excusing or exempting persons from responsibility as examples of what would count as the absence of such ‘‘fair’’ opportunity. The excusing and exempting conditions he cites are similar to those cited by Strawson in ‘‘Freedom and Resentment’’ under which ordinary practices of holding responsible would have to be modified in some ways, mitigated or abandoned—conditions such as incapacity, ignorance, coercion, compulsion, duress, insanity, mental impairment, and the like. It is not surprising, therefore, that appeal to Hart’s much-cited ‘‘fair opportunity to do otherwise’’ criterion for assigning responsibility in ordinary practices of law also lands one squarely in the center of traditional metaphysical debates that have concerned incompatibilists about whether causal determinism rules out the freedom to do otherwise and whether the freedom to do otherwise is required for moral responsibility. To illustrate the point further, it is instructive to consider the following question: Does an agent have a ‘‘fair opportunity to avoid wrongdoing’’ in certain circumstances, if the agent’s avoiding the wrongdoing in those circumstances is causally impossible? On the face of it, it seems not. Does one have a fair opportunity to avoid doing something that it was causally impossible in the circumstances for one to avoid doing? But note that, if an occurrence—any occurrence, including an act of wrongdoing—is causally determined, then, by definition, it’s nonoccurrence in the circumstances is causally impossible. And if an agent’s avoiding wrongdoing in certain circumstances was causally impossible, it would certainly appear that the agent lacked a ‘‘fair opportunity’’ to avoid doing it. But it might be objected that the causal impossibility of doing otherwise is not among the usual excusing or exempting conditions commonly cited in ordinary practices of holding responsible, such as incapacity, ignorance, coercion, duress, insanity, mental impairment, etc. Nor, one might add, does the causal impossibility of doing otherwise necessarily imply that any of these particular commonly cited excuses or exemptions obtain. In fact, a line of argument of this kind is used by Strawson himself in ‘‘Freedom and Resentment’’ in the attempt to show that the causal impossibility of doing otherwise (which follows from determinism) does not function as an excusing or exempting condition in ordinary practices of holding agents responsible; and it is a line of argument that has influenced many others. Yet it is a more problematic line of argument than is usually realized, for several reasons. First, one may grant that the causal impossibility of doing otherwise is not among the commonly cited conditions excusing or exempting agents from responsibility,
123
Moral Responsibility, Reactive Attitudes and Freedom of…
233
such as those mentioned. But it does not follow that the causal impossibility of doing otherwise cannot and does not often imply one or another of these commonly cited excusing conditions, such as incapacity or lack of opportunity. There are numerous examples showing that it can and does so function. Consider an accident at an airport in which an incoming aircraft has run off the end of the runway, injuring many passengers and seriously damaging the aircraft. The initial assumption is that pilot error was involved. But further inspection shows that there was a subtle defect in the design of the aircraft that made it causally impossible for the pilot to have avoided the disaster, given the conditions of the runway and the weather at the time. The pilot, it is concluded, consequently lacked a ‘‘fair opportunity to avoid the outcome’’ and could not be held responsible or blameworthy for the accident. Similarly, if an explosion occurs in an empty factory one evening, the security guard in a control room in charge of monitoring conditions throughout the factory may be legitimately excused of responsibility or blame for not preventing the explosion, if an investigation shows that it was causally impossible in the circumstances for him to have acted to prevent it because of a defect in the monitoring systems which gave him no warning. In the circumstances, he lacked the capacity to prevent it. It is a mistake, therefore, to assume that the causal impossibility of doing otherwise cannot and does not imply any of the commonly cited excusing conditions in ordinary practices of holding responsible. It can and often does so function. But it is equally important to point out that the causal impossibility of an agent’s doing otherwise in certain circumstances and the consequent lack of capacity or opportunity to do otherwise, do not always imply that the agent is excused from moral responsibility. They do so only conditionally. And this is an equally crucial fact, I believe, that has a bearing on the nature of moral responsibility in general and its relation to the freedom to do otherwise. To illustrate, consider another example. Suppose on a dark street and a rainy night, a drunk driver strikes and kills a pedestrian, and it could be shown that given the circumstances, including the weather, the visibility on the road and the condition of his nervous system, given the alcohol, he could not possibly have avoided hitting the pedestrian. This fact alone—that it was not causally possible, given all of these circumstances existing at the moment of the accident, for him to have avoided it— will not excuse him of responsibility. For one also wants to know whether he was responsible by virtue of earlier actions or omissions on his part for the existence of some of those crucial circumstances that made it now causally impossible for him to avoid hitting the pedestrian, such as his prior decisions and actions to drink as much as he did and then to get in his car and drive. Even if it was causally impossible at the moment for him to avoid the outcome, so that he lacked a fair opportunity to do so, he may be held morally and legally responsible for doing so, by virtue of these earlier decisions and actions, to the extent that he had a fair opportunity to avoid these earlier decisions and actions. This italicized qualification is crucial for understanding when the causal impossibility of doing otherwise is, and is not, a valid excusing condition in our ordinary practices of holding morally and legally responsible. And it leads us directly to a third way—and a particularly important way—of highlighting problems
123
234
R. Kane
with the insulation thesis. This particular challenge to the insulation thesis is discussed by a number of writers, most recently by Nichols (2015), which focuses on the implications of new research in empirical psychology and experimental philosophy for traditional philosophical debates about free will and moral responsibility. The passages of interest here are where Nichols discusses Galen Strawson’s claim, mentioned earlier, that our ordinary practices of holding responsible, and the reactive attitudes related to them, ‘‘enshrine’’ incompatibilist intuitions about freedom and responsibility rather than being ‘‘insulated’’ from them. In discussing what he takes to be important arguments supporting this claim, Nichols introduces two examples that play a pivotal role in his discussion. One of these examples is from Gary Watson’s (1987) well-known and much-discussed account of the ruthless murderer, Robert Harris, on death row in California for multiple murders. The other example Nichols considers is taken from my own writing about the trial of a young man who assaulted and raped a teenage girl. (Kane 1996: 84; Kane 2005) The examples have similar import. But I will focus on my own example for brevity and because it brings out some key points that Watson does not emphasize. As Nichols points out, my example is roughly based on my own experience, triggered by the trial of the young man accused of the assault and rape. My initial reactions attending the trial of this young man were filled with anger and resentment against him, since we knew the family of the teenage girl who lived in our neighborhood. But as I listened daily to the testimony of how the young man came to have the mean character and perverse motives he did have—a sordid story of parental rejection, sexual abuse, bad role models and other factors (not entirely unlike the story of Harris in Watson’s example, who was abused by his father and rejected by his mother)—some of my resentment toward the young men decreased and was directed towards other persons who abused and influenced him. But—and here is a key point—I was not yet ready to shift all the blame away from the young man himself. I resisted this ‘‘transference of responsibility’’ and ‘‘blame’’ entirely to others and wondered whether some residual responsibility might not belong to the young man himself. My questions became: To what extent was he responsible for becoming the sort of person he now is? Was his behavior all a question of bad parenting, neglect and abuse, social conditioning, and like factors, or did he have any role to play in choosing it? These are crucial questions about what might be called the young man’s ultimate responsibility. We know that parenting and society, genetic makeup and upbringing, have an influence on what we become and what we are. But were these influences entirely determining, or did they ‘‘leave anything over’’ for the young man to be responsible for? The question of whether he was merely a victim of bad circumstances or had some residual responsibility for being the way he is—the question, that is, of whether he became the person he is to any degree of his own free will—seems to depend on whether these other factors were or were not entirely determining. In short, it seems to depend on whether or not it was ever causally possible for him to have resisted the influences of his upbringing and to have acted differently at some points in the course his lifetime to make himself different than he
123
Moral Responsibility, Reactive Attitudes and Freedom of…
235
now is. But if determinism were true, acting differently than he actually did at any time in his lifetime would have been causally impossible. Now one might argue here that my particular reactions at this trial to the young man, the fact that my reactive attitudes of resentment and blame towards him were mitigated to some degree and transferred to others when I learned about his sad history, were the reactions of a ‘‘philosopher’’ that involved ‘‘overintellectualizing’’ the issues, and not the reactions of ordinary folk. But this was far from being the case. My wife and I sat in this courtroom with friends and other neighbors of the young girl’s family, none of whom were philosophers. They were firemen, businesswomen, store owners, high school football coaches, teachers, and many others; and all had similar reactions to mine. Keep in mind that, like me, they all resisted mightily transferring responsibility entirely away from the young man. But their reactive attitudes, including retributive ones, were nonetheless mitigated to some degree and influenced by hearing the sordid stories of his history. Moreover, if there were any persons in that courtroom whose retributive attitudes were not in any way influenced by listening to the history of the young man (as I am sure there were), then I would not want to see them anywhere near a jury deciding the fate of persons I cared about, or any other persons whatever. For they would not be capable of responding in ways I believe would be fair to those they judge. They would not be capable of responding fairly, if they were not capable of appreciating that, to the extent that the young man’s sad history made it causally impossible for him to have turned out differently, to that extent he would not have had a ‘‘fair opportunity to avoid wrongdoing.’’
3 Freedom to Do Otherwise and Determinism: Compatibilist Responses 1 and 2 These reflections help to explain why I agree with critics of the insulation thesis that ordinary practices of holding responsible, and the reactive attitudes associated with them, such as resentment and blame, cannot be entirely insulated from traditional philosophical concerns about determinism and the causal impossibility of doing otherwise. But rejecting the insulation thesis alone does not settle the question of whether determinism and its various implications, including the causal impossibility of doing otherwise, really do, in the final analysis, threaten or require significant changes in our ordinary practices of holding persons morally responsible. Relevant here is the fact that many critics of the insulation thesis (including many of those mentioned earlier) are compatibilists about moral responsibility and determinism. They agree with Peter Strawson that ordinary practices of holding responsible and related reactive attitudes are compatible with determinism. But they argue that claiming these ordinary practices can be entirely insulated from philosophical concerns about determinism, as Strawson does in ‘‘Freedom and Resentment,’’ is too simple and easy a route to compatibilism. One must engage with the philosophical debates, they believe, and offer further arguments why, for example, the freedom to do otherwise in senses related to ordinary practices is
123
236
R. Kane
compatible with determinism or that moral responsibility does not require the freedom to do otherwise in any senses that are threatened by determinism. There are many compatibilist arguments for such claims, some of which go back to ancient times. But they have become ever more complex and sophisticated in modern times, and especially so in the past half-century since the appearance of Strawson’s seminal essay. Nonetheless, despite the complexity and sophistication of compatibilist arguments, I believe they ultimately fail to show that the freedom to do otherwise in all of the senses related to ordinary practices is compatible with determinism or that moral responsibility does not require the freedom to do otherwise in any senses that are threatened by determinism. I will now try to show this by discussing the most important of these compatibilist strategies. 3.1 Conditional Interpretations of Freedom to Do Otherwise The most common strategy to show that freedom and responsibility are compatible with determinism employed by compatibilists in the modern era, from Hobbes and Locke, to Hume and Mill, and well into the twentieth century, has been to defend conditional or hypothetical interpretations of the freedom to do otherwise. According to this classical compatibilist strategy, as it is usually called, what we mean when we say that agents were ‘‘free to do otherwise,’’ or ‘‘could have done otherwise,’’ is that ‘‘they would or might have done otherwise, if the past (or the laws of nature) had been different in some way.’’ If, for example, the persons had had different beliefs or desires, or had reasoned or chosen differently, they would or might have acted differently. And claiming persons would or might have acted differently, if the past and laws had been different in some way, it is then argued, is consistent with claiming that their acting as they did was determined, given the past and the laws as they actually were. I believe this standard compatibilist strategy is deeply flawed. Immanuel Kant, as is well known, called it a ‘‘wretched subterfuge’’ and William James a ‘‘quagmire of evasion’’; and I think they were right. A number of cogent objections have been made against such conditional interpretations of the freedom to do otherwise since the mid-twentieth century; and even many prominent compatibilists today reject them. It may be true that persons would have done otherwise, if the past or the laws had been different in some way. But the past was not different in some way; it was as it was. Likewise, the laws were not different; they were as they were. Our freedom and responsibility must be exercised in the world that actually is, not in some hypothetical or merely possible world that never actually was. And if determinism is true of this actual world in which we live and act, then acting otherwise than we do in the circumstances we actually find ourselves would always be causally impossible.2 2
Some contemporary philosophers (e.g. Smith 2003; Vihvelin 2008; Fara 2008) have put forth so-called ‘‘dispositional’’ analyses of the freedom to do otherwise, which also take a conditional or hypothetical form, but are meant to be an improvement over the traditional conditional analyses of classical compatibilists. These new dispositional analyses do avoid some of the criticisms of traditional conditional analyses, but they are also subject to the objection made to classical analyses given in this paragraph. For further criticisms of these dispositional analyses, see Clarke (2008) and Berofsky (2011). There are other
123
Moral Responsibility, Reactive Attitudes and Freedom of…
237
3.2 General Powers and Their Exercise Another compatibilist strategy frequently employed in contemporary philosophy is to dispute the relevance of causal impossibility in the sense implied by determinism to ordinary practices of excusing persons from moral responsibility.3 Ordinary excuses, it is argued on this strategy, have to do with the ‘‘general powers’’ of agents—what it is impossible for them to do, given the laws of nature and facts about their basic constitution and capacities. Thus an agent might have the ‘‘general power’’ to speak English, but not Spanish, or to run a mile, but not in less than 8 min, or to jump 2 ft in the air, but not 6 ft, and so on. By contrast, the causal impossibility defined by determinism, it is argued, concerns the causal impossibility of exercising general powers at particular times in particular circumstances. According to this compatibilist strategy, it is the lack of general powers to do certain things that excuses persons from moral responsibility and not the causal impossibility (implied by determinism) of being able to exercise general powers in particular circumstances. I believe this compatibilist strategy is also deeply flawed, as numerous examples show. Consider the following example. An explosion has taken place in a chemical factory. The employees are rushing to escape the resulting fire and the noxious chemicals released. But a heavy cabinet has fallen on one of the employees and he cannot get out from under it. Several employees try to lift the cabinet, but it is too heavy for them. They therefore recruit one of the employees who is a big man with a reputation of being extremely strong and alert him to the problem. He comes and tries to lift the cabinet but, after several failed attempts, says he cannot do so. Seeing no more options, they all flee the burning building and the man dies in the fire. After the incident, suspicions are rampant among the employees that the big man deliberately failed to lift the cabinet, though he was capable of doing so. It turns out that he was having a long-term dispute with the man trapped by the cabinet and they were rivals for a foreman’s position in the factory. The cabinet was heavy, but seemed to be well within his general powers to lift. The big man, for his part, is deeply wounded and upset by these accusations by fellow employees and insists that he tried as hard as he could to lift the cabinet and wanted to save the man under it, despite their past antagonisms. But he simply could not lift it. Friends and family of the dead man nonetheless charge that he is morally blameworthy for the man’s death, demand that he be fired and threaten to file criminal charges against him. In proceedings held by the factory owners to see if these charges are warranted, crucial facts come to light. First, it is established that the weight of the cabinet was indeed well within the big man’s general powers to lift. That supported his moral blameworthiness and perhaps even further charges. But second, it also came to light that the noxious chemical fumes filling the factory after the explosion had a Footnote 2 continued subtle conditional analyses of ‘‘could have done otherwise’’ in the contemporary literature, such as that of Lewis (1981), that I do not discuss here. I give a more thorough discussion of these views, including that of Lewis, in Kane (1996: Chapter 4). 3
The most sophisticated defense of this compatibilist strategy in contemporary philosophy is that of Wallace (1994).
123
238
R. Kane
significant effect on the nervous system and physiology of the big man, which made it causally impossible for him to lift the cabinet in those particular circumstances, even though doing so was normally within his general powers. If this second fact was true, and the big man had tried as hard as he could to lift the cabinet, as he claimed, then he would not be morally blameworthy for failing to lift the cabinet and save the man. He would be no more morally blameworthy for this outcome than the other employees who tried as hard as they could to lift it, but lacked the general power to do so. Yet the excusing condition in his case was not that he lacked the general power to do so. It was rather that it was causally impossible for him to exercise this general power in the particular circumstances he was in. It is false therefore to claim, as this compatibilist argument does, that the causal impossibility relevant to ordinary practices of excusing persons from moral responsibility has to do only with the lack of general powers to do certain things and not with the causal impossibility (implied by determinism) of being able to exercise general powers in particular circumstances. The big man lacked a ‘‘fair opportunity to avoid’’ failing to lift the cabinet and save the man just as much as the other employees. But the reason was not because he lacked the general power to lift objects of that weight, like the other employees, but because it was causally impossible for him to exercise this general power, and so he lacked the capacity to exercise it, in these circumstances. That fact, and not merely what general powers he may or may not have had, is relevant to assessing his moral blameworthiness.
4 Freedom and Moral Responsibility: Compatibilist Responses 3 and 4 4.1 Moral Responsibility and Freedom to Do Otherwise Traditionally, most arguments for compatibilism about moral responsibility, like the above, have tried to show that the freedom or power ‘‘to do otherwise’’ required for moral responsibility is not really incompatible with determinism. But another strategy that has become especially popular in recent philosophy is to argue more radically that the freedom or power to do otherwise is not required for moral responsibility at all. One influential version of this strategy is exemplified by a wellknown argument of compatibilist Dennett (1984), which is developed in terms of the historical example of Martin Luther. Dennett notes that when finally breaking with the Church of Rome, Luther is reported to have said ‘‘Here I stand, I can do no other.’’ Suppose, states Dennett, Luther was literally right about himself at that moment. Given his character and motives, he literally could not then have done otherwise. Does this mean he was not morally responsible? Not at all, Dennett answers. In saying ‘‘I can do no other,’’ Luther was not disowning responsibility for his act, but taking full responsibility for it, even though he could not at that moment have done otherwise. From this example and others like it, Dennett draws the general conclusion that the power to do otherwise is not required for moral responsibility.
123
Moral Responsibility, Reactive Attitudes and Freedom of…
239
Now Dennett is right here, I believe, to claim that Luther could have been, and likely was, morally responsible for his act, even if he could not have done otherwise at the time and even if his act was determined by the state of his will and other circumstances at the time he performed it. But it does not follow from these facts alone that the freedom or power to do otherwise is not required at any times in an agent’s lifetime, if agents are to be morally responsible to some degree for the state of their wills when they act. Biographers of Luther have chronicled the many internal struggles of conscience requiring exercises of willpower he experienced over a long period of time that brought him to this point where he ‘‘could do no other.’’ Similarly, agents, in general, may often act as morally responsible agents from a will (character, motives and purposes) already formed and could not have willingly done otherwise at those times. But to be responsible to some degree for the state of the wills from which they act at such times, I would argue, their wills must have been formed to some degree by other choices or actions in their past for which they could have done otherwise and which were not determined. For if this were not so, it would always have been causally impossible for them to have done anything differently in their entire lifetimes to make their wills (characters, motives and purposes) different than they are and thus to ever willingly act differently than they do. This particular compatibilist strategy, in other words, involves a subtle ‘‘fallacy of composition’’: Inferring from the fact that we can be responsible for some, and even many, choices or actions for which we could not have done otherwise and which were determined by our wills when we performed them, to the claim that this could be true of all of the choices and actions in our lifetimes, if we are to be morally responsible for the state of our wills. What is needed here, as suggested earlier, is a distinction between freedom of action and freedom of will.4 Agents can be responsible for many acts in their lives where, like Luther, they are free to act at the time in the sense that nothing prevents them from doing what they will to do and their wills are already determined or ‘‘set one way’’ on doing what they will to do. And they can be responsible for such acts even when they could not at the time have willingly done otherwise and even if the acts were determined by the state of their wills and other circumstances at the time they performed them. But this could not be true of all of the acts in their lifetimes if they were ever to be responsible for the state of their wills when they act (that is, if they were to have freedom of will as well as freedom of action). For this, there would have to be some choices or acts in their lifetimes that were such that their wills were not already determined or ‘‘set one way’’ before they performed them. Rather, they would ‘‘set their wills’’ one way or another in the very act of making these choices or performing these actions in a manner that was not already determined by their preexisting wills at the time. I call such ‘‘will-setting’’ (as opposed to ‘‘will-settled’’) choices or actions ‘‘self-forming actions’’ (SFAs); and they are required I believe at some times in our lives if we are not only to be responsible for acting as we will, but 4
One other contemporary philosopher who makes a point of distinguishing freedom of action and will is Frankfurt (1971), who does so in terms of higher and lower order desires. While his distinction between different orders of desires is important for a full account of freedom of action and will, I have argued elsewhere (Kane 1996) that it does not give us the whole story about free will.
123
240
R. Kane
for willing to act as we do. If no acts of agents were will-setting or self-forming in this way, if their wills were already determined and ‘‘set one way’’ whenever they chose or acted, it would have been causally impossible for them to have ever willingly chosen or acted differently than they in fact did. 4.2 Moral Responsibility and Frankfurt-style Examples These reflections bring us to yet another and perhaps the most subtle compatibilist strategy for arguing that the freedom or power ‘‘to do otherwise’’ is not required at all for moral responsibility. This strategy, which has become especially popular among contemporary philosophers, is to appeal to so-called ‘‘Frankfurt-style examples’’ (FSEs), named after Frankfurt, who first put them forward. (Frankfurt 1969) Debates about these FSEs have become so complex in recent philosophy that it is impossible to discuss them in full detail here. I have done so elsewhere. (Kane 1985, 1996, 2013) But I will try to state as briefly as I can why I believe they also fail to show that the freedom to do otherwise is not required at all for moral responsibility. Frankfurt’s aim in formulating these examples was to refute a principle that he called the Principle of Alternative Possibilities (PAP): ‘‘Agents are morally responsible for what they have done, only if they could have avoided doing it, or could have done otherwise.’’ To refute this principle, Frankfurt imagines a controller, called Black, who has direct control over the brain of another man, Jones, and who wants to allow Jones to do only what Black wants him to do. But Black prefers to allow Jones to act on his own whenever possible. So he prefers to intervene, only if Jones is about to do something that he, Black, does not want. Frankfurt then imagines situations in which Jones is about to do what Black wants, so that Black does not intervene. In such situations, Frankfurt argues, Jones could be morally responsible for acting as he does, since he acted on his own, from his own motives, and nothing and no one (including Black) interfered with or prevented him from doing what he chose to do. Yet Jones in such situations could not have done otherwise, for if he had given any indication of doing otherwise, Black would have prevented him from doing so. Thus PAP, Frankfurt concludes, is false: It is not true that agents can be morally responsible for what they have done, only if they could have avoided doing it, or could have done otherwise at the time they acted. Now the first thing to be said about this argument is that it should not surprise us at all that PAP is false. For we can already see from the discussion of the Luther example and many others like it that agents can be morally responsible for actions that are determined by their wills at the time they acted and therefore such that they could not have done otherwise. One can be responsible for many ‘‘will-settled’’ actions, even if one could not have done otherwise than perform them. But not all of the actions in our lifetimes could be determined and already will-settled in this way when we act, if we are ever to be responsible for our wills being set the way they are when we act. For this to be the case, at some times in our lives, we would have to be capable of not merely ‘‘will-settled,’’ but of ‘‘will-setting’’ or ‘‘self-forming’’ choices or actions (SFAs), which were not determined by our existing wills when we performed them and which were such that we could have willingly done
123
Moral Responsibility, Reactive Attitudes and Freedom of…
241
otherwise when we performed them. And it can be shown that if all of our actions were under the control of a Frankfurt controller, there could be no such self-forming choices or actions (SFAs). The reason that such SFAs could not be under the control of a Frankfurt controller is the following: Such will-setting actions would have to be such that they were not determined by our existing wills when we performed them (hence not willsettled) and such that we could have willingly (voluntarily and intentionally) performed them and willingly done otherwise. Now consider choices that satisfy these conditions. Since they would be undetermined up to the time they are made, a Frankfurt controller could not know in advance which way the agents were going to choose before they actually did so. He would not know this until the agents had actually chosen one way or the other; and it would then be too late to prevent them from doing what he did not want. So if, on the one hand, the controller waited till they chose, he would no longer have control over which way they chose. But if, on the other hand, he did not wait until they chose, but intervened in advance to make them do what he wanted, he would be preventing them from having SFAs, and hence from having the power to set their own wills in one way or another. He would have preempted their choice. The agents could not therefore have the capacity and opportunity to make SFAs, and hence engage in their own ‘‘will-setting,’’ and at the same time be under the control of a Frankfurt controller.5 To summarize, Frankfurt-style examples (FSEs) were proposed to refute the principle PAP: ‘‘Agents are morally responsible for what they have done, only if they could have avoided doing it, or could have done otherwise.’’ Such examples do show that PAP is false. But we have reasons to believe from many other considerations as well that PAP is false. What FSEs do not refute, however, is another more complex principle that might be stated as follows: ‘‘Agents are responsible for willing to act as they do, only if at some times in their lives, they willingly (voluntarily and intentionally) perform certain (‘‘will-setting’’ or ‘‘selfforming’’) actions that it was causally possible at the time for them to have willingly avoided performing.’’ And such actions (SFAs) by which agents form the wills on which they subsequently act in a manner that is not determined by their pre-existing wills could not have been under the control of a Frankfurt controller, for the reasons given in the previous paragraph: Any action that is under the control of a Frankfurt controller is not such that the agent is both free to willingly perform it and free to willingly do otherwise. Even when the controller does not actually intervene, his presence (given his powers and intentions) makes it impossible for the agent to have willingly done otherwise. 5
I first suggested this line of argument against FSEs in Kane (1985) and further developed it in Kane (1996). Similar arguments were later independently developed by Widerker (1995), Ginet (1996) and Wyma (1997). My argument differs from theirs in several respects, however, not always noted in the literature: Their arguments are put forward as defenses of PAP, whereas mine is not; and my argument gives a role to SFAs that theirs does not. A number of new and more complex FSEs have since been put forward by various philosophers in response to this line of argument, including various FSEs involving ‘‘blockage’’ (e.g., by Mele and Robb 1998) and other so-called ‘‘buffered’’ FSEs. (e.g., by Pereboom 2001) In Kane (2013), I show that these new blockage and buffered FSEs are also incompatible with an agent’s being able to engage in will-setting SFAs.
123
242
R. Kane
5 Is an Undetermined Free Will Possible? Compatibilist Response 5 These thoughts bring us to the final line of argument for the compatibility of moral responsibility and determinism that complements all the other arguments and is regarded by most compatibilists and many others as decisive. It involves appealing to an ancient dilemma: If the ‘‘freedom to do otherwise’’ alledgedly required for moral responsibility (freedom of the will) is not compatible with determinism, it would not be possible at all, since it could not be compatible with indeterminism either. Undetermined events, it has been argued since the time of the ancient Stoics, would occur spontaneously and hence could not be controlled by agents in the way that free and responsible actions require. If, for example, a choice occurred by virtue of some undetermined (say, quantum) events in one’s brain, it would seem a fluke or accident rather than a responsible choice. Such undetermined events occurring in our brains or bodies would not seem to enhance our freedom and control over, and hence responsibility for, our actions, but rather to diminish our freedom, control and responsibility. Arguments such as these and many others, have historically led to a variety of familiar charges that undetermined choices or actions would be ‘‘arbitrary,’’ ‘‘random,’’ ‘‘irrational,’’ ‘‘uncontrolled,’’ or ‘‘mere matters of luck’’ or ‘‘chance,’’ and hence not free and responsible choices or actions at all. And so it would be, it might be argued, with the ‘‘will-setting’’ or ‘‘self-forming’’ actions (SFAs) of the previous section, which it was said would be required for us to be ultimate determiners of what it is we will to do and hence how we willingly act. If such selfforming actions had to be undetermined, their occurring one way or another would appear to be just a matter of luck or chance, and hence they could not be free and responsible actions. For such reasons, it has often been argued that for us to be the ultimate sources of our wills, as such an undetermined or libertarian free will would require, is an incoherent and impossible ideal. In response to these familiar charges over the centuries, defenders of such a libertarian free will, have invoked various unusual forms of agency or causation to explain how agents could exercise such a free will over events that are causally undetermined. They have appealed to immaterial minds, noumenal selves outside space and time, special forms of agent-causation that cannot be reduced to scientific modes of causation, transempirical power centers, uncaused causes, prime movers unmoved, and like factors that could not be accounted for by ordinary modes of explanation familiar to the natural and human sciences. But these common strategies have in turn led to charges of mystery and obscurity against their view. Strawson himself rejects all such libertarian strategems, referring to them as ‘‘the obscure and panicky metaphysics of libertarianism.’’ (Strawson 1962: 25) Can one make sense of the undetermined will-setting SFAs that I have argued are required at some times in our lives, if we had to be morally responsible for the state and quality of our wills? Can one make sense in general of free actions that are morally responsible and yet undetermined in a way that avoids their being reduced to matters of mere luck or chance, on the one hand, or to mystery, on the other? Though it is fair to say that the majority view among contemporary philosophers
123
Moral Responsibility, Reactive Attitudes and Freedom of…
243
and scientists is that this cannot be done, I believe it can be done; and I have attempted to do so in a series of writings over the past four decades.6 What I learned in the process, however, was that no simple answers to these questions would be forthcoming, if any answers were possible at all. Attempting to make sense of undetermined free and responsible actions that are not reducible to mere chance or luck, or to mystery, it turned out, would require piecing together a ‘‘complex tapestry’’ of ideas and arguments that would involve rethinking the relations of a host of interrelated notions: agency, choice, action, will, control, power, responsibility and others. Adequately explaining this complex tapestry in the present paper is not possible. So I must refer you to other writings (cited in note 6, above) for you to make your own judgments about whether a freedom of will that is incompatible with determinism is indeed intelligible. What I have attempted to do in this paper is show why it is important to try to do this, if we are to do full justice to our ordinary practices of holding persons responsible for their actions and their wills to perform them. What can also be stated here, however, about this complex tapestry is that many of the ideas developed in this paper have an essential role to play in it. These include, among others, the distinction between freedom of action and freedom of will; the important role of ‘‘self-forming actions’’ or SFAs; the distinction between ‘‘will-setting’’ choices and actions, on one hand, and ‘‘will-settled’’ ones, on the other; the distinction between ‘‘general powers,’’ on the one hand, and the capacity and opportunity to exercise such general powers in particular circumstances, on the other; the idea of a ‘‘fair opportunity to do otherwise’’; and the relation of all of these notions to our ordinary practices of holding persons morally responsible. In conclusion, I want to add one further thought about the role played by the reactive attitudes in this tapestry of ideas and in the arguments of this paper. To do so, return for a moment to the story of the young man on trial discussed earlier. I stated that when my wife and I and other neighbors and friends of the victim’s family heard the sordid details of the young man’s background in the courtroom, our initial resentment and anger toward the young man decreased to some degree and were directed towards other persons who brutally abused him. But I also added that we were not yet ready to shift all the blame away from the young man himself. We resisted this ‘‘transference’’ of responsibility, blame and other reactive attitudes entirely to others and questioned whether some residual responsibility might not belong to the young man himself. We wondered whether his behavior was all a matter of bad parenting, societal neglect, social conditioning, and like factors, or whether he had any underived role to play in choosing it.
6
See, e.g., Kane (1996, 1999, 2002, 2011, 2014), among other of my works listed in the references. Part of the task of making sense of such a free will involves considering the empirical and scientific question whether any indeterminism is there in the brain in ways appropriate for free will. No purely philosophical theory can settle that matter. It is interesting, however, that in the past decade there has been more openness and discussion on the part of some scientists and philosophers about this possibility. See, e.g, Bishop (2011), Baker and Gollub (1990), Hobbs (1991), Kellert (1993), Balaguer (2010), Heisenberg (2013), Glimcher (2005), Maye et al. (2007), Hameroff and Penrose (1996), Shadlen (2014), Brembs (2011), Stapp (2007), Tse (2013), Jedlicka (2014), and Briegel and Mueller (2015).
123
244
R. Kane
There was, however, a further feature of this story worth considering that was not emphasized earlier. It concerns this notion of transference of resentment and blame to other persons. Not only was there some resistance on the part of ourselves and our neighbors to transferring responsibility and blame entirely away from the young man, but when we did so, we looked to transfer our resentment and blame to some other persons, in this case to those who abused him. This natural tendency ‘‘to look for other persons to blame’’ reminds us of something significant that Peter Strawson emphasizes about the reactive attitudes. They are personal attitudes that are only appropriately directed toward persons who are responsible agents. But this feature of the reactive attitudes gives rise to an interesting question: What if there were no other persons to whom responsibility and blame might be transferred? Suppose it was learned that the young man’s vicious dispositions and behavior were determined by bad genes and defective fetal development, so that responsibility and blame could not be transferred to any other persons. The young man’s behavior was determined by these natural causes and no other persons were responsible. Some compatibilists argue at this point that there is an important difference between being determined by natural causes simply and being determined by the responsible actions of other persons. In the latter case, they argue, responsibility and blame can be transferred to those other persons. But in the former case, no such transference of responsibility from the agent to others is possible. For, one cannot blame nature, they argue, because nature is not a person. True enough. I think we must agree with that. But does it follow that, if we cannot transfer blame for the young man’s perverse will and behavior to nature or to any other persons, we can then blame the young man? That, I believe, would be perverse reasoning indeed. For, as argued throughout this paper, the crucial question in either case would be whether the young man himself was in any way responsible for the circumstances that determined him to do what he did, whether those circumstances were due to other persons or to natural causes alone. For, in either case, if these circumstances were determining, it would have been causally impossible for him to have avoided being the way he is and doing what he did; and he would thus have lacked a ‘‘fair opportunity to have done otherwise.’’
References Baker, G., and J. Gollub. 1990. Chaotic dynamics: An introduction. Cambridge: Cambridge University Press. Balaguer, Mark. 2010. Free will as an open scientific problem. Cambridge: The MIT Press. Bishop, Robert C. 2011. Chaos, indeterminism, and free will. In The Oxford handbook of free will, 2nd ed, ed. R. Kane, 84–100. Oxford: Oxford University Press. Brembs, B. 2011. Towards a scientific concept of free will as a biological trait. Proceedings of the Royal Society B: Biological Sciences 278: 930–939. Briegel, Hans J., and Thomas Mueller. 2015. A chance for attributable agency. Minds and Machines 25: 261–279. Brink, David O., and Dana K. Nelkin. 2013. Fairness and the architecture of responsibility. In Oxford studies in agency and responsibility, vol. 1, ed. D. Shoemaker, 284–313. Oxford: Oxford University Press.
123
Moral Responsibility, Reactive Attitudes and Freedom of…
245
Clarke, Randolph. 2008. Dispositions, abilities to act, and free will: The new dispositionalism. Mind 118: 323–351. Dennett, Daniel C. 1984. Elbow room: The varieties of free will worth wanting. Cambridge: The MIT Press. Fara, Michael. 2008. Masked abilities and compatibilism. Mind 117: 843–865. Fischer, John Martin. 2014. Peter Strawson and the facts of agency. In Oxford studies in agency and responsibility, vol. 2, ed. D. Shoemaker, and Neal A. Tognazzini, 93–116. Oxford: Oxford University Press. Frankfurt, Harry. 1969. Alternative possibilities and moral responsibility. The Journal of Philosophy 66: 829–839. Frankfurt, Harry. 1971. Freedom of the will and the concept of a person. The Journal of Philosophy 68: 5–20. Ginet, Carl. 1996. In defense of the principle of alternative possibilities: Why I don’t find Frankfurt’s argument convincing. Philosophical Perspectives 10: 403–417. Glimcher, P.W. 2005. Indeterminacy in brain and behavior. Annual Review of Psychology 56: 25–56. Hameroff, Stuart, and Roger Penrose. 1996. Conscious events as orchestrated space-time selections. Journal of Consciousness Studies 3: 36–53. Hart, H.L.A. 1970. Punishment and responsibility. Oxford: The Clarendon Press. Hobbs, J. 1991. Chaos and indeterminism. Canadian Journal of Philosophy 21: 141–164. Heisenberg, Martin. 2013. The origin of freedom in animal behavior. In Is science compatible with free will?, ed. A. Suarez, and P. Adams, 95–103. Dordrecht: Springer. Jedlicka, Peter. 2014. Quantum stochasticity and (the end of) neurodeterminism. In Quantum physics meets the philosophy of mind, ed. A. Corradini, and U. Meixner, 183–197. Berlin: DeGruyter. Kane, Robert. 1985. Free will and values. Albany: The State University of New York Press. Kane, Robert. 1996. The significance of free will. Oxford: Oxford University Press. Kane, Robert. 1999. Responsibility, luck, and chance: Reflections on free will and indeterminism. The Journal of Philosophy 96: 217–240. Kane, Robert. 2002. Some neglected pathways in the free will labyrinth. In The Oxford handbook of free will, 1st ed, ed. R. Kane, 406–437. Oxford: Oxford University Press. Kane, Robert. 2005. A contemporary introduction to free will. Oxford: Oxford University Press. Kane, Robert. 2011. Rethinking free will: New perspectives on an ancient problem. In The Oxford handbook of free will (2nd edition), ed. R. Kane, 381–404. Oxford: Oxford University Press. Kane, Robert. 2013. Frankfurt-style examples and self-forming actions. In Free will and moral responsibility, ed. I. Haji, and J. Caouette, 58–73. Cambridge: Cambridge Scholars Publishing. Kane, Robert. 2014. New arguments in debates on libertarian free will: Responses to contributors. In Libertarian free will: contemporary debates, ed. David Palmer, 179–214. Oxford: Oxford University Press. Kellert, S. 1993. In the wake of chaos. Chicago: The University of Chicago Press. Lewis, David. 1981. Are we free to break the laws? Theoria 47: 113–121. Maye, A., C.-H. Hsieh, G. Sugahara, and B. Brembs. 2007. Order in spontaneous behavior. PLoS ONE 2: e443. doi:10.1371/journal.pone.00066443. Mele, Alfred, and David Robb. 1998. Rescuing Frankfurt-style cases. The Philosophical Review 107: 97–112. Nagel, Thomas. 1986. The view from nowhere. Oxford: Oxford University Press. Nichols, Shaun. 2015. Bound: Essays on free will and responsibility. Oxford: Oxford University Press. Pereboom, Derk. 2001. Living without free will. Cambridge: Cambridge University Press. Russell, Paul. 2011. Moral sense and the foundations of responsibility. In The Oxford handbook of free will, 2nd ed, ed. R. Kane, 199–220. Oxford: Oxford University Press. Shadlen, Michael. 2014. Comments on Adina Roskies, ‘Can neuroscience resolve issues about free will?’. In Moral psychology, vol. 4, ed. W. Sinnott-Armstrong, 139–150. Cambridge: The MIT Press. Smith, Michael. 2003. Rational capacities, or: How to distinguish recklessness, weakness, and compulsion. In Weakness of will and practical irrationality, ed. S. Stroud, and C. Tappolet, 17–38. Oxford: Oxford University Press. Stapp, Henry. 2007. The mindful universe. Berlin: Springer. Strawson, Galen. 1986. Freedom and belief. Oxford: The Clarendon Press. Strawson, P.F. 1962. Freedom and resentment. Proceedings of the British Academy 48: 1–25. Tse, Peter Ulric. 2013. The neural basis of free will: Criterial causation. Cambridge: The MIT Press.
123
246
R. Kane
Vihvelin, Kadri. 2008. Foreknowledge, Frankfurt, and the ability to do otherwise: A reply to Fischer. Canadian Journal of Philosophy 38: 343–372. Wallace, R.J. 1994. Responsibility and the moral sentiments. Cambridge: Harvard University Press. Watson, Gary. 1987. Responsibility and the limits of evil: Variations on a Strawsonian theme. In Responsibility, character, and the emotions, ed. F. Schoeman, 256–286. Cambridge: Cambridge University Press. Widerker, David. 1995. Libertarianism and Frankfurt’s attack on the principle of alternative possibilities. The Philosophical Review 104: 247–261. Wyma, Keith. 1997. Moral responsibility and the leeway for action. American Philosophical Quarterly 34: 57–70.
123
