Galen Strawson (Reference Strawson1994, Reference Strawson and Kane2002, Reference Strawson2010, Reference Strawson and Craig2011) formulates in various ways his Basic Argument for the claim that true or ultimate moral responsibility is impossible, whether or not causal determinism obtains. Here is his briefest formulation: Because self-creation is required for being truly morally responsible and self-creation is impossible, it is impossible to be truly morally responsible for anything. Let us consider my simplified reconstruction of Strawson's most informative formulation:
Reasons Premise: An agent S’s intentionally performing an action A for which she might be truly morally responsible is explained by certain features of her mental constitution MC—namely, certain reasons for acting.
Responsibility Premise: S is truly morally responsible for an intentional action A only if S is truly morally responsible for the parts of her MC that explain her performing A, and S is truly morally responsible for her MC only if S is truly morally responsible for an earlier action A 1 in which S intentionally and successfully brought about those parts of her MC.
Iteration Premise: S is truly morally responsible for A by way of MC and A 1 as previously described only if S has performed an infinite number of even earlier free actions. (After all, S is truly morally responsible for A 1 only if S is truly morally responsible for the parts of her MC 1 that explain her performing A 1, and S is truly morally responsible for her MC 1 only if S is truly morally responsible for an even earlier action A 2 in which S intentionally and successfully brought about those parts of her MC 1. S is truly morally responsible for A 2 only if S is truly morally responsible for the parts of her MC 2 that explain her performing A 2, and S is truly morally responsible for her MC 2 only if S is truly morally responsible for an even earlier action A 3 in which S intentionally and successfully brought about those parts of her MC 2. S is truly morally responsible for A 3 only if and so on).
Impossibility Premise: It is impossible for finite beings like us to have performed an infinite number of past actions.
Conclusion: It is impossible for finite beings like us to be truly morally responsible for anything.
What these formulations have in common is that being truly morally responsible for something requires a kind of self-determination that is impossible for us to have.
Strawson (Reference Strawson1994: 5, 8) has lamented that his Basic Argument is underappreciated. Recently, however, the Basic Argument, or something relevantly like it, has received broad endorsement (Hendrickson Reference Hendrickson2007; Istvan Reference Istvan2011; Kershnar Reference Kershnar2015; Kment Reference Kment2017; Nagel Reference Nagel1986: 118; Smilansky Reference Smilansky2000: ch. 4; Waller Reference Waller2011: 19–42). Others have noted their sympathy for it (Henderson Reference Henderson2014: 328; Mickelson Reference Mickelson2015, Reference Mickelson, Griffith, Timpe and Levy2017).
Those of us, however, who believe that being truly morally responsible is at least possible must find fault with one of these premises. The reasons premise is uncontentious. After all, there is widespread agreement that actions for which we are truly morally responsible are actions that are performed for reasons. The impossibility premise is obviously true, and the iteration premise is just an entailment of the responsibility premise. Thus, the possibilist should take aim at the responsibility premise.
Possibilists have provided at least three reasons to doubt the responsibility premise. One reason is that agent-causal libertarianism is possible and agents with agent-causal power have the right kind of control over their actions to be truly morally responsible for them without also being truly morally responsible for features of their mental constitutions that explain their actions (Clarke Reference Clarke1997; Corabi Reference Corabi2017; O'Connor Reference O'Connor and Kane2011: 320–21; Pereboom Reference Pereboom2001: 65–68; Tucker Reference Tucker2007). Another reason involves a G. E. Moore shift against the responsibility premise: It is more plausible that possibly we are truly morally responsible agents than that the responsibility premise is true (Coffman Reference Coffman2015: 113–15; Fischer Reference Fischer2006: 112). A third reason is that that the responsibility premise is implausible, because it is demanding in a way that outstrips commonsense requirements of true moral responsibility (Clarke Reference Clarke2005; Fischer Reference Fischer2006; Kane Reference Kane1996; McKenna Reference McKenna, Trakakis and Cohen2008; Mele Reference Mele1995: 221–30; Russell Reference Russell2008; Wolf Reference Wolf1990, Reference Wolf2015).
My argument is situated roughly in the third camp. In what follows, I offer three arguments for the claim that the responsibility premise is unjustified and implausible, none of which appeal to agent-causal libertarianism or a G. E. Moore shift.
In the first section, I briefly explicate Strawson's (Reference Strawson1994: 6–7) idea of true or ultimate moral responsibility. In the second section, I argue that the moral luck debate shows that the responsibility premise is contradicted by part of our commonsense ideas about moral responsibility and appears to be supported by another part, and that this ambivalence undercuts the only reason that Strawson gives for thinking that the responsibility premise is true. In the third section, I argue that the responsibility premise has no foundation in common sense. The responsibility premise is so demanding that either it is an implausible requirement for the true moral responsibility that we take ourselves to have, or it is a plausible requirement of a species of true moral responsibility that we have never taken ourselves to have. In the fourth section, I suggest that the intuitive attractiveness of the responsibility premise is the result of a generalization from instances of constitutive luck that obviously undermine true moral responsibility to all instances of constitutive luck. But this is an overgeneralization precisely because certain kinds of constitutive luck are benign with respect to true morally responsible agency and other kinds at least partially enable it. The joint force of these arguments is that the responsibility premise is unjustified and implausible and that the Basic Argument is unpersuasive and unsound.
1. True Moral Responsibility
Strawson is specific about the kind of true moral responsibility that the Basic Argument targets. He writes,
What sort of ‘true’ moral responsibility is being said to be both impossible and widely believed in?
An old story is very helpful in clarifying this question. This is the story of heaven and hell. As I understand it, true moral responsibility is responsibility of such a kind that, if we have it, then it makes sense, at least, to suppose that it could be just to punish some of us with (eternal) torment in hell and reward others with (eternal) bliss in heaven. The stress on the words ‘makes sense’ is important, for one certainly does not have to believe in any version of the story of heaven and hell in order to understand the notion of true moral responsibility that it is being used to illustrate. … The story of heaven and hell is useful simply because it illustrates, in a peculiarly vivid way, the kind of absolute or ultimate accountability or responsibility that many have supposed themselves to have, and that many do still suppose themselves to have. It very clearly expresses its scope and force. (Reference Strawson1994: 6–7, italics in original; cf. Reference Strawson and Kane2002: 451–53; Reference Strawson2010: 2)
In other words, the ‘old story’ need not be true for Strawson's purposes. It merely illustrates the scope and force of the widely believed conception of responsibility that his argument targets.
Strawson (Reference Strawson and Kane2002: 452; cf. Reference Strawson2010: 2) appeals to heaven-or-hell responsibility to illustrate that the desert at issue is basic in the following way: ‘URD [ultimate responsibility-or-deservingness] is responsibility and desert of such a kind that it can exist if and only if punishment and reward can be fair or just without having any pragmatic justification, or indeed any justification that appeals to the notion of distributive justice’. Strawson (Reference Strawson2010: 2) calls the preceding quotation a ‘less dramatic but equally effective way of characterizing true moral responsibility’ in comparison with the heaven-or-hell story. Thus, heaven and hell are used merely to illustrate that what a person deserves is based on only backward-looking considerations. But then, reference to heaven and hell is not essential to characterizing the nature of true moral responsibility (Corabi Reference Corabi2017; Istvan Reference Istvan2011: 401). And since several philosophers have rightly noted that the appeal to heaven and hell gratuitously introduces confusion into the scope and force of the kind of responsibility at issue (Clarke Reference Clarke2005: 21–22; Fischer and Tognazzini Reference Fischer and Tognazzini2011: 25–30; Levy Reference Levy2011: 3), I consider true moral responsibility as a kind of basic desert and refer to it without reference to heaven or hell.
One more clarification about true moral responsibility is worth making at this point. Strawson (Reference Strawson2010: 15) contends that although compatibilist conceptions of moral responsibility, such as reasons-responsive views, can satisfy ‘many of the demands of our ordinary notion of freedom’, they cannot satisfy the demand for the kind of ‘self-determination’ that he believes is required for true moral responsibility. As Strawson (Reference Strawson2010: 15) writes, ‘The function of the qualifiers “true” and “truly” is simply to keep this point clear: whatever it may be, compatibilist responsibility is not true responsibility in the present sense’. I neither rehearse nor assess Strawson's (Reference Strawson2010: 107–47; cf. Reference Strawson1994: 16–17) arguments for these claims. For dialectical purposes, I grant that true moral responsibility is incompatibilist to show that my arguments have traction against the Basic Argument even if Strawson's arguments against compatibilism succeed. Nevertheless, my arguments do not presuppose incompatibilism; compatibilists can use them too.
In the next section, I undercut Strawson's justification for the responsibility premise by situating it within the broader debate about moral luck.
2. The Ambivalence Argument from Moral Luck
Strawson (Reference Strawson1994: 12; cf. Reference Strawson and Kane2002: 454, 445) asserts that the responsibility premise is a part of common sense: ‘We find, semi-dormant in common thought, an implicit idea that true or ultimate moral responsibility for what one does somehow involves responsibility for how one is’. Strawson (Reference Strawson1994: 11; cf. Reference Strawson1994: 17–18; Reference Strawson and Kane2002: 454) gives ‘some examples of people who have accepted that some sort of true or ultimate responsibility for the way one is is a necessary condition of true or ultimate moral responsibility for the way one acts’. His list includes Robert Kane, Immanuel Kant, and Jean-Paul Sartre among others. Thus, Strawson thinks that the responsibility premise is a commonsense requirement for true moral responsibility, and he illustrates that claim by appealing to the philosophy of Kane, Kant, and Sartre.Footnote 1
But it is not obvious whether common moral intuitions in fact justify the responsibility premise, and we can see why by putting the Basic Argument into the broader dialectic of the problem of moral luck, which is something that philosophers have neglected to do.Footnote 2
The problem of moral luck is that there is a contradiction in our commonsense ideas about true moral responsibility (Hartman Reference Hartman2017: 1). In one strand of our thinking, we believe that a person can become more praiseworthy or blameworthy by luck or in a way influenced by luck. To make this idea concrete, consider some cases of luck in result, circumstance, and constitution:Footnote 3
Resultant Luck Case: Two agents drive recklessly around a curb in the same way, and one but not the other kills a pedestrian (Nagel Reference Nagel1979: 29).
Circumstantial Luck Case: Two judges would freely take a bribe if one were offered. By luck of the courthouse draw, only one judge is offered a bribe, and so only one judge takes a bribe (Thomson Reference Thomson1989: 214).
Constitutive Luck Case: Two citizens would freely help a beggar if they had a good upbringing, but they were habituated differently. The citizen with good habituation stops to help the beggar, and the citizen with bad habituation ignores the beggar.
In each case, luck is the salient difference between the two agents. After all, the spatial location of the pedestrian is outside of each driver's control; being offered a bribe is outside of each judge's control; and being a recipient of good habituation is outside of each citizen's control. In our everyday lives, however, we make moral judgments that imply that the killer driver is more blameworthy than the merely reckless driver; the bribe taker is more blameworthy than the mere would-be bribe taker; and the helper is more praiseworthy than the mere would-be helper (Nagel Reference Nagel1979). (By the phrase ‘S is more blameworthy than S*’, I mean that S deserves more blame than S*, but perhaps it is also true that S is more deserving of blame than S*.) This is because we believe that certain kinds of luck in results, circumstance, and constitution can positively affect praiseworthiness and blameworthiness. Call the intuitions that represent this part of common sense the moral luck intuition.
Nevertheless, the moral luck intuition contradicts another feature of our ordinary thinking about true moral responsibility captured in this control principle: A person's praiseworthiness and blameworthiness is restricted to what is within her control, and her praiseworthiness and blameworthiness cannot be affected by that which is not within her control—whether it be the results of her actions, the circumstances she faces, or features of her constitution (Zimmerman Reference Zimmerman2002: 559; see also Nagel Reference Nagel1979). Bernard Williams (Reference Williams1985: 194) explains it this way: ‘There is pressure within it [our ordinary conception of morality] to require a voluntariness that will be total and will cut through character and psychological or social determination, and allocate blame and responsibility on the ultimately fair basis of the agent's own contribution, no more and no less’. True attributions of praiseworthiness and blameworthiness reflect only a person's pure or luck-free agency. Thus, the control principle implies that the drivers are equally blameworthy, the judges are equally blameworthy, and the citizens are equally praiseworthy, because the salient difference between the agents in each case pair is luck, and luck is irrelevant to true moral responsibility. Call the intuitions that represent this part of common sense the luck-free intuition.
Our commonsense ideas about true moral responsibility, then, appear to include both the moral luck and luck-free intuitions. A prominent way to understand the problem of moral luck is that it is not clear which intuition is correct. Thomas Nagel (Reference Nagel1979: 34), who is one of the founders of the contemporary moral luck debate (the other one is Williams Reference Williams and Williams1981), leaves the problem of moral luck as a ‘paradox’, because both parts of common sense appear to him to be undeniable.
Let us relate the problem of moral luck to the Basic Argument, which Strawson (Reference Strawson2000: 151; cf. Reference Strawson and Kane2002: 449) himself frames in terms of luck at a certain point: ‘One can put the point (somewhat contentiously) by saying that in the final analysis the way you are is, in every last detail, a matter of luck—good or bad’.
If the moral luck intuition is right, then it rules out the truth of the responsibility premise, because it implies that an agent's constitutive luck—that is, features of the agent's constitution that she did not bring about—may explain why she performs an action for which she is truly morally responsible. For example, the citizen is praiseworthy for helping the beggar even though his lucky constitution explains why he chooses to help. But the responsibility premise implies that the citizen is not praiseworthy for helping the beggar, because, on that view, she is not truly morally responsible for even part of the habituation that explains why she helps the beggar. Thus, at least some of our commonsense ideas about true moral responsibility contradict the responsibility premise.
If the luck-free intuition tracks the truth, then it would support the idea that the responsibility premise is true. The control principle implies that what is not within the agent's control, such as nonvoluntarily acquired parts of her constitutive mental life, cannot affect her true moral responsibility for the action, and the responsibility premise implies that the features of a person's constitutive mental life that saliently explain the action for which she is truly morally responsible must be constitutive properties that she has intentionally and successfully brought about by previous actions. Thus, at least some of our commonsense ideas about true moral responsibility appear to support the responsibility premise, and this fact helps explain why the responsibility premise is intuitively attractive.
The upshot is that our moral intuitions do not obviously justify the responsibility premise, because part of common sense implies that the responsibility premise is false and another part of common sense appears to support it. For this reason, it is not enough for Strawson merely to appeal to our commonsense ideas about true moral responsibility to justify the responsibility premise.Footnote 4 He needs to tell us why only the luck-free intuition is right. As his presentation of the Basic Argument currently stands, however, Strawson provides us with no all-things-considered reason to believe that the responsibility premise is true. The ambivalence argument from moral luck, then, shows that the Basic Argument is not persuasive, because the most crucial premise in the Basic Argument suffers from a stalemate of intuitions.Footnote 5
One might object that the moral luck intuition cannot track true moral responsibility, because the moral luck intuition is compatibilist and true moral responsibility is incompatibilist. But the objector would be mistaken. The claim that certain kinds of luck in results, circumstance, and constitution can affect true moral responsibility is consistent with both compatibilism and incompatibilism about true moral responsibility and causal determinism. The moral luck intuition is consistent with incompatibilism, because there is an important difference between cases of luck in circumstance and constitution, on the one hand, and luck in causal determination—that is, causal luck—on the other.Footnote 6 The difference is that causal luck is sufficient to bring about the action that it affects, but circumstantial and constitutive luck are mere causal contributors and thus are not sufficient to bring about the action that they affect. Thus, even if certain kinds of circumstantial and constitutive luck can affect true moral responsibility, it does not follow necessarily that the same is the case for causal luck. I should note, however, that there is an interesting relationship between the moral luck intuition and compatibilism. In my view, the moral luck intuition provides prima facie evidence for compatibilism because if luck in results, circumstance, and constitution can positively affect true moral responsibility, then there is at least some reason to think that causal luck can also positively affect true moral responsibility.
3. The Too Demanding Argument
I argue now that even the ambivalence argument from moral luck is too generous, because the responsibility premise is too demanding to capture adequately even part of our commonsense ideas about true moral responsibility.
Someone who is sympathetic to the Basic Argument is likely to object that this line of reasoning inevitably begs the question, because it is of course true that the responsibility premise is too demanding. This is why it is impossible to be truly morally responsible for anything.
Nevertheless, my too demanding argument does not beg this question. It does not involve the following G. E. Moore shift: It is more plausible that possibly we are truly morally responsible agents than that the responsibility premise is true. My too demanding argument also does not amount to this argument: Because it is plausible that we are truly morally responsible for something and the responsibility premise implies that we cannot be truly morally responsible for anything, the responsibility premise is implausible. Rather, the idea behind the too demanding argument is that once we explore sufficiently how demanding the requirements of the responsibility premise are, we will see that our commonsense ideas about true moral responsibility do not support those requirements. Subsequently, I explain why it is an understandable mistake to believe that the responsibility premise is based in common sense, and I do so in part by scrutinizing the control principle, which was introduced in the preceding section.
Consider, first, that the responsibility premise implies that to be truly morally responsible for an action, an agent must have chosen the part of her mental constitution that explains the action. For Strawson (Reference Strawson1994: 6; Reference Strawson2010: 29–45), the salient part of an agent's mental constitution is the agent's reasons for acting, because, as the reasons premise makes clear, the focus of the Basic Argument is on intentional actions for which agents can give reasons. The kind of reason for action that Strawson (Reference Strawson2010: 24) has in mind is the internalist belief/desire pair kind. Thus, we can put the responsibility premise in this way: An agent S is truly morally responsible for her action only if she is truly morally responsible for at least part of her beliefs and desires that crucially explain it, and S is truly morally responsible for at least part of those beliefs and desires only if S is truly morally responsible for an earlier action in which S intentionally and successfully brought about her having those beliefs and desires. As Strawson (Reference Strawson2010: 24) writes,
But to be truly responsible for how one is, mentally speaking, in certain respects, one must have chosen to be the way one is, mentally speaking, in certain respects. It is not merely that one must have caused oneself to be the way one is, mentally speaking; that is not sufficient for true moral responsibility. One must have consciously and explicitly chosen to be the way one is, mentally speaking, in certain respects, at least, and one must have succeeded in bringing it about that one is that way.
True moral responsibility, then, requires that one has consciously, explicitly, and successfully chosen in the past one's present reasons for action.
There are, however, familiar and weighty reasons to think that we cannot consciously, explicitly, and successfully choose our beliefs and desires relevant to free actions (cf. Strawson Reference Strawson2010: 37; Peels Reference Peels2017: 52–88).
Suppose, however, that we could choose our beliefs and desires in the strong way that Strawson requires to ground to true moral responsibility. Could we, then, be truly morally responsible agents? The answer is clearly ‘no’ because there is no way to stop the infinite regress of such choices and human beings live only about 130 years (cf. Strawson Reference Strawson2010: 42).
Furthermore, even if we were infinite beings who have performed an infinite number of self-creating past actions, it still does not follow that we could be truly morally responsible agents, because the conditions in the responsibility and iteration premises are mere necessary conditions on true moral responsibility. As it happens, an infinite being has no advantage for satisfying Strawson's conditions of true moral responsibility. As Strawson (Reference Strawson2010: 50) writes, ‘Not even God could be truly self-determining as to his motives for action; nor therefore could he be truly responsible for what he did in any ultimate way’. Why should we think that? True moral responsibility requires a buck-stopping action that introduces it in the first place. Strawson (Reference Strawson and Craig2011) puts it this way: ‘To be truly or ultimately responsible … there has to be, and cannot be, a starting point in the series of acts of bringing it about that one has a certain nature—a starting point that constitutes an act of ultimate self-origination’. Thus, having performed an infinite number of actions is no help for being truly morally responsible for something.
Perhaps the best description that Strawson gives of a regress-stopping action that introduces true moral responsibility to go up the causal chain is the causa sui (Strawson Reference Strawson1994: 5). A c ausa sui is a person's being the ultimate cause of who she is and what she does. Nevertheless, a causa sui is not merely metaphysically impossible but is unintelligible. As Strawson (Reference Strawson2010: 50) writes: ‘One cannot ultimately choose one's psychological condition; we cannot describe this. It's not simply that we don't know how to, but that we know we cannot’. The causa sui must both (i) bring about the foundational action for a person to become truly morally responsible for who she is and (ii) be truly morally responsible for who she is prior to the foundational action. In other words, the causa sui must act to become truly morally responsible for who she is at a time when she is not already truly morally responsible for who she is, and, for her to perform that buck-stopping action, she must already have been truly morally responsible for who she is. Thus, it is logically impossible and not merely metaphysically impossible to be truly morally responsible for something.
Strawson's requirements on true moral responsibility are, thus, extremely demanding. These descriptions should help us to see that the conditions of the responsibility premise are not recognizable in common sense. An open-minded agnostic about the soundness of the Basic Argument would, I think, agree. She would inchoately think that we need to be able to create ourselves in some sense to be truly morally responsible for our actions. But she would reasonably think that some weaker sense of self-creation is all that is required for true morally responsible agency, because the weaker sense is the only kind of self-creation to be found in common sense.Footnote 7
Why, then, do so philosophers believe that the responsibility premise or something relevantly like it is part of our commonsense conception of morality? One explanation is that they are erroneously extrapolating from ideas that are found in common sense.
We find in common sense the following:
Weak Control Principle: An agent is truly morally responsible only for what is within her control.
And one might think that the weak control principle implies the
Strong Control Principle: What an agent is truly morally responsible for cannot be affected by that which is outside of her control.
In other words, the strong control principle is the idea that true moral responsibility requires total control: ‘An agent has total control over X only if for any factor f which is a causal contributor to X and which is such that if f were not to occur, then X would not occur, the agent has control over f’ (Fischer Reference Fischer2006: 116). As I argued in the previous section, the strong control principle implies the responsibility premise. Thus, if the weak control principle implies the strong control principle and if the strong control principle implies the responsibility premise, it would follow that the weak control principle implies the responsibility premise. The responsibility premise would, then, have a foundation in common sense, because the weak control principle is a commonsense principle.
But the weak control principle does not imply the strong control principle. It is, after all, widely acknowledged that the strong control principle is too demanding to describe properly a requirement for true moral responsibility (cf. Coffman Reference Coffman2015: 103; Fischer Reference Fischer2006: 115–16; Hartman Reference Hartman2017: 49). For an agent to satisfy the strong control principle and so to have total control over an action, she would, for example, need to have had control over a meteorite's failing to demolish the earth or the sun's not flickering out, because these background conditions must have occurred for the agent to perform the action (Fischer Reference Fischer2006: 116). In fact, even philosophers who embrace the luck-free intuition agree that the strong control principle is too strong (Levy Reference Levy2011: 5; Zimmerman Reference Zimmerman1987: 377–78). One upshot is that we have a clear example of a too demanding style argument that does not beg any questions. Another upshot is that the responsibility premise cannot get its commonsense credentials from the strong control principle precisely because the strong control principle lacks such credentials.
Nevertheless, it is important to see that the responsibility premise need not depend on the strong control principle for its merits, because the responsibility premise is weaker than the strong control principle. As Michael Anthony Istvan Jr. (Reference Istvan2011: 405–406) rightly contends, the responsibility premise does not require total control over an action for an agent to be truly morally responsible for that action; rather, the responsibility premise requires only that the agent be truly morally responsible for the mere part of her mental constitution that explains her action (cf. Strawson Reference Strawson and Kane2002: 445).
One might, however, think that the responsibility premise receives its commonsense credentials directly from the weak control principle. But this does not seem to be the case either, because the weak control principle obviously implies only one conjunct of the responsibility premise. The weak control principle, the idea that an agent is truly morally responsible only for what is within her control, plausibly implies that a person is truly morally responsible for features of her mental constitution only if she had performed an action that brought about that mental constitution (or she had that constitution and chose to omit performing actions that would change it). Nevertheless, at the very least it is not clear that the weak control principle implies the contentious part of the responsibility premise—namely, that a person is truly morally responsible for an action only if she is truly morally responsible for the parts of her mental state that explain her performing that action. Whether or not the weak control principle implies that contentious part of the responsibility premise depends on whether a person can have the level of control over her action to make her truly morally responsible for it while lacking such a level of control over the features of her mental constitution that explain why she performs that action.
I claim that the weak control principle does not imply the contentious part of the responsibility premise. To defend this claim, I briefly sketch a commonsense picture of the relation between true moral responsibility for character and action. At the time when a person first has the relevant capacities for true morally responsible agency, such as moderate reasons-responsiveness, normative competence, and/or the ability to do otherwise, she performs actions for which she is truly morally responsible to some marginal degree (cf. Reference CyrCyr forthcoming; Mele Reference Mele2006: 129–33). In this way, true moral responsibility is not transferred from earlier actions as it is in Strawson's picture, but rather it emerges from the right kind of non-responsibility conditions. The more of these actions that she performs for which she is truly morally responsible to some degree, the more she becomes truly morally responsible for who she is to some degree. And as she becomes more truly morally responsible for who she is, she becomes more truly morally responsible for what she does until—again, under felicitous conditions—she crosses a threshold and becomes a fully morally responsible agent. The important point is that the commonsense idea of true moral responsibility involves a kind of moral emergence, and this point accommodates Strawson's idea that self-creation is relevant in some way to true moral responsibility for actions. If this emergence picture belongs to common sense, as I claim, it implies that Strawson's transfer approach deviates from commonsense ideas about true morally responsible agency, and consequently the responsibility premise lacks a foundation in common sense.
We should conclude, then, that the responsibility premise is so demanding that it is either false or otiose. If we construe the too demanding argument to show that the responsibility premise is false, then the responsibility premise is an implausible requirement for a species of true moral responsibility that we take ourselves to have. This horn of the dilemma amounts to a refutation of the Basic Argument. If we construe the too demanding argument to show that the responsibility premise is otiose, the responsibility premise is a plausible requirement for a species of true moral responsibility that we have never taken ourselves to have. This horn of the dilemma shows that the Basic Argument has missed its target and therefore is irrelevant. The target of the Basic Argument is a conception of true moral responsibility that is ‘widely believed in’ (Strawson Reference Strawson1994: 6) and ‘most people ordinarily and unreflectively suppose themselves to possess’ (Strawson Reference Strawson2010: 26). Either way, the Basic Argument does not yield a pessimistic conclusion. In the next section, I offer my main error theory for the responsibility premise.
4. Error Theory
If the responsibility premise lacks support from common sense, how else might we explain its appeal? I think that its intuitive attractiveness comes from an overgeneralization. I first explain the generalization, and then explain why it is a bad one.
When philosophers think about luck in constitutive properties, they immediately reflect on bad cases in which an agent's constitutive luck obviously mitigates her true moral responsibility. Some standard examples include severe emotional trauma, bad formative circumstances, systematic conditioning, and mental illness. One uniting feature of these examples is that the agent does not voluntarily have that kind of history or condition, and these histories and conditions at least partially undermine the kind of control required to be truly morally responsible for an action that issues from them. Consider a famous example from Susan Wolf (Reference Wolf and Schoeman1987: 63–64):
JoJo is the favorite son of Jo the First, an evil and sadistic dictator of a small, undeveloped country. Because of his father's special feelings for the boy, JoJo is given a special education and is allowed to accompany his father and observe his daily routine. In light of this treatment, it is not surprising that little JoJo takes his father as a role model and develops values very much like Dad's. As an adult, he does many of the same sorts of things his father did, including sending people to prison or to death or to torture chambers on the basis of a whim. … In light of JoJo's heritage and upbringing—both of which he was powerless to control—it is dubious at best that he should be regarded as responsible for what he does.
Wolf is correct that we tend to judge that JoJo is not truly morally responsible for his actions (or at least that JoJo is not fully truly morally responsible), because JoJo does not have the right kind of control over himself to be truly morally responsible for his actions due to systematic conditioning.Footnote 8
The generalization, then, proceeds from the claim that certain kinds of constitutive mental properties outside of an agent's control (that result from severe emotional trauma, bad formative circumstances, systematic conditioning, and mental illness) undermine true moral responsibility to the claim that all kinds of constitutive mental properties outside of an agent's control undermine true moral responsibility. That is, because standard excusing and exempting cases of constitutive luck undermine true moral responsibility, so do other cases of constitutive luck.
Strawson, of course, does not explicitly generalize in this way. In fact, Strawson offers no real argument for the responsibility premise but rather asserts that it is to be found in our commonsense ideas about true moral responsibility (Clarke Reference Clarke2005: 19–20; Fischer Reference Fischer2006: 118). Nevertheless, there is reason to think that this generalization may be in the background of his thinking—or at least in the thinking of others who intuit that the responsibility premise is plausible—because Nagel's (Reference Nagel1979) skeptical argument from luck is a more general version of the Basic Argument and because Nagel makes a generalization of this kind.
Nagel's skeptical argument aims to eliminate from true moral responsibility not only the luck that frames a person's mental constitution but also the luck that comes before it and after it. A rough summary of Nagel's (Reference Nagel1979) skeptical argument is as follows:
Prior to reflection it is intuitively plausible that people cannot be morally assessed for what is not their fault or for what is due to factors beyond their control (25).
Everything seems to result from the combined influence of factors, antecedent and posterior to action, that are not within the agent's control (35).
[Thus,] the area of genuine agency, and therefore of legitimate moral judgment, seems to shrink under this scrutiny to an extensionless point (35; recall that Nagel does not endorse this argument but leaves it as a paradox).
What reason does Nagel give for thinking that factors outside of an agent's control cannot affect that for which she is truly morally responsible? Nagel (Reference Nagel1979) offers this generalization: ‘the broad range of external influences [luck in results, circumstance, constitution, and causal determinism] seems on close examination to undermine moral assessment as surely as does the narrower range of familiar excusing conditions’ (26) including ‘involuntary movement, physical force, and ignorance of circumstances’ (25). Thus, Nagel begins with the explanation that various excuses outside of the agent's control undermine true moral responsibility, and then generalizes that all instances of luck—that is, factors beyond the agent's control—undermine true moral responsibility.
Let us return to the Strawsonian generalization, which is merely a convenient label for the generalization that I have outlined. I contend that this generalization is a bad one, because there are two broad kinds of constitutive luck that are relevantly different from the previously described excuses and exemptions.
We may call the first kind responsibility-enabling constitutive luck; this kind of constitutive luck at least partially enables true morally responsible agency (cf. Hartman Reference Hartman2017: 94–95). It furnishes its agent with the broad range of reason-giving cognitive abilities and reason-responsive volitional abilities required to have what R. Jay Wallace (Reference Wallace1994: 157–59) calls ‘reflective self-control’. (Wolf calls it ‘sanity’ [Reference Wolf and Schoeman1987: 381] and ‘normative competence’ [Reference Wolf1990: 124]). The cognitive part includes at least the ability to ‘grasp and apply moral reasons’, and the volitional part includes at least the ‘power to control or regulate his behavior in light of such reasons’ (Wallace Reference Wallace1994: 157). All agents who enjoy reflective self-control are necessarily recipients of this kind of constitutive luck, because they cannot make free choices to acquire this kind of constitutive makeup precisely because such powers are required to act freely. Many compatibilists and libertarians recognize that reflective self-control is a necessary condition for morally responsible agency (cf. Fischer and Ravizza Reference Fischer and Ravizza1998; Franklin Reference Franklin2018; Hartman Reference Hartman2017; Wallace Reference Wallace1994; Wolf Reference Wolf and Schoeman1987, Reference Wolf1990).
Responsibility-enabling constitutive luck differs from the kind of constitutive luck that excuses and exempts, because those excuses and exemptions diminish or damage at least some of the capacities required for reflective self-control. JoJo, for example, comes to acquire and endorse his father's values through systematic conditioning of a kind that cuts him off from adequately understanding the weight that moral reasons should have in governing his behavior. As a result, he is not morally responsible (or not fully so) for sending people off to the torture chambers on a whim (cf. Mason Reference Mason2015: 3055). Another familiar example is Robert Harris who was abused badly by both of his parents all throughout childhood (Watson Reference Watson, Fischer and Ravizza1993: 130–39), and this kind of severe emotional trauma (and perhaps psychopathy) almost entirely smothered his ability to see the weight that the moral reasons should have in his deliberations, which is why he is plausibly not morally accountable (or not fully so) for the horrendous double murder he perpetrated (cf. Mason Reference Mason2015: 3053; Wolf Reference Wolf, Wallace, Kumar and Freeman2011: 334). For convenience, refer to these standard excusing and exempting cases of constitutive luck as responsibility-undermining constitutive luck.Footnote 9
In addition to responsibility-enabling constitutive luck, there is a second kind of constitutive luck that differs from responsibility-undermining constitutive luck, and we may call it responsibility-banal constitutive luck. As the name suggests, responsibility-banal constitutive luck is a garden variety part of the agent's mental constitution that is not required for reflective self-control but also does not undermine it. This class of constitutive properties outside of one's control includes a broad range of mental dispositions that contribute to explaining why an agent does what she does. Such dispositions include being disposed to certain kinds of romantic partnerships, to eat food with green curry, and to daydream about getting an argument just right. In fact, most of an agent's constitutive luck falls into this banal category given the broad range of ways in which luck affects our dispositions to act. There is, however, a limited sense in which even responsibility-banal constitutive luck enables moral agency, because it provides internal boundaries that can generate reasons for action and acting for reasons is vitally important for true morally responsible action.
My explanation for why the Strawsonian generalization is a bad one is that there is an important difference between responsibility-undermining constitutive luck, on the one hand, and responsibility-enabling and responsibility-banal constitutive luck, on the other. The main difference is that the undermining kind damages or diminishes reflective self-control; the banal kind neither diminishes nor enhances reflective self-control; and the enabling kind makes possible reflective self-control. If philosophers think that the Strawsonian generalization is plausible, it is up to them to tell us why the differences that I point out between these three kinds of constitutive luck are differences that are not normatively relevant in the way that I suggest.
In the absence of that kind of explanation, I contend that this error theory dampens the intuitive attractiveness of the responsibility premise. And since the intuitive appeal of the responsibility premise is the only reason Strawson provides to embrace that premise, the error theory functions to diminish the plausibility of the Basic Argument itself.
It is worth mentioning briefly that there are other error theories in the literature too. First, Benjamin Hart (Reference Hart2012) argues that we should understand a distinction between responsibility-undermining and responsibility-enabling constitutive luck along the lines of proper function. Second, Matthew Talbert (Reference Talbert2015: 79) writes, ‘We might worry that [Galen] Strawson is running together causal and moral conceptions of responsibility. Strawson identifies true moral responsibility with ultimate responsibility, but the latter seems to be primarily a causal notion: a person is ultimately responsible for some outcome only if she is the causal source of that outcome. But why should we grant that this is what true moral responsibility or true blameworthiness comes to?’ Third, Wolf (Reference Wolf2015) offers an error theory that turns on her distinction between character and mental illness. For her, character traits (including moral virtues, vices, and neither) exemplify ‘active intelligence’ by being flexible and responsive to the agent’s perception, reason, imagination, empathy, and so on, but mental illnesses are stiff and unresponsive to them (Wolf Reference Wolf2015: 361–68). As a result, when a person acts from character, the person is necessarily involved in performing the action that her character explains. But then, the following rationale that might have made the responsibility premise appear plausible is debunked: ‘Once one’s character is established the person’s behavior is out of her hands’ (Wolf Reference Wolf2015: 369).
In closing, I highlight a limitation for any error theory—and so for mine in particular. An error theory for a proposition is merely an explanation of its intuitive attractiveness in a way that is consistent with the proposition's being false. It is not an argument that the proposition is false. In my case, that there is a kind of constitutive luck that at least partially enables true morally responsible agency does not itself show that the responsibility premise is false. It could be the case that responsibility-enabling constitutive luck is necessary for true morally responsible action in one way and is sufficient to rule out true morally responsible action in another way. Alfred Mele (Reference Mele2013: 241) thinks about the luck objection to libertarianism in a structurally identical way—namely, indeterminism is necessary for true morally responsible action in one way and is sufficient to rule out true morally responsible action in another way. (Mele is officially agnostic about the success of the luck objection to libertarianism.) Thus, my error theory is merely an explanation for the intuitive attractiveness of the responsibility premise; I rely on other arguments to show that the responsibility premise is implausible.
5. Concluding Thoughts
I have argued that the Basic Argument is unpersuasive and unsound. By putting the Basic Argument into the broader context of the moral luck debate, I argued that Strawson's reason for thinking that the responsibility premise is true is unpersuasive, because our commonsense ideas about true moral responsibility contradict the responsibility premise and appear to support it as well. Subsequently, I argued that common sense does not support the responsibility premise, because the latter is too demanding. That is, the responsibility premise is so demanding that either it is an implausible requirement for the true moral responsibility that we take ourselves to have, or it is a plausible requirement for a kind of true moral responsibility that we have never taken ourselves to have. Either way, the Basic Argument does not show that we lack the kind of true moral responsibility that we take ourselves to have and so does not license a pessimistic conclusion.
Additionally, I offered an error theory for the responsibility premise. The responsibility premise gets its intuitive appeal by a generalization from standard excusing and exempting cases of constitutive luck that obviously undermine true moral responsibility to all instances of constitutive luck being such that they undermine true moral responsibility. I argued that this is an overgeneralization, because there is a relevant difference between responsibility-undermining constitutive luck, on the one hand, and responsibility-enabling and responsibility-banal kinds, on the other.
The claim that the Basic Argument is unsound connects straightforwardly to my project of defending the claim that moral luck exists and is everywhere. Moral luck occurs when certain kinds of factors beyond an agent's control positively affect her praiseworthiness or blameworthiness (Hartman Reference Hartman2017: 2). One way to deny that moral luck exists is to claim that luck universally undermines the level of control necessary for true moral responsibility, because that claim rules out luck's ability to make a positive difference regarding the praiseworthiness and blameworthiness of the individuals in each case pair (the drivers, the judges, and the citizens). But if we can show that various ways of denying extant moral luck are implausible, then we have reason to think that moral luck exists. In my In Defense of Moral Luck (Hartman Reference Hartman2017), I argue that Greco's (Reference Greco1995), Levy's (Reference Levy2011), and Nagel's (Reference Nagel1979) skeptical arguments from luck are unsound.Footnote 10 (Greco does not endorse the skeptical argument that he offers.) If my arguments here show that Strawson's skeptical argument from luck is also unsound, then it would add even more weight to the negative part of my overall argument for the existence of moral luck, because yet another skeptical way to deny the existence of moral luck would be implausible.
