1. A strawsonian thesis
Following P. F. Strawson (Reference Strawson and Watson2003), a fair number of philosophers have adopted and defended a thesis that concerns the connection between hard incompatibilism or, as it’s more commonly called, ‘free will skepticism, ’Footnote 1 and our interpersonal relationships.Footnote 2 According to the Strawsonian Thesis (as I will call it), a presumption of our status as morally responsible agents is built into key elements of human sociality. Thus, on this view, the truth of hard incompatibilism would be one that adversely affects our relationships with other persons.
Of course, just what ‘adversely affect’ comes to differs among those who accept the Strawsonian Thesis. So, for example, Susan Wolf plainly tells us that if humans lack free will, ‘it is obvious why the words “friendship” and “love” applied to relationships in which admiration, respect, and gratitude have no part, might be said to take on a hollow ring, ’ (Wolf Reference Wolf1981, 391; emphasis added). Laura Ekstrom also emphasizes the superficiality of human relationships in a world without morally responsible agency when she claims that if ‘human beings are wholly without free will (of the sort required for moral responsibility) … we [should] give up some of the satisfaction that we derive from our relationships, ’ (Ekstrom Reference Ekstrom2000, 12; emphasis added). More recently, Seth Shabo has endorsed a slightly different version of the Strawsonian Thesis, claiming that, ‘our involvement in interpersonal relationships ensures our continued susceptibility to the [responsibility-entailing] reactive attitudes, ’ (Shabo Reference Shabo2012, 97; emphasis added). In making these claims, defenders of the Strawsonian Thesis are invoking a powerful but elusive intuition about the importance of free and responsible agency. But what intuition is this? And can we make it more precise, more exact?
I think we can. And I think that a first step towards this is to understand quite clearly what the Strawsonian Thesis isn’t claiming: the Strawsonian Thesis isn’t claiming that hard incompatibilism renders it literally impossible for us to engage in personal relationships of the sort that are quite familiar – friendships, collegial relationships, romantic relationships, and so on. No one – not Strawson, nor any of the defenders of the Strawsonian Thesis – think that we would be unable to have such relationships given the truth of hard incompatibilism. Rather, they each seem to mean that something about the value or justification of those relationships would be undermined, since, for example, relating to another individual as a friend seems to require that you first relate to her as a free and responsible agent.
Yet even this weaker statement of the Strawsonian Thesis is not without its detractors. Most notably, Derk Pereboom (Reference Pereboom2001, Reference Pereboom2014) has argued against several recent attempts to defend this claim (including the ones I’ve referenced above).Footnote 3 In short, the debate between Strawsonians and those who deny the link between responsibility and interpersonal relations can be summarized as follows. Those who accept the Strawsonian Thesis think that if hard incompatibilism is true, then participating in the rich network of human relationships that give meaning to our lives is no longer warranted. Strawsonians thus claim that truth of hard incompatibilism has radically revisionary entailments concerning the legitimacy and value of our relationships with others.
By contrast, hard incompatibilists like Pereboom reject this, arguing that the justificatory grounds of human relationships do not depend on morally responsible agency. Indeed, the world that Pereboom envisions is not the cold, nearly unthinkable world of the so-called ‘objective attitude, ’ i.e., a perspective of estrangement and detachment that is putatively opposed to human feelings and relations. Nor is it a world that is devoid of kindness, empathy, and compassion. Moreover, Pereboom tells us, other people are not to be manipulated to suit our needs; their personhood is not denied. Instead, we can be emotionally engaged with them, and in some cases, our lives can be deeply intertwined.Footnote 4
The truth of hard incompatibilism does entail, of course, that we must no longer regard others as apt targets of responsibility-entailing emotions like resentment or indignation, and that we must also recognize that they are not truly meritorious for their good actions (this is just what the thesis of hard incompatibilism comes to, after all). And this, perhaps, is a loss. But it’s not obvious that this is so, especially in the case of emotions like resentment and indignation, which just as frequently poison our relationships with others as they do restore them. If the Strawsonian Thesis is true, then its defenders must have something to say about the plausible alternative offered by hard incompatibilists.
In what follows, I defend the Strawsonian Thesis. However, instead of focusing on responsibility-entailing emotions like previous defenses of the Strawsonian Thesis, Footnote 5 I will argue that if hard incompatibilism is true, we cannot make sense of: (i) the possibility of promissory obligation, (ii) the normative significance of consent, or (iii) the pro tanto wrongness of paternalistic intervention. Because these practices and normative commitments are central to the structure and maintenance of our relationships as we currently conceive of them, it follows that hard incompatibilism has radically revisionary conclusions. By itself this does not mean that no form of free will skepticism is true, but it does suggest that some standard ways of framing the view rest on a mistake.
2. Hard incompatibilism and social practice
Even if we set aside issues concerning responsibility-entailing emotions, more needs to be said in order to vindicate Pereboom’s conclusion that hard incompatibilism poses no threat to normative bases of our personal relationships. After all, there are a great many social practices and institutions that affect the meaning and value of these relationships. Indeed, we understand and value these relationships not in isolation, but as part of a larger web of human involvement. But if the normative commitments that underwrite those social practices and institutions are themselves inconsistent with hard incompatibilism, then it will follow that skepticism about moral responsibility does in fact have significant implications for the meaningfulness of our personal relationships.
With that in mind, I now want to turn to three things that apparently require morally responsible agency and that would be ruled out on the truth of hard incompatibilism: the possibility of promissory obligation, the normative significance consent, and the pro tanto wrongness of paternalistic intervention.Footnote 6
2.1 Promissory obligation
Suppose I promise you that I’ll help you move but I then fail to deliver. Now suppose that you discover that I made that promise only under extreme duress. Plausibly, you would conclude that my utterance ‘I promise that …’ was infelicitous in some crucial way and that therefore, it wasn’t binding. That is, if I promised only as a result of extreme duress, then I couldn’t come to be obligated to keep that promise even though I (apparently) performed a speech act that, in other conditions, would be sufficient to create an obligation.
Similarly, if you find out that I’ve been coerced to make the promise, or if I only made the promise because I was forced to take a drug that made me especially susceptible to agree to whatever others asked of me, or … you’d probably conclude that I was not morally obligated to keep the promise. So too, if I was a young child, who was only at the very earliest stages of understanding the social and normative significance of ‘I promise.’
In all these cases, I think there’s a simple and unified explanation for why the outward utterance, ‘I promise to help you move’ fails to actually obligate: none of the agents’ promises were made of their own free will. Unlike categorical moral obligations, which (putatively) bind us simply because we are moral agents, promissory obligations are voluntarily undertaken. This means that the promisor’s will has to be implicated in her promise in the right way. And when the promisor is under extreme duress, or coerced, or …, then her will is not free, and she does not come to be obligated to follow through on the promise that she apparently makes. But notice: the hard incompatibilist tells us that, although normal adult humans are not identical to those under duress, subject to coercion, or …, their actions are similarly unfree. After all, on hard incompatibilism, we are no more the source of our actions than any of the characters assembled above. If hard incompatibilism is true, it seems true of no one that they have made a promise ‘of their own free will.’ However, this is just to insist that no one has ever genuinely been subject to a promissory obligation. Hard incompatibilism is apparently at odds with the very idea of promissory obligation.Footnote 7
It is no coincidence that the very sorts of impairments that ordinarily serve as excusing or exempting conditions on an agent’s status as free in the sense required for morally responsible (i.e., duress, coercion, childhood, etc.) also serve to severe the link between the speech act of promising and the creation of a promissory obligation. The best explanation for this parallel, I submit, is that the possibility of promissory obligation and an agent’s being morally responsible for some action each require that the agent’s will be implicated in the promise/action in just the same way.
Here the hard incompatibilist might reply that the explanation of why, e.g., duress, coercion, and childhood undermine responsibility is different than the explanation of why the truth of naturalism or causal determinism would undermine responsibility (i.e., the explanation of why hard incompatibilism is true). First, the hard incompatibilist can argue (plausibly, I might add) that duress, coercion, and the like undermine responsibility, perhaps, because agents who are, say, under duress or who are coerced are not suitably reasons-responsive. And since moral responsibility plausibly requires reasons-responsiveness, it’s this feature of duress, et al. that explains why they are excusing conditions. But notice: it’s plausible that being reasons-responsive is also a condition on being able to enter into promissory obligations. After all, if one lacks the general capacity to recognize and appreciate reasons for action, then it’s difficult to see how one can generate decisive reasons for actions for oneself through promising.
Second, the hard incompatibilist tells us that the truth of naturalism or causal determinism would rule out moral responsibility not because they undermine our ability to be reasons-responsive, but because our status as agents would disappear altogether (in the case of naturalism)Footnote 8 or because we would not be the ultimate causal sources of our actions (in the case of causal determinism).Footnote 9 If this is right, then hard incompatibilist has a principled basis for rejecting the apparent parallels that exist between the kind of agency that’s required for promissory obligation and the kind of agency that’s required for moral responsibility. She can simply accept that neither naturalism nor determinism would entail that we lack the kind of reasons-responsiveness that’s implicated by promissory obligation while maintaining that reasons-responsiveness alone is not sufficient for moral responsibility.Footnote 10
But consider what it would mean to accept this explanation for why the apparent parallels in the excusing and exempting conditions on moral responsibility and promissory obligation don’t commit the hard incompatibilist to rejecting the possibility of promissory obligation. It would mean that even if we were, say, causally determined to act in just the ways that we do in fact act, our agency would be significant enough to (in principle) genuinely obligate us to perform incredibly difficult, demanding, and burdensome tasks, and yet it would not be significant enough to render us deserving of attitudes like resentment or indignation should we fail to meet our obligations. This is puzzling.
To see why, recall that hard incompatibilists claim that no one is ever deserving of resentment because no one is ever the ultimate causal source of her actions. In other words, if hard incompatibilism is true then I am not able to exercise my agency in a way that would ground deserved resentment (or other forms of desert-entailing blame). Of course, here we might wonder why one must be the ultimate causal source of one’s actions in order to deserve resentment. Part of the story here is that if you can in principle deserve resentment, then how you exercise your agency can open you up to harsh and unwelcome treatment that is both unpleasant and burdensome. Since such treatment is otherwise objectionable, we’d therefore expect that only a robust form of agency – perhaps only an agent who is the ultimate causal source of her actions – has the power to act in ways that would license it.Footnote 11
However, the unpleasantness and burdensomeness of being targeted with resentment pales in comparison with that which we are able to incur for ourselves through the act of promising. Through promising, I can obligate myself to do things that cost me a great deal and that really set back my own interests. As a result, promising can be a very high-stakes affair.Footnote 12 What this means is that by exercising my agency in a particular way, I ground (or rather, I have the power to ground) genuine obligations for myself that might be quite unpleasant and burdensome. But why would the truth of causal determinism obviate the power to ground deserved resentment but not the power to ground genuine obligations? If there’s any discrepancy in when agents possess these two powers, then given the fact that promissory obligation can be much more unpleasant and burdensome that being targeted with resentment, one would expect that the power to ground promissory obligations would be more fragile – more in need of a very robust form of agency to get it off the ground – than the power to ground deserved resentment. It thus seems to me that the most natural explanation for promissory obligation’s and moral responsibility’s parallel excusing and exempting conditions is that the form of agency that’s required to create obligations is sufficient to render one deserving of resentment (and other desert-entailing forms of blame) for failing to meet those obligations.
This poses a real problem for hard incompatibilists who want to maintain that their skepticism about moral responsibility is a minimally revisionist position, since promises play an important role in our lives. Informal promises of the sort that we make on the playground as children are perhaps not that important – though don’t tell this to the person who’s promised not to tell your classmates who you have a crush on. But the promises we make as adults, to our partners for example, are quite significant. They are inter alia expressions of our love. But unlike other expressions of our love (like ‘Jimmy loves Sarah’ painted on the town water tower), the promises themselves constitute a deep form of commitment, since they, unlike other expressions of love, obligate us to continue in our love.Footnote 13
And it’s not just promises that we have to worry about. Contractual obligation serves as a formal, institutional analog to promissory obligation since its moral justification lies, in part, in the idea that promissory obligation is a special and significant form of obligation. But just a contract that is coercively executed cannot be enforced legitimately, it’s generally difficult to see why any contract that is executed by an agent who is not free or responsible can be enforced. No doubt, one could try to justify contractual obligation and contract law more generally in purely consequentialist considerations – it’s good, after all, to have institutions that enable people to make and enforce agreements of a certain kind. In such a case, perhaps strict liability can serve as the standard of enforceable contracts, and hard incompatibilism is consistent with that standard.
The problem with this, however, is that although a framework for contractual law that operates with the standard of strict liability for contract enforcements might be a good institution for a state to establish, it’s not obviously a fair institution, since it’s not clear that I should be responsible for bearing the burdens of a contractual obligation if I did not freely enter into that obligation. Further, if one admits that the only rationale for enforcing contracts is consequentialist, then one has no principled reason not to intervene in contractual situations in which doing so will greatly benefit one party without significantly harming the other. But this is just to deny that there is some special moral category of contractual obligation.
The hard incompatibilist is, of course, free to deny this, but such denials surely come at a cost, since one must then be prepared to revise large portions of tort law and actual social practice. (Or else, find a different justification for those portions of tort law.) If hard incompatibilism upsets the possibility of promissory (and so, contractual) obligation, then clearly, it upsets our personal relationships in a rather significant way.
2.2 Consent
You might think that this overstates things, since perhaps we make promises only quite rarely.Footnote 14 But the problem that emerges above generalizes in worrisome ways. Consider now the role that consent and agreement play in our personal relationships. Part of what makes our relationships reciprocal is that each party’s consent and agreement is important. Though we can do stuff together, we can’t be friends if I never take your consent (or lack thereof) to be normatively significant. But hard incompatibilism is at odds with thinking that consent is significant in this way.
To see this, let’s suppose I request that you let me live with you rent-free for a month while I look for a new place to live. Ordinarily, it’s wrong to just show up at someone’s house and squat there for a month, so I can’t move in solely of my own accord. However, if I secure your consent, then I’ve done nothing wrong by living there, since you’ve temporarily granted me that right. But now suppose that my parents, who don’t want me to move back in with them, were extorting you. They tell you that if you don’t let me stay at your place for a month, they are going to break all of your fingers and toes (and if that doesn’t work, they say they’re going to move on to other joints). In that case, the mere fact that you agreed to me moving in doesn’t seem morally significant. If your consent had been freely given, then it’d be permissible for me to live in your house, but because it wasn’t freely given, it doesn’t seem permissible. (Of course, if I didn’t know about my parents’ intervention, perhaps I’d be excused for living there.)
Or instead suppose that I secured your agreement by getting you really drunk, knowing that you’re exceptionally generous while intoxicated. Again we might imagine that you say, ‘yes’ in response to my request, but again it seems that I have failed to receive a morally significant form of consent, since mere verbal agreement is insufficient for that. And though hard incompatibilism isn’t the thesis that ordinary adults are no different than intoxicated agents or the thesis that ordinary adults are no different than agents who are being extorted, it does entail that none of us have free will.
Now it’s true that Pereboom himself focuses on the connections between lacking free will and deserving praise or blame, but as these cases suggest, lacking free will has further implications for our relationships with others. That I am not free means that my consent doesn’t change the normative status of your actions in the ways we typically take it to, since what really matters is not just consent, but freely given consent. But if hard incompatibilism is true, then no consent is freely given. Consequently, if hard incompatibilism is true, in assessing whether I should move in with you, what matters is not your will, but rather the overall goodness of that state of affairs. After all, if my moving in is a Pareto optimal situation, then, if we suppose that consent lacks the normative significance that we normally attach to it, it’s hard to see what reason I have not to move in if that’s what I want to do.
However, you might worry that this rests on an equivocation on my part concerning the nature of freedom. Hard incompatibilists like Pereboom can, it seems, claim that it’s another sense of freedom – one compatible with the truth of hard incompatibilism – that’s implicated by the thought that consent must be freely given. If so, then hard incompatibilism can accept that free consent is what matters without running afoul of their insistence that agents are not free in the sense required for desert-entailing forms of moral responsibility.
Perhaps this move offers the hard incompatibilist some room to maneuver, but ultimately, I doubt it. Here’s why.Footnote 15 Suppose that the sense of ‘freedom’ at stake is a forward-looking one, such that an agent is free with respect to some action just in case she has no barriers to performing the action that would prevent her from being able to adjust her behavior in light of forward-directed praise/benefits or dispraise/harms. In other words, suppose agents enjoying this form of freedom are free in just the sense J.J.C. Smart (Reference Smart2003) thinks is important. Notice that being in dire need is compatible with just the sort of freedom. A person starving in the desert might freely (in the sense at stake) consent to years of servitude in exchange for a meal. But intuitively, that consent is not enough to make it permissible to execute the deal. So too, relatively young children are free in the forward-looking sense at stake. However, their consent seemingly lacks the normative significance of the free consent of adults. These cases suggest that being free in the forward-looking sense is insufficient for freedom of the sort that grounds the normative significance of consent. What the hard incompatibilist needs, then, is a kind of freedom that cannot serve as the basis of deserved praise or blame but that tracks the conditions under which consent has normative significance. There might be a variety of freedom that has this property, but if so, it’s a surprising coincidence, since the kind of freedom that does serve as the basis of deserved praise or blame also tracks the conditions under which consent has normative significance. Such a stroke of luck merits explanation, and in the absence of a plausible explanation, the most natural conclusion is that freely given consent is consent given by an agent who is free in just the way that would render her deserving of praise or blame, given the valence of her actions.
This picture of the normative significance of consent has its roots in Kantian (or perhaps contractualist) accounts of morality. The idea that we should never treat others as mere means, but always as ends in themselves is one that puts a lot of weight on the importance of consent. After all, what makes it okay that I tell the waiter to get me a refill – i.e., what makes it true that I’m not using him as a mere means – is that he has freely consented to the conditions of his employment.Footnote 16 This suggests an even more worrisome point, viz., that hard incompatibilism is in deep, perhaps irreconcilable tension with Kantian-inspired moral theories. If it follows from hard incompatibilism than actual consent lacks moral significance, then we have reason to reject moral theories that simply presuppose or are otherwise committed to its centrality to the moral life.Footnote 17
In response to this way of framing the worry, Pereboom argues that the kind of agency that Kant requires for moral agency is importantly distinct from the kind of agency that is required for moral responsibility.Footnote 18 The former form of agency requires that we be capable of ‘setting ends and choosing means’ and ‘formulating [practical] principles and making commitments to them, ’ (Pereboom Reference Pereboom2001, 151). The latter form of agency, the hard incompatibilist insists, requires that we be ultimate causal sources of our actions. But even if no one is the ultimate causal source of their actions, it won’t follow that no one is an ends-setter or capable of normative commitment. ‘The capacities for these activities, ’ Pereboom says, ‘can remain intact [even if hard incompatibilism is true], ’ (Pereboom Reference Pereboom2001, 151).
But this, I submit, does not vindicate the normative significance of consent. The reason for this is related to a point made earlier (in §2.1.). Because the excusing and exempting conditions on moral responsibility are apparently isomorphic to the conditions under which consent has diminished significance, it seems that the kind of agency implicated by our ordinary conception of the significance of consent is no different than the kind of agency that underwrites our status as morally responsible agents. If there is a difference here, then the hard incompatibilist owes an account of this difference, but as I’ll argue later, in §5, it’s doubtful that there is such a difference.
Furthermore, there is a second reason to think that the agential capacities that give us reason to respect others’ consent are no different than the ones that underwrite morally responsible agency. Consent is normatively important because it can affect how I have reason to treat you. If you agree to certain kinds of treatment, then it means I can engage with you in potentially burdensome ways. Your agency thus directly licenses otherwise impermissible treatment. But this is precisely what happens when we’re morally responsible for our actions. When you act in, say, a blameworthy fashion, then you give me a reason to treat you in potentially burdensome ways – ways that would otherwise be impermissible. Most notably, you give me reason to resent you. In neither of these cases, however, are the reasons grounded in consequentialist considerations. Rather, the reasons for potentially burdensome treatment/resentment seem directly responsive to facts about your agency. If the hard incompatibilist wants to claim that there are important differences between the agential capacities that underwrite the normative significance of consent and our status as morally responsible agents, then, again, they owe us an account of why reasons to treat someone in potentially harmful and burdensome ways obtain in virtue of fundamentally different features of our agency than do reasons to blame.
2.3 Paternalistic intervention
Finally, related points can be made about paternalistic interventions. Typically, we think it’s impermissible (or weaker: that there are weighty pro tanto reasons not) to intervene in someone else’s choices, even when we think those choices are suboptimal, irrational, or just plain stupid. In such cases, it’s true that we can reason with, plead with, or even nag – so long as those interventions are not bypassing the agent’s reasons-responsive capacities. But interventions of this sort, even when they take a paternalistic turn, are different from the class of paternalist interventions that are typically prohibited. This latter class includes deception, manipulation, and in some extreme cases, physical restraint, and it’s precisely this class of activities that are ordinarily proscribed when others’ choices are their own. There are, of course, exceptions to this prohibition. You can use force to stop your child from putting her hand on the hot stovetop. You can take your friend’s car keys and phone when they’re drunk – they don’t need to drive or send texts to their ex. In the case of some medical interventions, it might (let me emphasize ‘might’) be permissible to ignore the patient’s wishes in order to better secure their welfare, at least when the stakes are sufficiently high. But in the case of ordinary interactions with other adult persons, these kinds of interventions – even when they make the person in question better off – seem objectionable.
The explanation of this is simply that if someone’s stupid choice is made of their own free will, we take ourselves to have reason to refrain from paternalistic intervention even if such intervention would be good for the individual all things considered. But if hard incompatibilism is true then these powerful reasons to refrain from paternalistic intervention don’t arise. Hard incompatibilism might not entail that we must intervene paternalistically when doing so benefits the individual, but it seems to entail that we would have reason to do so – that it would be permissible at least. And even this weaker entailment would have significant implications for our ordinary relationships. If I have reason to engage with other adults in the ways that I should engage with young children or drunk friends, then although I’m not regarding them as objects, I’m certainly not regarding others as people with whom I can have meaningful and mutually nurturing reciprocal relationships. After all, if I regard myself as having reason to paternalistically intervene in your life, even if you have not asked me to do so, then I see myself as (potentially) having reason to manipulate you in order to leave you better off. But insofar as I regard myself as potentially having reason to manipulate you, I cannot really regard myself as being friends with you, since that requires me to view you as an equal participant in our relationship.
Worse, it seems that paternalism of this sort is at odds with reciprocal love. Kyla Ebels-Duggan (Reference Ebels-Duggan2008) has recently argued that the practical dimension of reciprocal love – i.e., the dimension of love that accounts for its reasons-giving force – cannot be assimilated to a concern for the beloved’s welfare. Instead, to love someone we must regard them as having ‘selection authority’ over us as well as ‘authority in judgment, ’ i.e., we must regard their choices as being a source of reasons for us and their judgment as being prima facie valid. In practice this means, e.g., if you love your partner, you must see their desire to go to the opera as a reason to go to the opera, even if it’s the last thing you’d be inclined to do on your own. Of course, you might not end up at the opera. But if you fail to regard yourself as having some reason to be there, then you seem quite unloving in this case. And if this kind of normative indifference characterizes your relationship with your partner more generally, if you never regard their preferences and evaluative judgments as reasons-giving, then it’s implausible to think that you really love them at all.
These features of love explain why it’s typically impermissible (and certainly inadvisable) to intervene paternalistically within the context of reciprocal love relationships, even when doing so would make your beloved better off. A commitment to paternalistically intervene whenever it boosts your beloved’s welfare is a commitment to not regard him or her as giving you reasons to aid them in bringing about their ends. In that case, you are prepared to simply ignore the putative authority their choices have over you. But not treating their choices in this way is, at least according to Ebels-Duggan, necessary for loving them. So a commitment to paternalistic intervention, which seems warranted on skeptical grounds, is at odds with the practical component of love. Hard incompatibilism might therefore undermine the practical dimension of our reciprocal love relationships, even if we agree with Pereboom that one can have meaningful love relationships without responsibility-entailing emotions.
Here you might (again) worry that I’m running together two ways in which agents’ choices can be made ‘of their own free will.’ The first way in which an agent’s choices can be made of her own free will is the desert-entailing kind of free will that hard incompatibilists reject. The second way is more minimal. An agent chooses of her own free will just in case she is self-governed. And although self-governance isn’t sufficient for moral responsibility, it might be sufficient for generating the reasons we ordinarily take ourselves to have to refrain from paternalistic intervention.
In response to this worry, I’d point out two important things. The first is simply that the conditions under which paternalistic intervention seems to be warranted appear in precisely the cases of action in which excusing or exempting conditions are present. And for the reasons I adduced earlier (see §3.1), I think the best explanation for this is that what grounds the ordinary prohibition on paternalistic intervention is the kind of agency that is necessary for moral responsibility. The second thing to say in response to this worry is that there are cases in which agents meet proposed conditions on self-governance and yet it seems permissible to paternalistically intervene. If self-governance requires a mesh between your values and your motives or between your higher-order volitions and the first-order desires that move you to action, Footnote 19 then you might be self-governed even if you’re under extreme duress or suffering from genuine psychological maladies of the sort that undermine their responsibility. In such cases – say, in the case of a friend who self-harms (but, because she identifies with the motive to self-harm counts as being self-governing) – I’m inclined to think that paternalistic intervention shouldn’t be regarded as being off the table. And if this is right, then it shows that self-governance is insufficient to ground the pro tanto prohibition on paternalistic intervention.
3. Variations of the participant attitude
The relationships that make sense on hard incompatibilism are apparently ones that don’t involve promises, don’t take consent to be normatively significant, and don’t preclude extremely paternalistic forms of intervention and manipulation. I’ve tried to adduce support for this indirectly, but I think it is possible to make this case in a slightly more direct way.
In ‘Freedom and Resentment, ’ P.F. Strawson introduces the idea of the ‘participant attitude, ’ which is the general attitude we must take towards others in order to participate in interpersonal relationships with them. Strawson himself describes it almost exclusively in terms of the ‘reactive attitudes, ’ which are responsibility-entailing emotions like resentment and indignation (and also: guilt, gratitude, esteem, hurt feelings, etc.). However, by focusing almost exclusively on responsibility-entailing emotions, I think that Strawson (and those following in his wake) have unduly circumscribed the participant attitude. For Strawson, the more general participant attitude seems to be comprised wholly of our emotional engagement with others via the reactive attitudes.Footnote 20 But if this is right, then it’s very plausible that hard incompatibilism is consistent with, if not the participant attitude as Strawson himself conceived of it, an attitude of interpersonal engagement that involves genuinely personal emotions (albeit not responsibility-entailing ones). Hard incompatibilists like Pereboom would thus be within their rights to conclude that our relationships are, at most, only minimally affected by the truth of hard incompatibilism.
As I see it, however, the participant attitude is a much more general practical orientation. Plausibly, this does involve the rich forms of emotional engagement of the sort that Strawson discusses. Or at least, it involves the forms of emotional engagement that hard incompatibilists describe. But even more broadly, the participant attitude is also a lattice of mental attitudes, deliberative tendencies, and behaviors that structures and organizes our reciprocal relationships with others. That is, it is a much more general framework that facilitates interpersonal engagement with others. This means that even if we grant that the emotional attitudes that figure in this complicated network need not be responsibility-entailing ones, as hard incompatibilists maintain, it won’t follow that we can thereby understand other constitutive mental attitudes, deliberative tendencies, and behaviors as not being responsibility-entailing.
For this reason, the hard incompatibilist owes us an account of promissory obligation, the normative significance of consent, and the pro tanto wrongness of paternalistic intervention. Such an account will need to explain why none of these, each of which seems to be part of the lattice that makes up the participant attitude, is responsibility-entailing. Alternatively, the hard incompatibilist owes us an account of why pruning these practices from our lives leaves us with a recognizable and meaningful form of reciprocal social interaction.
Unfortunately for the hard incompatibilist, each of these tasks is difficult. The latter is difficult because it is hard to imagine what is left of our relationships after we prune promissory obligation, deemphasize the importance of consent, and come to regard paternalistic intervention as generally permissible. Though they might be satisfying in some ways and perhaps even a source of comfort, such relationships would resemble friendship and love as we currently understand them in only the barest ways.
On the other hand, the former task will be difficult because, as we’ve seen, the same conditions that undermine an agent’s responsibility (e.g., coercion, duress, youthful naiveté, drunkenness, etc.) also seem to undermine promissory obligation, the normative significance of consent, and the pro tanto impropriety of paternalistic intervention. This means that if there is some general condition that undermines our status as responsible agents, that condition should undermine these other things as well. But insofar as the hard incompatibilist wants to insist that the truth of hard incompatibilism is minimally revisionist, it seems that this is exactly the route they must take. In the remainder of the paper, then, I’ll consider one way that the hard incompatibilist might attempt to reply and explain why I think it fails.
4. The promise of compatibilism?
In a number of places, Pereboom has employed a powerful strategy for dealing with objections of this general kind. In short, Pereboom claims that when hard incompatibilism is pinned with a seemingly implausible consequence – e.g., that it is incompatible with genuine deliberation or that it can’t make sense of holding others to norms (which putatively require a willingness to blame), for example – compatibilist accounts of the phenomena will do the trick.Footnote 21 That is, he argues that although compatibilism fails as a theory of moral responsibility, it gives us the resources for fine theories of deliberation and even some forms of blame that are not desert-entailing.Footnote 22
I think this strategy really is quite brilliant. And by my lights, it’s successful (or very nearly successful) in the cases of deliberation and of holding others to norms by blaming them. Perhaps, then, Pereboom can simply adopt this strategy here, and give us a compatibilist account of promissory obligation, the normative significance of consent, and the pro tanto wrongness of paternalistic intervention. If so, then he can plausibly maintain that the truth of hard incompatibilism doesn’t commit him to a strained and impoverished conception of human relationships. I therefore want to conclude by offering an argument against this strategy in these particular cases. For although I think the compatibilist account of deliberation, e.g., is sufficient to account for deliberation as we conceive of it, I’m skeptical that one can endorse hard incompatibilism and a compatibilist account of promissory obligation, the normative significance of consent, or the pro tanto wrongness of paternalistic intervention. Here’s why.
5. Manipulating the four-case argument
Hard incompatibilists like Pereboom are sourcehood theorists. According to this view, agents are morally responsible for their actions only if they are the source of those actions. And though there are a variety of routes to being a sourcehood theorist, by far the most common (and most compelling) is via the manipulation argument.Footnote 23 Indeed, Pereboom’s rejection of compatibilism rests on precisely this sort of argument: the Four-Case manipulation argument. If this argument (or some other variant of the manipulation argument) is sound, then it shows that the conditions that compatibilists take to be sufficient for moral responsibility are not actually sufficient.
But notice: if it’s possible to argue against compatibilism (understood as a thesis about the conditions under which agents are morally responsible), it should be possible to run an identical argument against compatibilist accounts of promissory obligation, the normative significance of consent, and the pro tanto wrongness of paternalistic intervention.Footnote 24 And if the former argument is sound, as the hard incompatibilist says, then the latter ones must be sound as well. But then, hard incompatibilism would rule out not only desert-entailing moral responsibility but also promissory obligation, et al. If, on the other hand, the latter arguments are not sound (as I think is very clearly the case), then the compatibilist about moral responsibility has good grounds for rejecting Pereboom’s original Four-Case argument. The hard incompatibilist thus faces a dilemma: it’s either the case that hard incompatibilism requires very unhuman forms of social engagement, or, in an effort to avoid this unpalatable result, it sows the seeds of its own destruction by giving the compatibilist a principled explanation of why the Four-Case argument fails to show compatibilist accounts of moral responsibility to be false.
With this in mind, let me leave you with a schematic version of a modified Four-Case manipulation argument against promissory obligation. For reference, I’ll start where Pereboom does. I’ll then appropriate the framework that Pereboom gives us for my own purposes. Consider, then, Pereboom’s Case 1.
Case 1. A team of neuroscientists has the ability to manipulate Plum’s neural states at any time by radio-like technology. In this particular case, they do so by pressing a button just before he begins to reason about his situation, which they know will produce in him a neural state that realizes a strongly egoistic reasoning process, which the neuroscientists know will deterministically result in his decision to kill White. … his process of deliberation from which the decision results is reasons-responsive; in particular this type of process would have resulted in Plum’s refraining from deciding to kill White in certain situations in which his reasons were different. His reasoning is consistent with his character because it is frequently egoistic and sometimes strongly so. Still, it is not in general exclusively egoistic, because he sometimes successfully regulates his behavior by moral reasons, especially when the egoistic reasons are relatively weak. Plum is not constrained to act as he does, for he does not act because of an irresistible desire – the neuroscientists do not induce a desire of this sort.Footnote 25
Here Plum meets all of the purported compatibilist conditions on moral responsibility, and yet, the hard incompatibilist insists that he is not morally responsible for killing White because that action ‘is causally determined by the neuroscientists’ intervention, which is beyond his control, ’ (Pereboom Reference Pereboom2014, 77). At this point, it’s not essential that you agree with the hard incompatibilist in their judgment; what’s important is that you appreciate why Pereboom and other hard incompatibilists (reasonably, but erroneously in my opinion) take Plum to be exempt from moral responsibility.
With this in mind, let’s turn to a parallel case in which an agent is manipulated in precisely the way that Plum is manipulated, and consider what the hard incompatibilist must say about that agent.
Case 1*. Mustard is an agent who is just like Plum in Case 1 of Pereboom’s four-case manipulation argument. A team of neuroscientists periodically intervenes in his reasoning processes, but they do so in a way that leaves his rational capacities in tact.Footnote 26 Through their interventions, Mustard is causally determined weigh reasons in a way that leads him to promise Scarlet that he will help her move.
The day of the move comes, and we want to know, is Mustard actually obligated to help Scarlet – i.e., would he be doing something objectionable by failing to help? Or does he do nothing wrong if he decides at that time to refrain from helping?
For those of us who know about the neuroscientists’ work, is it plausible that through the speech act of uttering ‘I promise…’ Mustard has come to be obligated to help? I tend to think so, and yet, I can see why someone without a settled view of these matters would be agnostic. But whatever you, me, or an unsettled agnostic might be inclined to say about this case, it’s hard to see how the hard incompatibilist could accept this. For just as it’s tempting to say of Plum that he is exempt from blameworthiness because he is not the source of his action, so too, it seems tempting to say of Mustard that he is not obligated because the promise wasn’t his in the relevant sense.
Moreover, if one accepts the soundness of Pereboom’s Four-Case argument, one can also say of Mustard that in this case, he isn’t morally responsible for promising. But can one be obligated to keep a promise that one is not morally responsible for making? I don’t see how. As I’ve already noted, promises are optional but often very burdensome obligations. It doesn’t seem fair that someone come to be obligated in this way when you admit that she is not morally responsible for incurring the obligation.
Consider now Case 2*, which parallels Case 2 of Pereboom’s Four-Case argument.Footnote 27
Case 2*. Mustard is an agent who is just like Plum in Case 2. A team of neuroscientists has programed him from birth to reason in a particular way. This initial programing secures the neuroscientists’ preferred form of reasoning without interfering with Mustard’s rational capacities. Given his programing, Mustard is causally determined in the circumstances to weigh reasons in a way that leads him to promise Scarlet that he will help her move.
Again we can ask ourselves, is Mustard obligated in help in Case 2*? Well, as Pereboom points out in his statement of the Four-Case argument, it’s hard to see why added temporal distance to the neuroscientists’ intervention matters. So if he wasn’t obligated in Case 1*, then plausibly, Mustard isn’t obligated in Case 2*.
Cases 3* and 4* are just like Cases 3 and 4 of Pereboom’s Four-Case argument. They are therefore different than Cases 1* and 2*, but it isn’t obvious that their differences are relevant to promissory obligation.Footnote 28 And since Case 4* is a case of ordinary causal determination by blind physical forces, it seems that a compatibilist account of promissory obligation won’t work for the hard incompatibilism. Just as Pereboom’s original Four-Case argument putatively shows compatibilist conditions on moral responsibility to be insufficient, I think this modification of the four-case argument shows that, for hard incompatibilists at least, compatibilist conditions on promissory obligation are also insufficient.
Notice, however, this argument doesn’t pose a problem for the compatibilist, since she either accepts that Mustard is morally responsible for making the promise in Case 1*, and so is not committed to him being obligated even though he is not morally responsible, or she takes there to be an important difference between Case 1 and Case 4 of Pereboom’s Four-Case argument that allows that Mustard would be morally responsible in Case 4*, which would simply be a case in which Mustard is causally determined to make the promise by blind physical forces. The compatibilist, then, is not committed to the idea that one can be bound to keep a promise that one is not morally responsible for making. The hard incompatibilist, however, is committed to this – at least insofar as he or she wants to maintain that we are sometimes subject to promissory obligation. I find this a tough pill to swallow, and so it seems to me that the hard incompatibilist should instead just accept that we are never subject to promissory obligation.Footnote 29 By itself this doesn’t mean that hard incompatibilism is false – just that it requires a more revisionary conception of the participant attitude than previously advertised.
At this point, however, you might worry that the parody argument I’ve offered cannot succeed. For although it shares a similar structure to the argument Pereboom develops as a piece of overall argument for hard incompatibilism, there’s nothing that rationally compels one to have the same intuitions in each case, since they arguments concern different things – desert-entailing moral responsibility and promissory obligation, respectively.Footnote 30 In response to this objection, I’d first note that I’m not trying to move from the parallel structure of the two arguments to what I take to be the correct judgments in Cases 1* and 2*. Instead, I take myself to be proceeding in just the way that Pereboom (Reference Pereboom1995, Reference Pereboom2001) does: first from a considered judgment in a specific case to the best explanation for that judgment, and then to a more general claim that one arrives at after considering a range of cases that apparently similar. Pereboom claims that the lesson we’re supposed to learn from the original Four-Case Argument is that if your action traces back to factors wholly outside of your control (i.e., if you’re not its source), then you are not morally responsible for that action. This mirrors the lesson I think we should learn from these cases, which is simply that if you are not morally responsible for promising to x, then you are not morally obligated to x. Of course, just as compatibilists want to resist Pereboom’s Four-Case argument, hard incompatibilists can resist this parody of it. But in so doing they must either accept that you can be morally obligated to keep a promise that you weren’t morally responsible for making, Footnote 31 or they must identify a relevant difference between that and causal determinism. Second, I think it’s important to say that I’m actually happy to concede to this objection that although no one is rationally compelled to have the responses that I suggest are correct in response to Cases 1* and 2* (and others), I do think those responses (and the lesson a take from them) are reasonable ones to have. And if so, then it’s at least reasonable to think that these judgments can serve as planks in an overall argument that hard incompatibilism has more revisionary implications with respect to promissory obligation, the normative significance consent, and the norms of paternalistic intervention than has previously been recognized.
6. Conclusion
Although I’ve focused exclusively on promissory obligation here, similar arguments can be developed to show that it’s not clear that hard incompatibilists can give compatibilist accounts of the significance of consent or the norms of paternalistic intervention, but I will not pursue that here, since they are perfectly parallel to the one sketched above. Instead, I’ll conclude with the following. In response to other defenses of the Strawsonian Thesis, hard incompatibilists like Derk Pereboom have offered powerful arguments against the idea that responsibility-entailing emotions are necessary for meaningful personal relationships. However, even if we accept that those hard incompatibilist arguments are sound, there is still reason to think that hard incompatibilism threatens the legitimacy of some core elements of our shared way of life, since its truth threatens promissory obligation, the significance of consent, and ordinary prima facie prohibitions against paternalistic intervention. Hard incompatibilism may yet be true, but it is perhaps a great deal harder than its proponents are willing to admit.
Acknowledgement
For helpful feedback on these ideas, I’m especially grateful to Derk Pereboom, Neal Tognazzini, and two anonymous referees from this journal.
