The topic of free will continues to be a hotbed for discussion and disagreement. The most pressing questions remain: What is free will? And how does it affect the notion of moral responsibility? Here, Derk Pereboom and Dana Nelkin tussle over this issue, and disagree on whether we possess sufficient free will to hold us morally responsible. In spite of this, however, they both assert that we can maintain a stable criminal justice system in which punishment plays a prominent role.
In Defence of Free Will: Dana Nelkin
I believe that at least much of the time we act freely and in such a way that we are morally responsible for our actions and their consequences. This means that we are eligible to be blameworthy or praiseworthy. When you decided to read this article, you likely did so freely. When you expended some effort to help someone else just for their sake, you were likely morally responsible for it, and in a way that makes you praiseworthy; when you didn’t take as much care as you should have and thereby wronged someone else, you might also have been morally responsible, and blameworthy, for it. While we tend to simply take for granted that we are free and responsible in these ways, questions naturally arise, including whether we are simply under an illusion in assuming we are generally free and responsible for what we do. In answering these questions, I will need to say more about what I take moral responsibility to be, how skeptical arguments can be answered, despite their initially surprising strength, and what is at stake.
The notion of moral responsibility we are interested in has a central place in a whole set of interpersonal practices and ways we treat each other.
Let’s start at the beginning. What do we mean by free action and moral responsibility? For present purposes, I’m going to focus mostly on moral responsibility, and will take it that the kind of freedom we are interested in is whatever is required for us to be responsible agents. Though “responsibility” can mean different things, what I’ll focus on here is a notion of accountability, the idea that people can be accountable for their actions. In turn, this means that it is legitimate to make demands of others and of ourselves and to hold them and us to moral obligations. Being accountable is also generally agreed to entail that when one acts wrongly, it can be appropriate to respond with certain moral attitudes, such as resentment, indignation, or guilt. Though it is a source of more debate, I believe that being accountable also entails that one is a candidate for desert; if one acts wrongly while being accountable, and so is blameworthy, one deserves something of a negative valence, to have one’s interests set back in some way (which needn’t come to anything like the kind of treatment prevalent in the criminal justice system in the United States); if one is praiseworthy, on the other hand, one deserves something of a positive valence. In sum, then, the notion of moral responsibility we are interested in has a central place in a whole set of interpersonal practices and ways we treat each other. We hold each other to account, we have reasonable expectations of each other, and we respond to each other with both positive and negative moral attitudes when those are unmet, met, or exceeded.
If it is really difficult to do the right thing in the circumstances, one might still be blameworthy, but less so than if it had been easier.
What does it take to be responsible in this sense? We don’t think it’s appropriate to hold everyone responsible all the time. Young children do not seem appropriate targets of resentment or of demands, for example. And even normally responsible adults can sometimes have an excuse in particular situations. What conditions, then, make us responsible when we are, such that failing to meet them lets us off the hook?
I endorse the following Quality of Opportunity account. It is centered on the idea that whether one is responsible at all depends on one’s having a good enough opportunity to act for good reasons, and when one is morally blameworthy, a good enough opportunity to act well morally for good reasons. The account of blameworthiness is scalar, in that one’s blameworthiness can be mitigated to the extent that one’s opportunity is compromised or of a lower quality. For example, if it is really difficult to do the right thing in the circumstances, one might still be blameworthy, but less so than if it had been easier. Two kinds of factors can affect the quality of an opportunity. “Internal” factor comprises our skills and capacities for recognizing and weighing reasons and for translating that recognition into the adoption of reasons and action. (This package of capacities is often known as “normative competence”.) If one is skillful at identifying reasons, it should be easier in at least one sense to do the right thing, and that can make one’s opportunities better. On the other hand, even if one has incredible capacities, the world is not always so cooperative, and situational factors might still cause one to face a low quality opportunity for doing the right thing. For example, the existence of duress or challenging social factors can lower the quality of an opportunity from what it would otherwise be. Thus, there is also an “external” factor, situational congeniality, that is essential for determining the quality of an agent’s opportunity. This two-factor model has the resources to account for an otherwise heterogeneous set of possible mitigating factors, or excuses.
So far, so good. We’ve fixed the concept of morally responsible agency we are interested in, we’ve got an account on the table of the conditions under which people are responsible, and one that allows us to explain why some people might be more responsible and blameworthy than others. Now for the challenges!
For a long time, theorists were focused on the challenge of determinism. Everything that happens seems to be entailed by the state of the world in the past, together with the natural laws. If our own actions are determined along with all else that occurs in the world and we could not do otherwise, then how can we act freely or be responsible? And to deepen the puzzle, sprinkling in some indeterministic moments just seems to add randomness to the picture, and, as Derk Pereboom has vividly brought out in his work, that does not seem to add any desired control to the picture.
As important a challenge as determinism poses, the long and intense focus on it tended to obscure still other challenges to free and responsible agency. In recent decades, these have received more attention and that is a great thing not only because they are important in their own right but also because thinking about them allows for new insights about the nature of free and responsible agency, and, perhaps surprisingly, provides new tools for thinking about the challenge of determinism, as well. I’ll mention just a few here. Some challenges come from advances in social and cognitive psychology. For example, a large set of experiments, including the famous Milgram experiments on obedience, purported to show that the situations we find ourselves in, much more than anything special about us as individuals, explain our behavior. So, before we blame others for going along with (illegitimate and immoral) orders, we should take a moment to reflect that we, too, under similar circumstances, are likely to have done the same. Or consider the fact that situations are very often not at all congenial when they factor into the quality of people’s opportunities, poverty, lack of education, and oppression. Cognitive psychology has also revealed deep and systematic biases, some socially inculcated and some representing what appear to be built-in tendencies to irrationality and self-deception. All of these aspects of ourselves and the situations in which we find ourselves seem to undermine the idea that we have quality opportunities.
How to answer all of these challenges? A complete answer would take each one in turn. Instead, here, I will highlight just some of the resources of the account of responsible agency sketched earlier, and gesture at how they might each be deployed in providing at least partial responses to multiple challenges.
First, it is crucial to acknowledge that any account of responsible agency is just that. The account only gives verdicts when combined with the facts on the ground. This means that we should be open to a new and better understanding of the world, and to the possibility that not as many people are responsible or blameworthy as we thought, or that people are not blameworthy to the degree we previously assumed. And indeed, we already are. Children and even teens are a case in point; at some times, children were treated essentially like adults when it came to holding them to obligations and treating them as fully responsible agents. But that has obviously—and rightly—changed—in light of improved understanding of developing capacities, susceptibility to peer pressure, and more. And in fact, I believe that a number of the phenomena just mentioned—cognitive biases, difficult situations, and more--provide reason to shrink the scope of our blaming attributions. This is not a reason to reject either the idea that we are responsible agents in general or the Quality of Opportunity Account. In fact, it shows that the account is well supported for being able to explain why new understanding about incapacity and situational challenge can undermine attributions of blameworthiness. Importantly, too, we can use this knowledge of our situational obstacles and capacity impairments to work around them—on our own behalf and in helping others. This means that even as we acknowledge a restricted scope, we can at the same time use that knowledge to expand it by creating higher quality opportunities using new tools.
Second, the account offered above naturally explains how responsibility and blameworthiness are not simply “on” or “off”. They are a matter of degree. This allows for nuance; perhaps people are less responsible than we thought sometimes, but not fully off the hook. Note, too, that this is one way in which the previous focus on determinism—which itself is an “on/off” phenomenon—tended to postpone thinking through in more detail what a true-to-life account of responsible agency ought to include. In turn, as I’ve argued in more detail elsewhere, thinking about the scalar aspects of responsible agency can help motivate the idea that determinism is not in fact the key to the issue that once seemed so clear.
Third, it is important to be clear about what responsibility entails and what it doesn’t. For example, even if, as I suggested above, one’s being responsible and blameworthy entails that one deserves something, it does not follow that a person’s being deserving makes it good in and of itself to give them what they deserve. Rather, being deserving entails a kind of liability; if there are very strong reasons for someone’s interests to be set back, it is the deserving who come first. (This is a cousin of Joel Feinberg’s famous phrase, “fault forfeits first”.) But we leave no reasons on the table when we decline to give the blameworthy what they deserve absent other pressing goods. Thus, the account implies nothing directly about state punishment in general, and in fact, I believe that desert should play no role at all in the rationale of our criminal justice system; what should take center stage in the justification of such a system is the protection of rights. To my mind, this is an advantage for the account and forestalls at least some salient reasons offered on behalf of skepticism.
Finally, we come to the last question, concerning what is at stake. A number of theorists would point to the fact that it remains of central importance whether or not we are creatures who are apt targets of the particular moral attitudes such as resentment and guilt. But I am not convinced that this is what most matters. (See Derk Pereboom’s work for powerful arguments to this effect.) Still, on most days , I continue to believe that there is something very special about being the kind of creature who has qualities and capacities that might be presupposed by such attitudes. In particular, there is something especially valuable about being creatures who are subjects of obligations to which they can aptly be held. To be such a creature, one must possess a kind of autonomy and a kind of rational agency that deserves a special kind of respect. Further, special relationships like friendship, familial relationships, and relations between colleagues, are also partially constituted in part by special obligations that we have to one another and ones to which we can aptly hold each other. There is, thus, a constellation of interconnected reasons why there remains much at stake in the debate about whether we are free and responsible agents.
The Response: Derk Pereboom
Dana Nelkin argues that general conclusions about the nature of the universe, such as causal determinism, that everything that happens is causally determined by preceding events in accord with laws of nature, should not affect how we think about our agency and moral responsibility. By contrast, I contend that due to such general facts about the universe, we lack the type of free will required for the basic desert sort of moral responsibility, which is at issue in the traditional free will debate. That is, we lack the control in action required for our deserving, in a fundamental sense, to suffer the harm and pain of blame and punishment for immoral decisions, and the pleasure and benefit of praise and reward for those that are morally exemplary. We would not be morally responsible in this sense if our decisions were causally determined by factors beyond our control. One might think that features as large-scale as the causal determinism of the universe shouldn’t affect how we think about our agency and responsibility, but I think that we should take the possible implications seriously in everyday interactions, in particular if our lives would improve as a result.
We lack the control in action required for our deserving, in a fundamental sense, to suffer the harm and pain of blame and punishment for immoral decisions.
One way to see that that determinism has such implications is by drawing an analogy to universal causal determination by a situation in which we’re causally determined to act by other agents. Imagine that we’re agents in a massive computer simulation, controlled by intelligent programmers, in which our sense of free choice is an illusion. Would we then say that we basically deserve the pain and harm of blame and punishment for our immoral decisions? I say not. But universal causal determinism isn’t relevantly different, and so we should reject basic desert moral responsibility if determinism turns out to be true.
Given the truth of quantum mechanics as our best science of the physical basis of reality, it’s open that the universe is not deterministic after all. While some interpretations of quantum mechanics are deterministic, there are indeterministic interpretations as well. Whatever indeterminism there might be would need to result in indeterministic options for decisions on a regular basis. Whether this could be is controversial. Moreover, this sort of indeterminacy in our decisions may be as threatening to free will as determinism is, since the control we might exercise over quantum indeterminacy would seem to be lacking. I don’t claim that our having the sort of free will at issue is impossible. Nevertheless, since the only account on which we are likely to have this kind of free will is not credible given our best physical theories, we must take seriously the prospect that we are in fact not free in the sense required for basic desert moral responsibility.
At the same time, I maintain that we can live with a conception that rejects this type of free will and moral responsibility. Basic desert moral responsibility is backward-looking, since the aim of holding people morally responsible in this sense is to pay them back with harm or pain for their wrongdoing, and to reward them with pleasure or benefit for moral excellence in action. But our practice of holding morally responsible also has a substantial forward-looking component. We blame not only to pay back, but to protest wrongdoing in a way that aims at goals such as the moral formation of the wrongdoer and reconciliation in relationships. On both a personal and societal level we have an interest in the moral formation of the wrongdoer, and moral protest of wrongdoing can function as a step in that process. We often have a stake in a relationship impaired by wrongdoing, and protesting the actions of wrongdoers can function as a step in reconciliation with them. In addition to rewarding, praise has the function of celebrating and encouraging moral excellence.
The right to self-defense can justify the state’s practice of criminal punishment.
In dealing with more serious wrongdoing, such as violent crime, backward-looking moral responsibility features the notion of retribution, imposing on criminals the punishment they deserve in proportion to the seriousness of the wrong they’ve done. However, if we reject the free will required for basic desert, retribution against criminals would also need to be relinquished. Is there an alternative proposal for treatment of criminals that doesn’t appeal to free will?
Daniel Farrell sets out such an alternative based on the right to self-defense. It's widely accepted that we have the right to harm and to threaten harm to defend ourselves and others against wrongful aggressors, whether or not they are morally responsible and deserve such harm. Even if the school shooter is mentally ill, and for this reason not morally responsible, we believe that we have the right to harm him in defense of those who are threatened by him. Given this right to harm in self-defense, we also have the right to threaten to inflict this harm on a potential wrongful aggressor on the condition that he attacks. Furthermore, because each of us has these rights, the state, acting on behalf of us, may legitimately issue corresponding general threats to harm potential wrongful aggressors, and may also carry out these threats once their conditions have been violated. In this way, the right to self-defense can justify the state’s practice of criminal punishment.
Note, however, that the right to harm in self-defense is limited, since it is the right to inflict the minimum harm on a wrongful aggressor required to prevent him from harming you or someone else. This results in a limitation on the severity of punishment grounded in this way, and in many cases incapacitation, such as preventative detention and monitoring, is the measure that can be justified. But to this we can add that the state may not conceal the fact that it employs incapacitation as a preventative measure, and must publicize this information instead. Accordingly, incapacitation, coupled with a publicity provision, would to strengthen the general deterrence against crime that we expect from the state. To this we can add measures for rehabilitation and reintegration that would affirm our regard for the humanity of the criminal.
Still, given causal determinism, there are factors beyond your control – the distant past and the laws of nature – that made it inevitable how you deliberated, the choice to keep the promise that resulted, and the good effects it had.
Turning to the level of the personal, how would we conceive of our own agency if we came to accept that we lack free will in the sense at issue? If all deliberation and action are causally determined by factors beyond our control, then which option agents select upon deliberating is made inevitable by causal factors beyond their control. The effects of that selection are also inevitable by virtue of such factors. Thus, it would seem that how we deliberate and choose, and the effects of our choices, are beyond our control. Can we accept these consequences?
Dana Nelkin and I have each argued that given causal determinism there remains a sense in which how we deliberate makes a difference to how we decide, which in turn makes a difference to our futures. Suppose one spring you promise to help a friend clean her attic, but you consider breaking your promise and doing something more enjoyable instead. But upon deliberation – thinking of and weighing reasons for each option – you decide to keep the promise and help the friend. Imagine also that it was causally determined that you deliberated as you did and decided to keep the promise, and that as a result your friendship wasn’t impaired and instead benefitted from it. It yet may be that had you deliberated differently, you would have decided to break the promise, and your friendship would have been somewhat impaired. In this sense, your deliberation made a difference to your decision, which in turn made a difference to your friendship. This sort of difference-making control can be contrasted with what Eddy Nahmias calls bypassing – that our conscious mental activity is bypassed in making decisions and in acting. Bypassing would occur, for example, if a neuroscientist caused you to make a decision by direct neural stimulation without any deliberation on your part. But causal determination does not entail bypassing, and causally determined deliberation can make a difference to how you decide and act.
Still, given causal determinism, there are factors beyond your control – the distant past and the laws of nature – that made it inevitable how you deliberated, the choice to keep the promise that resulted, and the good effects it had. Thus the difference deliberation can make is limited. While causally determined deliberation can make a difference relative to other options being considered, those other options not being selected is made inevitable by factors beyond your control.
Thus, even if accept that we lack the free will required for basic desert moral responsibility due to universal causal determinism, or to the sort of indeterminism we see in certain interpretations of quantum mechanics, we can live with the practical implications. We can affirm the forward-looking component of our practice of holding morally responsible, focusing on moral formation and reconciliation rather than the deserved harms and benefits. We can accept a criminology focused on deterrence grounded in the right of self-defense and defense of others. And while, on the supposition of determinism, we would accept that our decisions are made inevitable by factors beyond our control, there remains a robust sense in which we can make a difference.