An Interview with T. M. Scanlon (Part I)
By Yascha Mounk
T. M. Scanlon is one of the most highly regarded moral philosophers working today. Besides his masterwork, What We Owe to Each Other, he has written transformative treatises on topics ranging from the significance of choice to the nature of permissibility, meaning and blame.
It is natural to expect a certain amount of pomposity from philosophers, perhaps especially from moral philosophers. Scanlon’s stature – and those old-fashioned, impersonal initials he uses in publications – makes him a particularly promising object for such suspicions. But readers of his works are quickly disabused of any such expectations. His prose is elegant, straightforward and accessible. At times, it is easy to miss the true import of what he’s saying, so calmly does he state even his most far-reaching conclusions.
To his colleagues and students at Harvard, T. M. Scanlon is, in any case, simply Tim: always friendly, always jovial, he is a tall, lanky man in his early seventies, with a long face and huge hands. There is something distinctly folksy about him. In seminars and lectures, his winning ordinariness makes his account of the spring for our moral concern – not an austere categorical imperative, he argues, but a deeply human desire to stand in relations of mutual recognition with each other – appear all the more reasonable. The vision of a world in which, like Tim, every moral agent is unpretentiously concerned to do right by his fellow man really does sound eminently appealing.
We meet for this interview on the morning of the 19th of February 2012, a Sunday, on the steps to the philosophy department where Tim was a graduate student, and is now a faculty member. Walking into the building, we pass scores of students in ill-fitting suits who are nervously pacing around Emerson Hall. He scans the corridor with irritation before realizing that the philosophy department has been repurposed as the venue of a debate tournament for the weekend. Tim, his loud, mischievous laugh reverberating up and down the staircase, says that he disapproves of these would-be Sophists: instead of scoring points against each other, they should, he tells me, use their smarts to interrogate their own beliefs.
Finally, we arrive at Tim’s spacious third-floor office and settle down at a long table in the back of the room. We talk well into the afternoon, overlooking Harvard Yard as a dreary day quickly fades. All the while, we are watched over by the happy citizens of Siena featured in a reproduction of Ambrogio Lorenzetti’s Allegoria del Buon Governo.
I: Free Will, Punishment and The Significance of Choice
I: Free Will, Punishment and The Significance of Choice
One of philosophy’s oldest worries is causal determinism: the fear that, if what we do and think is determined by physical processes beyond our control, then we should abandon moral categories like praise and blame and choice. But I take it that you’re less worried about that than many of your colleagues?
I think there are three ways in which this problem arises – the problem being the possibility that a causal explanation of a reaction we give would undermine its significance in one way or another.
The most commonly discussed is the problem of so-called moral responsibility: the fact that there’s a physical or causal explanation for what we’re like would undermine the appropriateness of moral evaluations – of blame, in particular.
The second, the more political version, is the fact that if our choices were caused by factors outside of us, over which we have no control, then that would undermine them as factors that can legitimate outcomes. The fact that I consented to something could not make it the case that I can’t object to it. Or the fact that I entered into a contract wouldn’t mean that I have to fulfill it.
And the third, what I call the personal problem of free will, is that the causal explanations would undermine the significance for me. My sense of ownership, or whatever – that I’m the person who’s represented, as opposed to these just being things that happened for reasons that don’t have anything to do with me.
And I think that these three problems, what I call the moral problem, the political problem (although it’s not solely political, but I’ll put it that way), and the personal problem – those three problems have increasing degrees of difficulty as you go down that list.
I think the moral problem, which has received probably the most attention, is the least problematic, at least as far as moral blame and other reactive attitudes or reactions, whatever they are, are concerned. Because there I think what we’re doing is reacting to the way the person is. The person is defined by his or her psychology. Here David Hume was right. Factors that make an action not revealing of my actual psychology undermine responsibility because I’m not really like that. Stimulating my brain so that I say something that I never would have said means that it doesn’t reflect anything about how I feel toward you. Such factors undermine what we can call the liberty of spontaneity. But the fact that there’s some explanation for how I am what I am doesn’t in general change the fact that I am that way.
So we can judge people for being a particular way, even if it’s not under their control that they in fact are like that…
Yes, they’re like that.
Now, the degree to which that’s true of course depends on what kind of response blame involves. If blame is purely evaluative, if it just involves judging me to be a good or bad person, then the conclusion I just mentioned follows immediately. That’s essentially what was going on in Hume’s treatise. He had in mind an essentially evaluative idea, a feeling of approval or disapproval toward that type of character.
I think there’s a lot – I don’t want to call it confusion – a lot of uncertainty and unclarity out there as to what blame involves if it’s not purely evaluative. As I said, I think the two poles that this idea of blame cycles between are, on the one hand, an evaluation and, on the other hand, some kind of punishment or sanction. But it’s always somewhat unclear what the sanction is.
At any rate, I think that the sanction idea isn’t a very good interpretation of blame: when we’re blaming people, we’re not acting as enforcers. But I also think that the evaluative idea of blame is too thin and weak. We must be doing more than that. Otherwise, the idea of there being a problem about what we do not being under our control wouldn’t seem like an issue, except in the Humean way.
So I’ve tried in the last chapter of my book Moral Dimensions to say: what more could there be to blame that is in this space? Now, I want to emphasize that I self-consciously chose not to address the problem exactly in that form. That is, I wasn’t doing it as a response to the problem of free will, as it were. I wanted to investigate what I actually think about blame, and what we’re doing. Once you have an idea what it is you’re doing when you’re blaming somebody, then you can address the question: okay, what has to be the case in order for that to be appropriate? If it’s more than your evaluation, then it’s not enough simply that it should be based on an accurate account of what the person’s psychology is. But depending on what the further content envisaged is, we’ll get different answers.
And I offer this relationship-based idea that to blame somebody is to decide that what he’s done impairs your relation with him in a way that justifies you changing your intentions about how you’re going to behave toward him, and how you’re going to understand your relation with that person. Now, a lot of people don’t like that view, in particular because they don’t think that it gives enough weight to moral emotions. That’s been one of the main criticisms of it. But anyway, it’s meant to occupy that space.
If you give that account of what blame involves in addition to a matter of evaluation, then there’s a first-order moral question about any particular way of revising my intentions about how I’m going to behave toward this person. It’s a first-order moral question whether that is justified, given what the person is like. And it may be that the free will problem will surface at that point, because it may be, with some particularly nasty ways of deciding how I’m going to treat this person, that they aren’t permissible unless the person had an opportunity to avoid running afoul of my deciding to treat him that way. But I think that, as I interpret blame, that first-order question doesn’t arise. It is a matter, as I say, of pure desert: that simply the fact that a person is like this can justify my withdrawing intentions to become his friend, to trust him to enter cooperative activities, to help him in certain ways, to care a lot about what happens to him, and so on.
So part of the reason why the free will problem doesn’t arise in the case of blame is that the burden imposed on other people is not that heavy. But what about cases where the burden is heavy – in punishment, for example?
Right. Punishment is much tougher. Punishment involves not only, in the characteristic case, some kind of expressive judgment – as Joel Feinberg rightly emphasized – but also, as they say, some kind of hard treatment. And then it’s a real question. I think there the permissibility of inflicting most of the kinds of hard treatment that are involved depends heavily on whether people have had an opportunity to avoid, of the relevant kind.
That brings us to the second kind of problem with free will. That has to do with the licensing effect of somebody’s having made a certain kind of choice. And there, I’m much closer – although not identical, and although I use it in a very different way – to something like J. J. C. Smart’s view. I guess this is the value of deterability for the person who is deterred: it’s a good thing for us if the prospect of having certain bad things happen is something we’re warned about under circumstances in which that’s likely to deter us. And if we haven’t been warned in situations in which we’re likely to be deterred, then we haven’t been given sufficient protection against having that bad thing happen to us.
I mentioned the comparison with Smart because Smart just looks at it as a problem with social policy. That is, it makes sense punishing people if they can be deterred because otherwise it isn’t socially useful. So that’s the social value of deterrence. But what I emphasize is its value for the individual person. Interestingly, Smart thought that what he thought of as the common sense understanding of blame had to be rejected because it presupposed free will. Whereas he thinks that the permissibility of sanctioning people as a deterrent strategy doesn’t presuppose free will. He didn’t ask whether the permissibility of threatening people and imposing these sanctions on them in order to deter others, depended on their having free will to choose to avoid this punishment. He didn’t address that question simply because he’s a utilitarian, so all he cared about was whether it was socially useful to do it. But I do care about it. I’m not a utilitarian.
So I have this basic theory: the so-called Value of Choice view. If certain bad things are going to happen to us, or if there’s a possibility of bad things happening to us, we have less of a complaint against that possibility if it’s set up so that we could avoid them if we so choose. Now, the value of that protection is very sensitive to the conditions under which that happens, and if a lot of people are raised under conditions in which it’s predictable and understandable that the threat of the law is not going to have any effect on them, that makes it much more difficult – or even impossible – to justify punishing them. So I think that opening for an objection to a system of punishment is clearly recognized on my view.
So if causal determinism were true, and if a person’s genes predetermined that they were going to commit a certain crime, then you’d be very worried about whether or not it would be legitimate to punish them? It would, perhaps, be legitimate to blame them, but not to punish them?
Well, you know, that’s what I was about to say, because the free will problem does come back. So I want to say: no matter how good we make the conditions under which a person has the choice to avoid crime – he’s not destitute, he’s been given a good education, and so on – nonetheless some people are maybe psychologically born in such a way that they’re going to commit crimes even if they’re not deficient in other ways. Perhaps. Who knows. So you could imagine that person saying: “Okay, so you’re saying that you offered me a valuable protection against the possibility of getting punished by offering me the choice, under any of these good conditions, to live right. It may have been valuable for some people. But it didn’t do any good for me.”
And I sort of think that’s like vaccination. Vaccination doesn’t work for everybody. But one of the things we’ve got to do to protect people against being sick is to give them the chance to get vaccinated. There’s a limit to how far we’ve got to go. It doesn’t work for everybody, but if you haven’t done it then you haven’t done it. If you’ve done enough, then you’ve done enough.
It’s a question of how much you’ve got to do. That seems sort of hard-hearted, but that’s the place where I bite the bullet.
So the starting point is that there’s a social need for a system of punishment?
You’ve got to have a social need, right. You’ve got to have a social need. It’s got to be an important need. This has to be a justifiable way of addressing that need, so there can’t be other less destructive ways of serving that need. And given our present penal system, it seems to me rather doubtful that there aren’t ways of reducing crime that don’t involve this monumental cost to individual people who are incarcerated, and to the communities from which they’re taken.
Mark Kleiman’s work in sociology and criminology is a good example of other strategies for trying to reduce crime. And there are other things out there that are trying to do that. So I think it’s just a horrible thing, the institution of criminal punishment that we have now, but obviously we can’t simply abandon it. In particular, we’d be abandoning the communities in which there’s high crime and leaving people vulnerable to being preyed upon more, perhaps. So you’ve got to do something. But I think most people simply aren’t aware of how bad the thing that we’re doing is. So that’s a serious problem.
What about the third problem – what you call the “personal problem of free will”?
I think it’s a tougher problem. To explain how I can continue to think of my reactions as mine, as reflecting me, if I know that there is a physical story which leads to my saying the things I’m going to say, my being like this, and so on. That seems to be the most perplexing of the three problems.
Although, for the reasons I’ve mentioned in talking about the person who was ultimately undeterrable, I think there is still a problem in the second case. But I think it’s more puzzling in the third case. How can I think of myself as a person if I also think of myself as somebody who’s caused to act the way I do?
And I think that is a puzzle, one that’s not going to go away. But, at the end of the day, I think I feel about it sort of the way I feel about the first one: that is, when I think about myself, what I’m thinking about is my present psychology. And an awful lot of that is just spontaneous.
I mean, it’s not only true that we do some things on impulse, or whatever. Some of those things maybe we don’t think of as really being us, as being uncharacteristic and anomalous. But much of what happens in conversation is un-premeditated. You asked me a question and an answer comes out. So: that’s me. (Laughs.)Sometimes I find out who I am by seeing what I say.
So I’m prepared to accept that, and I think: what alternative is there? It’s not as if there’s some other ectoplasm in there that’s really me – to me, that idea doesn’t seem to make any sense at all. So I think I have to be comfortable with that. But I can see why it’s a continuing, perplexing problem.
In the “Significance of Choice,” you gave an account of the different kinds of choice that retain their value even if causal determinism were true. With some of them, it’s obvious why that should be the case: the instrumental value of choice, for example. When you go to lunch in a restaurant, what food you get served should depend on your choice even if causal determinism is true. But what you call the “expressive” dimension of choice seems more problematic: that’s the idea that, rather than getting someone else to do it, it’s important that you choose the present you give to your wife. But you might think that if it’s causally predetermined that you were going to choose the present for your wife that undermines the significance of the gesture. She might say: well, it’s nice that you took the time but, after all, that you chose to take the time was just causally predetermined as well.
Well, that shows the interaction between these different forms of the problem, I think.
That is, my first response is to say that it’s more meaningful to my wife or my friend or my retiring colleague, or whatever, if I choose a present for them myself. Now, it’s partly because my going to the trouble of choosing it shows a level of concern. But also it’s that this being the thing that I thought of gives it an extra meaning – a meaning over and above the way in which my taking the trouble to think about it is evidence of a certain kind of concern. And I don’t think that that kind of significance is undermined by the fact that my exercising originality and insight in thinking of the appropriate gift is produced by a causal process.
That’s partly because – though this may just be seen as repeating the same problem somewhere else – as I said in talking about the justifiability of blame, our relations with other people are our relations with psychologically embedded people. Who they are to us is defined by the way they typically respond, and so on. What’s meaningful to me in relation to my friends is the way they actually are – not the way that something independent of what they actually are determines how they are. What they are as psychologically embodied and characterized creatures is what matters to me. So as for the significance for others of my exercising originality – that seems to me to go with the kind of story that I was telling about blame.
Another version of the significance of choice is the significance to me of my apartment’s decorations reflecting my taste and originality and personality and thought about it, and so on. Now, that depends essentially on the third problem there of how the fact that if I go into a store and choose this sofa rather than that sofa – or this kind of Corbusier-style furniture rather than a lava lamp (laughs)– why that should be significant. How can it be significant for me that it expresses me,given that it’s just like my conversation: there’s some causal mechanism that’s ticking in a way that leads to my saying one thing rather than another when you ask me a question, and it also leads me toward picking out one kind of furniture rather than another.
So these three problems are interrelated. The reasons that lie behind our concern with choice in the case of moral reaction and in the case of personal expression in a way presuppose something like the significance for a person of the third type, personal significance. So if you really couldn’t solve that problem at all – if you felt that the idea of a causal explanation of my response simply destroyed personal significance – then it would tend to destroy the significance of choice in those other cases as well.
Apart perhaps from the purely instrumental version. Of the three versions of the value of choice the instrumental would be the one that is most robust against that kind of undermining. But also it seems the most superficial in some ways. All in all not unimportant – I don’t want to find liver on my plate… (Laughs.)
I was very struck by the last sentences of “The Significance of Choice.” There, you say:
“moral and, if there is such a thing, legal indignation towards lawbreakers is entirely in order and the suffering we inflict upon them may be justified. But in justifying these sufferings and inflicting them we have to say not: ‘You asked for this.’ But: ‘There but for the grace of God go I.’”
What does this mean for somebody who has broken the law? Do they say, not, “There but for the grace of God go I,” but rather something like: “There for the lack of grace of God, I go?”
Well, as somebody who doesn’t take the idea of God very seriously – or finds it hard to understand in a way – I just sort of tossed off that line, and I wished afterwards that I’d just said: we have to see this as a matter of luck. Because if you take it literally, the idea of God’s grace isn’t what I would say. So I felt a little uncomfortable about that line.
But I think there’s an important thing right about Ronald Dworkin’s discussion of this in Justice for Hedgehogs. He says, essentially, that the person who’s actually punished can’t disown what he’s done completely. He can’t say: “Well, it’s just the cosmos acting through me; it’s got nothing to do with me.” He can’t do that without undermining his relation to everything he’s ever done and his relation to everybody else. It would be a kind of personal nihilism. It’s not just that it would be impossible to hold, but that it would be unattractive.
Now, you could say: “That doesn’t show that it doesn’t make any sense. It just shows it’s the only thing that makes sense of these other attitudes.” That’s why it’s hard to state this position in a way that makes it feel robust. But nonetheless it seems to me there’s something right about it. So I guess you need to put it the other way round, and say: “Ownership stands and fall as a block. Unless there is some special explanation of why you did the crime you can’t say: ‘it was just the Big Bang acting through me’.” It stands or falls as a whole unit. That’s I think the core idea behind Dworkin’s remark.
OK, so let’s accept that. You can’t do it piecemeal without some special excuse – which would be a whole different game. It stands or falls as a block. So why not let it fall? And Dworkin in effect says: but nobody would want to do that. It would destroy the meaning of your life. But then the next step is: “well, if it’s really correct that your life doesn’t make any sense if it’s all the Big Bang causing you – perhaps with a few randomizing elements built in along the way – then why hold a nonsensical view just because you’re attached to it?”
So then I think you have to put it in the inverse way at that point, and say: “Yes, that would be right if the fact that these things I do are causally explicable really did provide sound, undeniable reason for declaring it all to be meaningless. And why would it do that?” And then I get back to something like the moral problem of free will. It would make everything meaningless if it separated the things that I do from some kind of prior cause or condition which alone would make them mine, would make them make sense. And you want to say: well, no, it doesn’t do that.
If the only intelligible conception of the self were this kind of dualist conception in which there’s this kind of acausal me that is intervening in the universe and thinking these thoughts as the result of a self-caused entity coming up with these thoughts that then change the direction of the causal world – if that was the only thing that makes sense of our caring about these things, then we’d have to give it all up, right? But why think that’s how it is? Why not say something like what we said earlier: what makes it seem like us is the fact, a), that it comes from us and, b), that it comes from us in a way that’s enmeshed with all these other things – and reflects those things, on the whole.
So I think that something like the answer I was giving to the moral problems is relevant here: what we are to ourselves as well as to others just is this complicated set of psychological phenomena. There’s no reason to say that that’s something we have no reason to care about. So you’d need a more powerful spelling out of the story why the causal explanation makes that something we have reason not to care about than I’ve seen in order to say that, given the fact that it all stands or fall as a block, we ought to let it fall. That’s what I think. It’s a hard question – but I do think Dworkin’s discussion of it is rather good.
The impetus for modern conceptions of morality might be said to stem from a desire to protect ourselves from being responsible for things over which we have no control. Unlike most of the Ancients, we think it’s wrong to blame you for the status you were born into. Unlike some Christians, we don’t want to make what happens to you depend on whether or not you were born with grace. You can only be responsible for actions if you did them – and if you, in turn, were responsible for choosing to carry them out. Now, your response to all of this seems to be: well, if you go that way, then at some point you have to give up on all sense of personal agency altogether.
Galen Strawson has an excellent article on the impossibility of moral responsibility. And I read it backwards, as it were. That is: Strawson says that moral responsibility is impossible because, in order to be morally responsible for an action, we have to be morally responsible for being psychologically the way that we are, which led us to do the action. But we can’t be morally responsible for being that way unless being that way arose from a conscious choice on our part to be that way, which itself was a reflection of our being some prior way which we were responsible for. But we can’t be responsible for that… and so on. This regress can’t end anywhere, so this whole thing is incoherent.
And I want to say: what’s right about that, what drives the regress, is the idea that the self that is responsible would have to somehow be prior to any particular collection of psychological characteristics. But such a bare self would seem to me to be meaningless. What Strawson’s article shows is that you can’t imagine a self bare of psychological content making choices it would not have any reason to make. So every step you go back just replaces one given psychology with some prior given psychology. And the conclusion I draw from that is: yes, our appraisal of a person – and our understanding of a choice as a meaningful choice – are always against a background of some given psychology, which is what a person is like. And our moral assessment of a person’s actions is an assessment of it as revealing the person’s being like that, which is some fixed set of psychology. And the significance for the person of doing something lies in the fact that doing it reflects the way they really feel themselves to be, psychologically.
So whether you’re talking about moral praise or blame, or talking about what I call the personal problem of free will, what it shows is that the notion of the self that’s meaningful either as an object of appraisal or as a ground of individual meaning always has to have psychological content. And then I want to say: the fact that that’s to some degree arbitrary – in fact, is completely arbitrary in the sense of being caused – isn’t arbitrary for the person because it really is him. And it isn’t arbitrary for his friends, because that’s what he is actually like. So it isn’t arbitrary in the relevant sense – even though it may be caused by all these factors outside his control. That doesn’t stop it from being him. Strawson’s excellent article brings out that point – it’s a kind of reductio of this bare-self-view.
Now, when Strawson addresses the question of why responsibility should always require this stepping back, he isn’t very clear about what he means by responsibility. Except, he says, “it’s the kind of thing that would make the ideas of heaven and hell at least make sense.” So there we seem to be getting into a punitive idea – what I would think of as opportunity to avoid. He doesn’t literally say that, and I’ve never had a chance to ask Galen what he really meant by that, but that seems to me the way it looks, anyhow.
Click here to get to Part II: What We Owe to Each Other.