Galen Strawson’s Modest Requirement for Moral Responsibility.

It’s been a while since we’ve had a non-Smilansky related discussion (I wish I’d thought of Illusionism!), so I thought I’d post some thoughts I’ve expressed in comment threads but never really put out there for public criticism and ridicule. So here goes:

John noted that in his upcoming Hourani lectures he will accuse Strawson of having an unreasonably demanding requirement for moral responsibility, what he calls “total control.” Others have attacked Strawson for arguing that we need to be ‘wholly responsible’ for our characters in order to be morally responsibility for our actions, claiming that this is asking for too much. Now ‘total control’ and “wholly responsible” can mean a lot of things, but I believe that they are often interpreted in ways that do not do justice Strawson’s theory.

The source of the misconception, I believe, lies in Strawson’s use of the causa sui concept. The problem with the causa sui language (along the entertaining Nietzsche quote that inevitably accompanies it) is that there is rarely a discussion of how much of our character we would have to ultimately responsible for in order to be causa sui.

The claim “people have to be causa sui in order to be morally responsible for their behavior” sounds crazy if you take it to mean “people have to be ultimately responsible for their entire character in order to be morally responsible for their actions.” However, “people must be ultimately responsible for one small aspect of their character in order to be morally responsible for their behavior” sounds a lot more reasonable—at least at first blush.

Here’s what I take to be Strawson’s position. If we could be ultimately responsible for even one tiny aspect of our characters, then we could morally responsible for our actions. (note: I am substituting “character” for “the way we are” here for the sake of simplicity.) Which means: even if every feature of my character ccan be traced back to forces outside my control except one, I could nevertheless be morally responsible for actions determined by my character. Moreover, I don’t even need to be “wholly responsible” for that one feature of my character. I just need to be morally responsible for one of the determining factors that shaped that one aspect of my character.

Bottom line: as long as there is one determining factor of one aspect of my character that does not trace back to forces outside of my control, I can be morally responsible for my actions.

Framed this way, Strawson's theory sounds a lot more reasonable. Maybe a little too modest, in fact. Better still, I think it does reflect a commonsense view about the conditions for moral responsibility. Every thinking person recognizes that we are not morally responsible for most of what makes us who we are. We did not choose our DNA, we did not choose our parents, our early environment, how we are treated based on appearance, intelligence, and so forth. But, to use an analogy from one of John’s recent papers, we believe that though the hands that people are dealt differ widely, we can be morally responsible for how we play that hand. A very small part of how we handle all of the circumstances life throws at us is ‘up to us’--and that is how we acquire moral responsibility for our characters.

Of course, the whole discussion then turns on how to interpret ‘up to us.’ But one plausible interpretation, I suggest, is this: We believe that while most influences on our decisions can be traced back to forces beyond our control, there is a very small influence on some our decisions that cannot. We only need ultimate control over that tiny influence on certain decisions to be morally responsible for our actions. But we do need it if we’re going to acquire any moral responsibility for our characters and then actions which issue from our characters. Otherwise, luck really does swallow everything—even that one tiny influence we were desperately clinging to (in the way that Yankee fans cling to their dominance in the late 90s).

Again, at first glance, the requirement hardly seems that demanding, even if further inquiry reveals it to be logically impossible. But as anyone who has taught free will knows, most people have not engaged in that kind of inquiry. And so it would not be surprising if they do place that condition on moral responsibility.

One last thing to note: although compatibilists attack Strawson for his unreasonably stringent condition on moral responsibility, everyone else (philosophers and folk) attacks him for denying that this condition can be met. This suggests that everyone besides compatibilists seem to accept that Strawson’s modest requirement is a genuine requirement for moral responsibility.

(Note: I am not accusing John of misunderstanding Strawson on this point. See his “The Cards that are Dealt You” referred to earlier in the post for a subtle discussion of Strawson’s position. But I do think this misconception is out there, to some degree anyway.)

Comments

[Paul]

I wonder if this has gone to far the other way and is now too modest. It's seems odd to think that my being partly responsible for developing the character trait of being lazy makes me also responsible for being honest. I'm just not sure that responsibility for one gets you responsibility for all, unless perhaps there is some "unity of character" thesis that I've overlooked. I have no real argument other than "this seems odd" (which is not likely to win any arguments about free will and responsibility). Perhaps the oddness would go away if it was clear how being responsible for one part leads to changes (for better or worse) in other parts.

[Kip Werking]

Part (all?) of the problem is, as Tamler notes, and I tried to note in my thread "The Contours of Being Causa Sui", is that it's not clear what Strawson, or Nietzsche, meant by being causa sui.

Obviously, Strawson and Nietzsche makes a psychological observation, the gist of which immediately resonates with incompatibilists, or at least no-free-will types.

For example, Paul asks whether this immodest position goes too far? Well, hopefully it does---no-free-will types wouldn't want to win by such a thin margin. If Strawson is right that we don't even have (too)-modest responsibility, we certainly don't have actual responsbility---whatever that is (presumably somewhere between (too)-modest responsibility and Total Control).

But answering that incompatibilist consensus doesn't solve the larger problem of determining what (moral) responsibility is---it provides no knock-down argument against compatibilists.

[Tamler Sommers]

Paul,

I should have noted that this requirement is a necessary, not sufficient, condition for moral responsibility. I imagine that Strawson would add to this most of the normal compatibilist requirements for responsibility as well. Moreover, if this requirement were met, it would only grant responsibility for specific aspects of your character as well as the acts which issue from them. So, you would only be responsible for your laziness and any acts influenced by your laziness, unless you also were MR for other traits as well.

Kip,

Yes, you've been on the case for a while, as you note. (In fact, I urge everyone to check out his paper "You are the Cards that are Dealt You.")

[Container Gardening Tips]

I think the big thing about "moral responsibility" is that it's all subjective. One person may believe something is highly immoral, while another person believes it's perfectly fine. Take the abortion issue. Some people believe it's immoral to "kill a child", while other people think it's immoral to "bring an 'unwanted' child into the world". Who knows which side is really right? It's all about the morals each person decides for themselves. Unfortunately there's just no changing the moral ideas of most people. Once we believe something, it rarely changes.

[Jonathan Farrell]

A very quick note: this 'one tiny aspect' view outlined here looks very similar to that held by Robert Kane. Kane holds that in order to be Ultimately Responsible for an action it must be that we can trace at least one causal chain back from that action to the agent, and that the chain does not continue on back 'through' the agent to things outside of them.

[Paul]

Tamler, thanks for the clarification. But if he is just saying that we are responsible for the acts that result from the character we were partly responsible for developing, then I agree with Jonathan that this is starting to sound like a less-robust version of Kane's project.

Maybe I'm misunderstanding Strawson's project, but if he's saying we need at least these modest character forming actions for responsibility, then I've another worry. If determinism is true, then these just can't exist at all. Doesn't a little indeterminism about these types of actions (as opposed to quantum events) entail determinism is false? If not, then what's the argument that these modest actions are the only ones not determined by antecedent conditions?

[Tamler Sommers]

Jonathan and Paul,

I think that's right--Strawson's requirement similar if not identical to Kane's. More evidence perhaps that there is a misconception out there about Strawson's position. I don't remember Kane being characterized as believing that 'total control' is a necessary condition for moral responsibility, or that on Kane's view one needs to be "wholly responsible" for one's character in order to be morally responsible for one's action.

Paul, Strawson would concede that indeterminism is possible, even at the level of human behavior. But he would argue that indeterminism does not help agents acquire control for any aspect of their character.

[joe O.]

Perhaps on reflection, the incompatibilist will realize that certainly causa sui allows for "heaven and hell" responsibility, but responsibility that allows one to be an apt target of reactive attitudes isn't quite so 'needy'.

So, rather than asking too much of the causa sui concept, Strawon might be seen as asking too much of the moral responsibility concept.

I think this might resemble or be Randy Clarke's idea in his article in the midewest studies a couple years ago.

[Patrick Todd]

I think it's worth noting how the dialectic between (some, at any rate) compatibilists and incompatibilsts seems to unfold along these lines. The incompatibilist is seen by the compatibilist as making the demands on moral responsibility impossibly hard to meet, i.e. that it's impossible to be "wholly responsible" for our characters, or to be a "causa sui", etc. I feel like - though I don't have any examples off the top of my head - many compatibilists then move to the thought that such conditions are in all likelihood not required for moral responsibility. In other words, if a view postulates conditions that are impossible for us to meet, that's good reason to think that view has gone wrong.

Of course, this expresses the optimistic outlook that we are very plausibly morally responsible. If one is confident that we *are* MR, then if some conditions turn out to be impossible to meet, one will also be confident that those conditions aren't necessary.

And the incompatibilist could be seen as saying, "Well, if what you say is impossible really is impossible, then what we should conclude from this is that we're screwed, not that those conditions aren't required. They clearly *are* required. And I don't see why we should be so confident that we *are* responsible."

In any case, I like your points here, since I think they go some way in helping to bypass the move to say that the relevant conditions are impossible to meet. It's indeed hard to see how we could enjoy "total control" or be "wholly responsible" for our characters, but it isn't so hard to imagine that we're responsible for parts of them. Accusations of extravagance and immodesty are thus harder to make stick.

[Q the Enchanter]

This may be an insipid point, since I haven't read Kip's paper. (Plus, I'm just generally an insipid person.) But the analogy of "playing the cards that are dealt you" to freewill-under-constraints has always struck me as fundamentally flawed, because a key issue in contention is whether the ability to play the cards that are dealt you is itself one of the cards that are dealt you.

[Nick Hopkins]

Not an insipid point I think, but also not fatal to the idea of free will, as I'm not sure that it is anything more than another constraint.

Essentially we could view any card player above a certain level as having some card playing ability. There may be different levels of such ability across people, and within life histories across time. That does't mean that they don't play cards.

[Mark Smeltzer]

Tamler,

I'm curious about a comment you made above. You said, "But he would argue that indeterminism does not help agents acquire control for any aspect of their character."

What does it mean to be in control of our character? What benefit would that have? Moreover, what does that have to do with the way we react to and deal with people in the real world?

[Tamler Sommers]

Mark,

You're taking that quote a bit out of context. I was responding to a specific question about how quantum mechanics seems to undermine the truth of determinism. My point, one that's made very often, was that indeterminism of that kind cannot help us acquire responsibility for any part of our character. Because acts that shape our characters would still trace back to factors beyond our control (random ones perhaps).

[Mark Smeltzer]

Tamler,

One good accusation deserves another: I believe you're taking my comment too contextually.

I am genuinely interested in hearing your responses to the questions I ask. The comment I highlighted is what sparked my interest, but I don't think the context of the comment is relevant to the content of the questions it sparked.

My assumption is that you think if agents lack "control for any aspect of their character" that it would be a problem for someone. For whom? And what's the problem? (E.g. does this problem apply only to Kane and his view?)

[Tamler Sommers]

Mark,

Fair enough, but you won't like my answer. Lacking any ultimate control over any aspect of one's character is a problem for many people is this sense: it shows that we cannot be morally responsible for our actions.

Here's an example: when people argue that a criminal is not responsible for his actions because of a deeply troubled childhood, what's the most common response?

It's something like, "Well, other people who had equally troubled childhoods were able to overcome them and become decent moral people." The implication, I take it, is that while having a troubled childhood presents difficult moral obstacles and restricts a person's choices significantly, some small part of the outcome--whether we become cruel and vicious or troubled but ultimately decent-- remains "up to us."

See the original post for my interpretation of "up to us"--which involves an element of ultimate control. My suggestion is that if people thought we lacked that small element of ultimate control, most would think that moral responsibility goes out the window.

[Mark Smeltzer]

Tamler,

Can't we make a distinction to help in those situations? Here's the distinction I have in mind: there are people who really deserve to be treated a certain way and there are people who only appear to be people who deserve to be treated a certain way. Add to that a corollary that states that it is often hard to tell the difference.

If the distinction and the corollary are true, it suggests that are cases where we ought to reserve or suspend our impulse to act on more intense reactive attitudes.

Wouldn't that allow us to deal with the kinds of cases you just brought up without leaping to the "no one's ever responsible" conclusion?