Brittle People and the Varieties of Compatibilist Free Will not Worth Wanting

I want to use two compatibilist tenets against each other: (1) that we should understand free will in a way resulting in free will being valuable and (2) that in determining whether an agent has free will, we should focus on what the agent would have done in counter-factual worlds.

Neither of these tenets is necessary to compatibilism, although I believe both are popular.  In particular, I think Dennett has popularized both.

Of course, anti-realists about free will have challenged (1).  In an underappreciated passage from the introduction to Living Without Free Will, Pereboom directly questioned Dennett’s insistence that free will is valuable.  That is one way to attack the compatibilist position.

But there is another way to attack the compatibilist view.  What if we humor the compatibilist?  What if free will must be valuable?  Does the compatibilist version of “free will” even satisfy that standard?  In other words, are the varieties of free will that Dennett (and others) asserts are worth wanting really worth wanting?

Before going further, let’s distinguish between two kinds of compatibilist free will (cfw):

A.    Thin CFW: the agent acts as a normal person in the agent’s world, and does all of the things we associate with compatibilist free will: rational deliberation, reflecting on lower desires in light of values and higher desires.  But the agent doesn’t necessarily do any of these in other worlds.
B.    Thick CFW: the agent acts as a normal person in the agent’s world, as described above.  But, in addition, the agent also acts as a normal agent, and responds appropriately to reasons, in nearby counterfactual worlds.

To illustrate the distinction, consider normal Kip and brittle Kip.  Normal Kip is just like me.  Presumbly I have thick CFW.  I am typing this post right now, but if the world was slightly different, I would presumably be doing something different but still reasonable (understood in a rough compatibilist sense).  For example, if I had been thirsty, I might have made a cup of tea.

Brittle Kip is like normal Kip as far as this world is concerned.  He has thin CFW.  But brittle Kip doesn’t have thick CFW.  Thus, in this world, the lives (both internal and external) of brittle Kip and normal Kip are the same.  Brittle Kip types this post, just as I am typing this post.  Brittle Kip will wake up to go to work tomorrow, just as I will.

Unlike normal Kip, though, brittle Kip cannot handle other worlds at all.  The slightest difference in perception throws him.  Perhaps, for example, brittle Kip is configured so that, in response to any world other than this world, Kip takes a nap all day.  In these other worlds, brittle Kip is not necessarily rational, his higher order desires don’t mesh with his lower ones, etc.  In short, the slightest perceived deviation from this world “breaks” Kip’s normal functioning life.  That’s why we call him brittle Kip.

Now here’s my question: suppose brittle Kip actually lives in this world (the world where he functions just like normal Kip).  Does brittle Kip have anything to complain about, relative to normal Kip?

This is where I want to make (what I hope is) an interesting point: (C1) it seems to me that brittle Kip has nothing to complain about, and (C2) it seems to me that most compatibilists would disagree with me about C1.  Indeed, it seems to me that many or most compatibilists would deny that brittle Kip has free will.

Regarding (1), I say, “yes, brittle Kip would stop functioning in a different world, but who cares?  He doesn’t live in that world.  He lives in this world, where he functions just fine.”  He has nothing to complain about, because all of the problems that arise in other worlds don’t arise in the one world that matters—this world.  And because he has nothing to complain about, the extra features of thick CFW are not worth wanting.  They’re simply irrelevant to brittle Kip’s situation.

In saying this, I think I am echoing Frankfurt, who would say about his famous examples, “yes, the agent can’t do otherwise than he actually does, but he is still morally responsible.  The counterfactuals don’t undermine the moral responsibility.”  Perhaps this is why Pereboom adopts a Frankfurt-style view about sourcehood, while refusing to adopt compatibilism.  I don’t know, however, whether Frankfurt would say that brittle Kip has free will.

In sum, I think compatibilists place altogether too much weight on counterfactuals.  I have never understood how what an agent does in another, different world could be relevant to the powers and freedoms that the agent has in this world.

Gardeners (and especially compatibilists): is thin CFW enough to be free will?  Does brittle Kip have anything to complain about?  And if he doesn’t have anything to complain about, then is a kind of free will based on counterfactuals really worth wanting?

Comments

[Sofia Jeppsson]

Let me see if I get this straight. The difference between normal Kip and brittle Kip seems to consist in the fact that brittle Kip really lacks a lot of self-control, strenght of will etc. The slightest little thing is enough to make him go depressed and apathic, or freak out and become an irrational lunatic. But through extraordinary chance it so happens that during his entire life nothing that so upsets him ever happens, the environment always coincide in such a way with his current mood as not to throw him off balance.

Brittle Kip would then be analogous to a person who's born with, say, severe brittleness of the sceleton. But through extraordinary chance he still never breaks a bone, so this health problem is never detected.

I think one can grant that if a person is suffering from a severe weakness of some kind, but due to extraordinary circumstances this never have any consequences, well, he's not worse off due to the weakness. Still, in general, in about 99,9% of the cases, this kind of weakness WOULD be a problem for the person having it. So there's nothing strange in saying, as a general statement, that it's worth wanting to be free from such weaknesses.

[Kip]

Thanks for the comment, Sophia.

You write:

"Still, in general, in about 99,9% of the cases, this kind of weakness WOULD be a problem for the person having it. So there's nothing strange in saying, as a general statement, that it's worth wanting to be free from such weaknesses."

I disagree. The 99.9% of cases you will mention will happen with 0% probability. So they're irrelevant. A person might desire to be free from the weaknesses that happen in those worlds. It might even be natural or intuitive to do so. But those weaknesses will never happen to the agent in this world, so it's pointless to desire them.

[Kip]

Sorry for misspelling your name.

[Eddy Nahmias]

Kip, you suggest that Brittle Kip has the same capacity to deliberate and act rationally as Kip, but that's not right. Suppose both of them act rationally at time T (e.g., they both decide to exercise to improve their health). This sentence is true of Kip and, I think, *constitutive* of his being rational: If Kip believed at T that exercising would kill him, he would not decide at T to exercise. But that sentence is, given your description, not true of Brittle Kip. That makes Brittle Kip lack an essential capacity to be free and responsible according to compatibilists (one which incomaptibilists presumably think is necessary as well). Counterfactuals matter to anyone trying to understand the nature of certain properties, including rationality.

[Paul Torek]

I agree with Eddy's point and would like to illuminate it with an analogy. Philosopher K thinks that knowledge is valuable. Philosopher T claims that only true belief is valuable. T points to Brittle Kip, who believes things on shabby evidence but just happens to get it right every time.

It seems to me that K can admit that the lucky believer gets some of what's important, but deny that he gets all of it. K can reasonably hold that the reliability (or other counterfactual properties) of knowledge makes it better than mere true belief.

[Kip]

Eddy:

You write:

"Counterfactuals matter to anyone trying to understand the nature of certain properties, including rationality."

I'm not saying that Brittle Kip is rational or irrational. I'm saying that, if "rationality" depends on how Brittle Kip acts in different worlds, Brittle Kip shouldn't care about that kind of rationality.

All that matters to Brittle Kip, all that he should value, is how he reacts in this world. That's the only world he'll ever live in. All other worlds are irrelevant.

If that makes him "irrational," so be it. His life is identical to, and just as rich and rewarding, as normal Kip's. So it would be irrational (!) for him desire cross-world rationality as you define it.

I'll add an example to make my point:

Suppose Brittle Kip is walking along the street one day. Suppose he sees a dog. Brittle Kip likes dogs, and so he reaches down and pets the dog.

Now further suppose that, if the dog had been a cat, Brittle Kip would have taken out a knife, cut out chunks of rubber from a tire, and started eating them, even though he hates the taste of rubber.

Suppose you live in this world with Brittle Kip. You walk up to him. The following dialogue ensues:

Eddy: Brittle Kip, if that dog had been a cat, you would have done something really irrational and stupid and embarrassing.
Kip: So? It wasn't a cat, it was a dog.
Eddy: But doesn't that bother you?
Kip: It would bother me if it had been a cat. But it was a dog. So I petted it, just as I like to pet dogs. All is well.
Eddy: But don't you see, that makes you irrational. Rationality must be understood in terms of counterfactuals. So, because you would have done that ridiculous thing in other worlds, you're irrational.
Kip: So? Why should I care about that?
Eddy: Because irrationality is bad!
Kip: How is it bad, other than you saying it is?
Eddy: Well, if the dog had been a cat, you would have eaten the tire, you could have gotten sick and been taken to the hospital.
Kip: But it wasn't a cat. It was a dog.
Eddy: But if you're irrational, you can't have free will.
Kip: But I thought you said free will was worth wanting?
Eddy: It is worth wanting!
Kip: Well, I'm confused. You say the following is worth wanting: acting reasonably in fake counterfactual worlds. But why do you care what happens in fake counterfactual worlds? You don't live there. So what happens there doesn't matter to you. So, if free will requires that kind of rationality, how is this free will worth wanting? I don't think it is.

What's worth wanting is acting reasonably in the world where you live. But I already do that.

[Kip]

Paul,

You write: "K can reasonably hold that the reliability (or other counterfactual properties) of knowledge makes it better than mere true belief."

You cite "reliability" as making knowledge (as defined) better than true belief. But you say that "lucky believer" gets it right every time. So lucky believer's beliefs are as reliable as the knowledge holder's. So reliability is no reason to prefer one to the other, because true belief is 100% reliable.

[Justin Coates]

Kip, a true story:

I played a round of golf this morning, and I had one drive that was just spectacular. Now I'm sitting at home watching the PGA Championship and all the pros are hitting equally good (and in most cases better) drives on just about every hole. Of course when I'm setting up for my second shot (a 3 iron sliced into the short rough in case you're curious), it doesn't matter that in all the nearby possible worlds I probably only drove it 150 yards and deep into the right rough. Similarly, when the pros are setting up for their second shots, it doesn't matter that in all the nearby possible worlds, they're in the same position.

But surely you wouldn't want to evaluate my shot in the same way you would evaluate the pros' shots. After seeing my shot, you'd say things like "lucky," or as fellow Garderner Garrett Pendergraft put it, "miraculous." But when Tiger Woods hits the same shot, we say things like "good" or "great shot" (never "lucky" or "miraculous"). Why? Well, because pros are competent golfers in a way that I'm not. And competence is what grounds our positive evaluative assessment of their play in a way that doesn't ground positive evaluative assessment of my play. Don't get me wrong, it's cool to hit a great shot, but one great shot doesn't make me reliable or competent--only facts about nearby possible worlds determine my competence.

Like golf competence, normative competence also grounds a class of evaluative judgments, and that seems valuable to me. By appealing to counterfactuals, we can explain why Tiger deserves more criticism for hitting a bad shot than I do, and we can also explain why I deserve more blame for my wrongdoing than Robert Harris would. Maybe you don't think that's valuable, but I do.

So I think Brittle Kip should care. Since he's not competent (in the relevant sense), he's not subject to the same sort of praise and criticism that Normal Kip is subject to. And being subject to this sort of assessment seems wrapped up in what it means to be a person.

So maybe that's a reason for thinking that facts about other possible worlds could be valuable.

[Kip]

Justin,

I think there's a disanalogy between your case and the one I originally suggested.

You're making the following comparison:

Justin-hitting-lucky-shot = Brittle-Kip-Acting-Rationally

Here's the disanalogy: You only hit one lucky shot. It was truly lucky. If you went to hit another one, you probably wouldn't make it, and if you tried another 100, you wouldn't with almost 100% certainty.

Brittle Kip's rationality is not like that. He doesn't do one rational act, and then try to eat tire rubber with a fork for the rest of his life. His entire life is, for all appearances, rational. Because Brittle Kip gets "lucky" over and over again, I think he's more like Tiger Woods that you suggest.

The only ground for saying Brittle Kip is irrational is that he does ridiculous things in worlds that are different than this one. But Brittle Kip will never be in any of those worlds, so I'm pretty sure he shouldn't care.

[Aaron Boyden]

I'm not completely sure that the brittle Kip scenario is coherent. Or at least, I think that if it were spelled out enough to make it so, it would become apparent that the case is so extraordinarily strange that we probably shouldn't trust our intuitions about it. I think that the case involves a serious (and sadly common) confusion about modality.

As described, what's different about brittle Kip is that different worlds are close to his world. But what makes a world close is similarity. Thus, it's not possible to have Kip's world be extremely similar and have the nearby worlds be a completely different set; similar worlds are near similar worlds.

Of course, the scenario is very sketchily described, and if it isn't possible to have a world just like Kip's with very different worlds nearby, it is of course possible to have a world quite similar to Kip's in some respects and quite different in others with entirely different nearby worlds. But in that case, it makes a great deal of difference what the differences are; perhaps if we knew what they were, we'd have different intuitions about the case, and it hardly seems safe to assume that there's some set of differences which won't matter which will give the desired result. After all, to drastically change which worlds are nearby, the differences have to be quite significant in some ways.

To take an example, blockhead is like brittle Kip, but we wouldn't even be tempted to think blockhead is meaningfully free; blockhead isn't even sentient. It seems possible (indeed likely to me) that all brittle Kip scenarios would end up being at least somewhat like blockhead if they were more thoroughly spelled out.

[Sofia Jeppsson]

Kip, you misunderstood (but I might have been unclear in my post). I meant like this. If you, in THIS world, have 1000 different brittle people, about 999 of them would probably suffer all kinds of bad consequences on account of their brittleness, while 1 of them might be as lucky as brittle Kip in your example. That's why we could say as a general rule that brittleness is a bad thing, alhtough Kip might be an exception.

[Kip]

Sofia:

Let's be precise. By a "Brittle Person" I mean someone who acts normal, and gets by just fine, in their actual world, but who doesn't function properly in other worlds. I am saying that those Brittle People, as I've defined them, have nothing to complain about.

You seem to be using "brittleness" in a more general sense to mean that the person functions properly in one world, and doesn't function properly in many other worlds---but doesn't necessarily live in the world where everything works. I agree with you that that general kind of brittleness is generally a bad thing (at least it is in all of the worlds whether the person tries to eat tire rubber with a fork).

[Neal Tognazzini]

Kip,

I'm not sure I understood what Aaron said above, but I agree with him that the issue here is a misunderstanding about modality, and in particular, the function of possible-worlds discourse.

You are trying to describe a situation in which everything is fine and dandy in the actual world but really weird in other worlds. But the whole point of possible-worlds discourse is to make it easier and more illuminating to talk about what's up in the actual world. The order of ideas here goes like this: "Presumably having free will is a matter of having some capacity or ability or something like that, but what is it to have an ability to do something?" "Well, it's at the very least to be such that if you were in certain circumstances, you would do it." "Okay, but how should we understand subjunctive conditionals like that?" "Perhaps if we talk about "ways the world could have been" that will help -- possible worlds."

Unfortunately, possible-worlds discourse sometimes takes on a life of its own and it's easy to think of it as completely divorced from what actually goes on. But part of its point is to give us information about how things actually are. So the simple reason why Kip should care about what he does in other worlds is simply this: he should care about what he does and what properties he has in this world, and how he is in other worlds is inseparable from that. (Because talking about "how he is" in other worlds JUST IS a way of talking about what's up with him in this world.)

You're right that part of Frankfurt's point was to show that there are certain things that go on at other possible worlds that don't matter to us. But all he was saying is that the other-worldly stuff most incompatibilists seem to care about doesn't matter. Other other-worldy stuff certainly matters, and even Frankfurt would agree with that.

[Justin Coates]

Even if Brittle Kip gets lucky over and over and has a career that rivals Tiger's, he's still importantly different than Tiger: namely, BK's the luckiest man in the world (the Ringo Starr of golf), and Tiger's the most competent golfer in the world. And it's competence--not mere success--that grounds evaluative assessment. Of course, if BK was so lucky as to match Tiger shot-for-shot for a lifetime, we'd probably never know that he wasn't competent, but that's obviously irrelevant to the debate at hand. What matters is that were we to discover BK's lifetime of lucky shots, we'd have to radically revise our view of his golfing skills. I can only imagine the ESPN special...

Think of it this way. Take two poker players. A makes 2 million in her career, and B makes 1 million in her career. Who's the better poker player? Well, to answer that question we'd need to know more information. Suppose we discovered that A went all in whenever she had an inside straight draw--a low percentage play. But as a matter of luck, she always sucked out, and won. Further suppose that we discover that B always made the correct play, but in general, was unlucky, and frequently ran into boneheads like A. If we found this out, we'd say B was better than A, even though A's life was as full as successful poker as anyone else's has ever been. She just wasn't as good as B, and the best explanation for this is the fact that in all the nearby possible worlds, A goes broke all the time, and B's high-percentage plays regularly pay off big dividends. The fact that A was more successful than B seems irrelevant to our evaluation of skill. I take it the compatibilist wants to say that moral responsibility involves a type of skill or competence, and Brittle Kip lacks the relevant skills despite his seemingly normal life. We'd simply be mistaken to evaluate BK the way we evaluate NK (the same way we'd be mistaken to evaluate BK the same as TW or A the same as B). To avoid this conclusion, you'd need an account of 'competence' that doesn't appeal to possible worlds, and I'm not sure how that would go.

So of course Brittle Kip can complain. He lacks normative competence, and without it, he's not morally responsible. But Normal Kip is a morally responsible agent, and that's a pretty important difference.

[Kip]

Justin,

As in my response to Eddy, I think it's important to stick to my original claim: Brittle Kip has nothing to complain about.

Let's work with your poker analogy. You say that A *does* have something to complain about: A isn't really skillful, compared to B, just lucky.

Returning to my original claim: I don't think this is something for A to complain about. Why does A care if he's just "lucky" as you describe it, instead of skilled?

From the context of A's actual world (the only world A will ever live in), he at least appears quite skilled. And he reaps tremendous rewards from it (millions of dollars). So it appears that he doesn't have much to complain about.

You say that he is still lacking certain skills. But these are only skills that would help him in other worlds, worlds where he doesn't exist. So why would he care about those skills?

B might say: "Well, sure you're lucky in this world, but if we were in other worlds, you would fall flat on your face." And A's retort should be: "We don't live in other worlds. We live in this world, and in this world, I have all of the skill that I need. It would be irrational for me to desire more skill."

[Kip]

Neal,

You say: "So the simple reason why Kip should care about what he does in other worlds is simply this: he should care about what he does and what properties he has in this world, and how he is in other worlds is inseparable from that. (Because talking about "how he is" in other worlds JUST IS a way of talking about what's up with him in this world.)"

That's one traditional way of understanding ability. In that tradition, it is asserted that ability should be understood in terms of counterfactuals, and that such ability is essential to free will and moral responsibility. I'm not ignorant of that tradition, I'm just challenging it.

You say that the following are the same (one "just is" the other):

A. what is up with the agent in this world
B. what is up with the agent in other worlds

I disagree. In this world, I am typing this right now. In a little while, I will be drinking a Mojito. In another world, maybe I will be doing those things, and maybe I won't. The other worlds don't necessarily tell me anything about this world.

Now, I know what you intended by your statement in bold above. You meant: to understand whether the agent is able to X, or whether the agent is rational in X, we need to look at other worlds. In response to that, I am saying either:

1. You don't need to look at other worlds (e.g. if you want know what I am able to do in this world, I can only do what I actually end up doing. It's impossible for me to do anything else); or
2. Consider an ability that allegedly requires counterfactual analysis to measure. Lacking that ability is nothing to complain about. (E.g. if rationality means that I act reasonably not just in this world, but also in nearby worlds, who cares how I act in nearby worlds?).

[Mark Young]

Kip:
Suppose Brittle Kip is walking along the street one day. Suppose he sees a dog. Brittle Kip likes dogs, and so he reaches down and pets the dog.

Now further suppose that, if the dog had been a cat, Brittle Kip would have taken out a knife, cut out chunks of rubber from a tire, and started eating them, even though he hates the taste of rubber.

Kip should care about that -- unless he knows that he's never going to meet a cat (under sufficiently similar circumstances). How could he know such a thing?

You care about having fire insurance not because you know your house is going to burn down, but because you don't know that it's not.

A character with more-than-human knowledge of his future may not care about things that we humans with our all-too-human ignorance of our futures do (and should) care about. That that character's indifference would be rational says nothing about what is rational for us. It seems to me that the most your story shows is that the value of free will is contingent -- some entities (maybe) need not value it.

[Justin Coates]

A might care that she's not a good poker player, despite her winnings. She might care for the same reason any of us who have won a pot by getting lucky care (I find myself apologizing, etc.)--someone outplayed us.

Similarly, Brittle Kip should care that he's not morally responsible, despite the fact that he appears to be morally responsible.

As far as I know, everyone--compatibilist and incompatibilist--thinks that some kind of normative competence is required for moral responsibility, so unless you can offer an analysis of competence that doesn't appeal to possible worlds, then I've given you a reason to think other worlds are relevant.

Lastly, like Neal and Aaron have pointed out, I think you're confused about the nature of possible worlds. When you're ordering your mojito tonight, we typically think you could (should) have ordered a scotch. What makes it true that you could have ordered a scotch? Only that there is a maximally consistent state of affairs in which you--Kip Werking--order a scotch. So by thinking of another possible world, I have learned something about you-in-the-actual-world--namely, that you--Kip Werking--could have ordered a scotch.

[Cihan]

Besides all the discussion about modality, I think part of the debate hinges on what we might call internalism/externalism about value/happiness.

Consider the following examples - first two are Kane's and the next two are from Nozick -:

*In a possible world W, Alan the painter is really successful and revered for his art. His wealthy friend Sharon buys his art.

*In a possible world W*, Alan the painter isn't really a good artist. Nonetheless, his friend Sharon bribes the critics and friends so that it appears to Alan as if he is a good artist. Moreoever, Sharon is so good at this manipulation that the experiences of Alan in W and in W* are indistinguishable.

*In world X, Ravi leads a happy and successful career as a mathematician.

*In world X*, John is hooked up to an experience machine in such that the experience the machine is simulating for John is identical to that of Ravi. Experiences of John and Ravi are indistinguishable.

I think partly the reason why people might care about moral responsibility granting mechanisms that depend on counterfactuals is that similarly, most people would prefer W to W* and X to X*.

We might call this externalism about value/happiness. We don't just want to be the subject of good feelings - we also want our experiences to be connected to reality in some non-trivial way that'll capture the intuition behind preferring W and X over W* and X*.

In the same sense the experiences of agents in W* and X* are not well-connected to their reality, Brit Kip's mental events and his behavior may not be well-connected.

[Sofia Jeppsson]

Okay Kip, so you're definition of brittle includes that one is as lucky as "Brittle Kip". I think I finally follow you.

(But I think you confused things by letting Brittle Kip discuss dogs vs cats - in that discussion, he certainly doesn't seem rational in the way he talks and thinks. And it was part of the original example that he DOES seem that way in both speech and thought.)

So what you're saying is basically that someone who has a severe psychological disorder MIGHT be so lucky with his environment that he goes through life SEEMING like a perfectly normal person, both to himself and others. And only if he is that lucky does he count as "brittle", according to your definition.
You could still make an analogy with physical handicaps, and say that some severely handicapped people (like people suffering from an abnormally weak skeleton for example) could go through life without this handicap ever being detected, and if they do so, they're the physical analogy of "brittle Kip".

There's still, in THIS world, an objective difference between brittle people and others. They'd have to have some in principal detectable malfunction in their brains, here in THIS world. Otherwise they wouldn't act as weird as they does in the closest possible other world where you just change some little thing about the environment. It's just that in THIS world the malfunction is never detected, due to extraordinary circumstances.

Well, I think we can grant that if there are such people (the odds are astronomous when you think of it... much higher than my original 1000/1), they are no worse off just because they have this disability. In general though, people who are mentally disabled in this way ARE worse off, they lack very important mental capacities. The possible existence of a "brittle Kip" in our world doesn't change that general evaluation.

[pidda]

What if BK's wife started cheating on him (unbeknownst to BK), but in all other behavior towards him were the same; she loves him as much etc., but just feels like having a physical affair. BK isn't aware of it and isn't suffering in any way from it, but our intuitions are, probably, that he is (though unbeknownst to him) worse off for it.

Isn't the other-worlds case quite close to this scenario?

[Kip]

Mark and Cihan,

You both make brilliant points. Let me address them in turn.

Mark:

"Kip should care about that -- unless he knows that he's never going to meet a cat (under sufficiently similar circumstances). How could he know such a thing?"

Let's distinguish between:

A. The ability to respond reasonably to various situations is valuable, because any of those situations might be part of the real world; and
B. The ability to respond reasonably to situations that will never happen is valuable.

The way you've framed your question, it seems that you are defending A. Fine. That seems reasonable enough to me.

I was not trying to attack A, though. I was attacking B. I think many or most compatibilists would still defend B.

Cihan:

I think your analogy is quite apt. Of course, I'm perfectly satisfied with being plugged into the matrix. I think it's misguided to want more than that.

Sofia:

We seem to agree that brittle people, as I've defined them, have nothing to complain about. I'm glad we agree.

Justin:

You write: "A might care that she's not a good poker player, despite her winnings. She might care for the same reason any of us who have won a pot by getting lucky care (I find myself apologizing, etc.)--someone outplayed us."

But why should A cares if someone "outplayed" her? It seems that A is the one doing the outplaying.

You explain your assertion by saying that you find yourself apologizing sometimes in similar situations. But just because you happen to apologize in such situations doesn't necessarily mean that you should so apologize. I don't think you necessarily have anything to apologize for. Your behavior might have made you incompetent in other fictitious worlds; but in this world, your behavior was perfectly skilled and competent, as demonstrated by your winnings. You can just take the money and be happy.

You also write: "Similarly, Brittle Kip should care that he's not morally responsible, despite the fact that he appears to be morally responsible.

There is so much to say here:

1. It's not clear that moral responsibility is valuable. Maybe it would be better if we're all not morally responsible like Galen Strawson and Derk Pereboom say. So the fact that Brittle Kip lacks moral responsibility isn't necessarily a reason for him to complain.
2. Further, presumably thick CFW is valuable for reasons beyond securing moral responsibility (e.g. it secures greater happiness). Does the compatibilist really want to hinge the desirability of thick CFW on its securing moral responsibility? See also my fake dialogue with Eddy.
3. It's not clear that Brittle Kip is not morally responsible. Suppose that moral responsibility depends on freedom. The whole point of my thread is that thick CFW secures no more freedom for Brittle Kip than he has already. So Brittle Kip is just as free as compatibilist agents have always been. That much freedom has always satisfied compatibilists. So maybe, instead of saying "Brittle Kip isn't morally responsible, because he lacks thick CFW," compatibilists should instead say "Brittle Kip is still morally responsible, we were wrong to think that CFW has to be thick---thin CFW is enough."

As far as I know, everyone--compatibilist and incompatibilist--thinks that some kind of normative competence is required for moral responsibility, so unless you can offer an analysis of competence that doesn't appeal to possible worlds, then I've given you a reason to think other worlds are relevant.

I honestly don't know what you mean by "normative competence," or why you think Brittle Kip doesn't have it. People hardly ever use the term "moral responsibility" and philosophers cannot settle on a definition of moral responsibility so that they can even agree about whether people ever have it. Normative competence seems to be an even more suspect invention of the ivory tower.

[Cihan]

Kip,

I think partly the problem hinges on preferences. It almost seems like a matter of taste what kind of control is worth wanting. Most people would want their subjective experiences to have the right causal connection.

You say you'd want to be plugged into matrix but then also consider the case of Alan the painter. Would you also like to be duped?

You seem to endorse the claim that one's subjective experience is all that matters - what we may call some sort of an extreme internalism. But consider the following agents:

*A manipulated agent vs. an authentic agent
*A lover who's being cheated on vs. a lover whose significant other is faithful
*Someone whom people deceive into thinking that they are well liked vs. someone who is genuinely well-liked.

You could construct stories in such a way that all these agents could have the same subjective experience. But it seems disturbing to be someone who is actually cheated on and yet has the same subjective experience (say through great deception and manipulation) as someone who isn't being cheated on.

Are you willing to bite the bullet and endorse equivalence in all these cases?

To underlie my point: The difference between Brit Kip and Normal Kip could be like the differences between these agents who have the same subjective experiences but different underlying causes for them.

And lastly, I think you misunderstand Justin's point:

"But why should A cares if someone "outplayed" her? It seems that A is the one doing the outplaying."

A is not outplaying anyone. If you are lucky and constantly beat others through bad beats, then that's not outplaying others.

[Sofia Jeppsson]

Cihan, I agree that in the end it becomes a matter of preferences, a matter of taste...

I found myself agreeing with Kip that Brittle Kip isn't any worse off for having his psychological handicap if it's never detected, neither by others nor by himself. But I find it hard to generate any really strong intuitions at all regarding these kinds of cases, since they're so far off from our normal reality.

Sure, people who are decieved exist. But people whose lives are EXACTLY as if they had genuine friends (for example), although their so-called friends are decieving them, and this continues for a whole lifetime... well, that's pretty far off.

And since the idea of these deception cases where from the perspective of the "victim" everything is EXACTLY as it would have been had the situation been genuine are very unrealistic, I'm not sure what kind of conclusions one should draw from them. Suppose I agree that my life would go equally well for me whether my friends are genuine or just completely genuine-seeming deceptors. How is that supposed to influence my attitude to my real life and real people? I don't know... since it's so unrealistic, it seems pretty irrelevant actually. And the same goes for Brittle Kip, I think.

[Sofia Jeppsson]

Kip, you wrote:

"I honestly don't know what you mean by "normative competence," or why you think Brittle Kip doesn't have it. People hardly ever use the term "moral responsibility" and philosophers cannot settle on a definition of moral responsibility so that they can even agree about whether people ever have it. Normative competence seems to be an even more suspect invention of the ivory tower."

But what I think Justin is after could easily be re-stated in laymen's terms. Most people think that one can be excused for bad behaviour on grounds of insanity or mental handicaps. As you've described BK he does have some severe mental handicap or psychological disorder (that's really the only explanation of him going bananas in the nearest possible worlds). So it doesn't seem odd to say that if the people in BK:s world knew all the facts about him (which by hypothesis they don't), they shouldn't blame him for anything he does.

[Justin Coates]

Kip, several things:

First, "normative competence" is a general way of describing the sort of normative capacities agents must have appropriate targets of blame. Fischer and Ravizza analyze it as moderate reasons-responsiveness; Wallace analyzes it as a general capacity for self-control in light of reasons for action; Wolf analyzes it as sanity or a sensitivity to the True and the Good. I could go on...

Second, think about two really common excusing conditions: "she didn't know better" and "he couldn't help it." Why do these excuse? One explanation is that "she didn't know better" excuses because, if true, it shows that she wasn't appropriately receptive to reasons for acting, and "he couldn't help it" excuses because, if true, it shows that he wasn't appropriately reactive to reasons for action. This seems like a good explanation for an often inchoate activity, and if it's correct, then it shows that the folk have some idea of normative competence even if no one except a few hundred professional philosophers ever utter that term. Here I'm simply reiterating Sofia's point.

Third, I think BK doesn't have normative competence because as you stipulate, in all nearby possible worlds he does crazy stuff. This shows that he's not appropriately receptive or responsive to reasons. Normative competence, like all other competences, involves a modal component--that is, unless you can offer a plausible analysis of competence that doesn't appeal to possible worlds. As I've already said, I seriously doubt any such analysis is possible.

Fourth, the fact that people rarely say the words "moral responsibility" or "normative competence" is a red herring. So long as those concepts play a role in governing our behavior, it doesn't really matter how, or even if we can articulate those concepts. Furthermore, I seriously doubt that the philosophical project is one of defining terms, so this sort of objection doesn't strike me as particularly worrisome. I'm not interested in getting into a methodological debate here. Instead, I'm simply pointing out that maybe our divergent methodological tendencies are contributing to some of the disagreement here.

Fifth, just because you win a poker hand doesn't mean you've outplayed your opponent. However, when someone does make a bad play, but gets lucky and wins, you often hear someone else at the table say that they'd rather be lucky than good. That sentiment seems connected with your objection--BK is lucky that he's in circumstances in which he succeeds as an agent, and how could the compatibilist complain about that? Well, it's because we want to be more than lucky, we also want to be good.

Sixth, I don't read Pereboom as saying that free will in the sense required for moral responsibility isn't valuable. He concedes that it is, but rejects the Strawsonian claim that viewing others as unfree should lead us to adopt the objective attitude towards those individuals. So although we lose something valuable by not being free, we don't lose everything. But I understood the original post as proceeding from the idea that the compatibilist didn't have the resources for explaining the value of FW, so I'm not sure how citing G. Strawson or Pereboom is germane to the original post. Of course, I might have been misunderstanding your original argument.

Seventh, although I usually hate it when people do this, I think I'm going to have to withdraw from this discussion. I think the various views have been articulated and defended, and I'm not sure how much more I have to say. But if you have some more to say, I'll read it with interest.

Eighth, Mad Men season 3 started last night, and everyone needs to be watching that. It's compelling--in the sense that doesn't undermine your moral responsibility!

[Kip]

Cihan,

You say you'd want to be plugged into matrix but then also consider the case of Alan the painter. Would you also like to be duped?

I would not object to receiving the exact same sensory perception, by being duped, as if I hadn't been duped. If all else is truly equal, then whether you are duped or not doesn't matter. If nobody else gets hurt, and you don't get hurt, and you can't tell the difference, then it makes no difference. Not any difference that matters.

You could construct stories in such a way that all these agents could have the same subjective experience. But it seems disturbing to be someone who is actually cheated on and yet has the same subjective experience (say through great deception and manipulation) as someone who isn't being cheated on.

Are you willing to bite the bullet and endorse equivalence in all these cases?

I am not saying they are equivalent. They are different, by definition.

I am saying that all you have is your sensory perception. So if someone gives you sensory perception X by duping you, and someone else gives you sensory perception X by being faithful, they've both giving you the same thing. Neither one has been nicer, or meaner, to you. What you receive from them is exactly the same. And you will get nothing more, and nothing less, from them, because all we have is our sensory perceptions.

A is not outplaying anyone. If you are lucky and constantly beat others through bad beats, then that's not outplaying others.

Maybe we're using "outplay" in different senses (perhaps because I'm not so familiar with poker). I thought that the best way to determine who outplays who, is to see who ends up with the biggest pile of cash.

You say that "outplay" refers to something else. In other words, you're talking about a situation in which the agent is repeatedly acting unreasonably, or repeatedly screwing up.

But then the case is no longer analogous to Brittle Kip. Brittle Kip does not act unreasonably in this world. He acts reasonably over and over again.

The analogy originally made sense because Brittle Kip kept winning money. In other words (A) winning money was compared to (B) acting reasonably in this world. That's a fair comparison.

But then you guys started saying that Brittle Kip keeps winning money despite making bad bets, over and over, again. Then the situation is no longer analogous to Brittle Kip.

[Never forget that our knowledge of our own brains right now is still limited enough that, for all we know, we are all Brittle People. We don't seem to mind.]

Sofia:

Why shouldn't they blame him? The only reason would be the mental defect that you cite.

But not all mental defects result in exculpation. For example, colorblindness (generally) won't.

So the question would be: why is lacking thick CFW exculpating but not colorblindness?

From a compatibilist perspective, I don't think lacking thick CFW would be exculpating. This is what the debate would go like in the court room:

Defense: Your honor, Brittle Kip is clearly not guilty. He has a serious mental defect.

Prosecutor: That's strange, your honor, because Mr. Kip appears, to all appearances, to be a perfectly rational, ordinary and intelligent young man. And a rational, ordinary young man who has stolen Ms. Sofia's ice cream cone! The people demand justice!

Defense: But, clearly, your Honor, our neurosurgeons assure us that, if things had been just slightly different, Brittle Kip would not be able to function in society at all. He's just on the cusp of being deranged.

Prosecutor: But he isn't deranged, your honor! Our own psychiatrist has examined Mr. Kip's brain and assures us that, although the brain contains significant abnormalities, none of these abnormalities will ever result in misbehavior in this world! In other words, all of these abnormalities are latent and not used.

Judge: So, counselor, you're saying that, because these abnormalities are latent, and not exercised, that Brittle Kip is, effectively, a rational and intelligent person, as far as this world is concerned.

Prosecutor: Yes, your honor.

Judge: Well, I don't see why he can't be responsible in that case. He shouldn't get off the hook just because he has some abnormality that's unrelated to his theft of Ms. Sofia's ice cream cone.

[Mark Young]

Kip:

You wrote: "Let's distinguish between:

A. The ability to respond reasonably to various situations is valuable, because any of those situations might be part of the real world; and
B. The ability to respond reasonably to situations that will never happen is valuable."

Let's also distinguish between

B1: The ability to respond to things you know will never happen is valuable; and
B2: The ability to respond to things that you don't know will never happen is valuable.

I was trying to point out that, even if B1 is false, B2 is still true, and so B likewise remains true (B1 v B2 -> B). A person who does not know whether their house will burn down values fire insurance -- and rightly so.

Value is not an absolute. If you know that your house is not going to burn down, the value of the fire insurance drops to zero (for you). If you know that it will burn down, the value of the insurance rises to the value of your house and its possessions. (To you, of course. For the insurance company, the directions are reversed.)

When a compatibilist asks for the term "free will" to be understood in a way that results in it being valuable, it goes without saying that the value is to us -- regular human beings living our regular lives. Brittle Kip doesn't seem to fit into this category -- he seems to have an inhuman knowledge of his future.

In your earlier message (August 14th) you asked "why do you care what happens in fake counterfactual worlds? You don't live there." We care about what happens in "fake counterfactual worlds" because those worlds are the tools we use to shed light on this one -- including the future of this one. If, in a nearby fake counterfactual world, Kip is eating a rubber tire, it's because in this real world Kip has a propensity for eating rubber tires. If Kip is a normal human being (rather than a lesser relative of the Omniscient), that propensity to eat rubber is something that -- so far as Kip knows -- could cause him grief in the future. It is valuable to him now to act in ways to change that propensity -- even if the Omniscient knows that that propensity will never manifest.

It is Brittle Kip's knowledge of his future -- knowing that his propensity to eat rubber will never manifest -- that makes counterfactual reasoning useless for him. He knows not just that those worlds are fake, but that the changes made to create those worlds are irrelevant to his future -- he never will run into a cat under sufficiently similar circumstances.

In summary, BK's situation is too far removed from ours to shed any light on whether we should value CFW, or care about counterfactual worlds.

[Kip]

Justin:

Let me address your points in turn.

1. My confusion about "normative competence" is not so much about your use of the term. It's clearly related to what we regularly talk about at the Garden. It's just not clear to me how normative competence is supposed to be distinct from moral responsibility (if it is).

For example, Fischer distinguishes between moral responsibility and blameworthiness. But he also agrees that the distinction is pushing it (regarding how far we can split hairs about these terms). Is that the distinction you are trying to capture?

Further, it's not clear what an "appropriate" target of blame would be. Appropriate according to whom? Social custom? The word "appropriate" makes me think of Emily Post's Etiquette.

Perhaps the problem, for me, is that words like appropriate seem to introduce an ethical question into a debate that, as far as I can, need only be about the descriptive.

2 and 3. Third, I think BK doesn't have normative competence because as you stipulate, in all nearby possible worlds he does crazy stuff. This shows that he's not appropriately receptive or responsive to reasons. Normative competence, like all other competences, involves a modal component--that is, unless you can offer a plausible analysis of competence that doesn't appeal to possible worlds. As I've already said, I seriously doubt any such analysis is possible.

How about this proposal: normative competence is the ability to respond reasonably to the reasons that are actually presented to you---reasons that are never presented to you be damned. In other words, all you need is thin CFW (as I originally defined it).

[I only say that with my compatibilist hat on, which doesn't fit me well. As I explained in point 3 of my previous comment, I think that thick CFW is no better than thin CFW, so the compatibilist is faced with a dilemma: (1) accept that normal Kip doesn't have free will or (2) accept that Brittle Kip does have free will. I'm generally an anti-realist about free will, so I'll be happy if you accept (1). But I'll also be happy if you accept (2).]

4. I agree that our different methodological inclinations helps explain our difference. But I find the lack of use of the terms "moral responsibility," etc., to be much more telling than you do. It's the canary in the coal mine.

5. Well, it's because we want to be more than lucky, we also want to be good.

You should only want to be good because you believe that you won't always be lucky. Brittle Kip is not like that. Brittle Kip is always lucky. He, therefore, has no need for the skills that you otherwise desire. They add nothing to his ability to win at poker. It's irrational for him to desire those skills, in a way that is not irrational for you to desire them.

(You have to give me at least some credit for playing with your poker analogy for as long as I have.)

6. I never said that Pereboom or Strawson said that free will, or moral responsibility, isn't valuable. I said, "Maybe it would be better if we're all not morally responsible like Galen Strawson and Derk Pereboom say." That is, I was citing them for the proposition that we might not have fw/mr; I wasn't citing them for the proposition that fw/mr isn't valuable.

If anything, I agree that their inclinations are in the opposite direction. Pereboom's book is named "Living Without Free Will" for goodness' sake!

Still, if I wanted to read Pereboom as saying that fw/mr isn't valuable, I could find some support. For example, he directly challenges the assumption that fw is valuable in the intro to LWFW, as I discussed in the original post.

[Kip]

Mark,

What you are saying boils down to this:

If you could guarantee that we compatibilists live in a "lucky" world where our latent dysfunctions never manifest, you could limit us to having thin CFW, and we would agree that we have nothing to complain about. We would further state that we still have free will, and are morally responsible, for our actions.

This is just an empirical question. I think most compatibilists are so wedded to a counterfactual understanding of free will, and find the idea of thin CFW so disturbing, that they would not agree with the above (italicized). I don't think they would agree, even if you and Sofia would. But I could be wrong about that.

[Mark Young]

Kip:

It is an empirical question as to whether most compatibilists would agree with your re-statement of my position. Unfortunately, I don't know if I agree with your restatement of my position.

In particular, I'm not sure that Thin CFW makes sense. You say that BK does rational deliberation -- presumably in the course of choosing an action. You also say that BK is irrational in all the nearby possible worlds.

So here's BK, looking at a full house and a very confident-looking Justin Coates, who's just raised by $1,000 (the house limit). How does BK go about deciding what to do? Well, he has three options: call, raise, or fold. He needs to think about what will happen under each of those three options. Sadly for him, he can only choose one of those options -- and so the the other two are in counterfactual worlds, where he is completely irrational. The results of two of those actions would be that he'll end up eating rubber. Clearly he needs to choose the option that retains his sanity, whichever one that is.

How does he know which one it is? It can't be by rational deliberation -- since he's never been irrational before, he has no knowledge of what actions will make him irrational. His only basis for deciding would have to be that he was in precisely this position before, and the action he chose then left him rational (lucky guy!). In that case he could choose the same action again, and be safe (but utterly predictable -- a very bad thing to be when playing poker).

But whenever he's in a new situation he has to trust to his luck to get him the correct action. That's not rational.

Even if your guarantee that he's in a "lucky world" were a rational basis for decision making (which it is not!), he would get no rational guidance from it. He would know that whatever he decided would leave him sane, but would still have no rational basis to choose between them. He would be reduced to assuming (contrary to stipulated fact) that all three actions would leave him sane and figuring the outcome of the actions under that (counterfactual) assumption. Well, in two of the cases, he'd get the outcome wrong -- his conclusions about what the results would be are not what the outcome really would be. He would be working with counterfactual worlds that are "far away" (by your metric) because they look closer. And why do they look closer? It's because they are much more like the world he is familiar with -- the world where he has always maintained his sanity, regardless of the action he did choose.

So we're left with a mismatch between how far you say the worlds are apart, and how far apart they seem to be. Since you cannot provide a reasonable definition of your metric (please correct me if I am wrong), I'm thinking that the "magnitude of the apparent change" metric is the proper one to use when doing rational deliberation -- and so, from the point of view of rational deliberation, the nearest possible worlds are all worlds where we retain our sanity, in spite of your stipulation to the contrary.

Since thin CFW makes no sense to me, I really can't say what I'd agree follows from assuming it true. (Other than that anyone in that world would have no way of knowing that they were.)

[Kip]

Mark,

I hope my restatement of your position, as described in your previous comments, wasn't too far from the mark. I strive to avoid attacking straw men. Your most recent comment raises significantly different objections than you had previously focused on, so my previous restatement doesn't really apply to your new objections.

I don't see a concrete objection in your most recent comment. The objection seems to be something like "the idea of Brittle Kip is incoherent." You elaborate on that by describing BK playing poker. But we seem to have very different understandings about what BK playing poker would be like.

Here's how I image BK playing poker: it's exactly like when I play poker. So, for example, I can call, raise or fold.

Let's assume that, if BK calls or raises, he will go chew tire rubber. But if he folds, he will continue to act reasonably.

Because BK lives in a lucky universe, he folds. And the rest of his life continues in that ordinary, uneventful way, for the rest of his life, always choosing away from the worlds where he chews tire rubber.

There doesn't seem to be anything incoherent about that scenario.

Even if your guarantee that he's in a "lucky world" were a rational basis for decision making (which it is not!)

Perhaps your real objection is that Brittle Kip couldn't be in a lucky world (as suggested by the above quote). I don't see why not.

BK is an extremely strange, fictitious character conjured up in a philosophical thought experiment to prove a point, and nothing more. As creator of the thought experiment, I can modify the conditions of the scenario as I please (assuming I don't violate any laws of logic).

I can do this because I'm challenging the following statement: a necessary condition of valuable compatibilist free will is cross-worlds reasonableness/rationality. Because the compatibilist wants to say that this condition is necessary, all I need to disprove it is one, single counter-example in which the condition doesn't obtain, but the agent still has all of the compatibilist free will will worth wanting (it would be irrational to desire more). That's what BK is intended to do. He's that one, strange counterexample showing that cross-world reasonableness is not a necessary condition of compatibilist free will worth wanting.

If that's my aim, and all I need is one counter-example, then my counterexample can be in a Lucky World. He can be in any world. We can suppose that a blind Kip-making-machine makes Kips in all possible worlds, including the Lucky World. And in that case, we still get our counter-example, produced through a sort of natural selection. So it's not unreasonable to suppose that BK lives in a Lucky World.

Even if your guarantee that he's in a "lucky world" were a rational basis for decision making (which it is not!), he would get no rational guidance from it.

He doesn't need rational guidance. He's not even thinking about which worlds will result in irrationality. He might not even know he's brittle. And even if he did know, he's might not be concerned about it. For example, if he has no idea which options cause irrationality, it's a waste of time to worry about it. He can't figure it out, and he probably figured out that he can't figure it out a long time ago. He probably also figured out that he never chooses the irrational option a long time ago, and stopped worrying about it. Or perhaps a reliable Oracle told him that he always chooses right. Or perhaps he doesn't know he's brittle.

[Note, earlier you said that BK would have to know that he's brittle and lives in a lucky universe for my thought experiment to work. I don't think that is true. BK could know nothing about his condition. We, as observers, could know about his condition. And in that case, I believe that (C1) we, as observers, should say that BK has nothing to complain about relative to normal Kip (C2) most compatibilists would disagree with me about that. So I think the thought experiment doesn't require BK to know about his sitatuation.]

Perhaps that helps clarify the idea. I know it's very strange.

[Mark Smeltzer]

Kip,

I don't see what your analogy has to do with undermining the significance of counterfactuals in (compatibilist) discussions of freedom and responsibility.

When I wrote my previous post, "Desert and Responsibility", I said this at the beginning of the second section:

While it is easy to imagine a whole dissertation being written on the tension between ultimate [(D)] and warranted desert [(D*)], I want to move past that tension for now.

In the terminology of that post, your scenario is exactly the type I had in mind when I made that statement. Simply put, there are often times when (D*) and (D) will point to different conclusions about what S deserves and (more interestingly) both are right.

The chasm between (D) and (D*) is the difference between how we ought to treat S based upon the warranted beliefs we have about S and how S deserves to be treated based on the kind of person that S actually is. I see no reason to think that we couldn't have (potentially false) warranted beliefs about what Brittle Kip would do in this world based on our observations of him. And yet, by stipulation, in this world we aren't getting the full picture and according to (D) our judgments about Brittle Kip may be incorrect.

However, I think there is a far simpler scenario for you to consider. It is one of the scenarios that first got me thinking about the lines of (D) and (D*):

• Suppose a man happens to have been born on a deserted island where his parents had crash landed.
• His parents died while he was a teenager leaving him alone on the island.
• Although he lacks human companionship, he enjoys the company of dogs that were on his parents' boat that have produced several healthy litters of puppies.
• He has plenty of fruits and vegetables to eat.
• He catches fish from the sea.
• He would describe himself as a happy person.
• If you were to observe him, you would agree that he is genuinely happy (though you may feel sympathy and pity for him due to his isolation).
However, the fact is that he is a devilish, murderous fiend and (as fate would have it) just happens to be stuck on a deserted island.
• In fact, he is the kind of person that if he ever got off the island he would go on a heinous killing spree: raping, torturing, maiming and ultimately murdering vulnerable women.

From a (D*) perspective, this S deserves (something like) praise for caring for the puppies and coming out on top and sympathy and pity for having been alone for so long, and yet from a (D) perspective S deserves to be despited and hated -- though he has never actually killed, he would do so, and would do so in inexcusable ways.

That's the rub between (D) and (D*), and any account of freedom and responsibility that seeks to describe the full picture will need to address that distinction.

[Sofia Jeppsson]

Although I don't think the Brittle Kip scenario as originally described is incoherent, I do think it's incoherent to discuss what BK ought to care about, which several people in this thread does. So let's separate various ways of regarding BK, or you could say various epistemic positions one could be in regarding BK.

1. THE NOBODY KNOWS SCENARIO:
I live in this world. I regard BK, living in a completely different possible world from myself, through the lense of my imagination. I'm fully aware of all facts about BK (including the fact that he'll be lucky throughout his entire life), but BK himself is not. BK and all other people in his world thinks he's just like everyone else, since his mental defect has never been detected. And BK is happy.
Well, in this scenario it strikes me as unintuitive to say that BK:s life is worse due to a fact that he doesn't know, and that has no consequences for his life. Others have different intuitions, and I don't think there's a way to settle whose intuitions are "better".

2. THE SCIENTISTS DISCOVERED ABNORMALITIES SCENARIO:
In BK:s world, some scientists examine his brain, and find strange anomalies in the parts of his frontal lobes responsible for rational thinking. (This is a little like Kip's court room scenario, in his response to me earlier.) Still, he's always appeared normal. Now the scientists would, if they're anything like scientists in our own world, consider two possibilities. Either that what they previously believed about neurologics was wrong, and those parts of the frontal lobes aren't that crucial for rational thinking after all. Or they'd say that although BK has SEEMED rational so far, appearances are decieving in this case. Any second now he might go completely bananas, so he'd better spend his life in an institution from now on.
They can't possibly say that it doesn't matter since he lives in a lucky world, since that conclusion wouldn't just require certain intuitions on their part but also omniscence.

3. THE OMNISCENT ALLMIGHTY GOD TALKS TO BK SCENARIO:
So let's introduce an omniscent allmighty God into the picture. Suppose to start with that nobody ever sees any reason to doubt the explicit word of God, and suppose next that God makes an appearance before these scientists and BK. God explains that the abnormalities in BK:s brain really does affect his rationality. He's so irrational that was he ever to encounter a blackbird singing from a birch-tree at exactly 10:15 am, or was he ever to see a blond five-year-old playing with a Barbie doll on a Tuesday when there's a full moon, or if he was ever to encounter one of several million other situations of that everyday and mundane kind, he'd go bananas in response. But now God hasn't only created BK to be this way, he's also (cause God moves in mysterious ways) created the entire universe to ensure that BK will never encounter any of these situations. BK will go through his entire life seeming completely normal, and feeling completely normal himself.
Well, these news means it's unnecessary to lock him up in an institution. How he should be regarded by others in general still seems to be a question one can argue about a lot.
And BK himself? How should BK regard these news? It seems to me that this question doesn't even MAKE SENSE. If, among the nearest possible worlds, there are those where BK thinks completely random and weird thoughts, then the fact that he in his actual world happens to think thoughts that fit the context and comes in the right order is just a matter of coincidence. And in that case, this strange creature can't be said to be rational at all. But if one discusses how someone should regard something, that presupposes that the object of the discussion is rational.
If asked how he took these news, BK would of course say something that seems to make sense in the discussion, since by hypothesis he always does. He might say either that these are sad news for him, since he valued being an autonomous and rational agent, or he might say that these news don't matter to him since they don't make any difference to how his life goes. But since he wouldn't be uttering sentences that express some judgement he came to by weighing reasons, rather he just happened to have certain thoughts popping up in his head and then he happened to say them aloud, it seems pretty pointless to discuss whether this is what he ought to have said or not.

Finally, we have the 4. THE OMNISCENT ALLMIGHTY GOD TALKS TO PERFECTLY NORMAL KIP SCENARIO:
Suppose God instead approached a Perfectly Normal Kip and told him that he will MAKE him into Brittle Kip. Afterwards, BK won't remember anything about the encounter with God, nor that he was ever different. Now, should PNK be bothered by these news?
Despite my reaction to scenario 1, I think I'd be hugely bothered if I was in PNK:s shoes. But I don't think my reactions to 1 and 4 are necessarily inconsistent, since they come from completely different perspectives. It doesn't seem odd to value rationality. And still, it can only be valued as long as one to some extent have it, since otherwise one cannot even comprehend it. While I still have my rational thinking I could value it, once it's gone it still seems weird in my eyes to say that some fact that I'm not even aware of or affect my feelings in any way could make my life go worse FOR ME.

Be that as it may with my intuitions regarding this last point... I still think it's pretty important to point out that there's something wrong with considering how BK should regard the fact that he's BK.

[Eddy Nahmias]

I haven't been following all the discussion in these posts, but I happened to notice Kip's claim: Never forget that our knowledge of our own brains right now is still limited enough that, for all we know, we are all Brittle People. We don't seem to mind.

This is not right. You might say that science works precisely by jittering conditions to see what happens in various possibilities. In neuroscience and psychology, one of the goals of doing so is to determine what capacities humans have, in part because of the capacities of our brains. And the discovery of the capacities of a system is precisely the discovery of what that system would do in various conditions (some of which may never be actualized, certainly not by individual tokens of that system).

It's not clear how to make sense of the idea both that we are Brittle People and nothing about neuroscience has shown (or could show?) that. Rather, the sciences of the mind are showing how it is that we have the capacities to be reasons-responsive and also the (often significant) limitations to those capacities.

Finally, it's even crazier to say "we don't seem to mind." Surely, we believe of ourselves that we would never do certain crazy things (e.g., shoot our grandmother if she said the words "death panel") even though we also believe that we will likely never actually be in such a situation. And surely, we believe that *if* we would (or might) do such crazy things, we would be worse off. Personally, if I believed I was so "brittle" I'd go crazy (except I'd already be crazy), and I'd certainly never discuss health care with grandma.

[Kip]

Eddy,

I admire the confidence/decisiveness with which you contradict me. But I disagree.

I am not a neurosurgeon. But I know that:

1. There is still a huge amount that we don't know about the brain.
2. "Software" can be encoded in the brain at resolutions small enough to not yet be detected by us, if the software never manifests itself to us in behavior.
3. All of the data we have about brain responses are based on observations of how the brain works in this world. They don't necessarily tell us anything about how the brain would work in different worlds.

So, I don't think the absoluteness of your statement is justified.

Think of it this way: suppose a trickster God exists. Suppose that, as a trick, he wanted to create us all as Brittle People, without us knowing it. So the trickster God configured our brains in very subtle, hidden ways, so that we are Brittle. The brittleness is not detectable except by very powerful microscopes looking in the just the right places, and we do not have those microscopes or know where to look.

Are you honestly telling me, not just that the above situation is unlikely (which it certainly is), but that it is impossible?

I don't think it's impossible. Therefore I think your statement that my previous comment was incorrect is too quick.

[Mark Young]

Kip:

You wrote "I hope my restatement of your position, as described in your previous comments, wasn't too far from the mark."

My response to it was an attempt to show why I found it problematic. If I were editing it directly, I'd change it to:

A person living in a "lucky" world where their latent dysfunctions never manifest, who knows that they are living in such a world, would value FW and rationality less than we do -- perhaps even not at all. That would be right and proper for them, but would tell us nothing about whether we should value FW and rationality.

If Kip knows his house is not going to burn down, he does not value fire insurance. If Brittle Kip knows that none of his actions are going to make him worse off, he doesn't care about choosing his actions.

Given that, I don't think your example does what you want it to -- namely, show a counterexample to the claim that "a necessary condition of valuable compatibilist free will is cross-worlds reasonableness/rationality." CFW is not valuable for BK; that he has no cross-worlds rationality is irrelevant.

My earlier posts were pointing to the fact that if BK did not know that he was in a "lucky" world, then he would (or should) care about CFW -- it would have value to him.

My post last night was pointing out that, if he did not know that he was in such a world, then the counterfactual worlds that are nearer to him are the ones where he retains his rationality. How far away a possible world is depends on what you know. In your example we know more than non-omniscient BK does, so when we measure the distances we come up with different answers than he does. Since we are omniscient (or close enuf) w.r.t. BK's world, it makes sense to say that the distances for us are the "real" distances. But as Mark Smeltzer points out, in interesting ways we are both right about how far away these possible worlds are. We stand in different places, so the distances are different.

And it is that last point that prevents non-omniscient BK from playing the role omniscient BK was (also) unable to fill above. NOBK has cross-worlds rationality: the possible worlds close to him (for him) are worlds where he is rational.

[Sofia Jeppsson]

"I am not a neurosurgeon. But I know that:

1. There is still a huge amount that we don't know about the brain."

But it's not neurologics per se that says we can't all be brittle people. It's the whole enterprise of empirical science.

Either you understand laws of nature as just patterns that we can find in nature. If that is so, the idea of a pattern that is 'really' just a huge mass of coincidences is self-contradictory, which means that Brittle Kip is an incoherent idea to start with.
If we understand laws of nature as something more than that, it becomes logically possible to have a world where what looks like laws of nature at play really are just a massive amount of coincidences. For example, a world without gravity, but where everything constantly falls downwards rather than upwards, as a matter of pure chance. And if so, Brittle Kip also becomes a logical possibility.

But the whole enterprise of empirical science presupposes that we don't seriously consider the possibility that the constant correlations and patterns we see might mean NOTHING and just be a massive amount of coincidences. All gathering of statistics, all experimentation, presupposes that this is not the case.
So taking serious the idea that we might all be brittle people, is tantamount to taking a complete scientific scepticism seriously. Which one is of course free to do, I just want to point out that this is what's at stake here.

[Sean]

I haven't followed all of the discussion, but I'd like to throw in my two cents.

I think a compatibilist can keep the two tenets and still admit brittle Kip has nothing to complain about. Compatibilists can say:

'Brittle Kip's free will is severely defective (from tenet 2), and free will is valuable (from tenet 1). But Brittle Kip can't complain because he has something just as valuable as free will: the good luck to always act as if he had free will. Since both brittleness and free will give the same outcome, they are both equally valuable.'

I.e. If we want our lives to go well, we can either be competent decision makers, or brittle, incompetent decision makers with the good luck never to have our incompetency discovered. Both constant good luck and competence give the same good outcome.

However, its better for us to strive for competence, since luck is beyond our control. And an important part of judging and improving our competence (decision making abilities, rationality, mediating between higher/lower desires) is thinking about what we would do in other possible worlds.

[Kip]

It's amazing how many different philosophical ideas my thought experiment has touched on:

1. Mark Y. points out the knowledge condition on value;
2. Cihan points out (similarly) internalism/externalism about value (and Nozick's machine);
3. Mark S. points out moral luck (and reminds me of Locke's locked door);
4. Sofia reminds me of the Humean problem of induction.

All from one thought experiment!

Mark Y: My earlier posts were pointing to the fact that if BK did not know that he was in a "lucky" world, then he would (or should) care about CFW -- it would have value to him.

That statement implies that, if BK did know he was in a lucky world, he shouldn't care. Would you agree with that? Note that Eddy clearly doesn't.

[Mark Young]

Kip:

You wrote "That statement implies that, if BK did know he was in a lucky world, he shouldn't care. Would you agree with that?"

I agree that, if he knew he was in such a world, caring would be optional.

You added "Note that Eddy clearly doesn't."

Either you gave the wrong link, or one of us is mistinterpreting him. I took him to be saying that if he knew he was mentally defective (the way BK is) that he'd care about it. I don't think he was assuming that he knew that he was in a "lucky world". As I asked in my first response, how could he know?

[Kip]

Mark:

I agree that, if he knew he was in such a world, caring would be optional.

Then I have the single counter example that I needed. I'm trying to attack the following claim:

1. It's a necessary condition of the compatibilist free will worth wanting, that we have cross-world rationality/reasonableness.

You agree it isn't worth wanting, if BK lives in a Lucky World and knows that he lives in a Lucky World. So it's not a necessary condition.

Either you gave the wrong link, or one of us is mistinterpreting him. I took him to be saying that if he knew he was mentally defective (the way BK is) that he'd care about it.

My remark was based on comments from Eddy such as: "Personally, if I believed I was so 'brittle' I'd go crazy (except I'd already be crazy), and I'd certainly never discuss health care with grandma."

He doesn't say "I would go crazy unless I knew that I lived in a Lucky World." I'm pretty sure Eddy thinks that BK has something to complain about, even if he lives in a Lucky World. But I could be wrong about that (Eddy?).

As I asked in my first response, how could he know?

There are a variety of different ways.

1. Perhaps a reliable oracle tells him.
2. Perhaps a godlike being (or the creator of his universe) tells him, as Sofia and others discuss.
3. Perhaps he figures it out from induction, the same way we assume that the law of gravity will still apply tomorrow.

Those are just off the top of my head.

[Mark Young]

Kip wrote: "Then I have the single counter example that I needed."

I explained in my August 18 post why I think that's wrong. In summary: either BK has cross-world rationality (in spite of your hypothetical), or CFW is of no value to him. Either way, he's not the counter-example you need. I don't think you addressed those points.

[Cihan]

Rethinking some of this, I have the following question: How is Kip's scenario different from a glorified Frankfurt case?

Suppose that there is a very complicated and able neuroscientist that's going to intervene if BK does anything different from what he does actually (i.e. what the intervener wants BK to do and what BK wants to do for all of his life coincide) and make BK irrational in all sorts of ways.

Surely, the compatibilist doesn't want to say that BK is lacking free will that's worth wanting?

Isn't this the scenario we are discussing?

I think partly there is some relevance to Lewis' fink cases - i.e. a sorcerer intervening to a salt in water just before it dissolves.

Thinking it over, actually Fischer's semi-compatibilism wants you to "hold fixed" the mechanism that produced the actual decision and then apply a counterfactual test and this is where the first BK and the BK with Frankfurtian intervener have a crucial difference: there is no "mechanism" that can be held fixed for the first BK - it's all a matter of luck.

So maybe this sort of thought experiment doesn't work against more sophisticated brands of compatibilism.

[Kip]

Mark,

I explained in my August 18 post why I think that's wrong. In summary: either BK has cross-world rationality (in spite of your hypothetical), or CFW is of no value to him. Either way, he's not the counter-example you need. I don't think you addressed those points.

The above is too quick for me to understand. For example, you say that "BK has cross-world rationality (in spite of your hypothetical)." But BK can't have cross-world rationality, by definition. If he had cross-world rationality, he wouldn't be Brittle (wouldn't be BK).

So already I'm confused.

Then you say "or CFW is of no value to him." But that doesn't contradict me. I'm arguing that thick CFW would not be valuable to him. I'm arguing against the proposition that thick CFW is valuable (more than thin CFW). So I'm not afraid of that prong.

[Kip]

Cihan,

I agree. That's why I drew the analogy with Frankfurt's cases in the original post. Neal T. wasn't impressed by the comparison, but maybe there is more to the analogy than he suggests.

For example, perhaps BK is just a victim of a Frankfurt scenario in which the prevention is internally built into the agent, rather than externally called upon (hypothetically). In other words, let's contrast:

Original Frankfurt: Black lies in wait to initiate a fail safe mechanism if the signal happens. The signal never happens, so Black never initiates the failsafe mechanism.

BK: Black, like Diana in the Zygote Argument, designs BK from scratch so that he is Brittle. Any other sensory input than the input he actually receives makes him irrational. The other input never happens, so he never acts irrationally.

Essentially, the fail safe mechanism is woven into BK's being.

The moral of the story is, I think, the same: the counterfactuals are irrelevant.

Note also that, in the case of BK, there does not appear to be even a "flicker of freedom." This is consistent with those who argue that subjects in Frankfurt cases need not always have such flickers for the the thought experiment to work. I believe Pereboom defends that position, but I don't know the Frankfurt case literature very well (help?).

[Mark Smeltzer]

Kip,

Fischer makes the distinction between being morally responsible and being held morally responsible. Being morally responsible requires holding the mechanism fixed, but being held morally responsible requires that the agent could be subjected to a hypothetical interview and that our intuitions about the agent survive that interview: what Fischer is getting at here is that being held responsible presupposes warranted counterfactual beliefs about the agent.

If our picture of the agent falls apart during that hypothetical interview (as it surely would with Brittle Kip), then the agent cannot be held responsible. That does not mean that the agent is not responsible.

With this distinction Fischer is attempting to honor a corollary of the distinction between warranted desert and ultimate desert (which I discussed in my post "Desert and Responsibility").

In other words, I don't think your case presents any problems for Fisher's account. If that is the case, I do not think your case is relevant to debunking Frankfurt cases (since Fischer's account is modeled heavily after Frankfurt cases).

[Trevor Pisciotta]

Apologies for entering the conversation so late, and thanks for such an interesting debate.

Kip, I'm not sure I agree with the way you're using possible worlds talk.

For example, you say:
"All of the data we have about brain responses are based on observations of how the brain works in this world. They don't necessarily tell us anything about how the brain would work in different worlds."

This is true, in a sense. Other worlds are independent from ours, and the goings on there, don't depend on the goings on here. But facts about our brains in this world do determine the relevant closeness relation; i.e. such facts determine the set of worlds that is relevantly close to ours. And what goes on at these worlds is interesting and informative precisely because it tells us something about our brains, here in this world. Namely properties (albeit counterfactual ones) that they possess/instantiate.

Unless I'm mistaken, you're arguing that what happens other worlds isn't relevant to our reason-responsiveness. But reason-responsiveness just seems to be a counterfactual property; it describes how we would respond, were we faced with certain reasons. And surely counterfactual properties are legitimate. Its a little like saying when we ask whether a glass is brittle, we're not asking about how it would behave, if hit (i.e. in another possible world), we're just interested in whether the glass breaks in this world. But even if the glass never breaks in this world, it still is brittle, and it seems reasonable to care that it has this property.

So I'd be interested to know why (or whether) you think we can maintain the legitimacy of counterfactual properties more broadly, but not with respect to rationality?

[Sofia Jeppsson]

Trevor, thanks for the glass analogy. That's what I was trying to get at, by saying that if we take seriously the possibility that we're all brittle people, it amounts to total scientific scepticism.
Surely we'd regard it a scientific fact that the glass in my cupboard is more fragile than the rock in my front yard, and that it would be easier to break the glass in my cupboard than the rock in my front yard. It's a scientific fact, we'd think, that because of the different molecular structures in the rock and the glass, the glass is more fragile. But saying that the glass is more fragile than the rock IMPLIES that in some of the closest possible worlds where I happen to drop the glass on the floor, it breaks. While the rock doesn't break in any of the nearest possible worlds where it is subject to equal force.

Statements about what qualities something have here in THIS world, can be TRANSLATED into statements about what would happen in the nearest possible worlds. Therefore, saying that science only tells us something about what things are like in THIS world, but nothing about what goes on in the nearest possible worlds, is contradictory.

One can of course think that science doesn't tell us anything period, and be a total sceptic. But thinking that it gives us valuable information on this world but nothing on the nearest possible worlds is contradictory.

[Mark Young]

Kip:

"The above is too quick for me to understand. For example, you say that "BK has cross-world rationality (in spite of your hypothetical)." But BK can't have cross-world rationality, by definition. If he had cross-world rationality, he wouldn't be Brittle (wouldn't be BK)."

I explained in my Aug 18 comment why unknowingly brittle Kip has cross-world rationality in the way that matters to him.

"Then you say "or CFW is of no value to him." But that doesn't contradict me. I'm arguing that thick CFW would not be valuable to him. I'm arguing against the proposition that thick CFW is valuable (more than thin CFW). So I'm not afraid of that prong."

I seem to have misunderstood you, then. I thot you were arguing that thick CFW is not valuable at all. (You originally asked whether it was "really worth wanting".)

I agree that you have presented a possible world with a person who has no value for free will. (Note: not just thick CFW -- no free will is of any use to knowingly brittle Kip.) But I think that possible world is too far removed from our own to make any difference in whether thick free will is valuable to people like us.