Blog Coordinator

« Book Announcement | Main | Workshop Report »



Feed You can follow this conversation by subscribing to the comment feed for this post.

It might also be because in Ernie's case (but not deterministic Bernie's), your venom and vengeance have another target, Diana, who is a more apt target since she is the ultimate causal source of Ernie's nasty actions towards you. She had the power to prevent the harm to you in a way Ernie did not (and in a way the distant past sufficient for Bernie's actions did not). At least that's the most natural way to read the case, assuming she cares more about Ernie's doing A than preserving his compatibilist capacities for control over A.

It might also be that your soft heart towards Ernie is driven in part by your sympathy for him. We feel bad for people who are systematically manipulated. Of course, real world manipulation typically changes people from who they 'really were' or bypasses their compatibilist control, etc., in a way that the Ernie case tries to strip out, but our intuitive responses to the case (perhaps based on our emotional responses) may be less than fully sensitive to the differences between this creation case and real world manipulation.

One might say that both of these responses are driven by misreading the case, failing to accept its stipulations, or introducing information that should not be included (though its not misreading the case to say that Diana has power over whether Ernie harms you). But just as Pereboom and Mele ask us to consult our intuitions about these cases, you are consulting your own intuitive and emotional reactions, and asking us to do the same. It'd be surprising (to me) if we could do so in a way that strips the case from its real-world context or if we could have much certainty about what specific features are or are not influencing our reactions.

Of course, we could try to do controlled experiments to test what's influencing our reactions (as you know, Robyn Waller's doing some there at FSU). To start, I'd suggest that my (our?) venom and vengeance towards Ernie increases significantly if we are told (preferably without order effect of having the standard case presented first as it has been here) that Diana aimed to create Ernie to do something nice (B) towards you but she made a little mistake in the lab such that she deterministically created Ernie such that he does A (with compatibilist capacities intact).

Thanks for the interesting post, Randy! (And congrats on the book's birth!)

Eddie, at some points it seems that you are taking the "awareness of Ernie's backstory distorts our response" view, which might support rejecting premise (1). At other points it seems you are rejecting premise (2). Some of both?

How do you think you would respond to Ernie's mistreatment of you, were you vividly aware of the backstory? Do you think your response would be distorted by that awareness?

By the way, if Diana is in the world, then she, too, is subject to determinism. So I don't see that she has a power to prevent harm to me in a way that Ernie doesn't.

Hi Randy and Eddy:

I find the issue fascinating, and I don't find much to passionately disagree with. You both make great and smart points. I just want to add

1. The Ernie example works, to the extent that it does, because it reverses a moral illusion that humans naturally suffer from: self-creation. Whether we think we're entirely self-created, or merely partly self-created, we tend to think of ourselves as more self-created than, in fact, we are. And we see this when we realize that it's actually possible, theoretically, for Diana to have total control over Ernie's like simply by manipulating initial conditions, without undercutting his compatibilist capacities. On our natural, intuitive, pre-scientific and pre-philosophical view, Diana's level of control would simply be impossible, and the Ernie story shows that it actually is possible. Thus, the Ernie example works like a lens that, instead of distorting, actually clarifies, our self-understanding. It clarifies the limits of our freedom, options, and capabilities. There are a variety of illusions in the social sciences that are candidates here: the fundamental attribution error, the illusion of control, reactance, etc.

2. Eddy makes a great point that the Ernie example, and similar examples, raise questions about personal identify. Suppose that Diana alters the initial conditions so that Ernie kills Paul at T1. Or alters the initial conditions so that Ernie eats a candy bar at T1. Or alters the initial conditions so that Ernie gives a dollar to charity at T1. In all of these cases, Diana preserves the general compatibilist powers of "Ernie." But, even though she keeps constant those powers across all instances of "Ernie," maybe she doesn't keep constant the personal identity of this "person" - maybe there is not one "Ernie," but several: Ernie, Ernie*, Ernie**, etc. Maybe, in a meaningful way at the metaphysical level, these are simply different people. Then the example no longer works as an argument for incompatibilism. This is, in my view, the strongest argument for compatibilism, and one argument that doesn't get nearly enough attention in my view.

One problem with this argument, I think, is that it makes personal identity incredibly sensitive to inter-world differences. In fact, arguably, any inter-world difference will prevent Ernie and Ernie* from being the same person. If there is some degree of tolerance for inter-world difference, the compatibilist hasn't explained what it is, or why it differs from intra-world differences. Notably, we tolerate HUGE amounts of intra-world differences without sacrificing personal identity. Ernie can be hit by a bus, paralyzed, bullied, brainwashed by Nazis, etc., and we still call him Ernie (not Ernie*) and generally hold him legally responsible for his actions. If we tolerate so many intra-world differences, it's not clear why we should tolerate basically no inter-world differences for the same reasons.

3. Randy says, based on the Ernie example, that moral responsibility cannot be more than your watered-down reaction after hearing Ernie's back story. Why not say, instead, that moral responsibility required more, and that Ernie's back story made you realize that moral responsibility is not possible? That's what Pereboom says, and that's what Galen Strawson says with his skeptic hat on, and that's generally what Neil Levy says about constitutive luck, and that's the view I've always been most attracted to. You seem to reject it without much consideration?

"Ernie lives in a deterministic world" but "I've asked several folks whether…they think that knowledge…would alter how they do…respond". They say "no", but their reasons seem to me to miss the point, which is that nothing could alter their response.

Why when determinism is posited is it thought to apply to evil doers like Ernie but not necessarily to nice folks like us?

If determinism held, none of our disagreements about moral responsibility or for example penal reform would make any difference.

Jim, of course you're correct that if determinism is true, it covers everything in the world, including folks naughty and those nice.

However, what do you mean when you talk of something "making a difference"? Suppose that x happens and causes y. Suppose that determinism is true. Still, it might be that if x hadn't happened, y wouldn't have happened. x might still make a difference in this respect. Determinism doesn't imply otherwise.

Perhaps you mean to say that if determinism is true and x happens, then x couldn't have failed to happen. For it was determined that x happen.

Consider: if the laws could have been different in certain ways, then x could have failed to happen. And if the entire past (prior to the occurrence of x) could have been different in certain ways, then x could have failed to happen.

Determinism itself doesn't say anything about whether laws of nature could have been different, or whether the state of the world at every time could have been different. It states a conditional necesssity: Necessarily, if the laws are the actual laws and the world-state at some time t is the actual world's state at t, then the world-state at any other time t' is the same as the actual world's state at t'.

Perhaps determinism (with some further truths) implies that agents are never able to act otherwise than how they actually act. However, the claim that this is so is disputed.


This is a good angle from which to approach these cases. I'd like to add another (while keeping yours): what are the valid ethical reasons for accepting and retaining punitive emotions and practices? I think our ethical viewpoints will do a lot more work on the Ernie/Bernie cases than our metaphysics.

Let me focus on practices, and leave emotions for later. I like a view on which punishment is justified *analogously to* (*not* the same as "based upon") self-defense. That is, it respects each individual, protects vital values, gives each person a wide range of liberties, etc. Some obvious questions to ask here, then, are: are the kinds of people who do what Ernie does (assault, say), dangerous? Do the jail terms we have for assault tend to deter it to significantly lower levels?

On this last question, Diana might make a difference. It depends how we fill in the details of the story. Is Diana strictly a Leibnizian goddess: setting up the universe and then letting it unwind? If so, we still want to punish Ernie. The deterrent effect will work on him as well as anyone else. He could do worse next time if we don't punish.

On the other hand, suppose Diana is very interventionist, far more powerful than we are, and willing to override Ernie's status as a free agent if that's what it takes to get him to wreak her desired mayhem. In that case, the issue gets complicated. Ernie looks more like a typical case of manipulation, and less like an ordinary human being making his own decisions. I'm still thinking that he should be punished in that case too, but I'm less clear.

Hi Randy, part of the problem is that I'm not sure how to respond to the backstory and what counts as distorting my responses, because we don't know everything I think we need to know to respond to it without 'distortion'. (I'm not even sure the backstory is physically possible, but I'll leave that worry aside for now.) For instance, are we supposed to explicitly put aside any notion that Diana is interventionist (as Paul describes it), such that if she saw that her calculations were slightly off such that Ernie would not A, she would intervene closer to the time of A to ensure he still did it? Or is it supposed to be impossible for her to make such a mistake?

In any case, you ask how I'd respond to Ernie harming me were I aware of his backstory. Again, I think it would depend how the backstory is filled in (or how I fill it in). But I'm inclined to respond to the story the way I think I would respond to this story: A very powerful and smart mob boss Dan recruits a poor orphan teenager Ernie into his Family and raises him like a son, ensuring that he is devoted to the Family and that he comes to understand and appreciate its ways and mores. Dan does not brainwash Ernie or undermine his compatibilist capacities. Hell, Ernie can have whatever agent-causal or indeterminist powers you want. But Dan knows how to get Ernie to believe and feel the "right" things to ensure he'll do what Dan wants when asked.

When the time comes, Dan asks Ernie to break my legs for failing to pay up (what was I thinking taking a loan from the Family!?). Fill in the details a bit, and I think I would feel significantly less moral anger (venom and vengeance) towards Ernie than I would towards various other characters who might break my legs (characters with the same capacities for free will as Ernie but with different backstories, and perhaps different opportunities to exercise their free will).

Part of my mitigated reactive attitudes would, I think, be based on my transferring much of my anger to Dan (is that a mistake or distortion? And yes, I understand that more than one agent may be fully responsible for the same outcome). And part of my mitigated anger would be based on my feeling sorry for Ernie and the way he was raised (is that a mistake or distortion?). It's starting to sound like the Robert Harris case a bit, though I think there are important differences too. To be clear, many of the features that mitigate my attitudes and judgments of responsibility towards Ernie would apply to many real world wrongdoers. That's OK by me. I believe in degrees of responsibility based in part on degrees of possessing free will and opportunities to exercise it.

Kip, I've given long consideration to impossibilism about moral responsibility. It's not entirely clear to me why I find it untenable, but I do. I think the explanation isn't a short one.

I think the difference between my view and that of some responsibility deniers, such as Derk, is razor thin.

Perhaps one difference is that I'm inclined to think that desert is compatible with determinism. But I'd say something about it similar to what I say about responsibility: it comes to less that what some folks think.

Eddy, I'm curious why you think it matters whether Diana WOULD intervene if she saw she hadn't set things up right to begin with. Do you think that in the Frankfurt case the readiness of Black to intervene diminishes or somehow alters the degree to which Jones is responsible?

Thank you Randy for the kind response to the confusion in my comment, which I think lies in my missing the possibility of difference in *multiple* deterministic worlds in your thought experiment.

My concern remains that we live in in actual single world. If it is deterministic, that applies to everyone, scientists not just criminals. An example where I suspect this point is missed is Coyne's argument in his short piece "You Don't Have Free Will"*. He dismisses free will, in particular for criminals, on effectively deterministic grounds. However, then (without, apparently, relying on a compatibilist stance) he discusses whether "we" should change the penal system, and says "we" should discard the idea of moral responsibility. It seems to me that there is an elitist error here: is Coyne thinking that the behaviour of the criminal is determined, but "we" can make choices about our own practices and beliefs?

This is dangerous because of the prospect that deterministic or neuroscientific dismissals of human agency will be used in justifications for societies of control.

I'm reassured that discussion from at least some other commenters seems to indicate they feel that we do have real ethical choices, and perhaps therefore that our single world is not deterministic. But I fear the consequences if the opposite view takes a wide hold.


This from John Fischer:

Right, I would deny premise 1. I think whether Ernie is morally responsible is a matter of whether he meets the epistemic and freedom-relevant conditions along the actual sequence that issues in his behavior. What the distal intentions of another agent, even a creating agent such as Diana, don't seem to matter for moral responsibility. So I would be as venemous as a rattlesnake of the sort one finds on the trails here in Riverside, CA.

I ask you this. What if we changed your story, after you have told it, in the following way: "Oh, it turns out that Diana didn't have that intention after all. She did create Ernie in just the way we said she did in the original telling of the story, and the story of his entire life played out just exactly as in the original version. But it was a mistake to suppose that Diana had the intention in question; in fact, she didn't really give a darn." Would you now change your view? Would the venom creep back in? I don't see why this change should make a difference.

Indeed, if you do think it makes a difference, it seems to me that you are committed to denying what I take to be a plausible supervenience principle: one's moral responsibility supervenes on the actual story of one's life (after one's conception), assuming that one's genetic endowments are within the "normal" range. Well, this isn't the most elegant formulation, but I think something like this must be right.

By the way, the reason a comment from John Fischer is posted by me is that the security box for comments isn't working properly. Sometimes it doesn't give you any letters or numbers to retype. If that happens, keep clicking the Continue button until you get some letters or numbers.

John, I agree that the change you describe doesn't make a difference with respect to Ernie's responsibility. Indeed, I accept premise 2!

It seems to me that whether Diana intended this or that, or played no role at all, if determinism is true, then no very venomous or vengeful response to Ernie is fitting.

Thanks, Randy.

Ok, so we both agree with the "no-difference" contention.
Our difference of opinion though remains, and I wonder how we could resolve it or at least illuminate it.

At least however it seems that the Diana scenario is not *motivating* a view to the effect that if causal determinism is true, we should be less venomous. Is this correct? Your "less-venom under determinism" intuition is now floating free of the Diana scenario, right? I suppose that's progress.

In this regard, Patrick Todd has an interesting argument in his first Phil. Studies paper on manipulation that concludes that a compatibilist couldn't accept "mitigated" responsibility. I agree with him that one is in danger of a slippery slope if one goes in the "less venom" direction. Trails are dangerous enough with rattlesnakes around--better to avoid slippery slopes.

John, I've never seen you do anything remotely venomous or vengeful, so I doubt there's a great deal of distance between you and the position I recommended!

I think the role of Diana is that of showing me something that I might otherwise not have seen, a certain thinness to responsibility given determinism. Diana has nothing to do with making this so, though her inclusion in the story helps me see that it's so.

In fact I don't think that determinism mitigates responsibility, if that means that it makes responsibility less than it might otherwise be. I don't think it could possibly be any greater.

Maybe not venom. And I'm not into revenge (except when the 49ers are next playing the Seahawks...)

But still. When I think of what the Boston Marathon bombers did, I have robust indignation toward them. I think I hate them, and I think that the guy who is still alive fully deserves a harsh punishment. This would not change if I came to believe in the truth of causal determinism. That is, if scientists of the future jiggled the equations of the "theory of everything" so it came out deterministic, this would not make me think that the bomber did not act freely in doing what he did. Nor would it make me less indignant or make me think that he doesn't fully deserve harsh punishment.

Same with people who rape, murder, and so forth. So, for example, the fellow in Oklahoma who was recently put to death (but not as efficiently as some would have preferred). He shot an innocent young woman and then buried her alive. He only stopped poring dirt on top of her when he noted that she stopped choking and trying to breathe. We can legitimately worry about the mechanism of the death penalty itself and this particular way of seeking to implement it. But would you really change your views about what that guy deserved, if the scientists at MIT and Cal Tech jiggled those equations around a bit?


Please allow me a question. How exactly does the story of Diana "show you something you might not otherwise have seen" about the "thinness of responsibility" in a determinist world? How exactly does this work, given that you concede that it would not have mattered at all, if Diana had *not* had the intention in question? Given this, I don't see what the role of Diana is. How is it that telling the story of Diana with the specific intention included in the original story can make you see something important, if upon reflection you admit that her intention does not matter? Is it like Wittgenstein's ladder that we throw away?

For further development of this sort of point, see my VASTLY UNDERAPPRECIATED (ha/ha) paper in *Analysis", "The Zygote Argument Remixed".

Argh. I didn't put my point well. You have already pointed out that you wouldn't "change" your position, if under causal determinism. I take it your view is that robust moral responsibility (of the fully venemous kind) is impossible, no matter whether determinism is true of not.

My point in this context: I have the opposite view. As long as the individual in question acts from his own, suitably reasons-responsive mechanism, what does it matter whether determinism or indeterminism is true? He controls his behavior in the way that it is reasonable to require for full, robust moral responsibility. To ask for more is, in my view, to ask for too much.

But perhaps we are an an impasse of intuitions here. I recognize that many people have yours. How, I wonder, might we make progress? Or should we simply follow Franz Kafka's advice: "When you are up against the wall, start describing the wall."

I really appreciate that paper!

Biologists sometimes apply stains to tissue samples to make certain structures of cells visible. An effective stain doesn't create the structures that it makes visible; it makes visible structures that are already there.

I think of the role of Diana in the story as similar. I'm not, of course, presenting an argument to the effect that she plays this role. But you asked how I see it, and that's it!

I entirely agree with you that nothing that might be discovered about the basic equations of physics ought to lead us to deny that we act freely.

At the same time, it wouldn't take a discovery from Cal Tech or MIT to convince me that that the death penalty is barbaric.

As for hating murderers, I go in for that from time to time. I'm not sure I should. But it's not one of my greatest worries.


Thanks. And some, including Patrick Todd, have shown their appreciation for my Analysis paper by refuting it! But despite Patrick's very insightful critique of aspects of it, I have the temerity (or bull-headedness) to cling to the basic insights.

I'm not sure about the stain analogy. I worry because it doesn't seem to reflect a feature of the Diana scenario, that is, the feature according to which one is willing to concede that one's initial intuition--that Diana's intentions matter--really does not survive critical scrutiny.

Allow me to stress: I'm not defending the death penalty. I don't however agree with you that it is obvious that it is barbaric. I don't have a clear view one way or another. I do think that it is administered in a problematic way (or ways), that leads to inequities based on race and also gratuitous cruelty. A defense of the death penalty is not my aim. I do however aim to defend the view that causal determinism would not diminish my robust indignation say at the five boys who raped and humiliated a woman in India, and then used a metal rod to increase the damage, as a result of which she died. I don't know if they deserve the death penalty, but I think they fully deserve harsh treatment (assuming, what I don't know for sure, that they were acting from their own, reasons-responsive mechanisms)...

Is this venomous? If so, then I'm a snake. But I think that it is more accurate to say that this is a deeply human reaction, and any adequate theory of moral responsibility should at least take it very seriously.


1. I'm going to go read that paper RIGHT NOW.

2. In watching you (politely) duke it out with Randy, I will offer the following modest suggestion. Consider this premise:

P. Our pre-philosophical, pre-scientific, natural folk views of ourselves is that we are (partially or wholly) self-created in a way that makes the Diana story impossible. In other words, when we make a decision, that decision cannot be exhaustively explained by antecedent events. That would make us seem like puppet actors, playing out a script that someone wrote a long time ago. Instead, according to this natural, instinctive, naive, folk view, we think of ourselves as adding something, some new decision, not revealed from antecedent events - we add a decision that simply did not exist and could not be explained until it actually occurred, so that we have a fundamental, ultimate freedom from the past, and how it shaped ourselves and our circumstances. In this way, when people blame us for what we do, they blame us because we uniquely carry responsibility for ourselves, our self-creation, and our environment. We cannot point fingers at our past, or our genetics, or our upbringing, because we have a radical kind of freedom, albeit one that is difficult to articulate, with blurry boundaries, and arguably incoherent.

This premise P is, roughly speaking, the intuition that libertarianism is true. And, depending on the framework of the study, there is a lot of support for the widespread allegation that libertarianism is naturally intuitive.

Here, then, is the way in which the Diana story is illuminating: it shows how false P is, at least if determinism is true.

Without P, determinism is messy and complicated. The story of our lives naturally and slowly evolves, emerging from these various background conditions. Over time, our parents reproduce, our zygote is formed, we finally are born, and then we are raised as children. In the meanwhile, our brains develop, grow, learn, etc. We experience new events and sensations, and we're educated according to our circumstances.

In the scenario of mere determinism, then, it remains obscure how self-created we are. You can point to mathematical equations and explain to someone that you're not self-created, but this is difficult to feel in your bones. Our whole life story doesn't seem baked into the initial conditions of the universe - rather it seems to slowly emerge and rise forth, slowly and chaotically, from a soup of earlier events. In this way, it doesn't look *that* different from the libertarian story.

In the Diana story, however, we have something that smacks us in the face with predestination - Diana and her intentions. Way, way, way back, long before we ever could have done anything about it, there is a clear, simple, direct explanation for everything we do in our lives. Diana was there, long ago, and in a single moment, designed our entire lives. This makes the level of fate, predetermination, programming, etc., simple, clear, and undeniable.

Yet, as you, me, and Randy all equal - whether Diana is actually there or not makes no difference. What Diana does is provides a simple and clear example to show how little freedom and responsibility we have in all situations (Diana or no Diana). The situation of mere determinism is the complicated one, the tricky one, the obscure one, the muddled one - it is the simplicity and clarity of the Diana story that *reveals* just how little freedom and control we have in all these scenarios. In short, it shows us just how false P is.

Hi John

You suggest 'a plausible supervenience principle: one's moral responsibility supervenes on the actual story of one's life (after one's conception), assuming that one's genetic endowments are within the "normal" range.'

Does this mean that you think it'd make a difference to Ernie's moral responsibility if we changed when Diana's interference took place? We could stipulate that Diana modified a pre-existing zygote (shortly after it was conceived or perhaps later) so that the resulting agent A-ed at some point in the future. Or we could stipulate that Diana waited until Ernie was born and then modified in his brain so that he developed into an agent who will A later in life. (This would make it similar to Pereboom's Case 2.) I can't see how *when* Diana's interference takes place would make a difference, but your proposed supervenience principle seems to suggest that you think otherwise.

Hi Randy (if I may)

I'm with you on the hard-line, soft-heart position. I think compatibilists are fine to say that Ernie deserves *some response* for his action. But I don't think they need to accept that particular actions have particular responses.

John, I don't concede that my initial intuition was mistaken. I don't think I ever had the intuition that Diana's intention mattered.

I presented a case just about exactly like Erie's in my dissertation. The point there was to contribute support to a kind of partial incompatibilism: some form of responsibility survives determinism, some possible form does not. I've since rejected the second half of this.

Toward the end of Threepenny Opera, the chorus sings, "Don't punish wrongdoing too much." I confess that almost all of my moral outlook comes from that work. But the view I'm suggesting here isn't exactly that one. It's more like, "Don't punish wrongdoing with the wrong attitude." Who could disagree?

Dan (the mob boss) is almost positive Ernie will break my leg when that's asked of him, but if Ernie shows any hesitation Dan will give him a compliance drug to ensure he carries out the gruesome task. Does this potential intervention make Ernie less responsible when Dan doesn't deploy it? I'm not sure. But I feel my venom diminishing a bit more when we add this to the backstory. Dan has even greater control over whether my leg gets broken. Ernie, seems to have less.

Unlike Dan (and Diana), Black is not causally implicated in Jones' action, much less the causal source of Jones' decisions.


I think I know the answer to this question, but I'll ask it anyway. Is the reason you deny the possibility of the more robust form of responsibility because you deny that we can have any more robust (forms of) control over our behavior than we actually do? So, for instance, does Ernie have as much (or as many forms of) control as agents like us can possibly have?

Those who answer "no" to this second question might consider adopting a position according to which determinism mitigates (without eliminating) responsibility. This suggestion is developed in another vastly under appreciated paper in PPR called "Mitigating Soft Compatibilism"!

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan