God, Jones, and Black
December 12, 2014 — 13:15

Author: Michael Almeida  Category: Uncategorized  Tags: , ,   Comments: 12

Harry Frankfurt is credited (by some) with having shown that alternate possibilities are not necessary for freedom and moral responsibility. There are any number of Frankfurt-style counterexamples (FSC) to the principle of alternate possibilities (PAP) and any number of (more or less relevant) versions of PAP. There are, further, any number of (more or less cogent) replies to each of the counterexamples. The literature is vast. I want to suggest a rather direct reply that (as far as I know) hasn’t been suggested and that does not depend on any specific version of the FSC. I will consider a brief counterexample with just a few bells and whistles, add the ornaments that you think matter. First, the sort of PAP that matters to the discussion is something like PAP0.

PAP0. S is free and morally responsible for what he has done only if S could have done otherwise.

It’s important to note that PAP0 does not apply to cases where we had choices to do otherwise but now cannot do otherwise (I could have tied myself to the mast, but now I cannot resist the Sirens). Cases like the following aim to show that PAP0 is false (I borrow liberally from Fischer here, but change the story a bit)

Black has secretly inserted a chip in Jones’s brain that enables Black to monitor and control Jones’s activities. Black can exercise this control through a sophisticated computer that he has programmed so that, among other things, it monitors Jones’s voting behavior. If Jones were to show any inclination not to rob the bank then the computer, through the chip in Jones’s brain, would intervene to assure that he actually decides to rob the bank.

Jones of course decides on his own to rob the bank, and the device never causes his actions. But is it true that Jones could not have done otherwise? Some have advanced “flicker of freedom” replies and “weak flicker of freedom” replies wherein Jones can at least initiate the action of not robbing the bank prior to the intervention. I don’t find these unreasonable responses, actually. But suppose that we add to the example that Black is a perfect predictor who predicts everything you will do. So, no need for the device, Black already knows whether you will rob the bank or not, and he decides on the basis of this information whether to cause you to rob the bank.

Ok, here’s the direct reply: Consider a possible world W in which Black predicts what Jones will do. Black predicts that Jones will rob the bank at t, and it is in fact true that Jones will rob the bank at t. But is it true in W that Jones cannot fail to rob the bank at t? Here is the question to ask in W: What would happen were Jones not to rob the bank at t. Here are some answers.

1. Were Jones not to rob the bank at t, then it would have been the case earlier that Black predicted correctly that Jones would not rob the bank and caused Jones to rob the bank.

(1) effectively tells us that had Jones not robbed the bank at t, then a contradiction would have been true (Jones would have both robbed and not robbed the bank). But that’s not true: the closest worlds to ours in which Jones does not rob the bank are not impossible worlds. (1) is a backtracking counterfactual, and so requires (in this context) an unacceptably non-standard resolution. We would rather say that (2) is true.

2. Were Jones not to rob the bank at t, then it would have been the case earlier that Black predicted correctly that Jones would not rob the bank at t but Black’s device failed to cause Jones to rob the bank.

Or, more likely, (3),

3. Were Jones not to rob the bank at t, then Black would not have predicted earlier that Jones would not rob the bank at t.

Black is a perfect predictor, but not an essentially perfect predictor. (2) & (3) are compatible with Jones being able in W not to rob the bank.

Ok, but what if it is not Black, but God who wants to ensure that you have no alternative to robbing the bank? If God predicts that you will not rob the bank, then he causes you to rob the bank. This raises all sorts of freedom & foreknowledge problems, but set those aside. Could you fail to rob the bank?

4. Were Jones not to rob the bank at t, then it would have been the case that God predicted correctly that Jones would not rob the bank at t and caused Jones to rob the bank at t.

But again, this amounts to saying that the closest worlds to our in which Jones fails to rob the bank are impossible worlds (worlds in which Jones both robs and fails to rob the bank at t). That just isn’t true. More plausible candidates for closest worlds include the worlds in (5).

5. Were Jones not to rob the bank at t, then God would not have wanted Jones to rob the bank at t.

Certainly God does not want Jones to rob the bank in every world in which Jones exists, otherwise Jones necessarily robs the bank. In that case he is clearly not free. But (5) is compatible with Jones freely robbing the bank.

So, even under the assumption that God = Black, and God wants you to rob the bank, you still have the alternative not to do so! If you rob the bank, you had an alternative. That’s why you’re free and responsible for doing so.

ADDENDUM

Suppose that (4) above is true. If so, there is a simpler reply to the counterexample. Here again is (4).

4. Were Jones not to rob the bank at t, then it would have been the case that God predicted correctly that Jones would not rob the bank at t and caused Jones to rob the bank at t.

(4) entails (6).

6. ☐~(Jones fails to rob the bank).

With a small inference from (6) we have the reply to the counterexample.

7. ☐(Jones robs the bank).

According to (6), there is no possible world in which Jones fails to rob the bank. According to (7), it is metaphysically necessary that Jones robs the bank. But no doubt, if it is metaphysically necessary that Jones robs the bank, then he does not freely rob the bank. This reply is also available for those who insist that (1) is true.

 

Comments:
  • WH

    I doubt this claim: If P necessarily does A, then P does not do A freely. How can someone defend this claim without PAP? I know this claim is regularly taken to be true, but I don’t know of any argument supporting it.

    December 13, 2014 — 18:56
  • Michael Almeida

    WH,

    That’s a great question. Here’s how to defend that view without PAP. Take necessitarianism to be the view that there exists exactly one world, W. Everything that happens in W happens as a matter of metaphysical necessity. The actions of agents in W is as strongly determined as any action could be, more strongly than any form of fatalism entails. I don’t know anyone–compatibilist or not–who thinks that the agents in W are free. That is probably because the agents in W are not the source of their own actions. But also, even strong compatibilists (and semi-compatibilists) believe that free agents must enjoy at least the metaphysical possibility of failing to do what they do. So, there is a broad sense of having an alternative that everyone thinks is required for free action, at least as far as I can see. Further, most who concede that the agents are free in Frankfurt cases, would deny that the agent is free in those cases if they learned that the agent’s actions were in fact caused or metaphysically necessary. So, it is not the strong form of PAP that explains these different intuitions.

    December 14, 2014 — 9:08
    • WH

      I guess it depends which compatibilist account of free will one takes to be true.

      Fischer’s account does seem to require the metaphysical possibility of failing to do what agents actually do in order to prove reasons-responsiveness. But perhaps one can circumvent this problem by using per impossibile counterfactuals, but I don’t know if something like that can really work.

      But it seems to me that neither Frankfurt’s nor Mele’s accounts require this metaphysical possibility of failing to do what one actually does. Mele’s conditions for freedom are ideal self-control, reliable deliberation, the lack of compulsion, and the presence of sufficient beliefs/information conducive to effective deliberation. On Mele’s account, there would be no compulsion even if all of the agent A’s values are unsheddable as long as these values are subject to critical evaluation by A and/or the unsheddability of these values is arranged by A. Mele’s account has no mention of possible failings of action. Thus, it seems to me, but I’m not sure, that Mele’s account is consistent with necessitarianism.

      Apart from Mele’s (or Frankfurt’s) account, I have the intuition that God is always the source of his actions even if he has no possible alternatives for action, and thus God can still be free if necessitarianism is true. I admit it’s not a very strong intuition though.

      December 14, 2014 — 10:56
  • Michael Almeida

    Mele’s conditions for freedom are ideal self-control, reliable deliberation, the lack of compulsion, and the presence of sufficient beliefs/information conducive to effective deliberation.

    I don’t think there is any self-control in cases where my actions are metaphysically necessitated. I have no control at all in such cases. In Fischer’s case of controlling the direction of the car you’re driving, I think you do have control of what you’re doing (self-control as well). But that’s because you are not being causally or, worse, metaphysically necessitated.I cannot see how we can properly attribute control to someone who pleads that his actions were metaphysically necessitated. I’m not sure what to say about Frankfurt’s second-order desires if all of those are metaphysically necessary. In any case, I wouldn’t burden Mele, or anyone else, with defending the view that Spinoza’s agents are really free.

    December 14, 2014 — 11:15
  • John Alexander

    Good morning Michael
    If one accepts PAP, exactly what is implied? Does it mean the S must actually be able to do x or simply think that she can do x? To use a common example, imagine that t1 S enters a room, closes the door, and sits at a table. Unbeknownst to her someone lock the door so that she cannot leave. We know that S cannot leave the room, but does this mean that she is not ‘free’ to leave the room? At t2 (S sitting at the table) it seems clear that S thinks she can leave the room and based on the information that she has she gets up from the table goes to the door and turns the knob. Lo and behold the door is locked and she cannot leave the room. At t3 she knows that she cannot leave the room, but does this new information entail that at t2 she was not free to leave the room. It seems to me that from her conceptual perspective at t2 she is free to leave the room because she does not know that the door is locked and has no reason to think that it is. I think that ‘freedom of will’ is the ability to form a judgment about how one is going to act that is based on the information that the agent has at the time she is making the decision to act. At t2 S is free to leave the room even though she cannot leave the room because the door is locked. At t3, S will realize that she is not able to leave the room, but this does not imply that she was not free to leave the room at t2 even though she now realizes that it was not possible for her to leave the room at t2. I think it matters from which perspective, the 1st person (S) or the 2nd person (those who know the door is locked) we are trying to understand what ‘freedom to act for S’ means. My guess is that Jones will think the he is free to rob the bank because he accepts PAP (I can either rob the bank of not) and because he does not know what Black is doing (Jones will rob the bank). From his conceptual perspective he is correct to think this even though we know differently.

    December 20, 2014 — 9:44
  • Michael Almeida

    Hi John,

    The case of the locked room is typically invoked (I think this is Locke’s example) to distinguish between what we do “voluntarily” and what we do freely. If S is in the room and has no desire to leave the room, then S is in the room voluntarily (even if the door is locked). But S is not in the room freely, since S cannot leave the room. PAP is not typically construed as context dependent. If, from the first person perspective, it seems to you that you could leave the room, then you might believe you’re free in remaining in the room, but you’re not. That’s what PAP entails, construed as a condition of freedom (rather than responsibility). You may nonetheless be in the room voluntarily, if you like that distinction.

    December 20, 2014 — 9:55
  • John Alexander

    Hi Michael
    You state in the intro to your initinal 5 possible scenarios that beside Black predicting that Jones will rob the bank at t, “it is in fact true that Jones will rob the bank at t.” If it is in fact true that Jones will rob the bank at t, is it a fact because Black predicts it or it is a fact independent of what Black predicts? I take it, from your wording, that you mean the latter. But, if this is so, how can it be possible for Jones not to rob the bank at t? I take it that PAP maintains that from some initial starting point Jones can either rob the bank or not – both options are ones that Jones can, in fact, do. But, if at that starting point (t-1) it is a fact that at t Jones will rob the bank then it is not possible that Jones will not rob the bank at t or it would not be a fact that at t-1 that Jones will rob the bank at t. I take it that a fact reflects the way things are.

    I also find this puzzling: “Certainly God does not want Jones to rob the bank in every world in which Jones exists, otherwise Jones necessarily robs the bank. In that case he is clearly not free.” The puzzle does not have to do with the concept of necessity, but rather the relationship between God wanting x and x occurring. Are you suggesting that in every possible world where God does not want Jones to rob the bank that it is God wanting Jones not to rob that causes Jones not to rob the bank? Is God’s wanting, God’s ‘chip’?” It seems that you are suggesting that if God wants x then God gets x. But this seems wrong in so far as in this world it certainly appears from Scripture and the ‘logic of God’ that God does not want us to violate, say, the Ten Commandments or the Golden Rule but in fact many do. The fact that many do violate them seems to suggest that we are free to either follow them or not regardless of what God wants. It also seems that if God wants x in W then He want it in all possible worlds or God would not be God in all possible worlds. You state that in one PW Jones is right-handed and in another PW he is left-handed. But, this seems to indicate the existence of two Jones not one Jones in different PW’s. From this it seems to follow that if we know what God wants of us in this world then we do not need ‘possible worlds.’

    This brings up the interesting problem of what a rational agent would do if he or she knew what God wants. If Jones knows that Gods does not want him to rob the bank, and if Jones is rational, can Jones rob the bank? Based on some of your other writings, it seems that if God does not get what He wants then he severely punishes those who do not abide by His wants. This being the case then besides Jones knowing what God wants, he also knows the he will be punished if he does not do what God wants. It seems that a rational agent would not want to be needlessly and avoidably harmed. Knowing what God wants, and being rational, Jones does what God wants. Does this mean that if Jones is rational he necessarily does what God wants in the sense that the option to rob the bank if not a real option for a rational agent who knows what God wants and what He will do if He does not get what He wants. If Jones necessarily does what God wants then he is not free (Epictetus and Spinoza aside) to perform the alternative action of robbing the bank.

    Meanwhile back to the distinction between voluntary and freedom. Consider that Smith wants to walk to the bank and he knows that at some point in his walk he will have to choose between two routes that he can take to get there. One of them is shorter, but more crowded with people and the scenery is less enjoyable, the other is longer, but with less people and better scenery. He decides at t to take the later route. At t2 he is at the point where he has to make a decision, but finds that the route he wanted to take is closed and has been closed prior to t. I take it that PAP maintains that Smith at t is not free to take either route because one of them is closed off. (I assume from your earlier response that Smith is acting voluntarily at t, but not freely.) This is where I find the distinction not helpful. How can Smith voluntarily to x without being free to do x? Is it more important to be able to act voluntarily or to act freely? Is there really a difference? I think not.

    It seems to me that at time t Smith is free to act towards doing what it takes to make a choice at t2 because from his standpoint he thinks, maybe incorrectly, maybe not, that both options are available to him. Smith is not going to think that there is a Black (God) somehow controlling what he does (Black wants him to take the shorter route). But, even if he did think about ‘Black-like scenarios’, how would this affect his decision to act. What would be the advantage of thinking that this is a real possibility? There is certainly no logical contraction in thinking that Black might exist (Descartes Evil Demon comes to mind), but if we think that this is practically possible then how would we ever decide how to act? It seems to be a non-starter. He has a choice; to act or not to act. Both of them seem to be real options that Smith is capable of performing. Smith cannot possibly know what Black wants of him other than any ‘either-or’ scenarios that might pertain relative to his proposed action, e.g., Black either wants him to rob the bank or Black does not want him to rob the bank. So, he would reasonable discount Black in his decision-making and act based on what he thinks is the case – that there are two routes that are available for him to take and that he is free to take either one. It may turn out that what he thought was the case is not the case, but that has not stopped him from acting at t as if it were the case. How could he find out he was wrong if he did not act? When things turn out differently than he thought would be the case, he simply makes an adjustment in his beliefs to account for this new information, but this does not mean that his action was not free.

    Anyway, thank you for a thought provoking post.

    December 22, 2014 — 8:56
  • Michael Almeida

    if at that starting point (t-1) it is a fact that at t Jones will rob the bank then it is not possible that Jones will not rob the bank at t …

    I don’t know how you reach this conclusion. There are many things it is a fact that I will do, but which I might not have done.

    It seems that you are suggesting that if God wants x then God gets x. But this seems wrong in so far as in this world it certainly appears from Scripture and the ‘logic of God’ that God does not want us to violate, say, the Ten Commandments or the Golden Rule but in fact many do

    That’s a fair point. I’d substitute ‘God wills x’ for ‘God wants x’. Nothing turns on it being a desire of Gods.

    . I take it that PAP maintains that Smith at t is not free to take either route because one of them is closed off.

    I don’t think PAP entails that. You could simply not take either route, so there is an alternative.

    Is it more important to be able to act voluntarily or to act freely? Is there really a difference? I think not.

    Well, I gave you the distinction as it is classically made. Whether you find it useful, I don’t know. It does seem to me to capture an interesting distinction.

    It seems to me that at time t Smith is free to act towards doing what it takes to make a choice at t2 because from his standpoint he thinks, maybe incorrectly, maybe not, that both options are available to him

    I think this confuses the epistemology with the metaphysics. Thinking you’re free is a function of your knowledge, actually being free does not (in the cases we are discussing) depend on what you know. There are cases in which the epistemology has metaphysical implications, for instance, it’s probably true that I cannot repair the garage if I don’t know how to. And since I cannot, I am not free to do so. On the other hand, knowing how to repair the garage does not entail that I can or that I’m free to: maybe I’ve lost my limbs in an accident.

    December 22, 2014 — 9:49
  • John Alexander

    It seems that I may be confused regarding how you are using “alternative.’ Are saying that there are two options even if one is closed off because it is still an alternative to the one that is open because I could take if it were open? If this is so, then are you arguing that being free means that I could take either option if those options are available? The fact that one option is closed off means there must be an alternative to be closed off. This seems plausible, dare I say correct, but is one committed to using ‘free’ in this manner? Why can’t one simply say that one is free if one is actually able to take the alternative and not free if the alternative course is closed for some reason? Does it just come down to how we are defining our terms? I may be confusing epistemology with metaphysics again (which I probably do quite often not being a metaphysician).

    “I don’t know how you reach this conclusion. There are many things it is a fact that I will do, but which I might not have done.”

    OK, so it is a fact that Jones robbed that bank and it is also a fact that he might not have robbed the bank given the options available to him. But, it seems that you did not say this in you intro. You said that in fact he will rob the bank at t? I understood this to mean that there are facts about the future describing what will occur, not what might occur if certain things happen prior to t. Are you saying that it is a fact that Jones will rob the bank if Jones decides (is forced to) rob the bank?

    December 22, 2014 — 10:35
  • Michael Almeida

    In the case you describe, there are two routes, one of them is blocked. You ask whether the agents can take the open route freely. I said yes, because there is an alternative. He could for instance, just stand still, or turn around and walk back or stand on his head for the rest of his life. There are lots of alternatives.

    You said that in fact he will rob the bank at t? I understood this to mean that there are facts about the future describing what will occur, not what might occur if certain things happen prior to t. Are you saying that it is a fact that Jones will rob the bank if Jones decides (is forced to) rob the bank?

    I’m certain that there are facts right now concerning everything I will do from now till I die. I have no doubt about it. So, that’s exactly what I will do. Absolutely nothing follows from this concerning what I might do. Same for Jones. We both have alternatives, but in fact we will not take them.

    December 22, 2014 — 11:40
  • John Alexander

    Thanks Michael. I think I understand your argument now. I hope that you and yours will have a good Christmas and that 2015 will be a healthy and rewarding year for you.

    December 22, 2014 — 16:21
  • Michael Almeida

    Thanks John! Merry Christmas to you!

    December 22, 2014 — 17:13
  • Leave a Reply

    Your email address will not be published. Required fields are marked *