Wednesday, January 23, 2013

Compatibilism and Regret

This will be quick. Sometimes Calvinists hear that their belief in divine determinism makes regrets unintelligible. One Arminian puts it like this:
It would seem that regrets can only make sense, however, if we hold to a libertarian view of free will. Regrets are nonsensical if we believe that all of our actions are determined by decree and circumstances which are beyond our control. There is no point feeling regret for something you could not possibly have done otherwise; yet we still feel regret. Do Calvinists feel regret? How do they work such feelings into their worldview? Do they temporally shelve their worldview when confronted with the experiences of daily human life? Do they somehow train themselves to have no regrets so as to conform their feelings with their belief in determinism? I am curious to know.
I've had regrets, most of us have. Mine, and I'm assuming this is so for most people, go roughly like this: This situation I've brought about is somehow undesirable. If I knew then what I know now, I'd do differently. But this is consistent with determinism: Same past, same future; different past, different future (perhaps). What's the alternative? Do I say, "I wish I would have done otherwise given the exact same circumstances, the exact same information, the exact same reasons, the exact same belief-desire complexes, etc.? Would I have done differently? That seems quite odd to me. Why would I have done differently? Why think I would have done otherwise? I don't regret something by looking back and saying I would have done differently given the exact same situation. I don't think many others think that either. They might say, "I noticed X back then, but didn't think it was relevant. I wish I would have seen the relevance of X to my situation, then I would have done otherwise." And this, of course, is fully consistent with determinism.

30 comments:

  1. Perhaps it would be better for them to say it makes no sense to feel regret given our belief in God's wisdom and meticulous providence.

    ReplyDelete
  2. The only problem with that is that it's not any better.

    ReplyDelete
    Replies
    1. It looks better to me because given God's wisdom and providence we know that events unfolded according to His perfectly good plan. So how can we regret His wise plan coming to fruition? Not that the Calvinist has nothing to say in return, but it looks more appealing than "regret makes no sense when wecan't do otherwise."

      Delete
    2. I think Jonathan's point can be put as follows. The following challenge:

      1) If determinism, then regret is unintelligible.

      is made stronger if you introduce the following as an additional condition:

      2) God's divine decree issues in the best of all possible worlds

      So that it looks like this:

      3) If divine determinism, and if the divine decree issues in the best of all possible worlds, then regret is unintelligible.
      4) Divine determinism holds and the divine decree does issue in the best of all possible worlds.
      5) Regret is unintelligible.

      Delete
    3. I'm not trying to make (1) stronger. I don't think it works, as Paul explained. And I'm not really concerned about whether this is the best possible world.

      Delete
    4. I'm not attributing a belief or desire to you one way or another. I'm merely considering a claim, a claim which you prefaced by saying "Perhaps it would be better for them," which, to me, seems like a setup for a suggestion for how their challenge could be made more difficult to deal with. What you personally believe isn't relevant here.

      If by "God's wisdom and meticulous providence" you didn't mean something like "best of all possible worlds," then that's fine, but I don't think the concept is too far off from what you had in mind.

      Delete
    5. If my beliefs aren't relevant here then it's not "Jonathan's point"

      Delete
    6. BB,

      First, we need to be careful to distinguish between the best of all possible *worlds* and each true proposition of the world being "the best."

      Second, we can question the assumption of "the best" of all possible worlds, if there's a *set* of *equally* best, then in some of them, assuming I even exist, I do other than I do in the actual world.

      Third, you've just given us a bunch of conditionals, I don't see any reason to believe any of them.

      Fourth, I don't doubt there's some metaphysically loaded, dubious, and highly questionable libertarian model on which regret is unintelligible given determinism. I just don't care. I think more ordinary uses of 'regret' don't commit us to those models. I think most people mean by 'regretting X' either "wishing X did't happen," which doesn't commit you to accepting bloated libertarian metaphysics or "If I knew then what I know now, I'd have (perhaps) done differently." Both of these are consistent with determinism.

      Fifth, when you say, "The objector suggests that regret is only intelligible if the person who is experiencing regret over an action could have chosen otherwise than he or she in fact did," that doesn't strike me as right. What are the times we're holding fixed? If the regretful choice is at t2, do we hold fixed all times t1 ≤ t2? Do we hold fixed *everything* right up to the millisecond before choice? It seems we'd need some pretty whacked out indeterminism to get us to do otherwise. Or do we hold fixed all times up until 3 years old? I just don't think people think this way. Surveying ordinary uses, for example when I think about my own regrets, I wish I would have *seen* or *known* or been *appraised* or had been * familiar* with some relevant piece of information because I know that had I been I would have acted differently. This is why so many cases of regret that are seen in, say, psychologists' office start by saying, "If only . . ." e.g., "If only I hadn't had that 4th drink, " "If only I had seen that car coming," "If only I didn't mess up on that calculation, the company wouldn't have gone under," etc. For example, if we don't like where we are, but we still think it is the best place to be given what we know, we don't *regret*. If you are unaware of relevant information that could have changed the dire outcome and, *had you been aware of it* you would have acted on it, we then *regret*. Again, this is consistent with determinism.

      Last, when you say, "In other words: I don't care how I could've gotten there, I just wish I would've gotten there," that is consistent with determinism. Determinism is consistent with our *wishing* some state of affairs had been otherwise.

      Delete
    7. Your first and second points seem right. I agree. Still, given our epistemic condition, regret could either be had over a feature that plays a role in the world being best, or over a feature that, if it were different, wouldn't eject the world from the set of best possible worlds. But even if our regret over a feature of the world is of the former kind, since God allows that we often unwittingly pray for things not in our best interest, there is no problem with the regret itself. It would only be a problem if we somehow knew, or should have known (as in the case of Peter rebuking Jesus for suggesting that dying is an important part of his mission, only to get strongly rebuked himself), that what we wish for is inconsistent with God's plan.

      Paul wrote:
      "Third, you've just given us a bunch of conditionals, I don't see any reason to believe any of them."

      You're committing the same mistake as Jonathan, though in the other direction. I'm not suggesting that you should believe any of the claims I've put up. It's not an argument that I necessarily accept. I'm responding to your claim that Jonathan's proposal is "not any better" by suggesting how it might be seen as being better than the proposal of the original objector.

      Also, I'm confused as to why you'd say that all I've given are conditionals, when 4) and 5) are straightforwardly not conditionals. The form of 3-5, as I'm sure you recognize, is of a sort of "compound" modus ponens. The form isn't important, anyway, since it doesn't need to be framed that way in the first place. Moreover, you need to recognize what the argument is intending to do in order to assess whether it says enough or not. If I were giving that argument because I believe its conclusion, I would need to make a case for 3. But since the argument is only intended to suggest how it is stronger than the initial objection, there is a far different, and lighter, burden of proof to discharge. I only have to show that the introduction of 2, that is, that 3 (which is a modified 1) generates a greater problem than the initial objection by itself. Given your comments under "First" and "Second," though, I'm not sure I can do that.

      Paul wrote:
      "Fourth, I don't doubt there's some metaphysically loaded, dubious, and highly questionable libertarian model on which regret is unintelligible given determinism. I just don't care. I think more ordinary uses of 'regret' don't commit us to those models. I think most people mean by 'regretting X' either "wishing X did't happen," which doesn't commit you to accepting bloated libertarian metaphysics or "If I knew then what I know now, I'd have (perhaps) done differently." Both of these are consistent with determinism."

      Not sure what this is in response to. I was countering the suggestion that the libertarian objection is incoherent.

      Actually, regarding your fourth, fifth, and last points, it seems to me that you've misunderstood the point of my comments. My point wasn't to suggest regret is incompatible with determinism.

      Of course regret can be taken in the way you suggest -- I'm only saying it need not be. Regret can also take as its subject an outcome, or a consequence, with no particular preference specified vis-a-vis the probabilistic or influential factors leading to the more desirable outcome or consequence. I'm suggesting a kind of consequentialism about regret, at least for some cases. That seems to be all I need to ward off the incoherence suggestion. For if regret can be had over an outcome, without that regret involving the further desire that the preferred outcome come about via the later held reasons, then the objection doesn't fall prey to the incoherence that Steve alleges.

      Delete
    8. Berny, the context of dialogue is that regret is incompatibile with determinism. Even if there's *some* model of regret on which regret is incompatible with determinism, that doesn't show that regret is incompatible with determinism. So, for all you've said, I can't see how you've "strengthened" the initial case.

      As far as your point about when regret is a "problem," I'm not sure I follow. The use of "problem" is vague. I'm not sure what work it's doing. It wasn't an obvious feature of either my argument or the one I was responding to. To that that regret would be *inconsistent* with determinism if we knew God's plan is false. For we could still regret something even if we *shouldn't* regret it. But we're not talking about the *propriety* of regret but it *possibility* given other things. So, even granting your claim that "if S knew God's plan that X, then regretting X would be improper," this wouldn't show that we couldn't "regret" and do so intelligibly, especially as unglorified. To say "regretting X is improper" is different than saying "regretting X is *non-sensical*," the Arminian's claim.

      Berny wrote:
      "I'm responding to your claim that Jonathan's proposal is "not any better" by suggesting how it might be seen as being better than the proposal of the original objector."

      I'm not committing any error, I'm saying I don't think your suggestion helps.

      Berny wrote:
      "Also, I'm confused as to why you'd say that all I've given are conditionals,"

      I didn't; I said "a bunch" of conditionals. And I was saying that I don't see how that made the case better.

      Berny wrote:
      "Not sure what this is in response to. I was countering the suggestion that the libertarian objection is incoherent."

      I never said the libertarian objection was "incoherent." That was *their* contention. I *granted* their could be some highly dubious account on which the objection is coherent. But if you haven't noticed, I'm approaching this from the perspective of ordinary language philosophy. I'm merely trying to make the modest suggestion that ordinary uses of 'regret' fit perfectly fine with determinism and I reject the idea that there's some single "thing" or whatever that "regret" *means* and *that* is "inconsistent" with determinism.

      Berny said:
      "Of course regret can be taken in the way you suggest -- I'm only saying it need not be."

      But I never suggested otherwise. *I'm* saying 'regret' need not be taken to assume that: [](If S X's at t1 at regrets X at t2 (t2<t1), then S must have been able to do otherwise at t1 *holding everything fixed* for all times ≤ t1), or something there abouts.

      Berby said:
      "That seems to be all I need to ward off the incoherence suggestion."

      There's no incoherence suggestion. My objection stems from the claim that regret presupposes that we must have been able to do otherwise. I think the libertarian can give necessary conditions here—and they admit compatibilists can—but when they try to give sufficient conditions of 'able to do otherwise' they run into problems. This—they best of them—will involve all the sorts of questions I've been asking, viz., about times up the the action or decision and how much we hold fixed.

      Delete
    9. Your distinction between propriety and possibility is helpful. I guess what I had in mind is that the case against regret having a *place* in a deterministic system is stronger when Jonathan's consideration is added. But that's because I was taking "intelligibility" or "compatibility" to be broader than the modal logical reading you were (rightly) giving it. I was taking it in such a way to allow for a normative-theological challenge of regret, which as you note is an ethical, not a logical, consideration.

      Beyond that point, which is well taken, is a confusion about the other part of my comments tonight. I'll give three quotes to hopefully show this.

      Paul wrote:
      1) "I never said the libertarian objection was "incoherent.""
      2) "But I never suggested otherwise."
      3) "There's no incoherence suggestion."

      To respond to all three: I never said that you said that. I never said you suggested otherwise. And, yes, there is an incoherence suggestion.

      The suggestion is Steve's. I think the confusion here stems from you taking my reply to Steve to be a reply to you. To reiterate: that part of my comment had to do with responding to Steve's suggestion that the libertarian objection is incoherent to begin with. Now maybe I've misunderstood Steve, but if I haven't, then what I've done is show how the objection brought forth by libertarians isn't incoherent given that "regret" can be taken purely consequentially, and thus doesn't require that the antecedent conditions for any regretted actual state of affairs be different in order to wish that a different state of affairs had obtained. Or if the antecedent conditions are different, the causally influential conditions need not be supplied by the reasons, decisions, or actions of the agent in question.

      Delete
    10. I still don't think regret is always improper on the condition we were appraised of God's plan. Suppose I yell at my kids in a harsh and unjustified manner, even if God told me the bigger good purpose this served, I can still regret it. If we hold to double causation (e.g., Markosian appeals to this in his paper on compatibilist agent causation), regret applies to my level of causation. I wish *I* hadn't done that. Or say I make some stupid business decision. I still can wish that *I* hadn't done that. I can still say, "If only I had not extrapolated from the data, I would have . . .", for even if God uses this for some good reason, I can still regret my stupidity.

      As far as Steve's objection and coherency, I think you're focusing on the wrong thing. I think "do otherwise" needs to be focused on. There's a lot here and I'm not sure we both really want to get into it all. Some of the things we'd need to say are: we'd have to severely limit the scope of intelligible regret, for I have no reason to believe that a lot of the things we regret, even on libertarianism, we *couldn't* have done otherwise. We've *already* set are character. Suppose I libertarian freely set my character 25 years before I did some action I now regret. Does regretting that action involve at all the idea that I could have/should have done otherwise 25 years earlier? And why would we do that? Maybe the character trait is *normally* good, so I want it. I don't regret my forming it 25 years before the regretful decision. How are we analyzing this "doing otherwise?" Suppose I don't A and I regret it. So, I didn't A and I was able to A. What does that mean? Maybe: Smith is able at t1 to do A at t2 (where t1 ≤ t2) _iff_ there is a possible world W, and W is an extension of the actual world, @, at t1 where Smith does A at t2 by determining himself to exercise his general ability to do A" (this idea is taken from a forthcoming paper by Campbell). One question that arises what time is the important time to specify? What true propositions about the past matter when considering proper extensions of @? A good case can be made for *all* of them, but then this leads to worries of fatalism or at least restrictivism, which has always been a problem for indeterminism (Campbell argues that indeterminism *entails* fatalism (forthcoming). Among the issues that arise here are those of luck, contrastive explanations, &c. So I don't think the incoherency has been clearly escaped. We can't say what you say to avoid *this* incoherency problem. You say, "In other words: I don't care how I could've gotten there, I just wish I would've gotten there," and if we mean that to apply to the doing otherwise, here's *one way* I "wouldn't gotten there." I regret turning right when I needed to turn left, for this ended up taking me an hour out of the way. Do I say, "I don't care if an indeterminist ray shot from the sun causing me to wildly jerk to the left"? Do we say, "I wish I had had a mild aneurism that reversed my neural structure making me think that left was right and vice versa"? Do we say, "I wish I had irrationally turned left for no reason whatsoever, and then regained my senses"? Does libertarianism require something like *that* to avoid incoherency charges?

      Delete
    11. Paul wrote:
      "I still don't think regret is always improper on the condition we were appraised of God's plan. Suppose I yell at my kids in a harsh and unjustified manner, even if God told me the bigger good purpose this served, I can still regret it. If we hold to double causation (e.g., Markosian appeals to this in his paper on compatibilist agent causation), regret applies to my level of causation. I wish *I* hadn't done that. Or say I make some stupid business decision. I still can wish that *I* hadn't done that. I can still say, "If only I had not extrapolated from the data, I would have . . .", for even if God uses this for some good reason, I can still regret my stupidity."

      That's fine, but I have the intuition that regret would be improper if the focus of the regret is on the outcome, i.e. the thing that has happened. I can regret having decided something, or having acted in a certain way, but if I think that the event to have happened is better than any alternative having happened, then it doesn't make sense to regret the event's having happened, even though it can make sense to regret having done the act or made the choice that caused the event.

      I've tried a couple of different approaches to argue for the conclusion that it is not incoherent, on libertarianism, to feel regret. (This may be different from Steve's initial point -- Steve may have had in mind the kind of regret that requires acting differently due to favoring different reasons, or something close to this.)

      You're right, we don't "want to get into it all." My interest level in pursuing this any further is threatening zero. But let me just give a final explanation of why I'm taking libertarianism to be able to accommodate regret. If regret can be construed along the lines I've suggested, viz. (a)-(c), then it isn't necessary for the mental representation featuring in the regret to include any content about how the outcome comes to be accomplished. It isn't incoherent for a libertarian to feel regret in this way simply in virtue of the fact that there is no feature that a regret *must have* that involve the outcome being generated in any particular way whatever. I can regret what I've done, but have no content to my thought whatsoever regarding how the preferred outcome should have come about. If outcome regret is possible, that is, if the consequentialist version of regret is possible, then a libertarian can have regrets of this kind without inconsistency. At this point you may have qualms about accepting this consequentialist reading of regret; you may think it's not genuine regret. But my intuition is that it does characterize a genuine kind of regret.

      Delete
  3. Paul, isn’t the libertarian objection analogous to those time-travel paradoxes where, knowing what we do today, if we had a chance to go back in time, we’d do some things differently? But, of course, having the benefit of hindsight would, itself, change the original settings (as it were), so that we wouldn’t really be acting under the same conditions. Isn’t the libertarian objection similarly incoherent? I’d only act differently in retrospect, but that interjects new information into the past situation, in which case it’s not the same situation redux. By inserting my future self into the past, I change the past, and thereby change my future self, in which event it's not the same future self that traveled back into the past...and so on.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
    2. The objector suggests that regret is only intelligible if the person who is experiencing regret over an action could have chosen otherwise than he or she in fact did. But this doesn't require that the desire to have chosen differently involve a desire to have chosen differently *based on the later held reasons.* For example, I could wish that I would've chosen to go to a different college, given the exact same antecedent conditions. This is because my regret may come from not having arrived at a better outcome, and not necessarily from not having arrived at a better outcome *by possessing better reasons, or exhibiting better reasoning, or some such, to lead to the outcome.*

      In other words: I don't care how I could've gotten there, I just wish I would've gotten there.

      Delete
    3. But that's a retrospective viewpoint. Judging the past by the present. The regret involves additional considerations you didn't have at your disposal back then. Had you known otherwise, you would have done otherwise. Changing the initial conditions is what changes the outcome.

      Delete
    4. True, regret is inherently backward-looking.

      But you're assuming that in cases of regret, the representation of the counterfactual in the person's mind always involves the person having knowledge of how the outcomes turn out, and that knowledge being the difference-maker for the generation of the better outcome. But why must this be built into the representation? Why not require the representation to simply include the agent making a different choice irrespective of the reasons why? The point is the different outcome, however it comes about.

      Libertarianism says that there is a world A and a world B, and the only difference between them is that an agent Z chose p in world A, and q in world B. Regret could simply involve wishing that B had obtained rather than A, which is the actual world. Conditions are the same, but choice or outcome are different. The person regrets the choice they've made, but all that is included in the content of their regret is a different outcome. So the regret has these components: (a) feelings of sadness over actual state of affairs; (b) feelings of guilt over actual state of affairs; (c) wish that a different state of affairs had obtained. I don't think the knowledge of experiencing (a) or (b) have to be components in the representation of (c).

      Delete
    5. Berny, why saddle poor libertarians with a "representation to simply include the agent making a different choice irrespective of the reasons why? The point is the different outcome, however it comes about." Suppose we hold *everything* fixed a millisecond before the regretful decision was made. It seems to me we'd require some *radical* indeterminism to do otherwise. Would this ground control and responsibility? Doesn't seem so. So are you saying the libertarian escapes Steve's charge through a freedom and responsibility undermining bringing about of the different outcome? Doesn't that seem like a Pyrrhic victory?

      Delete
    6. Berny Belvedere

      “True, regret is inherently backward-looking. But you're assuming that in cases of regret, the representation of the counterfactual in the person's mind always involves the person having knowledge of how the outcomes turn out, and that knowledge being the difference-maker for the generation of the better outcome. But why must this be built into the representation? Why not require the representation to simply include the agent making a different choice irrespective of the reasons why? The point is the different outcome, however it comes about”.

      I don’t follow your argument. Even libertarians don’t think we may choices in a vacuum. It’s not a coin flip. Take libertarians who define freedom as access to alternate possibilities. To make choices, we must choose from a range of options. We contemplate the perceived advantages and disadvantages of the various options. We take into account the benefits or negative consequences of each option, as best we can anticipate.

      Of course, because we’re shortsighted, our choices have unforeseen consequences. And this can lead to subsequent regret. Looking back, we’re sorry that we didn’t take the other fork in the road.

      Or maybe we make an impetuous choice, on the spur of the moment. And we may live to regret our rash action–precisely because our choice was thoughtless and heedless of the consequences.

      The choices you make depend on the opportunities you have, or imagine you have, as well as the anticipated payoff.

      Delete
  4. Regrets, I've had a few. But then again, too few to mention.

    ReplyDelete
    Replies
    1. Do you regret not having more regrets?

      Delete
  5. Steve, yeah, there looks like an analogy there (though time travel cases are always puzzling). The basic point asks why in the world the libertarian thinks that in order for "regret" to be intelligible, we must believe that we could do otherwise holding fixed *everything* (?) right up to the moment of action or decision or choice? Or is it not that. Does it mean holding everything fixed up until a week before the action? This is all just odd. It seems to me most people think regretting X means that they would change X if they could, i.e., if they could go back with the knowledge they had, they'd have acted differently. And this is consistent with determinism. Or is the idea that some "regrets" X they would *hope* they would have done otherwise holding *everything* fixed, even though they'd most probably (certainly?) *repeat* the error. That just doesn't seem to be how most of us think about these things.

    ReplyDelete
  6. "There is no point feeling regret for something you could not possibly have done otherwise; yet we still feel regret."

    Part of the problem with the question is equivocation in the use of "not possibly have done otherwise." Calvinists think that people could have done otherwise on some hypothesis or other.

    Moreover, a Molinist would say that he *would* not have done otherwise. Why "regret" something you would not have done otherwise than?

    The answer, of course, lies in the fact that one simply wishes that a different outcome had obtained. That's just as consistent with Calvinism as Molinism.

    That said, a Calvinist has greater reason to feel comfort for his regret.

    -TurretinFan

    ReplyDelete
  7. "It is certain that man must utterly despair of his own ability before he is prepared to receive the grace of Christ." Luther, Heidelberg Disputation

    I would say that we must be brought to fully regret unto despair so that we can fully trust in what Christ has already done for us!

    ReplyDelete
  8. BTW, Paul, shoot me an answer to the question I posed in my latest blog post, if you would. It seemed to me that in Craig's case for the existence of God in his debate book with Sinnott-Armstrong, he required that God have agent-causal powers in order to generate the conclusion that God is both a person and free.

    Whenever you get time.

    ReplyDelete
    Replies
    1. Berny, Craig can't believe that, for Craig doesn't believe that *GOD* is a person. :) So does he think the three persons have agent causal powers? Or do most philosophical theologians think he so has? Probably at a pre-intuitive level. But it is of course hard to spell out what, exactly, "agent causation" is. Does God have the agent causal powers in the sense Chisholm described said powers? Why think that? His view faces serious criticism too. Current writers who have written on God's freedom (O'Connor, Timpe) have seen it as something sui generis. So there's not going to be some isomorphic mapping between our agent causal powers and Gods. Finally, it doesn't settle matters between compatibilists and libertarians, for there have been compatibilist accounts of agent causation offered (e.g., Markosian, Nelkin). Not sure if that helps at all, I wasn't entirely clear on what you were asking.

      Delete
  9. For what it's worth, Paul, I think you are absolutely right here. Regret most certainly is "intelligible" under determinism; one can *know* the meaning of the word and feel the qualitative experience of it without affirming libertarianism. Maybe the emotional response of regret is not normatively appropriate (please don't ask me to analyze that, I just made it up while typing this comment) under the belief that God ordains all things for his glory and the good of elect, but so what? Doesn't Jesus died for that? I hope so.

    Cheers,

    Adam

    ReplyDelete
  10. Hey Adam, thanks for the comment. As far as the "normatively appropriate" comment, I think it can be appropriate here if we allow for double causality, of a real sense in which *I* caused something. We should also grant that God's plan is summed as good, but this doesn't mean each individual part is good (in and of itself). I might go further, too, and claim it *is* (at least sometimes) normatively appropriate. God commands us to repent of our sins, and I assume at least some small part of that involves regretting our sins. It is normatively appropriate to do what God commands. It's normatively appropriate to repent. Thus, it's normatively appropriate to regret. But I know you weren't taking a hard stand on this (as prefaced by 'maybe'), but I just wanted to throw that out there. :)

    ReplyDelete