Puddleglum’s Wager
October 20, 2009 — 10:10

Author: Jeremy Pierce  Category: Afterlife Atheism & Agnosticism Existence of God Religious Belief  Comments: 41

We’ve been listening to C.S. Lewis’ Narnia Chronicles on CD. I read them when I was about ten years old, and I never got around to re-reading them, so some of it is almost as if I’m experiencing them for the first time. When I got to the following scene from the Silver Chair, it struck me as a strange argument, sort of like Pascal’s Wager, but something rubbed me the wrong way about it. The main characters were in the Green Witch’s underground domain and had fallen under her influence, which was causing them to lose their belief in the above-ground world. Puddleglum the marsh-wiggle then gives the following speech:

Suppose we have only dreamed, or made up, all those things-trees and grass and sun and moon and stars and Aslan himself. Suppose we have. Then all I can say is that, in that case, the made-up things seem a good deal more important than the real ones. Suppose this black pit of a kingdom of yours is the only world. Well, it strikes me as a pretty poor one. And that’s a funny thing, when you come to think of it. We’re just babies making up a game, if you’re right. But four babies playing a game can make a play-world which licks your real world hollow. That’s why I’m going to stand by the play world. I’m on Aslan’s side even if there isn’t any Aslan to lead it. I’m going to live as like a Narnian as I can even if there isn’t any Narnia. So, thanking you kindly for our supper, if these two gentlemen and the young lady are ready, we’re leaving your court at once and setting out in the dark to spend our lives looking for Overland. Not that our lives will be very long, I should think; but that’s a small loss if the world’s as dull a place as you say.

What rubbed me the wrong way was that it sounded as if he didn’t care whether the world was real. He was going to believe in it anyway, because it’s more pleasant to believe in it. How can the upper world be so much better than the underground world that its mere finite value of being better would be worth believing in a lie if it’s not true?
When I raised this issue with a friend, he said, “But it’s Pascal’s Wager!” I said, “No, it’s not!” He insisted that the upper world is Aslan’s world, which I’d been thinking of as the place at the end of the world that they went to in the previous book, and the upper world was just Narnia, which is the analogue of Earth. But we were interrupted and never managed to finish the conversation.
I realized later, when teaching Pascal’s Wager, what Lewis must have been up to, and it’s actually a neat trick. If he was seeing Narnia as a placeholder for the eternal reward of Pascal’s Wager and the underworld as a placeholder for this life, then you have an interesting argument that isn’t quite Pascal’s Wager. Pascal’s Wager concedes for the sake of argument that life in this world is more pleasant if you don’t believe in God but then argues that the chance of eternal reward in heaven compensates for that in terms of rational decision theory. You shouldn’t even need 50% likelihood of God’s existence for the wager to be worth it given that the reward is infinite and the cost merely finite if you bet wrong. But Lewis’ Wager is different in exactly one way: it doesn’t make the concession. It takes the finite value of life in this world to be better if you believe in God than if you don’t. So life is finitely better if you believe in God, and the afterlife is infinitely better if it turns out there is one. Therefore, it’s a no-brainer. You might as well believe in God. If it turns out you lose the bet (i.e. God doesn’t exist), you still end up finitely better off, and if you win (i.e. God does exist) then you get an infinitely better result.
One interesting result of Puddleglum’s Wager is that it easily avoids the problem Mike Almeida raises against Pascal’s Wager. Mike’s problem (which I’m not taking a stand on at this point) relies on its being better in this life not to believe.
[cross-posted at Parableman]

Comments:
  • Apparently, Lewis intended Puddleglum’s argument to be an ontological argument akin to Anselm’s and Descartes’s. Victor Reppert has a portion of one of C. S. Lewis’s letters that gives us the author’s intent behind Puddleglum’s argument. I have to admit, I didn’t see an ontological argument in Puddleglum’s argument when I re-read it about a year ago, but I guess that just shows that there’s more to Lewis’s children stories than I realized.

    October 20, 2009 — 10:48
  • Jeremy Pierce

    Or perhaps it just means that Lewis tried but was unsuccessful in capturing that argument in Puddleglum’s speech.
    It’s not Anselm’s argument, by the way. It’s a different one from Descartes. Look at the comments in that thread for an explicit statement of it.

    October 20, 2009 — 11:14
  • Mike Almeida

    It takes the finite value of life in this world to be better if you believe in God than if you don’t.
    Jeremy,
    What you are decribing (it seems) is William James’ Wager, which is a lot like Pascal’s, but adds that belief pays in this world, too. As it happens, Jeff Jordan also defends James’ Wager in OUP 2006 which he thinks better solves the many-gods objection.
    Maybe this sounds strange, but this quote I’m going to live as like a Narnian as I can even if there isn’t any Narnia struck me as Kantian more than anything else. There might be no kingdom of ends, but if there isn’t, there isn’t anything morally valuable either. Better to act as if we live in the moral world, since it is only such a world that would confer value on what we do.

    October 20, 2009 — 12:18
  • Here’s an argument that might be what Lewis is thinking of when he talks of his argument as inspired by the ontological argument:
    1. If ~Narnia, then fiction > reality.
    2. ~(fiction > reality)
    3. Therefore, Narnia.
    But the text is richer than this argument, as the text also has the idea that if fiction > reality, then it is better to live by the fiction.

    October 20, 2009 — 14:17
  • There were various attempts in the early and middle twentieth century to interpret the ontological argument as performative or quasi-performative, so it might be possible to regard the Puddleglum argument as an attempt to paraphrase the argument along such lines.
    I find the Pascal’s Wager angle interesting, though, because I think the way you laid it out makes it actually quite a bit closer to the way Pascal himself builds the Wager argument, despite the divergences from the decision-theoretical versions that have become popular since: Pascal’s notes on the Wager are actually broken up into different parts, only some of which deal with eternal reward in heaven, and the decision-theoretical interpretations consolidate these parts into a single stage. But it’s also possible to see the notes on the Wager as the beginning of a multi-stage argument, each part of which is dealing with a different kind of objection; and the Puddleglum argument seems to me to fit this fairly well.
    And Mike is right that it ends up looking a lot like Kant as well. Kant himself uses a wager analogy to talk about what he is trying to do (around CPR A825, I think).

    October 20, 2009 — 14:40
  • Jeremy Pierce

    Alex, the text might also be less rich than that argument if there’s nothing in the text that suggests that fiction can’t be better than reality other than saying it’s a funny thing. But I took that to be just the irony of the play game being better than the real world, not a suggestion that it’s not just odd but impossible.
    Brandon, there is one key difference. What Kant thinks he’d be betting if he stopped believing in God is morality itself, and that’s not something he’s willing to give up. It’s a pragmatic betting argument, but it’s nothing like the afterlife-based argument of Pascal.
    Also, I think it should be clear that Puddleglum’s argument isn’t really a wager at all. There’s nothing he’s giving up to make the bet, as far as he’s concerned. He’s not risking anything, if it’s win-win.

    October 20, 2009 — 16:05
  • That is true if we take the common decision-theoretical interpretation, which consolidates all the parts of Pascal’s notes on the Wager into a single decision matrix. But, as I said, there is an alternative interpretation in which the parts are each supposed to be doing something different — in which case the argument is not afterlife-based, because immortality is only brought in to deal with a particular agnostic objection (that, since the Wager involves wagering one’s life, we are wagering too much), and even with this Pascal also points to the possibility that belief can contribute to becoming “faithful, humble, grateful, generous, a sincere friend, truthful”. But certainly Pascal’s argument (like Puddleglum’s) is not transcendental — he isn’t arguing for theism as a practically necessary postulate, but simply denying the claim that there is no good reason to believe.
    And it actually seems to me to be here that the real difference between the three makes itself felt. Kant is angling for a practically necessary conclusion; Pascal is simply arguing that the agnostic has no good objection against believers; and Puddleglum is saying that the life of Narnia and Aslan is a superior life to one without, even if there is no Narnia and Aslan. They are all three arguments based on practical rather than theoretical reason, which is no doubt why people’s minds tend to put them together: they really are cousins. And while Puddleglum is not making a wager, unlike you I’m inclined to think it could easily be put in such form — betting on Aslan and Narnia is more rewarding even if there is no Aslan and Narnia. Pascal makes a similar point, although briefly. But they are being deployed for very different purposes in the face of very different interlocutors. The interlocutor in Pascal’s notes is not someone who denies that God exists, but merely denies that Christians are reasonable in believing He does; but Puddleglum is facing a persuasive enchantress denying the very existence of Aslan and Narnia.

    October 20, 2009 — 22:17
  • Jeremy Pierce

    Incidentally, that Kant passage has an odd admission. Apparently Kant would have wagered quite a lot on the claim that there is alien life on one of the planets visible to us. Was such a belief common in his day? I wouldn’t have thought so.

    October 20, 2009 — 22:31
  • DL

    Of course, Pascal also wanted people to “test-drive” Christianity because living it would, in time, become a more convincing argument for it than any logic. Similarly, Puddleglum isn’t just choosing to sit there and believe in Narnia, but he sets out looking for it. If he didn’t believe in it, he wouldn’t go looking for it, and thus he’d never find it even if it did exist.
    The ontological side struck me first. If you cannot get something for nothing, and if the dream-world has something that you can’t get from the dismal underworld, then the “dream” world, or something even more real must exist somehow, somewhere.
    But specifically to address the point that Puddleglum seemed prepared to believe in something without caring whether it was actually true, we need to know: why believe in reality? Why should we value true beliefs higher than false ones? However Lewis’s argument works, it has something to do with the proposed “truth” not being worth believing over the “dreams”. If the truth were dingy and pathetic, maybe we shouldn’t want to believe it.
    I’ve always thought this was an interesting argument for theists to run against atheists: if God is associated with reality (Christianity specifically identifies God as the Truth), then there are reasons why we should seek truth. But if reality is only atoms spinning in the void, then so what? If “truth” is an accident, why should anyone blame a person who doesn’t feel it’s worthwhile?
    (This also ties in to idea of how we could ever know what’s true, if everything including our brains are accidental, etc.)

    October 22, 2009 — 1:08
  • Sébastien Réhault

    “Pascal’s Wager concedes for the sake of argument that life in this world is more pleasant if you don’t believe in God”
    I am not sure it’s a proper construal of Pascal’s argument. Pascal does in fact argues that life is better for the believer event if it happens God does not exist. In that respect Pascal’s wager is very close to James’s wager in “The Will to Believe”.
    See what Pascal says about the benefits of belief in God if God doesn’t exist: “But what harm will come to you from taking this course? You will be faithful, honest, humble, grateful, doing good, a sincere and true friend. It is of course true; you will not take part in corrupt pleasure, in glory, in the pleasures of high living. But will you not have others? I tell you you will win thereby in this life”. (Pensées, 680, Lafuma edition).

    October 22, 2009 — 2:46
  • Mike Almeida

    “But what harm will come to you from taking this course? You will be faithful, honest, humble, grateful, doing good, a sincere and true friend. It is of course true; you will not take part in corrupt pleasure, in glory, in the pleasures of high living. But will you not have others? I tell you you will win thereby in this life”. (Pensées, 680, Lafuma edition).
    There are several versions of Pascal’s Wager, but that’s one that no one seems to advance. It’s effectively the argument that belief dominates unbelief. But having a seriously mistaken metaphysical view is, for lots of people, a source of considerable disutility (i.e. dissatisfaction), whether they know they’re mistaken or not. To run this version of Pascal’s argument you’d likely have to treat utility as something other than preference-satisfaction; as something like Millian pleasure or happiness.

    October 22, 2009 — 9:52
  • Mike Almeida

    But if reality is only atoms spinning in the void, then so what? If “truth” is an accident, why should anyone blame a person who doesn’t feel it’s worthwhile?
    This is where practical reasons pulls apart (or might pull apart) from theoretical reason. But there is something repugnant about ducking what’s true.

    October 22, 2009 — 9:57
  • DL

    “But there is something repugnant about ducking what’s true.”

    I would put it like this: is it better to be good, or to be right? Judeochristian theology answers: (a) it’s better to be good, and (b) but don’t worry because the Good and the True are the same thing (or rather, the same Person). I think that rejecting the truth seems to repugnant to us precisely because reality is such that to do that means you are also rejecting the Good.
    But to an atheist (or perhaps even to a different kind of theist?), why should Duty be Truth, or Truth Duty? Is that something we can know on a random earth? Or is it just a few millennia of cultural baggage? Atheists face certain issues trying to explain morality without God, but I don’t think similar problems apply to explaining truth. The problem is that once divorced, why should we care?
    Or to put it the other way around, is the fact that everyone everywhere seems to share this love of truth really just a learned prejudice, or is it strong evidence that something more is going on?

    October 22, 2009 — 10:24
  • Mike Almeida

    Or to put it the other way around, is the fact that everyone everywhere seems to share this love of truth really just a learned prejudice, or is it strong evidence that something more is going on?
    I’d deny that you have to be a Christian or theist to care about what’s true. I’d deny as well that there is any decent inference from atheism to some form of relativism, moral or otherwise. There are, for one example, extremely good arguments from the rationality of cooperation to compliance in widely recognized moral principles according to which truth, among other things, is valuable. You don’t need God, or anything like God, for that.

    October 22, 2009 — 11:32
  • DL

    I’d deny that you have to be a Christian or theist to care about what’s true.

    Certainly, anyone can seek truth; and you can care about it as a personal preference; but how can it be morally incumbent all by itself? It may be valuable towards some end, e.g. co-operation — or maybe not, if everyone else is equally ignorant. (I suppose in that case, the truth might make it easier to cheat, but that is presumably more repugnant than being uninterested in the truth like everyone else.)

    October 23, 2009 — 0:08
  • Mike Almeida

    Certainly, anyone can seek truth; and you can care about it as a personal preference; but how can it be morally incumbent all by itself?
    I guess I’m not sure what you mean. If the truth is morally incumbant all by itself–I take it you mean that it is an intrinsic property of ‘the property of truth’ that we ought to care about it–then it has that property whether or not theism is true. If it is not an intrinsic property, then it might well be something we care about for any number of theistic or non-theistic reasons. Either way, it’s having the property intrinsically or not does not depend on theism being true.

    October 23, 2009 — 9:43
  • Mike:
    I wonder if the distinction in DL’s argument is not so much between theism and atheism but between non-materialism and materialism? If all is atoms and the void, it is very puzzling how some segments of the atoms-and-void could have something be encumbent upon them–it seems, at least prima facie, that everything would just be as it is, and there would be no “shoulds”. But bringing in God does not by itself do anything to alleviate this. It seems just as hard to explain how a cloud of particles or a bump in a field could be obliged to do what God commands as it would be to explain how a cloud of particles or a bump in a field could be obliged to pursue truth in the absence of God.
    But if the distinction is between non-materialism and materialism, I think there really is a plausibility in the idea that the pursuit of truth wouldn’t be encumbent on it if materialism is true. And it might be that non-materialism explanatorily requires something like God in a way that is more obvious from the way materialism does (parallel: design arguments are more obvious to people than cosmological ones).

    October 23, 2009 — 10:19
  • Heath White

    We are getting off topic, but I have been thinking about something related to Alex’s point for a long time. There is a strong intuition that no clump of particles has a duty to do anything. But adding God to the mix doesn’t seem to change that. Question: what would change it? Under what circumstances would something have a duty to care about the truth? Alex suggests that something non-material would make a difference, but I don’t really see why.
    One might say that one has a duty to one’s happiness (or good), and with God on the scene, one’s happiness or good is bound up with (caring about) truth. But then, if everything so far is to be consistent, one is committed to the view that no clumps of particles have happinesses or goods. It is hard to see how adding God to the mix would change that either. And again, I don’t see that adding a different kind of substance to the mix fundamentally alters the situation.
    This leads me to think that the initial intuition is wrong. But I’m not sure that’s right either.

    October 23, 2009 — 12:51
  • Jeremy Pierce

    Just in terms of Lewis exegesis, I think it’s important to bear in mind that he did think morality is hard to explain without God and especially bad to abandon just because you’ve stopped believing in God, so I think he’d be fond of a Kantian transcendental argument. That may be why he’d find a life of belief in God to be better than a life without it, because it allows you to have a foundation for something that it would be horrible to get rid of but harder to explain in a world without purpose.
    As for the philosophical issue, I don’t think it’s either mere theism vs. atheism or mere materialism vs. non-materialism. I think it’s teleology that makes the difference. If we’re created for a purpose to serve the aims that God intended for us, and those purposes have shaped us in such a way that what’s good for us is aligned with our moral obligations, then there’s a much deeper foundation for ethics than you can provide with theories that don’t have such a thing.
    You either get an error theory or fictionalism about morality, where moral claims are just false, or you get a less robust view than most of us hold, where the truth-makers for moral claims are only facts about what is enjoyable, what we might rationally do given the aim(s) of survival and/or pleasure, or the brute facts about what we happen to prefer (with no deeper notion of what we should prefer).
    All of those accounts seem thoroughly inadequate to me, but I’m not sure I’ve seen a meta-ethical foundation for morality that goes much beyond that into the realm of those deeper moral truths that the transcendental argument relies on that would be awful to give up.

    October 23, 2009 — 12:59
  • DL

    I wonder if the distinction in DL’s argument is not so much between theism and atheism but between non-materialism and materialism?


    You’re right, that is a better distinction. You can care about truth for any number of reasons, but to be justifiably upset at my not caring depends on something bigger than both of us. (You might consider me foolish for neglecting something useful for various ends, but you’ve no reason to be outraged unless I’m under some higher obligation.)
    I also agree not just any deity would do. If God isn’t interested in us (let alone interested in whether we’re interested in him), then I don’t see a way to make truth obligatory. However, given the Christian God, say, to whom we are obligated (however that may work), then (respect/gratitude/obedience/etc. to God) + (God=Truth) —> (obligation to seek/respect/acknowledge truth).

    And it might be that non-materialism explanatorily requires something like God in a way that is more obvious from the way materialism does


    Yes. Of course, this idea was prompted by typical arguments between typical materialst atheists and typical western theists: if there really does exist the Judeochristian God (or a deity sufficiently similar in the right kind of way), then the theist has an easy way to justify caring about the truth, and therefore ought to belive in God. However, if no gods exist, the theist could still choose to believe in God. It’s not clear to me how the naturalist could object that one shouldn’t believe in falsehoods.
    But I guess this is just a specific example of an argument from morality. (How can the naturalist explain an authority binding on all of us apart from what rules we individually choose to follow, or explain a morally free will.) Of course, some materialists are happy to say that there are no such obligations, but conveniently it works out well for everyone to co-operate and not kill, steal, etc. anyway. Such a person wouldn’t have a reason why you shouldn’t believe in God, even if not true. (But perhaps such a person wouldn’t care anyway, as long as whatever you believed led you to co-operate in society.)
    There is also a teleological aspect: if our minds were made in order to know truth (again, God is an obvious but non-exclusive explanation for that), then not caring about the truth would be an abuse of your mind. However, if our minds are accidental, a naturalist couldn’t object to using them to believe falsehoods. (If indeed, there is a way to make a materialist epistemology work.)

    October 23, 2009 — 13:48
  • Heath White

    As for the philosophical issue, I don’t think it’s either mere theism vs. atheism or mere materialism vs. non-materialism. I think it’s teleology that makes the difference. If we’re created for a purpose to serve the aims that God intended for us, and those purposes have shaped us in such a way that what’s good for us is aligned with our moral obligations, then there’s a much deeper foundation for ethics than you can provide with theories that don’t have such a thing.
    I think this, and the comments that follow, are very helpful.

    October 23, 2009 — 21:09
  • Gordon Knight

    Jeremy,
    You wrote that without divinely grounded teleology:
    “You either get an error theory or fictionalism about morality, where moral claims are just false, or you get a less robust view than most of us hold, where the truth-makers for moral claims are only facts about what is enjoyable, what we might rationally do given the aim(s) of survival and/or pleasure, or the brute facts about what we happen to prefer (with no deeper notion of what we should prefer).”
    I don’t think this is true. First, there is the alternative of Aristotelian teleology–why beg the question and suppose there cannot be teleology in nature w/o God?
    And then there is the wide variety of moral realisms that do not depend on God (e.g Plato, Moore, Prichard, Ross)

    October 24, 2009 — 9:22
  • Mike Almeida

    .. . if there really does exist the Judeochristian God (or a deity sufficiently similar in the right kind of way), then the theist has an easy way to justify caring about the truth, and therefore ought to belive in God
    I don’t see that the theist has any better reason to care about the truth than the non-theist. What is confused in these observations, I think, is exactly which counterfactual situation we are considering. In considering a world in which God does not exist (I’d call the relevant conditional a counterepistemic, others would call it a non-trivial counterpossible, but the atheist would call it a mere subjunctive), we are not considering a world in which there are no values. We are considering a world in which there are values, but they are metaphysically different from (assuming theism for the moment) what they actually are. Perhaps in (conceivable) Godless worlds, values–such as the value of truth–are best understood in terms of informed and laundered preferences. I find such analyses very promising. Why do people–whether they turn out to be souls or arrangements of atoms or organisms or whatnot–value truth? One obvious reason is that we (we human organisms, I mean) together intersect in our preferences for the huge cooperative benefits of an institution of truth-telling and truth-pursuing. This approach gives a very nice, parsimonious explanation of why we value truth. It is not that God commands that we seek truth or that he instills a desire for truth, or anything of the kind. It is a matter of its obvious benefits, whether or not God exists.

    October 25, 2009 — 14:17
  • I, too, think that teleology is at least a large part of what makes the difference. However, in order to make the difference, it has to be an intrinsic Aristotelian teleology, not an extrinsic teleology constituted by the designer’s intentions. Imagine two purely material beings, A and B. A arises at random when lightning strikes a swamp. B is made by a smart and good designer to live a human-type life. As it happens, A and B are molecule-by-molecule exactly alike. It seems to me that the extrinsic historical fact that B was made to to live a human-type life is not enough by itself to render B’s life meaningful, a life that ought to be responsive to duties, reasons and goods. (Actually, essentiality of origins suggests that we need to be cautious about describing the causal history as “extrinsic”. Nonetheless, there is an important sense, I think, in which it is extrinsic.)
    Now, once we are dealing with an intrinsic teleology, then we’ve gone beyond atoms and the void. Maybe some sort of “materialism” is compatible with an intrinsic teleology, but it won’t be a naturalistic or physicalistic materialism in the modern sense.
    Mike:
    I see no reason to care about my preferences, even sanitized ones (and the accounts of the sanitization all really do fail, unless they build in something more primitively normative), unless these preferences line up with something that is good or unless it is good to fulfill preferences. Moreover, since on a naturalistic view, facts about the good as such are explanatorily irrelevant, they are not even partly explanatory of our beliefs about the good. But, if so, then (a) naturalistic theories of reference the most promising of which are causal theories of reference aren’t going to yield successful reference to the good, and (b) if our beliefs about the good happen to be right, that’s a mere coincidence.

    October 25, 2009 — 22:59
  • DL

    This approach gives a very nice, parsimonious explanation of why we value truth.

    But my concern wasn’t why we value truth. I value money — it too is a very useful thing, it enables many co-operative benefits — but if you told me you were forsaking all your wordly goods to live in the desert as a hermit, I wouldn’t be disgusted. In fact, I’d probably admire you in some ways. Yet if you told me you were going to forsake the truth, I would share the common reaction that there is something repugnant about that. Since I can’t see a naturalistic justification for that, either the universe is more than matter, or else we should stop blaming people if they give up truth.
    (I suppose you could have a reason to care whether I care about truth on the grounds that if I didn’t, I can’t co-operate with you as well as I might, and that puts you at a disadvantage; but the fact that you don’t like it doesn’t by itself put a moral obligation on me to start caring about the truth. I think most people would say that our natural discomfort at truth-rejecting is something that should apply even if you lived in isolation. That is, it’s immoral itself, not merely impolite or unsociable.)

    October 26, 2009 — 10:28
  • Mike Almeida

    Moreover, since on a naturalistic view, facts about the good as such are explanatorily irrelevant, they are not even partly explanatory of our beliefs about the good.
    As far as I can see, that’s either false or question-begging. On naturalistic views, what is good JUST IS what is preferable (or preferable under certain conditions). So the good as such JUST IS the preferable as such. Again, in the relevant worlds, you have to make the right metaphysical changes. Talk about the good as such as though it were obviously different from the preferred plainly begs the question against the naturalist. It is clearly part of what is at issue.
    I see no reason to care about my preferences, even sanitized ones (and the accounts of the sanitization all really do fail, unless they build in something more primitively normative), unless these preferences line up with something that is good or unless it is good to fulfill preferences
    To say that you do not care about your preferences (I take it you mean that you do not care about the object of your preferences) is just to say that you do not care about what is good. For that just is what preferences are, according to this metaphysical view. I see no big worries about having such an attitude, since there is the question (most famously put in Hobbes and Hume) of why anyone should (rationally should, that is) pursue what is good or right, no matter what view of value one endorses.
    It is just wildly overstated to assert that all of these theories fail. They all have problems, sure, but no interesting philosophical view does not have problems. No need to remind anyone that ‘intrinsic Aristotelian teleology’ is not exactly problem free, for instance.

    October 26, 2009 — 11:12
  • Mike:
    First of all, it’s not that I don’t care about the objects of my preference–that would be a self-contradiction–but that I don’t see reason to care for them just because they are preferred.
    “On naturalistic views, what is good JUST IS what is preferable (or preferable under certain conditions).”
    If “x is preferable” means “x should be preferred”–and that seems to be the usual meaning of the term–then the naturalist has a problem with that, because of the “should”. If “x is preferable” means “x is preferred”, then we have a lot of nasty problems. Sure, every theory has problems, but this one seems to be particularly problematic. (In my thinking about this, I am much influenced by William Lauinger’s recent dissertation at Georgetown.)
    Actual preference theories fail to be extensionally correct in cases of pathological preferences, such as preferences in the light of false non-normative information, or preferences in persons whose rationality was impaired in the process of forming the preferences (e.g., preferences formed due to mental illness or brainwashing). This requires a move to hypothetical preference theories on which the good for x is that which x would prefer if x were to go through some procedure P, such as cognitive psychotherapy or being informed of all non-normative facts. But those approaches have a whole slew of problems:
    1. They have all the standard problems of the “conditional fallacy” that face counterfactual accounts of just about anything. I think there is good reason to reject counterfactual accounts of just about anything. (Simplest case: Imagine that your brain has been rewired by aliens so that as soon as you go through cognitive psychotherapy or are informed of all non-normative facts, you start preferring the eating of slugs to everything else.)
    2. Defining the procedure P without making use of normative or teleological concepts the naturalist is not entitled to is dicey. For instance, cognitive psychotherapy is as a procedure designed to return the mind to normal functioning. I doubt that a naturalist account of this normalcy is forthcoming–it is part of the constellation of teleological concepts.
    3. Merely being informed of all non-normative facts is not enough to dislodge deep-seated preferences that were formed over many years on the basis of false non-normative beliefs. One needs a condition of being “vividly” informed. But once we are talking of being “vividly” informed, the results all depend on the order and method of presentation. For instance, I can vary the methods of vivid presentation of the details of (a) a particular crime and (b) the criminal’s nasty upbringing, and I can do so in such a way as to render it highly probable that you will prefer to forgive the criminal or that you will prefer to have the criminal severely punished. The kind of vividness the presentation must have is a normatively sensitive vividness–it must make appropriately vivid the salient features of the situation. Again, we go back to the need for normativity.
    4. Both actual and hypothetical preference theories have nasty problems with diachronic issues. At age 20, when fully sanitized, you would prefer that at age 40 it should be the case that p. At age 35, when fully sanitized, you would prefer that at age 40 it should be the case that not-p. Which of these is really good for you, then? (Lauinger really nicely presses this issue in his dissertation.)
    5. It may very well be that were I to undergo P, I would be a very different kind of person. But what motivates preference theories is the idea that what is good for me should be somehow grounded in my preferences. But if the P-sanitized person were to have radically different preferences–as might very well be the case–then this motivation for the preference theory is undercut. (Lauinger makes this point very well in his dissertation.) I might as well adopt Aristotle’s theory that I ought to prefer that which the phronimos prefers.
    6. There are miscellaneous little technical problems with hypothetical theories, such as that one might want to find out whether the Riemann Zeta Conjecture is true, but in a situation where one is fully informed of all non-normative matters, one would have no desire to find it out, because one would already know it.
    7. In a radically indeterministic world, unless Molinism is true (and I think it’s not), we don’t really have much in the way of binary counterfactuals. What we have are distributions of conditional probabilities. This means that instead of saying that the good for me is what I would prefer were I to undergo P, I’d have to take into account the probabilities. For instance, maybe the conditional probability of my wanting to remain a philosopher if I given cognitive psychotherapy is 0.8, but the conditional probability of wanting to become a full-time embedded-systems computer programmer on the same condition is 0.2. Unless we’re going to posit an arbitrary probability cut-off, the way to handle this is by calculating utilities using weights coming from these conditionals. If so, then the theory inherits all the standard problems with assigning numbers to preferences.
    “there is the question (most famously put in Hobbes and Hume) of why anyone should (rationally should, that is) pursue what is good or right”
    I am afraid I don’t understand the question… I can’t think myself into seeing a distance between “should” and “good or right”. I know that for some people, inserting the term “rationally” helps to create such a distance, but not for me.

    October 26, 2009 — 11:58
  • Let me add a further point about 7. On the naturalist account, what is good for me surely supervenes on the state of my mind, the physical constitution of the world, etc. But if we take the Molinist way out of the problem in 7, then the supervenience will fail, because the truth-values of Molinist counterfactuals do not thus supervene (and indeed some of them surely go against what you would expect from the state of mind and physical constitution of the world). So Molinism won’t help the preference theorist.

    October 26, 2009 — 12:08
  • Mike Almeida

    This probably could have been put in a paragraph of two. But I’ll wade through to make my point.
    If “x is preferable” means “x should be preferred”–and that seems to be the usual meaning of the term–then the naturalist has a problem with that, because of the “should”. If “x is preferable” means “x is preferred”, then we have a lot of nasty problems.
    I said nothing at all about meaning, so I don’t know what this is about. This issue is not a semantic one, it is a metaphysical one.
    Re (1): Your simple example describes a world that is just not anywhere near the closest world in which you consider your preferences under the relevant conditions.
    Of course there are problems with the hypothetical worlds under which you consider what you prefer, but most of the alleged problems are too artificial to be worrisome for a naturalist. It is a little like raising Cartesian skeptical issues with a naturalist psychologist. Those worlds are irrelevant.
    Re (2): I said nothing about cognitive psychotherapy. Nor was I talking about cases of those who are in various ways pathological. Why would I test the adequacy of a theory by how well it works under assumptions of such pathology? We have good scientific ways of determining the presence of pathology, and I concede that the preference of such individuals are not in general relevant to what has value.
    Defining the procedure P without making use of normative or teleological concepts the naturalist is not entitled to is dicey.
    The naturalist is perfectly entitled to teleological concepts (why would anyone think not??). There are perfectly good sense in which someone is healthy and functioning well psychologically that naturalists can help themselve to.
    Re (3): I’m not claiming that the theory applies to the preferences of schizophrenics. Obviously, there are cases of preferences that simply cannot be cleansed and so are irrelevant. To repeat, the theory obviously applies to the wide range of more or less psychologically healthy human beings. Such a stipulation is perfectly fine for naturalists to make, since we have perfectly good naturalistic methods for detecting the psychologically unhealthy.
    Re (4): Current preferences are what matter in any case of conflict on any such view that I’ve seen.
    Re (6): Essentially repeats (1).
    Re (7): The problem you describe in indeterministic worlds is a problem for current analyses of counterfactuals. It is not a problem for the truth of the relevant counterfactuals.

    October 26, 2009 — 12:54
  • Mike:
    “There are perfectly good sense in which someone is healthy and functioning well psychologically that naturalists can help themselve to.”
    What is that sense?

    October 27, 2009 — 11:37
  • Mike Almeida

    “There are perfectly good sense in which someone is healthy and functioning well psychologically that naturalists can help themselve to.”
    What is that sense?
    Go to the DSM IV for the current list of psychological disorders that inhibit functioning well psychologically. Psychological well-being is understood in terms of exhibiting certain natural properties relevant to psychological health: properties contributing to happiness, adjustment, etc. Note I don’t say that what is psychologically healthy is morally good, though that is often the case.

    October 27, 2009 — 14:07
  • DL

    Yes, naturalists can help themselves to the DSM IV and so on, but I would claim that they are ultimately helping themselves to non-naturalist ideas and not realising it. For a similar example, naturalists can help themselves to the Ten Commandments, but I still claim the commandments come from God. The fact that you can follow them without believing such obviously doesn’t indicate that they didn’t come from God.
    (I am thinking specifically of following the rules set forth in Genesis without believing that Moses literally received stone tablets from God, but the analogy works just as well if you think of it in terms of God as being the only source of morality — the possibility of your actually acting morally might require God to exist without requiring you to believe that God exists.)

    October 27, 2009 — 16:39
  • DL

    Oops, I replied to Mike Almeida’s reply a few days ago, but apparently it didn’t go through….

    This approach gives a very nice, parsimonious explanation of why we value truth.

    But my concern wasn’t why we value truth. I value money — it too is a very useful thing, it enables many co-operative benefits — but if you told me you were forsaking all your wordly goods to live in the desert as a hermit, I wouldn’t be disgusted. In fact, I’d probably admire you in some ways. Yet if you told me you were going to forsake the truth, I would share the common reaction that there is something repugnant about that. Since I can’t see a naturalistic justification for that, either the universe is more than matter, or else we should stop blaming people if they give up truth.
    Now if I see where you’re going, I think you would reply along the lines that if we assume there is a naturally-based morality (wherever it comes from — I remain unconvinced that such a thing is possible, but let’s assume it for the sake of argument), then pursuing truth will help me to achieve that morality, and thus truth can be morally mandated as a means to moral ends (though maybe not for its own sake).
    In that case, let’s consider a world where I believe in God (falsely) but end up coming to the same moral laws that you do (by following the truth). In other words, my untrue beliefs happen to matter only in non-moral issues, so in a practical sense, my false beliefs do not hinder my acting morally. That leaves no ground to obligate me to seek the truth in those other areas. However, I think our common instinct is that we ought to desire truth even in cases where it has no practical effect (or at least no negative practical effect). Again, one simple answer is to claim that that is a bad instinct, and we simply shouldn’t care. An equally simple but opposite answer is available to Christians: that since God is the ground of both morality and truth, they are inseparable from each other. There may be other possibilities in between.

    October 27, 2009 — 18:14
  • Puddleglum (Lewis) is using the ‘my ability to have faith in a better thing proves there’s a better thing that I can have faith in,’ enlarged by ‘even if I’m only dreaming the fact that I can dream bigger must mean that there’s something to the dream that’s more real than the reality I know.’
    C.S. Lewis says in the “Abolition of Man,” I believe, that imagination is the precursor to faith, because in order to believe in something one must be able to imagine believing in that something first. However, this goes against most conservative Christianity’s adherence to a line-in-the-sand argument for total depravity.
    Can you just imagine the Christian Right running around telling people to “Pretend Jesus is real!” instead of “Believe now or be damned!” Just not as catchy, I guess.
    None of this equating why one should believe in acting morally, that’s another topic.
    I think that one’s ability to have faith in something, or the belief a human can have faith in something, begs the question naturally, “Should I then believe?” And “if I do, then what does that look like on the outside?”
    How then does one end up believing in Jesus vs something else? Well, C.S. Lewis wrote the story of the Telmarine who makes it into the reality behind the real at Narnia’s end where all the worlds are united. You may read that for yourself.
    Thank you. I’m not an expert, but really enjoy the discussion.

    October 27, 2009 — 22:56
  • Mike Almeida

    Yes, naturalists can help themselves to the DSM IV and so on, but I would claim that they are ultimately helping themselves to non-naturalist ideas and not realising it.
    Right, and naturalists would claim that you are unknowningly ultimately helping yourself to naturalistic values when you believe the ten commandments. Neither of these observations furthers discussion.
    The question is whether a naturalist has a moral reason to value the truth. I’ve given a reason that is based on the most promising approach to morality–the social contract approach–that a secularist has available. They do offer a plausible account of the origin and content of morality in non-god-worlds, and that account explains why we would value truth in such worlds. Take the quasi-social contract approach in Harsanyi in which, he claims, some form of consequentialism would be the outcome of the social contract. I doubt that such a principle would be the outcome, but even here consequentialists have good moral reason to value the truth.

    October 28, 2009 — 9:58
  • Jeremy Pierce

    DL, your comments are getting through. One of your last few comments appeared twice, but I only approved one copy of it. Another appeared five times. Just because a comment doesn’t appear immediately doesn’t mean it wasn’t received. Unless you log in to comment, your comment gets held for approval.
    Tasiyagnunpa, do you mean the Calormene called Emeth who appears at the end of The Last Battle? Presumably quite a number of Telmarines served Aslan explicitly, since many of them remained in Narnia at the end of Prince Caspian. It’s almost certain that Caspian and Rilian make it into Aslan’s country, and they’re Telmarines.
    As for total depravity, I’m not sure how that prevents anything. The traditional view of total depravity may imply that no one can believe without a work of divine grace, but it’s not as it those who accept total depravity deny God’s grace operating among those who will eventually believe.

    October 28, 2009 — 12:57
  • DL

    Unless you log in to comment, your comment gets held for approval.

    Ah, thank you. (After one comment appeared and the other didn’t, I guess I got carried away!)

    The question is whether a naturalist has a moral reason to value the truth.

    Yes, and I accept that, insofar as there are truths that affect moral principles. If I go beyond that — say by believing in a religion that shares the same morality — there can’t be a moral objection to my extra, untrue beliefs. Now many people would in fact not care; it’s no skin off their noses if I want to believe in God, as long as I don’t murder, steal, etc. I can even see a quite plausible naturalistic explanation of why we would feel disturbed at someone’s rejection of truth in any part: because normal people (as opposed to philosophers or mathematicians or quantum physicists) believe things that affect day-to-day activities, we could develop a moral attachment to truth in general. By the time we get to (non-)truths that are beyond the realm of everyday morality, we’ve already developed a righteous instinct in favour of “truth”. But surely any theology that comes to the same moral conclusions as the naturalist must in fact be unobjectionable.

    October 28, 2009 — 16:04
  • Mike Almeida

    By the time we get to (non-)truths that are beyond the realm of everyday morality, we’ve already developed a righteous instinct in favour of “truth”. But surely any theology that comes to the same moral conclusions as the naturalist must in fact be unobjectionable.
    I’m not sure what you mean. If a theology T that comes to the same conclusions about the value of truth via false beliefs must be unobjectionable? It must be objectionable, no, since it is based on false beliefs, and (by hypothesis) we care about what’s true. In any event, I think we’ve more or less exhausted the topic.

    October 28, 2009 — 18:47
  • Ah…yes. Emeth of Calormen, my bad.
    Total depravity, as defined by most evangelical Christians, is that people are all bad and have no good in them whatsoever…but if that’s true how can one even imagine good? How can they pretend to believe, explore the possibility of belief if they’re really all bad, yet how can God’s grace operate in them if they’ve not accepted Him or pretended to even ponder that He could be true.
    If God exercises grace prematurally of their belief that thoroughly then why can’t everyone just go to heaven under the same operation of grace?
    And if it’s just a smidge of grace or just a little bit of enlightenment to give them some kind of chance, still how do they recognize it if they’re really that bad…God doesn’t choose for them.
    And to suggest that God only operates for those who will eventually believe isn’t scriptural even to a non-innerent. The rain falls on the just and unjust…to all He gave the power to become sons of God…for God so loved the world He gave His only begotten son…while you were yet sinners, Christ died…etc.
    Plus, for those who haven’t heard about Jesus, but operate as though they have bound by that ‘law on their hearts,’ what about them? Hell?

    October 28, 2009 — 20:58
  • Jeremy Pierce

    I don’t see this blog as the best place to discuss details of positions that are as much to do with exegesis and hermeneutics as philosophy, but you don’t seem to be working with an accurate picture of the views you’re criticizing.
    Total depravity, as defined by most evangelical Christians, is that people are all bad and have no good in them whatsoever
    No, not at least by those who actually hold the view. What you’re giving is in fact what people who hold to total depravity insist is not their view. I would in fact consider it a version of hyper-Calvinism to hold something like that.
    Total depravity is the view that every element of our being is affected by the fall. It’s not that there’s no good in us or that we’re as bad as we could be but that every aspect of us is tainted.
    Total depravity is also a view of what human beings are like apart from God’s grace. With God’s grace working to regenerate someone, it’s no longer true that the person is totally depraved.
    If God operated with saving grace on any individual, that person would be saved. So any individual could be saved if God’s grace were present. Some people obviously don’t believe, though, so salvific grace must not be present in those people. You might object to why God doesn’t bestow saving grace on all. Paul’s interlocutor in Romans 9 does that. His only response is that we’re not in an epistemic position to have a clue about how to evaluate God’s choices in such matters.
    The rain falling on all is the classic example of common grace, which applies to everyone. That verse is often even seen as the chief example demonstrating that there is such a thing and the easiest way to refute the hyper-Calvinists who deny common grace. I’m not sure why God allowing blessings to flow on everyone in that sense means that everyone has a work of grace within them that will bring them to salvation.
    The only place I know of in the entire New Testament that talks about those who haven’t heard is when Paul considers them and simply uses it as a good reason to go tell them the good news. If Lewis’ view were correct, telling them would remove from them the default salvation that they have until they’re told and then made magically responsible for now having to respond to the message of Christ.

    October 28, 2009 — 22:39
  • Personally, I’ve always thought that the true meaning of this passage is all in Plato. What do they teach them at school these days?
    I don’t think we can overlook the fact that Puddleglum is not merely engaged in a philosophical argument, he is trying to battle against an enchantment. He knows that either he is being befuddled by a fake memory of Narnia, a mere imaginative construct, or else he is being befuddled by the witch’s enchantment – but, in his befuddlement, he doesn’t know whether or not she is, in fact, a voice of reason.
    So, he seeks out things that cannot be doubted. The sensation of pain simply cannot be ignored – it demands attention. So too with the memory of Aslan as someone worth following. It makes sense to doubt that there is such a thing as the Sun, a single source capable of lighting the whole world, because they are, after all, in a place where the sun’s light is invisible. But there is no place where the light of moral obligation is invisible – even in the underworld, the same moral demands that apply in Narnia would still have to apply. The demand to follow Aslan is as non-negotiable as the sensation of pain.
    Now, what would it mean to follow Aslan if there is no Aslan to follow? It would mean doing all that Aslan demands without any confidence that Aslan would ever be able to rescue you in time of need. In fact, Puddleglum has just made a similar decision – he freed the prince even though he expected to die as a result, knowing that by disobeying instructions, he and his friends had lost any right to protection from Aslan.
    So, following a non-existent Aslan is like following a powerless Aslan. It involves separating the values Aslan represents from the power Aslan has to uphold those values. From a Platonist perspective, this is, of course, an impossibility: nothing could be more unified than God, who does not have different qualities that can be separated. But our knowledge of God is necessarily fragmentary, and our vision distorted.
    In summary, I think there’s an element of reformed epistemology about what Lewis is doing here. He has a Platonic theory about our source of knowledge of God, and a Christian theory about how that source can be befuddled, and a theory about how best to overcome such enchantments.

    November 2, 2009 — 10:30