Doubting Thomas and Pious Pete

Religious people like to say that religion provides an objective basis for morality and values and such, whereas without religion these can’t be anything more than just opinion.  Of course, even if this claim were true, it would prove nothing about the existence of God or about the goodness (in any sense that is meaningful to human beings) of any values that God might prescribe.  It’s no argument to say that something must be true because it would be bad if it weren’t.

But the main point here is not about why the argument is false, but rather about why it’s persuasive (I assume it’s persuasive because I hear religious people making it all the time).  I think the reason is this.  Suppose Doubting Thomas tries hard to figure out the truth about these matters, and has some not completely conclusive and only partially satisfying answers to show for his efforts, and he’s had little fun along the way.*  His buddy Pious Pete points out that Thomas did a lot of struggling and still doesn’t even claim to have a rock-solid foundation for his values whereas he (Pete) does.  That is, Pete correctly points out that Thomas faced a hard problem, struggled, and still didn’t find a fully satisfactory solution, while Pete went through none of this.  Of course Pete didn’t solve the problem either, he just avoided it by refusing to take it seriously and essentially assuming it away.  But it sure looks like Pete not only has the better deal, but is also righter; after all, he dealt with a lot less muss and a lot less fuss than Thomas did.  Misdirection, just like a magician, and it works.

*I think it’s fair to say that most people who either don’t believe in God or don’t accept God’s values don’t believe that values are just opinion (some of my views on this are in the comments to this post).  But some do, and even those who don’t probably had to do some real work and give up some cherished beliefs to get to whatever resolution they arrived at.

GD Star Rating
Tagged as: ,
Trackback URL:
  • Stuart Armstrong

    More bizzarely, if Thomas claims he has a moral rock to stand on, but it isn’t a religious one, people don’t believe him and point out the contradictions immediately.

  • Stuart Armstrong

    In the same way, people wondering “how could this so complicated universe have come into existence” conclude that “god did it”, and leave it happily at that – explaining nothing.

    Something really stops us from looking at religious arguments the same way we look at other arguments.

    Correction to my previous post:
    There do seem to be certain ideologies that are taken up with the same fervour as religion, marxism for one. Some freudian followers also seem to have the same certainty.

  • The double-standard is to charge that there is something irrational about having moral convictions (for which I can say something about why I hold them, where this still falls short of an absolute proof), but nothing irrational about having religious convictions (which are, at best, in the same boat).

  • David J. Balan

    Matthew, I think a lot of religious people would be satisfied if they could just show that their beliefs were no *more* irrational than those of a rationalist. And it really would be a significant vindication for them if this were so. So while you are right that religious beliefs are “at best” in the same boat as rationalist ones, the rationalist position really depends on being able to defend the claim that religious beliefs are in a worse boat. I think this standard is easily met, but that is the standard.

  • Nick Tarleton

    Of course, Pete is still begging the question of why he should do what God commands. He could appeal to hell, but then his ‘morality’ is not true morality, just the commands of a tyrant. And even that still begs the question of why he wants to avoid hell. No value system, not even self-interest, can be derived without taking some values as axiomatic.

    Personally, I don’t think this is a huge problem because normative claims aren’t necessarily worse off than non-normative ones; even deductive logic has to be taken as axiomatic, apparently. So there may be normative claims as obvious and reasonable as ((A->B)&A)->B, like “death is undesirable, all other things being equal.”

  • joe

    The prospect of eternal damnation in which you are forever punished juxtaposed with the proposition of a fairy-tale life in heaven…. is that not a very powerful idea to hold over someone’s head, even for someone with a strong prior against the possibility of an afterlife? Of course, if you place 0 probability on the possibility of an afterlife, then it doesn’t matter what someone says. But we all know what kind of trouble those kind of priors can get you in… this could potentially be the worst.

  • LP

    David —

    I think religious people make this argument not because it seems persuasive to them, but because it provides justification for beliefs they are already committed to having. Most religious people I’ve talked with have had some kind of personally transformative experience that they associate with God, and only bring out these sorts of arguments when challenged by the non-religious. It’s sort of like Pascal’s Wager — no religious person regards these practical kinds of reasons as sufficient for religious belief, but they trot them out because atheists are sometimes taken in by them.

  • Stuart Armstrong

    Maybe this may be due to another form of politness bias – it’s rude to criticise religion. So people don’t, at least not in public. So the arguments are not put to the test.

    The result is that religion is left with intensely weak arguments to defend themselves with.

    On a practical note, I’ve seen religion vs religion debates (christians vs muslims, in this case), and the arguments (on both sides) were much more persuasive and coherent than Pete’s. Maybe since they’re not expecting defference from another religion, they hone these arguments to a sharper edge?

  • David J. Balan

    Nick, You’re exactly right, and that is why I put in the bit about goodness in any sense that is meaningful to human beings. Atheists often make the point that there is no reason to believe that God exists, but seldom make the (at least) equally powerful point that if he does exist in anything like the form described in the religious sources, it is impossible to avoid the conclusion that he is a homicidal psychopath, unless you are going to accept as an axiom that if God said it it must be good (which is precisely what you are expected to do). There are deeper questions here too, that philosophers have been kicking around for a long time. Is what God says good *because* God said so? If so, that’s pretty arbitrary; God could have flipped a coin. Is it that there is some good that’s independent of God, but God is just a better moral philosopher than you? That’s at least possible, but then you have to give up rock-certainty in exchange for deference to someone who you think is (much much) smarter than you in these matters. That would be sensible enough if there were any evidence that it was true, but of course there isn’t.

    LP, you’re right that this is more how religious people talk to skeptics than how they talk to each other, but they still say these things presumably because they are effective in either defending their own faith or in spreading it to others.

    Stuart, You’re right that we are incrediby deferential towards religion in a way that we’re not towards other things. This causes a lot of obvious problems (Dawkins is forever talking about this). I’m basically on board with less deference, although I should note that the norm that you don’t criticize someone’s religion probably has some utilitarian value in allowing very different people to get along. As for the inter-religious debates, I’m not surprised if they sound smarter, but I think that’s just because the most indefensible beliefs of both parties are precisely the ones that neither of them are called upon to defend, and so they appear to be on more solid ground.

  • Daniel Greco


    Let me try to play Devil’s advocate. Here’s a valid (though, I think, certainly unsound) argument:

    Premise 1. If there exists a basis for objective morality and values, that basis must be God.
    Premise 2. There exists a basis for objective morality and values.
    Conclusion. God exists.

    Perhaps this is the way the argument is intended?

    It’s a bit strange to imagine using this argument to come to believe in God. If I were convinced of the two premises, but was not convinced that God existed, I’m inclined to think I might start to doubt one of the premises, rather than accept the conclusion. But I’m not sure whether that must be the right response. Perhaps I just can’t imagine being in a state of mind in which I was convinced of both premises of this argument, but not convinced of the conclusion. Perhaps, if I were convinced of the premises, and it was brought to my attention that they entailed the conclusion that God exists, I would conclude that God exists.

    Maybe what strikes me as bad about the argument (beyond the fact that I think its premises are false) is this: It seems strange to use moral premises to infer a metaphysical conclusion. I’m inclined to think that, in some sense, metaphysics is prior to ethics. You figure out what you think about God’s existence first, and then make whatever the relevant moral inferences are–to do things otherwise would be to get things back to front. I’m not sure that I’m willing to back this principle in general, but it strikes me as intuitive. Do you think this might be what’s wrong with the argument (besides the false premises)?

  • Nick Tarleton

    Joe, I would certainly take heaven over hell, and if there were some course of action that I had good reason to believe would get me into heaven, I would follow it. But my point was that “what gets you into heaven” does not equal “what’s objectively good”, because of what David said, and additionally that the preference of heaven over hell is not necessarily rational. All the time in moral discussion, it seems most people assume that without objective values the rational course of action is pure self-interest, but self-interest is no more justified by pure reason than is altruism or suicide.

    Daniel, the glaring omission in that argument (as in most arguments for God, especially of the First Cause type) is that “God” is a completely free variable that the arguer is surely about to make unjustified assertions about. Certainly if objective morality exists, it has a base and that base exists, but we can no more attribute any specific properties to that base than we can attribute to the First Cause.

  • If there were a caring, all powerful, all knowing God, then it would certainly make sense to believe this God’s claims about morality; he would be by far the best source one could hope for on such a topic. His claims would not be true by definition, but they would be very good evidence. So it seems to me that such a God, if he was talkative, would in practical terms provide something close to an objective morality.

  • David J. Balan

    Daniel, I agree with Nick’s response to your comment.

    Robin, That’s the point. It might be possible to rehabilitate God as the ultimate moral philosopher (as opposed to the *source* or morality). But the description of him in the religious sources suggests that he is either a way worse moral philosopher than I am, or that I am so bad at it that things that seem obviously evil to me are in fact good. Which of those is more likely (conditional on him existing and having done the things he is said to have done)? The only thing favoring the presumption that he is right and I am wrong about, say, the moral virtues of genocide and slavery is that he’s a super-badass and I’m not and he created the world and I didn’t (this is basically God’s argument in the Book of Job). But that’s not much of a reason; certainly not enough to overturn my belief that genocide is evil and that I’m right and he’s wrong about it. After all, human beings have accumulated lots of power and soon will probably even be able to create life, but that fact is only loosely related at best to our degree of moral development.

  • TGGP

    This post seems to approach a previously discussed topic: why should we try not to believe false things? Although I’m not a believer myself, I don’t really care if others persist in believing in a manner that makes them feel good, provided there aren’t significant downsides to it.

    Regarding just how good God is, there are some such as Vox Day who claim that God can make it okay to murder little children. When I was religious that would have made a certain sort of sense. Nobody other than God can determine what is right or wrong, so whatever He says goes. But then why be “good”? Perhaps the afterlife, but I think most Christians would not agree with Vox’s way of thinking. Nerds are nuts, but not everybody is like that. I am also reminded of Max Stirner’s view of God (though I don’t think he actually believed in him). God is clearly an egoist who places his divine plan above all other considerations, why not be likewise and serve one’s own whims rather than God’s?

  • David, judging yourself to be a better judge of morality than God, if he exists, seems to me an extreme example of overconfidence. If the two of you disagree, you should just admit you are most probably the one who is wrong.

  • David J. Balan

    Robin, That’s only true if you interpret his ability/willingness to create the world as a very strong signal that he is a good judge of morality, so strong as to make me utterly overturn my own judgments. I can see no reason to do so.

  • David, even if abilities are only weakly correlated, God’s vast physical abilities surely suggest his moral abilities are greater than yours. What could you possibly point to to suggest that his moral abilities are less than yours? The fact that your moral intuitions conflict with his moral judgments is not such evidence.

  • Friendliness assessment has to be carried out independently. Just because an entity has got a whole big bunch o’ power is no reason for me to believe that it occupies the same moral frame of reference as myself. If “good” is defined as “Whatever God says”, why should I care about this particular definition of “good” any more than I care about paperclips? This “good” runs skew to good as I understand it, a concept distinctly incompatible with genocide. (For which God would surely be responsible, no need to consult the Bible to show this; just look around you.)

    “David, judging yourself to be a better judge of morality than God, if he exists, seems to me an extreme example of overconfidence. If the two of you disagree, you should just admit you are most probably the one who is wrong.”

    Robin, if a Friendly AI that I personally designed and proved correct, tells me that torture is “good” as I defined goodness – for reasons much too complicated to explain to me – then I think I’d probably doubt my own proof of correctness, or question my own construal of the FAI’s metamorality. Do you think this shows extreme overconfidence on my part?

  • Eliezer, if you think that you and God mean different things by “good,” then you don’t think you actually disagree about good. And in the Friendly AI scenario you describe, if you mean a strong “doubt” then it does seem like you look overconfident. I can imagine many situations where I might plausibly think torture to be good. How can you be so confident it is not?

  • Nick Tarleton

    Eliezer, just for clarification, would you say that you’re “right” and God is “wrong” for thinking genocide is “good”, or just that you and God have different goal systems and neither of you could convert the other by rational argument? (Should this go on the consolidated morality thread instead?)

  • Robin, we assume the “Friendly” AI has informed me that torture is intrinsically good rather than consequentially good – perhaps asserting that, after meditating upon the issue for a few centuries with full information, any non-brain-damaged human being will conclude with a flash of surprising insight that torturing little children is beautiful in the same way that freedom, music, and love are beautiful.

    It seems to me that human-built Friendly superintelligence is a good way to state this question, because the causal origins of the supermind and our reasons for trusting it may thus be fully defined. I find it hard to imagine what a mere God could do to make me trust It half so much as a mind I built and verified myself. Even so human reasoning can be flawed and human engineering can fail. If I am willing to question my belief in my own AI’s Friendliness after seeing the “torture output”, I would certainly question my belief in a hypothetical God’s goodness on much less evidence.

    I ask you why you, Robin, would put so much confidence in this hypothetical God’s hypothetical goodness. Or to put it another way, how much initial evidence would you have to see of a hypothetical mind’s hypothetical goodness before you’d trust it about torture?

  • Eliezer, it is possible that usually-reliable evidence of a creature’s superior info and reasoning on some topic might be defeated by the fact that one of its opinions was very obviously wrong. But this possibility should not be generically invoked, as David seemed to do, as an excuse to usually refuse to defer to such a creature’s differing opinions.

  • Austin Cartwright

    Your main point cuts both ways. Lets change the situation a little to make a point. Lets say that we are talking about man made climate change and not morality. Now the Pious Pete is going to look at the evidence and say this looks like it points to the objective truth of man made global warming and because of this he believes it. Pious Pete makes the changes due to his beliefs and tries to change his behavior and society for the better. Doubting Thomas, on the other hand, will look at the evidence and ask a lot of question. He will deconstruct the models and look at all the assumptions that are being made. He will find a thousand difficulties and reasons to doubt. He will find that it is a theory and there is still room for uncertainty . He will face the hard problem, struggled with the questions and find all unsatisfactory solutions while Pete did none of these. But in so doing, Doubting Thomas does not have to change. He can go on living how he wishes with no added burdens. He does not have to change, using perfect misdirection, just like a magician, and it works. Now one day the truth will be known by both Pete and Thomas but then it may be too late to change.

    The problem with the original argument and the one I just used is that it assumes that the problem is the ill motives of the other side. Usually this is just our human bias since we can not read others heart’s.

  • Robin, superior info and reasoning does not imply within my moral frame of reference and you’ve got to evaluate evidence for the second above and beyond evidence for the first. I believe this is the point that Balan, TGGP, Tarleton, myself, and others have been trying to make.

  • Eliezer, I agree that one has to consider the possibility of miscommunication, that you and they are talking about different things, when trying to come to terms with a disagreement. This applies to any disagreement, about morals or anything else.

  • Eliezer, just for clarification, would you say that you’re “right” and God is “wrong” for thinking genocide is “good”, or just that you and God have different goal systems and neither of you could convert the other by rational argument? (Should this go on the consolidated morality thread instead?)

    I replied in Consolidated Nature of Morality.

    Robin, you might want to take a look at this, as I don’t think that ending up in different moral reference frames is properly described as “miscommunication” or even “disagreement”.

  • Stuart Armstrong

    Even if God/the AI is friendly, and has a moral system close to mine, that doesn’t mean I should trust him.

    For example, he may be paternalistic, and just telling me what would be best for me to hear. This doesn’t make it the truth. Since GodAi is so intelligent, he can be paternalistic at a level far beyond my imagining, and hide it from me with total efficiency.

    Words like ‘good’ or ‘caring’ are also dangerous here – if GodAi is so superior, he can make sure to appear good or caring in my eyes, even if he’s not. In fact, GodAi’s supreme intelligence is a problem here – I can’t trust him at all because of it. Someone of less fabulous intelligence I can trust more, because there is a chance that if he lies, he may get caught out. But GodAi’s lies will be perfect.

    In Eliezer’s words, not only does ‘superior info and reasoning‘ not imply ‘within my moral frame of reference‘, but the first actually makes it harder to assess the truth of the second. If GodAi is perfect, we should ignore him entirely.

  • Stuart, if GodAi doesn’t like you, you were probably disassembled into component atoms fifteen minutes ago. If GodAi wants to deceive you (why?) and doesn’t care how, it can just directly rewrite your brain. In some cases, superior info and reasoning makes it very easy to assess benevolence – if you’re still alive.

    Not that I’m saying you should trust GodAi, just pointing out that if GodAi is “paternalistic” you’ve got bigger problems than just GodAi’s words being optimized to deceive you.

    Naturally, all this would also apply to God. You know, it really is sad how religious people don’t ask the most basic questions about religion.

  • Stuart Armstrong

    Stuart, if GodAi doesn’t like you, you were probably disassembled into component atoms fifteen minutes ago. If GodAi wants to deceive you (why?) and doesn’t care how, it can just directly rewrite your brain. In some cases, superior info and reasoning makes it very easy to assess benevolence – if you’re still alive.

    There are many more possibilites that these – GodAi may behave like a scientist, experimenting with the human race. GodAi may be trying to overcome his own dislikes. GodAi may have a constantly shifting moral system along with a general unwillingess to act. He may keep me alive because he believes life is a misery and I deserve to suffer.

    The point is that we know nothing about GodAi’s motives (execept, if he is an AI, we may have some idea how he was initially designed). Talking about a “caring” or “good” GodAi is moot – we have no clue if he posseses these qualities. Since he is so ultimate, he can hide any moral system – his plan might be “benefit humanity for a thousand years, then torture them.” We can argue that’s an unlikely moral system, but we have no reasonable probability distribution to be talking about what moral systems GodAi could go for.

    just pointing out that if GodAi is “paternalistic” you’ve got bigger problems than just GodAi’s words being optimized to deceive you.
    Indeed. But there is a reverse pascal’s wager here – I have to behave as if I could ignore GodAi’s influence. Because if I can, that’s the only way to do so (I categorize “I can ignore GodAi’s influence by doing something complicated that I could never figure out” as “I can’t ignore his influence”). And if I can’t, the point is moot whatever I do.

    Naturally, all this would also apply to God. You know, it really is sad how religious people don’t ask the most basic questions about religion.
    I think the fact that there is no evidence for God is the saviour of religion. Imagine if there was a real God, who confirmed his opinions in clear terms… and the havoc that would wreak on every religion.

  • Stuart Armstrong

    Just a thought, possibly relevant to Friendly AI, on how we could get an impression of trust on a GodAi-like intelligence:

    Have a chain, a hierarchy of AI’s, progressively more intelligent, streaching to from human to GodAi. At each level, ask the AI “can the next level of intelligence be completely trusted, and have they answered this very same question, to you, in the affirmative?”. As the next level is not sufficiently above them to be sure of deceiving them with confidence, an answer of “yes” from the AI’s just above us would confirm (to high probability) a chain of trustworthiness up to GodAi.

    An answer of “no” is much less informative, of course, just confirming a slight lack of trust somewhere in the chain.

    If we are pretty sure that nowhere is there an overlarge leap of intelligence and power, and that AI’s find it hard to deceive intelligences just slightly below themselves (we might build up these assurances, step by step, as the GodAi increases in intelligence), then we may be able to ask more nuanced questions.

  • David J. Balan

    Boy, this got out of my depth in a hurry!

  • joe

    “Imagine if there was a real God, who confirmed his opinions in clear terms… and the havoc that would wreak on every religion.”

    Imagine if there was a real god, who has already tried to express to us his opinions, but many rejected him because they didn’t agree with his views.

    I don’t think this would wreak havoc on every religion. Most would probably call the god a false god… claiming that he couldn’t possibly be the god in whom they had placed their faith and worshiped.

  • joe

    “You know, it really is sad how religious people don’t ask the most basic questions about religion.”

    You know what I find more upsetting? Not only do they not ask the most basic questions, but most probably don’t think that they should. Too busy trying to spread the answers they think they already have. Not intelligent enough to see through the veil that has been pulled over their eyes…. or perhaps they simply don’t want to because of the way their religion makes them feel.

  • TGGP

    joe, if God is omnipotent and truly desired for people to believe something, shouldn’t his will have already been done? That seems to be what others have been getting at above.

  • joe

    Why would it be natural to conclude that god would use his powers to force people to believe something? There is a big difference between expressing your views for the sake of teaching and imposing your opinions on someone. Parents understand that no matter how much guidance they try to give their kids, that there are some things that kids just have to learn on their own if they are going to truly learn and understand.

  • TGGP

    Parents are not omnipotent and omniscient. God (assuming He actually exists) is. God was able to foresee whether or not we would believe in him. He created the conditions that caused people to not believe in him. The obvious conclusion would be that God doesn’t actually want many people to believe in him at all, or some of our assumptions (God’s existence, omnipotence or omniscience) are wrong.

  • joe

    I think you missed the point of the analogy… maybe god, like good parents, knows the best way for people to learn certain lessons about life. Having the knowledge would not necessarily mean that we would understand. I would say this is analogous to students being able to memorize formulae and methods from textbooks, but not really understanding the material. Perhaps an omniscient god knows the best way for us to understand.

    Clearly, we have some, if not complete control of our lives. Thus, if there is a god, even if he is omnipotent and omniscient, he is allowing us to make at least some decisions.

    “He created the conditions that caused people to not believe in him.”

    Sure, if you want to believe that he is omnipotent and uses this power to control absolutely everything, then sure, I will go along with your above statement. However, I think that is a point on which I can agree to disagree with you. I don’t think that an omnipotent god would necessarily use his omnipotence to control everything. What would be the point?

  • TGGP

    joe, God knew in advance what would result from the initial conditions of his creating the universe. By virtue of his role in creating the universe He MUST “control everything”. Everything that happened must have been acceptable to him. If he did not know what was going to happen, he was not truly omniscient and if he could not bring about his desired state of affairs, he was not truly omnipotent. Parents and teachers have limited ability to make children “truly understand” (although I think much of that talk is just bull among educators to avoid accountability, which is discussed in depth at this blog), GOD DOES NOT. If God wanted us to “truly understand” then we WOULD. I once believed in a somewhat inactive God, but that conception MADE NO SENSE.

  • joe

    “Parents and teachers have limited ability to make children “truly understand”, God does not”

    I think that is precisely the point. Notice that I draw a distinction between truly understanding and the transfer of knowledge/facts from parent/teacher to child.

    “If he did not know what was going to happen, he was not truly omniscient and if he could not bring about his desired state of affairs, he was not truly omnipotent.”

    I see that I am getting schooled here by using terms that lay-people toss around without much thought to their implications.

    So, I believe you are saying that if god is truly omnipotent and omniscient, then when he created the universe, he must have known every result of his actions or else he is not truly omnipotent….good point. I guess I don’t believe that if there is a god, he is either omniscient or omnipotent, nor do I think that any religion truly does either.

    If god is omnipotent and omniscient, then we must live in a completely deterministic world for how could he be otherwise. If that is the case, then humans do not have free-will or the power to make any decisions. However, there also many implications to this view including the pointlessness of heaven and hell. If god controls everything, then there is no point to either punish or reward someone in the afterlife since all of one’s actions while alive were determined by god himself. Say goodbye to any point in talking about good and evil, morality, or responsibility and accountability for one’s actions… doesn’t life necessarily become completely meaningless?

  • Nick Tarleton

    joe, libertarian free will is logically incoherent regardless of what you believe about God, and plenty of people realize this and don’t consider life meaningless.

  • joe

    Could you please expound on your statement “libertarian free will is logically incoherent regardless of what you believe about God”

  • TGGP

    joe, the theological view I am expounding is just good-old fashioned Calvinism. It’s as American as witch-burning. For more on different views of free-will and what science says about them, read Joshua Greene and Jonathan Cohen’s “For the law, neuroscience changes nothing and everything“.

  • Pingback: Overcoming Bias : I Had the Same Idea as David Brin! (Sort Of)()