Believing in Todd

Imagine that the kids in a family blamed broken items discovered around the house on "Todd."  When the parents ask more about Todd, and ask the kids separately, they get conflicting answers about Todd’s height, skin color, personality, and so on.  These facts would count as evidence against Todd: 

  • The parents have never seen Todd, though they are around often.
  • The kids want everyone to believe in Todd so less blame will fall on them.
  • The kids give conflicting stories about Todd’s features.

This seems similar to disagreements about God, and addresses the question Nick raised about Hal’s comment.  The facts that religions disagree about God’s features, that they have reasons to want to believe in God even if he did not exist, and that skeptics find it hard to find independent evidence for God beyond what supporters say, all suggest that religions are independently making up this story of God.

Added: I wonder how well a similar structure could rationalize other situations, such as in morality or paternalism, where some want to take disagreement as evidence for skepticism.

GD Star Rating
Tagged as: ,
Trackback URL:
  • Matthew C

    The question of God is not like the question of Todd.

    Whether a particular Todd exists or not is an emperical question — we gather the facts and make a determination based on the weight of the evidence.

    The question of God is a question about the nature of ultimate reality.

    The basic disagreement between atheists and all those with some form of God belief is whether ultimate reality is conscious, creative, and purposeful and that the regularities we see in the universe are a result of that, or whether ultimate reality is instead unconscious, accidental and non-purposeful natural laws. This is a much more interesting question to me than whether a particular tribal anthropomorphism of “God” and religious practices is absolutely real and correct, and all the other tribal anthropomorphisms are wrong. I suspect 95%+ of the readers of this blog have already decided that fundamentalist tribal religion is not absolutely and exclusively true, while the question of the ultimate nature and basis of reality seems much more unsettled.

  • Matt C, many people claim to have regular contact with God, and as I discussed here, the kind of God I’m skeptical about is this sort, one who intervenes frequently in our lives.

  • “Oh, sure, everyone thinks two plus two is four, everyone says two plus two is four, and in the mere mundane drudgery of everyday life everyone behaves as if two plus two is four, but what does two plus two really, ultimately equal? As near as I can figure, four. It’s still four even if I intone the question in a solemn, portentous tone of voice.”
    The Simple Truth

    Whether or not a particular God exists is an empirical question. We gather the facts and make a determination based on the weight of evidence. And the answer is still the same – “no” – even if you intone the question in a deep, portentous tone of voice.

    What would you do if, heaven forbid, you actually went and found your ultimate reality? Would it immediately become boring to you, and would you start looking for ultimate-ultimate-reality? This is a chain that never breaks, as it is written: “Alas for those who turn their eyes from zebras and dream of dragons! If we cannot learn to take joy in the merely real, our lives shall be empty indeed.”

    There is no ultimate reality. Just reality. That’s all there ever is.

  • What other things could this argument form be applied to?

    Suppose we substitute “psychopath” for “parents”, and “moral truth” for “Todd”…

    Or we could substitute “nihilist” for “parents”, and “value” for “Todd”…

    Or “philistine” for “parents”, and “aesthetic merit” for “Todd”…

    Robin, are psychopath, nihilist, and philistine on the same footing as atheist?

  • Nick, “God” is supposed to make things happen – that is, the hypothesized “God” is an instance of the class of “things that make other things happen”.

  • Nick, just as you wrote your comment I added a similar question to the post above. 🙂 It would certainly seem that one could try an argument with a similar structure; the question I guess is how well the details work out.

    Many actually have taken disagreement about morality as evidence for moral skepticism. But conversely one could take high levels of agreement about specific morals as evidence that those morals are real. So the skepticism would apply better to the kind of morals that vary widely.

  • Eliezer, so does that qualification need to be added to Robin’s argument structure? If so, why exactly does the argument fail without the extra assumption that “Todd” makes other things happen?

  • If God can observe, think, believe, judge, decide, act, affect or be affected, then God is part of the causal network. You can call it the Great Web of Causality, if you can’t get excited about things without speaking about them in an ominous voice, indicated in writing by capitalized letters.

    We, too, are part of the Great Web of Causality, and we learn about other, more distant nodes when their effects touch us, whether directly or indirectly. For complicated philosophical reasons, it is not proper to hypothesize a part of the Great Web of Causality unless it is the most compact explanation for the effects that have touched your own node. (Complicated philosophical reasons = you can’t draw an accurate map of a city by sitting in your living room with your blinds closed; you have to actually look at the city, or have it interact with you some other way.)

    Thus, if the kids start talking about a part of the Great Web of Causality called “Todd”, their only legitimate reason would be that one of Todd’s direct or indirect effects touched their node of the Web, and the kids reconstructed a model of Todd by observing His effects – in other words, their eyes saw Todd walking through the house and their brain correctly guessed that the cause of the image was Todd.

    The kids are part of the same Web as you – you know this because you can see the kids and talk to them. So you’d expect that, with a bit of effort, you could manage to intercept some of Todd’s rays of causality with your own node. In fact, on the hypothesis that the kids are telling the truth, you have already been indirectly affected by Todd, because Todd affected the kids who are affecting you. And yet you, yourself, can’t seem to see Todd directly. This leaves the question of whether Todd is the simplest explanation for the kids’ effects on you. Hence, Hanson’s observation that other non-Todd explanations exist for the kids’ behavior is salient. (If we have the causal structure A -> B <- C, then observing A can explain away B which then is no longer evidence for C.) Utilities, values and aesthetics do not have correspondence theories of truth. When I say a flower is beautiful, this statement does not correspond to a little invisible "beauty" tag attached to the flower. I am not stating here that morality does not exist, but rather, that morality is not a member of the class of "things that exist or don't exist". Rather, moral statements belong to a class of possible answers to calculations that are only carried out by physical objects that are congruent with them, the way that 4 is the only correct answer to 2 + 2 = ?, but not every possible calculator is computing that particular question, and yet given that you're asking 2 + 2 = ? the "correct" answer doesn't appear to depend on your own state of mind.

  • On second thought, maybe the last paragraph should be rephrased as follows:

    1) Even small children distinguish between alterable and unalterable moral statements; the teacher can make it all right to stand on your chair by saying so, but the teacher cannot make it all right to steal from a fellow kid’s backpack by saying so.

    2) There is no tablet on which the words “Killing is wrong” is written such that if you erased those words and wrote in “Killing is good”, it would become good because that was written on the tablet.

    3) General moral statements have no correspondence theories of truth because there is no state of affairs that would make them false, the way the statement “there is white snow on the ground” becomes false if I go out and melt the snow.

    4) General moral statements are not part of the Great Web of Causality – there is no node in the Web corresponding to a tablet on which we could write different things and thereby change morality.

    5) Hence, unfortunately, general moral statements don’t affect the world; even though death is bad, people still die.

    6) Hence, you cannot adduce general moral statements by observing the current state of the world because the current state of the world is not affected by general moral statements.

    Of course, your own thoughts are also part of the state of the world – so how can you treat them as evidence about morality?

    That’s when you have to start thinking of morality as a Platonic calculation which humans reflect as calculators. The answer to 2 + 2 = ? does not depend, even counterfactually, on the calculator; you can’t make the answer 5 by tinkering with the calculator. There is also, so far as we know, no tablet on which 2 + 2 = 4 is written, whereby we could make it 5 by writing a different number instead. The tricky part is that 2 + 2 = 4 is not part of the laws of physics and it does not directly affect the state of the calculator; it is computed by the calculator in the course of obeying the laws of physics. It’s a lot easier to imagine different laws of physics than 2 + 2 = 5. It’s in this sense that a calculator can be treated as evidence about numerical statements, which are not the sort of things that exist or don’t exist.

    Likewise, your thoughts are evidence about moral calculations, which are not the sort of things that exist or don’t exist – they aren’t nodes in the Great Web of Causality, any more than 4 is a specific object in someone’s closet somewhere. The reason the psychopath comes up with a different answer is that the psychopath is performing a different calculation – not that the psychopath looked in the closet and found that it was, in fact, empty.

    But insofar as God is asserted to see things, undergo complex internal interactions like thinking, cause observed events such as the creation of the universe, etc., God is an alleged part of the Great Web of Causality, is the sort of thing that either exists or doesn’t exist, and we did look in the closet and find it empty.

    Hope that clears everything up.

  • Eliezer, we can do inference with impossible worlds as well as with possible worlds, and so we can use the fact that certain error-prone calculations produced particular results as evidence. So in this way we can reason about math and morality.

  • I’m not totally sure of how to do causal inference about computations (it’s an open field with lots of potential solutions kicking around), but yes, obviously we can and do reason about math and morality. My point is that a psychopath coming up with a different decision is not analogous to looking in the God closet and finding it empty, it’s analogous to a calculator saying 5 instead of 4 because it’s calculating 2 + 3 instead of 2 + 2.

  • I’m a fairly surprised that you would say that there are platonic moral truths, Eliezer. Are you claiming that morality exists independently of sentient beings?

  • Once a sentient being asks a morality, the answer is counterfactually independent of the sentient being’s attributes or even existence. Like, I’m not claiming that numbers exist independently of mathematicians, but I am claiming that, once you ask “2 + 2 = ?”, the answer is fixed regardless of anything that happens to the mathematician including their ceasing to exist.

  • mitchell porter

    May I just point out that we do not know why anything exists, why anything causes anything else, or even that anything causes anything else. Or at least I don’t know any of those things, and I see no evidence that anyone else does, either. OK, this is basic epistemology, and somewhat removed from the question of how to rationally judge the particular theologies one shares the planet with, but if people are going to talk about “ultimate reality” and “the Great Web of Causality”, a reminder of the actual epistemic situation is appropriate.

  • Porter, you may not know why gravity works, what gravity is, or even whether or not such a thing as gravity exists, but if you walk off a cliff you shall fall, and I think we can all agree on that.

  • mitchell porter

    What is your point – that this discussion should take no notice of the actual uncertainty of such inductive judgements? And I don’t mean the quantified faux-uncertainty involved in probabilistic judgement, I mean the fact that the ultimate inputs to a probabilistic worldview, such as a universal prior, themselves entirely lack foundation.

  • TGGP

    Moral truth, value and aesthetic beauty all have no objective existence. The difference between them and God is that God is an entity like others that could possibly observed but (according to atheists/agnostics) has not been, whereas the others are recognized as not having concrete existence outside our subjective thoughts but are nevertheless are insisted upon being treated as if they existed outside of everyone and can thus be applied the same (unless the particular morality you advocate categorizes people in some way) to everyone. Even math describes existing things in a manner that can be checked against reality (when I was young and learning math I had to see if old concepts could replicate new ones where they overlapped, and at the crudest you can count the result of adding two foozles to three to see if it is really five or use a balance-scale to test equality). The mere facts in front of us can never say if something is “good”, and introducing it does not enhance our understanding (by, say, helping us to make correct testable predictions) of what exists. That is why there is not and will never be a moral/virtue/aesthetics calculator nor a parallel to Euclid’s Postulates that can have a general acceptance as a system in which to work. Preferences can not be removed from those areas and with multiple people preferences are bound to conflict.

  • A point I didn’t think of at the time of my earlier comment: if humans are calculators for some true underlying platonic morality, on what basis can we argue that these calculations show any correspondence to the true morality, if “you cannot adduce general moral statements by observing the current state of the world”? It sounds like you are attributing some kind of supernatural insight to human beings.

    Also: I was going to criticize what I thought was a terrible analogy between morality and mathematics, but on reflection there may be some utility in it: you seem to be making analagous errors on both. The question “does 2+2=4?” really only has a fixed, timeless answer given an assumed formal system for arithmetic. Of course you can compare the utility of two formal systems for describing the real world, but that is an empirical matter.

  • Whoa, I’m already losing count of how many difficult philosophical problems, which are subject to ongoing dispute and discussion by lots of clever professional philosophers who devote their careers on these things, seem to be regarded as trivial by the commentators here. Each commmentator appears fully confident in his own solution, and believes that a few sentences suffices to explain and demonstrate its correctness. Of course, even among the few people who have posted so far, we already begin to see disageement over these “obvious” philosophical truths.

    As so often when “easy” topics like philosophy are discussed – Epistemic Modesty has left the building.

  • Okay, I’m being severely understood here and I guess that should come as no surprise, unless I want to write a lengthy paper about it, which I rather wouldn’t at the moment.

    Simon, I wasn’t postulating a “true underlying platonic morality”. As Gordon Worley puts it, “Morality is objective within a given frame of reference.” For a given utility function over world-states, if this utility function does not incorporate an explicit dependency on the cognitive state of the utility evaluator, the utility function will show that the utility is counterfactually independent of what the evaluator “might think” about utility. In other words, if you judge killing to be wrong, then if you imagine a counterfactual universe in which you think killing is right, you will judge that in this counterfactual universe killing will still be wrong and you will judge that your subjunctive brainwashed self will be incorrect about this apparent fact. Thus, it seems like moral facts are independent of what we think about them, and moreover, we can imagine comparing a belief about a moral judgment to a moral judgment. You might say that, from inside a morality, it behaves like it is “objective”, and that this is why morality feels universalizable –

    – except that’s not the real explanation, of course. The real explanation is more complicated than that, and has to do with evolution reusing truth-finding cognitive architecture to represent the later-developed architecture of explicit linguistic argument over “should”-questions. A true daemonic Bayesian expected-utility maximizer would never even ask a incorrectly structured question like “Is my utility function really true?”

    That’s why I keep trying to say that psychopaths are answering a different question, not giving a different answer to the same question.

    One thing I want to emphasize is the rule: It all adds up to normality. When theories have nothing better to recommend themselves, they evolve to be somewhat but not utterly counterintuitive – it’s why, when people encounter Freudian psychology, they say “How strange! How interesting!” Whereas if you try to say that evolutionary psychology predicts that males will woo and females choose, they say, “Well, duh, everyone knows that – how boring.”

    When you’re through axiomatizing Peano arithmetic, it should make boringly normal-sounding statements like 2 + 2 = 4. When you’re through working out an economic model, it should make boringly normal-sounding statements like “People would prefer higher salaries.” When you’re through arguing about the nature of morality and have worked out a philosophy, it should make boring prosaic statements like “Killing is wrong” and “Flowers are beautiful”. Not fascinating counterintuitive statements like “Morality doesn’t exist, it’s all arbitrary, the universe is devoid of meaning.” That’s just people trying to show off how cynical they are. It seems a shame to throw away all our morals and aesthetics because, in a fit of confusion, we misunderstood their structure.

  • Nick, just because something isn’t easy to understand or explain does not mean the answer has to be complicated when viewed from the inside.

  • PS: To amplify on that last statement, what I mean is that confusing questions are not supposed to have confusing answers. It is perfectly realistic to suggest that “the nature of morality” (itself an ill-posed problem) has a resolution which is simple when viewed from the inside, but very counterintuitive to humans (for some reason or other). Problems about which people argue about a lot do not thereby need to be complicated in their resolutions; it makes just as much sense to suppose that the resolution is simple, but crosscuts more than one intuition.

  • PPS: While the internals of the resolution may be counterintuitive – like human psychology being shaped by natural selection, rather than a god, strikes many people as counterintuitive – the outputs of the resolution should not be counterintuitive; the outputs should appear boringly normal. (E.g. General Relativity has hard-to-understand internals, but it outputs the boringly normal prediction that if you walk off a cliff, you fall.)

    Okay, I’d better sign off for the night.

  • TGGP

    It is counter-intuitive that the world is not flat, that we all evolved from single-celled organisms that arose by chance in the “primordial soup”, that there is no magic or supernatural beings and that we not have free-will (the last two seem as much like “outputs” as your examples). It was not necessary for people to know the truth, so throughout human history we believed things that while not true were not terribly harmful to our ancestor’s reproductive success. That something violates our intuitions is not sufficient reason to reject it. It would certainly be counter-intuitive and foolish to claim I am absolved of all responsibility for my actions and then not expect people to become upset at some of them, but it is not necessary to believe in an objective morality for this to be the case. In other words, emotivism does not lead to testably false predictions. “Killing is wrong” is not something it would be wise to dispute if you don’t want people to suspect you might kill them, but our intuitions cannot really be reduced to that simple statement as we find it morally acceptable in situations ranging from eating plants and animals (if you are not a Jain) or appeasing the sun so it will rise (if you are an Aztec). Is there any objective way to determine whether the Jain or the Aztec is correct? You might believe so if you have already accepted a belief system you think to be true and one of them seems closer to yours, but it begs the question of what beliefs you should have chosen in the first place. I would expect that finding two people that will actually give the same response to all morality questions would be like trying to find identical utility functions (which I am not confident are even necessarily consistent for one person in real life).

    I don’t know if this makes a difference to you Eliezer, but I don’t say these things to try and “show off how cynical” I am. I say they do not exist for the same reason I say God does not exist and I am stating my actual beliefs.

  • Eliezer, you have got me confused about what you think. Do you agree with TGGP, or not? If yes, then I think you express it very badly, but I have no further objections. If not, please explain how you disagree.

    Nick, while I am not as comfortable disagreeing with philosophers on philosophy as I am about disagreeing with astrologers on astrology (for example), I am a lot more comfortable than with disagreeing with any sort of hard scientist. I see relatively little evidence that philosophers, in general, base their views on any sort of physical evidence or sound reasoning.

  • This discussion has gotten away from the topic of the post, and with Nick I am uncomfortable with how quickly people jump to make strong claims on controversial issues. To the extent that I anticipate the claims I make will be controversial, I’ve tried to post on them one small piece at a time, so those who object have a full chance to make their case. I think this gives our best chance for a reasoned discussion.

  • john

    Bias? How about comparing religious leaders to lying children?

  • michael vassar

    Nick: Ironic parallel to the religion discussion, isn’t it. Everyone disagrees about what the truth is, but everyone agrees that the truth is simple or obvious. As with Atheists, Professional (analytic) philosophers seem to constitute a class of self-declared ‘experts’ who’s expertise is not generally recognized by those outside of the class, but at least in the case of philosophers we have some posters on this blog, such as myself, who recognize the alleged expertise without belonging to the class.

    Simon: I think that you need to learn about the difference between ‘analytic’ and ‘continental’ philosophy.

  • michael vassar

    It seems to me that while there is some truth to the point about outputs being normal, the case is fairly overstated. Newtonian Gravity tells us that dense objects all fall at fairly similar speeds and move along parabolic trajectories. More dramatically, measurement or artistic training, not to mention developmental physiology, tell us that children have very different physical proportions from adults. Astoundingly in retrospect, these fairly simple observations were not made until the Renaissance. Slightly careful observation tells us that the normal conclusions are wrong. In economics this is often the case today, and many educated people still believe that government can basically dictate collective economic outcomes by, for instance, creating jobs for the unemployed, and almost everyone mis-predicted the outcomes of Milgram’s authority experiments. The possibility that people can be wrong about the fairly ‘obvious’ is a big part of why we have formal reasoning mechanisms, as well as formal methods of empirical examination.

    Of extremely high practical relevance, the conclusion that the highest priority for almost any sort of ethically concerned person is to reduce the probability of existential risk by encouraging the development of Friendly Artificial Intelligence does not constitute a ‘boringly normal’ output, but rather, a counterintuitive one.

  • Matthew C

    There is no ultimate reality. Just reality. That’s all there ever is.

    Let me explain what I mean by “ultimate reality”.

    For example last night I was dreaming and in that dream, a terrorist detonated a nuclear bomb in an airplane flying over Honolulu. Fortunately, it was just a “fizzle”, so only a few tons of nuclear energy were released.

    But then I woke up, and discovered that what I thought was reality was actually just me dreaming.

    So in that sense, compared to the dream I was having, that dream was unreal, and this waking reality is more real (although perhaps not ultimately real).

    In the same way, perhaps this seeming world we live in is itself similar to a dream. In fact, Robin has speculated on the possibility that we are living in a computer simulation. So if that is the case, it might be possible to discover that while still “within” the simulation. That would be akin to the phenomena of lucid dreaming, where one is aware of the fact that one is dreaming while the dream is happening.

    In that event, what we know of as our lives here would be similar to my dream, and ultimate reality would be something quite different.

  • Considering how long the comments have been getting, I’m going to give up for now. Simon, I agree with TGGP that there’s no morality-stuff floating out there, but I think I reacted to this elementary epistemological realization in a perhaps different fashion than TGGP does – I said, “Oops, I messed up my understanding of morality, how silly of me,” instead of “the universe is devoid of meaning, alas woe and despair”. That’s what I was trying to convey by the it all adds up to normality principle.

    I guess you can just ignore the morality-as-computation stuff for now, that would probably take a paper to explain if I wanted to express it. Nick, Robin, just because an opinion is comparatively briefly expressed doesn’t mean it’s a jump to a conclusion. It can be something that was worked out over quite a long period, but heavily compressed. I don’t know in advance what will or will not create an Aha! experience in others – apparently this didn’t, but it was still worth a shot.

  • Michael, ok, I don’t really know much about the difference between analytic and continental philosophy, but if this paper by Nick is analytic philosophy (correct me if I’m wrong) then I think I have a low opinion of analytic philosophy as well.

    Eliezer, I am still suspicious that your different way of putting things may be an indicator of some real disagreement, it would be helpful if you actually wrote that paper to explain.

  • TGGP

    This might have gone on beyond what Robin wished it had, but I’d like to note that my reaction to the lack of “meaning” in the universe (independent of the subjective beliefs of individuals), is not “alas, woe and despair”, just as we need not despair that there is no God.

  • Eliezer: no Aha! for me this time. For lovely Aha!s other times, my sincere thanks.

  • Hey Robin,

    You make an interesting point, but I think it is flawed. You do present some strong reasons why Todd might not exist, but comparing Todd with God is not an apt approach to take. Todd is an individual within a concrete set of boundaries: the family home, with limited access to that space. Thus, the differing descriptions of one distinct physical individual would be sketchy to anyone investigating this case.

    Religious believers, on the other hand, claim distinct experiences with different manifestations of God at different times and in different circumstances. These experiences do seem to show some consistency with the idea of God and thus, are not contradictory. The contradictions tend to come from the interpretations of these texts as done by the believers in their own situations.

    As an example, all three monotheistic religions recognize Jesus:
    – Judaism recognizes Jesus as a heretic (or a prophet, depending on who you talk to – I don’t want to split hairs here though).
    – Christianity recognizes Jesus as the Messiah.
    – Islam recognizes Jesus as a major prophet.

    Thus, there is some consensus around the main claims of Jesus. Does that mean that Jesus does or doesn’t exist? No. The evidence isn’t sufficient enough either way here.

    So, all three faiths make mention of Jesus. They interpret him differently. Other dogmatic discussions occur for numerous other religious topics. Does disagreement mean that the subject doesn’t exist? Not at all… in fact, based on the great deal of discussion and early source documentation pointing to the subject (God), I’d argue that there is some strong justifications to believe.

    Ultimately, the question becomes one of epistemology. And, with all things dealing with truth, we never can be too sure. As a Christian, reason helps me, but ultimately there is some level of faith involved. But, isn’t faith also involved when you look at other worldviews?

    Feel free to check out my blog ( for plenty more discussion on this topic.

    Thanks and talk soon!

    Todd Dow

  • Neil C. Reinhardt

    So Robin,

    Why for your e-mail address does not work?

    At least this one did not:

    I wrote you an email about “ViewQuakes” and requested a response only to have it come back. So, PLEASE E-me at: religionsucks@webtv,net



    “A View Quacker”

    P.S. “Reason” helps a believer? GIVE ME A BREAK!

    Reason Child, is why MILLIONS & MILLIONS & MILLIONS or FORMER Christians have STOPPED believing in your
    childish myth and become Atheists.