Essay talk:Russell’s Teapot and the Burden of Proof/Archive1

This is an archive page, last updated 8 August 2011. Please do not make edits to this page.

Archives for this talk page:  , (new)(back)

Röstigraben's remarks

OK, first, it would've been much better if you had bothered to reformat and upload your paper as an actual essay. I will have to quote you frequently, and while I can copypaste that from the document, it will make it difficult for others to see the context of the quote. And I'll state up front that I skipped the math portions because I don't think I'm qualified to comment on them, I hope they didn't contain an integral part of your argument (it didn't seem that way). Anyway, here goes in a rather short manner because I don't want to waste my whole lunch break on this.

1. First argument from simplicity - It is generally acknowledged that the simpler the theory, the better; and thus, if all things are equal, we should adopt the simplest theory possible. This insight is often referred to as Ockham’s Razor...Let us apply that principle to the question of burden of proof. Occam's razor is supposed to be applied to theories and their assumptions, not "positions". Apart from that, your own theory about how the universe came about involves a noticeably larger and more complex set of assumptions than the naturalistic one and should thus be rejected according to your own principle.
2. Second argument from simplicity - There is another way in which the principle of simplicity supports the uniform rather than existential position. This way involves, not the simplicity of the positions themselves, but the simplicity of the results we reach by following the position. No, this is irrelevant. The "simplicity" of results is not in any way related to their truth. Occam's razor is just about not introducing unnecessary assumptions, but conclusions of high complexity can result from very simple assumptions and need not be incorrect.
3. Attempts to rescue the existential position - God is generally not claimed to be a physical object. Huh? God is assumed to be supernatural, I wouldn't say that's equivalent to non-physical. God's interaction with the physical world is a standard claim of theists, and how would it do that without being physical itself (or at least able to manifest itself at will)? Theists frequently claim that their god can exceed physical limits, but that's different from having no connection to the physical world whatsoever. What you're outlining is an extreme version of deism, not mainstream theism.
4. Is existence univocal or multivocal? - Too lengthy to quote here, but a quick point: your "uniform" position would run into the exact same problems as soon as you apply it to questions of existence (as it is supposed to), so this is not an argument that shows its superiority.
5. Russell’s Teapot - Again, no real choice quote I take issue with, just the general stance, but I'll use this one: So, Russell’s teapot is actually the wrong case to use to derive a rule about the burden of proof in a case where we truly have no evidence either way. Yeah, I can see where you're going with this. The reason why Russell's teapot is, in fact, a very good analogy is because actual theist positions often take the form of similarly unreasonable claims that contradict prior evidence.

Lastly, I seriously wonder why you are outlining and arguing for an agnostic position. You're a self-avowed theist and have spent considerable time developing your religion - was that all wasted, since you can't produce the kind of evidence you consider necessary to back this claim up? Röstigraben (talk) 10:59, 29 June 2011 (UTC)

Agnosticism: Basically the purpose of my essay is to argue for what I call weak agnosticism, yes. My basic position could be summarised as In the absence of evidence for or against a proposition — in this case the proposition "God exists" — we should withhold judgement and neither belief nor disbelieve it. Now, I personally believe I have good reasons in favour of the proposition "God exists", which is why I am a theist; but, those reasons are outside of the scope of this essay. (One must travel one step at a time.) The purpose of this essay is to argue against the form of atheism which goes "There is no evidence, so disbelieve unless and until there is any", and argues that absence of evidence is no good reason for atheism, and should rationally result in agnosticism instead. Now, some atheists will say, that they actually have evidence for the non-existence of God (e.g. they might claim the problem of evil as an argument against God's existence) - this essay does not attempt to address any such arguments, so that is also out of scope.

Mathematics: Actually, the mathematics section is rather important - my point is the "existential position" (as I call it) is not followed in mathematics, and in fact when applied to mathematics produces a contradiction. So, the existential position only makes sense if mathematical objects are excluded from it - but, if we are to exclude mathematical objects, why not exclude other objects also (like God)? I encourage you to study the section, it does not require any heavy maths. (In fact, all required background knowledge can be found on Wikipedia.)

Principle of simplicity: You say "Apart from that, your own theory about how the universe came about involves a noticeably larger and more complex set of assumptions than the naturalistic one and should thus be rejected according to your own principle". My theories of the origins of the universe are out of scope for this essay. You say they are more complex than naturalistic ones; I am not sure I agree, but that is not relevant to this essay. In any case, even if my theory A is more complex than theory B, the principle of parsimony does not mean prefer the simpler theory unconditionally; it means prefer the simpler theory unless there are good reasons to do otherwise. So, even if my own theories are more complex, I can claim there are good reasons to justify the extra complexity. But, anyway, out of scope for this essay.

You say Occam's razor is supposed to be applied to theories and their assumptions, not "positions". But how are these two "positions" not theories? They are theories about the burden of proof. Following the principle of parsimony, given two theories about the burden of proof, if we have no other good reasons to choose between them, we should prefer the simplest. Some would argue they are not theories because they are not physical theories or scientific theories — but, the scope of Occam's razor is surely not limited just to physical or scientific theories — given that William of Ockham was a mediaeval Catholic theologian and philosopher, he would doubtlessly have disagreed with such a limitation.

"Occam's razor is just about not introducing unnecessary assumptions, but conclusions of high complexity can result from very simple assumptions and need not be incorrect." - Ockham's Razor is commonly quoted as entia non sunt multiplicanda praeter necessitatem ("entities must not be multiplied beyond necessity"). The existential position greatly multiples the number of propositions I should believe compared to the uniform position, for no apparent gain. If I have no evidence whether A exists, the existential position would have believe "A does not exist". The uniform position does not have me believe any proposition about A, beyond purely self-referential propositions of the form "I don't know whether A exists". (And, the existential position produces just as many such self-referential propositions.) Multiply A for any and all possible purported existents for which we lack evidence. So the existential position, vastly multiples entities - as in, propositions to be believed - compared to the uniform position.

Now, you probably disagree with my interpretation of Occam's razor. But, there is no doubt that Occam's razor is a vague principle that has many competing precise formulations. You may disagree that my version is right; but surely it is reasonable. And, faced with multiple competing formulations of Occam's razor, how do we determine whose is correct? Is there some test or experiment we could perform, to determine whose is right?

Is God a physical object? - You are confusing the question of whether God acts on physical objects with whether God is a physical object. A theist will claim God acts on physical matter (e.g. to work a miracle). But most theists do not claim that God is a physical object. God is claimed to be a non-physical object which acts on physical objects; but that is not the same thing as saying God is a physical object. Some rough analogies: a lamp emits light, but the lamp is not light; a speaker emits sound, but the speaker is not sound; a laser beam cuts metal, but the laser beam is not metal; the lumberjack chops wood, but the lumberjack is not wood. Just because A acts on B does not mean A must be the same type of thing as B. Likewise, if X acts on a physical object, there is no logical necessity that X must be a physical object.

Is existence univocal or multivocal? - your "uniform" position would run into the exact same problems as soon as you apply it to questions of existence (as it is supposed to), so this is not an argument that shows its superiority. What difficulties would the uniform position have? The uniform position is not defined in terms of the concept of "existence", so it is unaffected by any philosophical difficulties in defining existence. The fact that it can be applied to questions of existence doesn't make it subject to those difficulties, since it advises the same result whatever answer one has to them. Whereas, the existential position defines itself in terms of existence, so philosophical difficulties around what is existence are much more problematic for it...

Russell's teapot - You say "The reason why Russell's teapot is, in fact, a very good analogy is because actual theist positions often take the form of similarly unreasonable claims that contradict prior evidence." The point is, Russell's teapot is not a good argument for any theory about what the burden of proof should be. "claims that contradict prior evidence" is not a good case for determining the burden of proof, because it is too easy to muddle the evidence we have into the issue. A good case for determining the burden of proof would be "we have no evidence either way - what evidence would we need to go one way or the other?" And Russell's Teapot is not such a case. Since it is not a good argument for the burden of proof in general, it's not a good argument for what the burden of proof should be with respect to the existence of God either. (((Zack Martin)))^™ 10:33, 1 July 2011 (UTC)

Agnosticism - you see, the problem with agnosticism is that it's always a very reasonable position when applied to "god" in the abstract, just referencing a very vague notion of a deity. It is trivially easy to come up with possible gods - conceptualizations that usually are a variation of deism, and were set up with maximal unfalsifiability in mind. In a way, I'm agnostic towards these deities in that I simply don't bother with them at all, since they have no impact whatsoever on reality. Neither does anybody else, however - atheists like me deny the existence of highly specific supernatural beings, the ones that are actually worshipped by humans, who have acquired a vast mythological baggage over the years. The basic claim about their existence is complemented by a staggering number of secondary assertions, most of which tend to take extremely unreasonable forms that very often either contradict themselves or available evidence. I'm sure you know this problem very well, because you evidently designed your own goddess with the aim of avoiding such problems.

Occam's razor - I'm familiar with how it's used in a scientific setting: when in doubt between competing explanations, select the one that makes the least assumptions. Or in other words, in the absence of conclusive evidence, favor a hypothesis or proto-theory that has the best explanandum/explanans ratio. It's only to be used in cases where it is impossible to discriminate between hypotheses on the basis of evidence or logic. I agree that this highly specific and narrow usage is probably not what Occam himself had in mind when he formulated that sentence, but science is an obvious application for it. There are good reasons for keeping explanations as simple as possible - it's more economical, reduces possible error sources and makes theories both easier to test and understand. However, I'm unconvinced about its applicability to what you refer to as more general "positions". You just state on faith that we should prefer the simpler one, without giving a good reason for it. The problems that introducing unnecessary (the key issue here) clutter into explanations can pose in science does not present itself in a similar fashion when the explanans-explanandum structure isn't there. We don't even have any way of judging what's necessary and what is not.

Secondly, one thing that definitely does not make sense is to abuse Occam's razor by discriminating between general principles based on the number of propositions that we would have to accept as a result. Just think this through to the end - the simplest, and therefore best, position would be one from which absolutely nothing follows, say a paradox (I was tempted to substitute "postmodernism" here, but didn't want to go off a tangent). If we simply denied the possibility of knowledge or the validity of logic, we wouldn't have to accept or reject any propositions, which is a very simple result, but certainly not a desirable one for either of us.

Physical god - I guess we've again reached the familiar materialism-idealism junction here. I hold it to be impossible for a (completely) non-physical subject to interact with physical reality, but your mileage may vary. Theists typically do not hold a consistent stance on this - they claim that their deity does frequently interact with (and manifest itself in) the physical world, but whisk him away to some other plane of existence as soon as skeptics show up and try to investigate these interactions. Your analogies don't work here, because all of them involve two objects that are clearly physical. Interactions between physical entities and forces are all around us, while the non-physical is per se just another unproven, unfalsifiable, unreasonable assumption akin to "supernatural".

Univocal/multivocal existence - point taken, I misunderstood your argument there. I speed-read over the first paragraph and wondered why you'd hold the ambiguities of the word "exist" against one position featuring it, when both are to be applied to a problem featuring it anyway. Reading it again more carefully, that's not the issue here. This argument also seems to hinge on the mathematical one preceding it, which I still haven't tackled, so let's put it aside.

Russell's teapot - The whole essay is supposed to present a position on the burden of proof applied to the existence of god, so this is closely related to my first point about actually imagined versus possible deities. True theist assertions, as opposed to vague abstractions, happen to come in forms that do clash with evidence as well as logic. You can't blame Russell for straying beyond a purely hypothetical situation when he was presenting an analogy for very prominent real-life assertions. You say there is prior evidence against Russell's teapot, and you're right. But the same thing holds for the assertions of every believer except the most extreme deist. Since you're also trying to apply your theory about the burden of proof to the existence of gods, the reality of such claims needs to be taken into account. Röstigraben (talk) 15:41, 1 July 2011 (UTC)

Agnosticism - well, forget about my own religious beliefs, consider liberal Christian theism (the kind of characters you might find at the liberal ends of the big Protestant denominations), or the Christian component of Unitarianism/Universalism/UUism. They still believe in God (although some of them have moved over into agnosticism/atheism, I'm talking about the ones who keep that belief). But they will drop or strongly doubt many of the traditional Christian beliefs that non-believers reject - they'll reject the idea of the Bible being fully the word of God (maybe God inspired some bits, but the objectionable bits are of purely human origin), be doubtful of miracles, etc. Maybe some of them are closer to deism than theism, but deism and theism actually exist on a continuum (their "God intervenes rarely" theism is pretty close to deism, the "God intervenes at our 9am service every Sunday" theism of the Pentecostal is much further from deism). So, anyway, my point is that there are real forms of theism out there (other than my own) which avoid a lot of the baggage to which you object, and hence it sounds like you are saying you must be agnostic towards them.

Likewise, ask the proverbial woman/man on the street for their opinions on religion, I'm sure you'll get quite a few people who say things like "I believe in God, but not sure about the rest of it". Ask them a question like "What happens after we die?", one is likely to get a response from such a person like "I'm sure we go somewhere, I'm sure we don't just cease to exist, but I don't know what happens exactly". I think this kind of vague theism/deism is quite common in contemporary society, although you are unlikely to find it by reading books, etc., because these people don't have strong opinions to express and are unlikely to write a book (or even a webpage) about them. So I think this is another example of a belief in God which avoids a lot of the baggage to which you object.

Anyway, my principle is "Neither believe nor disbelieve in God unless you have good reasons to do otherwise". The baggage to which you refer could well be a good reason to do otherwise, and hence you'd be justified by my principle in disbelieving in God defined in that way. But, other versions of God lacking that baggage, you aren't so justified. If you agree with that, you actually agree with the principle I've been expounding.

Occam's razor - However, I'm unconvinced about its applicability to what you refer to as more general "positions". You just state on faith that we should prefer the simpler one, without giving a good reason for it. OK, let me put it this way - which position better describes your attitude towards the burden of proof - the existential or the uniform? And, what justifies your choice of that position rather than the other?

Let me justify Occam's razor in this way - we have two choices A or B. We have no strong evidence to favour one over the other. But, we have to pick one of them, or else we'll end up in a Buridan's ass situation. One way out of a Buridan's ass scenario is that sometimes the most rational thing to do is to make a random choice. But, here we actually have a better way - if there is something objectively better about one of A or B, even if that factor is not in itself evidence for A or B, the rational thing to do is choose the better one. Arguably, simplicity is one way one of them could be objectively better. So, applying Occam's razor is more rational than choosing at random or based on purely personal whim or fancy. And we have to make a choice here, because how can one actually operate as a rational agent if one suspends belief on which is the correct theory of the burden of proof? A rational agent can successfully suspend belief on very many topics, but not on one so fundamental as this.

You say that discriminating between general principles based on the number of propositions that we would have to accept as a result is an abuse of Occam's razor, since the simplest/best position is believing nothing (you suggest postmodernism; how about Pyrrhonian scepticism). But you are missing an important point - Occam's razor is used to choose between positions ceteris paribus, i.e. when we have no good independent reason to choose one over the other. To avoid ending up in the believe-nothing case you describe, all we need is a good reason to believe at least one thing, and there is nothing in my interpretation of Occam's razor which says we can't have good reasons to believe many things (reasons independent of any application of Occam's razor). Occam's razor is essentially a fallback principle, something you only use when you've got nothing else to go on.

Physical god - I hold it to be impossible for a (completely) non-physical subject to interact with physical reality. How do you justify your belief in that - what possible evidence could you have? If you want to claim I am aware of no good evidence that any completely non-physical subject has ever interacted with physical reality, fair enough. But when you strengthen that claim to add imposibility - how do you justify that strengthening? To say something is impossible (as opposed to just undemonstrated, or unlikely) is a very strong claim, and should require very strong evidence to justify.

My point was more, we can acknowledge the existence of different types of objects. Matter is one type of object. Mathematical objects are another. Others are ideas, languages, works of literature, computer programs, etc. (The works of Shakespeare are not a physical object, since they cannot be identified with any one copy of any of them.) Now, if you want to argue that some of these depend on physical matter for their existence, well, then you are asserting a non-physical object depends on a physical object for its existence. But, does the number 17 depend on the material universe for its existence? And, since the works of Shakespeare can be converted into a number, if the number 17 exists independently of the physical universe, don't in some sense the works of Shakespeare do too?

If, as I've argued, mathematical objects must be excluded from the existential position, then why can't other types of objects be excluded too? If God is yet another type of object, why not excluded God also?

Russell's teapot - "True theist assertions, as opposed to vague abstractions, happen to come in forms that do clash with evidence as well as logic". Well, I've pointed to at least two forms of theism that arguably don't - many forms of liberal Christian theism, and many forms of popular theism.

"But the same thing holds for the assertions of every believer except the most extreme deist." As I've mentioned, deism and theism occur on a continuum. There are forms of theism for which there is a lack of contrary evidence. For example, "God intervenes in the world in ways subtle and undetectable". Suppose I was to suggest, that on rare occasions, God intervenes in quantum indeterminacy, in order to achieve a providential outcome. Such intervention would actually be impossible for us to detect. Surely that is a form of theism? Theism can proceed by two ways, big miracles, or tiny providential interventions. We can doubt the first, but have no way of proving or disproving the second. (And, concerning that which can neither be proven nor disproven, we ought to suspend judgement.)

Suppose God intervenes eschatologically at the end of the world, but not right now (when will the world end? maybe next week, maybe ten million years hence). Is that deism or theism? That claim is not contrary to any available evidence, but it seems to me to be more theism than deism. (((Zack Martin)))^™ 22:55, 1 July 2011 (UTC)

We have a very liberal variant of mainstream Protestantism in Germany, and additionally, many people who are nominally members of a church are either ignorant about major tenets of their faith or explicitly reject them. Daoism might be another suitable example - while it does feature belief in supernatural entities, I've always taken that as optional, and the core tenets of this worldview are almost pure metaphysics. So yes, they do avoid much of the baggage (sometimes at the cost of inconsistency), but as long as they don't fall back on a completely featureless deity, there's always something that can be checked against prior evidence. I don't know if you're familiar with Victor Stenger's work, but his concept of possible gods is what I was referring to. These entities lack any properties that would allow us to question the inner coherence of the concept, are situated in a niche that makes it impossible to obtain information about them regardless of technology, and never interact with the world. They also lack any worshippers, precisely because they're obviously unable to provide what humans look for in a deity. In such a hypothetical case, I wouldn't even bother to raise the question about burden of proof, because I'd know that the very possibility of proof has already been explicitly excluded from the properties of the deity. That's a position that does come rather close to agnosticism, but it does not entail giving up on the principle of burden of proof as applied to religious statements. You will find that in order to attract believers, you have to offer them something tangible, leading sooner or later to the formulation of testable statements about the deity. The price you pay for inoculating Maratrea against objections is irrelevance - atheists won't bother with it because there's no concrete theist position to begin with, just a vague array of "maybes". Or, yet more general - where there's no real assertion, there can be no negation.

Apart from that, introducing an untestable, noninterventionist god into a model of the world that does not require one is a violation of the very principle we were talking about, which makes for a nice transition to the second major issue. I provided you with a description of how Occam's razor works in a scientific setting, where there is a clearly delineated aim: explain the phenomenon. That is the only basis on which we can judge which assumptions are necessary and which ones are not, which leads us in turn to selecting based on the number of such assumptions. None of this works when there is no clearly defined goal, so using the simplicity of a starting proposition as the main criterion for its selection strikes me as unreasonable. Nevertheless, time is a finite resource, and economical concerns are important, so we should use the one that allows us to deal with the claims we encounter in the most efficient fashion. Now, what we (or me, at least) are interested in is increasing our knowledge about the world. Entertaining propositions is not a goal in itself, but just a step along the way, eventually we want to be able to make a judgment on them. What you call the "existential" position will arrive at such judgements very quickly for all cases where assertions are provided without evidence, and there are quite many of those. The "uniform" position will have us set all of them aside, not judging either way, being inundated with clutter without reaching judgements. The first one strikes me as vastly more economical. You can make any number of claims about the world that are not backed up by evidence (indeed, humans are quite prolific at this), and the vast majority of them will be wrong. Should we not strive to equip ourselves with a toolkit that will serve as a defense mechanism against bullshit and increases the number of concrete judgements about the world? Now I'm sure you will argue that it's the correctness of judgements that counts, and the first position breeds the danger of producing false negatives. True, but consider that in science, no judgement is final - a premature negation can be reversed if evidence does pop up after all. Having to do this in a few cases is a small price to pay for the advantage of not wasting time with the rest of the nonsense out there.

Apart from this, I will have to go on about your idea of applying the razor to results, because it is seriously inconsistent. You use it to discriminate between your two positions under conditions of absence of evidence - surely the same thing holds for all other choices about our axioms as well? Which "independent reasons" for selection do suddenly crop up so we don't have to fall back on the razor in these other situations? It is very important to pay attention to the consequences of affirming a principle, and this one would have us set a desirable number of zero for the propositions resulting from our very first axiomatic choice, which has to occur under the same circumstances. So ideally, we'd start out with something like "There is no truth", which would also be the end point of our reasoning.

I reject the existence of the non-physical for two reasons, and lack of evidence is one of them, yes. But the other is the simple problem that you'd have to come up with some sort of interface for physical/non-physical interactions, which I find inconceivable. Maybe that's due to lack of imagination, because I simply don't dwell on such speculations. Yes, I do think that all of the "non-physical objects" (which I'd actually subsume under "ideas") you described depend on matter for existence, if for no other reason than human brains are made out of matter, and the intellect that emerges from it is in turn the foundation for these ideas (this realiy is another argument that is quickly unraveling into a falling tree question). Is it proper practice to apply the same notion of existence and "position" to all of these classes of objects? While that is an important question, it would lead off on a tangent. This early in the morning and without coffee, I can only offer you a conditional No. To me, mathematics is a concept to make sense of (material) reality, one that we can justifiably assume to be true because of its internal consistency and usefulness, so we judge it by different standards than claims about physical existence. I still don't see how it matters, though - again, religions tend to claim that their gods are, at least sometimes, physical. At that point, it's completely justified to apply your existential position, regardless of these considerations.

Regarding your final point, you will again have to explain why you suddenly abandon Occam's razor when trying to shoehorn theist assumptions into models that do not require them. If quantum mechanics suffices to explain a phenomenon, we do not need to assume a god that tinkers with it. The same thing goes for theistic evolution, a more prominent example - I have no real quarrels with people who affirm the historical reality of evolution, but say that god guided the process, I just wonder why they have to include that latter part at all. So why suspend judgement all of a sudden? You used Occam's razor to arrive at your position ("concerning that which can neither be proven nor disproven, we ought to suspend judgement"), but you're apparently unwilling to take the next logical step and use the same tool in a situation where you have nothing else to go by. That is another inconsistency, and coupled with the one above, makes the parts of your argument that rest on the razor unsalvageable. Röstigraben (talk) 06:10, 2 July 2011 (UTC)

Relevant Gods - You suggest that all possible deities must be either able to be disproved or else irrelevant to any possible believer. One issue you ignore is an afterlife. If there is a non-interventionist, deist God, who nonetheless offers us a life after death, that would be a God offerring something of value to many people, but which cannot be disproved. The value is not just in not wanting to cease to exist, but also that an afterlife might remedy many of the wrongs of this world, provide an accounting or justice. Yes, we should do everything we can to alleviate suffering and prevent evils in this world of ours; but no matter how hard we try, there will be things we cannot change. So, the idea of a God who will remedy these injustices in the afterlife and/or eschaton, and ensure that good really does prevail in the end, is offering believers somethign of value, but not something you can disprove. Another potential value for a believer is that God provides an ultimate objective standard of goodness, which is something difficult to support on a non-theistic worldview. God expresses her nature of perfect goodness through eschatological justice.

As I mentioned, if God intervenes in the world, there are different ways in which she could do so. She could do in blatantly obvious ways, either big miracles, or smaller miracles on a regular basis. Claims of such intervention are arguably testable - if the intervention is large-scale enough, or repeated with sufficient frequency, we would notice that. On the other hand, many God does really intervene in the world, but the intervention is a subtle and undetectable nudging. Such an intervention would be untestable. What is the benefit for the believer? Well, if God providentially guides the world, then it is possible that the suffering and evils and misfortunes we endure, contribute, in some way presently unknown to us, to the greater good; that all the evil is necessary for some good; that this is, as Leibniz says, "the best of all possible worlds"; how the evils we suffer is necessary to the greater good is presently unclear to us, but God will in the end reveal her plan. So that is the attraction to the believer.

Also, I think you misunderstand 'worship', maybe from looking at it too much from the outside and not enough from the inside. Sometimes worship is about give and take; God, if I worship you, you will give me all these good things. But, that is not the only approach to worship. There have been moments in my life, when I have seen the beauty of a painting or music or poetry, or a woman's face, when for a fleeting instant I felt like that beauty was somehow a reflection of something transcendent and divine. And the natural response in me to that feeling, is one of worship and reverence. Do I expect to get anything from such worship? No; such an expectation would be irreverent. I worship, not because I have anything of expectation in return, but because I feel I must, like the great beauty imposes upon me a duty to worship it. My point is, not all understandings of worship are about expectation of anything in return, and so a God which offered us absolutely nothing (not even an afterlife) may still draw some worshippers, although obviously not as many as those offering more may.

And, even if one worships in expectation of a return, whether that requires any testable supernatural action to receive the return depends on what the expectation is. Suppose an alcoholic decides to go to church because they feel that is the only way they can kick the habit. It is not implausible to suppose that might actually work for them. Of course, I don't want to say that one has to go to church to overcome alcoholism; I am sure there are many who have overcome it with no help from religion whatsoever. But, it is a way, and it is a way that works at least some of the time. And maybe, for this particular alcoholic, they may feel it was the only way that would work for them, and they may well be right in that belief. Now, we don't need God to explain how this works; we can point to elements of community, belief, values, spiritual practices (prayer can have entirely natural psychological benefits), etc. And, the alcoholic might well say, God was working through the medium of those natural factors. But, take the belief in God out of the equation, it may not have worked for that person, even as without God it might work for others. Maybe that person has deep psychological needs (e.g. for an afterlife, for an ultimate system of values, for the universe to ultimately respect those values, for suffering and misfortune to serve a providential purpose) which only religion can fill. That is not to say that someone else might not have those needs, or not have them to the same degree, and thus have no need for religion.

Occam's razor: Why should Occam's razor only be restricted to scientific settings? As I said, Ockham himself clearly meant it to have much broader application - he expounded it in a work on Christian theology. Now, maybe you want to have your own principle, that is like Ockham's, but only applies to the context of science. Fair enough then, but it should not really be called Occam's razor, it should be given another name.

Nor surely did Ockham justify his principle in the way in which you justify your version of it. Your justification for his principle assumes a context of scientific experiment; he expounded it in the context of a Catholic theology textbook. Given that William of Ockham was a theist, it seems strange to deploy his razor against theism - was he deeply self-contradictory, or have you failed to understand what his principle actually means?

Ockham's razor is commonly given as Do not multiple entities unnecessarily (entia non sunt multiplicanda praeter necessitatem). You interpret unnecessarily as unncessary to providing an explanation to phenomenon, but that supposes the only purpose of human thought is to explain phenomena. If human thought also serves other purposes, then an entity may not be necessary to explain phenomena, but may be necessary to some other end. Given that Ockham was a Christian theist who believed that God was a necessary being, I think he would have agreed that necessity functions in a broader sense than the narrow sense you try to restrict it to.

Entertaining propositions is not a goal in itself, but just a step along the way, eventually we want to be able to make a judgment on them. Why? Why must we make a judgement on them? Why not just suspend judgement until we get further evidence? How is making a judgement better than withholding one? Sure, if we have a good reason (such as evidence) to reach a conclusion, let's reach one. But, we shouldn't jump ahead to conclusions just out of some opposition to withholding judgement, because there is nothing wrogn with doing so.

The "uniform" position will have us set all of them aside, not judging either way, being inundated with clutter without reaching judgements. How does not having an opinion on a topic cause us to become cluttered? If there is some particular reason why I need to make a decision, it may pose a problem. But, if there is no pressing need to decide, there is no harm to me in refraining from making a decision. For example, I have heard many theories about who killed JFK; but I have no opinion on who actually did it. Since it makes no difference to my life who killed JFK, there is no harm in me refraining from deciding on this question. Likewise, many agnostics will say, whether or not there is a God will not change how they live, so they have no pressing need to decide whether there is one. Why must I make up my mind about every controversial issue I am presented with? Why not just restrict myself to those issues in which I have an interest, or which are relevant to my life, and which I feel I understand well enough to feel confident about making a decision?

You can make any number of claims about the world that are not backed up by evidence... and the vast majority of them will be wrong. You have to pay attention to the structure of possibilities in different cases. Suppose the question is about physical objects that may or may not exist, or historical events that may or may not have happened. Any possible claim picked at random is almost certainly wrong, so by choosing to reject unsupported claims, rather than withhold judgement, you are indeed likely to increase the number of true propositions you believe. However, not all issues have that structure of possibilities. Consider the claim "parallel universes exist". There are only two possibilities for that claim - it is true or false. It is not like there are 50 billion highly similar claims that could be made. So, in that case, picking an answer is just as likely to add falsehood to your beliefs as it is to add truth. The same applies to other claims, like "an afterlife exists" or "a deity exists". (Yes, there may be a gazillion possible parallel universes, or afterlifes, or deities; but the question is not about which of them exists, it is simply whether any of them do; so while the question of "which one" has enourmously many answers, the question of "whether any" has only two.)

Apart from this, I will have to go on about your idea of applying the razor to results, because it is seriously inconsistent. You use it to discriminate between your two positions under conditions of absence of evidence - surely the same thing holds for all other choices about our axioms as well? Some positions do not require evidence, because they are self-evidently true or false. For example, "1+1=2" is true, "1+1=3" is false. Ockham's razor is only for use when (1) the decision between the options is not self-evident and (2) there is no other evidence. William of Ockham said the same thing when he wrote “For nothing ought to be posited without a reason given, unless it is self-evident (literally, known through itself) or known by experience or proved by the authority of Sacred Scripture.” (Commentary on the Sentences of Peter Lombard, as quoted in Stanford Encyclopedia of Philosophy).

Physical objects - I do have responses to what you say, but I think this is getting tangential from my original point. My original point was, if we are justified in using different burdens of proof for different types of objects (i.e. mathematical vs physical objects), why not treat yet further categories (e.g. deities, souls) differently. I don't agree with your materialism, but if we are talking about burden of proof, we can't assume materialism, because logically the burden of proof must come prior to materialism. Otherwise, we could not talk about what materialism's burden of proof is. So, to establish rules about burden of proof, we have to assume it is possible for genuinely non-material things to exist. Maybe, once we have gotten our burden of proof, we can move further along with our reasoning and maybe disprove that assumption.

Divine quantum intervention - If quantum mechanics suffices to explain a phenomenon, we do not need to assume a god that tinkers with it. Again, if our only purpose is to explain things, such a supposition is useless. But if we have other purposes than just providing explanations, such a supposition may help one of those other purposes. Do not multiply entities unnecessarily - if this "multiplication" meets a need (even if it is not a need for explanation), then it passes Occam's razor. So, contrary to your claims, I am not inconsistent in my use of Ockham's razor after all. And my point in this case, is to show, there are ways that God could really intervene in the world (thus being theism), yet do so in such a way that no one could detect it scientifically. (((Zack Martin)))^™ 10:56, 2 July 2011 (UTC)

1. The afterlife that many prominent religions foresee for their followers is not only a promise of eternal bliss, but very often comes in the form of an explicit threat of eternal torture. If the nice afterlife offers believers something of value, what about the nasty one? And is it morally justifiable to instill fear of such a fate in small children, just so they will stick to arbitrary and often seriously outdated moral codes?
2. I'm not someone who thinks that we need an "ultimate objective standard of good", I even think finding something like that is impossible. But in any case, the idea that the deities described in "sacred" stories all over the globe could provide such a standard is ridiculous, because their actions are very often evil by almost any modern standard.
3. "Is it not enough to see that a garden is beautiful, without believing that there are fairies at the bottom of it, too?" - Douglas Adams. We all experience moments of perfect serenity, love, amazement and the like. In me, that instills not a reverence for any deity, but for my fellow man and the world surrounding us. If you want to worship something, might as well worship the human spirit, especially one which is willing to take the world and our place in it as it is, without entertaining small-minded comforting illusions. There's something grand about beings who face infinity, insignificance and oblivion, yet just shrug and say "Oh well, let's make the best of what we've got."
4. I am well aware that their might be other applications for Occam's razor apart from the scientific one, it's just that I supplied you with the justification for this use and asked for a similar one, which you still do not provide. Simplicity is not a self-evident value like "correctness", you need to show the particular benefits of your usage. If we're talking about how to obtain better explanations (which is the job of science, after all), the usefulness becomes obvious, but if all we have to go by is a vague reference to "other purposes", it's hard to see how we would select our tools at all.
5. I am very much interested in increasing my knowledge about the world, and that necessarily entails making judgements about what to accept as facts. A small number of false negatives that can be corrected later is an absolutely acceptable price to pay for increasing the number of judgements we can work with in the meantime. You don't have to accept this - in fact, there are many people who want to maintain a sense of wonder and deliberate non-knowledge. But I've never understood them, and they sure as hell are not the ones who lifted billions of people out of desperate poverty, sickness and misery.
6. A proposition about which you remain agnostic represents uncertainty. Perfect agnosticism entails either assuming a fifty-fifty chance (for dichotomous possibilities) or saying that you are unable to calculate any likelihood. I suppose your version is closer to the latter, because the first one would involve concrete cost-benefit-analysis and have an impact on your behavior (like Pascal's wager).
7. The main reason why it is possible to come up with an infinitely larger number of false claims is because objects have a set of properties that is limited compared to the number of statements you can make about their features. There is only one correct statement which you can make about the color of a red car, but a vast array of false ones. Likewise, you can claim the existence of a number of universes ranging from one to infinity, and only one will be right. The existence of simpler, dichotomous possibilities does not make the more complex ones and their myriad additional possibilities for errors vanish.
8. You're not getting one of my central points, or deliberately evading it. How do you discriminate between "There is truth" and "There is no truth" based on evidence? That is a very fundamental axiomatic choice (the rejection of solipsism, which you claim to have performed as well, would be another one). One undoubtedly results in a much simpler system of results, yet is clearly undesirable. While affirming a principle´s universal applicability, as opposed to the limited one I provided, will allow you to make easier judgments (ironic, no?), the drawback is that you will keep running into problems like this. In fact, another way of thinking about this is to introduce yet another position, let's call it the "denialist" one: "All claims are false." Simple, beautiful, self-refuting if it were not for the non-validity of logic that it also entails. According to your conception of Occam's razor, it's superior to the others.

And forgive me if I will keep further responses rather short. I have comprehensively outlined my position above, and I believe we have again reached unbridgeable divides regarding many central points. As much as I love arguing, I can't invest several hours on a regular basis to debate points with no hope of an eventual meaningful conclusion. Röstigraben (talk) 18:11, 2 July 2011 (UTC)

1. Part of why people want to believe in a nasty afterlife, is suppose you or a loved one have been gravely wronged (e.g. rape,torture,murder) by someone who got away with it without punishment, or you felt the punishment was insufficient. The idea that God will punish them in the afterlife is attractive. But, I don't believe in everlasting punishment for the wicked. I believe, essentially, that everyone goes to heaven, without exception. Now, people who have done gravely wrong things (not little wrong things, but big ones like rape and murder) may go to Hell to be punished, but only for a limited time. How long they will stay there depends on how serious their wrongdoing - Hitler will probably be there for several million years, at least. But even Hitler's crimes, while very evil, are not infinitely evil, so at the end of however many million years, Hitler will be let out of hell and into heaven. So, the objections you make about people scaring children with threats of hell don't apply to my beliefs. And I am not alone in such a belief, belief in universalism is common among liberal Christians (see Christian Universalism). The fact that many theists have objectionable beliefs is not actually an argument against theism, when other theists don't have those beliefs.
2. If you don't feel you need an ultimate objective standard of good, fair enough then. But the fact is, other people feel they do. Remember, I was explaining to you how a God which cannot be disproved by science/history/etc. could still meet the needs of religious believers, because you claimed that a God who cannot so be disproved could not possibly meet any religious believer's needs. The fact that traditional religious stories or texts contain questionable moral standards is no argument against theism, or the claim that God can be a source of morality. It is simply an argument against those texts/stories and the claims that they are of divine origin.
3. My point was to explain how one can worship a deity without needing any expectation of a reward. It is opposed to your claim that an untestable God can't offer us anything any human would be interested in. The fact that similar experiences do not move you to worship is no argument against the fact that those experiences do move me, and others like me, to worship.
4. So, you actually only accept Occam's razor applied to scientific questions (or at least, explanatory questions), because for you it is not a self-evident principle, but rather a principle requiring justification, and it is only for those types of questions that you are able to see a justification for its use? Well, on the contrary, to me it is a self-evident principle; I am quite sure William of Ockham would have agreed with me on this. So, you aren't actually following Occam's razor at all, but another principle which is actually rather different from his, and you should find a new name for your principle then. Why use Occam's razor? Suppose we have two possibilities A and B, and equal evidence between them. As I have said, the best choice is to withhold judgement. But supposing that is not possible, we have to pick one. We could pick one totally at random, say by flipping a coin. We could pick one based on our own personal prejudices or preferences or whims. We could pick the simplest one. Now, of these three choices, which one is the most rational? Intuitively, choosing the simplest sounds more rational than choosing at random or based on whim. That is the justification for Occam's razor.
5. As I said, I have no problem with acting to maximize the number of true propositions you believe, provided one does not bring too many false beliefs along with it. (Believing 50 true propositions and 1 false proposition is better than believing 1 true proposition and 0 false propositions; but not so believing 500 true propositions and 500 false propositions.) But, as I said, one has to look at the types of propositions one is talking about, and the structure of possibilities associated with them. If there are 500 possibilities of which only 1 can be true, then by disbelieving all but one you believe at least 498 true propositions and at most 2 false ones. But, if there are only 2 possibilities, then by disbelieving one, you have somewhere between 0 and 2 true propositions and somewhere between 0 and 2 false ones. So, what you are suggesting, is only appropriate when the possibilities are structured in a particular way, but the possibilities with regard to "does a God exist?" are not structured in that way.
6. If I am agnostic with respect to a proposition, that would mean I adopt a Bayesian uniform prior. So, if there are only two possibilities, 50-50 each way. If there are more than two possibilities, the probability I assign to each one will be less, but still all the probabilities will be equal.
7. You are confusing some separate questions: "is there an X?", "how many Xs are there?", "which Xs are there?". With respect to parrallel universes, gods, whatnot, you are right that the second and third questions admit a near infinity of different possibile answers. But the first question admits only two. We shouldn't treat the first question as if it had an infinity of different possible answers, because it doesn't, it only has two. You say The existence of simpler, dichotomous possibilities does not make the more complex ones and their myriad additional possibilities for errors vanish. Right, but I'm not claiming that. The different questions have different structures of probabilities, and thus must be treated differently. Your approach may well be the right one to the second and third questions. But, that does not mean it is the right approach to the first one.
8. How do I distinguish "there is truth" from "there is no truth" on the evidence? The first is self-evident and requires no evidence. As I mentioned earlier, Occam's razor does not apply to a decision about self-evident propositions, only about propositions which are not self-evident. Same goes for your position "All claims are false" - that claim is self-evidently false, and hence Occam's razor will not be used with respect to it. (Alternatively, we might say it is self-refuting, and that it is self-evident that self-refuting claims are false.) Self-evidence is a type of evidence, and Occam's razor is only applicable to cases where there is no evidence.

(((Zack Martin)))^™ 00:14, 3 July 2011 (UTC)

So, after all of this back and forth, we find that your argument simply rests on a highly idiosyncratic version of Occam's razor. Let me summarize:

• In situations of empirical uncertainty, your razor would have us discriminate between statements based on general simplicity - not just their own, but also that of the results they produce. Despite the sweeping consequences of such an application, the only justification given is that it is "self-evidently" reasonable, "sounds rational", and that you believe Occam himself would've agreed with you.
• You'd like to apply the razor to the choice between positions which you outline in your essay, but it is apparently to be discarded after one use. Indeed it has to, because you redefined it in a way that would have drastically negative consequences in other settings. You would like to spare our axiomatic foundations from receiving this treatment and insist that these choices can also be made on evidence. How you would do that when faced with intractable dilemmata like rejecting solipsism, the materialism/idealism choice, or establishing or rejecting the very validity of logic itself is beyond me. Not all of our axioms are self-evident, in fact - as you no doubt have found out for yourself - you will have take some of them on faith, which is usually the ultimate cause of disputes. And the simpler ones are not always the more reasonable.
• You reject the principle's scientific usage as overly narrow, and when I think about your own theories that you have outlined elsewhere, I think you'd also like to implicitly discard it at all, because of what it would do to them. It is rather strange to occupy a position that directly contradicts some of your other beliefs, but then, so is the spectacle of a theist promoting agnosticism.

Seriously, when facing criticism, you should at some point take a step back, not be too hung up on defending your conclusion, and reconsider some of your argumentative steps instead. It's a bad sign if you're forced to adopt nonstandard definitions of key concepts in order to make an argument work. Don't take my word for it - I'm sure there are a lot of discussion groups out there that would love to take up a topic like this. A formal (and consistent, of course) rejection of the burden of proof could even get you published if you bring it into the right shape. But I would be massively surprised if people with more experience in this did not immediately zero in on the same deficiencies, and most likely find a few others as well.

But I take heart in that Maratrea will be granting me eternal bliss no matter whether I deny her, blaspheme her, or imagine her in hot goddess-on-goddess action with Aphrodite. That is, however, not a very good selling point when you're trying to recruit new believers. Priests all over the world have consistently shown that the carrot-and-stick approach works much better. Röstigraben (talk) 08:12, 3 July 2011 (UTC)

Well, to give my own summary:

1. You argued earlier that all gods must either (1) contradict the evidence available to us, or (2) be so irrelevant that no one could possibly be interested in worshiping them. I have indicated multiple ways in which a deist God, or an theist God whose intervention is so subtle as to be undectectable, can nonetheless give benefits which would attract believers. Since you haven't responded further to that argument, I take it you agree your original contention was wrong?
2. I think it is fair to say we have different understandings of Occam's razor. My own view on this, is that most people agree that in some abstract sense, when all else is equal, the simpler idea is more likely (but not certain) to be the correct one. But, when we try to move from that general principle down to more precise formulations which can be applied in practice, we get disputes about which precise formulation is the right one.
3. One issue here, is how to decide which formulation of Occam's razor is correct? Since I understand Occam's razor as a basic principle of rationality, I think the correct formulation must be established by judging it against an objective standard of rationality. However, like an objective standard of morality, it is easy to agree that there is such a standard, much harder to agree on what its actual contents is. By contrast, you seem to see Occam's razor, not as a basic principle of rationality, but rather as something which is justified pragmatically.
4. I think, if we are arguing about "Occam's razor", we need to pay attention to what William of Occam's actual views were. How did he understand it himself? Of course, he is no inherent authority, and you don't have to agree with his conception of it. But, if your conception differs too much from his, it is fair to suggest that you don't actually believe in Occam's razor, but rather some other razor, maybe one drawing some inspiration from Occam, but not actually Occam's.

Maratrea doesn't mind if you deny her or blaspheme her. If you don't believe in her yet, it's because she doesn't want you to yet; you'll start believing in her when she wants you to, no sooner and no later. As to imagining her in hot goddess-on-goddess action, what makes you think she would be offended by those thoughts? Maratreanism doesn't hold there is anything with homosexuality.

Historically the stick may have had its successes, but I believe in being honest and teaching what I really believe, rather than just making stuff up to attract an audience. I have no stick, and I'm not going to make one, no matter how much it would attract people, because I don't believe in it. I'm not selling out. Besides, I think in today's world there is a big demand for religion with less stick and more carrot. (((Zack Martin)))^™ 08:37, 3 July 2011 (UTC)

No, I still think that such gods do not offer enough to spawn religions. Case in point: Maratreanism, number of known followers: one. You say there's a big demand for such a religion, so why aren't these people flocking to your faith? It can't really be the heavy competition, because the major players among the religions all make extensive use of threats. And a last point regarding Occam's razor - I'm not too concerned with divining the original thoughts of someone who's been dead for the better part of a milennium. The principle he outlined works great in the setting I described above, but I remain unconvinced about treating simplicity as an end in itself. Rationalism, as opposed to traditionalism, does not ascribe any value to ideas just because they were handed down by an authority or have been around for a long time. Röstigraben (talk) 08:49, 3 July 2011 (UTC)

Well, it's not just Maratreanism that is like that. Deism, many forms of liberal Christianity/Judaism, etc., fit the bill too. Universalist thought, while it has always been a minority in Christian history, has a tradition going back to the early church. The explanations I have given, as to how a deist or subtle/undetectable theist God, could attract worshippers, are ones many liberal Christians, Jews, Muslims, etc., would agree with. My religion does not have very many followers (you are underestimating, but not hugely so), but these other religious traditions have many more. So I think they disprove your claim. (((Zack Martin)))^™ 09:04, 3 July 2011 (UTC)

The thing is that many of these belief systems are close to the kind of pure deism that I mentioned in the context of "possible gods", but not quite the same. A possible deity is one that never interacts with the world in any manner, period. Even the most liberal Christian or Jewish theist will amend that concept to at least include some small measure of intervention and providence. And another notable feature of such strands is that the more liberal they get, the more you also have a collection of individuals who pick and choose their beliefs, instead of the strong organizational features and high doctrinal unity that are usually a hallmark of what we call "religions". For example, I remember discussing religion with my (nominally) Protestant great-aunt, who is the only self-described theist even in my extended family. It turned out that while claiming to be Christian, she casually rejected even the core tenets of this faith - most notably, she does not think Jesus was the son of god. I think there are lots and lots of such people in post-religious societies, even among the ranks of the major congregations you will find many who actually are "spritual, but not religious", they just can't be bothered to take their beliefs to a logical conclusion and exit the organization. Or maybe they've just grown too attached to the social experience and setting that surrounds the meaningless rituals and prayers. Anyway, what they typically hold is not a concrete theist concept, but an extremely vague and constantly shifting set of beliefs about entanglements between nature and the supernatural, with no clear delineation between the spheres or a clear idea about how they interact. In a word, it's just the vague notion that there's something more to this world than what they experience. Is that religion, or a concrete idea of god that is offering them something? I wouldn't say so. Röstigraben (talk) 09:34, 3 July 2011 (UTC)

Even the most liberal Christian or Jewish theist will amend that concept to at least include some small measure of intervention and providence - yes, but if the measure is small enough, the results will be the same - for divine intervention to be detectable, it must be either (1) major and obvious (work a big miracle), or (2) smaller but frequently repeated (e.g. our church heals people every Sunday). It also must have a clearly discernible purpose - if it is supposed that God intervenes to work miracles to prove the identity of his true prophet, then we know what kind of intervention we are looking for, and can test if the signs of such an intervention are present. If instead we suppose that God intervenes to push history in a particular direction (i.e. to conform with the divine plan), for reasons which cannot yet be clear to us, it is impossible to test to if such an intervention is going on, since we don't know what to look for. If one limits the interventions to being sufficiently small, sufficiently subtle, or sufficiently occasional, you can end up with a claim of intervention which it is impossible for science to disprove. As I mentioned, one can support a theory of providential intervention on the basis of a series of small undetectable nudgings. Not proveable, not disproveable, but enough to satisfy the need to believe in God's intervention in the world - well, maybe not everyone who feels that need, maybe not even most, but certainly some.

I don't think your great aunt is the best example to use of liberal religious thought. What about liberal religious ministers? Or liberal theology academics/students? They exist. I don't think they would agree with your characterisation of them as too lazy to take their beliefs to the logical conclusion. Obviously, if one is a minister, or a theology student or academic, one is serious enough about your religious beliefs to have chosen to devote a big chunk of your life to it - you aren't just blindly going to walk down that path simply out of habit or lack of thought, you have very likely thought rather deeply about your choice to go down that path.

The point remains, there are versions of theism (not just deism), which do not conflict with any proveable claims of science in their implications, and which manage to attract a reasonable number of believers (not a huge number, but not a miniscule number either). (((Zack Martin)))^™ 09:59, 3 July 2011 (UTC)

Oh come on, now it's really getting repetitive. I never doubted that it is possible to bring theist conceptions in line with scientific knowledge - as opposed to the method used to derive it. You can do that in two ways - resort to a god of the gaps argument and commit theologians to constantly update their ideas as science advances, or shove the deity into a niche that will forever be removed from scientific knowledge. The latter position is what I was talking about, but it is usually unavailable to established religions because it would necessitate massive changes to existing doctrines and myths. It is obviously possible to do when you're starting with a clean slate - that's what you have done, after all. The scientist will still reject it as an unfalsifiable hypothesis, and the prospective follower will not be too thrilled with a god who works through quantum-level events. I've read apologetics that go in a similar direction, and haven't really seen them catch on. Which is unsurprising - your average Christian believer, for example, is not just ignorant of theological "innovations", but even the foundational texts. There aren't many Christians who have actually read significant parts the bible and could tell you what's written in there, let alone offer an interpretation. I bother with the reality of organized religion and its influence, not some idealized version of it or tiny cults with next to no adherents.

And the point of my example was simply that in many modern societies, religion is unravelling into a chaotic tangle of individual conceptions of the supernatural, many of which no longer qualify as theism. As much as I'd like to believe that people are getting more rational, most of them just latch on to a different set of unfounded beliefs instead and grab whatever they can find from an inexhaustible pile of irrational ideas. That, however, is no longer religious thought in a meaningful sense of the word, and I strongly doubt that any would-be ministers, rabbis or imams seriously entertain thoughts that are as detached from the central tenets of their faith as my aunt's. Why should they embark on this career path when they don't believe most of what they're going to be taught and are expected to teach in turn? If you can reject the divinity of Jesus and still be a Christian, the word has no meaning. Röstigraben (talk) 11:03, 3 July 2011 (UTC)

If you can reject the divinity of Jesus and still be a Christian, the word has no meaning - well, then the word really does have no meaning, because there has been a long history in Christianity of a minority doing exactly that... read e.g. WP's Unitarianism article for a historical overview. There have been Christians who have denied the divinity of Christ all the way from the days of the early church to the modern times. It has always been a minority, but clearly the idea of Jesus as divine is problematic enough for many people that attempts have been made to drop it from Christianity even from the earliest days.

and I strongly doubt that any would-be ministers, rabbis or imams seriously entertain thoughts that are as detached from the central tenets of their faith as my aunt's - well, how then to you explain the phenomenon of religious ministers and theologians who become even more detached than she has? Consider Death of God theology. Or consider Anglican bishop John Shelby Spong's statement of his basic beliefs, which includes such standards of orthodoxy as "Theism, as a way of defining God, is dead. So most theological God-talk is today meaningless. A new way to speak of God must be found" and "Since God can no longer be conceived in theistic terms, it becomes nonsensical to seek to understand Jesus as the incarnation of the theistic deity. So the Christology of the ages is bankrupt".

The problem with your approach here, is you ignore and discount (possibly from unfamiliarity) entire traditions of serious religious liberalism. (((Zack Martin)))^™ 11:21, 3 July 2011 (UTC)

I stand corrected re:Unitarianism. While that is interesting, it reminds me of a Communist party embracing the idea of private ownership of the means of production, while still retaining its name. It's doable, but definitions tend to erode when they are stretched. The other part, however, is exactly what I was talking about - in fact, the "Death of Theism" is a perfect name for the trend I was describing: people reject established doctrines and make up new ones that often no longer feature concrete supernatural entities. I didn't know established theologians were willing to go as far as to explicitly reject theism, but it strikes me as rather strange that they're still willing to maintain a belief in god and the authority of parts of scripture at the same time. If we understand theism as the most elementary affirmation of the existence of a god, then that is surely not what they want to deny. Statements like "Theism, as a way of defining God, is dead. So most theological God-talk is today meaningless. A new way to speak of God must be found." strike me as exactly the kind of vague generality that is so characteristical for modern irrationalism. They are aware of the evidently nonsensical nature of traditional religion and its highly specific claims, but they do not want to go all the way and deny this position. So what they do is to simply liquidate all central concepts and definitions that made it a clearly distinguishable position in the first place. Where there's no position, there can be no counterposition, so it's impossible to argue against it. If they're willing to at least take a stand on human rights and rid their teachings of the most harmful commandments, more power to them - I don't have a problem with religionists whose actions are indistinguishable from those of humanists. But, again, they are perpetuating a trend that is undermining the very concept of a religion. Well, good luck with finding followers for your own beliefs, but given your position, it's going to be hard. The "spiritual" people will just make up their own beliefs and follow them, the fundies occupy the opposite extreme of rigid dogmatism and literalism, and the rationalists will shake their heads at this whole spectacle. There's not much space in-between, so you'll have to settle for a very small niche. I still think your time and abilities could be put to much better uses, but if you're content with that, fine. At the very least, this discussion was definitely more enjoyable than watching the latest HCM unfold. Röstigraben (talk) 13:25, 3 July 2011 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ I still think your time and abilities could be put to much better uses - I'm sure my time and ability could be put to better uses than editing RW, but then I'm sure yours could too :P (((Zack Martin)))^™ 05:10, 4 July 2011 (UTC)

I was referring to the design and promotion of new religions. Debating ideas is always time well spent, and this is a pretty good place to do it. Functionally, RW is mostly a chatroom/discussion forum with an attached article section (and a well-staffed drama department). But that's actually a strong point - discussions can spring up in lots of places, and can be triggered by article changes, news items or even some idiocy that happened at CP. Röstigraben (talk) 06:11, 4 July 2011 (UTC)

Are the the design and promotion of new religions and debating ideas really that different? The debate of ideas advances by people coming up with new ideas (design), people stepping forward to defend them (promotion), others stepping in to oppose them, and on it goes. New religions just means new ideas of a religious nature. If the debate of ideas is good, surely the debate of new ideas is good; if the debate of ideas is good, surely the debate of religious ideas is good; thus, surely, the debate of new religious ideas must be good too? (((Zack Martin)))^™ 06:18, 4 July 2011 (UTC)

Depends on your interests - we've been through your religious ideas before, and I don't really care much about theology. Your essay was free of such references, and we mostly avoided them during this discussion as well. But apart from that, you also need to keep the likelihood of success in mind. Converting people to Maratreanism is a tough job for the reasons I gave above, to the point where I'd say it's a waste of time. On the other hand, when you're debating a non-religious idea, success depends on the soundness of your argument and the quality of evidence provided. You'll have a much easier time winning people over to your position on agnosticism than to your religion. Even though your essay is flawed, it makes a whole lot more sense than the latter. Röstigraben (talk) 06:36, 4 July 2011 (UTC)

To me, my position on agnosticism is actually a step on the way to my theology. I have let people know what my theology is, but I haven't put much effort into trying to convince people of its truth, because I realise that a direct approach like that isn't going to work. Instead, I try to establish some philosophical/theological ideas (e.g. idealism, theism, etc.), which aren't really specific to my theology at all, a lot of others (e.g. many Christians) believe them. But, I need to establish them, before I can establish any of my more unique beliefs, which is why I concentrate on arguing for that common ground rather than the unique beliefs.

The argument about burden of proof is actually going somewhere. The next step is to justify belief by faith. But, my view, is that faith is only permitted where the evidence is balanced. So, in order for my argument in favour of faith to advance, I first have to reset the burden of proof to be equal for each side. Once I have validated faith, then I can try to justify by faith theism, and even other more particular views of mine. (((Zack Martin)))^™ 06:48, 4 July 2011 (UTC)

Fuck this.

Some PDF link and run? Must . . . Resist . . . Urge . . . To . . . Vape . . . . MtD^{Pinko Scum} 12:49, 29 June 2011 (UTC)

ListenerX's remarks

I Wiki-formatted your essay as a way of reading through it closely. Naturally I have a text-wall's worth of remarks. First, the minor points:

1. You seem to have copy-pasted part of your definition of "atheism" from the previous bullet-point. Also, your definition includes only a subset of atheism, strong atheism. As most atheists like to pretend they are not strong atheists, you might want to alter that; a few years ago, I caught a lot of slop here on the Wiki when I tried to use strong atheism as a starting definition for an argument.
2. You say "contradiction" in several places where you mean "negation."
3. In the "Russell's Teapot" section, I believe you meant to say "in fact the burden is on the theist to disprove the atheist."

And the more substantial ones:

1. Your definition of "God" excludes most pagan deities, including all of mine.
2. There is a problem with your position of "weak agnosticism" as the default. The distinction between "thinking" and "knowing" X is dodgy, and often dependent on whether or not X is actually true; but at the very least, people do not "know" something that (1) has no hard supporting evidence, and (2) is questioned by a large segment of society.
3. Using Occam's Razor as you do is questionable; while the "uniform position" may be "simpler," the "existential position" is stronger in terms of what it can explain. Hence, while entities would be multiplied in switching from one to the other, they are certainly not multiplied beyond necessity.
4. I am skeptical of the arguments from mathematics, if for no other reason than that Edgar Allan Poe said in the person of C. Auguste Dupin: "I dispute the availability, and thus the value, of that reason which is cultivated in any especial form other than the abstractly logical." Poe's argument was that there were many axioms available in mathematics that were not available generally, but there is also the argument that in mathematics and logic one must treat these questions with a rigor that is ridiculous when treating questions about the real world, which is not all nice and tidy like a Platonic form.
   Specifically, the ways in which one can use inductive reasoning is much more limited in mathematics. One cannot conclude, from the general lack of observed counterexamples to Goldbach's conjecture, that it is true; but in the real world, where somewhat less rigor is needed, given the general lack of observed miracles, it would be a reasonable inductive conclusion that miracles are impossible.
5. The ambiguity you refer to in the "second argument from mathematics" (concurrently having to assume a statement is true because of the lack of evidence for its falsity and vice versa) is usually resolved by assuming the broader statement, e.g., "It is not true that $40,000 in cash appeared in my garage for exactly one second at 2:43 P.M. on Friday, December the eleventh."
6. The argument involving the difficult and obscure philosophical questions surrounding "several different types of existence" cuts both ways, in that one could then adopt the "uniform position" for mathematics and the "existential position" in other matters.
7. Concerning the "legal arguments," I happen to agree with one of Ayn Rand's Atlas Shrugged villains, Robert Stadler, when he said, "It is only in the realm of pure science that truth is an absolute criterion. When we deal with applied science, with technology — we deal with people. And when we deal with people, considerations other than truth enter the question." I think the world would be a better place if more philosophers paid heed to those "pragmatic and ethical considerations," and thought about what might happen in the event that they are promoting wrongheaded ideas.
8. You seem to have forgotten one relevant argument; viz., that the "existential position" (i.e., logical positivism) was thrown into very serious doubt by Gödel's incompleteness theorems.
9. Overall, you should work on how you present your ideas; here we come back to the remark from Atlas Shrugged. Admittedly, most of the Wiki's user base need to be instructed in these concepts using small English words and a large measure of patience for when they run out of actual objections and start slinging ad hominems; but even when addressing a general audience, dense mathematical arguments do more to confuse people than to communicate the point. A few sentences would probably have sufficed for each argument.

ListenerX^TalkerX 05:18, 1 July 2011 (UTC)

First of all, thanks for pointing out those errors, I have corrected it.

Atheism - The definitions I have always favoured are that an theist is someone who claims to know that God exists, an atheist is someone who claims to know that God does not exist, an agnostic is someone who claims not to know whether God exists. The 'knowledge' of which I speak here does not have to be certain knowledge, just to hold that one position is significantly more likely than the other. Now, I am aware of so-called "weak atheism", but I don't see why that should be called atheism rather than agnosticism. Why call such a position "atheism", when we already have a perfectly good term—agnosticism—to describe it. Giving atheism such a double meaning destroys the symmetry in the definitions I use.

Pagan deities - yes, agreed the definition of "God" here excludes many (but not all) conceptions of pagan deities. This essay is about a debate around a particular idea of 'God'. That idea is common in religions like Judaism, Christianity, and Islam. It is not unknown to paganism, since much of its ideas originated from the works of pagan Greek philosophers such as Plato and Aristotle. That said, the paganism of Plato and Aristotle had very different ideas about deity than many other pagans (especially even many of today's pagans). So, yes I am deliberately excluding from the scope the sort of pagan conceptions of God (gods within rather than outside of this universe) to which you refer.

Weak agnosticism: The traditional definition of knowledge: to know something you must believe it, it must be true, and you must be justified in believing it. So, the weak agnostic doesn't know whether God exists - they are lacking the belief component. Then they ask, "Does anyone else know?" Suppose someone else says "I know God exists" or "I know God does not exist". To conclude that person is wrong, the agnostic must conclude that that person is lacking either the truth, the belief, or the justification components. The agnostic, being an agnostic, can't conclude that person lacks the truth component. And, it seems fair enough to assume, that if they say "I know God exists/does not exist", they are truthful in reporting their own beliefs. So, the only way the agnostic could doubt the (dis)believer's claim to know whether God exists, would be to doubt the presence justification (you believe it, it may well be true, but you don't have a good reason for believing it.) And, if the weak agnostic concludes that justification is unlikely to be present, they have moved from weak toward strong agnosticism. But, my point is simply that, the claim that someone else lacks justification for their own beliefs requires evidence; in the absence of evidence, we should withhold judgement on whether they are justified in their beliefs, and hence remain in weak agnosticism. Furthermore, strong agnosticism is in fact a very strong claim; not just that a particular person is unjustified in their belief in the existence or non-existence of God, but that everyone having such a belief is. That is a very strong claim, requiring very strong evidence, and I don't believe evidence of such strength is available.

Occam's razor: Using Occam's Razor as you do is questionable; while the "uniform position" may be "simpler," the "existential position" is stronger in terms of what it can explain. Hence, while entities would be multiplied in switching from one to the other, they are certainly not multiplied beyond necessity. - How can the existential position explain more? It certainly would lead us to reject the existence of things which the uniform would have us withhold judgement on the existence of, but how is that explanation? You say they are certainly not multiplied beyond necessity; but what necessity does insisting upon the non-existence of something for which we lack evidence actually serve?

Mathematics: your basic point seems to be, that we can't use examples from mathematics to infer generally applicable rules of reasoning, since mathematics is so unlike the physical world, and rules inferred from the world of mathematics will be the wrong rules for reasoning about physical reality. Actually, in a way my point is the same. Because the existential position doesn't work for mathematics, it can't be a generally applicable rule of reasoning. A generally applicable rule of reasoning would work on anything, including both mathematics and physical objects. Now, the existential position may well be a good rule of reasoning to apply to physical objects. But, why should we apply a rule specific to physical objects to God, when God is (most commonly) claimed not to be a physical object? Mathematical objects are non-physical objects; it would seem that, if God is claimed as a non-physical object, the right rule to apply to God would be closer to the one applied in mathematics than to the one we apply in the real world.

When you say The ambiguity you refer to in the "second argument from mathematics" (concurrently having to assume a statement is true because of the lack of evidence for its falsity and vice versa) is usually resolved by assuming the broader statement - I am not sure what you mean by this.

Types of existence: The argument involving the difficult and obscure philosophical questions surrounding "several different types of existence" cuts both ways, in that one could then adopt the "uniform position" for mathematics and the "existential position" in other matters. Yes, but if we must assume the uniform position in one matter, why not in further matters? If a different rule for maths, why not a different rule for God too? Or, why not just go the simpler route, and adopt the uniform position for everything, rather than arbitrarily applying different rules to different questions?

Gödel's incompleteness theorems: I don't think the existential position is the same thing as logical positivism, and I don't see what Gödel's work has got to do with either of them. Feel free to enlighten me.

In terms of your comments on style, how can one present deep arguments save at length? If one aims everything one says at the lowest common denominator, then one is left at only being able to present very shallow arguments. I assume my readers are well-read people with broad general knowledge, and who have a basic knowledge of all major disciplines - or at least, people who are willing to look things up when they encounter words or ideas or references they don't comprehend. (((Zack Martin)))^™ 23:43, 1 July 2011 (UTC)

Atheism: The difference between agnosticism and weak atheism is that weak atheism entails a lack of belief in God, and agnosticism adds to this with an explicit belief that one does not or cannot know whether God exists. But in any event, it rarely helps to re-define words in this manner, which tends to confuse people, even if it brings greater symmetry to the arguments.

Occam's razor: Being able to make any assumption at all allows more valid conclusions to be drawn. In a somewhat related example, as the shysters on the RATE committee take much glee in pointing out, there is no direct observational evidence that the rate of radioactive decay has stayed constant for the last 13 billion years, just because we were not around 13 billion years ago to make measurements; if one could not take the "existential position" and extrapolate current data backwards, one could not do any radiometric dating at all.

Mathematics: My point was not so much that the mathematical proof style is invalid for real-world use — it is not, so long as one takes care not to make the wrong assumptions — but that one cannot apply the high standards of rigor (e.g., requiring an actual number as a counterexample to Goldbach's conjecture) for proofs in the mathematical world outside that world.

Preferring the broader statement: A very narrow or specific claim is "strong," logically (i.e., many other statements can be inferred from it), while its negation is proportionally "weak."

Types of existence: In the previous section, you had argued against the idea of making mathematics a special case; I was highlighting an apparent inconsistency.

Gödel's incompleteness theorems: Logical positivism holds that if there is no proof of some statement A, it is "meaningless." With regard to statements about the existence of God, "meaningless" was generally used as a synonym for "false," which seems to approximate your "existential position." Gödel's incompleteness theorems proved that, on the contrary, there are true statements with no proof to them.

Style-wise, regardless of whether or not you must ratchet up the word-count in order to explain a "deep argument," if your audience starts saying "tl;dr," you will not have gotten your point across. Scientists have this exact same problem when writing papers and reports; some of the ways they handle it are: (1) writing abstracts for their papers; (2) making the first chapter or section in a longer paper into a summary of the remaining sections; (3) cutting out the really thorny minutiae and putting them in technical reports that they then reference from the original paper.

Your essay could probably have benefitted from all three strategies, except that instead of a technical report, you could have provided the details on the talk-page if questioned by a very thorough reader like Röstigraben. To give you an idea of where most of our editors are in terms of "broad general knowledge" and willingness to look things up, I once was in an edit-war for something like two weeks because no one else on the Wiki knew about, or wanted to look up any information on, the role of the Mosaic Law in Christianity. ListenerX^TalkerX 03:25, 3 July 2011 (UTC)

I think we need to draw a distinction between different types of agnosticism. A weak agnostic says "I don't know, and I don't know whether anyone else knows, nor whether anyone will ever know". Stronger agnosticisms say things like "I don't know and you don't either, even though you think you do" or "No one knows, and no one will ever know". I suppose "weak atheism" and "weak agnosticism" are bascially two names for the same thing, but I think "agnosticism" is the more appropriate term here than "atheism". Ask someone a yes/no question (Does God exist?), the valid answers are "Yes", "No" and "I don't know". "Yes" is the theistic answer, "No" is the atheistic answer, "I don't know" is the agnostic answer. Defining "I don't know" as a form of atheism doesn't make much sense.

I think, in terms of radioactive decay, a constant rate is simpler than a varying rate, so by Occam's razor, in the absence of other evidence, we should prefer the constant rate. However, is "God does not exist" simpler than "God exists"? I am not convinced. The question about radioactive decay is a question about something that is agreed to exist, we are just asking in what form it exists, it makes sense to label some possible forms simpler and other as more complicated. The question about whether God exists, is not a question about what form something agreed to exist takes, but about whether something exists at all. The cases aren't comparable, so I don't think you can infer things about the God case from the radioactive decay case.

As to your Mosaic Law edit-war experience, I think it is sad that on a supposedly rational wiki, so many editors are so unwilling to do even basic research, and are so quick to rush to judgement on matters which they have not even made the most basic efforts to actually understand. (((Zack Martin)))^™ 07:45, 3 July 2011 (UTC)

A weak agnostic says "I don't know..." And therein lies the difference. Weak atheism also encapsulates implicit atheism, which entails saying nothing whatsoever on the topic; all newborn babies are implicit atheists, for example. "Weak atheism" may not make any sense, but it occupies a place in the taxonomy of atheism, partially because it allows atheists to sling broadsides at "faith" without obviously catching themselves in the crosshairs.

I think that the Mosaic Law business actually underscores the rational nature of the Wiki, that its editors are skeptical of new claims and insist that they be backed up with a bevy of evidence. One wishes sometimes that one's fellow-editors were better informed; but this is something to be dealt with rather than bawled about. ListenerX^TalkerX 05:20, 4 July 2011 (UTC)

To me, a baby isn't an atheist. An atheist is someone who consciously rejects the existence of God; a theist is someone who consciously accepts it. A baby can neither accept or reject the existence of God, since the baby hasn't even encountered the concept yet. A baby isn't even really an agnostic, since an agnostic, even though they honestly don't know the right answer to the question "Does God exist?", they at least know the question exists, and have at least some (however limited) understanding of it. A baby is a pre-agnostic. I think, trying to define atheism in such a way as to turn babies into atheists, is actually a cunning rhetorical strategy on the part of atheism, and I refuse to endorse it.

I think positively rejecting a claim, when you know you know very little about the topic, is not rational. The rational thing is to be honest, and say "I don't know enough about the topic to know if what you say is true". One can even make plain one's doubts, "What you are saying sounds fishy to me, but to me honest I don't understand the topic well enough to know if what you say is right or not". That is rational. Crying, "Bullshit!" whenever someone says something you can't understand, is not rational. (((Zack Martin)))^™ 05:29, 4 July 2011 (UTC)

Bashing of Discussion with Maratrean from CP:WIGO

Enjoy^img, or perhaps don't... it all depends on how much you enjoy headdesking... --Eira ^{OMTG! _{The Goat be Praised.}} 06:30, 2 July 2011 (UTC)

So we're now outsourcing RW debates to Conservapedia's mainspace? Their content acquisition is geting stranger every day. Well, still beats Ken's drivel, Andy's insights, Poor stubs, or Karabirds. Röstigraben (talk) 06:40, 2 July 2011 (UTC)

Nah, it's just he can shove that shit in to CP without about a million people dog-piling on to tell him how much of a moron he is. Or indeed anyone noticing at all. Isn't CP a gem? --JeevesMkII ^{The gentleman's gentleman at the other site} 07:52, 2 July 2011 (UTC)

Yikes, it's depressing when you're listening to this old bitch WHILE reading that edit. Yes, I cannot prove Odin doesn't exist. I'm kind of confused, though - such blatant parody and logical fail wasn't yet inserted into the article until now? Pick up the pace, poes, CP has shown that it's even too stupid to destroy itself. Norseman^{Cyser Melomel} 13:44, 2 July 2011 (UTC)

She's easier to tolerate though just about as intellectually slovenly as Maratrean. Two thumbs up. 14:29, 2 July 2011 (UTC)

Me, intellectually slovenly? Nutty, in the intellectual slovenliness stakes, you win hands down. You can't even be bothered to respond to my arguments, with counterarguments, just dismissal out of hand. (((Zack Martin)))^™ 09:09, 3 July 2011 (UTC)

You deserve nothing more than derision and dismissal in my view. I'm smart enough to have noticed that a crank like you isn't worth getting into a discussion with. You misuse key terms, multiply the length of your posts arithmetically as you struggle to refuse to yield, and introduce progressively more and more irrelevant tangents. And it all seems to be in service of propping up the philosophical bases of your insane made-up religion. I can't even be bothered to respond to your "arguments" because you're a mentally I'll attention whore who never shuts up, not because I'm particularly intellectually slovenly. It's the same reason nearly everyone else ignores you. Most people's time is better spent not being pissed away engaging the likes of a cancer like you. Why won't you go away? 11:06, 4 July 2011 (UTC)

You say I "misuse key terms", "introduce... irrelevant tangents". You provide no evidence for these assertions. Which key terms are you talking about? How is my use of them wrong? Which "tangents" are you talking about? Are they really irrelevant? You provide no evidence, no examples, no argument, for these your claims, just assertions. And then you move on to issue baseless accusations against my mental health, which since you (1) don't know me personally, and (2) don't demonstrate any qualifications or knowledge of psychiatry or clinical psychology, you cannot have any basis for either. Tell me, if I'm "mentally I'll", which diagnosis (under say DSM-IV) do I have? And, its strange you claim "nearly everyone else ignores you", because quite a few people talk to me, just look at this page. No, this post of yours just proves your intellectual slovenliness once more. (((Zack Martin)))^™ 11:29, 4 July 2011 (UTC)

Whatever makes you feel better, crank. I can't wait for more people to tire of your endless attention whoring so you move on to plague some other site. 11:37, 4 July 2011 (UTC)

If you don't like me, why not just ignore me? But, you can't resist the urge to respond. (((Zack Martin)))^™ 11:57, 4 July 2011 (UTC)

The tea pot got there after it was blasted into orbit when the Flood waters erupted, duh. --Sid (talk) 13:53, 2 July 2011 (UTC)

That's a ridiculous lie. The moon was created perfect and cratered due to continuing Angel vs. Demon fighting. --JeevesMkII ^{The gentleman's gentleman at the other site} 14:58, 2 July 2011 (UTC)

Oh, people. The man is making an example just as Russell was. He's questioning the backstory of a statement and it's content. IT IS NOT THAT DUMB OF A REMARK. --uhm, t! 16:49, 2 July 2011 (UTC)

It's pretty silly to refuse to accept the conditions of a hypothetical in order to argue against it using your own inaccurate and self-serving definitions. But I don't expect much more from creationists and their ilk in these kinds of burden shifting arguments. 17:47, 2 July 2011 (UTC)

How am I "refusing to accept the conditions" of Russell's hypothetical? I am simply arguing it doesn't demonstrate what he seeks to demonstrate with it; that it doesn't prove what Russell thinks it proves. (((Zack Martin)))^™ 08:48, 3 July 2011 (UTC)

I like how so many of you just reject my ideas about the burden of proof as false without providing any argument as to why they wrong. I challenge you to prove your allocation of the burden of proof is the right one and mine is the wrong one. What's the burden of proof in determining the correct burden of proof? (((Zack Martin)))^™ 08:13, 3 July 2011 (UTC)

conservative also posts voluminous pages of... um, "argument," but we don't worry ourselves about proving him wrong every time, either.

Still, I'll oblige for a moment.

We do have good reasons to suppose that God does not exist. Our knowledge of the laws of nature does not require a God - and indeed seems to make him superfluous.

Further, when considering the existence of an object or entity, the default position is not "undecided." I am not undecided about the teapot nor am I undecided about the existence of a new white table in my apartment. I have no reason to believe in a white table or an orbiting teapot, so I don't. Similarly, since I have no reason to believe in God - a proposition so outlandish that "reason" requires rather more than the unbacked assertions of a wide variety of contradictory sources - so my default is that I don't believe in him.--^talk 08:33, 3 July 2011 (UTC)

You say the default position is "No". I say the default position is "Undecided". You haven't actually given much reasons why your view on the default position is right and mine is wrong; you've cited two examples, and attempted to interpret them according to your view, but haven't made the case why the interpretation of them according to my view is not equally valid.

Taking the case of the teapot specifically, I say the default position is "Undecided", but we can say "No" in the case of the teapot, because we actually have good positive reasons (based on the laws of nature and our knowledge of history) to suppose it isn't there. Likewise, the white table in your apartment, you have good positive reasons to suppose it is not there (you haven't seen it; the laws of nature are such that white tables don't just materialise for no reason, if one was there someone would have to have put it there; given your knowledge of human behaviour, it is unlikely anyone would have put one there without your knowledge/approval, etc.) This is why both the teapot case and your white table case are bad examples, because the default position is about the case where we truly have no evidence either way, but for both the teapot case and the white table in your apartment, we have positive reasons to assume they aren't there. (((Zack Martin)))^™ 08:46, 3 July 2011 (UTC)

Russel's Teapot was put there by aliens, that's why human history doesn't record anyone putting it there. And I know you're going to say that we've looked for it, but haven't seen it, well it turns out that it's so small that we haven't been able to get the necessary accuracy to be able to observe it. The whole point of Russel's Teapot is that the existence of the Teapot is not falsifiable. But you're asserting that we have good reason to not believe in the Teapot... but yet assert that we do have good reason to believe in a deity. What the fuck is this, special pleadings week? Your deity and the Christian deity are just as unfalsifiable as Russel's Teapot, because it is not necessary for humans to have put it there. And if we find anything there, apologists can just say "we used the word 'teapot' because we didn't have the right words to describe its shape as it actually is". I mean, seriously. Your willingness to argue stupid shit is amazing beyond that of any apologist I've ever had the misfortune to see. --Eira ^{OMTG! _{The Goat be Praised.}} 10:49, 3 July 2011 (UTC)

Do aliens exist somewhere in the universe? We have no good evidence that they do, no good evidence that they don't, so we are best just to suspend judgement on that question. If they exist, have they visited our solar system? We have no evidence that they have. Could they have visited without us being aware of it? Yes, that is possible. But, even if they have visited, why would they leave a teapot there? Although we can only really guess at alien psychology, we'd have to assume that they do things for a purpose, and it is hard to conceive of a reason why they'd dump a teapot in that orbit. Maybe it was an accident; maybe it was a case of garbage disposal; maybe it was an attempt at humour (the aliens have read Russell's essay, and find the irony of dumping a teapot there amusing.) But adding all the pieces together, it seems that, given what we know, it is unlikely that aliens put a teapot in that orbit, and added to our knowledge of the laws of nature and human history, the overall conclusion is that is unlikely that one is there. So, once more, the teapot is a bad example of a case where we have no evidence either way, since we have good reasons to suppose it is unlikely it exists. Whereas, in the case of a deity, no good evidence for non-existence can be found. (((Zack Martin)))^™ 11:06, 3 July 2011 (UTC)

Sorry, all of the above simply means that, if we apply the standards you're applying to the existence of God to the possibility of the existence of said teapot, we have to conclude that said teapot might exist, or might not. You are simply asserting that this is not the case based on the fact that YOU believe it's unlikely that the aliens actually left a teapot there. OK, I believe it's unlikely that God exists. Why is what YOU believe a better standard than what I believe? Answer: it isn't. As such, either we must conclude that the existence of Russell's Teapot is undecided, and therefore so is God's, or we can default to the position that both do not exist, until and unless we get solid evidence otherwise. 86.161.120.49 (talk) 16:43, 3 July 2011 (UTC)

As I said, we have good reasons to suppose the teapot does not exist. If you think the reasons I have proferred are invalid, you are welcome to argue against them. And, if you claim to have good reasons to suppose God does not exist, you are free to present them. My point is simply that (1) in the absence of evidence, we should assume neither existence nor non-existence (for the teapot or for God) (2) we have good evidence that the teapot does not exist (3) we don't have good evidence that God does not exist. If you think you have good evidence that God does not exist, by all means produce it, I am waiting. Likewise, if you believe my reasoning that the teapot is unlikely to exist is in error, please point out the error of my reasoning. (The fact that it is my reasoning rather than your reasoning is totally irrelevant to the question of whether it is the right reasoning or not.) (((Zack Martin)))^™ 05:14, 4 July 2011 (UTC)

present evidence that X does not exist ? really ? Hamster (talk) 05:55, 4 July 2011 (UTC)

Yes. I've presented evidence that Russell's teapot does not exist. Can you present evidence that God does not exist? Evidence that something does not exist doesn't need to be absolute proof of its non-existence, just positive reasons to think its existence is unlikely. (((Zack Martin)))^™ 06:07, 4 July 2011 (UTC)

Could you clarify for me what a "positive reason" is and how it contrasts with a negative reason? You use the phrase a lot so I assume it has a specific meaning. By the way, it seems from this ((1) in the absence of evidence, we should assume neither existence nor non-existence) that you reject the scientific principle of using the null hypothesis. Anyway, I have a belief of my own. The Teapot is itself supernatural - it came into existence by its own will alone (being as it is, not dependent on the natural laws that normally bring things into existence). Do you accept that my particular version of the Teapot is just as likely to exist as a deity? If not, why not? ^{ONE / TALK} 08:53, 4 July 2011 (UTC)

What I mean by a "positive reason" for proposition P is any reason other than a lack of reasons for ~P. So, a positive reason to believe the teapot does not exist, is any reason beyond "we have no evidence for its existence". So, "the laws of physics imply teapots don't start orbitting the Sun unless an intelligent agent puts them there", "our knowledge of human history implies no one human has put a teapot there yet", etc. But, "there is no evidence for the teapot's existence" is not a positive reason.

In my view, absence of evidence is not evidence for abscence, in and of itself. Now, if we have good reason to expect that, were X true, then we'd have evidence of some particular type E, then absence of evidence of type E is evidence for the falsehood of X. But, absence of evidence of evidence for X is not by itself evidence that X is false, only if X implies we should expect some particular type of evidence, which we don't get. As to the null hypothesis, I don't think my views pose any issue for acceptance of null hypothesis testing as it is actually used in science in practice. But, if you disagree, give me a realistic example of the use of null hypothesis testing, and we can discuss whether I am right in that belief or not. As to your third version, an explicitly divine, self-causing Teapot - well, I would say that all the reasons I have given so far, do not apply to your Teapot. So, based on the reasons I have so far provided, I have to be agnostic with respect to your Teapot. (Now, actually I don't believe in your Teapot, but that's for further reasons I have not shared here.) But, your Teapot isn't Russell's, since he never claimed any divine self-causing properties for his own. (((Zack Martin)))^™ 09:34, 4 July 2011 (UTC)

So you do concede that my Teapot is as likely to exist as a deity? (personal beliefs aside). I know it's not strictly Russell's teapot, but I wanted to propose a new hypothetical. ^{ONE / TALK} 10:09, 4 July 2011 (UTC)

Well, I think your Teapot is less likely than some other sorts of deities. However, as yet, I haven't offered any reasons for my belief, so as the conversation stands, yes, I must concede your Teapot is equal in likelihood to any other deity. (((Zack Martin)))^™ 10:36, 4 July 2011 (UTC)

the laws of physics imply teapots deities don't exist

our knowledge of human history implies no teapot deity

Logically there's no difference. Jack Hughes (talk) 09:38, 4 July 2011 (UTC)

As discussed below, the laws of physics or human history don't imply all deities don't exist. They possibly imply some deities don't exist, but there are plenty of possible, relevant, deities who they don't in any way preclude. (((Zack Martin)))^™ 10:40, 4 July 2011 (UTC)

(unindent) Maratrean, let's break down your 'evidence for the non-existence of the teapot':

• Do aliens exist somewhere in the universe? We have no good evidence that they do, no good evidence that they don't, so we are best just to suspend judgement on that question.

Under your standards for the existence of God, correct.

• If they exist, have they visited our solar system? We have no evidence that they have. Could they have visited without us being aware of it? Yes, that is possible.

So, applying the same standards as your standards for determining the existence of God, we have to conclude they might have.

• But, even if they have visited, why would they leave a teapot there? Although we can only really guess at alien psychology, we'd have to assume that they do things for a purpose, and it is hard to conceive of a reason why they'd dump a teapot in that orbit. Maybe it was an accident; maybe it was a case of garbage disposal; maybe it was an attempt at humour (the aliens have read Russell's essay, and find the irony of dumping a teapot there amusing.)

So, under the standards you're applying to the existence of God, again, we have to conclude that it is possible the aliens have put a teapot there, for any of these given reasons, or, indeed, for another reason you have not thought of.

• But adding all the pieces together, it seems that, given what we know, it is unlikely that aliens put a teapot in that orbit, and added to our knowledge of the laws of nature and human history, the overall conclusion is that is unlikely that one is there.

But you cannot definitively say that one is not there, so, under the standards you are applying for the existence of God, the existence of the teapot is undecided. And I should point out that, under the standards you're applying for the existence of God, this conclusion doesn't actually follow from your argument. It DOES follow using more sensible standards, but not the ones you should be applying, if you want to be consistent. 81.151.250.24 (talk) 10:55, 4 July 2011 (UTC)

Your argument ignores the fact we do have some information on the likelihood of alien life existing, on aliens being able to visit us, etc. Not a lot of information, but at least some. A few points worth considering:

1. The universe is a very big place, in both space and time; even if intelligent alien life exists, it is quite possibly very far away from us, spatially and/or temporally
2. Interstellar (say nothing of intergalactic) travel is very difficult, and it may well be so difficult that intelligent life never spreads very far from its origin
3. On the other hand, if affordable interstellar travel is possible, the exponential expansion of one lifeform may quickly occupy all the available evolutionary niches in the galaxy, and thus prevent any other intelligent life from evolving

We can argue about the above with respect to our knowledge of astrophysics, biology, and other matters. The point is not that we have definite answers to these questions, we don't. The point is, "did aliens create the Teapot?" is not a question for which we have zero information; not a lot of information, yes, but not zero either. One can construct a plausible argument, based on the above considerations, that us being visited by aliens is unlikely. Whereas, "Does God exist?" is a question for which we have zero information. So, that is why the two questions truly aren't comparable, and we can't treat them in the same way.

What we are talking about here is essentially the Fermi paradox, the Drake equation, etc. They are active subjects of scientific research. That's because the question of whether aliens are likely to exist is the kind of question science can answer. The existence of God is not an active subject of scientific research. That is because whether God is likely to exist is not the kind of question science can answer.

You misunderstand my standard. My standard is not "believe X only if we can be certain it is true". My standard is "if we have zero evidence for/against X, neither believe nor disbelieve it". So, to prove the alien-teapot-question is different from the God-question, it is enough to show it is unlikely the aliens did not put the teapot there; there is no equivalent evidence that it is unlikely that God exists. (((Zack Martin)))^™ 11:56, 4 July 2011 (UTC)

So, basically, you're saying we have virtually no evidence about the existence of aliens, so we can safely conclude that aliens probably do not exist, and therefore the teapot does not. We have even less evidence of the existence of God (namely, none at all) so...we have to conclude He might exist? I'm sorry, your position is simply not logical, not least because 'aliens put it there' is only one possible explanation for the presence of the teapot. Another one is that, in fact, this is an entirely natural phenomenon, and every single solar system in the universe has such a teapot, but we are simply entirely ignorant of any evidence of that fact - in much the same way you are saying we are entirely ignorant of any kind of solid evidence of God, and so we have to conclude He might exist. 81.151.253.86 (talk) 23:48, 5 July 2011 (UTC)

We observe the existence of patterns, regularities, laws, in this universe. From the existence of those patterns, we expect if X exists (say here, aliens), we'd likely see Y. The fact we don't see Y, reduces the probability that X is true. But, the God of Deism, existing entirely external to this universe, we can't infer any claim of the form "if X exists, we'd likely see Y". Since we have no Y, we have nothing expected lacking by which to reduce the probability of Deist God's existence, so P(Deist God exists) remains at its original value of 0.5. So it is not "no evidence" which leads us to infer the non-existence of aliens, but lack of reasonably expected evidence. Since we don't reasonably expect any evidence for God's existence, there is no evidential basis to infer he doesn't exist. (((Zack Martin)))^™ 10:48, 8 July 2011 (UTC)

Special pleading

As far as I can tell from the endless bollocks above the reasoning goes

• Russell's teapot doesn't exist because that would be silly
• God might exist because I want him to.

I have yet to see any reason why these may not be reversed. Jack Hughes (talk) 09:28, 4 July 2011 (UTC)

No, because we have actual reasons based on our knowledge of physics, history, etc., to justify the belief that the teapot does not exist. We have evidence for absence other than mere absence of evidence.

We do not have any actual reasons, based on our knowledge of physics, history, or any other discipline, to justify the belief that God does not exist. We have no evidence for absence other than mere absence of evidence.

So that's the reason why they can't be reversed, because the situations aren't parallel. For the teapot we have evidence of absence beyond mere absence of evidence, for God we don't. (((Zack Martin)))^™ 09:39, 4 July 2011 (UTC)

See above. There equal evidence in any scientific discipline you like against the existence of god - unless, of course, you're going for god-of-the-gaps which you appear to be. All the claims made about an interventionist deity, by definition, defy the laws of physics. Jack Hughes (talk) 09:44, 4 July 2011 (UTC)

You assume above that I define God to be interventionist. I don't. I define God as a being who (1) creates the universe, (2) provides for an afterlife (which exists outside the universe). I don't think any present time intervention in the universe is implied by my definition, only intervention at the beginning and the end - if there is any intervention at the present time, it is so subtle that is not detectable. So, there is no scientific evidence against the God which I have described. (((Zack Martin)))^™ 09:48, 4 July 2011 (UTC)

As the section says, special pleading. Because deities - as defined by practically everyone else - are as unlikely as Russell's teapot you have to define your own brand which is "outside the universe" which carefully avoids any reason. Sorry, there is no god outside the universe, there's no room because of all the wall-to-wall invisible unicorns. Jack Hughes (talk) 09:52, 4 July 2011 (UTC)

This isn't special pleading. It isn't just my idea of God which lacks the features you object to. Look to 17th and 18th century deism for a very similar view; or what about the God of Socrates, Plato and Aristotle? (Remembering that they were inclined to interpret the Greek myths in an allegorical fashion, and so likely didn't believe in them literally.) Why must everyone have the same concept of God as traditional Judaism/Christianity/Islam? Just because their concept of God has problems, doesn't mean that other people's must too. You are committing a fallacy of your own: because one version of an idea is bad, all versions of the idea must be bad, even versions that don't include the bits that make the bad one bad. (((Zack Martin)))^™ 09:57, 4 July 2011 (UTC)

But a god such as you describe is irrelevant. So he, she, or it created the universe? Even if this is correct, so what? According to you they've now fucked off and left us to it. Irrelevant. So he, she, or it runs sugar mountain - so what? That's totally irrelevant. Your god has no relevance and is meaningless. Jack Hughes (talk) 10:07, 4 July 2011 (UTC)

A God as I describe isn't irrelevant. I did mention an afterlife. A God who offers an afterlife is not irrelevant, but rather is offering something of value to human beings. (((Zack Martin)))^™ 10:33, 4 July 2011 (UTC)

Sorry, how is that relevant to my life now? This is very fundamental. So, there may or may not be a god of sugar mountain. So what? How does this affect my life right here, right now? Not only is there no evidence for such a being there is no evidence for what, if anything, I need to do to appease such a being so as to ensure my place on sugar mountain. The only way there could be any evidence is if this god were interventionist and then, well, bang go the laws of physics. So, right here, right now, you're entitled to believe in whatever you like about an afterlife but it simply has no relevance to how I should live my life. Jack Hughes (talk) 10:41, 4 July 2011 (UTC)

Well, first of all, you misunderstand my position. I don't believe in a hellfire and brimstone kind of God. I believe in the God of universalism - everyone goes to heaven, except maybe the really, really bad people (like rapists and murderers), but even if they go to hell, it isn't forever, but only for a limited period (but if you are Hitler, maybe a few million years), and after that they go to heaven too. So, as long as you don't do really bad stuff like rape and murder, you get your "sugar mountain" straight away, and even if you do the bad stuff, you get your "sugar mountain" a bit delayed. I think, you ought to want to do good, because its good, not because of bribes and threats. And that is why, God doesn't deal in bribes and threats, in my belief.

Now, if you have see no relevance in the possibility of an afterlife, that's fine. But don't expect everybody else to see things the same way as you do. For other people, faced with deep desires they cannot find fulfilment for in this life, deep tragedies from which they cannot escape, the prospect of an afterlife is not some candy sugar mountain, it is a real source of solace and hope. You are not one of those people, fine; but, not everyone is you. (((Zack Martin)))^™ 10:51, 4 July 2011 (UTC)

So you seriously suggest that some sort of unfounded fairy story is OK because it brings solace. FFS! Live for today, it's here, it's real. Don't spend your life yearning for some wish fulfilment fantasy because it won't somehow exist just because you want/need it to. Needing it simply doesn't make it real. Jack Hughes (talk) 11:01, 4 July 2011 (UTC)

Well, that gets us past the topic of the burden of proof, and on to the topic of faith. Is faith unconditionally a bad thing, or might it be good in certain circumstances? But I don't want to address that issue on this page, let's save it for another day. The point is, we can construct a concept of God, which cannot be scientifically/historically disproven, and will be relevant to some people (but not everyone - e.g. not to Jack Hughes, or Röstrigraben, or many other people...). If my allocation of the burden of proof is correct, we ought to be agnostic with respect to the existence of such a God. And my allocation of the burden of proof is what I am trying to argue for on this page. Russell introduces his Teapot as a counterexample to my proposed burden of proof, but his Teapot example doesn't work, because while we don't have any good scientific/historical evidence that God (as I've described him) doesn't exist, we do have good scientific/histroical reasons to suppose that the Teapot (as Russell describes it) doesn't exist. (((Zack Martin)))^™ 11:23, 4 July 2011 (UTC)

To be fair to Maratrean, he probably is "living for today". If the only things his belief in an afterlife compel him not to do on Earth is rape and murder, I think that can go unpitied. It's not like he's banning himself from sex or anything. ^{ONE / TALK} 11:46, 4 July 2011 (UTC)

Maratrea's rules when it comes to sex are simple: (1) consenting (no rape) (2) adults (no kids, unless you yourself are a kid, and your partner is of similar age) (3) no animals (4) no incest (5) nothing too extreme ("safe, sane and consensual", as the BDSM folks like to say) (6) take sensible precautions against STDs and unwanted pregnancy (7) no cheating (but open relationships, polyamory, etc., are fine, so long as all parties involved are aware and freely consent.) Pre-marital sex, no problem. Masturbation, no problem. GLBT, no problem. Prostitution? Pornography? Sex and capitalism should not be mixed, but if you want to take naked pictures with your partner go ahead (but no sharing them with others without his/her permission). Follow these rules, and you'll go to heaven. Break the biggies (like no rape and no kids), you'll go to hell. Break the smallies (like cheating on your husband/wife), she'll probably let you in to heaven straight away anyway. Actually, about rape and murder, the point of hell is not to scare you in to not doing these things. The hope is you won't want to do them anyway. Hell is more to meet victim's demands for justice, justice which often cannot be found on the earth. (((Zack Martin)))^™ 12:09, 4 July 2011 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘To go back - you are prepared to change your behaviour on the basis of a god about which you can - by definition - know nothing. It's your definitions of 'good' and 'bad' which are deciding who gets to sugar mountain and who doesn't, not this god's. You can wrap this up in faith all you like but there is no possible evidence that, even were your god to exist, then this would be a morality that it would work by. for all you know the only people who get to sugar mountain are the rapists and child molesters. Why not, no one said that a god has to operate by a human morality.

More likely this sort of religion is an attempt to wrap up the unknown and unknowable in some sort of security blanket and a belief that, as long as you're a good boy in this life then everything will work out fine in the next. If that's what you want to believe then fine but it's just a belief and has no more foundation than Russell's teapot. As such, if you want to persuade me to believe the same way then, as with Russell's teapot, the burden of proof is on you to persuade me that this imaginary being has any foundation outside your head. Jack Hughes (talk) 12:36, 4 July 2011 (UTC)

if you want to persuade me... the burden of proof is on you to persuade me - obviously, if A wishes to persuade B, then the burden of proof in practice is on A. That is true regardless of who the persons are or what the belief is. But when we say the burden of proof, we normally mean a different question - not where the burden of proof is in actuality, but where it should be in theory. This is the objective burden of proof, not the subjective burden of proof.

I have not made any attempt to convince anyone of the truth of my own religious beliefs here. I have simply been arguing that the (objective, not subjective) burden of proof is equal for both sides. I have been mentioning my own beliefs, not to try to convince anyone of their truth (how could anyone be convinced of their truth when I haven't given reasons for them?), just to try to make clear that (1) there is no actual evidence they are wrong, and hence I believe the burden of proof for them is equal to the burden of proof for their rejection (2) contrary to a number of claims, they are not irrelevant to all persons (although it is agreed they are irrelevant to some)

You misunderstand my attitude towards morality. I don't believe in doing what is right to get rewarded, and not doing what is wrong to not get punished. I believe in doing what is right because it's right and not doing what is wrong because it's wrong. As a universalist, I believe everyone gets the same reward; even Hitler. Really bad people, like him, get "punished" first, before they get their reward; but the "punishment" is not to deter the living from doing similar deeds, it is to answer the cry for justice of the victims. (That is why I put "punishment" in quotes, because it doesn't entirely match up with our ordinary idea of punishment, e.g. because deterence is absent.)

I think God is good, so I don't believe the prospect you've described (God rewards rapists and punishes non-rapists) is possible. Why do I think that? I believe that by faith. But, as I said, I want to leave discussion of faith to another day. I want to limit this page to trying to balance out the burden of proof; once the burden of proof is even, then I think we can use faith to move forward further. (((Zack Martin)))^™ 09:05, 5 July 2011 (UTC)

(Ab)use of skeptical tropes

This essay and the flood of comments above have proved an interesting point: the fuzzy verbal tropes that most people think of as "skepticism" aren't always precise enough to prevent their own abuse. In the interest of saving time I'll discuss the two most important examples:

• Occam's Razor. Numerous people have accused Maratrean of misusing Occam's Razor in the sections above, and while I don't completely agree with any particular commenter, the overall thrust of their argument is correct. If you let "simplicity" be determined by how complex something sounds when you type it up in English, then your definition of complexity is going to be severely warped. This opens the door to obvious abuses, because you can decide that basically anything is the "simpler explanation."
• Uncertainty. Imagine I select a person from your city at random and ask you to guess their full name. You could answer "I don't know," but you actually do have a lot of information about the question: you know that the person's name is much more likely to be "Mike" or "Sarah" rather than "Xpchtl Vaaaaaarax." Likewise, saying "I don't know whether God exists" hides so much information that it's almost dishonest. It's uncertain whether there is a God, sure, but unless you express the level of uncertainty you are making the question seem much more up-for-grabs than it actually is. To use a tired old example, it's as if I said "it's uncertain whether gravity exists." Precision is a virtue, and while I'm not suggesting that we should - or can - assign a numerical probability to the question of God's existence, we should strive to express the strength of our uncertainty as much as possible.

I don't want to get into a detailed discussion of Maratrean's essay, but if you would like to discuss or critique the examples I gave, listen to me prattle on about other ones, or suggest examples of your own, I'd be happy to talk. Tetronian ^{you're clueless} 13:12, 4 July 2011 (UTC)

Occam's Razor - well, elsewhere I've suggested that complexity should be defined as the Kolmogorov complexity of a theory's prediction set (that is a very rough definition, BTW). Trents suggest that instead it should be the number of elements in the theory's model. Two comments:

1. Two theories can be completely different yet have the same number of model elements. Two theories can have the same entities in their model, but apply different properties/predicates/attributes to those entities. I don't see how Trent's approach can work in that case. My approach, so long as the two theories have different prediction sets, those prediction sets could have different Kolmogorov complexities, and hence one could still apply the principle of parsimony
2. Trent and I have proposed different formalisations of the principle of parsimony - how do we decide whose is right?

Uncertainty - What I have been calling agnosticism (I don't know whether God exists, informally), take that to mean a uniform prior. P(God exists) = 0.5. Here are some rough definitions: Theism ===def P(God exists)>0.5, Atheism ===def P(God exists)<0.5, Agnosticism ===def P(God exists) = 0.5. Probably want to finesse that, to say if P(God exists) is near 0.5, that is agnosticism, near 1, theism, near 0, atheism. But not sure how near to make near. Maybe, dividing evenly, we should say Theism ===def P(God exists) > 2/3, Atheism ===def P(God exists) < 1/3, Agnosticism ==def 1/3 <= P(God exists) <= 2/3.

(((Zack Martin)))^™ 08:50, 5 July 2011 (UTC)

I'm with Trent on this one. Let's say that the universe is governed by a set of physical laws that can be expressed mathematically (which appears to be the case). We can make an uncountably huge number of predictions based on this, but it doesn't make sense for the K-complexity to be that large because we can output the laws using a much simpler Turing machine. I understand the appeal of your approach, but I just don't think it makes sense when you think about the fact that nearly any theory's prediction set is going to be enormous.

As for uncertainty - there's just no way you're going to get a prior of .5 for this problem. It's an abuse of the uniform prior, for a handful reasons:

• If you are only going to be discussing one particular piece of evidence in your essay, then you should present your findings as a likelihood ratio, not a posterior, and ignore priors entirely.
• If you are going to provide a posterior, then it is blatantly wrong to use an uninformative prior and then update solely on the one piece of evidence that you discuss. Instead, you should incorporate all of the other evidence into your prior or update on it before/after you update on the pieces of evidence you discuss in detail. (Because there certainly is a lot of other stuff to talk about.) That way, your posterior is at least honest.
• There are not two mutually exclusive hypotheses "God exists" and "God doesn't exist." Some other god(s) exist(s) is also an option. You have to be precise with what you mean by God - does it have to be something ontologically irreducible? Would super-powerful aliens count? (And so on and so forth). And because this definition would be extensive and contain a lot of information, you would be able to come up with a prior that is a hell of a lot more informative than 1/2.
• If, however, by "God" you mean "something roughly resembling the monotheistic god of the world's current major religions," then the uniform prior will not get you .5 - it'll get you something orders of magnitude smaller, because "something roughly resembling the monotheistic god of the world's current major religions" is a really, really small fraction of the search space.
• Why use a uniform prior when you can use Solomonoff Induction? (Especially if you think you have a better way of using K-complexity.)
• Even if your prior (and posterior) does turn out to be 1/2, you should specifically say so - "I don't know" doesn't express your level of belief, it just expresses generic uncertainty (this was my original point).

Shit, I'm getting long-winded again. Oh well. Tetronian ^{you're clueless} 19:46, 5 July 2011 (UTC)

We can make an enormously huge number of predictions based on the laws of nature, but what is the K-complexity of that prediction set? The length of those laws themselves, if those laws are expressed with perfect precision. So, we are back to my proposal, theory complexity = K-complexity of prediction set.

As to the uniform prior:

• You argue, that if only present one piece of evidence, I should use a likelihood ratio rather than a uniform prior. Actually, I present zero pieces of evidence (that is the whole point of my essay, assuming there is no evidence available, what should we believe), and if you'd rather a 1:1 likelihood ratio than a 0.5 uniform prior, I'm not bothered
• As I've argued repeatedly, as a general rule, "absence of evidence is evidence of absence" is false. In some particular situations, it can be correct, but that is because of the particular nature of those situations. Some of those particularities may be very common (e.g. arguments about physical objects), but that doesn't make them applicable to totally different situations (e.g. arguments about claimed non-physical existents)
• As to the "other Gods exist" argument, let me define "God" as follows: "An omnipotent being, existing external to the universe, who created it". Now, there can't be more than one such being, because more than one omnipotent being is impossible. Whether this being is best described by Judaism or Christianity or Islam or some other religion, is actually moving on to a separate question (not "Does God exist?", but "Which religion describes God the most accurately"?, etc.)
• Again, the search space only contains 2 options - yes God (as defined) exists, or no God (as defined) doesn't. There aren't any other answers, the way I've scoped the question. Also, your argument from "other Gods" ignores the possibility of inclusive monotheism (maybe all these different gods are just different names or accounts or somewhat distorted versions of the one God)
• I'd be open to trying Solomonoff Induction, if you can suggest how to use it in this particular case
• I agree "I don't know" is a bit vague. I think the rest of my argument made clear that I meant by it 50-50. I agree with you that probability theory/Bayesianism is a useful approach, but personally I am not as absolutely convinced by it as you are (I think it has logical problems, e.g. as I mentioned in my posting on your talk page about "metaprobability"), and I don't want to exclude readers who are not so keen on it.

(((Zack Martin)))^™ 11:02, 8 July 2011 (UTC)

I'm going to try to make this as concise as possible.

I guess you weren't using the phrase "prediction set" the way I thought you were - can you elaborate on what you mean by a prediction set? A technical definition would be preferred, if possible.

You are flatly wrong about absence of evidence. The idea that absence of evidence is evidence of absence is a law of probability; a good explanation of the derivation can be found here, and I'd be happy to elaborate if you like. The point is, you can't ignore it just because it's inconvenient for you, and you can't wave it away just because your beliefs are about "non-physical existents" (whatever that means).

What exactly is an omnipotent being? The concept seems reasonable when you look at it from a distance, but when you think about it in detail you run into some problems. It's quite possible that the category you are trying to make inferences about cannot even exist because it is not logically possible.

As I said above, your definition of "God" will inevitably contain information that will let you make a much better prior distribution than 1/2. "One omnipotent being" is a small cluster in the space of godlike things that could exist (particularly because, as you pointed out, it's possible for all logically allowable combinations of non-omnipotent beings to exist).

Solomonoff Induction tells us one very important thing about this problem: it would take an enormously more complex Turing machine to describe the mind of an omnipotent being than it would take to describe a set of physical laws.

As for metaprobability - in the interest of keeping this short, I'll reply on my talk page.

Tetronian ^{you're clueless} 23:56, 9 July 2011 (UTC)

Let me just respond to your point about the absence of evidence for now, and I'll respond to the rest a bit later. As I understand Yudkowsky, he is talking about P(H|E) v.s. P(H|~E). Now, what is H here? In our case, H is the claim "God exists". What is the evidence E? Well, my point is, there is no relevant evidence E. We can't talk about P(H|E) when there is no relevant E. Or, well we could have irrelevant evidence E. If E is irrelevant to H, then P(H|E) = 0.5, and P(H|~E) = 0.5 too. As I said, if we are in a situation where evidence is expected, then absence of expected evidence is evidence of absence. But if we are in a situation where a hypothesis implies no particular evidence is to be expected, then we can't make that inference. Or, slightly more accurately, H does imply some E (e.g. an afterlife), but none of that E is testable now, so for the E for which P(H|E) ≠ 0.5, we can't say anything much about E or ~E (maybe just, P(E)=0.5.) (((Zack Martin)))^™ 02:35, 10 July 2011 (UTC)

If the clouds parted, the sky turned bloodred, and a booming voice came down from the heavens declaring, "I am the one God of the universe", that would be evidence (E) in favor of your H. Thus, failing to observe E is (extremely weak) evidence against H, because P(E|H) > P(E|~H) and P(~E|H) < P(~E|~H). This example, though completely true, is a admittedly a bit extreme, but I'm just trying to get the point across. The best relevant evidence is the fact that the universe has never been seen to obey the whims of an omnipotent deity, and it is this absence of evidence that makes the falsity of H - which, as I have insisted above, was already favored by our prior distribution - appear enormously more likely. Tetronian ^{you're clueless} 03:57, 10 July 2011 (UTC)

Well, the problem is from the idea of God, to infer what evidence we would expect. You are right, maybe God could reveal herself as a booming voice from the clouds, and the fact she hasn't done so could be evidence against her existence. But, why would we assume she would want to do that? Of all the countless things she could choose to do, we have no particular reason to assume she would want to do that one, so the fact that she hasn't doesn't count for much at all. So, maybe you are right, this evidence pulls the probability beneath 0.5, but only very very slightly, so I doubt it makes much of a difference.

The universe has never been observed to obey the whims of an omnipotent deity? How do you assume to know what the whims of an omnipotent deity are? How do you know it isn't obeying those whims, right now? If we claimed to have some particular expectation for an omnipotent being to act in a certain way, then the failure of the universe to be that way would be evidence against the existence of such a being. But do we have any such expectation?

Yudkowsky suggests the problem of evil. I believe God is omnibenevolent, and I think I have a solution to that problem. However, the way I've framed the question, I have framed it in terms of an omnipotent being, but not an omnibenevolent one. Since the world contains a fair amount of both good and evil, it would seem reasonable to suppose that God as creator of the world is neither purely good nor purely evil, but a bit of both. More to the point, if we have no idea how good God is, the amount of good or evil in the world is not evidence for or against God's existence. (((Zack Martin)))^™ 05:40, 10 July 2011 (UTC)

I'm going to simplify the problem somewhat to make my point clearer. Let's take the hypothesis that a Really Really Powerful Being exists; call this H₁. H₁ can be represented as a group or cluster of hypotheses inside of an incomprehensibly large space of possibilities. As you pointed out above, various gods and beings inside of H₁ all have different motivations and would not all act the same. Some of them would use their powers in a way that would make their presence known to humanity, and others would not. Now, let's examine the "booming voice from the heavens" event, which I'll call E. We know a handful of things about H₁ and E:

• P(E|H₁) > P(E), or "the probability of the booming voice event is greater if a Really Really Powerful Being exists." We could also phrase this as, "E is more likely if we are inside the cluster H₁ than if we are at a random location anywhere in possibility-space." Even though not all of the gods in H₁ would perform the booming voice event, some of them would, and this makes more likely for E to occur inside of H₁ than it does at a random location in the space of possibilities.
• P(E|H₁) >> P(E|~H₁), or "the probability of the booming voice event is much greater if a Really Really Powerful Being exists than if one doesn't exist." Hopefully this one should be obvious - no matter what P(E|H₁) actually is, it's clear that P(E|~H₁) is going to be significantly less.
• P(~E|H₁) << P(~E|~H₁), or "the probability that we will not see the booming voice event is much lower if a Really Really Powerful Being exists than if one doesn't." This is similar to the above, and hopefully it will be obvious for the same reason.
• P(~E|H₁) < P(E), or "the probability of the booming voice event is lower if a Really Really Powerful Being doesn't exist," or "not observing the booming voice event is more likely if we are at a random location in possibility-space than if we are inside of H₁."

Therefore, ~E is evidence against H₁. Even though not all of the gods in the space would commit E, it's still more likely that E would occur inside of H₁ than outside of it or in a random location. Likewise, ~E is more likely outside of H₁ than inside or in a random location. This means that observing E would shift our prior P(H₁) upwards, and observing ~E would shift it downwards. If this still isn't clear, I suggest playing around with Bayes' Theorem until you get an intuitive understanding of my argument. Tetronian ^{you're clueless} 14:53, 10 July 2011 (UTC)

If event E is a miracle, then may I say that an event E has occurred in the form of me understanding that. ADK^{...I'll delete your railing!} 15:21, 10 July 2011 (UTC)

Bravo. I realize that a technical explanation can be hard to understand, but I figured it would be the only way to put things in precise enough terms to convince Maratrean. Tetronian ^{you're clueless} 15:28, 10 July 2011 (UTC)

Well, I was going to ask a sensible question and then, for the first time, I think I understand the difference between Bayesian and Frequentist and said sensible question was backwards. ADK^{...I'll coax your bassoon!} 15:31, 10 July 2011 (UTC)

If you still have questions, ask away! I try to explain things as clearly as I can, but sometimes that just isn't clear enough. Tetronian ^{you're clueless} 15:36, 10 July 2011 (UTC)

Well, it was a question about the first point and whether event E was more likely to occur outside H1 because even if it's low probability, the whole space is infinitely larger than H1. But I think that's because I was slightly misreading the actual clauses in that bit. The statement is saying that it would be more likely inside H1 than any other similarly sized sample of the probability space, but not necessarily saying that E is good evidence for us being inside H1. If you're inside H1 probability of E is, say, 0.2, but over the whole probability space E might be 0.01, and this is a constant proportion that doesn't care about the relative size. So perhaps you can perhaps see what I mean when I said it was "backwards". ADK^{...I'll baste your domino!} 16:01, 10 July 2011 (UTC)

You're exactly correct. Admittedly the phrasing was a little bit ambiguous; it makes more sense this way: "the probability of E occurring at a randomly selected point in H1 is greater than the probability of E occurring at a randomly selected point in the whole space." Tetronian ^{you're clueless} 16:04, 10 July 2011 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ Tetronian, say your argument is correct. Then P(H₁) < 0.5. But, maybe only very very slightly less. All the logic you've given so far is compatible with the conclusion P(H₁) = 0.5 - 10^-100. And, we can round 0.5 - 10^-100 up to 0.5. In conclusion, P(H₁) = 0.5 (or close enough for all practical purposes). (((Zack Martin)))^™ 06:59, 12 July 2011 (UTC)

See my reply below. Tetronian ^{you're clueless} 20:41, 12 July 2011 (UTC)

visual explanation

Perhaps a visual explanation will make things clearer. Here's how to read the diagram:

• The green circle represents the space of possible universes.
• The orange circle represents the group of universes that contain Really Really Powerful Beings (H1 in my explanation above).
• The red dots represent universes in which we observe "booming voice events" (E in my example above).
• The blue dots represent universes in which we do not observe E.

Do you see how observing E makes it more likely that you are in the orange circle, but failing to observe E makes it more likely that you are not in the orange circle? Tetronian ^{you're clueless} 17:35, 10 July 2011 (UTC)

Yes, but the increase in likelihood may be so tiny so as to be irrelevant in practice. You have an argument that P(H1) can't be 0.5, but it could be arbitrarily close to 0.5, say 0.5 - 10^-100. Which we are quite justified in rounding off to 0.5. (((Zack Martin)))^™ 07:10, 12 July 2011 (UTC)

A few things:

• Yes, it "may be compatible" with that figure, but what if it isn't? You can't make excuses to reality, so it makes little sense to dismiss a piece of evidence just because you can imagine a universe in which that evidence does not matter. So, if we're going to make up numbers, let's at least do so in way that isn't so prone to rationalization. Here's what I suggest. Answer the following questions as honestly as possible. I will use your answers to compute a P(H|~E) that more closely reflects what you believe using a sort of "revealed preference" method.
  • Would the observation of a booming voice event be strong evidence in favor a Really Really Powerful Being. Approximately how much do you think it would make us update?
  • How would you compare the probability that a booming voice event could occur in the absence of a powerful being (a "false positive," if you will), with the probability that a booming voice event could occur given the existence of such a being. If you think you can give estimates or give some kind of ratio, do so.
  • How would you compare the probability that a booming voice event would not be observed given the existence of a powerful being (a "false negative," if you will), with the probability that a booming voice event would not be observed given that a powerful being does not exist. If you think you can give estimates or give some kind of ratio, do so.
• Let's keep in mind the context of this particular tangent. The "sky turns red, loud booming voice is heard" example is but one small empirical cluster inside of a huge collection of (counterfactual) pieces of evidence that would have lent credence to the Really Really Powerful Being hypothesis had they been observed. We're not just updating on the booming voice event, we're updating on all such events. So, even if we grant the (improbable) assertion that the lack of a booming voice event will only let us update a fraction of a fraction of a percent, there are trillions and trillions more pieces of absent evidence that we have to update on. (This is probably the most important point, much more so than the actual strength of ~E.)

Tetronian ^{you're clueless} 20:40, 12 July 2011 (UTC)

Well, I think I see where you are taking the argument, so let me first explain my basic disagreement with that approach, and then I'll answer your specific questions.

Is Bayesianism right? - You seem to me to be a very convinced Bayesian. I see value in Bayesianism; but I also think it has some limitations. The most basic intutition behind the Bayesian approach is entirely right - our beliefs aren't just a binary "I believe X" vs "I believe not-X", or even a ternary "I believe X" vs "I believe not-X" vs "I neither believe nor disbelieve X"; there are degrees of certainty as well, from "Really not too sure about X, but reckon its slightly more likely than not" through to "Absolutely certain X is true". Then, Bayesianism goes one step further, and suggests that degree of confidence can be quantified as a real number between 0.0 and 1.0.

But, this is part of where I see Bayesianism coming undone. Suppose I say "Really not too sure about X, but reckon its slightly more likely than not". Here I have given a qualitative description of my own confidence. How sure can I be I have described my own confidence accurately? Reasonably certain. But now, try to make it quantiative. Well, 0.5 > P(X) > 1.0. How certain can I be about that? Again, reasonably certain. I can also add, with confidence, P(X) - 0.5 << 1.0 - P(X). (i.e. my confidence in X, although between 0.5 and 1.0, is much closer to 0.5 than 1.0.) I would even say, I could say with a lot of confidence, 0.5 > P(X) > 0.75. OK; but let's try to define P(X) more precisely. Can we assign a specific real number to it, rather than a broad range? Well, is P(X) 0.51? or 0.501? or 0.5001? I am much less confident a specific real number is a correct description of my own degree of confidence in X than I am that any real number in some broad range is. So, P(0.5 > P(X) > 1.0) ~ 1.0. But what about P(P(X)=0.51)? Much less. I am not sure at all that 0.51 is right, rather than 0.501 or 0.5001. (Given a range of probabilities, the smaller the range, the smaller the metaprobability; arguably, any specific probability assignment has metaprobability zero.)

So, while I feel comfortable using Bayesian reasoning to derive some broad heuristics, I am uncomfortable a lot of the time with using it with specific numbers. My metaprobability is much higher for the former than the later. Now, when we are dealing with real world scenarios, like say a gambling game, or a quantum physics experiment, where we can apply some (frequentist!) principles to derive specific numeric probabilities, I am much more comfortable with dealing with specific numbers - my metaprobability will be much higher. But, with abstract philosophical issues like this, use of such principles is a lot harder, the more precise you try to be quantitatively, the less the metaprobability your choice of probability is correct.

Bayes law is OK as a first approximation, but it can't be completely right, because it ignores metaprobability. If all our metaprobabilities are nearing 1.0, Bayes law is probably close enough. The further they get away from 1.0, the less appropriate Bayesianism becomes. Then I think we need to look at alternative approaches.... WP has some good background on this issue - imprecise probability, Upper and lower probabilities, Possibility theory, Dempster-Shafer theory... although I really don't claim to know the answer, just that the orthodoxy of LessWrong is far from being right...

Booming voice - Would a booming voice be strong evidence in favour of an omnipotent being? Well, one has to consider the alternative explanations. I've developed psychosis. I've been drugged. Technologically advanced aliens are zapping me with their mind control ray. By itself, I wouldn't say a booming voice is particularly strong evidence of an omnipotent being. If I had some more evidence to exclude the other possibilities, then that might make the booming voice stronger evidence. I repeat my objection to trying to put a specific number on it, but I'd suggest we shouldn't update very much as a result.

You then argue their are trillions of trillions of other pieces of evidence similar to E. However, the reality is, we have very little idea what P(E) should be (very low metaprobability), and very little idea of exactly how many pieces of evidence similar to E there are. And what matters here, is what is the order of magnitude between probability of an individual E vs. how such E there are. We really have very little idea what the order of magnitude of either is, so we really have very little idea whether the argument you are trying to sketch at the end is valid or not. Which brings us back to, 0.5. (((Zack Martin)))^™ 11:57, 13 July 2011 (UTC)

I'll reply in full tomorrow. For now, though, my reply is this: If you're not convinced by Cox's Theorem or the derivation of Bayes' Theorem from the definition of conditional probability, just look at the picture. See how observing a particular event (being in one of the red circles) makes it more likely that you are in the orange circle, see how not observing it (being in one of the blue circles) makes it less likely that you are there. You can say as much as you want about the philosophical underpinnings of it all, but you can't bundle everything together and slay it with one huge paragraph - the fact remains, Bayesian updating is the correct way quantitatively update on evidence. Still don't agree? Look at the picture.

As for your concerns about metaprobabilty paradoxes: well, the first thing that comes to my mind is "The novice goes astray and says: "The Art failed me." The master goes astray and says: "I failed my Art."" The fact that you can take a model and twist it into something paradoxical is not proof that the model is flawed - it's like the classic lemma proving that 1 = 0. But more on this tomorrow.

That said, you are exactly correct about psychosis - it's something I failed to consider, and it does indeed significantly lower the strength of each piece evidence. I maintain, however, that there is a huge amount of evidence, a point I'll elaborate on in the future. Furthermore: you can't dismiss this point entirely by saying that we don't know the order of magnitude of this evidence - what you are literally saying is, "I don't know the strength of this evidence, so I'm justified in ignoring it entirely."

Why this unholy desire to be locked in exact uncertainty? At this point I have to ask: what evidence would it take for you to significantly update your beliefs? Tetronian ^{you're clueless} 02:37, 14 July 2011 (UTC)

The problem about your diagram, is we actually have no real idea how big the circles are comparatively, and the relative frequency of the various dots in each. The ratios between the sizes of the circles, the ratios of the frequencies, are very important in determining how much this potential evidence will move us from a state of zero information (0.5). It could move us a lot, it could move us a very little. Because we have very little information on the real size of these ratios, we have very little information on how far that evidence could move us from 0.5 either.

You maintain there is a huge amount of evidence, but none of us can really have any idea of how much evidence there is, or what most of those pieces of evidence are, let alone assign probabilities (conditional or unconditional) to them. So, I agree there is a lot of potentially relevant evidence, but not knowing how much, or what exactly it is, or how relevant precisely it is, we really can't draw any conclusions.

I don't have this "unholy desire to be locked in exact certainty?". I am quite open about the fact that I don't think there are many things that can be known with certainty; and one thing which can't be known with certainty is my own degrees of confidence. I can have only a rough idea of my own probability judgements, but can't know them with sufficient certainty to assign a particular real number to them, only a very broad range of real numbers. I can't say for certain how I would update, or how a rational agent should update, given certain evidence; I can only give broad sweeping statements about that, and can't be very confident in them. If anyone here has an "unholy desire for exact certainty", it's you with your desire to assign specific numeric probabilities when that isn't really possible, to reduce the qualitative to the quantitative when we can't really do so, and in thinking that human rationality can be reduced to a single quantitative model (ignoring the fact that Bayesianism has its critics, who have proposed alternative mathematical formalisms which they believe to be improvements on it.)

And, let me ask you a question - what is the probability you assign to "Bayes Theorem is true"? (((Zack Martin)))^™ 10:46, 14 July 2011 (UTC)

Apologies for the rudeness in my last post. You've given me a lot to reply to - I'll try to get to it in the next few days. Tetronian ^{you're clueless} 19:12, 15 July 2011 (UTC)

No worries, I didn't think you were being rude. Maybe a bit argumentative, but I'm exactly the same. You didn't insult my intelligence, or ask if someone whacked my head as a child, or call me by the name of an organ in the ano-genital region (all of which others here have done)! (((Zack Martin)))^™ 00:07, 16 July 2011 (UTC)