Is the Universe Rational?

Is the Universe Rational?

About Taner Edis

Professor of physics at Truman State University

  • http://www.blogger.com/profile/06394155516712665665 CyberKitten

    No.

    Nor is it irrational….. [grin]

  • http://www.blogger.com/profile/09934402450298584577 Charles Sullivan

    Of course, even if the universe is intelligible it in no way implies that it was designed.

    Nice article, Taner.

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    The intelligibility of the physical universe, a fact that sustains all natural science, is a surprising issue if you are a naturalist. One question is how come our brains, which are the result of a long evolutionary process that selected those brains best able to perform tasks necessary for survival such as avoiding being eaten by tigers, would also be able to perform tasks such as discovering general relativity within gravitational phenomena. The question here is not so much how a brain optimized for basic survival skills should be capable of performing highly abstract tasks with no relation to survival (it is quite common in evolution that an organ which evolves to do A ends up doing B), but rather how come the properties of the physical universe are such that blind natural selection would produce a brain capable of discovering their deep structure. That is a really strange and surprising property for an uncaring universe to have.

    The most surprising fact though is one to which Taner alludes and which Nobel laureate physicist Eugene Wigner termed “the unreasonable effectiveness of mathematics in the natural sciences”. I don’t think the issue here has something to do with Platonism, for it seems that all mathematical propositions refer to properties of the physical universe, so no extra-physical realm for mathematics is necessary. After all even the most abstract mathematical theorems describe properties of what will happen (or will not happen) if one pushes symbols around physical paper following a particular set of rules. Rather Wigner points out that time and time again in the history of science, mathematics that was developed independently and in the abstract found unexpected applicability in the scientific modeling of phenomena. There are many examples for this, but to mention just one: Non-Euclidean geometry first developed by Arab mathematicians and completed during the 19th century found applicability in 20th century general relativity. In conclusion it’s a demonstrable fact that physical phenomena display an unexpected deep and elegant mathematical order.

    Now on theism this remarkable fact is easy to account for; not so on naturalism. When confronted with this type of unexpected and surprising fact about the physical universe naturalists tend to counter using the so-called anthropic principle as an explanation of last resort. Whether the use that principle can produce some kind of explanation in the first place is highly questionable (it certainly fails the gold standard of allowing for predictions), but in any case the anthropic principle is not applicable to the case at hand, because natural evolution (up and including the evolution of intelligence) is just as feasible in a universe which is *not* mathematically elegant. The fact then that our universe is so mathematically elegant, a fact which forms a key aspect of the universe’s intelligibility, represents a major surprise which cries out for an explanation. Which, as far as I know, no naturalist has been able to offer.

    As for the “God of the gaps” mantra, perhaps we should start talking about the “naturalism of the gaps”. After all it’s a historical fact that since the beginning of the twentieth century the number of naturalism’s explanatory gaps has grown with alarming speed. Not to mention several previously sacrosanct naturalistic intuitions had to be abandoned, for example determinism, locality, the epistemic principle that existential beliefs should be based on evidence (many naturalists now believe in the existence of parallel universes with no evidence whatsoever), Occam's razor, prima-facie plausibility, etc. That most of these problems for naturalism have originated with modern science is one more big surprise I suppose, given naturalism's reliance on science. It looks like if you try to build metaphysics exclusively on physics, physics turns around and smacks you on the head.

  • http://www.blogger.com/profile/16826768452963498005 Jim Lippard

    Dianelos: It seems to me that the history of science, even if you just restrict it to the history of mechanics from Aristotle to Einstein, shows that the kind of rationality and intelligibility the ancients expected of the universe isn't there.

    Hopes of unification of the sciences, and even of Hilbert's program in mathematics have been completely dashed.

  • http://www.blogger.com/profile/05693985638589020492 Mark

    Dianelos, I really don't understand your objection. Why is a true, "elegant" theory of physics more mysterious on naturalism than a true, "inelegant" theory? Does naturalism attach prior probability estimates to possible physical theories? Do we know for certain that the deepest laws of nature are intelligible to us? I don't think I've encountered any serious theistic philosophers who've argued this point, though if there are any I'm certainly open to being informed.

    After all it’s a historical fact that since the beginning of the twentieth century the number of naturalism’s explanatory gaps has grown with alarming speed.

    O.K., this sounds a little ridiculous, akin to creationist claims that discoveries of new fossils simply create new gaps in the fossil record. The proliferation of unanswered scientific questions is a direct result of the enormous progress that naturalistic scientific theories have made thus far. We're not discovering anything new that suggests our whole project needs to be overhauled and replaced with radically different theories having "non-natural" entities at the center.

    the epistemic principle that existential beliefs should be based on evidence (many naturalists now believe in the existence of parallel universes with no evidence whatsoever)

    This is terribly disingenuous. Many Universes is a philosophically viable (though I won't say "correct") inference from cosmological fine-tuning. See Darren Bradley's paper "Multiple Universes and Observation Selection Effects" (available online, check Google).

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Jim: I personally can hardly imagine a more interesting and intellectually challenging universe than the one we are inhabiting. If God exists and values thought then this is the kind of physical environment S/He would have created for us I think: not too easy as to be trivial, not too hard as to be impossible, and amazingly prolific. Conversely, if it were possible to unify science and thus reduce the scientific project to engineering the computation of the implications of a fixed set of rules, or, in the other extreme, if the universe quickly turned unintelligible, then how boring would a scientist's life be! Similarly for mathematics: The fact that Hilbert’s program was proven to be impossible by Goedel turned mathematics into an ever-expanding space of potential personal discovery.

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Mark wrote: “ Does naturalism attach prior probability estimates to possible physical theories?

    I don’t think it’s a question of prior probabilities. Suppose somebody throws 10 dice and gets the sequence 6124352414 and somebody else does the same and gets the sequence 6666666666. Both events are equally probable, but it is only the latter event that cries out for an explanation, for it is a surprising and unexpected result. The fact that the fundamental physical constants appear to be so incredible fine-tuned is not a scientific problem; rather it simply represents an aspect of the order that science has discovered in the physical phenomena it studies, and science’s job is just that: to discover this order. But the same discovery, namely that the properties of the universe appear to be precisely balanced on the point of a pin, was a highly unexpected and surprising result from naturalism’s point of view and hence cried out for an explanation – and surely enough naturalists have proposed an explanation for it (namely the multiverse hypothesis), albeit at a huge cost for naturalism’s epistemology, parsimony, and plausibility. My point in the previous post was that the universe’s intelligibility is a fact which, similarly but independently, represents the balancing of the universe’s properties on the point of a pin, hence also cries out for an explanation, but one which to my knowledge naturalists have not been able to provide.

    Mark wrote: “ The proliferation of unanswered scientific questions is a direct result of the enormous progress that naturalistic scientific theories have made thus far.

    I am not discussing science’s problems but naturalism’s. As far as science goes, it is indisputable that science continues to advance at a furious rate and succeeds in discovering ever greater order in the physical phenomena it studies. Incidentally, the “gaps” in the fossil record were expected and their existence is easy to explain, so it’s not like this analogy works. Indeed Darwinism is not so much a finished theory, but rather a theoretical framework which is continuously consolidated with new knowledge (e.g. the genetic drift). Now let us contrast this state of affairs to how the same scientific advance has falsified naturalistic assumptions that used to be pillars of the naturalistic worldview (such as determinism, locality, the eternity of time, etc), how it has forced naturalists to abandon their main epistemic principle (namely that existential beliefs should be based on evidence), how it has destroyed any semblance of agreement and objectivity (see the many mutually contradictory naturalistic interpretations of quantum mechanics, the many mutually contradictory responses to the hard problem of consciousness, etc), and how it has pushed naturalists to propose ontologies with fantastically implausible properties (causality backwards in time, the impossibility of suicide, panpsychism, and such). I don’t see science suffering from any of these problems. Indeed a theistic scientist need not search for a naturalistic interpretation of quantum mechanics, need not explain the apparent fine-tuning of the fundamental constants, need not explain the universe’s intelligibility, etc, simply because on theism these are not surprising facts.

    Mark wrote: “Many Universes is a philosophically viable (though I won't say "correct") inference from cosmological fine-tuning.

    This is like a theist saying: “That God has a moral justification for allowing evil is an inference from the fact that there is evil in the world.”

    Naturalists used to be the kind of thinkers who did not conjure up existents (never mind entire universes) with no evidence or scientific rationale whatsoever – just in order to shore up their metaphysical worldview. Nevertheless I concede that there exists a naturalistic explanation of sorts for the cosmological fine-tuning; my point in response to Taner’s piece is that there is no naturalistic explanation for the universe’s intelligibility. Or is there?

  • http://www.blogger.com/profile/16826768452963498005 Jim Lippard

    Dianelos: It sounds like you're advocating a Goldilocks view of the intelligibility of the universe. It's neither too intelligible nor too unintelligible, but is just right, somewhere in between. But what makes it just right, and aren't there arguments available for a theist to take *any* of those positions?

  • http://www.blogger.com/profile/05693985638589020492 Mark

    I don’t think it’s a question of prior probabilities. Suppose somebody throws 10 dice and gets the sequence 6124352414 and somebody else does the same and gets the sequence 6666666666. Both events are equally probable, but it is only the latter event that cries out for an explanation, for it is a surprising and unexpected result.

    I don't know what you mean when you say "cries out for an explanation." The difference between the two events you name here is that one of those two sequences is vastly more probable under the hypothesis of an unfair die than under the hypothesis of a fair die. We know this because these probabilities can be mathematically grounded on the fair/unfair dice models. However, you're not going to be able to ground the surprising-ness of the universe's intelligibility on any sort of naturalistic model. Naturalism alone – the simple position, roughly speaking, that there are no supernatural agents, whatever those are – doesn't tell us anything about whether the universe ought to be intelligible to humans or not. Nor does theism, for that matter, although perhaps certain forms of theism do.

    Now let us contrast this state of affairs to how the same scientific advance has falsified naturalistic assumptions that used to be pillars of the naturalistic worldview (such as determinism, locality, the eternity of time, etc), how it has forced naturalists to abandon their main epistemic principle (namely that existential beliefs should be based on evidence), how it has destroyed any semblance of agreement and objectivity (see the many mutually contradictory naturalistic interpretations of quantum mechanics, the many mutually contradictory responses to the hard problem of consciousness, etc), and how it has pushed naturalists to propose ontologies with fantastically implausible properties (causality backwards in time, the impossibility of suicide, panpsychism, and such).

    Giving up locality, determinism, etc. has simply led us to a different, more counterintuitive form of naturalism than before. It hasn't led us away from naturalism, so I fail to see how this is germane.

    You are again being extremely uncharitable of accusing proponents of the multiverse hypothesis of abandoning previous canons of evidence in order to maintain their naturalism, since they'll say fine-tuning is the evidence for the multiverse hypothesis, in just the same way that theists will argue that fine-tuning is evidence for theism. This aside, you give both camps (theists and multiverse-ers) too much credit. Whether fine-tuning counts as evidence for anything at all is highly contended by many philosophers. Mark Colyvan, Elliott Sober, John Norton and Bradley Monton, for example, all give plausible reasons against this assumption. Some of these reasons are actually mutually incompatible, but they are all at least live.

    There are some interpretations of QM that are strictly incompatible with naturalism, but I don't believe that these are the most favored by any stretch of the imagination. And the hard problem of consciousness, while a real problem for materialism, is not something new and not something that scientific advancement has uncovered.


    This is like a theist saying: “That God has a moral justification for allowing evil is an inference from the fact that there is evil in the world.”

    No, that's silly. I refer you again to the Darren Bradley paper, which you really ought to read. Bradley explains quite clearly that if observation method M is biased toward observing an event of type E, and we observe an E-event through M, this is confirmation of the hypothesis that there were many opportunities for an E-event to occur rather than just one. He then goes on to apply this to our observation of fine-tuning and multiverses, as well as to address certain criticisms that've been made to the argument. His argument does assume a certain view of confirmation, but it's one that's quite well subscribed to by philosophers of science nowadays.

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Mark:

    In my analogy the die is fair and that’s why both sequences are equally improbable, but only the latter one is surprising as it does not comport with our expectations, and therefore requires an explanation. My point is that given an assumption it is not improbable data that challenge it but rather data of an unexpected form, data that do not fit. That the universe’s fundamental constants are so incredibly fine-tuned is certainly unexpected on naturalism, for it does not fit with the idea of an undesigned universe.

    I still don’t understand your rationale for claiming that the cosmological fine-tuning is evidence for the existence of parallel universes, for I don’t see how a problem can count as evidence for the otherwise ad hoc explanation put forward to solve it. Consider the following stories:

    1) On the assumption that naturalism is true the fact that the fundamental constants appear fine-tuned is unexpected. A possible explanation, if one wishes to hold on to the assumption that naturalism is true, is that parallel universes with different fundamental constants exist. Therefore the apparent fine-tuning of the fundamental constants evidences that parallel universes exist.

    2) On the assumption that the butler is the murderer the fact that twenty people swear to have been at a bar with him when the murder took place is unexpected. A possible explanation, if one wishes to hold on to the assumption that the butler is the murderer, is that all twenty people are lying. Therefore the sworn testimony of the twenty people is evidence that they are lying.

    These two stories are analogous, and therefore the conclusion of the first one is as absurd as the conclusion of the second one. The example I mentioned in the previous post would be this:

    3) On the assumption that theism is true the fact that so much evil exists is unexpected. A possible explanation, if one wishes to hold on to the assumption that theism is true, is that God has morally sufficient reason for allowing so much evil to exist. Therefore the fact that so much evil exists is evidence that God has morally sufficient reason for allowing it.

    Darren Bradley in the paper you recommended argues as follows: Suppose we know that a pair of dice has been thrown either once or twice. Further we know that at least once a (relatively rare) double six was rolled. This latter fact counts as evidence for the hypothesis that the pair of dice has been thrown twice. So far so good. But then he argues that there is a “straightforward” analogy with the fine-tuning of the universe, and I quote: “Replace the dice throws with universes, and replace the result of getting a double six on a throw with getting a universe with the right constants for life. The probability of there being life given only one universe is very small. But if there are enough universes, the probability that there exists one with the right constants for life becomes very high. Thus the evidence that there is at least one universe with the right constants for life [namely ours] confirms the hypothesis that there are many universes.” But to go from dice throws to universes is a huge conceptual jump. After all we know that reality is such that either one or two throws of dice can take place, but we do not similarly know that reality is such that one or perhaps many universes exist. Indeed, by definition, if one universe exists then reality is not like that.

    One can see the fallacy of Bradley’s argument by considering the following close analogy: “The probability that a universe with the same laws and initial conditions of ours would produce the Mona Lisa painting is vanishingly small. But if there are enough universes with the same initial state as ours, the probability that there is at least one universe which would produce that painting becomes very high. Thus the evidence that there is at least one universe which produced the Mona Lisa (namely ours) confirms the hypothesis that there are many universes with the same initial state as ours.

  • http://www.blogger.com/profile/05693985638589020492 Mark

    You're confusing "unexpected" and "expected not to be the case." We actively expect a fair die not to exhibit simple patterns in a sequence of flips. The probability of such a sequence happening is very low relative to the hypothesis of an unfair die. But naturalism doesn't assign a low probability to the universe's intelligibility (to humans). This is because it doesn't assign any sort of probability at all to such a statement. So while it's true, therefore, that its human-intelligibility is "unexpected," it's not unexpected in the sense necessary to generate a problem. At least proponents of the fine-tuning argument claim to have a basis for their probability assignments in the Principle of Indifference applied to the values of physical constants (treated as random variables).

    On the assumption that naturalism is true the fact that the fundamental constants appear fine-tuned is unexpected.

    No, it's (supposed to be) unexpected under the single-universe hypothesis plus naturalism, not under naturalism simpliciter. That's why fine-tuning is purported to be evidence against single-universe naturalism and for multi-universe naturalism. Fine-tuning is only evidence against naturalism simpliciter if you can additionally show that fine-tuning is more likely under theism than under multi-universe naturalism (which, AFAIK, no one in the literature has really tried to do, at least without resorting to the subjective theory of probability). You can try to insist that the existence of multiple universes is low on naturalism, but for this assertion you will only be able to produce the wholly question-begging argument that multiple universes has no evidence speaking for it given naturalism. (Fine-tuning is supposed to be the evidence!) It's not like current scientific theories say that there can't be any other universes.

    So unlike your comparison cases, we're not saying, "A is evidence against B, but since we know B is true, A has to be false." We're instead saying, "We have evidence for B. If B, then either C or D. But A is much likelier if B&D; is true than if B&C; is true. Therefore we have evidence for B&D.;" Here, B = naturalism, C = single-universe hypothesis, D = multiverse hypothesis and A = the observation of fine-tuning.

    But to go from dice throws to universes is a huge conceptual jump. After all we know that reality is such that either one or two throws of dice can take place, but we do not similarly know that reality is such that one or perhaps many universes exist.

    Of course we know (tautologously) that either one or more than one universe exists. What you need to be saying is, "We do not similarly know that reality is such that many universes can exist." But exactly parallel worries come up for theism, and in perhaps nastier form. We don't know that reality is such that a universe-shaping deity can exist and play with the values of physical constants. Moreover, positing such a deity is less qualitatively parsimonious than positing multiple essentially universes just like ours but differing slightly in their parameters.

    Re: the Mona Lisa thing. You need to read Bradley more carefully. He's at pains to point out the crucial role of observation selection effects (OSEs) in the inference to multiple universes. But there isn't an OSE in place in observing the Mona Lisa. So I believe you're attacking a straw man.

  • http://www.blogger.com/profile/05693985638589020492 Mark

    The probability of such a sequence happening is very low relative to the hypothesis of an unfair die.

    Whoops, this should've said "low relative to the hypothesis of a fair die, but higher relative to the hypothesis of an unfair die."

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    One way in which the universe has the appearance of design is the fact that intelligent creatures came into existence in the history of the universe.

    There are three possibilities concerning this fact:

    1. Given the physical nature of the universe, the probability that intelligent creatures would arise is low.

    2. Given the physical nature of the universe, the probability that intelligent creatures would arise is moderate.

    3. Given the physical nature of the universe, the probability that intelligent creatures would arise is high.

    If the probability is thought to be low, then theists say "The physical nature of the universe cannot reasonably explain the fact that intelligent life arose, so we can infer the existence of an intelligent designer."

    If the probability is thought to be high, then theists say "The physical nature of the universe reasonably explains the fact that intelligent life arose, but the fact that the nature of the universe is such as to make this outcome probable, we can infer the existence of an intelligent designer."

    If the probability is thought to be moderate, then theists say "The physical nature of the universe explains the fact that life arose, and the fact that the nature of the universe is such as to make this outcome somewhat probable, we can infer the existence of an intelligent designer who respects our free will and so did not want to make the evidence for his/her existence so overwhelming as to make disbelief in the existence of the designer nearly impossible."

    In short, the theist says "Heads I win, tails you lose."

  • http://www.blogger.com/profile/05693985638589020492 Mark

    Of course, we can also claim that the life-friendliness of a designer is improbable and itself "cries out for an explanation." I'd say all of this hints at there being something very wrong with this type of reasoning. (And IMO, Elliott Sober points out exactly what the error is.)

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Mark:

    You are saying: "We have evidence for B. If B, then either C or D. But A is much likelier if B&D; is true than if B&C; is true. Therefore we have evidence for B&D.;" Here, B = naturalism, C = single-universe hypothesis, D = multiverse hypothesis and A = the observation of fine-tuning.

    How is that reasoning different from a theist saying the same with A = theism, C = the hypothesis that God does not have morally sufficient reason for allowing suffering, D = the hypothesis that God does have morally sufficient reason for allowing suffering, and A = the observation of suffering ? I don’t see any difference, so by the same logic the theist is justified in claiming that our observation of suffering is evidence that God has morally sufficient reason for allowing it.

    Of course we know (tautologously) that either one or more than one universe exists.

    Well, I disagree. A philosopher should consider the semantic content of a proposition beyond its syntax. The proposition “one or more than one universe exists” is a claim about how reality is, and reality may well not be like that. One cannot reasonably take a known property about things that exist within reality (such as one or many dice throws can exist) and project that property to things that potentially refer to the whole of reality (such as that one or many universes can exist), and moreover claim that the same probability analysis that the former conforms to also applies to the latter. That, it seems to me, is just playing with words, or rather letting words drive one's thinking.

    In conclusion I am not denying that the cosmological multiverse hypothesis successfully responds to the fine-tuning argument; in the same way that the hypothesis that God has morally sufficient reason for allowing suffering successfully responds to the argument from evil. My point is that as long as the naturalist or the theist cannot offer independent evidence for the respective hypothesis the whole thing remains a purely ad-hoc conjuring trick.

    Shifting gears into a more personal note: Doesn’t it trouble you that naturalists, in order to fit scientific discoveries with their ontological assumptions, find themselves obliged to multiply the universe beyond counting (thus grossly violating Occam’s principle)? And it’s not just the case of the cosmological fine-tuning. The most popular interpretation of quantum mechanics among naturalistic scientists is the many worlds interpretation, which is an even more exorbitant hypothesis. For which, again, no evidence exists.

    But exactly parallel worries come up for theism, and in perhaps nastier form.

    My contention here is not theism versus naturalism, but good reasoning versus bad reasoning, wherever it’s being used.

    But naturalism doesn't assign a low probability to the universe's intelligibility (to humans). This is because it doesn't assign any sort of probability at all to such a statement. So while it's true, therefore, that its human-intelligibility is "unexpected," it's not unexpected in the sense necessary to generate a problem.

    If so, it should. Any fact that makes the universe appear designed does not fit with the premises of naturalism and thus requires an explanation – whether it’s a fact about the apparent design of the species, or of the fundamental constants, or of the property of the universe to be such as to blindly produce an intelligence capable of understanding it, or of the property of the universe to be mathematically elegant. The last two facts refer to the universe’s intelligibility, and here not even multiplying the universe works as a possible solution. Conversely any fact that makes the universe appear imperfectly designed (be it moral evil, or natural evil, or animal suffering, or God's hiddenness, etc) does not fit with theism's assumptions and requires an explanation from theists.

  • http://www.blogger.com/profile/05693985638589020492 Mark

    How is that reasoning different from a theist saying the same with A = theism, C = the hypothesis that God does not have morally sufficient reason for allowing suffering, D = the hypothesis that God does have morally sufficient reason for allowing suffering, and A = the observation of suffering ?

    I take it you mean B = theism here. Well, I'd agree that A does indeed favor the hypothesis B&C; over the hypothesis B&D.; The difference is that it favors the atheistic hypothesis just as much as B&C;, and that the atheistic hypothesis has much more prior plausibility than B&C.; (It is, for instance, much simpler and invokes fewer unanswered mysteries.) The theistic design hypothesis, in contrast, isn't more parsimonious than the multiverse hypothesis, however, unless you confuse quantitative and qualitative parsimony.

    Well, I disagree. A philosopher should consider the semantic content of a proposition beyond its syntax.

    No idea what you're talking about here. The proposition "Necessarily, if something is X, then either one or multiple things are X" is about as close to self-evident as it's possible to be. This is not something I'm interested in defending.

    Shifting gears into a more personal note: Doesn’t it trouble you that naturalists, in order to fit scientific discoveries with their ontological assumptions, find themselves obliged to multiply the universe beyond counting (thus grossly violating Occam’s principle)?

    It's a much less radical addition to my ontology than a timeless, spaceless agent who can mysteriously warp physics with the power of sheer will. Multiple universes are basically just more of the same; an atemporal God would be something completely new.

    Any fact that makes the universe appear designed does not fit with the premises of naturalism and thus requires an explanation – whether it’s a fact about the apparent design of the species, or of the fundamental constants, or of the property of the universe to be such as to blindly produce an intelligence capable of understanding it, or of the property of the universe to be mathematically elegant.

    No offense, but I don't think anything you're saying here is grounded in anything but your own private sense of incredulity. At least, I haven't seen anything that gives me any reason to revise my beliefs.

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Mark:

    Well, I'd agree that A does indeed favor the hypothesis B&C; over the hypothesis B&D.;”.

    But this result, namely that the presence of suffering is evidence (or “favors the hypothesis”) that God has morally sufficient reason to allow it, is clearly absurd which shows that the reasoning that led to it is faulty. But this is the same reasoning you suggested for justifying the claim that the fine-tuning is evidence for the existence of parallel universes. Incidentally I don’t think it’s difficult to see where the fault in this line of reasoning lies: If A appears to contradict B then one can resolve the apparent incoherence by showing that there exists a *possible* C which is compatible with both A and B. Such reasoning does indeed resolve the apparent contradiction, but does not justify the claim that A is evidence for C. Rather C remains an ad-hoc hypothesis, the only worth of which is that it resolves an apparent contradiction. Plantinga in his discussion of the problem of evil is especially clear on this point.

    [The atheistic hypothesis is] much simpler and invokes fewer unanswered mysteries

    Well, I disagree. This might have been the case in the late 19th century, but no more. Today, both because of scientific knowledge and philosophical considerations, naturalism suffers from mysteries galore. Perhaps you’d like us to make a list.

    The proposition "Necessarily, if something is X, then either one or multiple things are X" is about as close to self-evident as it's possible to be.

    Not necessarily, for X may be a property of non-countable things. The universe is by definition the set of all physical things we observe, such as dice throws. It is not in general valid to apply properties of the members of a set to the set itself, so the fact that dice throws are countable does not imply that the universe is countable also. And even if one assumes the latter, the fact that probability rules apply to dice throws does not imply that the same apply to universes. So, far from us having a “straightforward analogy” as Bradley claims, we here have nested arbitrary assumptions. One cannot establish evidence for a hypothesis on such.

    It's a much less radical addition to my ontology than a timeless, spaceless agent who can mysteriously warp physics with the power of sheer will.

    I understand then you are troubled, but not sufficiently.

    Multiple universes are basically just more of the same; an atemporal God would be something completely new.

    Two things: First it seems you are changing Occam’s razor from “do not multiply entities” to “do not multiply types of entities”. Speaking of my own intuition of reason, even though I agree that multiplying either entities or types should be avoided, I don’t see why the latter carries infinitely more weight than the former. After all the naturalist is arguing that it is better to multiply the universe a gargantuan number of times than to multiply the ontological substances by 2. Secondly, not all theistic ontologies are dualistic. Indeed Berkeley’s subjective idealism is monistic, and this is in my judgment the by far more powerful theistic ontology.

    Incidentally it is clear that God has some atemporal properties; whether God has only atemporal properties is being debated and is indeed highly implausible in my view.

    No offense, but I don't think anything you're saying here is grounded in anything but your own private sense of incredulity.

    I, on the other hand, can hardly imagine a freethinker considering the intelligibility of the universe in its various forms and not being stricken by how utterly designed, or put-in-place-for-a-purpose, it all looks. Naturalists must face all apparent design in the universe and explain it on naturalistic grounds. Denying a problem does not remove it, as Daniel Dennett found out in relation to the hard problem of consciousness.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    Consider the third of three possible views of the probability of the appearance of intelligent creatures:

    3. Given the physical nature of the universe, the probability that intelligent creatures would arise is high.

    This appears at first to be a concession on the part of theists and a victory for atheism, but then theists push the explanatory issue back one step, from the origin of intelligent creatures to the origin of a universe that has the propensity to give rise to such creatures. The theist says:

    "The physical nature of the universe reasonably explains the fact that intelligent life arose, but from the fact that the nature of the universe is such as to make this outcome probable, we can infer the existence of an intelligent designer who ensured the universe would have just the right features and characteristics to make intelligent life a likely outcome."

    The fine tuning argument appears to make the same sort of inference about the physical nature of the universe, as theists previously made about the appearance of intelligent creatures. Specifically, the idea is that left to itself, apart from divine intervention, the probability is low that the universe would have a physical nature that has a propensity for producing intelligent creatures.

    Suppose, however, that further scientific investigation were to show this assumption to be false. Suppose that in the 22nd century it was established that the probability is actually high that the universe would have a physical nature that has a propensity for producing intelligent creatures.

    At that point, could theists make the same logical move that was made concerning the origin of intelligent creatures? Could theists push the explanatory question back one step further, and say, "Isn't it amazing that the factors that shaped the physical nature of the universe were such as to make it highly probable that the physical nature of the universe would end up with a propensity for bringing about intelligent creatures? Doesn't this give us good reason to infer the existence of an intelligent designer?"

    Is there any way to win this game?

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Bradley:

    That’s an interesting analysis. Indeed it seems that the theist can build a good argument from the intelligibility of the universe which is broadly independent from how probable it is, given the physical nature of the universe, for creatures intelligent enough to understand it to naturalistically evolve. Which is a good thing, because the scientific computation of this probability is probable intractable. After all we still don’t know how probable it is, given the physical nature of the universe, for creatures as complex as we are to naturalistically evolve, and this latter question is much simpler. (I personally agree with naturalists’ guess that the probability for both is high, but we really don’t know.)

    Incidentally, I think that the more difficult problem related to the universe’s intelligibility is about the mathematical elegance of the universe. Here I believe the estimation of probabilities is more straightforward. For each mathematically elegant universe in which intelligent beings can evolve via natural selection there must be a gazillion mathematically unintelligible universes in which intelligent beings can evolve just as well. It is then a remarkable fact that we should live in an elegant universe, no matter how many parallel universes may exist.

    You write: “Is there any way to win this game?

    According to the traditional theistic view God doesn’t just exist, but exist necessarily. By this I understand that God exists in all possible worlds that minimally comport with our experience of life. If that is so then the answer to your question can’t be positive. Indeed, if God exists then it would seem that God, being the foundation of reality, is ultimately an inescapable object of knowledge any way one engages with reality: be it through the intellect, or be it existentially in living the good life, or be it in contemplation, or even perhaps in artistic creation.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    Dianelos said:

    "Indeed it seems that the theist can build a good argument from the intelligibility of the universe which is broadly independent from how probable it is, given the physical nature of the universe, for creatures intelligent enough to understand it to naturalistically evolve."

    Response:

    The intelligibility of the universe is certainly a separate consideration from the consideration of the origin of intelligent creatures.

    But isn't the logic of the arguments similar? Does probability play no role in your argument from the intelligibility of the universe? If the intelligibility of the universe could (somehow) be shown to be a probable outcome of some prior natural conditions and processes, then would your argument still work?

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Bradley:

    You said: “If the intelligibility of the universe could (somehow) be shown to be a probable outcome of some prior natural conditions and processes, then would your argument still work?

    By “intelligibility” I suppose you mean mathematical elegance which I have been arguing is necessary for the known remarkable success of science, and which represents the most difficult version of the problem.

    Before commenting, I’d like to mention a parallel argument I have heard in relation to the apparent fine-tuning of the fundamental constants (as well as of the initial conditions) of the universe. The idea is that the universe appears so fine-tuned because there are many such "brute fact" numerical values, and each appears to be relatively fine-tuned so that the product of all factors becomes immense. But science, which has uncovered these numerical values, is an on-going process, so perhaps it is the case that these numerical values are not independent as we assume, but are internally related through some simple mathematical formula. If so, the apparent fine-tuning of the universe would be the product of less factors, and hence be much lower. Indeed, for all we know, all fundamental values may be related in a simple way with the value of pi, which is a simple non-contingent number. In this latter case only one physical universe of the type that science studies (i.e. which comports with the form of the scientific equations) is possible, and hence the numerical values present could not possibly be different.

    It seems to me that this argument does demonstrate that the apparent fine-tuning of the universe is perhaps illusory. But even if this argument works it only represents a very weak naturalistic answer, equivalent to a theist arguing that perhaps there are morally sufficient but as yet undiscovered reasons for God to allow the evil we observe.

    Coming back to your question above, I don’t think the same kind of argument can work in the case of the mathematical elegance of the universe. The blind Darwinian process which all naturalists believe explains the appearance of functional complexity in the universe, is well understood. We know that in order to work the Darwinian algorithm places some requirements on its physical environment, e.g. this environment must allow for material configurations to be stable enough for inheritance to be possible but not too stable, otherwise mutations wouldn’t be effective. On the other hand we can visualize a wide range of possible physical environments in which the Darwinian algorithm does work and which are not mathematically elegant. This range vastly outnumbers the range of elegant environments. If, as you suggest, the known elegance of our universe could be reduced to prior natural conditions then the elegance of our universe would actually increase and thus become even more rare among the Darwinian universes. (Indeed the hypothesis discussed above, namely that all numerical brute facts known to science are related to pi, vastly increases the elegance of our universe.)

    So the argument from the mathematical elegance of our universe against naturalism does not admit even an ad-hoc solution, such as that there are perhaps many parallel universes, or that there are perhaps still unknown scientific relationships to be discovered (for the already known mathematical elegance of the universe suffices). It seems that what science has removed with the one hand from the theistic argument from design (by discovering Darwinism), it has returned many times over with the other hand (by discovering the apparent fine-tuning of the universe's constants, the universe's property of producing intelligence able to understand it, and especially the universe's mathematical elegance).

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    Bradley said: “Is there any way to win this game?”

    Dianelos responded:

    "According to the traditional theistic view God doesn’t just exist, but exist necessarily. By this I understand that God exists in all possible worlds that minimally comport with our experience of life. If that is so then the answer to your question can’t be positive. Indeed, if God exists then it would seem that God, being the foundation of reality, is ultimately an inescapable object of knowledge…"
    =======

    Response:

    I'm unclear on how necessary existence relates to our ability to know that God exists.

    Even if "God exists" were an analytic proposition (like "All triangles have three sides") it might be an analytic proposition that is so subtle and complex that it is beyond the power of human minds to grasp, and thus is something that humans cannot know to be an analytic proposition (or know to be true).

    Furthermore, if God does exist in all possible worlds that "minimally comport with our experience of life" wouldn't it be the case that some of those worlds are such that the evidence available to human beings pointed to the non-existence of God? Can't we conceive of a world in which God exists (and has necessary existence) but the evience available to humans (that humans can know and understand) does not support belief in God, but rather supports atheism? If such a world is conceivable, then doesn't that sever the logical tie between God's necessary existence and human knowledge of God's existence?

    If we cannot conceive of a world in which God exists (and has necessary existence) but in which the evidence available to humans fails to point to God's existence, then the evidence for God's existence is indeed inescapable, but only if God exists and has necessary existence. So, this would still leave open the possibility that the evidence available to human beings fails to support belief in God's existence, because there is no God and thus no God with necessary existence.

    Am I missing something here?

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Bradley Bowen said: “ Can't we conceive of a world in which God exists (and has necessary existence) but the evidence available to humans (that humans can know and understand) does not support belief in God, but rather supports atheism?

    No, we can’t conceive of such a world, because such a world would be created by an all-good and all-powerful God and thus would be such that knowledge of God was available to humans (by “evidence” I assume you mean “good reasons for believing”). As Einstein in a different context said, “God is subtle but not malicious”.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    Dianelos said:
    "No, we can’t conceive of such a world, because such a world would be created by an all-good and all-powerful God and thus would be such that knowledge of God was available to humans (by “evidence” I assume you mean “good reasons for believing”)."

    Response:

    The concept of God need not imply that God created this universe, unless you build that into your definition of God. If we understand "God" to mean something like "the perfect person", then it does not logically follow that

    (A) If God exists, then this universe was created by God.

    Did God create the atomic bomb? No, we humans did. God, if he/she exists, permitted human beings to create an atomic bomb.

    Did God create this universe? Perhaps, but God, if he/she exists, might have permitted some imperfect intelligent designer to create this universe, just as God permitted us imperfect human beings to create an atomic bomb.

    Would an all-good deity permit a universe to be created by an imperfect intelligent designer? Why not? If an all-good deity permitted imperfect human beings to create an atomic bomb, then I don't see any reason why such a deity would balk at the creation of a universe by an imperfect intelligent designer.

    If you want to take the easy way out and simply stipulate a definition of "God" so as to make being the creator of this universe a necessary condition for that application of this term, then it is still not clear to me that your point follows.

    Why can't an all-good deity create a universe in which the evidence available to some intelligent creatures fails to point to his/her existence? Why should an all-good God be morally obliged to make all intelligent creatures aware of his/her existence?

    If knowledge of God's existence was essential to human happiness and well-being, then I suppose it would be unkind of God to create a universe with human beings such that the evidence available to those humans failed to make God's existence obvious or probable. But it is far from clear to me that knowledge of God's existence is essential to human happiness and well-being (even assuming that there is a God).

  • http://www.blogger.com/profile/17239457772830013242 tmdrange

    Bradley wrote: "If knowledge of God's existence was essential to human happiness and well-being, then I suppose it would be unkind of God to create a universe with human beings such that the evidence available to those humans failed to make God's existence obvious or probable. But it is far from clear to me that knowledge of God's existence is essential to human happiness and well-being (even assuming that there is a God)."
    I have two comments. First, it would be unkind of God to create such a world even if knowledge of him were not essential, but merely LIKELY to enhance happiness and health.
    And second, it is interesting that Christianity implies that knowledge of God is essential to happiness and well-being, since it claims that such knowledge is essential for salvation.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    tmdrange said:

    "…it would be unkind of God to create such a world even if knowledge of him were not essential, but merely LIKELY to enhance happiness and health."

    Response:

    I disagree. God would not be unkind to fail to provide one particular possible way of enhancing human happiness and well being. Swinburne has argued that God is not morally obligated to provide or perform every possible good thing or deed. As Kant pointed out, with imperfect duties (I think that was his term) one can always do a bit more (e.g. give a bit more money to the poor).

    Suppose that God provided humans with what was necessary to acheive a moderate level of happiness and well being (in this life) and God also provided us with means of enhancing that happiness and well-being apart from obtaining awareness or knowledge of God's existence.

    For example, suppose that practicing Transcendental Meditation provided the same level of enhancement to human happiness and well-being as the knowledge of God's existence. In that case, God would be providing us with a means of obtaining a higher level of human happiness and well-being without creating a universe such that the evidence available to humans made God's existence obvious or probable.

    I would not consider such a God to be "unkind" or even morally blameworthy for failing to provide a particular method of enhancing human happiness and well-being.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    tmdrange said:

    "it is interesting that Christianity implies that knowledge of God is essential to happiness and well-being, since it claims that such knowledge is essential for salvation."

    Response:

    If knowledge of God's existence was a necessary condition for humans to obtain eternal life filled with happiness and well-being and/or for avoding eternal life filled with misery and suffering, then God would indeed be unkind and even immoral to create a universe in which the evidence available to humans failed to make God's existence obvious or probable.

    We could then justly complain to God, as Bertrand Russell suggested, "God, why did you not make your existence more evident?"

    But what reason is there to believe that knowledge of God's existence is a necessary condition for eternal happiness and well-being, or for avoiding eternal misery and suffering?

    If one has to swallow all of the implausible assertions of traditional Christian theology first, in order for a case to be made that God is morally obliged to create only universes in which humans have ample evidence of his/her existence, then the game is over from my point of view, and the position has been reduced to absurdity.

  • http://www.blogger.com/profile/17239457772830013242 tmdrange

    Bradley wrote: "Suppose that God provided humans with what was necessary to achieve a moderate level of happiness and well being (in this life) and God also provided us with means of enhancing that happiness and well-being apart from obtaining awareness or knowledge of God's existence. … I would not consider such a God to be 'unkind' or even morally blameworthy for failing to provide a particular method of enhancing human happiness and well-being."
    I agree with that, but it is not relevant to what I wrote. Merely to provide a means of enhancing happiness and well-being is not the same as doing something that would make it LIKELY that happiness and well-being would be enhanced. Suppose that, on a scale of 0 to 10, the average human's happiness and well-being is at level 4, and that for God to provide good evidence of his existence would LIKELY raise that level to 6. I think that in that circumstance, it would be unkind of God to fail to provide the evidence.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    tmdrange said:

    "Merely to provide a means of enhancing happiness and well-being is not the same as doing something that would make it LIKELY that happiness and well-being would be enhanced. Suppose that, on a scale of 0 to 10, the average human's happiness and well-being is at level 4, and that for God to provide good evidence of his existence would LIKELY raise that level to 6. I think that in that circumstance, it would be unkind of God to fail to provide the evidence."

    Response:

    My point still holds when put in terms of God doing or providing something that "would likely raise" the level of human happiness and well being.

    If providing humans with citrus fruits would likely raise the well-being level from 4 to 6, and if the likelihood of this beneficial result was as high as the likelihood of this beneficial result from providing evidence of his/her existence, then there would be no unkindess or moral failure in using the former means instead of the latter.

    Of course, to the extent that various alternative ways (of making it likely that the level of human happiness and well-being is boosted) are causally and logically independent, adding more and more such alternatives would increase the probability of the beneficial results, but then we are back to the issue of imperfect moral duties.

    God can always add one more way and increase the probability a little bit more. It is questionable to assume that God must satisfy imperfect moral duties to an infinite degree.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    Back to my earlier question…

    Is there any way to win this game?

    Dianelos said:

    "If, as you suggest, the known elegance of our universe could be reduced to prior natural conditions then the elegance of our universe would actually increase and thus become even more rare among the Darwinian universes. (Indeed the hypothesis discussed above, namely that all numerical brute facts known to science are related to pi, vastly increases the elegance of our universe.)"

    This appears to confirm my suspicion that there is no possible configuration of evidence which would count against the intelligent designer hypothesis.

    I'm not sure I understand what Dianelos is saying above, so I will restate it in other words and see whether Dianelos agrees with how I interpret his comments.

    I think there are two claims operative here:

    1. If the natural conditions prior to the origin of this mathematically elegant universe do not make such a universe probable, then it is reasonable to infer that an intelligent designer intervened to ensure the mathematical elegance of this universe.

    2. If the natural conditions prior to the origin of this mathematically elegant universe do make such a universe probable, then it is reasonable to infer that an intelligent designer had established those particular natural conditions in order to ensure the origin of a mathematically elegant universe.

    Did I get that correct?

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Bradley said: “ The concept of God need not imply that God created this universe, unless you build that into your definition of God. If we understand "God" to mean something like "the perfect person", then [snip]

    “God is the perfect person” is the commonly used and shorter version of St. Anselm’s definition, namely that God refers to what nothing greater can be conceived. In that sense it does follow that God is the direct creator of the universe, for between two conceptions of God, namely 1) a quasi deist, kind of lazy and kind of not-really-here God, who perhaps via the help of others (as you suggest) creates an independently working universe, sets the whole thing in motion, and then only here and there supernaturally interferes, and 2) a “ground-of-all-being” God where all existence and all change are continuously contingent on Him/Her, so that the physical universe around us in all its order and splendor is nothing but one more manifestation of God – it seems clear to me that the latter concept is the greater.

    In general I would say that a common error that many people (theists and atheists alike) commit while thinking about God is to fail to consider the really greatest possible concept. If one applies even modest discounts to the concept of God then one is bound to do bad theology.

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Tmdrange said: “First, it would be unkind of God to create such a world even if knowledge of him were not essential, but merely LIKELY to enhance happiness and health.

    Only if that enhancement in happiness and health were not to the detriment of what God’s primary purpose in creation is. I learned this from reading your book by the way. So the issue turns on what a perfect person’s primary purpose in creation would be. What is that of the greatest value that God could possibly create? This is the fundamental question which theodicy must answer. In my judgment the best answer is the one that Irenaean theodicy gives, namely that God’s purpose is the creation of persons who, on their own merit, become as perfect as S/He is. Which immediately explains why the existence of evil is necessary. To mention just one example, one cannot be considered courageous if one has never acted courageously in one’s life, and one cannot have thus acted without facing some state of affairs in which courage is required, and such a state must be one where evil is present (and indeed one where one’s powers are limited).

    Coming back to the issue at hand, John Hick in his development of the Irenaean theodicy argues that a world optimized for God’s purpose must be one where there is some epistemic distance between God and us. I think this is doubtlessly true, for if the reality of God were absolutely obvious to us then all personal merit and opportunity for growth would disappear. To use the above example, what merit would there be in one acting with courage if the reality of God were absolutely obvious to us? What merit would there be in charity? Or in forgiving? Or in self-transcendence? Of course the other extreme, namely that of God being absolutely unknowable, would not be optimal either. So the only question is whether the relative cognitive difficulty of recognizing the presence of God entailed in our current condition is optimal or not. This is a complex question to think about, but I personally see no reason to doubt that our condition is optimal in this respect too. After all, there are more than one ways to recognize the presence of God. Even the intellectual way is, in my judgment, not really that hard either, as long as one centers one’s critique of theism on one’s judgment of perfection (rather than, bizarrely, on those parts of the Bible or on those theological claims which contradict that very judgment). And also as long as one applies the same epistemic criteria to both theism and naturalism and compares them directly. Ultimately, it seems to me, the most effective way to do ontology is by comparing one to one alternative ontological worldviews and finding out which works better.

    Tmdrange said: “And second, it is interesting that Christianity implies that knowledge of God is essential to happiness and well-being, since it claims that such knowledge is essential for salvation.

    The view that salvation is by faith alone (i.e. that holding true beliefs about God is both a necessary and sufficient condition for salvation) is a characteristic of the Protestant denominations, and not of Christianity in general. When I read the Gospels it seems to me clear enough that what Jesus asked is for us to follow a particular way of life. He said over and over again: “Do this, do that, love exorbitantly, have trust in God and do not fear” and not really "be careful to form the right beliefs about God". Actually all great religions teach the same: that there is a path of self-transformation by which one realizes the purpose of life. And all great religions describe that path in pretty similar terms, namely as the path of self-transcending love. I have personally no doubt that a non-religious person who nevertheless follows this path becomes transformed by it and thus finds salvation. And the fact that there are non-religious people who follow it only demonstrates how powerfully we sense the tug of the ultimate goodness of reality.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    Dianelos said:

    " 'God is the perfect person' is the commonly used and shorter version of St. Anselm’s definition, namely that God refers to what nothing greater can be conceived. In that sense it does follow that God is the direct creator of the universe… a 'ground-of-all-being' God where all existence and all change are continuously contingent on Him/Her, so that the physical universe around us in all its order and splendor is nothing but one more manifestation of God – it seems clear to me that the latter concept is the greater."

    Response:

    In Platonic philosophy, the creator of the physical world is a quasi-deity, a less-than-perfect being that is closer to God (or the form of the Good) than the physical world.

    This is partly because of the Greek view that matter is evil or imperfect and thus it was inappropriate for God to be the direct creator of material beings. Nevertheless, this illustrates how it is less than self-evident that a God as "the perfect person" or as the "ground of all being" must be the direct creator of our universe.

    My previous question still applies: Did the "ground of all being" create the atom bomb?

    If you answer "Yes", then you admit that God can create things through less-than-perfect intermediaries (such as human beings).

    If you answer "No", then you admit that God permits significant creation activity by less-than-perfect inteligent beings.

    Of course, the basic materials that we humans used to construct atom bombs already existed when we came along. But this universe could have had a similar origin, with the "ground of all being" providing some basic stuff, and some less-than-perfect intelligent creature shaping and altering that basic stuff into the mathematically elegant and life-friendly universe that we call home.

    I don't see how it follows from the concept of God as the "ground of all being" that God must be the direct creator of our universe.

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Bradley:

    In the previous post I tried to explain that when Ι say that “God created the universe” Ι mean this in a much stronger sense than when one says “a scientist created the atom bomb”, because in the latter case the bomb once created is out of the scientist’s hands, whereas in God’s case the very existence and order of the universe are continuously contingent on God’s so-called general providence. Creation can’t get more direct and immediate than that. Further I argued that the concept of God who creates an imperfect universe via the action of intermediaries is clearly less great than the concept of God who directly creates a perfect universe (in relation to His/Her primary purpose). It is on the latter claim that theism stands or falls.

    More specifically, you ask: “ Did the "ground of all being" create the atom bomb?

    I would answer, yes and no, for God created us living in this universe in which we are capable and can freely choose to create the atom bomb.

    In this context, I think it is useful to discuss an ambiguity in the terms “create” or “design”. Here is an example out of my personal life: As it happens I am a computer programmer. Once I needed a large number (32 if I remember well) of different hash functions. Hash functions are functions that approximate a clearly defined mathematical property. Instead of directly designing this large number of hash functions, which is extremely difficult to do by the way, I programmed a computer to randomly create functions one after the other, and to automatically measure how well they approximated the needed mathematical property. After literally hundreds of millions of tries the computer output the 16 best functions thus created, which I then used in my larger project. Now the question is: Did I actually design these functions, or not? In a sense I did because they were the result of my work. But in a sense I didn’t because I created them without knowing how they worked. Indeed I remember I picked one of these functions and tried to understand why it worked so well, but couldn’t. – My larger point here is that questions such as “Did God create evil?” or “What design purpose does our appendix serve?” are too simplistic to allow for a direct answer.

    In any case, according to Irenaean theodicy the universe is just a means, and we are actually God’s co-creators of what God wants to achieve. To use John Hick’s terminology, God built the universe with the purpose to give us the opportunity to build our souls.

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Bradley said: “ Back to my earlier question…

    Is there any way to win this game?

    In the context of the theistic argument from the mathematical elegance of our universe I cannot see what kind of solution naturalism allows. As I argued before, defenses such as the hypothesis that many parallel universes exist, do not work in this case. But there may be other possible solutions, who knows? Certainly a naturalist must assume that there are and keep trying to find them.

    Bradley said: “ This appears to confirm my suspicion that there is no possible configuration of evidence which would count against the intelligent designer hypothesis.

    Or at least, given the strength of the argument from the elegance of our universe, it is difficult to imagine what kind of evidence would weaken it. It’s even difficult to imagine any hypothesis, even one with no evidence speaking for it, that might help.

    Bradley said: “ 2. If the natural conditions prior to the origin of this mathematically elegant universe do make such a universe probable, then it is reasonable to infer that an intelligent designer had established those particular natural conditions in order to ensure the origin of a mathematically elegant universe.

    Did I get that correct?

    Well, I had not discussed the possibility of prior conditions to the origin of our universe, so let’s consider this possibility here:

    The strength of the argument from the elegance of our universe depends on the realization that the ratio of elegant Darwinian universes is very small. What happens if one moves the problem one step back and considers the natural conditions prior to the origin of our universe, which for all we know might have existed and might have caused our universe to come into existence? Well, given the second law of thermodynamics (which appears to be a necessary law of any mechanistic worldview rather than just a contingent law which applies only to our universe) we see that there are many possible natural conditions that would produce (or would probably produce) an inelegant universe, but only a few possible natural conditions that would produce (or would probably produce) an elegant universe. (For the same reason that there are many possible prior conditions that would cause a disorderly, high entropy, physical state, but only few possible prior conditions that would cause an orderly, low entropy, physical state.) Therefore the ratio of prior conditions that would cause an elegant universe to come into existence is even smaller than the one we had been considering. Which means that by moving the problem one step back one only gets an even harder problem for naturalism.

    For the same reasons given above, moving the problem even more back and hypothesizing that our universe was after all purposefully designed and created by very powerful, intelligent, out-of-this-world beings, but still by beings who function mechanistically within a mechanistic reality and hence conform to naturalism – does not work either.

    In conclusion it seems that the argument from the mathematical elegance of our universe (which property the natural sciences have by now so overwhelmingly demonstrated) represents an extremely strong version of the argument from design. It will be interesting to wait and see whether naturalists will be able to come up with even only an ad-hoc solution to it. I personally cannot imagine one.

  • http://www.blogger.com/profile/05693985638589020492 Mark

    The strength of the argument from the elegance of our universe depends on the realization that the ratio of elegant Darwinian universes is very small.

    This is the central confusion I was trying to point out from the start. There is no ratio of elegant universes to inelegant universes. There are infinitely many elegant universes and infinitely many inelegant ones. You could get around this if you could define some probability measure P on the space of possible universes and show that the subset of possible universes in which the laws of physics are "elegant" has a small value under P. However, I have no idea how such a measure could be defined without some way of parametrizing the set of all possible universes. Worse, even if you were able to do this, you still wouldn't have shown that your probability measure is the "correct" one, or the one that's the most reasonable under naturalism. You'll run into the same problem that the Fine-Tuning Argument does, namely, that there are many different ways of parametrizing a sample space which yield different probability measures when applying the Principle of Indifference. You'll then be forced to retreat away from all such measure-theoretic arguments into the territory of the subjective theory of probability. And once there, your argument will have lost all force for those who don't share your prior credences.

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Mark:

    First of all, the ratio of two infinite quantities is well defined in many cases. So, for example, the ratio of prime numbers to natural numbers is zero. As for there being an infinite number of mathematically elegant universes – this assumption may actually not be true. Perhaps there is only a limited number ways one can construct a universe as deeply mathematical as ours.

    In any case, the justification for the belief that the ratio of mathematically elegant Darwinian universes is small can be given on at least two levels:

    1) Our experience with artificial intelligence. It turns out that the Darwinian algorithm has been simulated many times in computers (see the field of “genetic programming”) and even though some care must be given in fine-tuning the parameters of the respective artificial environment there is no need whatsoever for that environment to be mathematically elegant. So there is actually experimental evidence for the belief that the ratio of elegant Darwinian universes is small. I will concede that this evidence is not absolutely conclusive (the current implementations of the Darwinian algorithm are relatively simple and are as yet incapable of producing intelligent organisms), but the combination of this evidence with our knowledge of Darwinism (which in no way entails a requirement for a mathematically elegant environment) is, I think, more than sufficient.

    2) In general, mathematically elegant universes are less complex (in the Kolmogorov sense). In other words data representing a mathematically elegant universe can be compressed better than data representing a mathematically inelegant universe. We know that compressibility is a rare property and hence that elegant universes are rare too.

  • http://www.blogger.com/profile/05693985638589020492 Mark

    The ratio of prime numbers to natural numbers is not well-defined. We can say that the limiting value of pi(x)/x is 0 (with pi(x) being the prime counting function), but this is in part an artifact of the way that we order the integers. We could reorder them in a different way such that the asymptotic value of pi(n)/n is 1.

    1) This is a case where we are physically constrained to a finite sample space (what can be stored in a computer) and in which the selection procedure is non-random.

    2) There are infinitely many universes with compressible laws of physics. You'd need to define a probability measure over these, as before, and show it's the correct one.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    Dianelos said:

    “What happens if one moves the problem one step back and considers the natural conditions prior to the origin of our universe, which for all we know might have existed and might have caused our universe to come into existence? Well, given the second law of thermodynamics (which appears to be a necessary law of any mechanistic worldview rather than just a contingent law which applies only to our universe) we see that there are many possible natural conditions that would produce (or would probably produce) an inelegant universe, but only a few possible natural conditions that would produce (or would probably produce) an elegant universe. (For the same reason that there are many possible prior conditions that would cause a disorderly, high entropy, physical state, but only few possible prior conditions that would cause an orderly, low entropy, physical state.) Therefore the ratio of prior conditions that would cause an elegant universe to come into existence is even smaller than the one we had been considering.”

    Here is my interpretation of this argument:

    1. If naturalism is true, then the second law of thermodynamics would be a law of any and every universe.
    2. If the second law of thermodynamics was a law of any and every universe, then there are many possible natural conditions that would probably produce an inelegant universe, but only a few possible natural conditions that would probably produce an elegant universe.
    Therefore:
    3. If naturalism is true, then there are many possible natural conditions that would probably produce an inelegant universe, but only a few possible natural conditions that would probably produce an elegant universe.
    4. If there are many possible natural conditions that would probably produce an inelegant universe, but only a few possible natural conditions that would probably produce an elegant universe, then the ratio of prior conditions that would probably produce an elegant universe to prior conditions that would probably not produce an elegant universe is very small.
    Therefore:
    5. If naturalism is true, then the ratio of prior conditions that would probably produce an elegant universe to prior conditions that would probably not produce an elegant universe is very small.

    1. If A, then B. (premise)
    2. If B, then C. (premise)
    Therefore:
    3. If A, then C. (from 1 & 2)
    4. If C, then D. (premise)
    Therefore:
    5. If A, then D. (from 3 & 4)

    Did I get your reasoning correct?

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Mark:

    You said: “The ratio of prime numbers to natural numbers is not well-defined. We can say that the limiting value of pi(x)/x is 0 (with pi(x) being the prime counting function), but this is in part an artifact of the way that we order the integers. We could reorder them in a different way such that the asymptotic value of pi(n)/n is 1.

    I suppose you mean that if we order the integers in such a way that they start with all the prime numbers then we shall never encounter a non-prime integer and it will seem as if there were only prime integers. But then, what’s the point of this argument? I mean, it’s a fact that prime numbers are rare (only about 5% of the first billion integers are prime) and as one goes up this rarity continuously increases, so that at some point one can count a billion consecutive integers without encountering one prime. Your argument, it seems to me, only shows that there are ways to confuse the truth.

    1) This is a case where we are physically constrained to a finite sample space (what can be stored in a computer) and in which the selection procedure is non-random.

    I don’t see the relevance of the fact that when one simulates the Darwinian process the computer’s memory is finite. Our physical universe as far as biological evolution goes is finite also. As for the second point, when we are capable to simulate the Darwinian evolution of intelligence, the selection procedure can easily be made as random as we wish. The evidence we already have is the result of a pretty random selection procedure, as the real world implementations of the Darwinian algorithm did not in any way take into account the issues we are discussing.

  • http://www.blogger.com/profile/09925591703967774000 Dianelos Georgoudis

    Bradley:

    Close enough, but the argument as you formalized it does not express the idea that if one assumes prior naturalistic conditions to the origin of our universe then the original argument from the elegance of the universe becomes even harder.

    The general idea is this: As our universe has the rare property of being a mathematically elegant Darwinian universe (which is a precondition for successful science to obtain) the whole thing appears designed, which creates an explanatory problem for naturalism. The seriousness of the problem is inversely proportional to the ratio p of mathematically elegant Darwinian universes to mathematically inelegant Darwinian universes.

    If one assumes that our universe’s origin was caused naturalistically by some prior condition, then, as the second law of thermodynamics applies, the average number E of possible prior conditions that will produce a mathematically elegant Darwinian universe is smaller than the average number N of possible prior conditions that will produce a mathematically inelegant Darwinian universe. Let q be the ratio of possible prior conditions that will produce an elegant Darwinian universe to possible prior conditions that will produce an inelegant Darwinian universe. Then q = (p*E/N), which means that q

    So if one assumes that our universe was naturalistically produced by some prior conditions then the appearance of design grows because the ratio q of prior conditions that will produce an elegant universe is even smaller than ratio p of elegant universes.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    Dianelos:

    Thank you for the nice clarification of your thinking about the rarity of elegant universes, as well as for your other thoughts and comments about the appearance of design.

  • http://www.blogger.com/profile/05211466026535549638 Bradley Bowen

    Response to Dianelos:

    Your explanation of the probability calculation seems fairly simple and straightforward, but I must admit I'm still a bit confused about what this means. So, I will attempt to make this even simpler, perhaps this will help you to see and explain where I'm going astray.

    I carry a pair of dice with me, not to gamble with, but for use in generating random numbers (and
    for help in thinking about probability). One die is red, and the other is white. Let the red die represent various possible natural conditions that could have preceded the origin of this universe. Let the white die represent various possible universes that could be produced by the various possible natural conditions.

    The white die is a standard die with six sides, and each side has a specific number of dots (1,2,3,4,5,6). Suppose that there are six different types of universes that could be produced by the various possible prior natural conditions, and suppose that only one of those types of universe would be a mathematically elegant universe (and that all of the universes of that type were mathematically elegant). To make it easy to remember, let's say that the side of the white die with one dot represents this elegant type of universe. So only one out of six possible universes is mathematically elegant (in this example).

    The red die is also a standard die with six sides (1,2,3,4,5,6). Suppose that the various natural conditions that could have existed prior to the origin of the universe can be divided into six types, and suppose that only one of those types of prior natural conditions will produce a mathematically elegant universe (and for simplicity let's suppose that prior natural conditions of that type always produce a mathematically elegant universe). To make it easy to remember, let's say the side of the red die with one dot represents the type of prior natural conditions that will always produce a mathematically elegant universe. So only one out of six possible types of natural conditions will produce a mathematically elegant universe.

    Normally, when I roll this pair of dice, the outcome of the roll of the red die has no influence on the outcome of the roll on the white die. These would normally be independent events. However, based on the above suppositions, the outcome of the white die is causally determined by the outcome of the red die. If the red die comes up 1 (representing the type of prior natural conditions that always produce a mathematically elegant universe), then the outcome on the white die will also be 1 (representing the production of a mathematically elegant universe). Furthermore, if the red die comes up anything other than 1, then (based on the above suppositions) the white die must also come up with something other than 1 (representing the production of a universe which is NOT mathematically elegant).

    In this hypothetical example, there are six different types of natural prior conditions, and only one out of six of these types will produce a mathematically elegant universe. There are also six different types of universe, and only one type of universe is mathematically elegant. Assuming that each of the sides of the red die has an equal chance of being the outcome of a roll, isn't there just a one in six chance of the the outcome of the white die being 1 (representing the production of a mathematically elegant universe)?

    Also, if we know that the outcome on the white die is 1 (representing the production of a mathematically elegant universe), doesn't this imply that the outcome of the red die must have been 1 (representing the occurrance of the only type of natural prior conditions that would produce an elegant universe)?