Fine-Tuning and Life Chauvinism

by Luke Muehlhauser on January 31, 2011 in General Atheism

Part 5 of my series on the fine-tuning argument.

One objection to the fine-tuning argument (FTA) for God’s existence is the charge of “life chauvinism,” which runs like this:

You say the existence of intelligent life is highly improbable, and thus requires an explanation, and the best explanation is God. But the existence of stars is also highly improbable, and so is the existence of iPads, and so is the fact that the resting mass of an electron is 9.11 x 10-31. So why do you think it is intelligent life that needs explanation, but not stars or iPads or the value of the resting mass of an electron?

The point is that any particular configuration of the universe is highly improbable given the variety of possible universes, so the theist needs to show what is special about a universe with intelligent life that needs to be explained.

Some FTA defenders reply that it is intelligent life that needs to be explained because intelligent life has intrinsic value – that is, life would have value even if it was alone in the universe with no agent to value it. But the only evidence presented for this assertion is that it feels like life has intrinsic value.

But now, the whole point of the fine-tuning argument is that it’s supposed to provide “scientific evidence” for God. If it all hinges on our feelings in the end, why not just say it feels like God exists and skip the whole examination of fine-tuning? Defenders of the FTA need a better reply to the charge of “life chauvinism” than this.

Versions of the argument

But first, note that some versions of the fine-tuning argument are not subject to this worry. For example, Robin Collins’ Bayesian formulation of the argument may not be subject to the “life chauvinism” reply, as Bill Craig explains:

Letting “FT” represent the fine-tuning of the universe for intelligent life, “T” represent theism, and “ASU” represent the atheistic single universe hypothesis (i.e., there is a single universe and no God), Collins argues that the fine-tuning is significantly more probable on theism that it is on atheism: Pr (FT/T) >> Pr (FT/ASU). Therefore, the observed fine-tuning confirms the hypothesis of theism.

On this version of the argument, it doesn’t seem that your question is especially pressing. We can calculate the probabilities of other observations as well to see if they similarly confirm theism. Take rainbow planets with fiery rings (X3). Is Pr (X3/T) >> Pr (X3/ASU)? It doesn’t seem like it. There’s no reason to think that Pr (X3/T) is very high or that Pr (X3/ASU) is very low—unless you’re thinking it to be naturally impossible, in which case such a miraculous phenomenon would be evidence of theism —similarly, for X2, singing gas, whatever you mean by that! So it seems to me that on a Bayesian approach, one can plug in any sort of observation we have and ask if it’s more probable on theism than on atheism, and if it is, then it confirms theism.

Still, many versions of the fine-tuning argument need a good reply to the charge of life chauvinism.

One possible response comes from John Leslie’s discussion of “tidy explanations” in his book Universes, to which we shall turn next.

Previous post:

Next post:

{ 155 comments… read them below or add one }

rosyatrandom January 31, 2011 at 4:48 am

Of course, if you don’t give the S in ASU much credence then much of the ‘problem’ goes right away…

  (Quote)

Forrest January 31, 2011 at 4:50 am

Luke,

I thought I read somewhere that you favored MWI of quantum mechanics. If that interpretation is correct, shouldn’t the fine tuning be effectively explained?

  (Quote)

Scott January 31, 2011 at 5:39 am

For WLC’s calculations up there, where the hell does he get the probabilities? Is there some fancy math/science I’m missing, or is he pulling numbers out the air?

  (Quote)

Patrick January 31, 2011 at 6:39 am

Why does Collins argue that fine tuning for life is more likely under theism than under atheism? That seems non obvious to me.

I am familiar with classical fine tuning arguments that claim that life is more likely under theism. But claiming that life is more likely under theism is not the same thing as claiming that fine tuning for life is more likely under theism.

Why not a universe in which life exists even though fine tuning doesn’t exist and even though the physical constants of the universe don’t permit life, and where the explanation for this is THEISTIC OMNIPOTENCE SO SUCK IT SCIENTISTS! Why is that probability low?

  (Quote)

PDH January 31, 2011 at 7:18 am

Forrest wrote,

Luke,I thought I read somewhere that you favored MWI of quantum mechanics. If that interpretation is correct, shouldn’t the fine tuning be effectively explained?  

There are different kinds of multiverse and the many worlds of MWI are the wrong kind. You need something like Eternal Inflation.

But generally speaking you’re right. Because if fine tuning is evidence for theism it’s also evidence for eternal inflation, which is the simpler hypothesis (note: I said simpler hypothesis, not simpler ontology. There’s no rule I’m aware of that says more complex universes are less likely to exist than simpler ones but there are rules in probability theory, like the conjunction rule, that place restraints on what it is rational to believe. Eternal inflation is a heck of a lot simpler than theism).

Even if the fine tuning argument works, it’s not impressive. Anyone can do this. Consider: what is the probability that the universe would be spaghetti permitting? It is even lower than the probability that it would be life permitting (since, as far as we know, you need intelligent lifeforms to make spaghetti). Therefore, the universe must have been designed for spaghetti. Plainly, that is much more probable under the hypothesis that the flying spaghetti monster exists than it is under theism. That is what I would expect to find if there exists a monster that considers spaghetti to be valuable and for some reason has to create an entire universe so as to make spaghetti instead of just making spaghetti.

This is the relevance of parody arguments. They demonstrate how trivially easy it is to come up with a hypothesis under which certain seemingly improbable events would appear more likely.

  (Quote)

Yair January 31, 2011 at 7:25 am

Life chauvinism is still a problem for the Bayesian argument in that it isn’t at all clear that consideration of other tunings-for doesn’t lead one to the conclusion that the values are fine-tuned for another purposes, not life, and the argument that God would want to fine-tune for these is rather slim. There is Smolin’s idea about fine-tuning for black-holes, for example, and there is an argument that the cosmological constant doesn’t have the value that maximizes life…

Of course, I don’t believe a non-ad-hoc explanation of why P(FT|T) is high or higher than P(FT|SAU) can be provided. I don’t believe these probabilities can even be meaningfully estimated, and Ignorance is not handled well by Bayesian analysis (especially when infinities are involved…).

Yair

  (Quote)

rosyatrandom January 31, 2011 at 7:32 am

I think it’s clear that, taken in isolation as a single cosmos, the universe is incredibly unlikely. Both in its evolution and parameters, there seems to be no reason to pick what we observe over all the other possibilities.

To me, it’s like saying only a single integer, with 10^LOTS of digits, is the only one that really exists. Without context, it just seems bizarre and ridiculous. And it begs the question: why does it exist? And that leads to existence being _difficult_ in some sense, requiring special circumstances and some behind the scenes justification. In short, God. But of course God doesn’t really abnegate those same problems except through mysterious ineffability.

The only solution, as far as I’m concerned, is to make existence _easy_. That arbitrary number exists? Fine. So do all numbers. So does all of the mathematics underlying the concept of number, and all the mathematics built from it. This arbitrary cosmos exists? Fine. So do all the other cosmoses, and all the other universes, and every possible permutation of the metaphysical structure.

As Tegmark noted, bigger and more general and encompassing systems are simpler and easier to swallow, metaphysically speaking. I find this leads straight to an Everything/Nothing equivalence that renders the entire debate here unnecessary.

  (Quote)

Steven R. January 31, 2011 at 8:49 am

For WLC’s calculations up there, where the hell does he get the probabilities?Is there some fancy math/science I’m missing, or is he pulling numbers out the air?  

That’s exactly what I wondered when I read Craig’s response last time. He’s merely saying stuff but providing no real calculations or anything.

I think that the argument from a Bayesian statistic is actually rather weak evidence for Theism (and yes, I have a poor understanding of Bayesian statistics so please tell me if there’s something wrong with my objections). This is because it posits a being that can create and do anything, therefore, the probability something happened at all under the existence of such a being is much greater than without it. But that’s the thing, it’s really not very convincing because we grant this being the ability to do anything and seems like a post-hoc observation.

Consider, for example an archeologist finds a temple in South America. We have a poor understanding of South American building techniques for that particular tribe. Assuming hyper-intellectual architectural aliens who love to build Temples very much like the one in South America and had the means to travel to the Earth and the materials to build such a thing, indeed the probability that it is the result of extraterrestrial travelers seems much greater than this tribe thinking up the plans and using old-fashioned building techniques. But is this a good enough reason to justify such a hypothesis? Indeed, I’m more inclined to say that those E.T. archeologists and much more probable than something that’s magically granted the powers to “fine-tune” the universe and which acts like a disembodied mind.

Not only that, but because Theism posits a personal agent, we should be able to explain this agent’s actions through goals. What goals does this Theistic God have? Clearly not to maximize life or even to make life more pleasurable for intelligent beings. And, if this God was fine-tuning for life and he had all the powers to change everything to fit his goal, why would he make life so hard to come by? Indeed, once we assign the goals that the F-T Argument implicitly assigns to God, we realize that the universe is hardly up to what that God would want. We would need to make God “weaker” (as in, he could only Fine-Tune the universe within some limits and it just so happened that one such limit allowed life on Earth to exist) or assign him goals like “do a half-assed job of fine-tuning life and for whatever reason create a sun that will expand and explode, destroying the life he fine-tuned in the process” to explain what we have. I wonder how many people who use the F-T argument if it required such a goal from the God it “proves.”

Getting a little ahead of this, I read Craig’s explanation of Leslie’s “tidy explanations” and can’t help but note that the examples Leslie provides of requiring an explanation involve interaction with personal beings. But if the F-T Argument is trying to prove the existence of one such being, and a “tidy explanation” makes us presuppose that we are interacting with a personal agent, it seems that now the F-T Argument, under life chauvinism, is now circular.

To clarify that if it wasn’t clear, take the example of the merchant at the bazaar. IF you know that the merchant is an intelligent being trying to maximize his profit, then you very well have reason to question where he places his finger. On the other hand, if we merely know there is a thumb but not a personal agent controlling a thumb, we have no reason whatsoever to think that the thumb is covering the moth hole for any particular reason, let alone say “hey, there’s an intelligent being that is connected to this thumb that defies all of our past experiences that is trying to rip me off!”

  (Quote)

Steven R. January 31, 2011 at 8:52 am

Indeed, I’m more inclined to say that those E.T. archeologists and much more probable than something that’s magically

Gr…E.T. architects, not archeologists.

  (Quote)

Kieran M January 31, 2011 at 9:11 am

To be fair, Craig’s argument is fine here, but does use a bit of a sleight of hand. If we compare mr A, who believes that a God who wanted humans to come about exists, then P(FT|A)=1, as God needs to fine tune the universe according to current evidence. While if we consider mr B, who believes that there is no God, and we just happen to have been created we have P(FT|B)=some small number based on various constants needing to be their current value. Of course what we want in both cases is the P(A|FT) and the P(B|FT), and there are lot of arguments to be made that our prior for A is a lot smaller than our prior for B. What prior we assign to B i’m not sure here… Bear in mind that Mr B could believe in the many universe hypothesis, which boosts the probability.

  (Quote)

Patrick January 31, 2011 at 9:25 am

“If we compare mr A, who believes that a God who wanted humans to come about exists, then P(FT|A)=1, as God needs to fine tune the universe according to current evidence.”

This seems false. Fine tuning is needed if the only way life can emerge is for physical laws to permit the formation of planets, complex chemistry, etc, etc, etc. If you posit an omnipotent superbeing this no longer seems true. The universe could, for example, have been formed in seven days by words moving across water, consist of an ocean below and an ocean above with land in the middle, and have a total scale smaller than Australia. Physical laws would not permit this to happen, and this universe would not have fine tuning (or at least, would not require many aspects of what is claimed to be fine tuning of physical constants). But omnipotence could surely pull it off.

  (Quote)

Brian_G January 31, 2011 at 9:46 am

I don’t think Collin’s version makes the life chauvinism go away, but it does specify what is needed to solve the problem. Since the argument claims:

Pr (FT/T) >> Pr (FT/ASU)

It’s not enough to show that fine tuning is unlikely under atheism. One also must show what it’s not nearly so improbable under theism. One still needs a reason for why God would choose to create a life permitting universe. If every possible combination of universes was just as likely given God, then the argument fails. (eg if God is just as likely to create a universe made entirely of black holes or entirely of helium.)

What isn’t needed is to show that God would only create a life permitting universe. All we need to show is that a life permitting universe is much more likely under theism then atheism. So if we consider the entire set of universes that God might create, it needs to be a smaller set of universes then we would expect to exist under atheism. We also need some reason to expect that a life permitting universe would be in the set of universes God might have created.
Can we do those two things without merely appealing to feelings? I think we can. We’d expect that the set of universes that God would create is smaller then the set of universes that might exist under atheism. Everything we know about intelligent agents is that they select choices from a much more narrow set then the entire range of possibilities. If we believed that the folks at AMD were just as likely to arrange their transistors in any of the possible configurations, we wouldn’t buy their processors. There chips would be indistinguishable from chance.

Is there any non-arbitrary reason an intelligent agent might want a life-permitting universe in his set of live options? Life permitting universes allow for the existence of intelligent agents. It’s natural for an intelligent agent to want to create other intelligent agents. We can find evidence of this in our observation of human intelligence. Humans are interested in AI. If humans could create intelligence, it’s likely that they would do so. Since God is proposed as an intelligent agent, one reasonable sort of thing he might want to create is other intelligent agents. I think one might object that it’s inappropriate to use observations of human agency to infer what God would do. Notice I’m not claiming that we can infer from intelligent agency that God would create a life-permitting universe. All I’m claiming is that a life permitting universe is among the universes that would be in the set of universes an intelligent agent such as God would create, and that this set is smaller then the total set of possible universes we might expect under atheism. So I think it’s entirely appropriate to use what we can observe about intelligent agency to help understand the set of universes God might create. We don’t have to determine which universe would be God’s favorite, we just need to know what’s on the table as live options.

  (Quote)

Patrick January 31, 2011 at 9:58 am

Argh, I’m going to keep worrying at this all day now, because I keep spot checking to see if anyone’s answered me, and instead I find people doing the same thing I asked about.

Brain_G now continues the trend of treating Pr (FTU/T) as equivalent to Pr (LPU/T) without explaining why.

  (Quote)

Brian_G January 31, 2011 at 10:09 am

Patrick wrote:
“Why not a universe in which life exists even though fine tuning doesn’t exist and even though the physical constants of the universe don’t permit life, and where the explanation for this is THEISTIC OMNIPOTENCE SO SUCK IT SCIENTISTS! Why is that probability low?”

I think the problem here is what would this universe look like from our perspective? Suppose that God created a world where humans existed and needed oxygen to survive, but there was no oxygen. Miraculously he allowed them to survive anyway. How would these people ever discover that they needed oxygen, if such a miracle is preformed all the time? Wouldn’t it just look like the laws permitted humans to exist without oxygen?

  (Quote)

rosyatrandom January 31, 2011 at 10:14 am

@Brian

A constant miracle would either look like another finely-tuned universe, or there’d be some part of it that could never be explained by the theories that work for everything else. In this case, it would be the physical lack of the theoretically-possible substance ‘oxygen’, which itself might be inexplicable. Furthermore, all the chemistry that incorporates oxygen would still go on as if it was there. The only way you might be able to tell that it wasn’t there, maybe, would be the fact that it had no mass. Unless the mass was also faked, in which case there would be no difference between ‘real oxygen’ and ‘miracle oxygenless pseudo-oxygen’.

  (Quote)

Steven R. January 31, 2011 at 10:57 am

I don’t think Collin’s version makes the life chauvinism go away, but it does specify what is needed to solve the problem. Since the argument claims:Pr (FT/T) >> Pr (FT/ASU)It’s not enough to show that fine tuning is unlikely under atheism. One also must show what it’s not nearly so improbable under theism. One still needs a reason for why God would choose to create a life permitting universe. If every possible combination of universes was just as likely given God, then the argument fails. (eg if God is just as likely to create a universe made entirely of black holes or entirely of helium.)What isn’t needed is to show that God would only create a life permitting universe. All we need to show is that a life permitting universe is much more likely under theism then atheism. So if we consider the entire set of universes that God might create, it needs to be a smaller set of universes then we would expect to exist under atheism. We also need some reason to expect that a life permitting universe would be in the set of universes God might have created.Can we do those two things without merely appealing to feelings? I think we can. We’d expect that the set of universes that God would create is smaller then the set of universes that might exist under atheism. Everything we know about intelligent agents is that they select choices from a much more narrow set then the entire range of possibilities. If we believed that the folks at AMD were just as likely to arrange their transistors in any of the possible configurations, we wouldn’t buy their processors. There chips would be indistinguishable from chance. Is there any non-arbitrary reason an intelligent agent might want a life-permitting universe in his set of live options? Life permitting universes allow for the existence of intelligent agents. It’s natural for an intelligent agent to want to create other intelligent agents. We can find evidence of this in our observation of human intelligence. Humans are interested in AI. If humans could create intelligence, it’s likely that they would do so. Since God is proposed as an intelligent agent, one reasonable sort of thing he might want to create is other intelligent agents. I think one might object that it’s inappropriate to use observations of human agency to infer what God would do. Notice I’m not claiming that we can infer from intelligent agency that God would create a life-permitting universe. All I’m claiming is that a life permitting universe is among the universes that would be in the set of universes an intelligent agent such as God would create, and that this set is smaller then the total set of possible universes we might expect under atheism. So I think it’s entirely appropriate to use what we can observe about intelligent agency to help understand the set of universes God might create. We don’t have to determine which universe would be God’s favorite, we just need to know what’s on the table as live options.  (Quote)

Well, that’s an interesting argument but I don’t think it works.

1. You haven’t shown that there are less universes that God would want to create as compared to those under an “atheist” universe. You merely say we may have reason to believe God would want to create other intelligent beings, but that doesn’t really tell us anything or eliminate other probabilities or universes. You yourself said it was inconclusive.
2. I question the validity of saying “everything we know about intelligent agents tells us X” and using it to surmise what God would do. Why? Because you’re more or less saying God is a disembodied mind. Why is God exempt from “Everything we know about intelligent agents” up until it becomes convenient to take into consideration “everything we know about intelligent agents”?
3. God could be capricious. He may not particularly care what he creates. He may have started out doing one thing and then halfway through decided to do something else. God may like his work to be indistinguishable from chance. God may not care about us but the AI robots we make or the iPads we create. God may hate intelligent life and created us only to watch us suffer and praise him as a cruel joke. At any rate, we can only reduce the number of possibilities once we assign specific goals to God. This is problematic in that we have no idea what God would actually want, as there are virtually infinite number of goals and aims we cannot even begin to conceive of (oh, I love the irony in how God’s mysterious ways can be used against “Him”).
4. Humans are interested in AI for a number of reasons but none that I would think apply to God. We are interested in knowing how our brains work and if we’re capable of replicating what we have, a God would know how his mind works and know that it is within his power to replicate it. We look to AI to alleviate us from manual labor and see if it can expand our knowledge of the universe. I don’t think many people believe God created us to bypass manual labor and obviously if God is omniscient it wouldn’t need other, lesser intelligent beings to understand the universe.
5. Why would a God that values intelligent life create things that restrict its growth? Moreover, why would he let chance, mutations, the process of evolution, etc. threaten the universe he/she/it supposedly fine-tuned for the purpose of creating intelligent life? And why would there be restrictions on how much this intelligent life can understand? Why not give us more intelligence and understanding?
6. It seems to me that the purpose of creating intelligent life would be to interact and converse with it, and morally speaking, even protect and preserve it. Why doesn’t God protect us? Provide us with leisure and free us of manual labor, if what he sought to maximize was intelligence, and labor and other factors restrict the amount of intelligent people in the world?

I can go on and on, but the simple problem remains: once you assign to God a specific goal, he usually fails, unless we tailor the goal to fit our observations. This is what impresses me about the F-T Argument.

  (Quote)

Patrick January 31, 2011 at 11:00 am

I think the problem here is what would this universe look like from our perspective?Suppose that God created a world where humans existed and needed oxygen to survive, but there was no oxygen.Miraculously he allowed them to survive anyway.How would these people ever discover that they needed oxygen, if such a miracle is preformed all the time?Wouldn’t it just look like the laws permitted humans to exist without oxygen?  

If P(FTU/T) is inscrutable under theism, and your response is to change the issue to whether we’d perceive that we had FTU given theism, then your point of analysis in the Bayesian equation should be the perception of a finely tuned universe.

But the probability that we perceive a finely tuned universe given atheism is 1.

That’s usually a stupid answer to the fine tuning argument, but its perfectly adequate if you want to argue that under theism fine tuning may or may not exist, but we’d think it did.

Also its not incredibly hard to imagine a life permitting universe that isn’t fine tuned but which works because shut up omnipotence. A disc resting on four elephants resting on a giant turtle, for example. Or a universe where all the planets are pushed around by angels. Or even a universe where the gravitational constant, relative to the size of the earth, is appropriate for keeping humans from floating away into space… but where the total mass of the universe is drastically smaller than that needed to accomplish a big bang , but where a big bang happened anyway. Or a universe where god signs everyone’s appendix with a “sorry, I forgot to take this out” note. Or a universe where every single solar system consists of exactly one planet, one sun, and one moon, all packed in so close that their orbits bring them close enough that they can launch spacecraft to each other every two years. Or a universe where knives and bullets work unless you try to use them on someone who’s death is morally unjustified, in which case the knives do not cut and the bullets do not fire.

The number of possible universes that could be sustained by omnipotence that makes up laws of physics and abrogates them at will is limited only by our creativity.

  (Quote)

ayer January 31, 2011 at 11:35 am

On a related note, interesting discussion at the physics blog “Not Even Wrong” on how the multiverse hypothesis (the String Theory version, not necessarily the QM version, which is distinct) was motivated by the desire to explain away the appearance of fine-tuning, particularly of the cosmological constant, due to its theistic implications:

http://www.math.columbia.edu/~woit/wordpress/?p=3419&cpage=1#comment-78544

http://www.math.columbia.edu/~woit/wordpress/?p=3419&cpage=1#comment-78694

The idea that the universe (absent the assumption of the nonfalsifiable multiverse hypothesis) is fine-tuned for life appears to be somewhat of a consensus position in the community of cosmologists.

  (Quote)

Adito January 31, 2011 at 12:11 pm

The probability of fine tuning on theism still depends on assuming some specific nature for God whether or not you consider the charge of life chauvinism to be relevant. At the end of the day we still have absolutely no reason to assume that the set of Gods desires would cause Him to make this world rather than some other world. We replace the problem of a finely tuned universe with the problem of a finely tuned God.

  (Quote)

PDH January 31, 2011 at 1:14 pm

ayer wrote,

On a related note, interesting discussion at the physics blog “Not Even Wrong” on how the multiverse hypothesis (the String Theory version, not necessarily the QM version, which is distinct) was motivated by the desire to explain away the appearance of fine-tuning, particularly of the cosmological constant, due to its theistic implications:http://www.math.columbia.edu/~woit/wordpress/?p=3419&cpage=1#comment-78544http://www.math.columbia.edu/~woit/wordpress/?p=3419&cpage=1#comment-78694The idea that the universe (absent the assumption of the nonfalsifiable multiverse hypothesis) is fine-tuned for life appears to be somewhat of a consensus position in the community of cosmologists.  

A big part of the reason that string theory (which no-one here is defending) is so lowly regarded is that people have attempted to defend it using various forms of ‘anthropic reasoning’ that is perilously similar to theism. But the reverse is also true. That this makes string theory seem so dubious reflects very poorly on the fine tuning argument! If it was a good argument then the ‘fine tuning’ of the constants could be used as evidence for string theory. In other words, the fact that the fine tuning argument is so horrendous and yet constitutes a big part of the evidence in many defences of string theory is partly why string theory has such a poor status.

Inflation, meanwhile, is backed up by shed-loads of evidence. In this paper Alan Guth identifies the size of the universe, the Hubble expansion, the homogeneity and isotropy of the universe, the (‘particularly impressive’) flatness problem, the apparent absence of magnetic monopoles and the anisotropy of the cosmic background radiation as lines of evidence for inflation. Note that nowhere is there any mention of the WAP, though of course if fine-tuning is evidence for theism it is also evidence for inflation. The reason for this, as mentioned, is that most physicists think the fine-tuning argument is crap.

Furthermore, inflation was hypothesised (many decades ago) to explain the origins of the universe, not the fine-tuning of its constants. That is just a bonus. Not that there is anything wrong with creating a hypothesis to account for surprising data, as you seem to imply. That’s what an explanation is for, after all.

We need to carefully distinguish between the different theories at play here. I’m sure you didn’t mean to but it is extremely misleading to imply that they all have the same status as string theory.

  (Quote)

PDH January 31, 2011 at 1:15 pm

Link to the paper I mentioned above: http://arxiv.org/PS_cache/hep-th/pdf/0702/0702178v1.pdf.

  (Quote)

Luke Muehlhauser January 31, 2011 at 1:18 pm

Forrest,

According to most versions of the argument, yes. But Luke Barnes, for example, resists this. Check my interview with him.

  (Quote)

Luke Muehlhauser January 31, 2011 at 1:21 pm

ayer,

The best link I can reply with is The Dilemma: Bayes or Science?

  (Quote)

Henry January 31, 2011 at 7:21 pm

The lottery winner might say, “You know, those lottery balls had to be finely tuned to drop as they did. A slight variation in any one of the precise bounces made by each ball would have drastically changed the outcome and my millionaire life never would have come into existence.”

  (Quote)

Steven R. January 31, 2011 at 7:21 pm

ayer,The best link I can reply with is The Dilemma: Bayes or Science?  

Wow, the more I read Eliezer the more I like every bit he says. Haha, what an excellent little article on all this. Luke, what’s your stance on this Bayes-Science gap?

  (Quote)

Luke Muehlhauser January 31, 2011 at 7:24 pm

Steven,

I advocate bringing scientific method in line with Bayes and still calling it ‘science.’ If you’re doing science that disagrees with Bayes, then you’re just mathematically wrong. But anyway, it’s already happening. That’s what the “Bayesian revolution of the sciences” is, and I don’t think it will be stopped.

  (Quote)

Kiwi Dave January 31, 2011 at 7:32 pm

“Everything we know about intelligent agents is that they select choices from a much more narrow set then the entire range of possibilities.”

Every intelligent agent we actually know requires a physical base for storing and processing information. Does this fact make an immaterial intelligent agent less likely, or is it not relevant to the fine tuning argument?

Intelligent agents generally prefer longevity and variety to brevity and monotony. Cockroaches have been around for about 300 million years and exist in about 4500 species. Does this mean our universe is fine-tuned for cockroaches?

  (Quote)

cl January 31, 2011 at 7:33 pm

Duh, all this stuff popped out of nowhere just perfect, like, you know, the luckiest roll of the dice ever! Don’t you stupid Christians realize this yet? [/SARCASM]

On a more serious note, Palonen concludes via Bayes that multiverse hypotheses are not adequate explanations for fine-tuning.

  (Quote)

Steven R. January 31, 2011 at 8:18 pm

Steven,I advocate bringing scientific method in line with Bayes and still calling it ‘science.’ If you’re doing science that disagrees with Bayes, then you’re just mathematically wrong. But anyway, it’s already happening. That’s what the “Bayesian revolution of the sciences” is, and I don’t think it will be stopped.  

Tsk, do you have any links/posts that can clarify how you can be “scientifically correct” and “mathematically wrong” and what that implies? I realize that the question may be pretty hard to answer, so any few starting places for Bayes that would be able to answer my question will be appreciated. I already have the “An Intuitive Explanation of Bayes” link.

PS:
If you’ll link me to the simplest explanation of the Bayesian Theorems and stuff it would be great since I’m VERY new to this sort of stuff. Actually just became aware of it thanks to this blog and it’s comments.

  (Quote)

ayer January 31, 2011 at 8:53 pm

Furthermore, inflation was hypothesised (many decades ago) to explain the origins of the universe, not the fine-tuning of its constants. That is just a bonus.

I’m afraid that’s not correct; Guth was motivated by the fine-tuning of “flatness” so as to be life-permitting; see pages 191-193 of “The Limits of Scientific Truth”, publicly available at Google Books here:
http://tinyurl.com/4ag5o58

  (Quote)

Brian_G January 31, 2011 at 9:00 pm

@Patrick
“But the probability that we perceive a finely tuned universe given atheism is 1.”

Why think that unless something already seems special about life? Why wouldn’t atheism predict that life could exist in just about any sort of universe? Under theism or atheism, life permitting conditions must obtain. That’s evident because we’re here. However, the fine-tuning argument is based on the fact that the conditions are astonishingly narrow. If we we’re the sorts of things that could exist under a wide range of conditions, then that would seem to support the conclusion that the universe wasn’t design.
There’s another way the universe could appear to be not fine-tuned under atheism. There are 4 fundamental forces: gravity, electromagnet, weak force, strong force. All these forces play a role in the fine-tuning examples commonly cited. Suppose we observed a world where there were 400 fundamental forces, four of them are necessary for life, and the remaining 396 forces had nothing to do with life at all. This universe would of course be life-permitting, but given that most of the laws have nothing to do with us, I think this would cast some doubt into the alleged design of the four finely tuned laws.

  (Quote)

ayer January 31, 2011 at 9:15 pm

The best link I can reply with is The Dilemma: Bayes or Science?

That link discusses only QM, not the String Theory multiverse. It is clear that the String Theory multiverse is a reaction to the theistic implications of fine-tuning; not so with QM, Everett’s work pre-dates the evidence for fine-tuning.

I advocate bringing scientific method in line with Bayes and still calling it ‘science.’

This would seem to make science the handmaiden of philosophy, a position I am sympathetic with, but somewhat surprising, since most atheists appear to embrace science (traditionally defined–i.e., experiment-based, falsifiable, etc.) as the supreme way of knowing (e.g., Dawkins most preeminently).

  (Quote)

Brian_G January 31, 2011 at 9:29 pm

@Steven R.

I agree that one could propose a hypothesis where God is indistinguishable from chance. That’s not the hypothesis that proponents of the fine-tuning argument propose. Suppose we compare this hypothesis (G2) with the hypothesis of God acting as designers typically act (G1). In this case G2 would be in the same position as atheism and thus:
P(FT|G1) >> P(FT|G2)

Secondly, as you point out there are many universes God might have created. Some of these may be better then this one. Perhaps he did design other universes. But whether he did or didn’t, isn’t at issue here. What we need is some reason to think this universe is on the table of possible universes a designer might create. We’re not seeking infallible certainty here, all we need is enough reason to out-weight the chance hypothesis. Given the chance of a fine-tuned universe is astonishingly low, we don’t need much to tilt the scale.

  (Quote)

Shane Steinhauser January 31, 2011 at 9:56 pm

Of course, if you don’t give the S in ASU much credence then much of the ‘problem’ goes right away…  (Quote)

Yep Craig seems to be ignoring the possibility of the multiverse.

  (Quote)

Larkus February 1, 2011 at 3:30 am

1) Suppose no gods exist. What is the probability that the relevant constants of the universe happen to be – out of all the possible constellations – such that they permit for the existence of life? More or less than under 2)?

2) Suppose a god exists. Without making any additional assumptions about the identity resp. the intentions of this god, what is the probability that this god set the relevant constants of the universe – out of all the possible constellations – such that they permit for the existence of life. More or less than under 1)?

3) Suppose a god exists. If you make additional assumptions about the identity resp. the intentions of this god, what is the probability that – out of all the possible identities resp. intentions – this god has the identity resp. the intentions, that you assume he has?

  (Quote)

rosyatrandom February 1, 2011 at 3:51 am

Can I get peoples’ thoughts on the following?

We know at least one universe exists, with what seems to be an extremely contingent history and parameters that also seem to be highly arbitrary.

Given then, that this exists, what existential obstacle do other universes (those with different histories/parameters) face?

We don’t know what existence _is_. Most people’s ideas of it seem to come from some incorrect notion of solidity, but really it just seems to amount to qualia (sensory, emotional, knowledge and logic) we can validate by corroboration with others. It’s information but we don’t know what the metaphysical substrate is. And every time you try to embed the information in some kind of structure, you snare yourself in the same ontological problem – what is this structure? What is it made of? What makes it real?

So, to me, any attempt to make anything other than information itself the basic foundation of reality and existence just leads to problems.

And so… what makes our universe different from the other conceivable ones? What possible property can it possess that makes it ‘more’ real, and what prevents us just retrofitting these other universes with that property? I ask this, in all honesty, because I simply cannot grasp what kind of thinking and logic could construct an alternative.

  (Quote)

Patrick February 1, 2011 at 5:54 am

@Patrick
“But the probability that we perceive a finely tuned universe given atheism is 1.”Why think that unless something already seems special about life?Why wouldn’t atheism predict that life could exist in just about any sort of universe?Under theism or atheism, life permitting conditions must obtain.That’s evident because we’re here.However, the fine-tuning argument is based on the fact that the conditions are astonishingly narrow. If we we’re the sorts of things that could exist under a wide range of conditions, then that would seem to support the conclusion that the universe wasn’t design.
There’s another way the universe could appear to be not fine-tuned under atheism.There are 4 fundamental forces:gravity, electromagnet, weak force, strong force. All these forces play a role in the fine-tuning examples commonly cited.Suppose we observed a world where there were 400 fundamental forces, four of them are necessary for life, and the remaining 396 forces had nothing to do with life at all.This universe would of course be life-permitting, but given that most of the laws have nothing to do with us, I think this would cast some doubt into the alleged design of the four finely tuned laws.  

Either I’m not understanding you, or you’re not understanding me.

I’m so so at Bayesian reasoning. I can do the math. It sometimes feels intuitive. Sometimes it doesn’t.

So I tend to latch on to the parts I understand like a wolverine. See, eg, my contempt for Plantinga. I don’t understand everything about the EAAN, but I do understand that you can’t do a bayesian analysis on the probability of a randomly selected pair of beliefs and desires being adaptational and then draw conclusions about the probability of adaptational belief forming mechanisms. If you wanted to do that, you needed to do bayes on adaptational belief forming mechanism. Full stop, he didn’t, therefore his argument sucks. Nothing else to see until he does, or until he comes up with a way to connect the two.

One of the things I understand is that variables must mean the same thing every time they’re encountered in an equation.

One of the things I notice in theistic efforts at using Bayes to demonstrate a fine tuning argument for god is that in the few rare cases that they’re smart enough to do their calculations based on Pr (FTU), they tend to swap out Pr (FTU) for Pr (LPU) when they want to argue Pr (LPU/T).

But Pr (LPU) isn’t Pr (FTU). Just on the face of it, its not the same thing. They might try to demonstrate that it comes out to the same thing, but that would take an actual demonstration and not a sleight of hand. And intuitively, seeing as how the argument that fine tuning is necessary for life is premised on life developing organically from a universe that originated in a big bang, Pr (LPU) doesn’t equal Pr (FTU) for the simple reason that a deity wouldn’t be limited to developing life organically from cosmological processes that take billions of years of natural, physical action to eventually produce life… and those processes are the thing that is alleged to be finely tuned. In fact, you can go further- there are loads of life permitting universes that seem to be possible under theism but not under atheism, and its not clear that they would all need to be fine tuned since magic could fix the details.

So if Pr (LPU/T) and Pr (FTU/T) are not equal, then any fine tuning argument that treats them as equal fails, full stop. In fact, if you’re interested in a coherent argument where you don’t have to do all the work on behalf of the apologist, if Pr (LPU/T) and Pr (FTU/T) are not demonstrated to be equal then you can just stop reading because they haven’t done their job of making a simple, mathematically proper prima facie case. You can tell them to come back when they’re willing to take things seriously.

That’s where I am on this.

In response to your point that in a theistic universe we might expect to think we had fine tuning even if we didn’t, which I took to be an attempt at justifying equating Pr (LPU/T) and Pr (FTU/T), I think that’s not a good argument: because if that’s the case, then you should be running your Bayesian analysis on Pr (PFTU), or the probability of perceiving a finely tuned universe. And at least according to advocates of fine tuning arguments for god, a finely tuned universe is the only atheistic one that can allow life.

Your most recent response steers things in an interesting direction, but one that I think is ultimately futile. If I understand you correctly, you are arguing that even if Pr (LPU/A) is high, Pr (FTU/A) may be low. But for that to work you have had to create a new definition of FTU, and once you do that, you have to go back to your original work and start over. No changing variables mid stream in a mathematical argument.

  (Quote)

PDH February 1, 2011 at 8:10 am

I’m afraid that’s not correct; Guth was motivated by the fine-tuning of “flatness” so as to be life-permitting; see pages 191-193 of “The Limits of Scientific Truth”, publicly available at Google Books here:
http://tinyurl.com/4ag5o58

Guth was interested in explaining a number of things, not just flatness. The absence of magnetic monopoles, for instance. It is about clearing up the niggles of the Big Bang Theory.

But, in any case, my ancillary point was that coming up with explanations for surprising data is the whole point of explaining things. We didn’t always know why the moon didn’t fall out of the sky and crush everyone. It was the sort of thing that gods took care of. Maybe there’d be a particular god who was in charge of it or a single god in charge of everything. Does that mean that Newtonian gravitation was just a wild excuse to avoid ‘theistic conclusions?’ From a theist who continued to believe in God and even considered his work to be further evidence for God?

Suppose that inflation was created specifically to address the fine-tuning argument. So what? You think this needs to be explained, right? Well, here’s an explanation that is miles better than theism. What’s the problem? Do you expect us to stick with the same scientific knowledge forever? Whatever the state of scientific knowledge was in, say, 1985, has to be set in stone for all time?

This is one of the problems with falsificationism that Yudkowsky is talking about (I think it’s more accurate to say that it’s a problem with falsificationism than that it’s a problem with ‘science’ like he does in that article. The argument is about how science should be interpreted). For instance, I once got into an argument with a guy about dowsing. I eventually got him to admit that it didn’t work and instead of owning up to it he said, ‘well, duh. That’s because dowsing finds the essence of water, not water itself!’ That is playing silly buggers. Moving the goalposts. I can see why there would need to be measures taken to avoid that sort of thing. If people had said, ‘well gravity is just the method that God uses to make the moon orbit the earth,’ that would be moving the goalposts.

So maybe if people think that ‘science’ is some kind of fixed hypothesis it would seem like scientists keep adding crap onto that hypothesis to protect it from falsification. ‘Did we say Newtonian gravitation? We meant General Relativity. Did we say ‘one universe’? Because meant a potentially infinite number’ etc. But, if not carefully defined, this would result in every single scientific explanation being considered ad hoc. Falsificationism arguably privileges things that are discovered first. It seems like MWI came along and made no testable predictions but it makes loads of testable predictions. It just doesn’t make unique testable predictions. But you could say the same thing about Copenhagen. The only difference is that one came along first and that’s no reason to think it’s more probable. If anything, the first theory to be discovered is usually the stupidest. It’s a first guess.

MWI is simpler and more probable than Copenhagen. Inflation is simpler and more probable than theism. It doesn’t matter what order they were developed in or what reasons motivated their development. You can’t say, ‘this is a real problem that needs to be explained’ and then ban people from coming up with explanations. If you succeed in convincing atheists that the fine tuning argument actually works, then you’ll still have a whole heap of superior ‘naturalistic’ explanations waiting and it will be no use complaining that some of them are unscientific, unfalsifiable or that they involve unobservable entities because so does theism, which has many other problems besides.

  (Quote)

Steven R. February 1, 2011 at 8:22 am

@Steven R.I agree that one could propose a hypothesis where God is indistinguishable from chance.That’s not the hypothesis that proponents of the fine-tuning argument propose.Suppose we compare this hypothesis (G2) with the hypothesis of God acting as designers typically act (G1).In this case G2 would be in the same position as atheism and thus:
P(FT|G1) >> P(FT|G2)

That’s irrelevant (and slightly incorrect, since God’s desires would “explain” why the universe is fine-tuned for life, just that the purpose would be to appear as it was made from chance) because my point was that you hadn’t reduced any of God’s possible goals or reasons for creating the universe and it is also conceivable that God would take delight in having his work indistinguishable from chance. Whether or not this puts this God of the F-T Argument in the same perceived “rut” of an atheistic universe is quite besides the point.

Secondly, as you point out there are many universes God might have created.Some of these may be better then this one.Perhaps he did design other universes.But whether he did or didn’t, isn’t at issue here.What we need is some reason to think this universe is on the table of possible universes a designer might create.We’re not seeking infallible certainty here, all we need is enough reason to out-weight the chance hypothesis.Given the chance of a fine-tuned universe is astonishingly low, we don’t need much to tilt the scale.  

Then that version of the F-T Argument is utterly worthless. What are the chances of you getting a particular combination of cards when you shuffle a 52 card deck? It’s some crazy number that way beyond even billions, its EXTREMELY improbable. Yet you have to get some value. And that’s the biggest problem. Sure, the chances of you getting any particular deck are much, much higher if we propose a spirit who controls the cards and for whatever reason wanted you to get that particular deck. But is this “Bayesian Proof” against the chance hypothesis? Please. This sort of argument doesn’t bring any insight whatsoever. You can’t really assign a goal to God conclusively and as somebody else pointed out, “What are the chances that a God that wanted to create a universe like this exists?” It’s an really silly argument in that it would as easily explain a universe full of helium as anything else (yep, with this response we’re back to square one), because the chances of that are also REALLY low, so a designer who loves Helium is oh so much more probable. So what?

  (Quote)

Brian G February 1, 2011 at 9:11 am

@patrick

I agree that a life permitting universe isn’t the same as a fine-tuned universe. When I made my arguments about God’s possible reasons for creating a life permitting universe, I wasn’t trying to ignore the distinction. All I was trying to illustrate is that there are plausible reasons a designer might create a universe that was life permitting. Universes that are fine-tuned for life would include all life permitting universes. However, a universe that’s life permitting need not be fine-tuned. Every since the human race has existed, we’ve known that the universe is life-permitting. We’re here! It’s only been recently that we’ve observed that the laws and constants must be fine-tuned for our existence. To say that the constants of nature are fine-tuned, just means that the life permitting range is very narrow.

I don’t see any good reasons to think that a hundred years ago an atheist would have bet that this is the case. It’s somewhat strange to say that only a fine-tuned universe could be expected under atheism. Surely an atheist could have expected there to be laws of nature, and that those laws would be consistent with our existence, but this isn’t the same as expecting a fine-tuned universe. This would just be expecting a life permitting universe.
If the theory of evolution makes is easier to be an atheist, (and I think it does) it’s not unnatural to conclude that the fine-tuning evidence makes it harder to be an atheist. The reason that evolution seems to have helped the atheist, is that Darwin proposed a simple mechanism which can produce designed-like complexity. Natural selection and random mutations are the sorts of things one might expect in a universe without design. (reproduction is a bit tricker, but we can set this aside for the moment.) Random mutations aren’t surprising, and neither is natural selection. The way Darwin helps atheism is that it makes it easier to get design-like complexity. Now enter the fine-tuning evidence. This evidence suggests that it is terribly difficult to get a universe that can even get evolution going. Given this background, it seems strange the sort of arguments used against fine-tuning. If were going to say ” so what, some universe or other must exist, if it wasn’t fine tuned we wouldn’t be here, their’s nothing special about us”, why not do this with biological complexity? Why propose that the complexity in my body is due to natural selection? All these chemical must have been in some combination or other. What’s so special about life? It seems that both sides of this debate have always recognized that their’s something special about life.

I’ve already mentioned a couple of ways we could observe a non-fine-tuned universe under atheism. Could there be non-fine-tuned universes under theism? I think this might depend on what you mean. Could God create a universe that didn’t look designed? Sure. He could create a universe where the vast majority of the fundamental forces were unnecessary for life. But that would be an untestable hypothesis and wouldn’t help theism. However, I think what you mean is a universe where instead of fine-tuning the laws, God interacted miraculously to create life. Sure God could have done this as well. In this case, I think we would still observe very improbable events necessary for life. This gets us the same thing as fine-tuning, that the conditions for our existence are extremely narrow. I don’t see how this helps the atheist case. The mere fact that God could create other universes, doesn’t count against the design argument for this universe.

  (Quote)

ayer February 1, 2011 at 9:12 am

Suppose that inflation was created specifically to address the fine-tuning argument. So what? You think this needs to be explained, right? Well, here’s an explanation that is miles better than theism.

I’m not sure why you are under the impression that Inflation solves the fine-tuning problem (even though that was at least partly Guth’s motivation, as you say); as of 2008, Discover Magazine still refers to fine-tuning as “the biggest problem in physics,” with the multiverse the only viable alternative to theism. See:

http://discovermagazine.com/2008/dec/10-sciences-alternative-to-an-intelligent-creator

This is one of the problems with falsificationism that Yudkowsky is talking about (I think it’s more accurate to say that it’s a problem with falsificationism than that it’s a problem with ‘science’ like he does in that article. The argument is about how science should be interpreted).

I have no problem with the idea that falsificationism is not the only road to knowledge; on the most important questions of existence, it is useless, and philosophy must take over (this is what adherents of “scientism” like Dawkins seem not to grasp). But when you leave falsification behind, you are no longer doing “science”; you are doing philosophy. Relabeling it “a different interpretation of science” may make you feel better, but you have left science behind.

  (Quote)

rosyatrandom February 1, 2011 at 9:15 am

I know I’m banging a drum here, but what would God create a universe out of? Also, surely God could imagine all possible universes. What is the difference between a universe imagined in the mind of God and one he makes with his Godly Creatifier? If none, then we are back at all universes existing….

  (Quote)

Steven R. February 1, 2011 at 9:20 am

I know I’m banging a drum here, but what would God create a universe out of? Also, surely God could imagine all possible universes. What is the difference between a universe imagined in the mind of God and one he makes with his Godly Creatifier? If none, then we are back at all universes existing….  

I think your first question is much more relevant to the Cosmological Argument, where the Theist goes “you can’t say the universe came from nothing!”…okay, but then how did God create matter or the universe? Oh right, from nothing. Great explanation! It really doesn’t have much to do with the F-T Argument though.

BTW, nice questions you asked on your other post. I can’t even come close to answering them and I doubt you’re going to get a relevant response :P

—-
@ Ayer:

What about simply denying the “problem”? You’ve yet to show that we should be particularly amazed that we happen to live in this universe in a way that requires some special explanation other than chance. To me, it’s still the same as marveling at whatever combination of 52 cards you get after shuffling them. You got a deck that never existed before and is highly improbable that you would get. Big deal.

  (Quote)

rosyatrandom February 1, 2011 at 9:25 am

The trouble with the ideas I subscribe to is that I am so taken by them, and they pull the rug from under so many other questions, that I have a hard time dealing with ‘separate’ philosophical issues on their own terms. I feel like I’m being presented with a bunch of different nuts… but they are all nuts, just the same, and here I am sitting with this handy sledgehammer resting temptingly on my lap.

  (Quote)

Brian G February 1, 2011 at 9:29 am

@Steven R.
Suppose we found some way of observing other universes. Suppose that the more we looked the more dead, lifeless universes we found. Some universes had only helium, some only with black holes, some where the big bang collapsed on itself within the first few seconds. Would such a scenario count against the fine-tuning argument? Sure it would. Atheists would be all over that evidence as proof that fine-tuning fails. They’d claim that this provides good reason to doubt God’s existence. Why? After all, perhaps God likes universe containing only helium or black-holes? Just the fact that this scenario is possible, gives us no reason to think it’s true. It simply isn’t what we’d expect under theism. Just because it could happen under theism, doesn’t me we should judge it equally likely with all other possibilities.

  (Quote)

Adito February 1, 2011 at 9:45 am

Here’s another reason fine tuning doesn’t point towards a God. The usual theist view is dualism which means minds can exist without bodies. So no fine tuned environment is needed for bodies. In other words God could have created a universe with any kind of laws and still managed to populate it with humans and whatever else He valued.

  (Quote)

Steven R. February 1, 2011 at 9:52 am

@Steven R.Suppose we found some way of observing other universes. Suppose that the more we looked the more dead, lifeless universes we found. Some universes had only helium, some only with black holes, some where the big bang collapsed on itself within the first few seconds. Would such a scenario count against the fine-tuning argument? Sure it would. Atheists would be all over that evidence as proof that fine-tuning fails. They’d claim that this provides good reason to doubt God’s existence. Why? After all, perhaps God likes universe containing only helium or black-holes? Just the fact that this scenario is possible, gives us no reason to think it’s true. It simply isn’t what we’d expect under theism. Just because it could happen under theism, doesn’t me we should judge it equally likely with all other possibilities.  (Quote)

Er…it would count as evidence against fine-tuning merely because there are so many universes, the probability of our own existing seems more likely, but again, if my objection tot he F-T Argument is that simply because something is improbable it does not require some sort of spirit to explain it, this isn’t evidence for or against a F-T Argument. Indeed, supposing only our own universe, the one with Helium and the one with blackholes existed, a Theist could say, “Wow, out of all the possible worlds that could exist under an Atheistic universe, only these three exist! Surely the probability of these three is conclusive proof of a God that values life, Helium and black holes!” and then we’d have to deal with this annoying argument.

But you (inadvertantly) point out an important thing: the F-T Argument isn’t really an argument for the probability of things, its one of personal incredulity. There is absolutely nothing that Theism expects, God could as easily like blackholes under theism as meteroids that threaten to obliterate intelligent life. But the theist finds it incredulous that life exists and thus demands an explanation for no discernable reason other than life chauvinism; of course a universe fine-tuned for black holes is as much evidence for a Theistic God as our own, but that isn’t the thrust of the F-T Argument, is it? No, it’s projecting your own desires and using it to guide “what we would expect under theism” because “theism” here doesn’t mean the Theistic God F-T “proves”, but some sort of human-favoring God from which to base religion, and that is more often than not what is implied by this argument. Is it any coincidence that what we’d expect under theism is what we would want if we oursleves had to fine-tune the universe? I’m sure you’re familiar with the quotation that goes, if horses had hands to draw, they’d draw gods that look like horses.

  (Quote)

ayer February 1, 2011 at 9:58 am

What about simply denying the “problem”?

I second Brian G’s response here; as they might say at “Less Wrong,” his reasoning seems “obvious and normal” to the community of cosmologists (i.e., that fine-tuning, absent a multiverse, is more likely under theism than under atheism). Of course, since neither position is falsifiable, you can continue to hold to the chance hypothesis, but it seems clear that Brian G’s reasoning is likely to remain persuasive absent additional countervailing evidence.

  (Quote)

Steven R. February 1, 2011 at 10:06 am

I second Brian G’s response here; as they might say at “Less Wrong,” his reasoning seems “obvious and normal” to the community of cosmologists (i.e., that fine-tuning, absent a multiverse, is more likely under theism than under atheism). Of course, since neither position is falsifiable, you can continue to hold to the chance hypothesis, but it seems clear that Brian G’s reasoning is likely to remain persuasive absent additional countervailing evidence.  (Quote)

Now you’re just making sound like an Appeal to Authority. “Well, I can’t actually prove there is a problem, but hey, these important people seem to think it is, so it must be.” If the F-T Argument can’t even prove there is a problem that needs to be explained then it is a collossal failure, logically speaking. Sure, it may catch-on by simply appealling to emotions, but other than that….

However, I do note that you said “that fine-tuning, absent a multiverse, is more likely under theism than under atheism” so maybe I misinterpreted your comment a bit. If we’re merely saying that fine-tuning is more likely under theism than atheism, then I am inclined to agree (ignoring, for a second, how absurd and highly unlikely it is that a disembodied mind that can control the laws of physics exists), but as I pointed out, it really doesn’t mean much. It’s the same thing as pointing to the 52 card deck and claiming it’s much more likely that you got that particular combination because a spirit or whatever wanted you to get that particular combination. I’m just repeating myself but I’ll say it once more: big deal.

  (Quote)

PDH February 1, 2011 at 10:26 am

ayer wrote,

I’m not sure why you are under the impression that Inflation solves the fine-tuning problem (even though that was at least partly Guth’s motivation, as you say); as of 2008, Discover Magazine still refers to fine-tuning as “the biggest problem in physics,” withthe multiverse the only viable alternative to theism.See:http://discovermagazine.com/2008/dec/10-sciences-alternative-to-an-intelligent-creator

I have no problem with the idea that falsificationism is not the only road to knowledge; on the most important questions of existence, it is useless, and philosophy must take over (this is what adherents of “scientism” like Dawkins seem not to grasp).But when you leave falsification behind, you are no longer doing “science”; you are doing philosophy.Relabeling it “a different interpretation of science” may make you feel better, but you have left science behind.

But part of the reason that people still consider it to be a problem is because of falsificationism’s hold on science. But falsificationism is not the boss of science! The philosophy behind science is hardly set in stone. People now regard logical positivism as a flat out disaster, for example. This was not always the case. Falsificationism didn’t even exist before Popper. I don’t think it’s unreasonable to anticipate a similar revolution.

If you want me to come out and say it, I concede that I side with the Bayesians on this one. If that means I part with science on this issue, so be it. Rationality is broader than science. I can’t perform a scientific experiment and submit my results to a peer-reviewed journal every time I need to find out whether I can safely leave my house without being attacked by an unfalsifiable dragon. I can, however, safely rule that out on the grounds that it’s utterly improbable. I have to bet my entire life on my ability to calculate the probability of being attacked by an unfalsifiable dragon. Science doesn’t settle that issue (though, it does give me successful models of biology and physics etc. that affect my calculations).

And I think Dawkins does grasp this point but isn’t philosophically literate enough to understand the issues. Remember, he doesn’t openly subscribe to scientism (I don’t know anyone who does), this is just a straw-man argument that rationalists are frequently subjected to. There may be reasons for thinking that he is but I think he is arguing for something much broader. For example, he states in the God Delusion that he thinks we still ought to be able to make ‘probability judgements’ about unfalsifiable phenomena, he has a scale of belief to show where people stand on God (much more Bayesian than the usual strong/weak atheist dichotomy), he openly criticises Gould’s separate magisteria idea etc.

I’m all for asking people who promote science as much as Dawkins does to learn more about philosophy (and philosophy of science in particular). We may have some common ground on that issue.

  (Quote)

Patrick February 1, 2011 at 10:35 am

Alright, Brian_g…

“However, a universe that’s life permitting need not be fine-tuned.”

two sentences later

“It’s only been recently that we’ve observed that the laws and constants must be fine-tuned for our existence.”

a few sentences after that

“To say that the constants of nature are fine-tuned, just means that the life permitting range is very narrow.”

The above is very confused. From reading the whole of your post, you are operating on your own definition of fine tuning that is not compatible with other fine tuning arguments. That’s fine, but as it stands, I’m not sure you realize that you’re doing it, and if you are, you haven’t actually elaborated what your argument is. So I’m not sure its worth me trying to discern and discuss. I’m going to stick with fine tuning as its meant by people like Robin Collins until you give me a different definition you wish to discuss.

“I don’t see any good reasons to think that a hundred years ago an atheist would have bet that this is the case. It’s somewhat strange to say that only a fine-tuned universe could be expected under atheism.”

And that’s completely misunderstanding me. I’m not going to explain it again beyond this: I was answering a specific comment you made where you attempted to equate P(FTU/T) with P(LPU/T), and I did so by pointing out that your proposed solution would require changing P(FTU/T) to P(PFTU/T), and propagating that change through the whole bayesian argument, which would render it pointless because P(PFTU/A) under the traditional definition of fine tuning is 1. This may not be the case under your definition of fine tuning. But I don’t know what that is.

  (Quote)

Alex Petrov February 1, 2011 at 11:05 am

I don’t understand how adhering to probability means anything. It doesn’t matter whether or not it’s probable, it matters if 1) it’s possible and 2) there’s evidence for it.

As another person said above, the chances of winning the lottery is significantly tiny. So much so that you can spend your whole life playing it and never win. But people do win, many people multiple times. It doesn’t matter that it’s unlikely what matters is that it is possible.

We currently have a universe which needs no god for explanation of any actual observed and scientifically verified phenomenon. It is to the point that people are arguing probability. It’s inane. Completely inane. Even if we could actually calculate the probabilities and found god to be more likely, does that make god’s existence true? No. It makes it more likely, but we still have no evidence to judge one way or another.

Also, as far as multiple universes go, it’s still just hypothesis. Though, unlike creationists, they do actually plan on testing it. Seeing as how observing these multiple universes is impossible given they would be moving faster than the speed of light away from us, scientists are trying to recreate the conditions of the big-bang to see what happens (I believe this is one possible use for the hadron supercollider).

But even if there is one and only one universe, that doesn’t mean anything special happened in the creation of this universe. We know: 1) this is a possible universe (it exists), 2) a universe has to exist (we have observed this one). Thus there is a chance that, given random circumstances, this universe could be the only universe to exist given there is only one.

But again, this is referring to probability, which is meaningless when considering the origins of the universe. Meaningless because we don’t have enough information to say what is and isn’t probable for the origins of the universe, and as such any argument given has no use to anything in reality. All it’s used for is arguments, I fail to see it create anything useful.

  (Quote)

Troy February 1, 2011 at 7:28 pm

It seems to me that, if we’re confident of the extremely low value of Pr (FT/ASU), it’s an incredibly weak, and so eminently plausible, claim that Pr (FT/T) >> Pr (FT/ASU). If I had never encountered the concept of God before and was just told “there’s an incredibly powerful and intelligent being who can create universes–what kind of universes do you think he’s likely to create?” I’m pretty confident that “ones with life in them” would be fairly high on my list. As I believe Brian noted some time ago, intelligent beings that we know of, had they the capacity to create universes, would be far more likely to create ones with life in them than ones with (insert random type of object here) in them. Other posters have pointed out that there’s all sorts of other possibilities of things God could create, and of course that’s true, but it’s no less true in the case of ordinary intelligent humans who had the capacity to create universes–and yet it seems plain that we’d expect such humans to create a universe with intelligent life before other types of universes, not withstanding that it’s possible “that they just want to create universes with black holes” or “planets with fiery rings” or whatever. I can’t peer into the minds of those humans to discern their desires, intentions, etc., but so what? There’s still enough of an analogy to my prior experience of intelligent beings (and my self-knowledge) to go on. And it seems to me that the same is true of a super-powerful intelligent being who may have created the universe.

In my estimation this is not a weak point at which one can attack the fine-tuning argument. Much more profitable is to argue for a multiverse hypothesis of some sort. (Although I should say that I’ve only skimmed the discussion up to this point so it’s entirely possible that I’ve missed some powerful objections to the reasoning I’ve sketched here.)

  (Quote)

Larkus February 2, 2011 at 4:22 am

@Troy

You are making a hasty generalization.

You base your observations only on human beings. You assume that a “super-powerful intelligent being who may have created the universe” has preferences similar to (what you believe to be) the preferences of human beings (in regard to “what kind of universes […] he’s likely to create”). But how likely is it, that your assumption is true?

  (Quote)

Patrick February 2, 2011 at 4:36 am

“If I had never encountered the concept of God before and was just told “there’s an incredibly powerful and intelligent being who can create universes–what kind of universes do you think he’s likely to create?” I’m pretty confident that “ones with life in them” would be fairly high on my list.”

Bzzt! You fell for the trick too!

“Finely tuned universe” and “universe with life in it” are not equivalent terms.

If the issue is “universe with life in it,” then you have to deal with the life chauvinism argument. You also have to deal with a number of other issues, such as the overwhelming amount of the universe that is apparently hostile to life.

If the issue is “finely tuned universe,” by which is commonly meant a universe with cosmological values and ratios between physical forces such that a big bang could occur, stars could form, planets could form around those stars, complex chemical reactions could occur on at least one of those planets, and life could emerge, then you have to argue the likelihood that this is what an omnipotent superbeing with magical powers would create if his goal was to create a life sustaining universe. Good luck with that. Most human cultures, when speculating on what an omnipotent superbeing with magical powers would do to create a universe, come up with something like “breathe on dirt,” or “find a cosmic cow and work something out,” or “murder another superbeing and then build stuff out of the corpse.”

  (Quote)

ayer February 2, 2011 at 7:49 am

If the issue is “universe with life in it,” then you have to deal with the life chauvinism argument.

It appears to me that’s what Troy just did.

You also have to deal with a number of other issues, such as the overwhelming amount of the universe that is apparently hostile to life.

The size of the universe is irrelevant when the odds against any universe being life-permitting at all are astronomical.

then you have to argue the likelihood that this is what an omnipotent superbeing with magical powers would create if his goal was to create a life sustaining universe.

No, to prevail over atheism, you only have to argue that such a scenario is more likely on theism than on atheism.

  (Quote)

rosyatrandom February 2, 2011 at 7:59 am

No, to prevail over atheism, you only have to argue that such a scenario is more likely on theism than on atheism.

No! If that was the case, as people have been trying to get across, then the best explanation for the exact number of hairs on my head is that a genie who likes that particularly number has been manipulating my development to make it just so. You cannot just remove all the improbability to a new conditional and then ignore it!

  (Quote)

ayer February 2, 2011 at 9:01 am

No! If that was the case, as people have been trying to get across, then the best explanation for the exact number of hairs on my head is that a genie who likes that particularly number has been manipulating my development to make it just so. You cannot just remove all the improbability to a new conditional and then ignore it!  

No, in your example there is a naturalistic explanation with higher likelihood; in the case of fine-tuning, there is not

  (Quote)

rosyatrandom February 2, 2011 at 9:07 am

I do not see the difficulty in having the universe be one corner of a more general system. There is also the possibility that there is some emergent force that encourages internally complex universes. To discard all the easy naturalistic answers in order to use a magical, ineffable one is more wrong than I can express easily and without swearing.

  (Quote)

Steven R. February 2, 2011 at 9:30 am

No, in your example there is a naturalistic explanation with higher likelihood; in the case of fine-tuning, there is not  

The chances of you getting that number of hairs on your head if a genie wanted it: 100%. Chances with naturalism? Who knows, REALLY small if we take into consideration how many genetic mutations, errors, etc. could have happened to alter that number. Guess genies that control the number of hairs in people’s head is a better explanation!

  (Quote)

Brian_G February 2, 2011 at 9:32 am

Alright, Brian_g…

“However, a universe that’s life permitting need not be fine-tuned.”

two sentences later

“It’s only been recently that we’ve observed that the laws and constants must be fine-tuned for our existence.”

In my first sentence I was referring to what could be the case before observing the evidence. In my second sentence I was explaining what we have in fact observed. All I’m saying here is that in our epistemic situation prior to looking at the evidence, we’d have no particular reason to expect the universe to be fine-tuned, unless of course we already thought there was something special about life. Under atheism the most reasonable expectation prior to looking at the evidence would be that the universe is not fine-tuned. In fact, atheist Victor Stenger, has attempted to defend this position.
Sorry for the confusion. When I wrote that post, I debated with myself whether I should add some clarifying remarks to that first sentence. I decided against it because I thought it would make the post unnecessarily confusing, but I seem to have gotten the result I hoped to avoid.
Furthermore, to the best of my understanding, my definition of fine-tuning is the same as Robin Collins.

Your claim that P(PFTU|A) = 1. You’ve provided no evidence for this claim. I can think of a number of ways we would not perceive a fine tuned universe, under atheism:

1) We could not exist. We wouldn’t be here to observe any kind of universe, fine-tuned or otherwise.
2) We could exist, but observe a universe like Victor Stenger defends, where the constants don’t require fine-tuning for us to exist.
3) We could live in a universe where only a small handful of laws and constants are fine-tuned with a much larger number of constants that could take on any value.

In order for there not to be any confusion, by “could” I referring to what is epistemically possible under atheism. I suspect that 2 is metaphysically impossible. However,1 & 3 seem metaphysically possible.

  (Quote)

ayer February 2, 2011 at 10:35 am

if a genie wanted it

That “if” is your problem right there, and where the analogy breaks down in comparison to fine-tuning and theism vs. atheism. Troy noted above the likelihood that an incredibly powerful and intelligent being who could create universes would want to create a life-permitting universe (based on our prior background knowledge of the preferences of intelligent beings we are familiar with), as opposed to, say, a universe with only black holes, or a universe with undifferentiated space. I would agree that detailing any preferences beyond preferring “life-permitting” to “life-prohibiting” is more problematic; but you only need the general preference for life to determine that fine-tuning for life makes theism more likely than atheism.

  (Quote)

Steven R. February 2, 2011 at 10:49 am

That “if” is your problem right there, and where the analogy breaks down in comparison to fine-tuning and theism vs. atheism. Troy noted above the likelihood that an incredibly powerful and intelligent being who could create universes would want to create a life-permitting universe (based on our prior background knowledge of the preferences of intelligent beings we are familiar with), as opposed to, say, a universe with only black holes, or a universe with undifferentiated space. I would agree that detailing any preferences beyond preferring “life-permitting” to “life-prohibiting” is more problematic; but you only need the general preference for life to determine that fine-tuning for life makes theism more likely than atheism.  (Quote)

…So what? We have a genie who loves controlling hair. There. Hell, let’s even say the genie has hair. Because it is an intelligent being with hair, and a lot of intelligent beings with hair like having hair, we can expect the genie to control hair to the extent tat it pleases him. That’s pretty much what Troy did. Make an assumption and as Brain_G noted, the assumption that God favors intelligent life isn’t even conclusive and based on projecting human desires on non-human entitites. Honestly, this argument (the F-T Argument in general) is a trainwreck.

  (Quote)

Patrick February 2, 2011 at 10:59 am

The size of the universe is irrelevant when the odds against any universe being life-permitting at all are astronomical.

I reject the obviousness of your assertion that

1: the likelihood of a magical superbeing would create a universe that is just barely short of 100% short of life if that magical superbeing’s motivation was to create a universe that contains life

is lower than

2: the chance of fine tuning.

If you care to argue otherwise, do so. Please make sure that when you do you use the same sorts of logical inferences about probability that underlie the assertion that the chance of fine tuning is low. You’ll need to begin by coming up with some estimation of the total sample space of universes a magical superbeing might create. That won’t be enough to finish the argument, but that will be a start at repairing the utter, disreputable philosophical malpractice that goes on in the fine tuning argument.

If this can’t be done, then the fine tuning argument doesn’t even deserve attention in polite company. This isn’t a refutation of fine tuning that I’m offering: I’m noting that the bare minimum prima facie burdens of actually explaining and arguing for the contentions implicit in the probabilistic fine tuning argument haven’t been supported.

  (Quote)

PDH February 2, 2011 at 11:07 am

ayer wrote,

No, in your example there is a naturalistic explanation with higher likelihood; in the case of fine-tuning, there is not  

Yes, there is. Eternal inflation.

  (Quote)

PDH February 2, 2011 at 11:12 am

Also, the size of the universe is relevant because when we’re comparing two hypotheses such as inflation and theism, this is one of the many reasons to prefer inflation, which does predict a big universe.

  (Quote)

ayer February 2, 2011 at 11:24 am

the assumption that God favors intelligent life isn’t even conclusive and based on projecting human desires on non-human entitites.

The likelihood that an intelligent designer of the universe would favor life over lifelessness is based on our experience with the preferences of other intelligent designers; the belief that, on atheism, random chance would produce a life-permitting universe is based on—nothing.

  (Quote)

ayer February 2, 2011 at 11:26 am

ayer wrote,
Yes, there is. Eternal inflation.  

I’m sorry, but eternal inflation does not address the fine tuning of all the constants and quantities built-in to the initial conditions of the big bang that are necessary for life to exist (other than, perhaps, “flatness”). Why do you think the multiverse is to go-to alternative theory to theism?

  (Quote)

Steven R. February 2, 2011 at 11:30 am

The likelihood that an intelligent designer of the universe would favor life over lifelessness is based on our experience with the preferences of other intelligent designers; the belief that, on atheism, random chance would produce a life-permitting universe is based on—nothing.  (Quote)

*Sigh* Answer me this then: why would an incredibly intelligent designer create us not-so intelligent human beings with very limited understanding and intellectual capacities? Our experiences with intelligent designers tell us that they try to maximize intelligence, not hinder it. So why would God create us when he could conceivably create much more intelligent beings? And why subject them to the whims of evolutionary mutations and the like? Also, refer to my post where I explain why intelligent designers are interested in intelligent life in the first place–none of the reasons seem to apply to God.

On Atheism, we know that chance can produce life-permitting universes based on our own experiences–we are in one, aren’t we? Or maybe it wasn’t by chance…who knows. But this uncertainty shouldn’t justify God anymore than the complete uncertainty over how many hairs we end up with justifies belief in our genie interested in hair.

  (Quote)

Patrick February 2, 2011 at 11:34 am

Your claim that P(PFTU|A) = 1. You’ve provided no evidence for this claim. I can think of a number of ways we would not perceive a fine tuned universe, under atheism:

1) We could not exist. We wouldn’t be here to observe any kind of universe, fine-tuned or otherwise.
2) We could exist, but observe a universe like Victor Stenger defends, where the constants don’t require fine-tuning for us to exist.
3) We could live in a universe where only a small handful of laws and constants are fine-tuned with a much larger number of constants that could take on any value.

Ok… If this is becoming an argument over who said what when, instead of about the issues being discussed, let me know and I’ll stop responding. Honestly, this isn’t a big issue for me- I’m more interested in the fact that your specific assertion (that fine tuning can be assigned a high probability under theism because even if fine tuning didn’t exist under theism we’d probably think it did, but that somehow this doesn’t require completely scrapping the entire fine tuning argument and starting over with a probabilistic analysis of a conjunction of multiple probability statements designed to add up to the chance we think that fine tuning exists) is plainly illogical.

As for the specific issues you’ve mentioned, 3 is a non starter. Adding 50 new cosmological constants wouldn’t change the probabilities involved in the values of the small handful. 2 is also a non starter because it doesn’t advance the argument you were attempting to defend. I mean, yes, technically, it is a defense of the specific attack I made against fine tuning to respond that the attack may be irrelevant because fine tuning isn’t true, but that’s not going to help the bottom line, so to speak. I’m not sure about 1 because I’m not sure what conclusion you’re arguing for anymore.

My original point still stands, however. Anyone who equates fine tuning with life permitting without an argument that all logically possible life permitting universes (including those that could be created by magic) are finely tuned universes is a chump, and anyone who makes this elision in published papers is either a crap philosopher or mathematician, or a charlatan. The point is non obvious, and needs defending.

  (Quote)

PDH February 2, 2011 at 12:05 pm

I’m sorry, but eternal inflation does not address the fine tuning of all the constants and quantities built-in to the initial conditions of the big bang that are necessary for life to exist (other than, perhaps, “flatness”).Why do you think the multiverse is to go-to alternative theory to theism?  

On a level 2 multiverse the constants can vary from universe to universe, depending on the model used.

The Level II multiverse is far more diverse than the Level I multiverse. The bubbles vary not only in their initial conditions but also in seemingly immutable aspects of nature. The prevailing view in physics today is that the dimensionality of spacetime, the qualities of elementary particles and many of the so-called physical constants are not built into physical laws but are the outcome of processes known as symmetry breaking. For instance, theorists think that the space in our universe once had nine dimensions, all on an equal footing. Early in cosmic history, three of them partook in the cosmic expansion and became the three dimensions we now observe. The other six are now unobservable, either because they have stayed microscopic with a doughnutlike topology or because all matter is confined to a three-dimensional surface (a membrane, or simply “brane”) in the nine-dimensional space. Thus, the original symmetry among the dimensions broke. The quantum fluctuations that drive chaotic inflation could cause different symmetry breaking in different bubbles. Some might become four-dimensional, others could contain only two rather than three generations of quarks, and still others might have a stronger cosmological constant than our universe does.

http://space.mit.edu/home/tegmark/PDF/multiverse_sciam.pdf

As for you second question, I’d say it’s pretty much a foregone conclusion that we live in a multiverse at this point and probably many different kinds of multiverse. Just today I was reading an article on sci-fi website io9 about the Level 1: http://io9.com/#!5749332/the-universe-is-at-least-250-times-bigger-than-it-looks

It’s time to get used to it, I think. A better question, IMO, would be why is everyone so reluctant to do this?

  (Quote)

Troy February 2, 2011 at 2:06 pm

@TroyYou are making a hasty generalization.You base your observations only on human beings. You assume that a “super-powerful intelligent being who may have created the universe” has preferences similar to (what you believe to be) the preferences of human beings (in regard to “what kind of universes […] he’s likely to create”). But how likely is it, that your assumption is true?  

I don’t think I’m assuming any such thing, any more than I’m assuming that other intelligent humans have preferences similar to preferences I’ve observed in the past, or know that I personally have. Rather, I’m assuming that this is sufficiently *probable* that Pr (FT/T) >> Pr (FT/ASU). As far as I can see, if all we know about the deity is that he is conscious, super-intelligent, and super-powerful, this analogy is about all we have to go on. Clearly the analogy is less than when I encounter a human being I’ve never met before and assume him to be like other humans, or even less than were I to encounter an intelligent alien. But it doesn’t need to be very strong at all, since Pr (FT/ASU) is so staggeringly low. It seems to me that the kind of arguments being presented against Pr (FT/T) being substantially higher than that are really arguments, not against fine-tuning in particular, but against induction, analogical reasoning, and predictions about the behavior of other persons in general–i.e., arguments for skepticism.

Incidentally, denying the possibility of any sort of analogical reasoning to discern the intentions of a God would make any theistic argument (except a priori ones, I suppose) impossible, so far as I can see. If a hundred people saw an apparent miracle performed before their very eyes, it wouldn’t matter how low the probability of that event is on naturalism, because its probability given the existence of God would be inscrutable. (Of course, you may see this is a virtue of your position, but it seems to me that it’s just unfairly denying theism any possibility of winning from the outset.)

  (Quote)

Troy February 2, 2011 at 3:07 pm

Bzzt!You fell for the trick too!

“Finely tuned universe” and “universe with life in it” are not equivalent terms.

If the issue is “universe with life in it,” then you have to deal with the life chauvinism argument.You also have to deal with a number of other issues, such as the overwhelming amount of the universe that is apparently hostile to life.

If the issue is “finely tuned universe,” by which is commonly meant a universe with cosmological values and ratios between physical forces such that a big bang could occur, stars could form, planets could form around those stars, complex chemical reactions could occur on at least one of those planets, and life could emerge, then you have to argue the likelihood that this is what an omnipotent superbeing with magical powers would create if his goal was to create a life sustaining universe.Good luck with that.Most human cultures, when speculating on what an omnipotent superbeing with magical powers would do to create a universe, come up with something like “breathe on dirt,” or “find a cosmic cow and work something out,” or “murder another superbeing and then build stuff out of the corpse.”  

Looking at the way I’ve formulated the argument again, I think you’re right that there’s some ambiguity, but I think that the argument can be reformulated such that your criticisms don’t really hurt it. Probably the simplest thing to do is to talk solely about the conditional probabilities that there exists a “universe with life in it.” If L = universe with life, then I think Pr(L|ASU) = Pr(FT|ASU), so the right-hand side of the equation stays staggeringly low. As I think your last paragraph shows, Pr(FT|T) is lower than Pr(L|T); but of course that doesn’t hurt the argument formulated this way. You say that this argument then has to deal with the life chauvinism argument; if that just means justifying giving a value many orders of magnitude above Pr(L|ASU) for Pr(L|T), then I think I’ve already done that.

As for the fact that most of the universe is hostile to life, there’s two ways one can respond. The first is to separate that out as another issue. If L is understood as simply the proposition that the universe permits life, then this is irrelevant to Pr(L|T), and so to the claim that Pr(L|T) >> Pr(L|ASU). That’s because L says nothing about how much life there is in the universe or what that life is like. Now, after we’ve updated our probabilities (and to do that, of course, we need to somehow get prior probabilities for T, which I’ll readily grant is another issue that needs to be addressed), *then* we can ask whether details about our universe disconfirm theism, and surely many do (e.g., the existence of horrendous evils). The question then becomes whether these other facts are enough to to lower our probability of T as much as observing L raised it. Maybe they do, but that doesn’t mean the fine-tuning argument fails, just that other anti-theistic arguments are weightier.

The other response is to to adjust L to take into account the specificities of our universe and see if that makes a difference. Certainly this makes Pr(L|T) much lower, but it makes Pr(L|ASU) much lower too (adding any sort of specificity will do that). The question is whether it effects Pr(L|T) to a sufficiently greater degree to destroy the argument. I don’t think it does. To begin with, it’s not clear to me that what I would expect if all I knew was that there was some super-intelligent designer who was going to create a universe is that he’d create a universe full of as much life as possible. If a super-intelligent human (or alien) was creating a universe, I don’t think I’d expect that. To be clear, I don’t think I’d particularly expect one with a small amount of life either; my experience with conscious intelligent creatures just doesn’t seem to give me much reason to favor either one. But be that as it may, I think the defender of the fine-tuning argument can grant that God is, say, a thousand times more likely (given what we know of him) to create a universe teeming with life than one where most of the universe is hostile to life; since Pr(L|ASU) is so staggeringly low, Pr(L|T) will probably still be many orders of magnitude above Pr(L|ASU). This, then, would decrease the amount of confirmation given to theism by the argument, but it would hardly eliminate it. (Maybe bringing in other facts, like the existence of evil, would do that; I don’t know. But, simply for the sake of being able to evaluate issues clearly, I think we should definitely treat that as a separate argument.)

One other point, on the issue of the prior probability of T: I think that what Steven R.’s genie analogy is getting at is really this issue. Where G = a genie who will, if he exists, see to it that you have exactly the number of hair in your head that you do, and N = you have that number of hairs, clearly Pr(N|G) >> Pr(N|~G). Of course, none of us would take this as good reason to believe that G. The standard way of explaining this is that G is incredibly ad hoc. It seems clear that T is not ad hoc in the same way: people have been advocating T long before we knew anything about fine-tuning, for instance. But one might still want some more satisfying explanation of why Pr(T) >> P(G). Unfortunately, priors are one of the difficulties faced in all Bayesian reasoning, and translating theoretical virtues and vices into them is a contentious matter. With that caveat in mind, my own suspicion is that the precise problem with a hypothesis like G is that it’s so specific. Indeed, it’s tailored precisely so that G entails N. I think that as we add (non-independently motivated) specificity to a hypothesis H to make it predict evidence E, we in effect lower the prior probability of H just as we’re raising Pr(E|H), so that they in effect cancel out (I haven’t tried spelling out exactly how this would work in Bayes’ Theorem, so I admit this is somewhat vague right now). That’s why we don’t take N as good grounds to believe G. Is T in a similar boat as G? I don’t think so. Even ignoring the fact that it was advocated before we knew that Pr(L|ASU) was so low, the hypothesis isn’t described in such terms as to conveniently explain the evidence. It’s not that specific of a hypothesis at all; it just says that there exists a super-intelligent, super-powerful conscious being with the capacity to create the universe. Nothing in the hypothesis specifies his desires, intentions, etc.; we’re able to get some grasp on that only by the kind of analogical reasoning I sketched earlier. Similarly, were we to discover complicated machines on another planet and infer that probably the hypothesis that intelligent life exists there is true, that would be a perfectly reasonable inference because we’d be judging that the probability of those machines given intelligent aliens is much greater than the probability of those machines given no intelligent life on the basis of an analogy with other intelligent life we know, and not on the basis of building a desire to create sophisticated machines into our hypothesis.

  (Quote)

Patrick February 2, 2011 at 3:54 pm

Troy- I just woke up from a nap, so I’ll write a full response later when I’m less fuzzy. I just want to make two points now.

1. I completely agree that arguments can be made on a number of issues that arise if the apologist is forced to stop using bad math in order to lie to his audience. My preoccupation in this thread (albeit not entirely expressed in the specific response to you) is that I really, really hate it when people make a bayesian argument and then cheat. And that’s how the fine tuning argument is always presented- first, they use different hypotheses depending on which side of the bayesian equation they’re on, and then afterward they claim they don’t need to address priors because the resulting number is so low. They’re hacks. They need to actually do bayes right, adjust the strength of their conclusion accordingly, and respond to issues that arise as a consequence of their line of reasoning, which includes responding to other bayesian arguments about the state of the universe versus their hypothesized magical superbeing.

2. Your point about the genie thing is off. The problem is that you’ve engaged in a bayesian argument in which the hypothesis is a magical superbeing who possesses the power and desire to bring about exactly the known fact that you are using as the evidence portion of your bayesian analysis. The probability of a particular fact given the existence of a magical superbeing who possesses the power and desire to bring about that particular fact is going to be 1 every time, which is a great way to generate spurious arguments that appear correct: do this, and then give a good dodge for why you don’t have to discuss priors. Oh! Hey! That’s exactly what fine tuning advocates do. The way bayes is supposed to account for this is in our priors, not in arguments about how ad hoc something is.

3. Arguing that a magical superbeing would be reasonably likely to have desires similar to a person, based on our prior experiences of people, is silly. Come on now. People like people because people are social animals who grow up in communities. The hypothesized magical superbeing contains none of those characteristics. In fact, it contains no characteristics at all. If you are going to IMPORT human characteristics, you must adjust your hypothesis accordingly, and as a result, adjust your priors accordingly.

4. People who want to argue that the probability of fine tuning under atheism is low AND that the probability of our particular universe under theism is not low need to use the same forms of logic about probability on both sides of the equation. It would be perfectly reasonable for me to answer them by listing out all the possible sizes and types of life permitting universes that are logically possible (not physically given the big bang or anything, logically a magical superbeing), and then putting that number under a 1. This is EXACTLY the logic used in calculating Pr (FTU|A).

  (Quote)

Troy February 2, 2011 at 5:48 pm

@Patrick:

A few brief points.

Re 1: I don’t think we’re really in disagreement here. If someone’s presenting the fine-tuning argument, or any other evidential argument, as decisive proof for God’s existence, they’re just confused; inductive arguments can’t give us decisive proof of anything. But that doesn’t mean the arguments have no merit, just that we need to be clear about what they are and are not doing. If we want to make a complete analysis of the probability of theistic propositions, then clearly we need to do more than just talk about fine-tuning, but that’s another project (and one that is probably more complicated than can adequately be discussed in a single blog thread).

Re 2: To begin with, the hypothesis I was evaluating mentioned nothing about the designer’s desires, and I have been quite explicitly disavowing the claim that Pr(L|T) is 1 or even anywhere near 1. The hypothesis also, as I noted, was not made up just for the purposes of accounting for this data; it’s a hypothesis that has been around for a while. And even if it hadn’t been, we have other experiences of conscious beings doing things intentionally, which would provide motivation for a similar hypothesis here. Now perhaps these just have to do with how “ad hoc” the hypothesis is, and not its prior probability. Maybe. But I’m skeptical that there are better methods out there for evaluating prior probabilities–although if you have some I’d love to hear them!

Re 3: I didn’t say “reasonably likely.” The probability could be 1/10^10, and the argument would still go through. Granted, if we wanted to do a complete analysis of the probability of God, we’d have to look at all the other facts out there too, and then we’d want to have a more precise estimate. But as far as the fine-tuning argument is concerned, a probability this low will still allow for a very high confirmation of theism.

At any rate, I don’t think your criticism pushes the probability even that low. Your point, as I understand it, is that we can explain why human beings would be interested in creating life, but that that explanation wouldn’t apply to God. It may be true that humans are social animals for the reasons you say. But in order to destroy the validity of any analogical reasoning from us to God, your confidence that that was the explanation, that that was the only explanation, and that that explanation did not apply to God, would have to be practically 1. Otherwise you can’t be sufficiently sure that the explanation isn’t something common between us and God, and if it is, then the analogy goes through. And I don’t think we know nearly enough about intelligent life, or about consciousness in general, to have such high confidence in something like that. To return to my alien analogy: suppose we encountered an intelligent alien who we knew to be the only one of his kind and to have never encountered any other living things. If we saw that alien tinkering around creating universes, I’d still think that he was more likely to create ones with life in them than ones with planets with fiery rings, etc. I’ll grant that I’d probably be less confident than if he was a member of a thriving alien community, but not by that much.

Re 4: I don’t have firm opinions about how to reconcile differing interpretations of probability, but I don’t think what you say here is right. Suppose we’re playing poker, and I draw 5 royal flushes in a row. It’s pretty clear that the probability of that is much higher given that I’m cheating than given that I’m not cheating (and that’s clear intuitively, even if you have no idea of how to do the relevant math). But suppose you accused me of cheating, and when I asked you to justify that, you showed me how unlikely it is that those hands would come up by chance, and then observed that it’s much more likely that they would come up if I’m cheating (even though, incidentally, it’s probably not *that* likely even if I am cheating). It’d be absurd for me to say “Ah, but you’re conflating two different types of probability–if you list all the possible combinations of ways that those hands could have turned up given chance and put that under a 1, then I can list all the combinations of ways they could have turned up given that I cheated, and look–it’s just as many!”

Looking back, those points weren’t really brief. But then, your points weren’t really 2, so I suppose we’re even. :-)

  (Quote)

Troy February 2, 2011 at 5:57 pm

Because I can’t resist adding this:

To continue with the poker analogy, suppose that after your accusation I pointed out that if I were cheating, I’d almost certainly be more discrete about it, and chosen for myself something besides five royal flushes, e.g., I’d have chosen hands that would only just have let me win without arousing your suspicion. That’s probably true, as far as it goes, but it doesn’t drive Pr(Five Royal Flushes|I’m Cheating) *nearly* low enough for you to start doubting that I’m cheating. It seems to me that your (and others’) points about what God would have been likely to do had he been aiming to create a life-filled universe are analogous. They may (although I’m skeptical of this in many of the cases, for reasons already expressed) reduce Pr(L|T), where L is understood in the narrower sense, but not nearly enough to defeat the fine-tuning argument.

  (Quote)

Troy February 2, 2011 at 6:04 pm

Because I keep thinking of more things to say;

I would be curious, Patrick (and perhaps you were already going to respond to this in your longer post), what you think about my incidental response to Larkus earlier, as I think it applies to you too: if you deny the possibility of any sort of analogical reasoning about a supernatural being, you are in essence making it impossible for the theist to present any evidence for his position, since we aren’t in a position to make any judgments about the probability of God doing one thing rather than another. No apparently supernatural event, no religious experience, no occurrence at all could confirm theism. To me this seems a reductio of your position, but one person’s modus tollens is another’s modus ponens, so I’d be interested in seeing (a) if you agree that this follows from what you’ve said and (b) if so, if you think this is right (nothing could confirm theism).

  (Quote)

rosyatrandom February 2, 2011 at 7:59 pm

The thing about poker is that we understand the mechanics of how it works – how cards may be drawn and the probabilities of hands etc.. We can calculate statistical distributions of expected hands and the exact unlikeliness of any particular sequence. We also know that cheating/manipulation is possible and thus, as the only alternative to random chance, the probability that this is going on.

With universes, we only have one to go on. We know from the weak anthropic principle that however ‘unlikely’ it is, we can only observe universes that contain us in the first place. So straight away we can see that if there are lots of universes and only ‘unlikely’ ones contain life, that is where we find ourselves. Given that at least one universe exists, I think there would need to be a damned good reason for others not to exist as well. We also know of no reason why a god should exist, what sort of god should exist, and what its nature is. All we have are entities posited from the top down that perform a function/embody some kind of principle but without any kind of reasoning as to what it really means or why it should be given any kind of metaphysical primacy. If we went back to poker, the theistic equivalent would be a wizard giving you royal flushes but in a way we can’t explain or observe, but have to give credence to no matter how much progress we make solving the problem naturalistically. And even further, people might be saying the wizard might still exist even if we found out the deck had just been tampered with.

In short, give me an explanation or at least show me where an explanation might exist. If you can show that there can be _no_ explanation, well, then we might have a cause to invoke ineffable, unaccountable, arbitrarily just-so beings from outside our reality. Otherwise… no. Just, no.

  (Quote)

ayer February 2, 2011 at 8:11 pm

On a level 2 multiverse the constants can vary from universe to universe, depending on the model used.

If by Eternal Inflation you are referring to the multiverse, I agree that it is a viable alternative theory to theism as an explanation for fine-tuning–but it’s a philosophical, not a scientific, explanation, since it, like theism, is non-falsifiable. And as a philosophical explanation it is inferior to theism under Occam’s Razor.

  (Quote)

ayer February 2, 2011 at 8:11 pm

On a level 2 multiverse the constants can vary from universe to universe, depending on the model used.

If by Eternal Inflation you are referring to the multiverse, I agree that it is a viable alternative theory to theism as an explanation for fine-tuning–but it’s a philosophical, not a scientific, explanation, since it, like theism, is non-falsifiable. And as a philosophical explanation it is inferior to theism under Occam’s Razor.

  (Quote)

rosyatrandom February 2, 2011 at 8:19 pm

And as a philosophical explanation it is inferior to theism under Occam’s Razor.

That depends on how you see things. You’re just setting the universe in context, allowing what we already posit to occur to happen on a wider and more general scale; it’s certainly arguable that this multiverse is a simpler entity than the universe, and implied by it. (In a similar way, if we know 7,540,498 exists, then we should not be surprised if other numbers – indeed, all other numbers – exist. For them not to exist would be strange.)

Theism, however, adds an entity of a qualitatively different kind (as opposed to extending one we already have), without existential justification.

  (Quote)

ayer February 2, 2011 at 8:23 pm

Also, the size of the universe is relevant because when we’re comparing two hypotheses such as inflation and theism, this is one of the many reasons to prefer inflation, which does predict a big universe.

Theistic fine-tuning also predicts a big universe, since the size of an expanding universe is a function of time, and galaxies, stars, planets, evolution of life, etc. unfolded over time in accordance with the initial conditions built in to the Big Bang. That is why the size of the universe is irrelevant, even if life only exists on Earth. (I’m not sure when the general atheist position shifted from Carl Sagan’s “the universe is teeming with life, so we are not special, thus there is no God” to “the universe is empty of life except for Earth, so life is not special, thus there is not God”, but it’s an interesting turn in the argument).

  (Quote)

ayer February 2, 2011 at 8:28 pm

That depends on how you see things. You’re just setting the universe in context, allowing what we already posit to occur to happen on a wider and more general scale; it’s certainly arguable that this multiverse is a simpler entity than the universe, and implied by it. (In a similar way, if we know 7,540,498 exists, then we should not be surprised if other numbers – indeed, all other numbers – exist. For them not to exist would be strange.)Theism, however, adds an entity of a qualitatively different kind (as opposed to extending one we already have), without existential justification.  

The mulitiverse posits billions of additional entities, in addition to some unknown naturalistic mechanism which pumps out all of these entities, to explain the unique characteristics of the only universe we know; theism posits a single entity.

  (Quote)

rosyatrandom February 2, 2011 at 8:31 pm

Saying ‘god happened to choose things this way that’s all’ does not retro-actively predict it. It just allows it to have magically happened. Inflation as a mechanism of generating universes does predict an expanding universe, and allows for ‘unlikely’ ones.

As to life teeming or not… it’s just a matter of scale, and not really relevant.

  (Quote)

rosyatrandom February 2, 2011 at 8:33 pm

The mulitiverse posits billions of additional entities, in addition to some unknown naturalistic mechanism which pumps out all of these entities, to explain the unique characteristics of the only universe we know; theism posits a single entity.

No, it’s just a bigger entity. It’s one system. Again, I submit my numerical analogy.

  (Quote)

ayer February 2, 2011 at 9:23 pm

No, it’s just a bigger entity. It’s one system. Again, I submit my numerical analogy.  

If you lump it all together and call it “one entity,” then you have a hideously complex entity with an enormity of composite parts and sub-entities, all needed to explain the existence of the “fine-tuned” sub-entity in which we exist; as opposed to the theistic view–one creating, non-composite entity that exists as pure mind with no physical parts, which creates a single universe. Theism wins on Occam’s Razor hands down.

  (Quote)

Eric February 2, 2011 at 10:18 pm

WLC –
Letting “FT” represent the fine-tuning of the universe for intelligent life, “T” represent theism, and “ASU” represent the atheistic single universe hypothesis (i.e., there is a single universe and no God), Collins argues that the fine-tuning is significantly more probable on theism that it is on atheism: Pr (FT/T) >> Pr (FT/ASU). Therefore, the observed fine-tuning confirms the hypothesis of theism.

I’m sorry but isnt this a logical fallacy. P(FT/T) >> P(FT/ASU) is NOT EQUIVALENT TO P(T/FT) >> (P(ASU/FT). I thought this was the whole point of Bayes’ theorem.
If they were equivalent, then the whole argument would be arbitrary. For example, take this poker analogy:
A: I draw a hand of a 3 of Diamonds (3D), 4C, AC, JS, 7H
B: I played the game fairly and drew those cards by chance
C: I cheated in order to draw those cards
Given a full deck:
P(A|B) =~ 1/(311 Million)
P(A|C)=~1
Using Craig’s logic:
” Pr (A/CI) >> Pr (A/B). Therefore, the observed hand confirms the hypothesis that I cheated in order to draw those cards.

Before you could jump to those conclusions, you would have to show the prior probability of C is high, just as William Lane Craig would have to show the prior probability of Theism is high.

  (Quote)

Eric February 2, 2011 at 10:53 pm

Also, its not just the prior probability of Theism, its the prior probability of Theism where we would have a god that WOULD PROBABLY CREATE THIS UNIVERSE. Else the problem is that a God could create ANY logically possible universe, while only metaphysically possible universes could exist under naturalism. So Pr (FT/T) << Pr (FT/ASU), which defeats the entire point of this argument.

  (Quote)

Eric February 2, 2011 at 11:36 pm

ayer –
If you lump it all together and call it “one entity,” then you have a hideously complex entity with an enormity of composite parts and sub-entities, all needed to explain the existence of the “fine-tuned” sub-entity in which we exist;

But this entity is not necessarily complex since every sub-entity is basically made of the same “stuff,” matter, energy, space, and time. No new entity is needed in this regard. If all these multiverses are reducible, then it is favored under Occam’s Razor.

Ayer-
as opposed to the theistic view–one creating, non-composite entity that exists as pure mind with no physical parts, which creates a single universe. Theism wins on Occam’s Razor hands down.

But a mind that is not reducible to simple matter is incredibly complex. Every possible trait of this mind has to exist as brute fact, which their are many triats, unless you can reduce a mind that WOULD CREATE THIS UNIVERSE DOWN TO SIMPLER PARTS. Because the “ontological” argument’s qualities of God do not deductively result in a God that would create this particular universe, then it sounds like this kind of mind would be incredibly complex and fail under Occam’s razor.
So I do not agree theism wins hands down, absent a word game.
At the very least, you have to show that this mind is simpler than more identical universes (identical in the fact it consists of the same stuff) before you can claim “Theism wins on Occam’s Razor hands down.”

  (Quote)

rosyatrandom February 3, 2011 at 3:10 am

Ah, good. I go to sleep and someone answers for me :)

The universe is more complex than the (hypothetical) multiverse that contains it. This is a subtle point, but worth hammering in. The universe, by itself, is far more specified and arbitrary; it has to happen ‘just-so’ for no real reason, without any kind of underlying ontology given to generate it and make it happen in the first place.

The multiverse, on the other hand, is an extension of our universe both spatially (not a problem for Occam) and structurally (simplifying the underlying rules by subsuming them into a more general one. Once again, the set of integers is simpler than any particular number by itself. This reasoning applies to pretty much any kind of multiversal theory.

And God is, as said above, just this highly specified and unexplained magical entity existing by fiat. I would also ask again that if God could create one universe, surely he could create more — all other possible ones, in fact — and that by nature of being God contains all these universes in his mind. Which is the same way that our universe exists theistically speaking, right?

Anyway, before I go I shall quote Tegmark from http://en.wikipedia.org/wiki/Multiverse:

“A skeptic worries about all the information necessary to specify all those unseen worlds. But an entire ensemble is often much simpler than one of its members. This principle can be stated more formally using the notion of algorithmic information content. The algorithmic information content in a number is, roughly speaking, the length of the shortest computer program that will produce that number as output. For example, consider the set of all integers. Which is simpler, the whole set or just one number? Naively, you might think that a single number is simpler, but the entire set can be generated by quite a trivial computer program, whereas a single number can be hugely long. Therefore, the whole set is actually simpler. Similarly, the set of all solutions to Einstein’s field equations is simpler than a specific solution. The former is described by a few equations, whereas the latter requires the specification of vast amounts of initial data on some hypersurface. The lesson is that complexity increases when we restrict our attention to one particular element in an ensemble, thereby losing the symmetry and simplicity that were inherent in the totality of all the elements taken together. In this sense, the higher-level multiverses are simpler. Going from our universe to the Level I multiverse eliminates the need to specify initial conditions, upgrading to Level II eliminates the need to specify physical constants, and the Level IV multiverse eliminates the need to specify anything at all.”

I would characterise your position as someone who, on going into a forest, would be continually surprised at every tree they see.

  (Quote)

ayer February 3, 2011 at 5:53 am

But this entity is not necessarily complex since every sub-entity is basically made of the same “stuff,” matter, energy, space, and time.

You are forgetting to include whatever naturalistic mechanism generates the mulitverse out of matter, energy, space and time. And even not counting that mechanism, pure mind is simpler than the four components of matter, energy, space and time.

But a mind that is not reducible to simple matter is incredibly complex. Every possible trait of this mind has to exist as brute fact, which their are many triats, unless you can reduce a mind that WOULD CREATE THIS UNIVERSE DOWN TO SIMPLER PARTS.

You are confusing the mind with its thoughts. Pure mind itself is completely simple; if the thoughts are complex, it could stop thinking those thoughts; they are epiphenomal and not brute.

  (Quote)

ayer February 3, 2011 at 6:02 am

The multiverse, on the other hand, is an extension of our universe both spatially (not a problem for Occam)

It certainly is a problem; these universes are completely independent of each other (indeed, are undetectable from our own); the multiverse can by no means be called “non-composite”

structurally (simplifying the underlying rules by subsuming them into a more general one

You have made no case that “the more general one” (the multiverse-generating mechanism?) is by definition simple; as a naturalistic mechanism, it would likely be as complex as the naturalistic sub-entities it is generating.

I would also ask again that if God could create one universe, surely he could create more — all other possible ones, in fact — and that by nature of being God contains all these universes in his mind.

As with Eric, you are confusing the thoughts of pure mind with pure mind itself. The thoughts are not necessary components of pure mind; they are epiphenomenal, while the mind is completely non-composite

  (Quote)

rosyatrandom February 3, 2011 at 6:17 am

You are forgetting to include whatever naturalistic mechanism generates the mulitverse out of matter, energy, space and time.

Err, no. MEST are facets of the multiverse, not the other way around. And without external time, the multiverse requires no creation. If you mean what drives inflation, well, it seems you are always going to be unable to look at naturalistic phenomena without trying to peek behind the curtain. What makes 1+1=2? It’s intrinsic to their very nature. what makes spacetime inflate? Its nature. I don’t know what that exact mathematical structure is, and I would love to know, but I’m not going to start looking for magical solutions instead.

And even not counting that mechanism, pure mind is simpler than the four components of matter, energy, space and time.

Really now? My personal view is that ‘mind’ is an abstract structure in no different a way than the physical realm. They exist independently yet at the same time fundamentally entwined – mind implies the physical, and the physical has mind embedded within it. But to say it’s simpler… I’m sorry, I have to call bullshit on this. You are in no position to understand mind, MEST or judge their simplicity.

You are confusing the mind with its thoughts.Pure mind itself is completely simple; if the thoughts are complex, it could stop thinking those thoughts; they are [epiphenomenal] and not brute.

The hardware is simple, the software is not? Mathematically speaking, there is no difference.

Here, let me try and convey what I consider an important thing here. You are very hung up on two things I consider major stumbling blocks: that there is some kind of existential barrier that makes it _hard_ for things to exist, and that there is some kind of privileged metaphysical viewpoint that delineates exactly _what_ things are.

The hardware/software distinction above? They are exactly equivalent, mathematically; you can posit a spectra of hypothetical computers split along hardware/software lines, with data structures encoded in different ways so that they are processed as-is or through algorithmic decompression… and they are all the same thing. It makes no difference to the output.

We do not see the multiple worlds of QM because our minds do not span them. They exist, but our perspective makes us see the world in a classical way. This is artificial. It depends entirely on viewpoint. Everything does.

  (Quote)

rosyatrandom February 3, 2011 at 6:29 am

It certainly is a problem; these universes are completely independent of each other (indeed, are undetectable from our own); the multiverse can by no means be called “non-composite”

I fail to see a problem here. I really do. From our internal, temporally-bound perspective, there might well be no detectable interaction with them, but a desitic god would have exactly the same problem. It is only our limited standpoint that makes us unable to see that the universe is merely a part of a greater whole, and I don’t think reality cares too much about our ignorance.

You have made no case that “the more general one” (the multiverse-generating mechanism?) is by definition simple; as a naturalistic mechanism, it would likely be as complex as the naturalistic sub-entities it is generating.

No. It is, as I quoted from Tegmark, a case akin to algorithmic complexity. Indeed, the simplest way of generating output containing complex structure is essentially a random number generator, and I don’t really make the distinction between process and output anyway.

As with Eric, you are confusing the thoughts of pure mind with pure mind itself.The thoughts are not necessary components of pure mind; they are epiphenomenal, while the mind is completely non-composite

It is amazing how you know these things!

  (Quote)

PDH February 3, 2011 at 7:28 am

rosyatrandron has covered most of what I would say but this really is a crucial point.

Ockham was a guy living hundreds and hundreds of years ago. I don’t mean to diminish his importance but things have moved on somewhat since then. Multiverse hypotheses typically come out remarkably well on modern treatments of parsimony like minimum message length and Kolmogorov complexity.

There’s no law that says we can’t have a big, complex universe. On the contrary, it would take additional laws to explain why inflation happened once and then just stopped. Because why would it? Any hypothesis that suggested such would be a more cumbersome hypothesis than inflation, with additional, ad-hoc rules accounting for why it happens once (und only vonce!) instead of eternally.

A mind/personal agent/intelligent designer – anything like that – is going to be horribly unparsimonious. You can call it metaphysically simple if you want but the amount of information it would take to describe it will still be monstrously vast. Because it still does all of the things that a regular personal agent does and that takes a lot of information to describe. Inflation, meanwhile, mostly relies on well-established physics. It only takes a small amount of additional information.

There is a lot more that I could say but, actually, the first thing I ever read from Yudkowsky (a couple of years ago when I didn’t really know who he was) was this article, which covers the issues much better than I could: http://lesswrong.com/lw/q3/decoherence_is_simple/

The notion that decoherent worlds are additional entities penalized by Occam’s Razor, is just plain mistaken. It is not sort-of-right. It is not an argument that is weak but still valid. It is not a defensible position that could be shored up with further arguments. It is entirely defective as probability theory. It is not fixable. It is bad math. 2 + 2 = 3.

(Also of interest are: http://lesswrong.com/lw/jp/occams_razor/ and http://lesswrong.com/lw/q4/decoherence_is_falsifiable_and_testable/ )

Now that’s a level 2 but the same reasoning applies.

  (Quote)

ayer February 3, 2011 at 7:35 am

It is amazing how you know these things!  

For someone who asserts knowledge of billions of undetectible parallel universes, that’s quite a statement of chutzpah

  (Quote)

rosyatrandom February 3, 2011 at 7:37 am

One universe, a billion universes… what’s the difference?

  (Quote)

Brian_G February 3, 2011 at 9:33 am

Ok… If this is becoming an argument over who said what when, instead of about the issues being discussed, let me know and I’ll stop responding. Honestly, this isn’t a big issue for me- I’m more interested in the fact that your specific assertion (that fine tuning can be assigned a high probability under theism because even if fine tuning didn’t exist under theism we’d probably think it did, but that somehow this doesn’t require completely scrapping the entire fine tuning argument and starting over with a probabilistic analysis of a conjunction of multiple probability statements designed to add up to the chance we think that fine tuning exists) is plainly illogical.

As for the specific issues you’ve mentioned, 3 is a non starter. Adding 50 new cosmological constants wouldn’t change the probabilities involved in the values of the small handful. 2 is also a non starter because it doesn’t advance the argument you were attempting to defend. I mean, yes, technically, it is a defense of the specific attack I made against fine tuning to respond that the attack may be irrelevant because fine tuning isn’t true, but that’s not going to help the bottom line, so to speak. I’m not sure about 1 because I’m not sure what conclusion you’re arguing for anymore.

My original point still stands, however. Anyone who equates fine tuning with life permitting without an argument that all logically possible life permitting universes (including those that could be created by magic) are finely tuned universes is a chump, and anyone who makes this elision in published papers is either a crap philosopher or mathematician, or a charlatan. The point is non obvious, and needs defending.

Have I said something to offend you? You seem upset. To try do address some of your concerns.

1. You said that you didn’t know what I was arguing. I’m defending the position that:
P(FTU | T) >> P(FTU|A)

Now if you want we can change this to perceiving a fine tuned universe (PFTU):
P(PFTU|T) >> P(PFTU|A)

I don’t see any way that this hurts the argument. We’d have to be able to perceive a fine-tuned universe, in order to make the argument. It may even help the argument somewhat. We could imagine a world that was precisely fine-tuned but the inhabitants had know idea how fine-tuned the universe is. Under theism, there are plausible reasons God might want us to perceive a fine-tuned universe. (For example, he might have wanted to leave clues about his existence.) However, under atheism, there’s no reason for us to expect the universe to make it’s fine-tuning known to us.

2. I don’t think that life-permitting is the same as fine-tuning. Life permitting means that the universe has conditions necessary for life. Fine tuning means that those conditions are astonishingly narrow. As I’ve said before, I think an atheist could reasonably expect that the universe is life permitting without expecting it to be finely tuned. One atheist, Victor Stenger defends this position. In theory, perhaps God could also create a universe that wasn’t fine-tuned, or at least one that didn’t look fine-tuned to us, but I have no reason to expect that under theism. In my view this is a mere possibility equivalent to the possibility that we were created 30 seconds ago by a gremlin who implanted memories in our brains and fossils in the earth to make it look as if the universe is a lot older. I have no reason to believe that, but I can’t deny it as a possibility.

  (Quote)

rosyatrandom February 3, 2011 at 9:39 am

In my view this is a mere possibility equivalent to the possibility that we were created 30 seconds ago by a gremlin who implanted memories in our brains and fossils in the earth to make it look as if the universe is a lot older. I have no reason to believe that, but I can’t deny it as a possibility.  

If we allow creation ex nihilo, what is the difference between an actually old universe and one that just appears so?

  (Quote)

Patrick February 3, 2011 at 10:03 am

Because I keep thinking of more things to say;I would be curious, Patrick (and perhaps you were already going to respond to this in your longer post), what you think about my incidental response to Larkus earlier, as I think it applies to you too: if you deny the possibility of any sort of analogical reasoning about a supernatural being, you are in essence making it impossible for the theist to present any evidence for his position, since we aren’t in a position to make any judgments about the probability of God doing one thing rather than another. No apparently supernatural event, no religious experience, no occurrence at all could confirm theism. To me this seems a reductio of your position, but one person’s modus tollens is another’s modus ponens, so I’d be interested in seeing (a) if you agree that this follows from what you’ve said and (b) if so, if you think this is right (nothing could confirm theism).  

Ok, I”m going to be bouncing around as I read this and respond.

I’m perfectly ok with you making arguments about what god would prefer.

But I have a list of rules I apply when analyzing apologetic arguments.

First, I actually remember the arguments, and feel its completely fair to point out when an argument in one context contradicts an argument in another. For example, once a theist argues that god would naturally want to create life because that’s what he thinks a person would do and people are our best source of information about sentient beings, then I’m going to hold the theist to that standard three hours later in the debate when we’re discussing the argument from evil and the theist doesn’t want to apply human standards to god’s behavior. In this context, once you argue that naturally the magical superbeing would want to create life because that’s what a human would do, and then in the same breath argue that the magical superbeing wouldn’t necessarily care about how much life it creates, nor about whether the universe created out of a motivation to have a life supporting universe is in fact actually a blasted lifeless wasteland in 99.9999…% of its volume, and your reasoning is that you haven’t got any experience with magical superbeings so you can’t say whether that’s unlikely… then I’m going to get very frustrated, because you’ve changed your standard of evaluation halfway through a paragraph. The assumptions that let you make the first argument preclude you from making the second.

Second, I’m going to insist that you update your bayesian argument every time you add new information about the magical superbeing. Just as running a bayesian argument on “a magical superbeing” generates one set of priors. Running a bayesian argument on “a magical superbeing who wants to create life” generates a second set of priors that takes into account the chance that a magical superbeing might not want to create life. These are only equivalent if you think the chance that a magical superbeing would want to create life is 100%. If so, you need to present an argument.

The fine tuning argument frustrates me because its always presented as such a slippery, dishonest piece of junk. Its supposedly an argument from probability, but it has at least three terms in it, and theists never defend two of them. Arguments with significant undefended premises are stupid arguments. To deflect attention away from this the theist just focuses as hard as they can on P(FTU|A), and hides the ball as much as possible on the other terms. Well, screw that.

  (Quote)

Eric February 3, 2011 at 11:26 am

ayer –
You are forgetting to include whatever naturalistic mechanism generates the mulitverse out of matter, energy, space and time.

Seems like rosyatrandom answered this. Also if our universe is reducible to what caused it, as the elements are reducible to quarks, then we have no violation of Occam’s razor. And if what our universe is reducible to results in more universes, then we yet again have no violation of Occam’s razor. We have just moved the set of brute facts back one level without increasing them. In fact, it may even decrease the set of brute facts (although I don’t know enough about the multiverse to tell this for sure), in which case it is more favorable to Occam’s razor to accept expansion theory and the resulting multiverse than it is to accept a single universe.

ayer –
And even not counting that mechanism, pure mind is simpler than the four components of matter, energy, space and time. You are confusing the mind with its thoughts. Pure mind itself is completely simple; if the thoughts are complex, it could stop thinking those thoughts; they are epiphenomal and not brute.

Please provide an argument that shows how mind is simple. Instead, as PDH puts, “A mind/personal agent/intelligent designer – anything like that – is going to be horribly unparsimonious. You can call it metaphysically simple if you want but the amount of information it would take to describe it will still be monstrously vast.” . Lets give pseudo code:
if (situation A)
{
Action A
}
Think of situation A as a thought and Action A as the action. Notice that this is one situation among a near infinite number of possible situations. Each one must be defined. Similarly, every combination (OR, AND, NOT) must also be defined. Now if the mind is reducible to a smaller set of physical matter, and all the actions of a mind are reducible to the smaller set of chemical properties of this matter, which are reducible to a small number of fundamental particles and their properties, then a mind would become much simpler to code. But this God Mind is not reducible, and therefore MUCH more complex. Note: Notice this code does not define thoughts, but how a mind works.

  (Quote)

ayer February 3, 2011 at 8:33 pm

Also if our universe is reducible to what caused it, as the elements are reducible to quarks, then we have no violation of Occam’s razor.

A multiverse which consists of billions of different arrangements of trillions of different quarks is the most extreme opposite of “non-composite” simplicity; indeed, it would be “composite” in excelsis, and would flunk Occam’s razor miserably.

But this God Mind is not reducible

That is the very reason it is simple–it is the ultimate non-composite entity.

  (Quote)

Eric February 4, 2011 at 1:11 am

ayer
A multiverse which consists of billions of different arrangements of trillions of different quarks is the most extreme opposite of “non-composite” simplicity; indeed, it would be “composite” in excelsis, and would flunk Occam’s razor miserably.

but all these arrangements are arbitrary and do not need to exist as independent brute facts. So I fail to see how they flunk occam’s razor. This sounds like a very trivial use of the concept of complexity. To program a quark as an object in a computer program would be MUCH more simple than programming an entire mind (total understatement). And if a process exists by brute fact which creates this object, you would need an additional brute fact to show why it only occurred once or multiple times. Either way, you don’t need additional brute facts to explain why more than one was created over just one.

ayer-
That is the very reason it is simple–it is the ultimate non-composite entity.

I guess you ignored the reason why being reducible makes it simpler. This God Mind must consist of an immense set of brute facts which JUST EXIST. If the mind is reducible, a much smaller set of brute facts need to exist. I already explained this but instead of responding, you just assert the opposite.

  (Quote)

rosyatrandom February 4, 2011 at 2:38 am

And futhermore, if one ‘simple’ mind exists, then why not many such ones? All with different character, capabilities, intentions….

  (Quote)

Larkus February 4, 2011 at 2:51 am

@Troy

“It’s not that specific of a hypothesis at all [theism]; it just says that there exists a super-intelligent, super-powerful conscious being with the capacity to create the universe. Nothing in the hypothesis specifies his desires, intentions, etc.”

Correct.

But you are offering an amended version. You add the hypotheses that

1) human beings would prefer to “create a universe with intelligent life before other types of universes” and

2) a super-powerful intelligent being who may have created the universe has preferences similar to (what you percieve to be) the preferences of human beings (in regard to “what kind of universes […] he’s likely to create”).

Reasoning by analogy is not bad. But not every reasoning by analogy is good. You make a hasty generalization from a single example of intelligence, human intelligence, to intelligence in general, including non-human intelligence.

From the Fallacy Files: Hasty Generalization

Exposition:

This is the fallacy of generalizing about a population [in your case the population of all possible kinds of intelligent beings] based upon a sample which is too small to be representative. If the population is heterogeneous, then the sample needs to be large enough to represent the population’s variability. With a completely homogeneous population, a sample of one is sufficiently large, so it is impossible to put an absolute lower limit on sample size. Rather, sample size depends directly upon the variability of the population: the more heterogeneous a population, the larger the sample required. For instance, people tend to be quite variable in their political opinions, so that public opinion polls need fairly large samples to be accurate.

How likely is it, that your sample is representative?

  (Quote)

ayer February 4, 2011 at 6:57 am

To program a quark as an object in a computer program would be MUCH more simple than programming an entire mind (total understatement).

This God Mind must consist of an immense set of brute facts which JUST EXIST.

You are completely failing to grasp the difference between “composite” and “non-composite.” I don’t know if you are a computer programmer, but your response regarding “coding” mind shows a fundamental misunderstanding of the concept of “mind” in the field of philosophy of mind. You have to break out of your materialist presuppositions to deal with the concept properly. You need to show how the naturalistic multiverse is non-composite in order for the multiverse to prevail under Occam’s razor, and since, as you admit, it would have to be composed of trillions upon trillions of quarks, it is composite in excelsis by definition. A mind of a personal agent, on the other hand, can produce complex thoughts, but those thoughts are not constituent “parts” of the mind; it can simply stop thinking those thoughts and think something else. The pure mind itself is non-composite and thus the very essence of simplicity.

  (Quote)

rosyatrandom February 4, 2011 at 7:01 am

Ayer, this is just wishful thinking on your part. You’re projecting your biases to make naturalism complex and theism simply by sheer assertion.

  (Quote)

rosyatrandom February 4, 2011 at 7:01 am

^typo: should be ‘simple by sheer assertion’.

  (Quote)

ayer February 4, 2011 at 7:26 am

Ayer, this is just wishful thinking on your part. You’re projecting your biases to make naturalism complex and theism simply by sheer assertion.  

Of course, it seems to be that you and Eric are the ones projecting your biases to make naturalism simple and theism complex by sheer assertion. Oh well, minds rarely get changed in these sort of blog exchanges.

  (Quote)

rosyatrandom February 4, 2011 at 7:56 am

Then let us try to genuinely get somewhere here. What does it mean for something to be simple or complex, metaphysically speaking?

Both god and a naturalistic system that is/contains the universe/multiverse are outside of time, so it is incorrect to think of them coming into being or changing – time is just an internal aspect.

We know from Relativity that spacetime is not an absolute, unambiguous thing – what we observe depends upon our perspective. And I would go further and say that EVERYTHING we witness of reality is the same; we see laws, behaviours and reductions into particles and forces that seem to be set-in-stone building-blocks of the universe… but these are really just things that have crystallised out of the true internal structure of the universe from the seed that is our minds.

Here’s a metaphor I find useful: the universe is a kaleidoscope, dense with order and complexity. The exact pattern you observe depends on where you twist the dial. This brings some relations into tight focus and hides others behind a mask of noise. What we see would be entirely arbitrary except for the fact that the dial position represents *you*. What we see is a kind of slice through the… uberverse? No, too pretentious. oververse. Our minds are like tangled webs that, when drawn together, pull the chaos of reality into an order reflecting the order within our minds.

So we have this chaotic oververse. Pretty complex and unlikely, right? No. I would view it more as a kind of fractal; an entity whose apparent complexity derives from very simple rules.

And this is the thing! Do you see a fractal as complex or simple? It is the same point people have been making with regard to Kolmorov Complexity – an entity’s true complexity is based upon how hard it is to define/construct.

So, for me, a single universe in isolation is highly complex because it lacks context – we would have to generate everything by specifying highly arbitrary contingencies for it all. In short, it does not make sense to me. But the multiverse… that is far simpler! You see a bunch of additional entities that each require as much explanation as the first, but we see a single entity that requires less.

Why do we have this difference in perspective? Well, as I said before, you seem to be hung up on existence being difficult and with a component-oriented rather than holistic view. You see all the bits in the universe and treat its complexity as the sum of them, but we look at it all together and see the less-complex picture.

Strangely, though, you do not apply this same skepticism to god. God just seems to exist, and its thoughts can be dismissed as being merely riders on the simple substrate of mind itself.

Now, if this god exists, its nature and thoughts are highly specific. Why not other gods and other minds? How do we draw statistical conclusions from such a morass of potential progenitor gods? And I would also say again that I don’t think you can just separate Mind and Thoughts like that. Like hardware and software, they are two aspects of the same thing, and without physical hardware and interacting apparatus the distinction is meaningless.

I would also like to wonder how a timeless mind works, apparently thinking thoughts in sequence.

  (Quote)

Steven R. February 4, 2011 at 9:11 am

Re 4: I don’t have firm opinions about how to reconcile differing interpretations of probability, but I don’t think what you say here is right. Suppose we’re playing poker, and I draw 5 royal flushes in a row. It’s pretty clear that the probability of that is much higher given that I’m cheating than given that I’m not cheating (and that’s clear intuitively, even if you have no idea of how to do the relevant math). But suppose you accused me of cheating, and when I asked you to justify that, you showed me how unlikely it is that those hands would come up by chance, and then observed that it’s much more likely that they would come up if I’m cheating (even though, incidentally, it’s probably not *that* likely even if I am cheating). It’d be absurd for me to say “Ah, but you’re conflating two different types of probability–if you list all the possible combinations of ways that those hands could have turned up given chance and put that under a 1, then I can list all the combinations of ways they could have turned up given that I cheated, and look–it’s just as many!”Looking back, those points weren’t really brief. But then, your points weren’t really 2, so I suppose we’re even. :-)  

The Poker Analogy doesn’t work.

1. The Fine-Tuning Argument seeks to establish that there is a personal agent adjusting the values of physics and life to generate life and this argument aside, we don’t have any evidence of what exactly this mysterious entity wants and we have no direct experience with it. On the other hand, with the poker analogy, we already know we’re dealing with a personal agent that we know quite a bit about and can thus predict that he may cheat. In other words, it’s two completely different scenarios. If we’re trying to establish that we’re playing poker with a personal agent, simply because the opposing player gets 5 Royal Flushes in a row doesn’t mean that it is an intelligent being cheating with the deck. In fact, we wouldn’t even know if the player even has any desire to cheat.

2.The analogy over-states what’s involved with Fine-Tuning. Yep, what we’re discussing the is the probability of getting one royal flush because, according to the argument, the whole universe is fine-tuned to life, at least to some extent. As such, it counts as ONE hand and we can’t accuse anyone of “cheating” for that.

3. As was pointed out by rosyatrandom, we already know the mechanics of the game of poker, but certainly not all the ins-and-outs of the universe and why we live in the one we do.

  (Quote)

ayer February 4, 2011 at 9:26 am

2.The analogy over-states what’s involved with Fine-Tuning. Yep, what we’re discussing the is the probability of getting one royal flush because, according to the argument, the whole universe is fine-tuned to life, at least to some extent. As such, it counts as ONE hand and we can’t accuse anyone of “cheating” for that.

Have you seen the precision required of the constants and quantities for fine-tuning for life? The odds are astronomically higher against such precision being generated by chance than the odds of getting a single royal flush. E.g., for just one of the constants:

“Another interesting example of a finely-tuned initial condition is the critical density of the universe. In order to evolve in a life-sustaining manner, the universe must have maintained an extremely precise overall density. The precision of density must have been so great that a change of one part in 1015 (i.e. 0.0000000000001%) would have resulted in a collapse, or big crunch, occurring far too early for life to have developed, or there would have been an expansion so rapid that no stars, galaxies or life could have formed.9 This degree of precision would be like a blindfolded man choosing a single lucky penny in a pile large enough to pay off the United States’ national debt.”
http://biologos.org/questions/fine-tuning/

  (Quote)

rosyatrandom (Michael R.) February 4, 2011 at 9:53 am

Either there’s enough blind men to pick up every penny, or there is an unknown reason for the pile to be smaller. Neither of these gives me even a twinge of philosophical crisis.

I shall add something else to the debate: say the universe _was_ created by a god. Now, it could have been created a few seconds ago and made to appear old. For parsimony’s sake, we choose the latter. It’s more elegant for the apparent processes underlying the cosmos to be traced backwards and for its evolution to be natural and unforced. But god didn’t have to have just created the universe; it could have also created a universe-spawning multiverse. Once again, it’s simpler and more parsimonious. So… why not? And if we keep doing this, what do we leave god to do? This is why god just seems arbitrary and superfluous to me.

  (Quote)

rosyatrandom (Michael R.) February 4, 2011 at 9:54 am

Oh dear, looks like today is typo day.

“Now, it could have been created a few seconds ago and made to appear old. For parsimony’s sake, we choose the latter.”

should have been

“Now, it could have been created a few seconds ago and made to appear old, or it actually is old. For parsimony’s sake, we choose the latter”

  (Quote)

Steven R. February 4, 2011 at 9:55 am

Have you seen the precision required of the constants and quantities for fine-tuning for life? The odds are astronomically higher against such precision being generated by chance than the odds of getting a single royal flush. E.g., for just one of the constants:“Another interesting example of a finely-tuned initial condition is the critical density of the universe. In order to evolve in a life-sustaining manner, the universe must have maintained an extremely precise overall density. The precision of density must have been so great that a change of one part in 1015 (i.e. 0.0000000000001%) would have resulted in a collapse, or big crunch, occurring far too early for life to have developed, or there would have been an expansion so rapid that no stars, galaxies or life could have formed.9 This degree of precision would be like a blindfolded man choosing a single lucky penny in a pile large enough to pay off the United States’ national debt.”http://biologos.org/questions/fine-tuning/  (Quote)

I fail to be impressed and you missed the point. No matter how low the probability of getting a particular hand is, the point remains that getting a good hand once is hardly indicative of cheating, even if chances are “astronomically low” of getting that hand because, after all, you had to get some result. The Poker Game analogy makes it seem as if we’re talking about various factors (presented as getting 5 Flushes in a row) when in reality, it is just one. Suppose the blind-folded man were to get the lucky penny. Is that indication of god’s existence beecause the chances of him getting that particular penny if god wanted it are 100%?

Not only that, but the only reason it’s noteworthy that the man picked the lucky penny is because we previously established that this was a good thing. On the other hand, the F-T Argument works like having a man randomly pick the penny and then declaring that penny the “lucky penny” and saying that the chances of getting this particular penny requires a particular explanation, even when other than our own desire to call this penny special there is no real rason to do so. And then Troy’s and Brian_G’s argument would be “well, we’re intelligent agents and we all felt like calling this the lucky penny, surely, this is reason enough to postit that god made the blind-folded man pick that penny which we later thought to call the lucky penny!”

Jeez, the more I think about this argument the more flawed it gets. Not even as a Theist did this argument come close to being convincing and now it’s just downright awful.

  (Quote)

Steven R. February 4, 2011 at 9:58 am

Oh, and I should add, if we’re projecting human desires to say what God will want, then one human desire is to MAXIMIZE the amount of intelligent life that there can be. So the argument that an intelligent agent will want to create other intelligent agents based on human desires also fails because said agent, based on human behavior, would also want to fine-tune the whole universe to have the most intelligent beings possible. Oh, and as a human, I’d also want to protect the intelligent life, which this God has failed to do (come on, putting us at risk of an asteroid collision? Hardly a human desire).

  (Quote)

Eric February 4, 2011 at 11:00 am

ayer –
You are completely failing to grasp the difference between “composite” and “non-composite.” I don’t know if you are a computer programmer, but your response regarding “coding” mind shows a fundamental misunderstanding of the concept of “mind” in the field of philosophy of mind. You have to break out of your materialist presuppositions to deal with the concept properly. You need to show how the naturalistic multiverse is non-composite in order for the multiverse to prevail under Occam’s razor, and since, as you admit, it would have to be composed of trillions upon trillions of quarks, it is composite in excelsis by definition.

The point of this kind of reduction is that these are fundamentally NOT COMPOSITE since they are made of the SAME thing. At its most fundamental level, its not composite. It sounds to me like you are playing word games, which is common when people say “break out of your materialist presuppositions to deal with the concept properly.” The coding example is not materialist, its an abstraction. I’m giving you a way to think about this that doesn’t just involve playing word games. I’m giving you a odels. You are giving me vague assertions.

ayer –
A mind of a personal agent, on the other hand, can produce complex thoughts, but those thoughts are not constituent “parts” of the mind; it can simply stop thinking those thoughts and think something else. The pure mind itself is non-composite and thus the very essence of simplicity.

Once again, I did not specify thoughts, but how the persona agent deals with thoughts. Its like a processor. Even if no program exists, the processor still exists. And the mind processor is incredibly complex. And that was actually quite generous of me not to include thoughts because, if these thoughts don’t directly result from a few simple facts, then they themselves are composite in excelsis. As long as these thoughts EVER EXIST, then you can’t just ignore them. Think of software still existing on a hard drive even if the hard drive is not turning.

ayer –
Have you seen the precision required of the constants and quantities for fine-tuning for life? The odds are astronomically higher against such precision being generated by chance than the odds of getting a single royal flush. E.g., for just one of the constants:

Its not about how unlikely it is to get 3 royal flushes in a row. its about playing mind games using analogies.
Take, for instance, a tank with 1 Million balls, each with a different number. You have to draw a ball with the number 1 on it to win. You blindly reach into the tank once and draw the number. OMG Its the correct number and you’ve won! Now you may have cheated, you had a 1 in a million chance to draw the right ball, but it may be hard to someone to be THAT sure.
Now lets take a different tank, one with 100 balls, each with a different number. You have to draw a ball with the number 1 on it to win. You blindly reach into the tank once and draw the number. OMG Its the correct number and you’ve won! But the game show host gave you another chance to double your winnings or lose it all by taking another blind plunge into the tank. OMG! Its the correct number and you won again. So he tries one more time. Double or nothing. OMG! you did it again. (Note that every time you did this, the winning ball was shuffled back into the tank). Does it look any more likely that you cheated? Its still a 1 in 100*100*100 = 1 in a million chance. But you had something unlikely happen to you MULTIPLE TIMES IN A ROW.
Now, if we had a tank of 10 balls and you drew the winning ball six times in a row. I think you can see where this analogy is going.
Its a mind trick. The odds are the same in all 3 cases, yet it would seem to the casual observer that the later analogies are bigger suspects of cheating than the first one. If you really want to give a good analogy (even if the odds are all that bad ::cough normalization problem cough::), then you make a gigantic deck, but with only 1 of every card needed to make a royal flush (Of course, you are still begging the question of intrinsic value)

  (Quote)

ayer February 4, 2011 at 11:21 am

So, for me, a single universe in isolation is highly complex because it lacks context – we would have to generate everything by specifying highly arbitrary contingencies for it all. In short, it does not make sense to me. But the multiverse… that is far simpler! You see a bunch of additional entities that each require as much explanation as the first, but we see a single entity that requires less.

Well, at least we agree that our single universe’s “highly arbitrary contingencies” cry out for an explanation–mine is theism, yours is the multiverse; we disagree on which explanation holds up best under Occam’s razor. But it seems there is a gulf between our position and that of Steven R. and those in his camp, who fails to see the need for an explanation at all:

No matter how low the probability of getting a particular hand is, the point remains that getting a good hand once is hardly indicative of cheating, even if chances are “astronomically low” of getting that hand because, after all, you had to get some result.

At least the multiverse theorists are making an attempt, instead doing what appears to me to be the equivalent of sticking their fingers in their ears and chanting “what problem?”

  (Quote)

Larkus February 4, 2011 at 12:03 pm

Concerning the poker analogy:

Some problems:

1) We have exactly one universe and no other universes to compare. As StevenR. already pointed out, the correct analogy would be, that we get exactly one hand of cards and no other hands of cards to compare.
2) We don’t know the composition of the deck.
3) We don’t know the identity of the card dealer (who could be anyone, even an alien, a super-powerful intelligent being, a bunch of monkeys or a perfect shuffling machine).
4) We don’t know the intentions of the card dealer. (we don’t know, how likely the card dealer is to manipulate the game).
5) We don’t know, which – if any – game is played. (life may or may not be a ‘winning combination’ in our universe)

  (Quote)

ayer February 4, 2011 at 12:29 pm

1) We have exactly one universe and no other universes to compare. As StevenR. already pointed out, the correct analogy would be, that we get exactly one hand of cards and no other hands of cards to compare.

We know the number of possible universes with different quantities and constants:
“In a new study, Stanford physicists Andrei Linde and Vitaly Vanchurin have calculated the number of all possible universes, coming up with an answer of 10^10^16. If that number sounds large, the scientists explain that it would have been even more humongous, except that we observers are limited in our ability to distinguish more universes; otherwise, there could be as many as 10^10^10^7 universes.”
http://www.physorg.com/news174921612.html

2) We don’t know the composition of the deck.

See above.

3) We don’t know the identity of the card dealer

We know it is an intelligent agent

4) We don’t know the intentions of the card dealer.

We know the preference for life over non-life based on intelligent agents with which we have experience.

5) We don’t know, which – if any – game is played. (life may or may not be a ‘winning combination’ in our universe)

See response to 4), above.

  (Quote)

Larkus February 4, 2011 at 12:54 pm

@ayer:

What relevance does the article have, that you cited. Do you now support the multiverse theory?

Concerning the rest:
You are making bare assertions.

  (Quote)

Eric February 4, 2011 at 1:05 pm

ayer –
At least the multiverse theorists are making an attempt, instead doing what appears to me to be the equivalent of sticking their fingers in their ears and chanting “what problem?”

As far as I can tell, our ears are open. You just have not told us what the problem is. As I explained to Al Moritz in a previous post, chance is not a-priori a bad explanation for low probability phenomena (despite how low it is). In fact there are many phenomena for which chance is an excellent explanation. If an arbitrary person wins the lottery, chance is an excellent explanation. If you draw an arbitrary hand in poker, as I explained earlier, chance is an excellent explanation. So for arbitrary phenomena, chance can be an excellent explanation. If life is an arbitrary phenomenon, then chance can be excellent explanation. That’s why we are asking you to show us why chance is NOT an arbitrary phenomena before we rule out chance. How is life similar to drawing a royal flush and not to drawing the random hand mentioned earlier? To prepare for what tends to be the usual response, lets pretend every possible hand you can draw has a name, and thus, independently specified parameters. I’ll call the hand from earlier “the dukes of hazzard.”

ayer –
We know the number of possible universes with different quantities and constants:
“In a new study, Stanford physicists Andrei Linde and Vitaly Vanchurin have calculated the number of all possible universes, coming up with an answer of 10^10^16. If that number sounds large, the scientists explain that it would have been even more humongous, except that we observers are limited in our ability to distinguish more universes; otherwise, there could be as many as 10^10^10^7 universes.”

Yes, but this is under an inflationary model of the universe, which itself would explain perceived fine tuning. This doesn’t apply to the single universe hypothesis.

ayer –
3) We don’t know the identity of the card dealer

We know it is an intelligent agent
4) We don’t know the intentions of the card dealer.

We know the preference for life over non-life based on intelligent agents with which we have experience.

5) We don’t know, which – if any – game is played. (life may or may not be a ‘winning combination’ in our universe)

See response to 4), above.

This is quite circular. How do you know it is an intelligent agent without assuming life has intrinsic value?

  (Quote)

ayer February 4, 2011 at 1:20 pm

What relevance does the article have, that you cited. Do you now support the multiverse theory?

Supporters of theistic fine-tuning don’t deny an overwhelming number of POSSIBLE universes (indeed, that is what makes the fact that the actual universe is fine-tuned for life so remarkable). What distinguishes the multiverse theorists is that they argue that all of those possible universes actually exist (even though that view is nonfalsifiable).

The argument that would defeat both the theistic and the multiverse view would be to establish that the constants and quantities in the initial conditions of the Big Bang are not arbitrary but are required by some “law of nature” and could not have been otherwise (not by asserting the absurd proposition that “there is nothing that needs to be explained”). It is the fact that there IS no “law” requiring the constants to have the values they have in our universe that drives atheistic theorists into the arms of the multiverse (no matter how bizarre it seems).

  (Quote)

ayer February 4, 2011 at 1:27 pm

How do you know it is an intelligent agent without assuming life has intrinsic value?

Because of our background knowledge (based on our own experience) of what intelligent agents prefer (I don’t see that “intrinsic” has anything to do with this particular argument, though I believe life has intrinsic value based on the existence of objective moral values grounded in God’s nature as that being greater than which none can be conceived–but that is a different discussion).

  (Quote)

Larkus February 4, 2011 at 1:39 pm

So it seems, you agree after all, that “we have only one universe”. The analogy would be one hand of cards. Of course there are other possible universes just like there are other possible hands of cards. But we get only one of those possible hands of cards.

Concerning whether we know the composition of the deck, from the article you cited:

As the scientists explain, the calculation of the number of universes is an important step toward an even larger goal: to find the probability of living in a universe with a particular set of properties. What are the chances that we live in a world in which the laws of physics are these laws that we currently observe? [emphasis mine] Answering this question requires finding probabilities that depend on knowing about other universes, among many other challenges.

Obviously the answer to the question is still open, according to the article.

  (Quote)

Larkus February 4, 2011 at 1:53 pm

@ayer

Eric: “How do you know it is an intelligent agent without assuming life has intrinsic value?”

ayer: “Because of our background knowledge (based on our own experience) of what intelligent agents prefer.”

From the Fallacy Files: Hasty Generalization

Exposition:

This is the fallacy of generalizing about a population [in your case the population of all possible kinds of intelligences] based upon a sample which is too small to be representative. If the population is heterogeneous, then the sample needs to be large enough to represent the population’s variability. With a completely homogeneous population, a sample of one is sufficiently large, so it is impossible to put an absolute lower limit on sample size. Rather, sample size depends directly upon the variability of the population: the more heterogeneous a population, the larger the sample required. For instance, people tend to be quite variable in their political opinions, so that public opinion polls need fairly large samples to be accurate.

  (Quote)

Eric February 4, 2011 at 1:59 pm

ayer –
(not by asserting the absurd proposition that “there is nothing that needs to be explained”)

You keep asserting this but do not back it up. Do you want to have a discussion or just keep throwing the same assertions around we’ve already responded to… Now I will say that, from a contingency point of view, we should always attempt to explain the universe. Specifically though, the reason why OUR PARTICULAR UNIVERSE exists is not necessary if it is arbitrary. Of course, if the explanation of our universe does show that it is necessarily the only kind of universe, or one of a small set of physically possible universes, then it solves a “problem” that doesn’t seem to actually be a problem. I have, in previous posts on this site, commented on the difference between the logical need for an explanation and the scientists desire for explanation.


Because of our background knowledge (based on our own experience) of what intelligent agents prefer (I don’t see that “intrinsic” has anything to do with this particular argument, though I believe life has intrinsic value based on the existence of objective moral values grounded in God’s nature as that being greater than which none can be conceived–but that is a different discussion).

So I see a person who won the lottery. I know this person wanted to win the lottery, so I assume he cheated? Maybe you need to produce some kind of syllogism cuz I cannot find a logically coherent argument in your statement for why an intelligent agent is any better than chance. All arguments I can think of contain:
If A then B
B
Therefore A
which is a non-sequitur.
Note: But the existence of intrinsic value based on objective moral values is a rather weak argument, as we cannot be close to sure objective moral values exist, only that we perceive a set of shared values, which, as WLC has said over and over in his debates, is explainable via evolution. He just claims we have objective moral values and not perceived shard values, but doesn’t seem to back it up beyond heuristic intuition.

@Larkus, I wanted to point that out, but I thought it didn’t mattered since the article only applies to the inflationary model, which itself already solves the “problem” of fine-tuning.

  (Quote)

rosyatrandom (Michael R.) February 4, 2011 at 3:32 pm

I get the impression it would be handy to have a whole new discussion as regards what I might call the ‘existential hurdle’:

What does it take for something to exist? Is it difficult, or easy? Should our default position be positive or negative?

In short, what is existence, what kinds of things can exist, and of those which do?

I used to be a physical realist, which I think is the default position for atheists — the world is real (in a ‘tap a hard object’ kind of way) and what you saw is what you got. This is much the same kind of view theists generally have, too, though with no confusion over souls and dualism. You have your universe, it exists, and anything else must be proven. Certainly, alternate universes are considered foolish sci-fi nonsense that is unjustified. In other words, existence is hard.

For me, though, as rational and common-sensical as that position is… it’s one I found started to fall apart the more I thought about things. Just like theism has these mysterious properties — Free Will, Souls, God… — more and more it seemed like Existence was another fuzzy concept that things just had, but without explanation or meaning of any kind. And I started reading Egan, and things clicked.

I no longer saw existence as something difficult, something unusual, that had to be bought about through some unknown process, but I came to see reality as timeless (time being an internal aspect of the universe, after all) and uncontingent upon any kind of fiat. It became clear that if something could exist, it did. It was a tautology, really. So the question is, yet again, “what can exist?”

I have been heavily persuaded by the Tegmark’s Mathematical Universe Hypothesis (“[his] sole postulate is: All structures that exist mathematically also exist physically“) and Greg Egan’s Dust Theory (anything you can represent abstractly in terms of another thing is just as real, which essentially means that all possible things exist within anything).

What we have here, then, is a framework that outlines the borders of existence — all possible (mathematical) things, and a generative process that can produce it from nothing (abstract representation). For me, this is a pretty complete metaphysical system. Oh, there are plenty more things to discover, and implications not yet fleshed out (I hope!), because we still do not know so much about how our minds work, which is a crucial aspect with massively significant entailments.

  (Quote)

Steven R. February 4, 2011 at 3:54 pm

Ayer said…”We know it’s an intelligent agent”

Uh, unless you’re referring tot he Poker Analogy (and that’s not what Larkus meant when he said card-dealer), in which case your comment is irrelevant, your argument is now circular. You’re trying to prove that an intelligent agent is involved, you can’t just assert that it is an intelligent agent just like that.

  (Quote)

ayer February 5, 2011 at 8:02 pm

Just viewed this interesting debate that took place about 10 days ago in Canada between William Lane Craig and George Williamson. In their discussion of the fine-tuning argument, they touch on several of the points made in this thread by each side:

http://saskskeptics.com/2011/02/04/video-william-lane-craig-vs-george-williamson-debate/

  (Quote)

Luke Muehlhauser February 5, 2011 at 9:43 pm

ayer,

Thanks for the link!

  (Quote)

rosyatrandom (Michael R.) February 5, 2011 at 11:58 pm

Oh man, I don’t have the patience to watch videos. I just get irked there’s no transcript I can skim, with perhaps some comments so I see what the important points were meant to be. That’s the semi-digested way I like to take my information in….

  (Quote)

Rosita February 6, 2011 at 4:20 pm

The whole fine tuning argument ignores the fact that the universe is much more “finely tuned” for the death of any life that arises to fit a niche somewhere.

  (Quote)

Rosita February 6, 2011 at 4:26 pm

Life on earth (or anywhere else) has a precarious, tenuous and temporary existence. The universe will eventually kill it.

If the universe was, indeed, willfuly fine-tuned to support reproducing intelligent self-ware life on earth like planets then where is the morality and intelligence in arranging the rest of the universe to be antithetical to this?

  (Quote)

Brian G February 8, 2011 at 10:09 am

Life on earth (or anywhere else) has a precarious, tenuous and temporary existence.The universe will eventually kill it.If the universe was, indeed, willfuly fine-tuned to support reproducing intelligent self-ware life on earth like planets then where is the morality and intelligence in arranging the rest of the universe to be antithetical to this?  

I think this is an interesting point which fits nicely with what I said earlier. We wouldn’t expect an intelligent agent to create just any sort of universe. There are specific things we could observe that would count against design. Design is clearly testable. The question is whether it has already failed the test. And this I think we’ll need to leave to a discussion at another time.

  (Quote)

Larkus February 9, 2011 at 12:59 am

@Brian G:

We wounldn’t expect an intelligent agent like us to create just any sort of universe, if he creates the universe. But what is the probability that a “super-intelligent, super-powerful conscious [agent] with the capacity to create the universe” and with unspecified desires, intentions, etc. is an intelligent agent like us.

An estimate would be a start.

Remember, we are talking about the hypothesis T. As Troy put it earlier in the thread: “It’s not that specific of a hypothesis at all; it just says that there exists a super intelligent, super-powerful conscious being with the capacity to create the universe. Nothing in the hypothesis specifies his desires, intentions, etc. [emphasis mine]; [...]”

If you want to amend T by claiming, that a super intelligent, super-powerful conscious being with the capacity to create the universe has some specific desires, intentions, etc. then you have to tell, how likely it is, that of all possible desires, intentions, etc. this being has these and no others.

  (Quote)

Troy February 9, 2011 at 9:33 am

While I think ayer has done a perfectly good job of addressing the criticisms of the poker analogy, I’d like to note that I never made that analogy with the intent of saying that it was exactly like fine-tuning in all relevant respects. I’d already made my argument in favor of T over ASU at that point; I was only using the analogy to illustrate what was *wrong* with a specific *objection* to the FTA—i.e., that it utilized two different kinds of probability. I then expanded it to indirectly address another objection, that God would probably have created a different kind of life-filled universe than this one. My point there was just that even if this is true, it doesn’t render the posterior probability of T nearly as low as some seem to think, just as noting that if I was cheating I’d probably have been more discrete about it (true though that may be) leaves the posterior probability of cheating high enough for you to continue sticking with that hypothesis.

It looks to me like much of what’s been said I’ve already addressed, but I’ll try to respond to some new things and to points that are continually coming up:

So, first, rosyatrandom (Feb 2, 7:59 pm) observes that with the poker game we have some way to get a grasp of the prior probability of cheating. Certainly true, although I would note that we’d never expect someone to be able to give an exact prior probability, and that we don’t think they’d need to be able to do so for the inference to be rational. Similarly, the prior probability of theism does not seem to me to be anywhere near the order of 1 in 10^100 (or however unlikely the life-supporting constants arising by chance is), although I couldn’t give you an exact prior. Of course, one might argue that we’re justified in giving cheating a sufficiently high prior only because we have previous experience of it, but I’m skeptical that that move can be pulled indefinitely, because at some point we need to move from knowledge of our sense experiences to an inference about the external world (there is a world of mind-independent objects and I am not being deceived by an evil demon, etc.). In that case one can’t appeal to prior experiences of a mind-independent world because that’s exactly what’s at issue. But it seems clear that the prior probability of a mind-independent world is high enough that the inference is valid; and when the prior for God only needs to be many orders of magnitude above something like 1 in 10^100, it seems clear to me that it’s high enough for the argument to work (assuming, as I have been this whole time, that what Collins calls the Atheist Single Universe hypothesis is the only alternative—I actually think the multiverse hypothesis is a live option, but I’ve been bracketing that).

(In that discussion of reasoning from our experiences to an external world, I am assuming some sort of strong foundationalist internalism here, but I’m guessing I’ll get less flak for that here than I would in most philosophy departments.)

Eric says (Feb 2, 10:53 pm):

Also, its not just the prior probability of Theism, its the prior probability of Theism where we would have a god that WOULD PROBABLY CREATE THIS UNIVERSE. Else the problem is that a God could create ANY logically possible universe, while only metaphysically possible universes could exist under naturalism. So Pr (FT/T)<< Pr (FT/ASU), which defeats the entire point of this argument.  

As I’ve emphasized multiple times, God’s being (sufficiently) likely to create this universe is not built into the God hypothesis as I’ve presented it. It’s inferred after by analogy. I don’t see why it should be built into the prior anymore than I’d have to build “wanting to build structures like the ones we find here” into the prior that alien life exists on this planet on which we’ve just found what look like buildings. The prior’s there already; we reason that aliens would be sufficiently likely to build such structures by analogy with our experience of humans.

@Patrick (Feb 3, 10:03 am): I don’t disagree with anything you say about having the same standards in different situations (I’m quite skeptical of the skeptical theist response to the problem of evil, for instance), and following the proper rules for Bayesian reasoning. I don’t think it’s as likely as you do that God would want to create a universe filled to the brim with life, but we’ve been over that. The main point I want to make in response is that I don’t think you’ve really answered my question. What I suggested was that denying the validity of any analogical reasoning from humans to God would make P(E|T), where E is any evidence and T is theism, inscrutable. And this would make any evidential argument impossible, or at least any Bayesian argument, which from my perspective is the only game in town. The reason is that with having P(E|T) you can’t say if P(E|T) > P(E|~T), because P(E|T) is inscrutable. And without that you can’t reason that P(T|E) > P(~T|E). So it doesn’t matter how apparently miraculous of an event you have as evidence; no proper inference to the existence of God can ever be made. Do you accept this as an implication of your position? If not, what’s wrong with my reasoning here?

  (Quote)

Rosita February 9, 2011 at 9:42 am

@Brian G: But what is the probability that a “super-intelligent, super-powerful conscious [agent] with the capacity to create the universe” and with unspecified desires, intentions, etc.is an intelligent agent like us.An estimate would be a start.

Everything that we know about the universe points to the conclusion that super complex things (including minds, intelligence and personalities) take eons to develop. The chance of something like that existing or blinking into existence in zero time is effectively zilch.

The sensible hypothesis is that the the universe began with something incredibly simple, such as a random fluctuation in potential energy like those observed in the realm of quantum physics, and gained increasing complexity from that point.

In other words, the theo-sophist arguments that posit causal complexity are based on scientific ignorance. It is time they were dismissed for that reason.

  (Quote)

rosyatrandom (Michael R.) February 9, 2011 at 9:55 am

Hmmm, this emphasis on the likelihood of god’s existence (or not), and of god’s preferences if it does exist, is prompting me to make some observations.

Say I specifed a very large number u, perhaps through an algorithm to derive it, then asked the question, ‘what is the probability that is u prime?’

Well, at first glance, we don’t know until we calculate it. Perhaps we could give a probability based around u‘s estimated size… but that isn’t the real answer. The real answer is 0 or 1, depending. Our ignorance about u doesn’t change whether it is prime or not, since it has been defined precisely already.

What is the probability that god exists? And if a god exists, what is the probability that it is a particular one with particular preferences?

If you’re in the existence-is-hard camp, which most theists and atheists are, you’re going to be reducing the question down to some kind of logical/aesthetic/empirical basis on what things should exist and what things shouldn’t. While I am not in that camp, I would be particularly hard-pressed to take the theistic side seriously. I find the whole notion of god and creation to be arbitrary and unexplained.

I’m in the existence-is-easy camp; if there’s no reason why a particular god should not exist, I’m OK with it. But I am an atheist because I see all these gods as, again, arbitrary. The cancel each other out and are inelegant, superfluous additions to any metaphysical framework.

  (Quote)

rosyatrandom (Michael R.) February 9, 2011 at 10:00 am

Everything that we know about the universe points to the conclusion that super complex things (including minds, intelligence and personalities) take eons to develop. The chance of something like that existing or blinking into existence in zero time is effectively zilch.The sensible hypothesis is that the the universe began with something incredibly simple, such as a random fluctuation in potential energy like those observed in the realm of quantum physics, and gained increasing complexity from that point.
In other words, the theo-sophist arguments that posit causal complexity are based on scientific ignorance. It is time they were dismissed for that reason.  

From a block-universe standpoint, the increase in complexity of the universe is merely something that happens from a particular perspective on it. The whole universe itself is atemporal and pretty much just sitting there. My view is that the best way to resolve this is to always generalise such that complexities are folded into a simpler framework. The easiest analogy is that any particular book is complicated and unlikely, but can be found in the library of all possible books, which is completely un-complicated.

  (Quote)

Rosita February 9, 2011 at 10:21 am

The problem with that line of thinking is that it assumes that each item in the library of all-possible-things is equi-probable. There is no evidence for that, and plenty to undermine it.

  (Quote)

Brian_G February 9, 2011 at 10:22 am

@Brian G:We wounldn’t expect an intelligent agent like us to create just any sort of universe, if he creates the universe. But what is the probability that a “super-intelligent, super-powerful conscious [agent] with the capacity to create the universe” and with unspecified desires, intentions, etc.is an intelligent agent like us.An estimate would be a start.Remember, we are talking about the hypothesis T. As Troy put it earlier in the thread: “It’s not that specific of a hypothesis at all; it just says that there exists a super intelligent, super-powerful conscious being with the capacity to create the universe. Nothing in the hypothesis specifies his desires, intentions, etc. [emphasis mine]; [...]”
If you want to amend T by claiming, that a super intelligent, super-powerful conscious being with the capacity to create the universe has some specific desires, intentions, etc. then you have to tell, how likely it is, that of all possible desires, intentions, etc. this being has these and no others.  

We don’t need to propose an intelligent being that is like us in all respects, all we need is one that acts in ways distinguishable from chance. This is a perfectly natural assumption which conforms with experience. We would require this assumption for any detection of intelligent agency. Suppose SETI finds an a signal in one of their radio telescopes. ( Currently they’re only looking for carrier signals, but if they found a carrier signal, the next step would be to build a more powerful radio telescope to try to decode the signal. ) Suppose the signal came back as a sequence of prime numbers, or a list of consents like the ratio of the proton to electron mass. That would be strong evidence that there was an intelligent agent sending the signal. Keep in mind that we don’t know anything about these agents. They could have a different evolutionary history. They could have completely different brain chemistry. There may even be a very strange form of intelligence that can come about without going through an evolutionary process. So how do we know that an intelligent agent would prefer to send a signal of prime numbers or consents of the universe? We don’t. But we would expect an intelligent agent to be distinguishable from chance. If we think that an unknown intelligent agent is equally likely to send any combination of one’s and zero’s as any other combination, then we have no way of distinguishing an unknown intelligent agent from chance. Someone could use this approach to reject extraterrestrial intelligence. But notice this approach doesn’t explain why we got a signal meaningful to us. It doesn’t make our observation more likely. All it does is try to argue that there is no reason to expect intelligence to differ from chance. I find this to be extraordinarily implausible.

With the fine-tuning argument, we don’t need to propose that it’s highly probably that God would create a universe roughly similar to this one. It may be that there are a lot of universes God might prefer equally. It could be that the probability of God creating a universe roughly like this one is less then one in a trillion. However, when it’s even much more improbable given chance, then the fine tuning still supports theism. The probability of a low entropy universe on chance is said to be 1 in 10^10^123. If that’s even in the ballpark, the probability of God creating a universe like this one can be extraordinarily low and still beat chance by a huge margin.

  (Quote)

Troy February 9, 2011 at 10:31 am

I had a very long post typed up in Word of which the above is only the first part; but every time I try to paste it in my browser freezes up. :-( Since I don’t have time to keep on trying this indefinitely, I’m going to give up for now, but I may try again tonight.

  (Quote)

rosyatrandom (Michael R.) February 9, 2011 at 10:32 am

The problem with that line of thinking is that it assumes that each item in the library of all-possible-things is equi-probable.There is no evidence for that, and plenty to undermine it.  

But it is easier to generate the library than any particular book. It could scarcely be easier. And once you generate the library, hey presto – there’s the book.

When it comes to the metaphysics of existence and likelihood, what do you think is harder to explain? A single universe, all by itself, or all possible universes? If the former… why? Consider all possible universes generatively as opposed to artificially splitting it up and looking at everything in isolation.

  (Quote)

Troy February 9, 2011 at 10:44 am

Larkus has asked multiple times how likely it is that the designer is like us, and my answer remains the same: It’s fairly unlikely. I think it’d be much too strong to claim, say, that the probability of God’s having desires and intentions like ours is above 0.5. (If that is what I was saying, then an accusation of hasty generalization might be warranted.) And I’d say the same about reasoning about the desires and intentions of a completely alien life form. But again, the probability doesn’t need to be high for the argument to go through. It just needs to be many orders of magnitude above P(L|ASU). In your most recent post you ask for an estimate. If 1/1^10^123 < P(L|T) < 0.5 is an estimate, then I’m happy to oblige, but if you want something more exact than that, I’m afraid my credences just aren’t that precise. But I couldn’t give you an exact probability of five royal flushes given that you’re cheating either, even though I’m exceedingly confident that it’s many orders of magnitudes above the probability of five royal flushes given chance (and substantially below 0.5 too, I should think). Nor could I give you an exact probability of buildings on another world given that intelligent aliens live there, but it’s still clearly many orders of magnitude above the probability of that happening by chance (I use this example because my probability estimate here is again done by analogy with beings that may very well be completely unlike us, but that doesn’t render it so low as to defeat the inference from buildings to intelligent aliens).

Finally, Steven R. and others have continually claimed that humans have something like a desire to maximize intelligent life, and so that we would expect an intelligent designer to create a universe brimming with life. Now, I’ve already noted that even if this is true it doesn’t defeat the FTA, any more than noting that if I were cheating I’d probably have been more coy about it defeats the hypothesis that I’m cheating. But I am curious as to why so many people seem to think that we even have reason to expect this. This may just turn into intuition-mongering, but this does not strike me as at all a plausible claim. I don’t think I have a desire to maximize intelligent life. Am I really that odd in this?

In addition, it occurs to me that it would be possible to maximize intelligent life without just cramming more into the universe. You could just do it by having the universe continue indefinitely and having intelligent life continually coming into existence. I have no idea whether this is what God would do, especially not if we’re building as few assumptions into the God hypothesis as I am in my reconstruction of the FTA, but it would seem to accomplish this purpose so many seem to think God would have just as well.

  (Quote)

Troy February 9, 2011 at 10:46 am

Well, there we go: I managed to get it posted after all. It looks to me now like the system was freezing up whenever I typed or pasted one < after another, so I took the second one out right before P(L|T). I have no idea why that was causing trouble, but it was rather annoying.

  (Quote)

Troy February 9, 2011 at 10:53 am

The SETI analogy is a good one, Brian.

Say I specifed a very large number u, perhaps through an algorithm to derive it, then asked the question, ‘what is the probability that is u prime?’

Well, at first glance, we don’t know until we calculate it. Perhaps we could give a probability based around u‘s estimated size… but that isn’t the real answer. The real answer is 0 or 1, depending. Our ignorance about u doesn’t change whether it is prime or not, since it has been defined precisely already.

But you can say that about anything. We need subjective epistemic probabilities because we’re not omniscient. It’s not clear to me that this is even a problem with necessary truths, like math (I’m not sure if you were making the analogy because God is sometimes claimed to be a necessary being, but I’ve heard others make the analogy for that reason). We can speak perfectly sensibly about the probability of a logical or mathematical proposition being true, from our perspective. For example: we have good inductive evidence that Goldbach’s conjecture is true. Now, if it is true it’s a necessary truth, but at the moment the rational thing to do, it seems to me, is to be confident that it’s true (because of the inductive evidence) but not certain (because we can’t actually prove it).

  (Quote)

rosyatrandom (Michael R.) February 9, 2011 at 11:07 am

Time is an internal facet of the cosmos; I don’t think any metaphysical theory can be taken seriously if it has time outside of existence itself. (If it did, we’d have a timeline of things that exist, which can be summed into a new timeless consideration.)

So we have a timeless reality. Already, this relegates any act of creation to a mere evolution-within-a-system than anything more profound, and makes anything that exists necessary. The universe, as it seems by itself, is hard to postulate as being necessary without also granting necessity to its neighbouring possible universes (in fact, I would say that Possible and Necessary are the same thing, fundamentally), and I see no reason to withhold that necessity from all the structures surrounding it. Fully extended, I don’t have any problems with the necessity of the universe.

But god? No… I don’t see it. God is not part of any system that can extend to contain it and make it necessary, nor based on any principles that cover it sufficiently and exclusively. And besides – the universe exists, and so is already necessary. What need for some kind of intentionality to ‘choose’ it?

  (Quote)

Rosita February 9, 2011 at 1:15 pm

Since we have no knowledge of any other universes there is no valid means of calculating the likelihood of another one or of doing anything but speculating on what another “possible” universe might be like or how many there could be.

The problem seems to reduce down to something akin to the empty speculation of the sum of all possible Beethoven’s. It’s a mind game with no substance in reality.

As for the SETI analogy, has anyone considered that humans might be the first or most intelligent form of life, at least by human definition? Something has to be first and “best”. Why not us? Why do humans have such trouble considering that they might be the only ones that are so far capable of seeding life on other planets and thereby “creating” a new race of humanoids, species of animals or forms of plant life elsewhere? Why are humans so reluctant to be “gods” and stand at the top of the tree? Is there an innate drive to be in the subservient role of the child for ever? Or is that just a personality trait that is common among those who gravitate to authoritarian type religions?

  (Quote)

rosyatrandom (Michael R.) February 9, 2011 at 1:27 pm

If there’s intelligent life, then (especially with the large scales involved) we can expect life to be fairly common, and intelligent life too. If it’s common, then we have no reason to expect ourselves to be unusual.

It’s the principle of mediocrity, and that’s also what I’m applying to the universe and existence.

  (Quote)

Patrick February 9, 2011 at 3:38 pm

Troy-

Wtf, man? Playing Dr. Craig on me with the crap false dichotomies? I either agree with you or am denying all possibility of debate?

I don’t care if you want to make analogies between human behavior and the behavior of a hypothetical magical superbeing with no characteristics at all. But you must do the following:

1. Every time you add a characteristic to the magical superbeing, you must adjust your probabilities. This is because adding characteristics to the magical superbeing is altering your hypothesis. The only way you can add a characteristic to the magical superbeing without adjusting your hypothesis is if you think it is 100% certain that the magical superbeing would have the characteristic, because that means you’re not adding it, you’re elaborating on what you’ve already calculated.

2. You must use the same assumptions about probability throughout your argument. For example, if you calculated the probability of fine tuned physical constants under naturalism by a tabulation of logically possible universes under naturalism, then you need to calculate the probability of fine tuned physical constants under theism by tabulating logically possible universes under theism. You’ve chosen logical possibility as your standard, now suck it up and own it.

3. Inferences about god, the universe, or whatever, have to remain the same throughout your argument. For example, if its fair to argue from human motivations to the motivations of a magical superbeing on the grounds that both humans and a magical superbeing are sentient, then one must remain consistent on this point throughout one’s argument. This isn’t even a matter of Bayesian analysis, its a matter of simple intellectual honesty.

  (Quote)

Troy February 10, 2011 at 7:46 am

Patrick:

I was articulating what I take to be a logical implication of your position on analogical reasoning about God. You might be completely willing, as a psychological matter, to discuss this or any other argument; but if you never allow analogical reasoning to determine the characteristics of the deity, then, as far as I can see, the probability of any event given the existence of God is inscrutable. And that means no Bayesian argument can work. Now, if you think your position doesn’t imply that, you’re welcome to argue that, and I’ve unfairly represented your original position, you’re welcome to correct me (I take it that’s what point 1. is supposed to be doing).

As for your specific points:

1. To begin, I don’t know why you say that the proposed being “has no characteristics at all.” None are built into the hypothesis; that doesn’t mean that there aren’t any there. The prior probability of a hypothesis should be determined by what is directly built into the hypothesis. After we’ve done that, we can articulate P(E|H) on the basis, not only of our definition, but also of prior experience with things like H. Imagine finding a building on another planet, and saying: “Aha–this confirms the hypothesis that intelligent aliens live on this planet.” This reasoning works because P(buildings|aliens) >> P(buildings|~aliens) and P(aliens) is not prohibitively low to begin with. It would be absurd for me to object to this reasoning by saying “If you’re going to suppose that aliens would have built buildings like these, then you’re going to have to build that characteristic of them into your hypothesis, and that will make the prior probability of aliens low enough to make the posterior probability negligible.” P(aliens) is already fixed; we reason by analogy with our experience that they’d be much more likely to build buildings like that than that would arise by chance. The hypothesis of aliens would not change, and so P(aliens) would not change, if we had found different kinds of buildings, or if we had found what looked like a large farm, or if we had actually seen what looked like aliens. The probability of all those events given aliens is fixed, so far as I can tell, independently of anything we’ve built into the hypothesis.

Analogously, I don’t see why I have to change P(T) when I reason, by analogy, that an intelligent designer would be reasonably likely to create a universe with life in it. The prior probability of the hypothesis is determined simply by what’s specified in the hypothesis: that there exists an extraordinarily powerful conscious being with the capacity to create a universe. The hypothesis says nothing about his desires, intentions, etc.; and when I reason by analogy to determine P(E|T) for some evidence E, I am not adding to the hypothesis.

To relate this to the earlier question of the implications of your position on analogical reasoning about God: It seems to me as if you’re setting up a dilemma: either P(E|T) is inscrutable, or P(T) is low enough to make P(T|E) negligible. I’m suggesting that P(E|T) can be determined (not exactly, but closely enough for our purposes) by analogical reasoning without affecting P(T). However, were this dilemma actual, it still seems to me that your position would have the implications I earlier spelled out. That’s because–no matter how apparently miraculous of an event E is–either P(E|T) will be inscrutable, making P(E|~T) irrelevant, or P(T) will be rendered so low (by adding things to the hypothesis like “God would want to reveal himself to the entire human race through a miraculous event”) that P(T|E) will still be negligible.

2. I’ve already responded to this with the poker analogy. Not all choices of a conscious agent are equally likely (there’s just as many ways for me to stack the deck as there are for a hand to come up randomly, but that doesn’t mean that each distribution of hands has the same probability given that I’m cheating as given chance.)

3. I’ve already said I agree with this; I’ve given reasons for why I don’t think that saying “but humans have a desire to maximize intelligent life” doesn’t defeat the argument, but the principle you articulate here is fine.

  (Quote)

Troy February 10, 2011 at 7:50 am

Addendum: It is, of course, not true that no characteristics of the deity are built into the hypothesis–I should not have said that. Rather, only a few characteristics (intelligence, consciousness, and power) are built into it.

  (Quote)

Patrick February 10, 2011 at 3:11 pm

Christ, man. How many times do I have to write that I don’t care whether you analogize from humans to a magical superbeing, as long as you do so consistently, before you stop translating that to “Patrick says I can’t analogize from humans to a magical superbeing.”? What’s the deal?

“1. To begin, I don’t know why you say that the proposed being “has no characteristics at all.” None are built into the hypothesis; that doesn’t mean that there aren’t any there. ”

That’s how hypotheses work. The only way you can say that a characteristic is “there” even though its “not built into the hypothesis” is if its 100% certain that the characteristic will be true if the hypothesis is true. If there’s a chance that the hypothesis as written could be true, but the characteristic could be not true, then that chance has to be factored in to the argument. Bayes gives two options for that, depending on how you want to word your argument: you can put it in the background data, or you can factor it in to P (E|H).

“It would be absurd for me to object to this reasoning by saying “If you’re going to suppose that aliens would have built buildings like these, then you’re going to have to build that characteristic of them into your hypothesis,”

Actually, this would be TOTALLY reasonable…

“and that will make the prior probability of aliens low enough to make the posterior probability negligible.””

…its just this point that would not. We’d have to actually discuss the probability that aliens would want to build buildings, and come up with an estimate of the probability of that claim. THEN we’d decide what effect it had on the rest of the argument.

“Analogously, I don’t see why I have to change P(T) when I reason, by analogy, that an intelligent designer would be reasonably likely to create a universe with life in it. ”

You’ve taken a hypothesis, and added to it a piece of data of which you are not 100% certain. The conjunction of the original hypothesis and this new piece of data is the product of the two. This will be a number less than the probability of the original hypothesis.

Of course, Bayes offers you a second way to do this. Instead of adjusting P(T) to account for the chance that the magical superbeing doesn’t want to create a life sustaining universe, you could insert that into P(E|T).

It doesn’t really matter which option you pick, except that if you pick neither, you suck at math.

Oh, and your poker analogy is non responsive to my point. I don’t know what else to do with that except refer you back to my earlier posts. The only thing I could possibly add is that if you have a credible argument for why its automatically more likely that a completely vanilla magical superbeing possessed only of sentience and a desire to create life would choose to create a finely tuned universe of the sort that atheism could produce (albeit, according to this argument, as a low chance event) would be more likely than a universe where, say, angels push the planets around through space, or babies have protective forcefields until they reach the age of responsibility, let me know.

  (Quote)

Luke Muehlhauser March 14, 2011 at 3:17 am

Lol @ “ugly bags of mostly water.”

  (Quote)

Leave a Comment

{ 1 trackback }