Malaclypse the Younger

Satan Oscillate My Metallic Sonatas

If it were truly independent, it wouldn't change the probability. However, The Anthropic Principle is very important here. If naturalism were true and the universe were not life friendly, you would not be around to observe this fact. Our ability to observe the life-friendliness of the universe is thus (under naturalism) dependent on its actual life-friendliness.

The case is even worse under supernaturalism, where our ability to observe the life-friendliness of the universe really is independent of the actual life-friendliness of the universe. If supernaturalism were true it would not-- by definition--be necessary for the universe to actually be life-friendly for life to exist. The fact that the universe actually is life-friendly thus argues against supernaturalism, since it is highly improbable that we would actually observe a life-friendly universe created by a supernatural designer.

Of course, if a designer were constrained to create a life-friendly universe in order to instantiate life, then this designer is, by definition, naturalistic (constrained by natural law).

[ February 26, 2002: Message edited by: Malaclypse the Younger ]</p>

tw1tch

tronvillain, Malaclypse and Dr. Retard,

Can ya help a theist out?


I've got some questions:

My understanding of this analogy is that the number '1063' means 'life-friendly universe'. That is to say that 1092 is not a 'life-friendly universe' as is 1094, 1095, etc. Thus the probability of 'life-friendly universe' is probability of rolling a 1063 out of 10^12 numbers
or 10^(-12). Are we on the same page with this?


When you guys say 'a priori'...prior to what event? The event of rolling the die IE universe coming into existence?

Also...
Is the main argument here:
A-I can't know X BEFORE the event (IE creation) which in this case is impossible.
OR
B-I can't know that the probability of God making a life-friendly universe is high. (IE I can't know that one die is biased towards 1063).

Are you saying there is no evidence that God would likely make a life-friendly universe?


No need. I completely believe the above...that if I don't know what number the die is biased towards
I won't be able to tell which die was actually rolled.

I absolutely, completely agree with some of the points Dr. Retard is making. That if one were walking along one day and were to simply find a box that said '1093'...that this would in NO WAY be evidence for a '1093 loving' troll.

Thoughts and comments welcomed,


Satan Oscillate My Metallic Sonatas

tronvillain

SOMMS:

I am simply exploring the thought experiment of two trillion sided dice, which as far as I can tell is not an analogy at all.

Malaclypse the Younger

tw1tch

When you ask nicely, sure.

We seem to be, yes.

Correctly accounting for what is or is not known a priori is one of the most challenging parts. A priori just means "prior to".

For instance, we know that the a priori probability of rolling a total of 12 on two standard (fair) dices is 1/36.

If we know a priori that the first die rolled is a 6, then the probability that the two dice will add up to 12 is 1/6.

Even a relatively simple case where the a priori knowledge is subtle can cause great confusion.

A parent tells you, "I have two children. One of them is a girl. What is the probability that the other is a girl?" Think about it for a moment before you look at the answer.

The answer, surprisingly, is 1/3 (and not 1/2). Given that for any specific child, a boy or girl is of equal probability, there are four equally possible possible cases:

1) boy boy
2) boy girl
3) girl boy
4) girl girl

What do we know a priori? All we know is that case 1 is impossible (one of the children is a girl). Therefore, there are three equally probable cases left, 2-4. In two of the remaining cases, the "other" child is a boy, in one of them the "other" child is a girl.

Now if the parent told you, "my eldest child is a girl", then we would have different [i]a priori knowledge, specifically that two of the cases (1-2) were excluded and only two cases would remain.

Let us now apply the concept of a priori knowledge to the FTA.

Suppose you are a genius who is capable, once you decide to do so, of quickly discovering all the laws of physics. Now suppose you somehow know, before you start investigating the laws of physics, that naturalism is true. You now know, even before you start, that you definitely will find that the laws of physics permit the development and existence of life. You now have a priori knowledge. In other words, you will not be surprised when you find out the laws of physics permit the development and existence of life.

Suppose, however, before you start investigating, that "supernaturalism is true" and "a supernatural god instantiated life." Supernaturalism entails that life could exist even if it were contrary to the laws of physics. Therefore you don't know whether or not the laws of physics would or would not permit the existence of life. You lack a priori knowledge about the laws of physics. To find that the laws of physics actually do permit the existence and development of life would be highly surprising.

Tercel

Malaclypse the Younger,
I’m a little stretched for time at the moment and I apologise for that. Hopefully I’ll be able to catch up on Friday with everything.
I see you’ve replied to a more recent post of mine. This is not my reply to that (which will hopefully be forthcoming on Friday), but rather a reply to one of your older posts which I was already half-way through responding too. So apologies if I cover information discussed too much in your most recent reply to me.
Anyway, a response to your post of posted February 24, 2002 09:27 PM:


Fascinating, because I have always found Koy’s writing style is to be as insulting and obnoxious as he possibly could be in between using the word “cult”. I have to seriously doubt I could be a quarter as bad if I tried. As for Koy’s logic… hmm, well lets just say I’ve heard more concise logic from 10 year olds.

Okay. However I still have the same question: Why do you want to find it “suprising”. The Fine-Tuning argument is not (at least my version of it is not) an emotional argument but an intellectual argument. It does not proceed “Wow, it’s so amazing that life exists!”, but rather by noting the existence of life and analysing the reasons for its existence.

You should have a read of my post to Wizardry dated February 25, 2002 05:37 PM in which I discuss this point. In summary, the difference is that an intelligent being would make a design decision to create a universe with life or not and then select a universe with the appropriate constants to carry out that decision.
Now it seems to me that your argument is that the idea of whether the created universe could contain intelligent life or not wouldn’t actually feature highly or at all in the mind of it’s creator. You suggest that its creator would be equally interested in any other specific physical feature of the universe.
Now I would agree that indeed a creator could be interested in other things besides whether its creation contained life, so that must of course be taken into account in estimating P(E|D), it might for example result in an estimate somewhat below 0.5.
However for this to be important to the argument the estimate has to be tiny. That is you would have to argue that the life/no-life issue is well down the creator’s list of priorities pretty much to the point where he’s not concerned about it whatsoever.
Now this indeed seems to be your argument and you are suggesting, if I understand you correctly, that the creator would likely be equally interested in the different non-living results.

I would have to say I seriously disagree with this suggestion. If there is an intelligent creator, then I think it would be extremely unlikely for him (or it or whatever) to be completely unconcerned about other intelligent life. It would seem extremely doubtful to think he finds a rock formation, chemical reaction or some other non-living event extremely hugely more interesting than he would find other intelligent beings like himself. Now a creator who is completely disinterested in intelligent beings is certainly possible. But is it very likely at all? I think, no. Since P(E|D) is effectively an average of all the conceivable possible feelings of all possible intelligent creators averaged by how likely they are, it is not going to be affected significantly by this consideration.

True. But the chance that an intelligent being created this specific arrangement of galaxies is similarly small. P(E|D) = P(E|C) Thus we have no reason to prefer design as an explanation to chance.

You could. But P(D) is then small rather than P(E|D). Since you need to assess P(D) * P(E|D), it doesn’t really matter whether you make P(D) the small one or P(E|D) the small one. The result of the multiplication is still going to be the same compared to P(C) * P(E|C). Thus, again, there is no reason to prefer design as an explanation to chance.

Utter rubbish. My argument is that you compare P(E|D) * P(D) with P(E|C) * P(C).
You shouldn’t be comparing P(E|D) with P(E|C) and drawing a result directly!!!! Haven’t you read anything I’ve written here?

Of course you can if you use your statistically-demented attempt at copying my argument.

All I need to show, as I outlined above is as follows:
P. “In all probability, an intelligent being who was creating a universe, would be at least somewhat remotely interested in whether the universe being created was capable of sustaining intelligent life. And furthermore would, in all probability, not be too likely to want to create a universe incapable of sustaining intelligent life as opposed to a universe capable of sustaining such.”

What experience I have of intelligent beings tells me that not only is the above true but that it is an extreme understatement.
It’s not really a question of whether life is “objectively special” but whether it is, in general, even remotely subjectively interesting to living beings. Certainly in my experience, living beings seem to be very interested in their existence and the existence of other intelligent beings.

I sense a little sarcasm here. Well if you’ve already introduced yourself under your new name I didn’t see it or at least don’t remember seeing it.

It annoys me though, as t prefer that any argument I am putting forward doesn’t get sidetracked by having you reconfigure it into something you think is better. Especially when “better” apparently means something that doesn’t work.
Changing it into a causal argument seems to me to introduce a whole lot of problems and one ends up defending causality rather than the Fine Tuning argument.

Hmm. I suppose that’s a compliment since you noted above that you thought Koy’s style was impressive. Clearly I’m far closer to Koy than I imagined… oh dear…

Similarly because you gave be so little to work with in your previous post I had to content myself with saying your statements were wrong. I couldn’t argue because no logic was presented to argue against.
However, in my first post to this thread, I outlined the basic steps of the Fine-Tuning argument as I see it and proceeded to flesh out the statistics behind the important steps in subsequent posts.

Actually, you didn’t. Your conclusion was “the life-friendliness of the universe is caused” which was clearly not my conclusion since, as noted above I completely avoid mixing causality and the FT argument.

I disagree. My argument is entirely rigorous. The values and computations I’m doing are the ones which drive Bayes’ Theorem. A statistical genius like you should have no trouble working out exactly what is being computed by my method and why it is valid.
If I used Bayes’ Theorem itself then it could be argued that it’s not entirely rigorous since P(D) &lt;&gt; 1-P(C) and thus Bayes’ the results of Bayes’ Theorem aren’t correct. Indeed, I note that you have argued this before when I did use Bayes’ Theorem. This is a slight change to avoid your objection and make the argument completely and utterly rigorous so I don’t need to keep qualifying my conclusions.

Come again? D is defined as: an intelligent being is responsible for the selection of the constants’ values. E is defined as: The subspace of the total space of scientifically analysable constant combinations which would allow life as we know it to exist in the universe. Okay, so those definitions aren’t exactly rigorous, but you should be able to get the idea. D does not seem to be connected to E in any way, it is certainly not defined in terms of E, as you suggest here.
I think you are confusing D (as defined above) with E|D. I have separated in this argument P(D) from P(E|D). E|D, is defined from the definitions of E and D (since it’s simply “E given D”) and P(E|D) is the probability that the constants take values from E, given an intelligent being is responsible for the selection of the constants’ values.

P(E|D) doesn’t equal one though – see above. Similarly you can’t have a set of {D1, D2, … Dn} – again see above.
I am not setting P(E|D) = 1 and incorporating the idea of P(E|D) = 1 into D! That should have been clear if you’d read my previous posts.
What I think your trying to say is that P(E|D) should equal P(E|C), which is the whole point I’ve been discussing with Dr Retard and Wizardry. Have a careful read of my posts to them.

You’re doing here the same thing you’re doing above. I let D be the hypothesis that a magic troll was the thing causing the result. Similarly P(E|D) was the probability that the number 1093 would be the result given that the result’s being caused by a magic troll. Since magics troll in general have no apparent reason to prefer 1093 to any other of the first trillion numbers I set P(E|D) at 10^-12. Setting P(E|D)=1 like you want to and incorporating the idea that the troll will pick 1093 into D is certainly a valid method. However, it is, I think, a waste since there are two ideas here: that the troll caused the result and that the troll would cause 1093 given that it was causing the result. We might as well have the ideas as separate probabilities. Certainly it is not fallacious to do so!

Say what? If P(E|D)=1, how can it equal ~10^-12 as well?

This objection, is, I think coming again from your confusion of how I’m using D and E|D. You want to be comparing E|D with E|C, and similarly D with C. My assertion is that the FTA works with a lower P(D) than P(C).

Again, don’t compare P(D) with P(E|C), because they aren’t talking about the same thing. Your complaint should be, I think, that P(E|D) is fifty or so orders of magnitude higher than P(E|C). I’ve tried to explain to Wizardry why this is.

I’m not definition it in terms of E so P(E|D) isn’t 1 by definition. The “designer” is defined only as being an intelligent being and selecting the constants by intelligent choice. I don’t know what you mean by “rational estimate” since one of your earlier comments suggested your opinion of what the word “rational” meant differed significantly to mine.
But we can certainly make an estimate to the best of our abilities and experience and knowledge as to the approximate value of P(E|D). Being intelligent beings ourselves, the probable decisions of the “designer” would certainly seem within our ability to understand and estimate.

What you mean by “indeterminate” I’m not sure. P(D) is the a priori probability that this universe is the result of intelligence. –Obviously including the idea that that intelligence exists. This can be reasonably estimated. Sure we don’t know its exact value. But all of us have our ideas about exactly how a priori likely it is that the world is the result of a Creator, we can estimate, approximate etc without too much bother.

Of course. I would hope that most people reading this or coming across the Fine-Tuning argument elsewhere comes to their own decision about what they think are reasonable a priori values. Your suggested value here appears to demonstrate clearly that you are a material presuppositionalist, that is to say your suggested value would appear to be so extremely unreasonably small that I am left to gather that you are effectively ruling out the idea of the possibility of a Creator a priori.
Well to some extent that is fine, you are simply doing the same thing as Theophilus but choosing the opposite side. On the other hand, I prefer to look at evidence before coming to a decision, and aren’t impressed by presuppositions to your position is not going to draw much respect from me. Certainly many of your fellow sceptics here seemed to have serious issues with Theophilus’ presuppostionalism so they might have something to say to you.

Okay, hopefully I’ve managed to convince you that P(E|D) is not equal to 1 and hence P(D) isn’t equal to P(E). I do though, wonder why on earth you’ve said P(E) = P(E|C). That would obviously only be true if P(C) = 1. (which would lead to problems by what you’ve written above, since P(D) would equal zero which would mean P(E) was zero which would be a rather interesting result!)
If you could rewrite the above values based on P(E|D) not being 1 and P(D) not being P(E) -as I’ve hopefully managed to convince you with this post- then that would be good.

Tercel

Dr. Retard

I'd just like to say that I'll participate in this discussion soon, but right now I'm experiencing intense stomach pain. Let's just say that my exercise of free will was not worth the consequences of Hormel Chili. Why, God!?

HRG

That's exactly the simple situation which I described (with "isomorphic" = mapped 1-1 onto each other by a function which is continous in both directions). It can even be made as smooth as you want.

This function would correspond to the transition of one set of physical constants to another, equivalent one.

But similar functions exist if C is not a square, but any shape which is contractible.

Regards,
HRG.

Malaclypse the Younger

Tercel

Don't start this shit, Tercel. You do not want to get in a flame war with me. I'll let this one go by.

I change it to the causal argument because causality is easier to define and argue than design. Since design entails causality, showing that belief in causality is not warranted entails that belief in design is unwarranted.

The problem with formulating the argument with design is that design and chance are not mutually exclusive options. It just requires more arbitrary values for the various constants. But I can go either way.

Now let's get to the crux of the bisquit.

First, your approximation is sufficiently accurate. But rather than bothering to do the arithmetic, it's vastly more instructive to look at how the assumptions affect the outcome.

We stipulate that P(E) = P(E|C) = 10^-50. Given that this constant is extremely small, it really doesn't matter what we use for P(E|D) and P(C); we might as well hold them arbitrarily close to 1.

Essentially we are comparing P(E) with P(D). Examining the equation analytically, we see a definite relationship between the assumptions and the outcome:

• If we assume P(E) greater than P(D) we conclude chance.
• If we assume P(E) less than P(D) we conclude design.
• If we assume P(D) equal to P(E) we can't draw a conclusion.

Restating this relationship in English:

• If we assume that chance is more probable than design, we will conclude chance
• If we assume that chance is less probable than design, we will conclude design
• If we assume that chance and design are equally probable, we can't conclude either.

The bottom line is that any value of P(D) is completely arbitrary. Assuming 10^-30 is just as arbitrary and not rationally justified as assuming P(D) = 10^-50, 1/pi, .26 or 1/(10^10^10^..^10).

Your choice of making P(D) 20 orders of magnitude larger than P(E) is arbitrary, and we see that the "conclusion" you derive is entirely dependent on that arbitrary assumption.

Your version of the FTA is thus circular and fallacious.

[ February 27, 2002: Message edited by: Malaclypse the Younger ]</p>

Clutch

[bells, buzzers, flashing lights, applause...]

Well said. End of story.

Satan Oscillate My Metallic Sonatas

Dr. Retard,
I will definitely pray for you.


However, it will take A LOT of faith to deal with Hormel.



Satan Oscillate My Metallic Sonatas