pug846

I don’t think so. There seems to be a difference between for what I’m arguing for and for a general event that has a low probability. For instance, you shuffle a deck of cards and give each player 5 cards. Obviously, the probability of getting any one particular set of 5 cards is pretty low. However, if someone were to have received the same set of 5 cards 100 times in a row, I don’t think it would be reasonable to assume that it was just chance. You seem to be arguing, and perhaps I’m wrong here, that we couldn’t validly conclude that the deck was stacked.

I’m not sure I understand – In the example I used in my previous post, there were good reasons to assume that a non-designed universe would not be “life-friendly.” I’m not claiming anything about the a priori probabilities. If there are in fact 10^10000 potential combinations and we could calculate using a computer how many of them would allow for a Universe to form. If only 17 of the possible combinations allowed for “life-friendly” universes, then I don’t see what we couldn’t conclude that something is amiss here and “something intentional must have been involved in the creation of the Universe.” In the same way we would conclude that if someone got the same 5 cards 100 times in a row, we could rule out chance and conclude that “something intentional must have been involved.” Perhaps I’m still glossing over something here, but I don’t see why you can’t apply a similar criterion in both situations. (The one with the cards and the one with the Universe).

And I still don’t see why P2 is false. By simply running the different constants through a computer, we can “observe” how other universes are non-life-friendly.

Malaclypse the Younger

The cases are different, because you specified a priori a particular outcome (or set of outcomes) that would constitute a stacked deck. However it is not possible to specify a priori that life is "meaningful"--at the most trivial it is known that designers (ourselves) design all sorts of non-life-friendly mechanisms, and events that have nothing to do with life are indeed caused.

A non-life-friendly universe could certainly exist. However, it is not possible for us to observe a non-life-friendly naturalistic universe, since we would not exist as a consequence of its non-life-friendliness.

This is a claim about a priori probabilities. We are stipulating that, based on our understanding of physics, before we look at a particular universe (a priori) the probability that it will be "life-friendly" is 17/10^10000.

However, we cannot specify a priori (before we look at the particular universe) that a "caused" universe will contain life.

You are conflating specifiability with improbability. We can indeed know before we examine a deck of cards (a priori) that we can observe certain conditions that cannot plausibly obtain. This is because the rules of poker are known before we look at the cards.

As a side note, even if it were valid, the Fine Tuning argument could determine only that the particular values were caused as opposed to chance--it cannot, by itself, determine that the values were teleologically intended.

For instance I can tell if I am in a gravitational field by observing grains of sand. If they all suddendly start moving in the same direction (a highly improbable and specifiable event), I can conclude I'm in a gravitational field. I have certainly not concluded that some teleological intelligence wanted those grains of sand to move.

P2 asserts that if a designer did not exist, we would observe a non-life-friendly universe. This assertion is necessarily false because if a nondesigned universe were non-life-friendly, we would not exist to observe that fact.

[ February 23, 2002: Message edited by: Malaclypse the Younger ]</p>

Tercel

In my opinion, the Fine Tuning argument has nothing to do with the objective significance or otherwise of life. (unless you straw-man it of course)

Pug846 is on the right track.
1. There are a large number of potential combinations of physical constants.
2. The margin of error of some of these constants for allowing any sort of life as we know it to form is very tiny
3. What is the most likely reason for this?
C. If there exists one and only one universe, then the probability that the reason is "design" far outweighs the probability that the reason is "chance".

I think it's not difficult to see that the argument is valid. With regard to its soundness, the options seem to be to deny 1 and insist the universe is a brute necessity or to deny the the conclusion by hypothesising multiple universes. Since both ideas are completely lacking in evidence beyond ways of escaping this argument, I have always liked the Fine-Tuning argument.

But this is wrong because the Fine-Tuning argument is a probablistic argument. The case is not "if small prior probability then design" but rather:
Any singular event with a small prior probability was, in all probability, designed.
(An example: An entire trained firing squad misses me at my excution, the chance that there is some designed plot not to kill me is immensely greater than that they all missed my chance.)
The inverse implication of which is clearly true. -eg the chances of me winning the lottery is tiny.

The probabilistic case is not complicated at all. We merely note the occurance of the event and ask for the most likely reason for that even.
In statistical terms we're comparing P(E|D) * P(D) with P(E|C) * P(C) -where E=event, D=design, C=chance- (with the condition in the case of the Fine-Tuning argument I described above that P(C) = P(~D)) which is not a difficult task at all.)


<strong>Dr Retard</strong>,
With regard to your initial analogy.
As I noted to Malaclypse we are comparing P(E|D) * P(D) with P(E|C) * P(C), so lets do that to your analogy.
We already know P(E|C) - it's one in 10^12.
(Your analogy differs slightly to the fine tuning argument because here P(C) &lt;&gt; P(~D), but that doesn't matter because this system of assessment to find the most likely cause will still work, it simply means we have to make up two values -P(C) and P(D)- instead of just the one)
Personally I would say that the existence of a magic troll who is interefering with your RNG is fairly arbitrary and random and extremely unlikely: P(D) = 1 in 10^12 perhaps to be generaous? (That is, one in a million million)
The probability of the troll liking 1093 as opposed to any other number is exactly normal random chance - ie P(E|D) = 1 in 10^12
The probability that your random generator is working by chance, I would put at about 99.9999% say: P(C) = 0.999999

Thus feeding the numbers through the equations we get:
P(E|C) * P(C) = just under 10^-12
while
P(E|D) * P(D) = 10^-24
That is to say: the troll hypothesis has been found to be a million million times more unlikely than chance (and I was being generous to P(D) too).

Let's face it the probabilities behind the Fine Tuning argument do work. They're just the normal way of assessing probability, nothing's been cooked up specially for the Fine Tuning argument - As you can see by my analysis of your analogy, the statistics are valid for assessing the most likely cause of any event.

Tercel

[ February 24, 2002: Message edited by: Tercel ]

[ February 24, 2002: Message edited by: Tercel ]</p>

Dr. Retard

But that seems to be very like the FTA. The existence of an immaterial designer somehow interfering with the cosmological constants seems very unlikely. And the probability of the designer liking life-permitting universes as opposed to any other kinds of universes seems to be exactly normal chance. Moreover, I'd put the probability of the constants just falling out naturally -- the "chance" hypothesis -- at a high likelihood. I see the two cases as alike; is there any important and relevant difference here?

[edited for typo]

[ February 23, 2002: Message edited by: Dr. Retard ]</p>

Dr. Retard

Why isn't the pro-troll argument just as good?
1) There are in fact numerous potential RNG results.
2) Only a few of those results are 1093.
3) Given that there is an extremely small chance that we see a 1093-result, there must be a 1093-loving, RNG-interfering troll.

This troll case also stipulates vanishingly low probability for the result in question. But surely we don't need to invoke any explanation for it. Why would we need to do so for a life-permitting universe? What is the (important and relevant) difference?

LinuxPup

If I may borrow your RNG analogy let me offer this for discussion:

A random number generator gives us 1 billion numbers, and each number can vary from 1 to 1 trillion. Now let's say after running the RNG only one time, every single number (all 1 billion of them) is 1093. Would it be more reasonable to say this happened by chance, or that the RNG was somehow made (intentionally or not) to do this. Obviously to believe this happened by chance is ridiculous. It would be more reasonable to believe that this wasn't mere chance.

This I think is a better analogy of the fine-tuning argument.

And that principle is correct. However this of course does not remove the unlikeliness of a universe suitable for life, but merely states that a life form *must* obvserve the attributes necessary for life, for in order for him to see it, they must exist. This is a common slight-of-hand argument, but it completely skips the argument of unique fine-tuning.

I agree. But the evidence for fine-tuning seems quite apparent. I recommend checking out this page:

<a href="http://www.reasons.org/resources/books/creatorandthecosmos/catc14.html" target="_blank">http://www.reasons.org/resources/books/creatorandthecosmos/catc14.html</a>

Dr. Retard

First, as per my original post:

Second, I can see why a series of 1093-results would be special or significant. But I don't see how a life-permitting set of cosmological constants could be special or significant. Just like I don't see how a single 1093-result could be special or significant. Can you give a reason why a life-permitting set of cosmological constants is special or significant?

Malaclypse the Younger

Tercel

The objective (or at least prior) significance of life has to be established to distingish a priori life-friendly universes and find it's occurence "surprising".

The problem with this argument is that the chance of any unique feature of any universe is vanishingly small. By this reasoning, we would have to accept that, if there is indeed one universe, no matter what the effects of the physical universe, it must be designed.

This is a false argument because there are many singular features of the universe (such as the particular arrangement of galaxies) which are also highly improbable, yet do not seem to demand the explanation of design.

Well, you don't actually explain the entire argument, because you don't make explicit the fundamental premise. Your argument, stated formally is:

P1: The probability of a the universe being life-friendly is very small.
P2: If a singular small probability event occurs, it must be caused
-------------
C: the life-friendliness of the universe is caused.

This argument is valid, but P2 is not true (and the burden of proof is on the claimant). Depending on your definition of "singular", either singluar small-probability chance events occur constantly without cause (and P2 is false), or no singular small-probability events occur (in which case P2 is indeterminate).

State the argument formally so we can see P2 above explicitly, and realize that it's an astonishing claim to be taken as an assumption.

First, this argument is not entirely rigorous (you have to use Bayes Theorem to compute P(D|E)), but (in either formulation) we have to figure out the values of P(D) and P(C). Since any combination of values could have been caused, there is no justification for assigning P(D) a value any higher than P(E|C); of course P(C) then is 1-P(D), and we find that the two sides of the equation are equal.

Only if you assume (without evidence) that the a priori probability of a designer is orders of magnitude higher than the a priori probability of a chance universe. If you make that assumption, it is not surprising that you will arrive at that same conclusion.

As noted, you cooked up an a priori probability for your designer 12 orders of magnitude higher than the a priori probability of a chance universe.

[ February 23, 2002: Message edited by: Malaclypse the Younger ]</p>

Malaclypse the Younger

LinuxPup

If we could observe 1 billion independent universes (or even two independent universes) and conclude they all had life, then one would certainly start to conclude that life-friendliness was somehow caused, and that the range of possible values was somehow more highly constrained than we now suspect.

Of course, it is the case that we can observe only a single universe, not a billion (or even two) of them.

The anthropic principle refutes the premise that it is surprising that we would observe a life-friendly universe; indeed it trivially proves that if naturalism is true, we will necessarily observe a life-friendly universe.

Of course, if we observed a non-life-friendly universe, we would instantly and rationally conclude the existence of a supernatural designer. We can use this fact to create an an evidential argument against a supernatural designer.

P1: If naturalism were true, we would necessarily observe a life-friendly universe.
P2: If naturalism were false (and supernaturalism were true), it would be highly unlikely to observe a life-friendly universe (since a supernatural designer has no need or particular reason to design a life-friendly universe to create and sustain life).
P3: We do indeed observe that the universe is life-friendly.
------------
C: Supernaturalism is highly unlikely, and thus naturalism is highly likely.

[ February 23, 2002: Message edited by: Malaclypse the Younger ]</p>

Satan Oscillate My Metallic Sonatas

Dr. Retard,

Your example, while well phrased, is not analogous to the fine-tuning argument for this reason: it assumes the answer...that the universe randomly happened.

Notice you start off: ‘I have a RANDOM number generator...I get 1093...is it working properly or did some troll mess with it?
Again, this argument assumes the answer...that 1093 came up randomly.

A more accurate analogy of the issue at hand (a life-supporting universe) is this:
You have 2 number generators:
Number Generator #1 is a trillion sided die, one side has ‘1093’ on it. (random)
Number Generator #2 is a 10 sided die, 9 of those sides have ‘1093’ on them. (troll)
Event: you choose one machine (you don’t know which one), turn it on, and up comes ‘1093’.

Now how likely is it that you chose the Number Generator #1...the completely random generator? Not very. How likely is it you chose Number Generator #2...the troll’s number generator? Very likely.


Thoughts and comments welcomed,

Satan Oscillate My Metallic Sonatas