Ojuice5001

Here is my formulation of the Fine-Tuning Argument:

1. There is some aspect of existence--call it A--that makes our facts happen one way and not another. This may be the reason our laws of physics are what they are, or the laws themselves.

2. All A can be classified into those with an intelligent being, and those without.

3. The vast majority of possible A without an intelligent being will result in a universe unfriendly to life. This is the point of all the statistics about how weakening the strong force would prevent elements other than hydrogen, etc. "Possible" A means "logically possible." For example, a situation that mandates every force to be half what it is, is a possible A.

4. A much greater proportion of possible A with an intelligent being will be friendly to life. This is so because there are plenty of reasons intelligence would want to create more intelligence.

5. The universe is life-friendly.

6. So the universe we have would be produced by relatively many A with an intelligent being, and relatively few without one.

Autonemesis

OK, so there's a supernatural A acting intelligently, tuning the fundamental constants that describe physics, so that a cosmos hospitable to life is possible. So What? What are you implying by advancing this proposition?

Is there a particular way in which you think this intelligent A should be regarded? Are you implying a relationship to any other claim or claims of supernatural intervention?

Of what utility is this theory of yours?

Datheron

Ojuice5001,

Ah, ze old fine-tuning.

I'm not too sure whether this can be well-defined, and whether A itself is subject to some other controlling factor (ad infinitum).

Not true.

If we do not have a sufficient data sample size, we cannot tell whether A is affected by intelligence or otherwise. I just watched an excellent movie where one of the plot lines was the flipping of a coin and its tendency to land on tails every single time; it was only after the 10th or so time that we start finding the coin suspicious. Similarly, given 100 tosses, statistically, there is a very high chance of flipping either 7 heads or tails consecutively, even though such odds would seem odd in isolation.

Another objection is the ruling of A being of an intelligent being. Have we ever met an intelligent being that has produced A, in order to make this claim? This is circular.

But by what means can you say that any other universe resulted from a different set of constants will for some reason be uninteresting? By assuming that life is somehow important, you are arguing something that you cannot have evidence for; the only way you can do is is by drawing connections with the theory of the intelligent creator-lifeform, which is once again circular. We do not know how any other configuration will result; if it still takes ~11 billion years for the simplest life to form with "intelligent constants", then wouldn't a set of constants that yields life in 5 billions years be even more intelligent? In 3? Who's to say whether such constants are possible?

Once again, an argument based on circular logic. Only by assuming an intelligent creator can you say that the universe is interested in intelligent life. If I deny that premise, then you cannot rationalize such a statement.

Not quite. The universe is horribly unfriendly to life...why are all arguments so homocentric? There is an entire universe out there, and we have not spotted any other signs of life anywhere else out there. We can say that at least 99% of the space out there is desolate real estate. You'd be more correct in saying that boiling lava is "life-friendly".

From my refutations above to your claims and their shaky assumptions, I would reevaluate this conclusion.

[ March 29, 2002: Message edited by: Datheron ]</p>

Malaclypse the Younger

Ojuice5001

Acceptable.

Unproven. There is no reason to believe this statement.

This premise is in doubt. The probability distribution may simply be an artifact of the way we arbitrarily choose to represent physical law. Without actual examples, we cannot know.

However, for the sake of expounding on probability theory, this premise is stipulated.

This premise is very much in doubt. The probability distribution of intelligently designed universe is simply unknown.

However, again for the purpose of understanding statistics, I will stipulate this premise.

Agreed. This fact is observable.

This is correct. However you have not proven that it is more likely that an intelligence thus exists.

To make sense of the FTA, you have to first understand probability notation.

Unadorned letters (A, B, C, etc.) are predicates, assertions that something might be true or false. The normal boolean operators, ~ (not), & (and) and $ (or) apply normally.

P(A) is the probability distribution of true statements of A; in other words, in what percentage of possible cases would A be true.

P(A|B) is the probability distribution of A given the truth of B.

On to the FTA.

For a more thorough description, see <a href="http://quasar.as.utexas.edu/anthropic.html" target="_blank">The Anthropic Principle does not Support Supernaturalism</a>

F is the predicate, "The universe is life-friendly". It is obviously true.

D is the predicate, "An intelligent designer exists." This is the predicate we wish to determine.

By our stipulations, P(F|D) gt P(F|~D); this is just your statement 6 written in probability notation.

However, we wish to determine P(D|F). To do so, we must emply Bayes Theorem:

P(D|F) = (P(F|D) * P(D)) / ((P(F|D) * P(D)) + (P(F|~D) * P(~D))

However, we have no way of knowing P(D) (note: P(~D) = 1-P(D) in this formulation). Therefore we can't draw any conclusions from the stipulated inequality.

Worse yet, we must take into account the Weak Anthropic principle:

L is the predicate, "Life exists."

We know that if there is no designer, and life exists, it is definite that we will observe a life-friendly universe. Thus, in probability notation

P(F|~D&L) = 1.

By another form of Bayes Theorem, we know that

P(~D|F&L) = (P(F|~D&L) * P(~D|L)) / P(F|L)

therefore

P(~D|F&L) = P(~D|L) / P(F|L)

since 0 less than P(F|L) less than or equal to 1,

P(~D|F&L) greater than or equal to P(~D|L).

Therefore, given the WAP, the observation that the universe is life-friendly does not reduce confidence in the nonexistence of a designer, therefore it is not evidence for a designer.

The FTA is simply flawed on so many counts, and fundamentally rests on an unsupported interpretation of probability theory.

[ March 29, 2002: Message edited by: Malaclypse the Younger ]</p>

Pomp

MtY,

2. All A can be classified into those with an intelligent being, and those without.

Unproven. There is no reason to believe this statement.

Out of curiosity, what is your objection to proposition 2? As far as I can see, all A can be classified according to whether or not an intelligent being is responsible. There's no a priori reason to assume that such a classification scheme is any more meaningful than other possible classification schemes, but it's certainly one possible scheme.

Malaclypse the Younger

Pompous Bastard

Well, he's making a truth claim, not arbitrarily establishing an objective distinction.

Let us use "lumps of clay" and "human potters". It is false that we can determine, without knowing the source extrinsically, that it is possible in principle to correctly classify all lumps of clay as being human-designed or naturalistic.

Pomp

MtY,

Well, he's making a truth claim, not arbitrarily establishing an objective distinction.

OK, if you read it that way, I see what you're talking about.

Malaclypse the Younger

It's really irrelevant anyway; that step is unnecessary. You can just skip to steps 3 and 4.

bd-from-kg

Ojuice5001:
This argument fails immediately for two reasons:

1. Steps 3 and 4 are meaningless. The subset AN(I) of AN which will result in a life-friendly universe is infinite, as is AN itself. Infinities cannot be compared unless one of them actually has a smaller cardinality than the other. But there is no reason to suppose that here; one would naturally expect AN(I) to have the same cardinality as AN. Similarly, the subset AI(I) of AI which will result in a life-friendly universe would seem to have the same cardinality as AI.

2. The conclusion does not follow from the premises. Even if it were true that AN is much larger than AN(I) (as asserted by step 3) and that AI is not much larger than AI(I) (as asserted by step 4) it would not follow that AI(I) is much larger than AN(I) (which is what the conclusion asserts). To see this, let’s pretend for the moment that these classes are finite so that their sizes can be compared. If AN is much, much larger than AI, it could easily be true that AN is much larger than AN(I), AI is about the same size as AI(I), and AI(I) is much smaller than AN(I).

HRG

The fine-tuning argument doesn't talk about cardinalities, but about probability measures defined on AN resp. AI and the probabilities they assign to AN(I) resp. AI(I). Two intervals of the real line have the same cardinality, but we still can distinguish their lengths.

Nevertheless, the non-finiteness of AN etc. is a major stumbling block for fine-tuners. The point is that no a priori probability measures exist on an infinite set (on a finite set we have the measure which assigns equal probability to every single point).

Thus the statement "AI(I) has a small probability" is meaningless, because it depends on an arbitrary choice of a probability measure on AI.

Regards,
HRG.