Automaton

Ojuice5001:

Assuming naturalism, why would the big bang be a unique event?

Assuming naturalism, why would (therefore) arbitrarily defined physical laws not be manifest in as different ways as there would be universes?

Assuming naturalism, why would the "fine-tuning" of one in an indefinite set be extraordinary?

As you can see, this argument is hardly convincing to anyone but true believers.

tronvillain

Interesting. You've almost presented the reverse of the standard fine-tuning argument. Of course, it is no more successful than any of the standard versions.

Now, your formulation amounts to taking the set of logically possible universes and examining the set of logically possible causes for those universes. You presumably claim that universes capable of supporting "life" have a higher proportion of "intelligent" causes than universes that are not so capable, and that this proportion is in fact greater than that of "unintelligent" causes. Apparently then you assume all causes equiprobable, and since we are in a universe capable of supporting life, conclude that it is more probable than not that our universe has an "intelligent" cause.

The first problem arises with the claim that universes capable of supporting "life" have a higher proportion of "intelligent" causes. While it's possible, it is by no means certain - intelligences uninterested in life are readily imaginable.

More significantly we have the second problem: the claim that the proportional of "intelligent" causes for a life-bearing universe is greater than the proportional of "unintelligent" causes. Just as there are far more ways of being dead than there are of being alive, it seems likely that there are far more ways of being unintelligent than being intelligent, regardless of the properties of the universe to be explained.

While one could theoretically take issue with assuming all causes equiprobable, since we are presumably considering every logically possible cause I don't see a problem with it. This only important when fine tuning proponents reduce the number of hypotheses to two and declare them equiprobable.

bd-from-kg

HRG:
The argument presented talks about the number of members of AN, AI, etc., not about probabilities. It's my custom to deal with the argument actually presented, not with what I think is a better argument.

I'm familiar with formulations of the FTA in terms of probabilities. If someone chooses to offer such an argument here I'm quite prepared to deal with it.

Duck!

Unless you can calculate the odds of the following....

1) The likelihood of a "God" universe resulting in intelligent life like humans.

2) The likelihood of a "Godless" universe resulting in intelligent life like humans.

3) The fraction of all potential universes that actually are "God" universes.

...then you can't really claim which is more likely, that humans are the result of natural processes or that they were deliberately created by an intelligent higher power.

And anyway, your argument only mentions the vague idea of a universe with "an intelligent being". Surely that covers the gamut from an omnipotent, omniscient creator God to demiurge type Gods to superintelligent aliens to human beings. The range of possibilities (and we don't know if they are possibilities) are limitless.

No sir, your argument don't cut ice with this atheist.


Duck!

HRG

You should, IMHO. Probabilities are numbers between 0 and 1, and the probability of a disjoint union is the sum of the probabilities. Thus, "equiprobable" makes only sense on a finite set - and the set of all logically possible causes is not finite.

In particular, fine-tuners have to assume that the physical constants may be tuned continously; otherwise they lose any basis for arguing that the domain of life-friendlyness is small. But this immediately raises the question: small according to which notion of probability ?

Regards,
HRG.

tronvillain

No, I still don't have a problem with it. If there is an infinite row of dominoes but every fourth domino is white in contrast to the black, what is the probability of picking up a white domino from the infinite row? It appears to be 0.25, though the probability of picking up any specific domino is infinitesimal. Perhaps I'm missing something, but I don't see why the fine tuning claim couldn't be similar.

bd-from-kg

tronvillain:

Yes, you're missing something. The actual probability is indeterminate.

Let's take another example. If you pick one number from the set of all positive integers, what's the probability that it's prime? Well, the density of the primes is zero in the sense that the percentage of primes in the first N integers approaches zero as N increases indefinitely. So by your reasoning, this probability should be zero. And yet, I pick a number: 23. Smith picks a number: 91. Both of them are prime. What happened? Well, the "probability density functions" for our choices were strongly skewed toward the smaller integers. In fact any actual probability density function for any actual choice of integer will be skewed in this way. For suppose that Jones claims to have chosen an integer in a way that is not so skewed. Now for any N, what's the probability that Jones's choice is smaller than N? Why, zero, of course, since there are infinitely many integers greater than N and only finitely many that are less. But since this is true for any N, Jones's number is larger than any given integer with probability 1. This is absurd, so we are forced to conclude that it is impossible to choose an integer in this way.

And this is a simple case. Typically there is no uniquely “obvious” or “intuitive” way to define a probability density function at all. For example, consider the following problem. We have a circle with diameter of 1, and choose at random a line that intersects it. The line will make the same acute angle with the tangent to the circle at either point of intersection. What’s the probability that this angle will be greater than 60°?

Well, let’s see. The possible angles run from zero to 90°, so the probability that the angle is greater than 60° would seem to be ⅓.

But wait. Any such line defines a line segment inside the circle, and every such line segment has a unique center. The centers of the line segments for the lines that make angles of 60° or more with the corresponding tangents are those that lie inside a circle with the same center as the original one, but with a diameter of ½. The area of this circle is ¼ the area of the original one. Since each point inside the circle defines a line segment (namely the one for which it is the center), and ¼ of these points correspond to lines that make an angle of 60° or more, the probability of choosing a line that makes an angle of 60° or more would seem to be ¼.

But wait. Each of these line segments is perpendicular to a diameter, and each line segment intersects its corresponding diameter at a unique point. Now for any given diameter, of the line segments perpendicular to it, the ones that make an angle of 60° or more are precisely those that lie in the middle half (that is, inside the circle with radius ½ described earlier). But exactly half of the points on the diameter lie on this middle segment. So by the same kind of reasoning as before, the probability of choosing a line that makes an angle of 60° or more would seem to be ½.

So the probability is ⅓, ½, or ¼. Take your pick. None of these answers is more “right” than the others. What we’ve actually done is to define three different, but equally plausible, probability distributions for the lines that intersect the circle.

In the case of things like AI and AN (even if they were actually sets, which they aren’t, so that such things could be defined), the number of different plausible probability distributions would clearly be infinite. (In fact, it would be an infinity of indefinitely large cardinality.) None of them would be more “objectively correct” than any of the others. And they would of course give a ridiculously large range of probabilities for the possibilities in question.

So this whole argument, as Bentham might have said, is not only nonsense, it is nonsense on stilts.

[ March 31, 2002: Message edited by: bd-from-kg ]</p>

tronvillain

Are you claiming that if one started actually drawing lines and measuring angles, the answer wouldn't converge towards a specific value? I may be missing something again, but to me it appears that it would converge to ½ and that the first two arguments are wrong:

1) While the possible angles run from zero to ninety degrees and we are looking for angles greater than sixty degrees, simply dividing sixty by ninety to give a probability of ⅓ ignores the mechanism by which the angles are produced. If the mechanism is examined, we find that in half of all possible cases the angle is greater than sixty degrees.

2) While the area of the circle with a diameter of ½ is ¼ of the area of the original circle, area is not representative of the probability. The number of tangents which can be drawn to a circle is infinite, but does not depend on the size of the circle, whether its radius is zero or one. Now, the number of pregressively smaller circles which can be drawn with a larger circle is infinite, but it does depend on the size of the circle (specifically on its diameter or radius). So, we see that while the circles from diameter ½ to one sum to an area of ¾ and the circles from diameter zero to ½ sum to an area of ¼, the number of tangent lines of each will be equal since they are summed over the same distance.

As I said, the answer appears to be that the probability that the angle will be greater than 60° is ½. Which is good, because it doesn't really seem to make much sense for all three answers to be equally right - just think about actually drawing lines and measuring angles.

[ March 31, 2002: Message edited by: tronvillain ]</p>

Theli

Ojuice5001...

ok. But why call it "A"? anyway...

Not physical laws, they are purely mechanical in nature. There is no sign of influence from any emotional being.

Like ours... Most of the time, that is.

Only if life existed before the universe, meaning that the universe adapted (with god's help) to life. Not the other way around (wich most evidence points at).

BTW, the intelligent being doesn't have to be friendly.

In cartoons, yes.

Aaaahhhh.... The core of the intelligent design argument. "intelligence is needed to create intelligence".

But tell me, what created the first intelligence?
It wasn't intelligence...

Malaclypse the Younger

tronvillain

What bd appears to be saying is that the probability distribution will differ as an artifact of how you choose your lines. The problem with the circle is that the naive formulation is not sufficiently well-specified to determine a particular probability distribution.

To extend it back to the FTA, without actual empirical knowledge of the distribution of physical constants in the universe, Physics is not sufficiently well-defined to agree on any particular probability distribution.

One could conjecture any number of additional mutually exclusive axiom-sets, each internally consistent with established physics, each entailing a particular probability distribution. However without actual observations of the physical constants of other universes, it is impossible to objectively determine which of these axiom-sets represents reality.

[ March 31, 2002: Message edited by: Malaclypse the Younger ]</p>