The Fine-Tuning Argument Gets Kicked In The Teeth

[_JohnStuartMill]

Physicists are saying that some cosmological constants might not be so "constant" after all.

[_asbestosman]

Once again proving that science doesn't know everything.
Therefore God. ;)

[_JohnStuartMill]

Hehe. The Argument from Ignorance conquers all!

[_Some Schmo]

Once again proving that science doesn't know everything.
Therefore God. ;)

True, science doesn't know everything. But I suppose I'd take that over religion which doesn't seem to know anything.

Therefore, no god. :D

[_Simon Belmont]

Once again proving that science doesn't know everything.
Therefore God. ;)

True, science doesn't know everything. But I suppose I'd take that over religion which doesn't seem to know anything.

Therefore, no god. :D

And I would accept almost anything over your unsubstantiated and often naïve opinions.

[_Some Schmo]

And I would accept almost anything...

You could have stopped right there.

[_Willy Law]

And I would accept almost anything...

You could have stopped right there.

Dammit Shmo, you just made me laugh so hard I spilled my Fruity Pebbles all over my MacBook.

[_Calculus Crusader]

Physicists are saying that some cosmological constants might not be so "constant" after all.

I don't see a kick in the teeth. Incidentally, this:

"The data for this 'dipole' model of alpha has a statistical significance of about 4.1 sigma, meaning that that there is only a one in 15,000 chance that it is a random event."

Is an absolutely confused statement.

I assume the author of the article meant the following:

The test statistic computed from the astronomical data was 4.1; under the null hypothesis (i.e., that the fine structure constant = 0.007297352569), the probability of observing that test statistic or one more extreme is 1/15000.

I'm not sure what the sampling distribution would be under the null hypothesis, though, because there is no variation for a physical constant apart from measurement error. (As opposed to, say, height in humans, in which we observe natural variation.) Assuming that the measurement errors are normally distributed, and that is what the test statistic of 4.1 refers to, then the two-sided p-value would be 2*P(Z >= 4.1) = 1/24204. They may very well have used a t-distribution instead, though, which has fatter tails (consistent with the higher p-value of 1/15000) to account for the population variance being estimated by the sample variance. In that case, the p-value would depend on both the test statistic of 4.1 and the degrees of freedom (i.e., the total number of observations -1).

[_asbestosman]

More seriously, I'm not sure how much this discovery hurts the fine-tuning argument. Before the discovery, were we even sure that it was meaningful to talk about variation in these universal constants, or was it more fixed in the way that pi is the ratio of the circumference of a circle to its diameter in flat Euclidean space? Now we at least have evidence that one the supposed constants can indeed vary.

The evidence that one constant can vary brings up the question about how and why it varies. Imagine if it varied quickly--that might make life much less probable. Does the variation happen randomly the way that some QM phenomena appear to be random, or is it tied to another aspect of the universe or law?

I'm not saying that the fine-tuning argument is solid proof of God, but I don't think evidence of variation in our cosmic constants (or even evidence against it) actually favors either side. Pretty neat though. It makes me wonder if something couldn't be formed in one part of the universe with particular "constants" and then move to another part with different constants (or more aptly, variables). Could different or changing parameters increase the probability of life forming while others increase the probability that it becomes what it is? Could it change what we'll find and how we'll look for it?