Bayesian inference and the teleological argument

Darwin didn’t argue for universal common descent.

1 Like

No, but I occasionally try asking anyway. :upside_down_face:

That’s actually much better than your usual answer - it is on track to what I was asking. I’ll go father than that: It’s better than Dembski, because at least you are thinking about the sort of values that could be used to test a hypothesis.

Any bell shaped distribution supports the design inference.

The Bell curve shape is very natural, arising from the sum of many small random variations (a proof every stats grad student must encounter). The shape itself would not indicate Design.

Perhaps the tighter the shape the stronger the inference.

So a normal distribution with a smaller standard deviation relative to something else? Good! You have stated a distribution and a parameter (the standard deviation) to be tested. I think we could find some context where this might show evidence of human design in certain objects.

The problem in ID is that it generally assumes the existence of the Designer, then concludes what was assumed. Posing a test for Intelligent Design that does not assume the existence of Designer is hard. It seems to require that the designer do something that nature cannot; something not merely unlikely by means of natural evolution, but actually “more likely” by some stated mechanism of Design.

So when you say things like "inference to the best explanation" without stating any mechanism to provide an explanation, people might think you are talking nonsense. :wink:

3 Likes

Close enough: “one or a few”.

Inference to the Best Explanation (ITBE) works wonders in the case of evolution. The alternative to evolutionary mechanisms is Design Intervention. The patterns we see in organisms have only a modest probability P (which is small) under ordinary evolutionary processes. But under Design Intervention they have a probability of 100%, as they are obviously what the designer wanted. So Design Intervention is always the best explanation. Which is why ITBE is often invoked by “design theorists”.

If elephants are large and gray and lumber about the savannah eating bushes, then that is exactly what the Designer wanted. So it has 100% probability. On the other hand if they are small, fuzzy and pink, and flit from flower to flower pollinating them, then that would be exactly what the Designer wanted, so it too would have probability 100%. ITBE eliminates doing a lot of work actually observing organisms, or thinking about evolutionary processes.

4 Likes

Hi Andrew
I think the closest example is a court case where the Jury decides based on a quantity of evidence either for or against the defendant. A part of the prosecutions case maybe testable blood evidence yet another may simply be the testimony of a trusted witness.

Another example is how an investor thinks about which of several alternatives is the best investment. The decision is also based on the weighing the evidence of which alternative will produce the best financial returns.

While probability calculations are valuable they are only occasionally available to help us make a decision.

So no actual method, then. You say there is some weighing of the evidence, but ultimately there is no quantification of any such weights you would have either the inclination to perform or the beginning of a clue how to. You just decide something is an explanation and then just decide it is the best one based on no criteria you are willing to disclose.

Bear in mind, I agree that probability calculations are not always available or helpful, but at least they are something. What you suggest to do instead is just to decide things based on intuition alone. Oh, sure, you say it’s something like how courts or investors work, because you think that standards and rigor is something exclusive to the science lab. It is not. A competent investigator will of course perform something quite like Bayesian analysis, if not in numbers, but a trial decides not what the “best explanation” is, but which of the cases brought before it is subjectively more compelling than the other. They are not in the business of deciding facts, they are in the business of adjudicating law. Investment firms actually do run the maths, do the statistical modeling and probability calculations, because, unlike creationists, they understand that it is possible for them to be wrong, and it will hurt their bottom line to be, so they actually put some effort into not being completely wrong quite all of the time.

3 Likes

I would disagree based on IDcreationists’ behavior.

None of them has ever tested an IDcreationist hypothesis. If they thought that they might even be right, they would be doing so with every penny and minute available. The fact that they, including the ones who used to do real science, don’t tells us all we need to know about their confidence. It’s the reason Bill will avoid computations, too.

1 Like

Oh dear. I think I am now beyond the point of pointing and laughing. I can only weep in sorrow. :sob:

Beyond any specific instructions they may receive from the judge, jury-members have no defined method, let alone a rigorous one.

Investors may, as some above have pointed out, attempt to use rigorous statistical analysis, they may ‘go with their gut’ … they may even throw their money away on Meme Stocks. Truth Social (or more specifically, its parent company) anyone?

My point is that ‘do as jury-members do’ or ‘do like investors do’ is no guidance whatsoever.

Surprisingly enough, the explicitly ‘potentially problematic’ guidance of my Stanford quote above, provides the best guidance to date:

… the [best explanation] must appeal to the so-called theoretical virtues, like simplicity, generality, and coherence with well-established theories; the best explanation would then be the hypothesis which, on balance, does best with respect to these virtues.

Now it could be claimed that a bald assertion of “design inference”/everything “is exactly what the Designer wanted” could be considered to possess “simplicity” and “generality” – albeit vacuously so.

It could not however be considered to be “coheren[t] with well-established [scientific] theories”.

It also runs into the problem that a bald assertion of “design inference”/everything “is exactly what the Designer wanted” could not be considered an “explanation” in its normal English sense, i.e. “That which explains, makes clear, or accounts for … a statement that makes things intelligible.” [OED]

Therefore it could hardly considered to be the best explanation.

4 Likes

Big bang, fine tuning and information storage and processing systems within cells are 3 major discoveries that has provided wind in the sails of theism.

This was @Giltil original claim that in itself was not an ID claim or a probability claim or some combination of both.

Yes, but as I pointed out in the parent thread, only one of these claims is an actual “scientific discovery”. The other two are simply claims made about science, of little (if any) probative value.

And even with the Big Bang, it is unclear why it makes it more likely that an omnipotent being created the universe than that it would occur naturally. Is there any reason an omnipotent being couldn’t, or wouldn’t, create a Steady State universe, for example?

2 Likes

How was it not an ID claim or a probability claim or a combination of both? What was it, in your opinion, a claim about seamanship?

Anyone with basic reading comprehension recognizes that Giltil is claiming that these things make theism more believable than it would be without them. And the claim is made on the back of a reference to fine tuning and information processing, two central arguments brought in defence of ID. What are you talking about?

2 Likes

Actually, it could be claimed that this wasn’t any sort of meaningful claim – just a piece of vague, vacuous, misleading, “windy” rhetoric.

To the extent that it is making any meaningful claim, @Gisteron’s interpretation above would appear to be as close to anything solid as is possible.

2 Likes

Oddball question: could even an omnipotent creator create a universe without the normal distribution?

The (most common form of the) Central Limit Theorem (CLT) states that as long as you have independent and identically distributed (iid) variables, the sample mean will converge to normal. As this is a mathematical proof, it is also a logical certainty. This means, short of preventing the existence of all iid (and thus the existence of dice, tossable coins, etc, etc – and even then some variants of the CLT don’t require iid), it is a logical impossibility to prevent the normal distribution. Even an omnipotent being cannot create a logical impossibility. Therefore even an omnipotent deity cannot prevent the normal distribution from occurring (short of preventing the existence of dice, etc, etc).

Is there anything wrong with my logic? :slight_smile:

1 Like

ID is a more specific discipline and similar to the argument @Joe_Felsenstein addressed.

@Giltil argument was simply that, as you point out, the evidence that IDers use to show evidence of design was helping support the theistic worldview. He was making an inference separate from any probability argument.
Psalm 19

The heavens declare the glory of God;
the skies proclaim the work of his hands.

There are some branches of apologetics literally claiming that without God the universe would be indescribeable with mathematics. Like, if God didn’t exist, counting would not obtain. A non-designed universe would be such that counting would be invalid, and nothing could be counted.

I find that I am unable to imagine such a thing. Any non-empty universe, in so far as it is non-empty, has countable content. I don’t care what the content is, if it exists it can be counted. Even if there is just 1 of them. But if there is 1 of them, we can imagine there could be 2. That means set theory obtains. All the rest of math obtains.

Apologetics is so stupid it is painful. Why do people believe such stupid claptrap?

4 Likes

Unless it’s possible to create a universe where logic itself is fundamentally different? – I thought about this a while; it’s either impossible or is doesn’t make any sense to ask the question. Any universe should end up with something equivalent to ZFC.

It’s pretty good. I would add that variance must not be infinite, which IIRC is required for all versions of the Law of Large Numbers. A universe where everything has infinite variance would be chaos, the very opposite of creation.

It is hard to imagine a body of work that includes Dembski’s fart video, Behe’s alien designers, Paul Nelson’s “ontogenetic depth” vaporware, Ann Gauger’s greenscreen, etc, etc being described as a “discipline” – the word has strong connotations of an orderly and systematic method that the ID clown-car is completely lacking in.

Intelligent Design Creationism is a subset of Christian Apologetics, a ‘field’ with low standards (they let Ray Comfort in after all), and an anything-goes mentality.

4 Likes

Yes, that is why I attempted to frame my argument in terms of logical certainties and impossibilities, in order to avoid such claims.

I find them likewise absurd, and bring to mind a mental picture of God having to be constantly vigilant to ensure that 2 + 2 doesn’t sneak off and become 5.

On the contrary I believe that some degree of naturally-occurring order is unavoidable, and would require constant divine intervention to prevent it – e.g. constantly changing the fundamental constants of a universe, to keep which atomic structures are stable in a constant state of flux, in order to prevent larger structures composed of atoms from forming.

If I recall correctly, God’s constant vigilance is a central tenet of Ed Feser’s universe.

1 Like

Two quotes from that piece:

Moreover, theism holds that the fact that there is any world at all is something that could not even in principle have obtained in the absence of divine creative action.

Or, to put it another way: anything at all exists therefore God

(Which is very close to @Roy’s " Everything supports the design inference")

The point is that the kind of argumentation involved will not be a matter of forming empirical hypotheses and then testing them (using Mill’s Methods, or appealing to probability theory, or whatever). That’s just a category mistake.

That does appear to cut @colewd’s & @Giltil’s ground from under them.