Malaclypse the Younger

According to my formulation of Evidential Arguments, the FTA is an unsound evidential argument for the existence of a designer.

According to my formulation, H represents an hypothesis (to be concluded), E represents a piece of evidence--a proposition that can be known directly (e.g. by perception).

P1: H -> E
P2: ~H -> ~E
P3: E
-------------
C: H

In the case of the FTA, to call the life-friendliness of the universe evidence for the existence of a designer, we must examine the argument in evidential form:

P1: If a designer exists then we would observe a life-friendly universe.
P2: If a designer did not exist, then we would observe a non-life-friendly universe.
P3: We actually observe a life-friendly universe
--------------
C: A designer exists, and created the life-friendliness of the universe.

The Weak Anthropic Principle (WAP) denies the truth of P2: If a designer did not exist, we would still (necessarily) observe a life-friendly universe. Therefore the argument is unsound and the conclusion does not follow.

This analysis does not prove that no designer exists, it merely rebuts the soundness of the FTA as an evidential argument for the existence of a designer.

We must also include the the "we would observe" qualifier. Both E and ~E must be knowable; if one were not knowable, then we could merely switch E and ~E and show that the second or third premise was not possible to know.

It is possible to formulate a somewhat different weak evidential argument against the existence of a supernatural designer.

P1: If supernatural designer does not exists then we would observe a life-friendly universe.
P2: If a supernatural designer exists then it is unlikely we would observe a life-friendly universe
P3: We observe a life-friendly universe
------------------
C: It is likely that no supernatural designer exists.

This argument has many weaknesses, of course; so many that it cannot do more than merely hint at the truth.

The most obvious weakness is that it presumes a supernatural designer. A naturalistic designer would, of course, be constrained to make the physics of this universe life-friendly in order to instantiate life, and the WAP would apply.

It is difficult to coherently define supernaturalism. Most importantly, if supernaturalism were true, then evidentiary arguments in general might not be known to be valid.

It is possible that a supernatural being created the universe for other reasons than the existence of terrestrial life. If so, then the WAP would apply.

P2 of the evidential argument against a supernatural designer might assume an arbitrary definition of physical law; it might be true that the "unlikeliness" of observing a life-friendly universe is merely an artifact of an arbitrary choice.

As Kenny correctly notes, Modal definitions of God make probabilistic arguments suspect on general principles.

We must conclude then, that the FTA is an unsound evidential argument for the existence of a designer, and only a very weak evidential argument against the existence of a supernatural designer, entirely dependent on arbitrary assumptions.

[ March 04, 2002: Message edited by: Malaclypse the Younger ]</p>

Kenny

Actually, that is not my position. I assume your are referring to my comments over in the <a href="http://iidb.org/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=50&t=000135&p=4" target="_blank">Fine Tuning Analogy Thread</a>

My comments over there were not intended to invalidate evidential arguments for and against theism, but to demonstrate the necessity of distinguishing between various philosophical notions of probability such as objective probability and epistemic probability (see link for explanation). Evidential arguments, I would allege, are generally primarily concerned with epistemic probability as opposed to objective probability. I believe that the FT argument is included among such arguments.

Anyway, I just wanted to clarify my position.

God Bless,
Kenny

Tercel

Malaclypse the Younger,
Are you still on about evidential arguments? I would have thought your, um... not entirely sucessful... post on the Philosophy Board some time ago might have convinced you that you really don't know what your talking about. But at least you've got the name right this time.

Is there any reason we should accept your formulation? That is to say, Evidential Arguments are a standard class of argument, and what makes you think that you have the authority to formulate how they should proceed?

This definition would seem to include only half of all Evidential Arguments: What about probabilistic arguments?
eg Something like:
P1: H =&gt; E is likely with probability X
P2: ~H =&gt; ~E is likely with probability Y
P3: E
-------------
H is confirmed to a degree calculable by Bayes' Theorem.

And of course the FTA doesn't work when formulated that way. It's a probabilistic argument and so when you formulate it non-probabilistically you wreck it. Big deal.

Malaclypse the Younger

Kenny

I'm not specifically referring to your position. I'm referring to the FTA in general.


Tercel

It should be noted that I do not take your opinions about logic and philosophy at all seriously; I doubt that many do--certainly few who do not share your biases.

It should be noted that the criticism on that thread mostly involved the metaphysical interpretation of successful instances of evidential arguments. The other criticisms seemed to address either arbitrary minutia or rebut points not actually made in the OP.

Indeed, bringing up minutia such as the arbitrary labelling of this form of argumentation is a not-so-subtle dig and an example of the fallacy of poisoning the well; it is unsurprising that you employ this tactic here.

The form of the argument is obviously logically valid and it clearly shows the incorrect assumption of the FTA.

It should be noted that no sound FTA has yet been presented which concludes it is true (or even more probable to believe) that the universe was designed. As such, the advantage of this form of the evidential at least makes the unjustified assumption clear.

The probabilistic formulations add little to the discussion except for additional complexity. Note that the probabilistic form can be represented in the predicate form by representing "highly likely" and "highly unlikely" as predicates. In which case, the predicate form can be represented as:

P1: If it is likely a designer exists then it is highly likely we would observe a life-friendly universe.
P2: If it is likely a designer did not exist, then it is highly unlikely we would observe a life-friendly universe.
P3: We actually observe a life-friendly universe
--------------
C: It is higly likely a designer exists, and created the life-friendliness of the universe.

P2 is still obviously false: It is false that if it is likely a designer did not exist, then it is that it is highly unlikely we would observe a life-friendly universe; it is definitely true that if a designer did not exist, we would observe a life-friendly universe.

The Baysian Probability arguments that conclude design are unsound. Most are rebutted in <a href="http://quasar.as.utexas.edu/anthropic.html" target="_blank">The Anthropic Principle Does Not Support Supernaturalism</a>. Kenny's Bayesian formulation (which defines "designer" in terms of life-friendly constants) provably requires that one hold the a priori probability of the existence of such a designer equal to the a priori probability that this universe arose by chance leading to the conclusion of uncertainty. Your non-Bayesian formulation requires the arbitrary assumption of parameters, which renders it trivially unsound.

The FTA for design fails according to both the probabilistic and predicate form of evidential argument. It's just easier to see the flaw in the predicate version.

[ March 04, 2002: Message edited by: Malaclypse the Younger ]</p>

Automaton

All the fine tuning argument can do is show us that if God existed and created the universe, he would need to fine tune it. It is flawed reasoning to say it demonstrates the existence of said God, as their argument against NC (naturalistic causation) rests on an unsupported assumption: That if the universe was formed naturalistically, it would be a unique and singular event. But if one assumes NC, there is absolutely no reason at all why it should be a unique event, and infinite big bang events means exhaustion of all possible physical law configurations, and thus the fine tuning argument against NC fails.

Tercel

That's a rather strange way of putting it. You seem formulating the argument in such a way that it's subjected unnecessarily to the Anthropic Principle objection... oh, wait... that's what you're trying to do.
You might want to try replacing the words "We observe" with "there exists". Not to mention that I think you've got one too many "likely"s in both P1 and P2.

Of course, IMO, the above argument would still not be particularly good as it attempts to prove too much and is vunerable to attacks of "necessity" and "many-many-worlds" against P2.
So I would formulate the FTA as follows:
P1: If chance was responsible for the creation of a universe then it is very highly unlikely the created universe would be life friendly
P2: If a designer was responsible for the creation of a universe then it is reasonably likely the created universe would be life friendly
P3: There exists one and only one universe
P4: The universe is life friendly
----------
C: The chance hypothesis receives a significant degree of disconfirmation.

Vorkosigan

Let's see....

P1: If chance was responsible for the creation of a universe then it is very highly unlikely the created universe would be life friendly

Numerous problems here:

False dichotomy: the choice is not between "chance" and "design" but between "processes unknown" and "Design"

highly unlikely: no one knows the probability of life arising in any universe. So no statement about probability may be made. Do you have a statistically valid sample of universes?

life-friendly: for all we know, this universe IS unfriendly to life. Other universes may be much more fecund. Do you know of the condition of life on a statistically valid sample of the trillions of objects in the universe?

life: Why is "life" the problem? We can reconfigure this to say If chance was responsible for the creation of a universe then it is very highly unlikely the created universe would be friendly to beautiful lightning bolts on Jupiter.

P2: If a designer was responsible for the creation of a universe then it is reasonably likely the created universe would be life friendly

Drivel. Nothing is known about the nature of the Designer. No one knows why the universe was Designed, or even if Design was the actual intention of the Designer. For all you know, our universe is a side effect of some other process of the Designer(s), and nothing in it is intentional.

P3: There exists one and only one universe

Prove it.

P4: The universe is life friendly

So far, we know of life on only one body in the trillions in our universe. Is the universe life-friendly? No one can answer that question.

In any case, you have not yet demonstrated that we should take the existence of a set of material objects and processes we call "life" as indicative of anything, let alone the reason the universe was created.

In any case, as I have repeatedly said, you misunderstand constraints and selection processes. Fine Tuning is the inevitable result of selection processes operating under natural law. Regardless of what kind of universe it is, as along as it has natural laws, it will appear Fine Tuned.

Michael

bd-from-kg

Tercel:

There are several problems here:

1. To say that the chance hypothesis “receives a significant degree of disconfirmation” is rather ingenuous. I note that you don’t say that the design hypothesis is more probable” than the chance hypothesis, even after taking FT into account, only that it’s more probable after taking FT into account than before. This could mean merely that the likelihood of its being true goes up from 1 in 10^1000 to 1 in 10^999. To make FTA a serious argument for theism you need to estimate the prior likelihood of the chance and design hypotheses and justify those estimates.

2. The design hypothesis looks very much like a “tailored” hypothesis. That is, it is designed (pun intended) to explain the supposed fine tuning and has no other function. Hypotheses of this kind have a dismal track record, for obvious reasons. For example, if we lived in a pre-scientific society, I could make a similar argument along the following lines:

P1': If natural phenomena (e.g., earthquakes, eclipses) were accidental (or “chance&#8221 byproducts of the way the universe happens to work it is highly unlikely that there would be such a thing as thunder.
P2': If a thunder god is responsible for thunder it is highly likely that there would be such a thing as thunder.
P4': There is such a thing as thunder
----------
C': The chance hypothesis receives a significant degree of disconfirmation.

This argument appears to be perfectly valid. The problem is that the prior likelihood of the existence of a thunder god is ridiculously small. Of course, it wouldn’t seem all that improbable to us if we lived in a pre-scientific society. But by now the absurdity of this kind of argument should be obvious to any halfway educated person.

Now you might argue that the “designer” is not a tailored hypothesis. After all, the supposed Designer is also the Creator and Lawgiver. But this is rather like saying that the thunder god is not a tailored hypothesis because, as it happens he is also the volcano god, the earthquake god, and the flood god. Tailoring a hypothesis to kill several birds with one stone doesn’t make it any less tailored. And there is no apparent reason why the Designer, the Creator, and the Lawgiver have to be the same entity any more than the thunder god, volcano god, etc. have to be. Each of these functions seems to be designed to “explain” some aspect (or supposed aspect) of the world. None of them has any evidential support other than the supposed need for an explanation of these aspects.

3. The FTA is just the latest version of the Argument from Design. If people in, say, the 18th century had been familiar with Bayes’ theorem they could have made the Argument from Design very scientific-looking by dressing it up with probability notation. But it was garbage. A whole series of scientific advances has demolished one version of the Argument from Design after another. The FTA is the Argument from Design’s last stand. Why should we believe that this one will not also be demolished by yet another scientific advance?

4. P3 is completely unsupported. There is no reason to believe that there is just one universe. The best that could be said is that we have no idea how many universes there are, and even this claim is highly debatable. More on this in a later post.

[ March 04, 2002: Message edited by: bd-from-kg ]</p>

Tercel

Whether design becomes more likely than chance after the FTA has been taken into account clearly depends on what a priori values one assigns to these things. What is clear is that the chance hypothesis is deconfirmed to a very very large degree. I suggest and justify some exact values for a priori probabilities in the thread <a href="http://iidb.org/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=50&t=000135" target="_blank">Fine Tuning Analogy</a>.

In general arguments from Design were against evolution and depended on emotional feelings rather than any particularly data. By contrast that FTA is neither and incorporates scientific data.

Well, duh. Obviously one can escape the conclusion by asserting that there exist sufficient universes that at least one of them must be able to sustain life. The whole point of the FTA is to render the “There exists only this universe and it is the way it is by pure chance” view untenable. Since this is the view taken by quite a large proportion of atheists, I’m happy to settle for demolishing this position. Obviously the atheist could adopt another position, thus escaping the argument but the beauty is that none of these alternate positions are particularly comfortable (hence, I assume, the large acceptance of the above position). The two main alternative positions are that the universe is a necessity, or -as you point out- that there are other universes. If the atheist accepts that there are other universes then they would appear to lose the oft-praded Occam’s Razor advantage over the theist. Not to mention they would be accepting the existence of an invisible, undetectable entity with no proof whatsoever - so they’d have to stop accusing theists of that. Furthermore a multiple-universe explanation would appear to be ad hoc, akin to the God of the Gaps of the theists.
Similarly, that the universe is a necessity is not an easy hypothesis to swallow. The scientific data supporting this idea would appear to be zero. Not to mention that the idea seems to show unwarrented model skepticism. Scientists can model other universes without any contradictions appearing. As far as it can be determined there are six basic physical constants whose relation to one another is entirely arbitrary. And finally, one of the basic ideas which fuelled the advance of science was that the universe was a free creation of God and not necessary: Thus implying that the nature of reality could not be determined by thought alone but required empirical observation.
Certainly if the atheist accepts one of these two hypotheses, they can hardly criticise the theist for accepting God as the explanation.

So, with this in mind, I aim to use the FTA only to destroy the single universe-chance hypothesis, and leave the rest hanging.
Tercel

bd-from-kg

Tercel:

It seems to me that you failed to answer any of my points very convincingly.

In reply to my first point you said:

Well, yes. That was my point.

Not so. You suggest some values; you do not justify them.

As to my second point, about D being a “tailored” hypothesis, you offered no reply at all.

Perhaps I failed to make this point clearly enough. Here are some examples of the kind of hypothesis I’m referring to:

(1) Some drugs tend to make people sleep because they contain the “dormative principle”.
(2) Some things shine on their own because they have “radiant potential”.
(3) Some people commit crimes because they have a “criminal nature”.

The list could be extended indefinitely. In each case, the “explanatory hypothesis” has certain tell-tale features:

(a) It was invented for the sole purpose of explaining the phenomenon in question; it doesn’t explain anything else
(b) The only evidence that something has the postulated quality (dormative principle, radiant potential, or whatever) is that exhibits the phenomenon purportedly being explained. No independent test is proposed for detecting this quality.
(c) The hypothesis yields no predictions.
(d) The hypothesis yields no insights into the nature of the phenomenon in question.

Now D has all of these features: It was invented to explain the supposed fine-tuning and explains nothing else. The only evidence for the supposed designer is the supposed design. It yields no predictions. And it yields no insights into the nature of the supposed fine-tuning. (That is, if the universe appeared to be fine-tuned to allow some other supposedly desirable phenomenon, the hypothesis would simply be modified to say that the “designer” had deliberately designed it that way. And, of course, it offers no clue as to why these specific values of the physical constants were chosen out of all the combinations that allow for life.)

Now hypotheses of this kind simply cannot be taken seriously. When such a hypothesis is proposed as an explanation for some phenomenon, the rational response is not to assign it a prior probability of, say, 10^-50, but to ignore it entirely. Such hypotheses always have the attractive property that the probability of the phenomenon in question given the hypothesis is essentially 1, but they have the less attractive property of always being false.

Now let’s move on to your reply to my third point:

Nonsense. The Argument from Design has a long history. Various versions of it were used long before anyone had ever heard of evolution. It is far from being an emotional argument; it was a fairly persuasive argument before scientists spoiled all the fun by finding naturalistic explanations for one seemingly “designed” aspect of the universe after another. Many of the Founders, such as Thomas Jefferson, rejected Christianity but accepted Deism largely because the Argument from Design seemed to them to be unanswerable.

And finally, my fourth point:

”One” need not assert anything. This is a premise of your argument. Since you’re the one making the positive argument, the burden is on you to offer convincing evidence that your premises are true.

In that case you can’t claim to be offering an argument for theism. Your argument at best shows that if certain unsupported assumptions happen to be true and if P(D) is arbitrarily set high enough, then P(D|E) &gt; P(C|E).

Also, I seriously doubt that this is the position taken by most non-theists. I suspect that the position taken by most of them is that they have no idea how many universes there are (beyond the fact that there is at least one). Perhaps a large proportion of non-theists who have “views” on matters about which they know nothing believe that there is just one universe. Those of us who are more scientifically minded have no views on questions for which no evidence is available.

However, this is not a question on which no evidence is available. Here, for example, are some interesting passages from the article Guth’s Grand Guess in the latest (April 2002) issue of Discover magazine.

In a sidebar the article adds:

[“Guth” is of course Alan Guth, the architect of the inflation hypothesis. To say that he is a reputable physicist is something of an understatement.]

Now I’m not quite as confident as Guth that there are infinitely many universes. But if the inflation hypothesis is correct (and the evidence is leaning strongly that way) it makes a lot of sense. It’s hard to see why a process similar to the one that created this universe would have happened exactly once. It seems far more plausible that any possible process would occur indefinitely often.

Since I said nothing about this idea I’m not sure why you bring it up. But I see no reason why the hypothesis that the universe is necessary should be “difficult to swallow”, especially for a theist. If the universe was created by a necessary being, it would seem to be necessarily necessary. But be that as it may, the proposition that the universe is contingent is yet another premise of the FTA, so it is up to the proponents of this argument to give cogent reasons for accepting it. “Difficult to swallow” won’t cut it. It’s “difficult to swallow” that space and time are interchangeable, that matter and energy are the same thing, that a single particle can interact with itself in such a way as to produce interference patterns, etc. Apparently it was difficult for most people to swallow the idea that the earth is not the center of the universe, but is rather a tiny speck in an unimaginably huge universe. It is still difficult for many people to swallow the idea that our species is an accidental byproduct of an unplanned, undirected natural process. Frankly, what’s “difficult to swallow” seems to have very little correlation with truth.

Also, “pure chance” and “necessity” are hardly the only possibilities regarding the physical constants. For example, it may be that there are relationships between the constants. That is, it might be that some of them can be computed from the others. Thus the number of free parameters might be less than it now appears. This would have a profound effect on P(E|C). For the FTA to work you have to add a premise to the effect that this isn’t the case (or else arbitrarily adjust P(D) upward by many orders of magnitude).

That’s absurd. In no sense is this assumed by the scientific method, nor did this idea in any way “fuel” the advance of science.

This is a perfect illustration of the fundamental difference between the way the minds of the typical theist and the typical nontheist work. The typical theist takes it for granted that for every datum one must accept some explanation. A simple admission of ignorance is just unacceptable. You have displayed this mindset repeatedly by insisting that in order to “escape” the FTA the atheist “must” accept this or that hypothesis. (SeaKayaker’s <a href="http://iidb.org/cgi-bin/ultimatebb.cgi?ubb=get_topic&f=50&t=000104" target="_blank">Impossibility of Neutrality</a> thread is a classic illustration of this same mindset: neutrality about the existence of God, he claims, is really just another way of taking sides.) The typical nontheist is perfectly willing to suspend judgment on questions which he does not have enough information to decide.

Atheists generally do not “accept” either the hypothesis of many universes or the hypothesis of necessity. (The former idea does have some support, but is still rather speculative – which is to say that there is some evidence for it, but not, IMHO, enough to justify rational belief.) And they are perfectly justified in criticizing theists for accepting a hypothetical entity as the “explanation” of anything in the absence of convincing evidence that it even exists.

[ March 08, 2002: Message edited by: bd-from-kg ]</p>