Thanks, Chris, for yet another excellent post, most of which I'm in full
agreement with. (I just wish I could have expressed the arguments so well
myself.) There are, however, a few points on which I differ with you.
>Chris
>Another little essay follows. This one is prompted by the fact that the
>drivel known as the "fine-tuning" argument is still popular with the sorts
>of people who read and believe the stories of aliens and God in the
>National Enquirer, and by the fact that Jones has repeatedly made approving
>mentions of it, despite my having pointed out some of its more obvious
>flaws after the first few such mentions. At the end, the reader will find a
>nice summary of some of my points and a list of "morals" to be gotten from
>the critique of the fine-tuning argument. If the reader keeps these
>"morals" firmly in mind when evaluating arguments like the fine-tuning
>argument, he will avoid (or correct) *many* grievous errors.
>
>- - - - - - - - - - - - - - - - - - - -
>
>The number of "universes" does not need to be infinite. It only needs to be
>a very *large* number of other universes to increase the probability of
>ones like ours to nearly one. Further, even if it *does* have to be
>infinite, it is *still* a less radical proposal than that there is a God
>(who is *also* claimed to be infinite (and, via infinite knowledge/
>intelligence, also infinitely complex). As I've pointed out before,
>*NOTHING* is gained, in Occam's Razor terms, by introducing God as an
>alternative to naturalistic theories, because the *same* considerations
>that might count against naturalistic explanations *also* count against
>theistic ones, with the theistic ones having a the additional burden of
>proof of their radical *metaphysical* claims. It is trivially easy to prove
>that NO theistic explanation can *ever* be simpler than the best
>naturalistic explanation, because the theistic explanation always instantly
>introduces the need for a *metaphysical* realm that naturalism does not
>need __ and nearly always a whole host of other claims about the alleged
>God that need proof or coherent conceptualization, as well.
I guess that the "metaphysical realm" you refer to is what is more commonly
known as the "supernatural realm". I think it's misguided to make a
scientific argument against a supernatural realm (which is what you're doing
when you invoke Occam's Razor) without having some definition of what
"supernatural" means. As far as I can determine, the term is generally used
as a label arbitrarily affixed to certain phenomena in an attempt to give
those phenomena a privileged status so that they don't have to be subjected
to the usual processes of rational/scientific thought. Most attempted
definitions of "supernatural" are circular, such as "that which is not
natural" (and "natural" means "that which is not supernatural"), and
"non-material" (where "material" is a synonym for "natural").
I can think of two possible non-circular definitions of "supernatural", but
neither is of any use to those who wish to claim that the supernatural is
some distinct realm.
Definition 1. The "supernatural" is simply that which is not yet explained,
or that which lies outside the current scientific paradigm. By this
definition, "supernatural" is a temporary state. As our knowledge improves,
supernatural phenomena will tend to become natural.
Definition 2. A "supernatural" phenomenon is one which does not have any
cause, and is therefore fundamentally inexplicable. It may be that the
Cosmos (meaning everything that has ever existed) is such a phenomenon, but
that would be the case whether the Cosmos started with God, the Big Bang or
some metaverse which spawned multiple universes.
A theist might argue as follows: if the origin of the Cosmos is
fundamentally inexplicable, why should we apply Occam's Razor to it? Isn't
it just as plausible that the Cosmos began in a complex state (with God) as
in a simple state (with a Big Bang)? My answer would be that
science should not give up, but should proceed on the assumption that a
solution to the paradox (of what came first) will eventually be found. If
there *is* a solution, then Occam's Razor should apply to it.
If anyone here can supply a better definition of "supernatural", I would be
delighted to hear it.
>Further, the *regularities* of a universe (ours or any others) never needs
>any explanation except as a *philosophical* issue. It is not even logically
>possible to coherently describe a universe without regularities. This
>derives almost directly from the law of identity. For things to be *truly*
>irregular (to obey *no* laws of physics of any kind), they'd have to have
>*no* causal interactions, no causal relationships between them or within
>themselves, no actual *particular* attributes of any kind. There would also
>have to be no such thing (within such a universe) as change or time; each
>"instant" or state would have to be *absolutely* disconnected from all
>other instants or states. If we could detect what was going on in such a
>universe, we would see *pure* informational "noise," with the only
>"regularities" being purely random occurrences of succeeding states that
>*we* might see as being similar to causal patterns or patterns that arise
>out of causation.
>
>What *is* true, of course, is that there is no apparent metaphysical
>requirement that a universe be such at any time as to support any form of
>life. But, if, as some suggest, *our* type of universe is the only
>logically *possible* type of universe (physically), then it would follow
>that the "fine_tuning" *does* have a metaphysical/logical basis, thus
>totally obviating the explanatory value of introducing God (since, given
>this basis, even *God* could not build a universe that was essentially
>different from ours, since any such universe would be logically
>contradictory).
This assumes that logic is an absolute, and that there is no possibility of
a universe with different rules of logic! But I think we're entitled to make
this assumption, aren't we?
I'm reminded of Carl Sagan's novel "Contact", in which the protaganist finds
a message encoded in the digits of the number pi. The value
of pi was chosen by the constructor(s) of the universe!
>About the fine_tuning argument as evidence for designConsider *any*
>imaginable universe, with or without life. It is obvious that we could look
>at such a universe and find facts about it that the fine_tuning argument
>would apply to, facts that are such that, if some basic aspects of said
>universe were different, would not be present.
>
>That is, given *any* set of basic constants and factors that make up a
>universe, there will be *some* fact or set of facts in or about that
>universe which would *not* exist if one of those basic facts were changed
>just slightly. In one universe, for example, the gravitational force (if
>any) might be *much* weaker than is needed to balance the speed of
>expansion, so that all matter is expanding at such a vastly higher rate
>that no galaxies and star systems form. Well, *by the logic of the
>fine_tuning argument,* this fact would *prove* that there is a God who
>*designed* the universe to achieve just this result. After all, if the
>expansion had been any slower, it would not be expanding at just the rate
>it is in this hypothetical universe. Other, much more sophisticated
>examples could be offered, but the reader with a background in physics can
>easily make them up by the dozens. The point is that the precise condition
>or state of *any* universe (and the things in it) depends on whatever
>factors bring that state about, and therefore any change in any one of
>those basic factors will bring about *different* states.
But the only result we're interested in here is the origin of intelligent
life (intelligent enough to ask why the universe is the way it is). If the
universe was not such as to produce that result, we would not be here to ask
the question. This is the argument made by the proponents of fine-tuning,
and I think that this part of their argument is valid. But they must allow
for *any* form of intelligent life, not just life-as-we-know-it.
>The fallacy of the fine_tuning argument is that it assumes that there is
>something "special" about life as we know it. But, life is another
>phenomenon that may or may not occur, along with millions of others that we
>could think up or generate by computer simulations based on varying
>whatever facts we take as basic to the nature of our universe, or creating
>new ones.
>
>There is no "obvious" inference to the existence of a designer, except in
>the primitive person's "obvious" inference that things he doesn't
>understand are controlled by gods. For one thing, we don't even *know* what
>the basic "constants" (etc.) of our universe are, and we also don't know
>how they are related. Unless it can be shown that our universe is special,
>we have no basis at all for any such inference. In fact, we don't even need
>to suppose that there *might* be (or have been) other universes. The
>fine_tuning argument is based on bad logic, arbitrary premises, and rampant
>speculation posing as established fact.
>
>Unfortunately for Jones' position, the article he quotes does *not* show
>that our universe "peculiar" in any relevant sense, so there is no need to
>invoke multiple universes. The purpose of invoking such a possibility to
>begin with is simply to point out the unsoundness of the fine_tuning
>argument's probability_premises, but such a refutation of the argument is
>not the only one, and it could be totally dispensed with if it turned out
>to have real problems.
>
>In any case, it is invoked only as a proof of the unsoundness of the
>fine_tuning argument, not as a proof of any positive claims. That is, no
>one who invokes it needs to actually *assert* that there are other
>universes. The point is that, if there are, and even if the *other*
>premises of the fine_tuning argument are accepted, the fine_tuning argument
>fails. Thus, the fine_tuning argument must prove that *our* universe is the
>only universe or that it is one among only a few that exist or ever *have*
>existed. It assumes this burden of proof by the very nature of its
>premises, one of which is to the effect that an infinite or very large
>number of other universes are *possible.* The question is then, Why is
>there *this* universe rather than one of the huge number of other possible
>universes?
>
>At this point, the atheist needs only to point out that the fine_tuning
>argument *itself* asserts that other universes are *possible*, and that all
>he's doing is suggesting that they might be *actual* as well. That is, the
>idea that there are possibly an indefinitely large number of other
>universes is a direct implication of one of the fine_tuning argument's
>*own* premises.
>
>This means that the *same* Occam's Razor "problem" that Davies claims for
>the multiple universes idea applies also to the corresponding fine_tuning
>argument premises. The fine_tuning folks have to prove what amounts to a
>contradiction: That an indefinite number of universes are in fact
>*possible*, but that not even a mere few trillion trillion trillion of them
>are *actual*. *If*, as they claim, such a large number of universes *is*
>possible, on what grounds can they possibly positively deny that at least a
>few (relative to infinity) of them might *actually* exist?
>
>Since the fine_tuning argument is a probabalistic argument that concludes
>that *our* universe is somehow too improbable to be a chance occurrence, we
>may ask, "Well then, what is the *probability*, for any of the other
>universes you claim are possible, that it actually exists? Obviously, if
>you make the probability too low, then it becomes pointless for you to
>argue that they are possible alternatives to *our* universe. That is, if,
>for one such possible universe, you claim that it has a probability of
>existing that is effectively indistinguishable from zero, then it is also
>*effectively* impossible. If you apply such a claim to *all* such
>alternative universes, and make your claim strong enough to make it
>effectively impossible that there *are* such alternative universes, then
>*our* universe effectively becomes the only *possible* universe, thus
>eliminating the need for fine_tuning.
>
>On the other hand, if you claim that the probability of each of these other
>universes actually existing is very *high*, then you implicitly *accept*
>the multiple_universes premise as being viable, in which case, the
>improbability of our own universe vanishes.
>
>After all, what it means to say that something (such as a universe) is
>genuinely possible is that it *might* actually exist or have existed or
>come to exist in the future. If the fine_tuning folks want to deny that any
>of these universes exist, they will need some positive evidence to restrict
>the implications of their own possibility premise. They need to say (and
>prove) something like: Though an indefinite number of other universes are
>individually possible, only *ours* (and perhaps a *small* number of others)
>does exist or ever has existed." This is necessary, because, otherwise, the
>implication of their possibility claim is that some or all of these other
>universes might actually exist.
>
>What, other than the kind of provincialism that led the ancients to believe
>that Earth was the center of existence, could we possibly use to prove such
>a claim?
>
>A special case occurs if they claim that an *infinite* number of
>alternative universes might each be possible. In this case, no matter *how*
>small the probability of any one of them existing is, if it is not *truly*
>zero, then there is almost *certainly* an infinite number of such
>alternative universes (the probability is *infinitesimally* less than one).
>
>In summaryWhile we can't prove (yet) that other universes exist, neither
>can the fine_tuning folks prove that other universes are *possible*. If
>they *do* prove that other universes are possible, they will *then* need to
>prove that they don't actually exist. They bring this burden of proof on
>themselves by adopting the initial claim that other universes are possible.
>
>Davies' (and Jones') claim is thus hoisted on its own little petard.
It seems to me that, in the above argument, you may be failing to
distinguish between two different hypotheses:
1. Other configurations were possible for the one and only universe.
2. There may be more than one universe.
Having said that, I'm not sure that hypothesis 1 is meaningful. If this
universe is all there ever was and ever will be, what does it mean to say
that there were other possibilities? I think we're back here to the question
of whether the universe had a cause. If it didn't, then I don't think it
makes any sense to say that there were other possibilities.
>As is the fine_tuning argument in general, though there are other major
>objections to it, such as the fact that two or more basic constants or
>other "design" factors might be varied in *complementary* ways to produce
>an indefinite number of other "universe designs" that might support life,
>and the fact that some of the claimed aspects of "fine_tuning" are or might
>be merely different ways of regarding *one* basic fact, not two or more.
>For example, if the *mass* of the universe is *inherently* related to the
>rate of expansion and/or the gravitational force, then varying whatever is
>the source of both may *always* produce a universe that can support life.
>The rate of expansion balances the gravitational force, but if the initial
>rate of expansion *creates* a matching gravitational force, then the
>probability of the rate of expansion "balancing" the gravitational force
>becomes one, and thus not remarkable at all. In this case, the fine_tuning
>argument would then have to claim that the underlying common causal
>relationship must be improbable. This would be fine, except that, if the
>number of basic factors becomes so small (such as one) that *their*
>particular values no longer sufficiently improbable, then the fine_tuning
>argument still fails.
>
>In general, arguments of the fine_tuning sort depend on a raft of premises
>and assumptions that need to be individually validated. In daily life, we
>can often accept such premises easily, because we have life_long experience
>with the relevant facts. However, even in such cases, people are
>*notoriously* likely to make seriously wrong judgments about such
>probabilities, assuming "design" or some similar conclusion when the facts
>involved in the particular case are merely coincidental. The assumptions
>that people carry around about probabilities are often incorrect.
>
>But, in the case of the fine_tuning argument for the existence of God, even
>the *ordinary* factual "controls" on error are missing, and the common
>assumptions of daily life do *not* have even the weak and sporadic support
>that they do when applied to daily life situations.
>
>What's the probability that of 23 randomly selected people there will be
>two or more with their birthday being the same month and day? Most people
>think that it's quite small, but it's actually slightly greater than 0.5.
>And this is an elementary example, where all the relevant facts are known.
>In the case of the fine_tuning argument, crucial premises are simply
>assumed or claimed on the weakest of factual bases, because we simply don't
>*have* the relevant facts, and do not seem to be on the verge of getting
>them any time soon (even with post_Hubble telescopes, the Large Hadron
>Collider, and so on). Part of this is simply a result of the current state
>of physics and cosmology, etc. But part of it is the result of a
>fundamental epistemological problem: No matter what we find about our
>universe scientifically, we still won't be able to logically exclude the
>possibility of something naturalistic *outside* of it. Except in the
>broadest sense of "universe" (to mean: everything that exists
>naturalistically), we cannot, from the inside, determine that there is
>*nothing* more than what we know about. We may not be able to claim that
>there *is* something, but we cannot claim that there *isn't* something,
>either. The fine_tuning argument and the "multiple universes" refutation of
>it *both* depend on something outside of the known universe (either other
>universes or God), but the empirical facts will not specify *which* might
>exist. (And, because of the radical metaphysical claim, the assumption that
>any such "something" is naturalistic is better from an Occam's Razor point
>of view than the conclusion that that "something" is an infinite and
>unexplained God who can create universes out of nothing.)
>
>In short, the fine_tuning argument fails in almost all imaginable respects
>in which such an argument *can* fail. Jones' continued support for this
>logically invalid and foundationally unsound argument is a testament to his
>unwillingness to analyze it critically. And Davies simply appears to have
>long ago sold out serious thought about such topics in favor of cheap
>popularizations and facile but ultimately flimsy sensationalism (it
>certainly sells more books to make such statements as the one Jones quotes
>than it does to point out that such claims are unjustified by any known
>rules of inference from any known facts).
Let me give you my own argument against the probabilistic fine-tuning
argument.
As I understand it, the fine-tuning argument is an argument against the
hypothesis that the parameters of the universe are random variables. The
argument is that, under this hypothesis, the probability of the parameters
being such as to support intelligent life is extremely small, and therefore
we should reject the hypothesis.
This argument, incidentally, follows the logic of Dembski's "Law of Small
Probability" ("specified events of small probability do not occur by
chance") , as described in The Design Inference. In Dembski's terms, the
origin of intelligent life is a "specified event". Dembski rules out the
relevance of multiple universes, but his argument for doing so is flawed.
The problems with the fine-tuning argument, as I've stated it, are as
follows:
1. We don't know that the parameters of the universe are random variables.
Indeed, if there is only one universe, then it's not clear that the concept
of probability has any meaning with regard to its initial state.
2. Even if the parameters *are* random variables, we don't know their
probability distributions and interdependencies.
3. The argument doesn't allow for the possibility of multiple universes.
We don't need to appeal to multiple universes, because points 1 and 2 are
sufficient to scupper the probabilistic argument.
However, there is another, more interesting (to me) argument. Let's suppose
that there is only one logically possible configuration for the universe.
Isn't it amazing, the argument goes, that this one possible configuration
happens to be just right for the origin of intelligent life? Since this
isn't a probabilistic argument, it's hard to know how to interpret "amazing"
in this context. Suggestions would be welcome!
>The morals of all this include
>
>1. Check the premises of arguments. The fine_tuning argument supporters
>seem to *love* to use unstated premises as established fact. Dig out these
>premises and examine them and their real implications.
>
>2. Check the *logic* of arguments. Unless the fine_tuning argument can both
>reasonably assert the *possibility* of an infinity (or very large number)
>of other universes *and* the claim that none (or incredibly *few*) of such
>universes *do* exist, the fine_tuning argument is invalid because it does
>not soundly exclude a perfecty viable alterntive to its own conclusion.
>
>3. Watch for gross *abuses* of current scientific theories and hypotheses.
>The ones most radically and virulently abused by "fine tuners" are "basic"
>constants and relationships among them and other such basic facts of our
>"known" universe (such as the amount of matter in it, and the speed of
>expansion, etc.) Most of these "constants" and basic facts are, at best,
>poorly understood. They don't provide support for *any* special
>metaphysical claims because we don't know *their* metaphysical basis, we
>don't know how "deep" they go, and we don't even know if they *are*
>constants (and, if they are, whether they are constant across all possible
>universes or just ours, etc.).
>
>4. Watch for attempts at subverting Occam's Razor by mis_comparing
>more_specific propositions ("There may be a large number of other
>universes") with vaguely stated alternatives ("Design"). Consider what the
>alternative claims *really* involve, and *then* ask yourself what Occam's
>Razor would say. In this case, the "design" hypothesis expands (minimally)
>to something like: An infinite being with infinite knowledge and infinite
>power (and an infinite *number* of distinct powers) and who can create an
>infinite number of universes out of nothing, and who utterly metaphysically
>transcends ordinary naturalistic reality, including space and time.
>
>*This* statement of the design hypothesis (even in this small form) does
>not compare well, in Occam's Razor terms with the simple hypothesis that
>there *might* be a very large number of other universes (or perhaps our own
>universe in various past "incarnations"). Whenever you hear a theist making
>an Occam's Razor claim for theism, consider what theism *really* entails,
>and the Occam's Razor claim will be seen not only to be unsupportable, but
>that Occam's Razor, insofar as it provides support for any theory, provides
>support only for non_theistic theories in all such cases (this is true even
>when the claimed "God" is *not* claimed to be infinite). Remember, by the
>usual definition provided by theists themselves, the existence of and the
>claimed nature of God is the most *extreme* of nearly all imaginable claims
>that are not outright and directly self-contradictory. Compared to the
>pathetic little process that evolutionists are claiming, the God claim is a
>*monstrous* violation of the entire *idea* of parsimony and reasonableness
>of hypothesis as related to the facts to be explained. It is *worse* than
>the police, upon finding a person who has been killed and who has what
>appears to be a bullet hole through his head, concluding that he was killed
>by psychic means because the *sentence* "He was psychically killed" is
>simpler than the *sentence* "Some human with a forty-five caliber pistol
>and some bullets shot him."
>
>Occam's Razor basically only applies in comparing two theories if the two
>theories *do* the same things, or in which the more *powerful* theory is
>compared favorably to another, weaker theory. Since design theory does not
>begin to explain evolution in predictive terms (for one example: on the
>basis of the design premise, ID folks can't predict that animal breeding of
>an as-yet-unknown animal will work, whereas evolution can make such
>predictions on the basis of its basic principles). Since evolutionary
>theory is *vastly* simpler in principle than the claim of theism, and since
>it is a *weaker* theory (in predictive terms), it will always lose to
>evolutionary theory.
>
>Incidentally, though it helps *immensely* to replace theistic design with
>intelligent design done by aliens or "metaverse" beings, etc., it doesn't
>help *enough*, because, given the current evidence for design (essentially
>none at all), it is still vast overkill for the facts to be explained.
>Jones, et. al., *still* don't grasp that the extreme radicalness of their
>theistic claim requires a *type* of proof that they simply cannot get from
>empirical means (for example, even if the fine-tuning argument were valid
>up to its conclusion, all we could conclude is that some being or beings
>designed the universe. We could *not* rationally conclude that the
>designer(s) were even remotely similar to God.
>
>
This archive was generated by hypermail 2b29 : Tue Jul 25 2000 - 10:04:01 EDT