Re: Examples of natural selection generating CSI

From: Richard Wein (rwein@lineone.net)
Date: Mon Nov 06 2000 - 08:05:38 EST

  • Next message: Bertvan@aol.com: "Great web site"

    From: Ivar Ylvisaker <ylvisaki@erols.com>

    >Richard Wein wrote:
    >
    >[snip]
    >
    >> I haven't read Paley. Do you know if Paley's argument is indeed a maximum
    >> likelihood one?
    >
    >Here are some excerpts from the book that keeps his memory alive on
    >creationist web sites:
    >
    >http://www-phil.tamu.edu/~gary/intro/paper.paley.html

    Thanks, Ivar. I really enjoyed reading Paley's argument. I think the main
    flaw lies in the following paragraph:

    "V. Nor, fifthly, would it yield his inquiry more satisfaction, to be
    answered that there existed in things a principle of order, which had
    disposed the parts of the watch into their present form and situation. He
    never knew a watch made by the principle of order; nor can he even form to
    himself an idea of what is meant by a principle of order distinct from the
    intelligence of the watchmaker."

    It's not entirely clear what Paley means by a "principle of order". But, if
    we take the theory of evolution to be a principle of order, then it's clear
    that this part of Paley's argument has been undermined by the advent of the
    theory. We *can* now form an idea of an unintelligent principle of order
    that has the potential to form biological mechanisms.

    If we ask ourselves why Paley reached an erroneous conclusion regarding the
    origin of biological mechanisms, I think the answer must be that he failed
    to consider the true natural explanation.

    Paley continues:

    "VII. And not less surprised to be informed that the watch in his hand was
    nothing more than the result of the laws of metallic nature. It is a
    perversion of language to assign any law as the efficient, operative cause
    of any thing. A law presupposes an agent, for it is only the mode according
    to which an agent proceeds: it implies a power, for it is the order
    according to which that power acts. Without this agent, without this power,
    which are both distinct from itself, the law does nothing, is nothing. The
    expression, "the law of metallic nature," may sound strange and harsh to a
    philosophic ear; but it seems quite as justifiable as some others which are
    more familiar to him, such as "the law of vegetable nature," "the law of
    animal nature," or, indeed, as "the law of nature" in general, when assigned
    as the cause of phenomena, in exclusion of agency and power, or when it is
    substituted into the place of these."

    Here, Paley fails to rule out the possibility of a law of nature that could
    be responsible for the watch, but appeals to an intelligent agent as the
    motivator of natural laws. This is really a separate argument, dealing with
    ultimate causes, which are beyond the scope of biological evolution.

    In Chapter Two, Paley goes on to consider the hypothetical possibility of a
    self-replicating watch:

    "II. He would reflect, that though the watch before him were, in some sense,
    the maker of the watch, which, was fabricated in the course of its
    movements, yet it was in a very different sense from that in which a
    carpenter, for instance, is the maker of a chair -- the author of its
    contrivance, the cause of the relation of its parts to their use. With
    respect to these, the first watch was no cause at all to the second; in no
    such sense as this was it the author of the constitution and order, either
    of the arts which the new watch contained, or of the parts by the aid and
    instrumentality of which it was produced..."

    This raises the question of whether an unconscious machine can be
    considered the "author" or designer of an object it produces. In the case of
    a machine which had only the capacity to produce objects of a pre-specified
    design, possibly with minor variations, I expect we would say "no", as Paley
    did. But, if we consider a robot with sufficient intelligence to create new
    designs, a possibility which Paley probably didn't foresee, then we probably
    would consider the robot to be a designer in its own right. It might,
    however, be difficult to draw a clear boundary between the two cases.

    I have some sympathy with Chris Cogan's position. He attempted to draw a
    distinction between intelligent and non-intelligent algorithms by defining
    the former as those which look ahead (in some sense). I'm tempted to apply
    this same criterion to the question of what is or is not a designer. But I
    see problems. Suppose we have a robot which incorporates a predefined design
    for an object with one variable parameter. Each time the robot produces an
    object, it selects a suitable value for the parameter, based on an algorithm
    that looks ahead to maximize the future utility of the object. This would
    meet Chris's criterion for intelligence, but I don't think we would consider
    the robot to be a designer. Or would we? If so, what if there was no
    variable parameter, and the robot was simply choosing between two completely
    static predefined designs?

    >I don't know when the term "likelihood" was first used but I doubt
    >that it goes to 1800. In any case, Paley was trained as an Anglican
    >priest, not as a statistician or mathematician. He was not aware, so
    >far as I know, of the writings of Hume.
    >
    >Many recent critics of Paley have claimed that he was reasoning by
    >analogy and that this is bad unless the things being compared are very
    >similar to one another.

    I don't think he was reasoning by analogy. Such an argument would have taken
    the following form: "Watches are designed. Biological mechanisms are like
    watches. Therefore biological mechanisms are probably designed."

    IMO Paley was simply using the discussion of watches to *illustrate* a mode
    of argument, which he then applied to biological mechanisms.

    "Every observation which was made in our first chapter concerning the watch
    may be repeated with strict propriety concerning the eye, concerning
    animals, concerning plants, concerning, indeed, all the organized parts of
    the works of nature."

    >Sober claims that Paley was reasoning using
    >an "inference to the best explanation" and that this is reasonable
    >except that evolution is a now better explanation.

    Given the definition of "inference to the best explanation" that you give
    below, I'm inclined to agree with Sober. But note that it's not clear
    whether evolution is a "principle of order" or a "law of nature" in Paley's
    terms. As I showed above, Paley hedges his bets by saying that laws of
    nature are signs of design anyway.

    [...]

    >Here is the definition in the glossary
    >of "The Philosophy of Science" edited by Boyd, Gasper, and Trout:
    >
    >"Inference to the best explanation: A pattern of reasoning by which
    >one infers that a hypothesis is true from the fact that the
    >hypothesis offers the most plausible or satisfactory explanation of
    >the evidence."

    Thanks for providing this definition. This reinforces my previous impression
    that "inference to the best explanation" is not a very useful term. Surely,
    by this definition *any* scientific conclusion is an "inference to the best
    explanation."

    [...]

    >I have thought some about the questions that would be raised if
    >various kinds of signals were received by astronomers. I don't
    >have any clear answers.

    Well neither do I, except that I'm pretty sure the process must include some
    intuitive method of provisionally rejecting all other possibilities (known
    and unknown).

    >Actually, I generate more questions than answers. Here are some
    >examples:
    >
    >When we are trying to determine if something was designed, are we
    >trying to detect purpose or a construction process or a combination
    >of the two?

    That's really a matter of definition. And I don't think we yet have a clear
    definition of "design."

    The best definition I can think of at the moment is that design is a process
    that looks ahead (in some sense). (Thanks to Chris!)

    Given this definition, the only processes that we know of so far which can
    perform design are conscious beings and machines constructed by conscious
    beings. Thus, barring some unfamiliar type of phenomenon, we could say that
    design necessarily involves the action of a conscious being, directly or
    indirectly. One example of a potential exception would be an intelligent
    organism which evolved naturally without developing consciousness.

    >How do we know that the following things are designed or constructed
    >or both or, alternatively, that they are not:
    >
    >An automobile (that we have not seen assembled)
    >The coliseum in Rome
    >Stonehenge
    >A bird's nest
    >A wasp's nest
    >The paw print of a wild animal in mud
    >Animal dung
    >Tree rings
    >A rock cracked by a tree root.
    >A rock cracked by freezing water
    >Lava flows
    >DNA
    >Genetically modified corn
    >Hybridized flowers
    >The universe
    >
    >If God makes things happen by magic, do these things appear to be
    >constructed?
    >
    >Alternatively, do we identify construction by asking if it looks
    >like something that a human being has made or could make?
    >
    >What is the status of things made by machines?
    >
    >Are machines always designed and constructed by intelligent beings
    >(assuming that the definition of machine does not require this)?
    >
    >Which came first, gods or nature or machines, and how do we know?
    >
    >If abiogenic processes are going on today, is there any way that we
    >could tell? Might not existing life eat the evidence?
    >
    >And, finally, what are the rules for deciding whether something is
    >designed? Words like "implausible" and "intuitive" and "cannot
    >image" are not very helpful as rules. If we are sure that something
    >is designed, then we should be able to tell other people what our
    >rules are if our reasoning is supposed to be scientific. Further,
    >rules form a scientific hypothesis that needs to be verified (or
    >refuted)?

    That's a very good set of questions, and I'm hoping it will spark off some
    more discussion. I've addressed some of them above.

    The last paragraph is asking the $64,000 question. (Actually, I'm sure an
    answer would be worth a lot more than that!) Frankly, nobody knows the
    answer. But note that this is a particular case of a more general problem.
    No-one has a general and well-defined set of rules for making scientific
    inferences. This is why the philosophy of science is such a controversial
    area. Like it or not, scientists *do* have to use their intuition.

    My impression is that the most advanced work on this question is being done
    by the AI community. They're the people who really need a rigorous answer,
    because they can't appeal to the intuition of a computer.

    Richard Wein (Tich)
    --------------------------------
    "Do the calculation. Take the numbers seriously. See if the underlying
    probabilities really are small enough to yield design."
      -- W. A. Dembski, who has never presented any calculation to back up his
    claim to have detected Intelligent Design in life.



    This archive was generated by hypermail 2b29 : Mon Nov 06 2000 - 08:04:49 EST