Re: Chance and Selection

From: Chris Cogan (ccogan@telepath.com)
Date: Sat Dec 02 2000 - 18:18:29 EST

  • Next message: Chris Cogan: "Eight respects in which evolution is neither intelligent *nor* random"

    >Hi Chris,
    >Descent with modification existed as a concept before Darwin. Darwin's only
    >contribution was the mechanism - tiny random mutations, which natural
    >selection created into complex biological systems by picking and choosing.

    Chris
    You're confusing a *metaphor* of picking and choosing with *actual* picking
    and choosing. Darwin, and many others (including myself on this list) have
    pointed out that this *is* a metaphor. In Darwin's theory, there is *no*
    non-metaphorical "choice" in the basic evolutionary process.

    >Darwin himself stated that if any biological system proved to be too complex
    >to be created in this way, his system would prove to be invalid. Some
    >scientists, such as Behe, now claim that many, if not all, biological systems
    >are too complex to have been "created" by natural selection. If, as you
    >claim, "most scientists" no longer regard Darwin's theory the explanation of
    >evolution, all they have to do is say so. Much of the controversy would
    >disappear, for it is not "evolution" that most critics question, but
    >specifically "the creation of complex biological systems by RM&NS". Some
    >people use the term "imperfect copying" (of DNA I assume) instead of random
    >mutation. This language certainly implies "accidental" and "lacking
    >intelligence or purpose".
    >
    >In your analogy, the monkeys are only allowed to make certain changes. Who
    >sets the rules for which changes are allowed?

    Chris
    In Nature, physics. In the case of the monkeys and PC, anything that makes
    the process sufficiently analogous to the known laws of physics as they
    apply to the chemistry of DNA-replication. The point of the analogy is to
    illuminate what happens in Nature. Since the analogy is otherwise obviously
    invalid, we have to introduce some sort of restrictions on the case of the
    monkeys and PCs in order to match the molecular physics (i.e., chemistry)
    of DNA-replication in the real world.

    >The machine itself you claim!!

    Chris
    Only on the basis of patterns *already* laid down by the physical nature of
    the atoms and energy involved.

    > Any machine or entity that sets rules, or makes decisions sounds to me like
    >it either includes intelligence, or it is itself the result of an intelligent
    >design. I realize yours is only an analogy, but below is a quote.

    Chris
    In the case of the very simplest organisms, the rules would be *only* those
    of physics. In the case of more-complex ones, ones where a more-complex
    "computer" had already evolved, the process could *then* gradually generate
    variations, some of which could serve to limit (or encourage)
    modifications. If such modifications occurred and proved to be beneficial
    to their own survival, they would tend to be perpetuated (and improved
    upon). If such modifications turned out to be harmful, they would tend to
    disappear.

    >Chris:
    > >So, we let the monkeys do their thing on a billion computers,
    > >and a billion new computers are produced based on their
    > >(mostly slight) changes. In some cases, the change might
    > >only an increase in power supply wattage. In other cases,
    > >it might be an increase in the amount of memory. In other
    > >cases, it might be a difference in some software subroutine.
    >
    >Bertvan:
    >We can guess the odds that an accidental modification made to billion
    >computers by a billion monkeys would actually result in increased power
    >supply wattage, or increased amount of memory, or change a software
    >subroutine. I suggest that the odds are even less that an equivalent
    >"improvement" would accidentally occur in a biological system, which is many
    >times more complex.

    Chris
    False. Modern supercomputers and the software that runs on them is much
    more complex than the simplest *living* things. Further, the existing
    living things that are much more complex than that have been selected for
    precisely their ability to generate variations/mutations/modifications that
    each, individually, fall within a range such that they have a (relatively)
    high chance of being beneficial modifications. They will have an advantage
    over similar organisms that *don't* have this ability.

    >The most unlikely of all, IMHO, would be that all this
    >random tinkering by monkeys would result in increased complexity, rather than
    >deterioration.

    Chris
    You forget that it could result in increased complexity *and*
    deterioration. But, no, in fact, depending on what is used to parallel the
    physics of Nature, the chances (on a population of a billion initial
    computers), would be quite high, especially if it was carried out over many
    generations of the billion computers. Your grasp of the probabilities
    involved matches, but unfortunately does not surpass, that of Stephen
    Jones, who makes all such judgments on the basis of his uninformed
    intuition, which is precisely the mistake you are making in this case. If
    you actually *do* the math, under reasonable "parameters" that match those
    of the physics of the components of living things, you get results that do
    in fact demonstrate that the monkeys will produce occasional beneficial
    changes.

    Keep in mind that to match biological populations, most replications would
    merely involve duplication (at the level of complexity of a PC), so that
    the population, supposing the existence of a suitable environment, would
    soon be not merely a billion PCs, but a population of many trillions, or
    trillions of trillions, with only a relatively small percentage of
    noticeable modifications. This provides a "pool" in which new variations
    can appear.

    >You say the system has the ability to repair itself. How
    >would it to that without intelligence?

    Again, the short answer is physics. Those mechanisms that occur that have
    even a slight tendency to cause the overall structure, or some significant
    part of it to be repaired are obviously biologically advantageous in many
    environments. We know that biological organisms are physical organisms, and
    we can *observe* (to a degree) the physical processes involved, not a
    single one of which is intelligent. There is nothing there that we can
    observe except the behavior of atoms and molecules.

    >I'm not sure if the "random mutations" you have in mind are point mutations
    >in nucleotides. In which case, each new biological function would require
    >thousands of mutations occurring in the proper order.

    Chris
    No. In the simplest case, it would require nothing more than, perhaps, the
    addition of a single atom. In the case of existing life, each new
    biological function requires *at most* only *one* mutation -- from the
    state of the genotype one step before that particular function is
    produced. You are misleading yourself by thinking of biological functions
    in terms of gross activities or major categories of functions (such as
    circulation or digestion as we tend to see them). But even the slightest
    genetic change that produces a behavior that is different from what would
    have occurred before that change is a *new* biological function (to that
    genome). For example, suppose there is an organism that does not respond
    *at all* to a certain color of light, but a slight change in its genes
    enables it to respond just *slightly* to that color of light. *that's* a
    new biological function, and one that can, obviously, be achieved in some
    cases with only a single small modification.

    >Since I find it
    >difficult to think that is what you mean, I assume that that you believe DNA
    >itself has the ability to "organize" various point mutations into coherent
    >information to specify a new protein. I've heard the suggestion that an
    >existing gene might be used for a new function. Accidentally? Without
    >intelligence?

    Chris
    Why not, *if it already* performs that function to some degree (perhaps in
    biologically useless way)? Again, you are *assuming* a major biological
    function, and forgetting that these "new" functions always connect with
    *existing* functions. The Panda's "thumb" is an example. The bone was
    already there. Why does it take *intelligence* for the environment to give
    an advantage to those pre-Pandas that have a slightly functional bump where
    the "thumb" is now? Just where, in this entire process of variation and
    selection, does intelligence have any use? Why mightn't that variation,
    along with millions of others, be the result of the physics of genotype
    replication in Pandas? A more obvious example would be heavy fur on furry
    animals that generally live in cold climates. Fur length normally varies
    anyway, and in an environment that rewards longer fur (and penalizes short
    fur by killing off the organism and its genotype), why *wouldn't* any
    accidental genetic increase in fur length be a survival advantage, and why
    *wouldn't* such genetically-based changes in fur length occur in
    cold-climate organisms as well as in hot-climate ones? Why would this
    require *intelligence*? How many possible ways are there to vary fur and
    yet still *have* fur? Does it require intelligence or merely an accidental
    change in some gene or two?

    And, why is it *intelligence* when a cold-climate animal is born with
    longer fur but a "mistake" when the *same* change occurs in a hot-climate
    animal? Why assume intelligence in *either* case? Obviously, if the same
    change can occur by accident in a creature that lives in a hot climate, it
    can *also* occur by accident in the same creature (or one just like it)
    that lives in a cold climate.

    You seem to be forgetting a fundamental fact: There is no Platonic,
    pre-definition of which specific traits will be beneficial and which not,
    and it also clear, empirically, that *many* of the changes that occur that
    are not beneficial in their circumstances *would* be beneficial in *other*
    circumstances. So, again, why is it a "mistake" (or accident) in one case
    and a sign of intelligence in another case, when the genotypes are the same
    in each case?

    If I stumble (by accident) and fall on my face, it is clearly not a sign of
    intelligence. But, what if I stumble and fall on my face and it just
    happens that I fall out of the path of a bullet that would have simply
    blown my head apart had I not fallen? Is the *same* stumble now magically
    converted into an *intelligent* action? I don't think so. It's a *lucky*
    event, but not an *intelligent* one.

    >What would make the choice to use a gene for another function,
    >if not some form of intelligence?

    Chris
    What *choice*? Why assume choice at all? If a rock falls and lands on your
    foot, why assume that it did so by choice?

    >Your view of life might not be too far from my version of intelligent design.
    > I know it is important to you that no god play a roll in nature, and I
    >wonder if that is the reason for your insistence upon "randomness".

    Chris
    I never insisted upon randomness, except in an extremely narrow sense
    (i.e., lack of direction by anything like a mind). I insist on this because
    there is, so far, no evidence for anything like choice or design.

    >However, your view of life wouldn't have to be in conflict with Theism.

    Chris
    It isn't. Most "Gods" don't exist because they are logically impossible.
    Others are simply devoid of any reason to believe in them. Others amount to
    abuses of the word "God" (as in Tipler's "Physics of Immortality").

    Bertvan
    >I have no interest in a theory of evolution that would prove materialism
    >wrong.
    >I doubt either materialism or its opposites can ever be proved to the
    >satisfaction of everyone. Materialists who are repelled by the concept of
    >God could attribute nature's ability to organize itself into rational systems
    >to an invisible force called "intelligence". Since a definition of
    >"intelligence" is something that makes choices, it will always be
    >unpredictable. Belief in whether or not God plays any roll in that
    >intelligence could be optional.

    Chris
    But, first, we have to show that there is any fundamental role in Nature
    for choice at all, except in a clearly metaphorical sense. Evidence of such
    choice is no more present than evidence of intelligence or free will (in
    the indeterministic sense) or "spontaneity." Obviously, things happen. Does
    the ball "choose" to roll down hill when placed on a smooth but quite
    sloping surface? I don't think so. Does the DNA-replication process
    *choose* to duplicate a stretch of genes? I don't think so. I think the
    physics of that particular instance are such that there was *no* choice,
    and the duplication was simply necessitated in the same way that sufficient
    heat and oxygen will burn ordinary dry wood. Neither the wood nor the
    replication processes have any choices to make, nor are any needed.

    Summary:

    Though you have repeated your claim that beneficial changes cannot occur by
    accident, you have not given any reason to think that this is so,
    particularly in light of the fact that what *counts* as beneficial or not
    depends on particular circumstances. If a particular change can be
    accidental, unintelligent, in one context, how does it get to be
    intelligent if it happens in another context? There seems to be no reason
    for dividing up changes into intelligent and unintelligent merely according
    to whether they are or are not beneficial, when the *same* change might be
    involved in both cases. Can you explain to us why a small genetic change
    that's beneficial in one case is intelligent and the *same* change in
    another case is not intelligent, or is a mistake, or is an accident? It
    would seem that, if the change is accidental in the case where it is *not*
    beneficial, it could *also* be accidental in the case where it *is*
    beneficial, but, by the terms of your argument so far, that change should
    not occur at all *except* by intelligence, since you claim that only
    *intelligence* can result in beneficial changes.

    And, yet, if the intelligence can make mistakes and produce changes that
    are harmful, why couldn't the same hold for *beneficial* changes?

    In short, your argument is asymmetrical. You switch from intelligence to
    failure of intelligence on purely ad hoc basis. This is inconsistent and
    logically indefensible, without strong *independent* support.
    Unfortunately, so far, your reasoning is essentially *circular*, and that's
    not a good thing in reasoning, because it divorces the whole procedure from
    reality and make it independent of truth. Conclusions of circular arguments
    may or may not be true, but, if they are, we can't tell it from circular
    reasoning.

    That is, your position (and your criticisms of the views of others) rest on
    the *presumption* that there is intelligence involved. This is why you use
    the word "choice" when a better description would be simply that something
    *happened*. Describing a genetic change as driven by choice without first
    *establishing* that choice is involved is clearly invalid. Yet, sadly, you
    do this routinely, as in the examples in your remarks above.

    The entire line of reasoning is unsound, and cannot be *made* sound. Your
    conclusions require a different *kind* of support than what you have so far
    offered. The burden of proof does not go away simply because it doesn't
    occur to you (it seems) to ask whether a particular genetic event really
    *is* a matter of choice, or simply a matter of the local physical causal
    conditions at the time of DNA-replication. And yet, I suspect you would not
    attribute choice to a rock that happens to come loose from its location on
    a mountain-side and fall. --So why do you attribute choice to the behavior
    of the DNA-replication process?

    I don't know. That you do this is a mystery, because none of the reasons
    that I know of that you have offered are strongly supportive of the
    conclusion, because they are not true, or because you have no valid
    reasoning to connect the facts specified to the conclusion, or both. Some
    of the problems in your reasoning are indicated above, and in my remarks on
    the oddly off-base analogy that I was commenting on, and in similar remarks
    you've made in past posts.

    Ultimately, by the way, the PC analogy is not a good one at all, because it
    requires too much fiddling to make it good. By the time it is made sound,
    it is too far from the normal situation with respect to PCs and technology
    generally to be valuable as a means of making points with respect to
    evolution. One of my points was that your initial version of it was clearly
    invalid, and that, as changes were made to make it more representative of
    the real world situation in biology, the less its value in criticizing NET.

    I hope that, for your next analogy of that general type, you will ask
    yourself whether it really *does* represent any significant principle or
    claim of NET. If it doesn't, perhaps all you need is a different analogy,
    but perhaps also you need to consider whether it is *possible* to make such
    an analogy that both seems to support your claims *and* is genuinely
    representative of NET. I doubt *very* much that you (or anyone) can do
    this, and the many failed past attempts at doing just this seem to suggest
    that it's because there is more wrong than merely the failure to find just
    the right analogy. What it suggests is that there *is* no such analogy, and
    that this is true because the conclusion you are trying to support with
    these analogies is false.

    That is, you are trying to indicate why, in your opinion, beneficial
    changes cannot occur by chance, and perhaps the analogies intended to show
    this fail because it is simply not true that they cannot occur by chance.

    It *may* be that intelligence is involved in such changes, but it the
    analogies that you have used do not support this conclusion because they
    are misrepresentative of the situations to which they are to apply. It
    seems clear that, unless we posit something to *prevent* beneficial
    changes, then they *can* occur by "chance" (i.e., without intelligent
    direction). If they *couldn't*, then such changes could not possibly all be
    chance changes. Put another way, *if* such modifications occur by chance,
    then some of them *must* be beneficial, because, by definition of "chance,"
    there could be nothing to prevent beneficial changes from occurring. It is
    logically impossible, mathematically, that *only* harmful or neutral
    changes could occur by chance, if beneficial changes are possible at all.



    This archive was generated by hypermail 2b29 : Sat Dec 02 2000 - 19:20:34 EST