Re: the definitions of evolution

From: Tim Ikeda (
Date: Tue Oct 09 2001 - 23:54:57 EDT

  • Next message: Mark Plain: "Here is the REAL solution to the problem of terrorism"

    Howard wrote:
    >>1. What kind of empirical evidence would contribute toward a knowledge
    >>of process?
    >>2. Does mechanism have to be known before process?

    Uko responded:
    >Good questions. With regard to the question concerning empirical evidence
    >to process, I think we need evidence for the transition (at the very
    >least, evidence that points to the transition) from ancestor to descendent.

    If there were no transition, and if two species were not related
    by descent, we'd wouldn't expect to see the same degree of biochemical
    or morphological similarities among the descendents than if they were
    related. Sometimes people get lucky and there is temporal evidence
    as well (e.g. fossils).

    >In other words, we need evidence that at least enables us to conceptualize
    >the development of another stage or acquisition of another structure.

    I think that's more a question of "mechanism", than "process". Pattern
    relates to spatio-temporal relationships among discrete data points.
    Process is what connects the points (descent with modification). Mechanism is
    how it happened (selection, drift). One may infer common descent (process)
    from the pattern of relationships knowledge that organisms reproduce as
    imperfect copies without knowing why a particular trajectory was taken

    >We will undoubtedly differ in our criteria for such conceptualization. But I
    >would judge, for example, that Ken Miller's account of the evolution of
    >the cilium in response to Behe's claim that Darwinian gradual evolution
    >is inadequate to account for the evolution of the cilium, is still nothing
    >more than an account of the pattern of evolution. He also has not given an
    >adequate explanation of the mechanism of the evolution of the cilium either.
    >I think he still fails to address Behe's basic thesis.

    I believe Behe's thesis was stated more strongly than the whether
    "adequately described pathways" have been presented. That, in itself is
    an unremarkable claim. His actual, interesting thesis is that "very
    irreducibly complex" biochemical systems *cannot* originate by natural
    mechanisms. Behe claims as evidence that exceptionally detailed, step-by-step,
    exactly reproducible or experimentally tractable explanations have not been
    described. But, FWIW the reality is that we can't even model _existing_
    systems at the level of detail that Behe requires for evolutionary
    explanations and predict that they could even work. For example,
    I would have great difficulty predicting what would happen to the rate of
    blood clotting if I changed a particular amino acid in Factor-VIII or blocked
    post-translational glycosylation of one of its surface tyrosine sulfates. It
    would be impossible to predict that would happen if I subsequently changed a
    few more in other factors. While many have studied the clotting cascade as
    interactions between individual components, but no one has ever characterized
    what happens to all the components under wide variety of truly physiological
    conditions. So, although Behe boldly asserts in his book that life's current
    process are in fact no more than chemistry, even he cannot support his
    statement at the same level of proof he demands of evolutionary proposals.

    His case, if I might paraphrase it, isn't even as good as an "I don't
    think 'X' is plausible" argument, it's actually, "I can't tell how 'X'
    occurred". Thus, to counter his thesis, which is that "'X' could
    _not_ have arisen naturally", is to suggest possible routes and support
    them with biochemical relationships. In that endeavor, Russ Doolittle and
    others have mapped out perfectly plausible intermediate steps for the
    emergence of the blood clotting pathway that can be evaluated from sequence
    and biochemical comparisons. At the level of detail that is experimentally
    approachable, we know of less complex, "irreducible complex" systems that
    have evolved under laboratory conditions. I've even cited one case in my work
    on nitrogen metabolism in bacteria (Search on "alanine fast grower" and my
    name in the ASA evolution list).

    >Miller is still left with pattern and he continues to convince himself
    >that he has accounted for the process.

    I've heard Miller talk. Miller certainly doesn't believe that
    he's accounted for the process in the sense that most of the essential
    details are there and that there's little more work to be done. He's
    offered one possible explanation (well, actually others have), and
    backed it up with the biochemical evidence that is currently available.
    This _must_ be understood in contrast to Behe's argument that IC systems
    such as the various cillia couldn't have evolved or otherwise we'd have
    detailed biochemical explanations by now.

    I think one tell-tale sign of the actual uncertainty of Behe's
    arguments is the way he's recently tended to focus more on flagella
    and ATPases -- ancient systems whose origins are more likely to be
    obscured over time -- than more recently emerged systems like
    the clotting cascade. In fact, I've asked Behe and his supporters
    numerous times to provide an example of the most recently emerged,
    unevolvable, IC system of which they are aware -- No responses to date.
    In other words, instead of attempting to study IC systems for which
    one might expect to gather the best evidence to address the
    question of evolvability (Recently emerged systems make better
    experimental models), he and others have focused on the oldest, most
    highly integrated systems and least scientifically tractable systems.
    This is a highly unorthodox and unlikely approach for anyone hoping
    to acquire definitive results, as we scientists know well.

    >Concerning the second question, does mechanism have to be known before
    >the process? I would say no. We might well be able to give an account
    >of the process without understanding the mechanism. Natural selection
    >as the chief proposed mechanism for the process of evolution falls
    >short of serving as an adequate explanation for much of the process of
    >evolution. Simply because natural selection does occur, doesn't mean that
    >it is sufficient to account for the extensive evolutionary processes
    >that have taken place. In my view, evolutionary biologists loose scientific
    >credibility when the continue to impose a mechanism on processes regardless
    >of whether there is sound empirical evidence for doing so.

    Credibility is likewise enhanced when evolutionary biologists also discuss
    the roles of other mechanisms such a drift, or consider the underlying
    biochemical contexts & historical contingencies (in biology, context is
    everything) when considering evolutionary processes rather than
    focusing on natural selection alone.

    >Extrapolation or speculation me be appropriate, but then it must also be
    >recognized as such. What is basically need is more scientific honesty.
    >Lack of alternative explanations is not a form of empirical evidence.

    That suggests Behe's arguments in support ICness and "unevolvability" are
    invalid, because "lack of alternative explanations" is his ultimate

    >Perhaps my comments raise more questions about my original comments
    >concerning the meanings of evolution. But I do think it is very legitimate
    >to ask for the empirical evidence for process as well as mechanism for the
    >pattern of evolution.

    Well, let's take the chimp/human split as an example. Do we have a
    process? Yes. First, we have sex. We have empirical evidence that organisms
    do not reproduce exactly. We have evidence that relatively few changes in
    the development pathways of organisms may have large effects of morphology
    -- Changes that even retain or enhance viability. We have evidence that
    chimp and human genomes exhibit _tiny_ differences, no greater than what
    has even been seen among organisms of the same species. We have observed
    numerous modes of speciation in the wild and a few in the lab. We have
    observed and even understand the biochemical basis of the generation of
    genomic variation. Basically, we have ample evidence that populations
    diverge genetically over time and frequently speciate. We may not
    know what drives or supports divergence, but it is an observed reality.
    Determining whether a process occurred is simple, knowing the exact
    the mechanism/s involved at any particular time is the problem.

    What don't we have? We lack a step-by-step picture of each mutation
    accumulated and a clear idea of the exact population and conditions
    under which any variations arose. In most cases we can't determine
    the effect of any particular combination of sequences on an
    individual, let alone calculate evolutionary trajectories across
    populations. It is beyond our capability to model a hypothetical
    organism from basic physical principles, let alone one we know exists.

    Does the "scant" evidence we have for the chimp/human divergence outweigh
    the theoretical limitations/impossibilities of modelling? Let's move to
    the more concrete: Does the scant experience we have with isolated,
    biochemical sub-systems and highly limited, whole-organism experiments
    outweigh the lack of theoretical modelling of whole-organisms such that
    we biochemists can confidently assume that extant life can be reliably
    studied through physical means without resorting to "supernatural"

    Clearly we are talking about shades of grey in both instances. Yet
    whereas few question the assumptions about the chemical basis of life
    as it operates today, where there are far more unknowns than knowns,
    the same suspension of disbelief is often not extended to older,
    biological phenomena. I think that biological evolution, because it
    involves the study of not just whole organisms but of whole populations
    of interacting organisms _over long periods of time_, may produce more
    doubt because it: a) Is unfamiliar territory for most biologists,
    b) Requires more integrative understanding of biology than most of us
    have the time or will to acquire, and c) Most clearly highlights our
    ultimate lack of understanding about the basic, physical basis of life
    (If we can't cure most cancers why should we even expect to have developed
    highly-detailed evolutionary pathways for multi-component biochemical

    [Were we to consider religious viewpoints, I would also add:
    d) It can appear as a lump in the "theological oatmeal" that some
    find unpalatable.]

    If the relative scientific "validity" of ideas like biological evolution
    and the chemical basis of life could quantified, I'd argue that they
    differ by a few degrees rather than by orders of magnitude. A few people
    will find themselves falling between the two, but most will ultimately
    fall on one side or the other of the pair.

    Tim Ikeda

    This archive was generated by hypermail 2b29 : Wed Oct 10 2001 - 00:11:39 EDT