Re: evolution and diet

From: Charles Hixson (charleshixsn@earthlink.net)
Date: Thu Apr 24 2003 - 10:26:14 MDT

  • Next message: Spudboy100@aol.com: "Re: Doomsday vs Diaspora"

    gts wrote:

    >Eliezer S. Yudkowsky wrote:
    >
    >
    >
    >>In matters such as these, the wise Bayesian attempts to accumulate
    >>enough evidence to make all reasonable differences of priors moot.
    >>
    >>
    >Presumably Mr. Hugh M. Species has already done so regarding the theory of
    >evolution. He accepts the theory of evolution, and accepts the fact that he
    >is adapted best to the diet upon he evolved.
    >
    >I'm starting to wonder if you really accept the theory of evolution.
    >-gts
    >
    But evolution is a process, not a snap-shot. At any particular
    time-place-(species&individual) there will be certain ways in which it
    is adapted, and others in which it isn't. These will summate to a
    probability for success (i.e., successful creation of
    great-grandchildren). But this success it a human perspective. The
    individual perspective is "here I am now, it feels like X, so I'm going
    to do Y." This is summated over a population. If Y is generally a
    "good enough" answer, the species persists. Since the individuals are
    different, those less adapted to the environment will be more likely to
    make a poor choice of Y (for themselves). But if someone lacks the
    ability to synthesize vitamin C, and their mate likes the taste of
    citrus, then grandchildren are more likely to result, with a combined
    lack of ability to synthesize vitamin C and preference for citrus. Not
    a huge change, but an adaptive one. (Adapting to synthesize vitamin C
    again is more difficult.)

    So, for example, if a population begins herding cattle, drinking milk
    will become an option. But it is only useful if you can digest it. So
    children whose ability to digest milk persists will have an advantage
    when food is scarse. This will cause a shift in the genes. But there
    is more than one factor involved, and they adjust on a "as it happens"
    basis. So the ability to digest lactose and the ability to tolerate
    high cholesterol diets are decoupled, and each adapts to the
    circumstances separately. But each adapts to the circumstances. After
    a few centuries, the population is mixed. And the values of each
    adaptation vary with the seasons (how much food of what kind is available).

    Note that circumstances never hold still. Adaptation is always to the
    current environment, and is never perfect. Adaptations to the old
    environment may easily get lost if they are no longer relevant. OTOH,
    if they aren't expensive, they may persist unaltered for a long time.
    Or they could persist in an altered form (since they aren't being used,
    changes wouldn't necessarily be maladaptive).

    In short, the only model in which the paleolithic diet is presumed to be
    optimal is one that assumes:
    1) the paleolithic diet was optimal for the people then living, and
    2) evolutionary change is quite slow
    Condition one is made more likely because normally the environment
    changes quite slowly, and evolution has time to keep up. But this
    presumes the entire ensemble of paleolithic conditions. Air pollution
    may well increase our need for vitamin C or E, to take one minor example
    of why it might not be optimal now. Condition two is much less
    probable. Demonstrating that we aren't optimally adapted to the current
    diet is easy, but that's not proof that we are adapted to any particular
    time in the past. Since the changes are largely decoupled, the way to
    bet would be that we aren't.

    Still, a better question might be, are we more nearly adapted to the
    paleolithic diet than to the modern diet. But do remember to adjust for
    body sizes, activity levels, chemical stresses, etc. So even that's not
    a simple question, though the probable answer begins to appear to be yes.

    -- 
    -- Charles Hixson
    Gnu software that is free,
    The best is yet to be.
    


    This archive was generated by hypermail 2.1.5 : Thu Apr 24 2003 - 10:35:39 MDT