Re: evolution and diet (was: FITNESS: Diet and Exercise)

From: gts (gts_2000@yahoo.com)
Date: Sat Apr 19 2003 - 14:54:21 MDT

  • Next message: gts: "RE: evolution and diet (was: FITNESS: Diet and Exercise)"

    "Eliezer S. Yudkowsky" <sentience@pobox.com> wrote:

    > gts wrote:

    >> "Eliezer S. Yudkowsky" wrote:
    >> Or rather, it doesn't matter what the null
    >> hypothesis or Bayesian prior or whatever
    >> *was*, because there are now enough specific
    >> cases of modern diets being detrimental because
    >> of violating ancestral invariants that I would,
    >> indeed, tend to take as the *new* working
    >> assumption that the ancestral diet is better until
    >> proven otherwise.

    >> Actually that is what I mean by the paleodiet
    >> being the null or default hypothesis.

    > No, that *would* be just playing burden-of-proof
    > tennis. The paleodiet hypothesis is an interesting
    > one, but if your sole justification is "because
    > that's what our ancestors ate", in the absence of
    > any evidence

    Eliezer, that is NOT my "sole justification."

    Apparently I haven't made myself clear. I wonder how
    many more times I will need to repeat myself on this
    particular point. I've done so at least three times
    here already with Harvey and others, in this thread
    and in other diet-related threads populated by the
    same people.

    The evidence is clear that paleolithic foods are more
    nutrient-dense and healthier than non-paleolithic
    foods, and this evidence comes not only from
    archaeology but also from entirely outside of
    paleodiet theory, directly from nutritional science
    itself. As I wrote to Harvey, I shouldn't need to
    remind anyone interested in nutritional science that
    the literature is replete with evidence that
    paleolithic foods are among the healthiest foods
    available. Who with any basic knowledge of nutrition
    can argue that fish, lean meats and poultry, fruits,
    and vegetables are not among the healthiest and most
    nutrient-dense foods in existence? And that foods like
    ice cream, pizza, candy, milk-shakes, bread, and
    pancakes have by comparison very little food value?
    These basic ideas should not really be a matter for
    debate here, I would think.

    I was agreeing with your *entire* paragraph at the top

    of this post, including and especially your words
    "especially because there are now enough
    specific cases of modern diets being detrimental
    because of violating ancestral invariants that I
    would, indeed, tend to take as the *new* working
    assumption that the ancestral diet is better until
    proven otherwise."

    What I'm trying to convey here is this: when you say
    as you did above that you would...

    "tend to take as the *new* working assumption that the
    ancestral diet is better until proven otherwise"

    ...you are saying precisely what I am also saying.

    I restate the exact same idea in statistical research
    terms:

    "I tend to think paleodiet theory is the null
    hypothesis which needs be disproved by some competing
    hypothesis, as opposed to being one of those competing
    hypotheses."

    As for so-called "Burden of proof table tennis," I
    really don't see what the problem is. Burden of proof
    is not some kind of dreaded term to be avoided. It is
    an emotionally neutral idea. It does not alone imply
    that either side is right or wrong. It's merely an
    artifact of statistical research. For scientific
    progress to take place, someone must compare and
    discredit the null hypothesis by providing proof that
    a competing hypothesis better explains the empirical
    data.

    In terms of paleodiet theory testing, the experimental
    setup would look like this:

    Definition: "'Optimal Diet' causes a, b and c." (where
    a, b and c are some measurable parameters of good
    health and longevity agreeable to both camps.)

    The Null Hypothesis:
    "The Optimal Diet is Paleolithic."

    The Competing Hypothesis:
    The Optimal Diet is paleolithic except with respect to
    adding [insert non-paleo item(s) here] and/or removing
    [insert paleo item(s) here].

    If a test on humans showed that the health parameters
    associated with the competing hypothesis were
    healthier than those from the null hypothesis, with
    statistical significance > 95%, then we would reject
    the null hypothesis and accept the competing
    hypothesis.

    This has already been done successfully, for example,
    with respect to adding non-paleo antibiotics. The
    tests were not directly against the paleodiet but few
    would dispute that they pertain also to the paleodiet.

    -gts



    This archive was generated by hypermail 2.1.5 : Sat Apr 19 2003 - 15:04:19 MDT