Randomness and complex organization via evolution

From: Chris Cogan (ccogan@telepath.com)
Date: Wed Jul 12 2000 - 09:41:44 EDT

  • Next message: Bertvan@aol.com: "Randomness and complex organization via evolution"

    To: Bertvan@aol.com
    Subject: Re: Randomness and complex organization via evolution

    At 12:22 PM 07/11/2000, you wrote:

    Chris:
    >One reason is that randomness *is* complex organization.

    Bertvan
    Hi Chris,
    Science is what we know about the universe. What we know about the history
    of life on earth is meager. We know that different organisms have existed at
    different times, that they are in some respects similar and perhaps related,
    but that the differences are also great enough to avoid explanation. At the
    moment any theories of how this came about is speculation, speculation
    usually designed to accommodate some particular philosophy, such as theism or
    atheism. I haven't included agnosticism because one definition of an agnostic
    is the ability to live comfortably with unexplained phenomena. We also
    indulge in speculations but I think we are a little less likely to claim them
    as "thruth".

    You have accommodated your philosophy (no plan, purpose, design or teleology)
    with the facts by defining randomness as "complex organization". To me that
    is a little like defining up as down , or defining black as white, but I have
    no objection.

    Chris
    No, I have not *defined* randomness as complex organization, though it is
    complex organization. Pure randomness is more complex than even the most
    complex organism, for the same amount of matter. That is, if you take the
    atoms that make up even the most complex organism, if you randomize them,
    the result is even *more* complex (by a vast amount, as well, not just a
    slight margin). Living organisms are often complex, but not hypercomplex
    for the material involved. The complexity of living organisms is merely a
    small "sampling," as it were, of the complexity of the forces acting on
    life. What makes living things simpler than random groupings of the same
    atoms is that there is a vast amount of regularity and repetition of both
    small structures (molecules) and much larger structures (blood cells, the
    DNA in cells, etc), and much larger structures still (in the kidneys, for
    example). Consider the fact that we can, to large degree, understand how
    our bodies work, but we could not achieve such a degree of understanding of
    the various internal details of a random structure of the same atoms as are
    in a human body without *vastly* more research, because we'd have to study
    each small area in detail. We could not generalize from a sampling of blood
    cells to billions of others like them. In information terms, the random
    structure would not usually be significantly compressible in terms of
    expressing its structure, whereas a human being made from the same atoms
    would be similarly definable in a vastly smaller amount of space, because
    so much of the information about certain substructures could be simply
    referenced where needed. For example, we could completely specify one red
    blood cell and then use that information and a small addition for each
    other red blood cell.

    Further, I frequently worry about the term "random," because it is so
    diffuse without strong contextual restriction. In a literal, metaphysical
    sense, I don't believe in randomness *at all*. But the alternative is not
    intelligence. The alternative is lawfulness, order determined by the nature
    of the various things that exist. Round things can roll smoothly, square
    things cannot (on smooth terrain). If something intrudes on the genome of
    an organism during DNA replication, it does so according to the laws of
    physics (whatever they may be), not in any absolutely or metaphysically
    random way. However, because we are generally several orders of magnitude
    away from having sufficiently detailed measurements or understanding of
    what's going on in any particular case, it appears random to us, and it
    *is* true that there is no known *systematic* relationship between a genome
    during DNA replication and the intrusion of a cosmic ray that happens to
    hit a certain atom in the DNA at a crucial moment. Even if we *could* get
    the details, we could not make any nice "laws" that would say that such
    cosmic rays would uniformly intrude at that point in the process and hit
    the corresponding atom in other cases. Each individual instance is a case
    unto itself.

    I would define *true* randomness in terms of indeterminism. But for
    ordinary purposes, a *mathematical* characterization is better. Such a
    definition says that there should be no overall repetitive patterns, there
    should not be a major preponderance of some values over others (assuming
    that the randomness is supposed to be "flat" rather than gaussian, etc.),
    and so on. Even this has problems, because patterns that are clearly not
    random in one sense may appear to be random according to such tests. For
    example, the digits of pi seem to be random according to such tests, but we
    know that they are strictly determined.

    (Worse, pi is immensely compressible, at least for many uses, by specifying
    how it is defined or how further successive digits may be computed. This is
    infinite compression, just as the laws of physics represent infinite
    compressibility in that they apply to a potentially infinite number of
    cases within certain parameters. And yet, randomness is commonly thought of
    as non-compressible in information theory terms.)



    This archive was generated by hypermail 2b29 : Wed Jul 12 2000 - 09:43:27 EDT