On Mon, 22 May 1995 12:38:27 +0930 you wrote:
>
>I am a new subscriber. I am quite interested in the theological
>ramifications of various views of life origins, but I will
>leave these questions to another post.
Welcome aboard Mark, especially another Oz-ite!
>>Note the key word: *functional* intermediates. It is one thing to
>imagine a smooth continuum of intermediate stages (Darwinists often
>can't even do that), but it is quite another for *all* those imagined
>stages to be *functional*. Clark points out:
>
>"In order to build up a structure by natural selection, it is
>essential that each stage in the building process must make an animal
>better fitted to its environment than the one before it. An eye that
>is half developed must be more useful to an animal than an eye that is
>49 per cent developed, and this in turn, than one, the development of
>which has proceeded to only 48 per cent, and so on. The graph of
>usefulness against the extent of structural organization must show a
>steady upward rise-otherwise progress must inevitably stop, hindered
>by natural selection itself. If the graph is not a steady upward
>rise, but has ups and downs, then natural selection (which selects
>usefulness and adaptation), working from either direction, will force
>the organism to the nearest maximum.
>This is a half truth.
Disagree. A good example is Microsoft's difficulty in providing
a smooth transition from Windows 3.xx to Windows '95. It now looks as
though they can either have "smooth" or "transition" but not both.
>Evolution, basically, is a process which
>attempts to solve an optimization process, namely, to maximize
>fitness.
Evolution does not attempt to solve anything. It is a blind,
non-teleological process.
>The problem with lots of optimization procedures is that
>they get stuck in what are called "local maxima", ie locally (among
>variations which aren't too far away from each other) a certain
>genotype may be optimal, but on a grand scale, it may fare quite
>poorly.
Agreed. Micro-evolution is accepted by all but not macro-evolution.
>The trick to developing a good optimization algorithm is
>to give it the ability to get out of local maxima. The problem is
>best illustrated by the "greedy algorithm", which says: "only
>change to something different if it improves the fitness". This
>strategy is highly likely to get stuck in local maxima. For once it
>reaches a local maxima it won't want to leave as all the local
>alternatives seem worse. But evolution is _not_ the greedy algorithm.
>It overcomes this problem to an extent by introducing a probabilistic
>element. Instead of saying, "only change if fitness improves", it
>says, "a change occurs with higher probability if fitness improves".
>This means that while there is a general tendency to improve fitness,
>there will be times when fitness decreases.
Yes. And Clark's point is that selection would weed it out at that
stage.
>Local maxima still have the potential to cause the evolutionary
>process problems (indeed, some evolutionists say that this is why
>some species remain relatively unchanged for long periods - because
>they are stuck in a local maxima), but these problems are not
>necessarily insurmountable.
I believe they are.
>Today, with our much greater knowledge of and familiarity with complex
>systems, we know that steady upward rises of the kind demanded by
>materialistic evolutionists are unknown to science. Isolated
>This assertion is not at all clear. Have you evidence for this claim?
I was quoting Clark. It seems self-evident to me. I am not a
scientist. If you have examples of "steady upward rises of the kind
demanded by materialistic evolutionists", please give them.
>Neither is it obvious that the "extent" of local maxima is "small"
>enough to make the process of evolution a viable explanation for
>the life we have today. I have not seen evidence either way. Does
>anyone have evidence?
I do not rule out evolution. I just am not convinced from what I have
seen that it is sufficient to account for the living world. Many
eminent biologists agree, eg. P.P. Grasse, etc.
Stephen