On Tue, 17 Jun 97 21:18:47 CDT, Wesley R. Elsberry wrote:
WE>I mentioned earlier that replacing the now-missing "weasel"
>program described by Dawkins in "The Blind Watchmaker" would
>not be a difficult task. I'm happy to announce the
>availability of a "weasel" replication. Programmed in Delphi 2,
>the executable runs under Win95 or NT. The source code ended
>up at 154 lines, rather than topping the predicted couple of
>hundred lines. The target string, character pool, mutation
>rate, and population size are under user control.
>
>A population size of 100 and mutation rate of 6% seems to
>provide similar results to those reported by Dawkins.
This "mutation rate of 6%" seems to agree with one of ReMine's criticisms
of Dawkins "Methinks it is like a weasel" simulation that has an
unrealistically high mutation rate:
"Dawkins did not say, but he must have chosen the mutation rate to
optimize the speed of evolution. If he had chosen a low mutation rate,
(such as 10^-8 as in humans) then the simulation would require
roughly 50 million generations. On the other hand, if he had chosen
too high a mutation rate, then it would cause error catastrophe and
the target phrase would never be reached. Dawkins picked the
mutation rate that produced the fastest evolution." (ReMine W.J.,
"The Biotic Message", 1993, p233)
WE>The executable is available via anonymous ftp at
>inia.tamug.tamu.edu in the pub directory as "weasel00.exe".
>
>I think that I can easily demonstrate an interest in the "live"
>issue of test by computer simulation.
The question is whether it is a realistic "simulation" of the
biological world. If Dawkins' simulations really did simulate
real-world evolution, they would be included in every biology
book, but they are simply ignored.
The following are some criticisms of Dawkins' "weasel" simulation
from non-Darwinists (apart from ReMine):
Berlinski calls it "an achievement in self-deception" because
it uses "a target phrase":
"What more, then, is expected; what more required? Cumulative
selection, Dawkins argues-the answer offered as well by Stephen Jay
Gould, Manfred Eigen, and Daniel Dennett. The experiment now
proceeds in stages. The monkeys type randomly. After a time, they
are allowed to survey what they have typed in order to choose the
result "which however slightly most resembles the target phrase." It is
a computer that in Dawkins's experiment performs the crucial
assessments, but I prefer to imagine its role assigned to a scrutinizing
monkey-the Head Monkey of the experiment. The process under way
is one in which stray successes are spotted and then saved. This
process is iterated and iterated again. Variations close to the target
are conserved because they are close to the target, the Head Monkey
equably surveying the scene until, with the appearance of a miracle in
progress, randomly derived sentences do begin to converge on the
target sentence itself. The contrast between schemes and scenarios is
striking. Acting on their own, the monkeys are adrift in fathomless
possibilities, any accidental success-a pair of English-like letters-lost
at once, those successes seeming like faint untraceable lights
flickering over a wine-dark sea. The advent of the Head Monkey
changes things entirely. Successes are conserved and then conserved
again. The light that formerly flickered uncertainly now stays lit, a
beacon burning steadily, a point of illumination. By the light of that
light, other lights are lit, until the isolated successes converge,
bringing order out of nothingness. The entire exercise is, however, an
achievement in self-deception. A target phrase? Iterations that most
resemble the target? A Head Monkey that measures the distance
between failure and success? If things are sightless, how- is the target
represented and how is the distance between randomly generated
phrases and the targets assessed? And by whom? And the Head
Monkey? What of him? The mechanism of deliberate design purged
by Darwinian theory on the level of the organism, has reappeared in
the description of natural selection itself, a vivid example of what
Freud meant by the return of the repressed." (Berlinski D., "The
Deniable Darwin", Commentary, June 1996, pp27-28)
Dawkins himself responded to Berlinki's article, but significantly
he did not rebut Berlinki's crticisms of his simulation. However,
other correspondents brought it up, so Berlinski responded:
"In inserting a Head Monkey into Richard Dawkins's thought
experiment, my aim was to show how the mechanism of design,
purged on one level of Darwinian analysis, makes a stealthy
reappearance at another...If the mechanism of Darwinian evolution is
restricted to changes that are favorable at the time they are selected, I
see no reason to suppose that it could produce any fancy structures
whatsoever. If the mechanism is permitted to incorporate changes
that are neutral at the time of selection, but that will be favorable
some time in the future, I see no reason to consider the process
Darwinian. This is hardly a matter of semantics. A system conserving
certain features in view of their future usefulness has access to
information denied a Darwinian system; it functions by means of alien
concepts. But this is precisely how Dawkins's experiment proceeds.
My estimable Head Monkey conserves certain alphabetic changes
because he knows where the experiment is going. This is forbidden
knowledge; the Darwinian mechanism is blind, a point often stressed
by Darwinian theorists themselves..." (Berlinski D., "Denying Darwin:
David Berlinski and Critics", Commentary, September 1996, p25)
and
"When I observed that Richard Dawkins was unable to write a
computer program that simulated his linguistic thought experiment, I
did not mean that the task at hand was difficult. It is impossible. Mr.
Wadkins commends the discussion in Keen and Spain's Computer
Simulation in Biology as a counterexample; its no such thing. What
Keen and Sain have done is transcribe Dawkins's blunder into the
computer language Basic. Here are the steps they undertake. A target
sentence is selected- BASIC BIOLOGICAL MODELING IS FUN.
The computer is given a randomly derived set of letters. The letters
are scrambled. At each iteration, the computer (or the programmer)
compares the randomly derived sequence with the target phrase. If
the arrays-sequences on the one hand, target phrase on the other-do
not match, the experiment continues; if they do, it stops. There is
nothing in this that is not also in Dawkins, the fog spreading from one
book to the next. The experiment that Keen and Spain conduct is
successful inasmuch as the computer reaches its target; but
unsuccessful as a defense of Darwinian evolution. In looking to its
target and comparing distances, the computer is appealing to
information a biological system could not possess. This point seems
to be less straightforward than I imagined, so let me spell out the
mistake. Starting from a random string, suppose the computer
generates the sequence BNDIT DISNE SOT SODISWN
TOSWXMSPW SSO. Comparing the sequence with its target, it
proposes to conserve the initial "B." But why? The string is gibberish.
Plainly, the conservation of vagrant successes has been undertaken
with the computer's eye fixed firmly on its future target, intermediates
selected not for what they are (gibberish, after all), but for what they
will be (an English sentence). This is a violation of the rule against
deferred success. Without the rule, there is nothing remotely like
Darwinian evolution. What the computer has in fact done is to match
randomly selected items to a template, thus inevitably reintroducing
the element of deliberate design that was banished from the
Darwinian world." (Berlinski D., "Denying Darwin: David Berlinski
and Critics", Commentary, September 1996, p31)
Mike Behe points out that it is an analogy not of Daewinian evolution
but of intelligent design:
"Dawkins' analogy (which is slightly different in details in his book
versus Sober's rendition), though transparently false, appears to have
captured the imagination of some philosophers of biology. Besides
Sober, Michael Ruse has used a similar example in his book
Darwinism Defended, as has Daniel Dennett in Darwin's Dangerous
Idea. What is wrong with the Dawkins-Sober analogy? Only
everything. It purports to be an analogy for natural selection, which
requires a function to select. But what function is there in a lock
combination that is wrong? Suppose that after spinning the disks for a
while, we had half of the letters right, something like
MDTUIFKQINIOAFERSCL (every other letter is correct). The
analogy asserts that this is an improvement over a random string of
letters, and that it would somehow help us open the combination
lock. But if your life depended on opening a lock that had the
combination METHINKSITISAWEASEL, and you tried
MDTUIFKQINIOAFERSCL, you would be pushing up daisies. If
your reproductive success depended on opening the lock, you would
leave no offspring. Ironically for Sober and Dawkins, a lock
combination is a highly specified, irreducibly complex system that
beautifully illustrates why, for such systems, function cannot be
approached gradually. Evolution, we are told by proponents of the
theory, is not goal-directed. But then, if we start from a random
string of letters, why do we end up with
METHINKSITISAWEASEL instead of
MYDARLINGCLEMENTINE or MEBETARZANYOUBEJANE?
As a disk turns, who is deciding which letters to freeze and why?
Instead of an analogy for natural selection acting on random
mutation, the Dawkins-Sober scenario is actually an example of the
very opposite: an intelligent agent directing the construction of an
irreducibly complex system. The agent (Sober here) has the target
phrase (lock combination) in his mind and guides the result in that
direction as surely as a fortune-teller guides a Ouija board. This
hardly seems like a secure foundation upon which to build a
philosophy of biology. (Behe M.J., "Darwin's Black Box", 1996,
pp220-21)
Selection Requires Immediate Function
In this analogy, function (success) is achieved only when a meaningful
English sentence, or something closely approximating that, is
reached. After ten "generations" of valiant phrases, however, we still
do not observe a single English word, much less an entire phrase.
Thus, none of these early variants could possibly be selected for
function on the basis of the properties they immediately possess -
unless the illegitimate device of a distant target phrase, or goal, is
employed, to guide the sequence through the space of its initial non-
functional variants . To see that the lines 1 through 10 are actually
non-functional, cover lines 11 to 43 with a sheet of paper, and ask
someone unfamiliar with this analogy to describe the syntactical
significance (i.e., function in English) of any of the first ten lines.
Origin & Design 17.2 has a `computer print-out' of Dawkins "weasel"
program. It points out that Darwinian selection procedure can have no
target and natural selection must find some function immediately in any
variant:
"...the use of a target phrase strips this analogy of any resemblance to
biological reality. A strictly Darwinian selection procedure can have
no target. Rather, natural selection must find some function
immediately in any variant. In the analogy, however, the trick of
"evolving" the meaningless phrase WDLMNLT
DTBKWIRZREZLMQCO P depends entirely on the distant target
phrase METHINKS IT IS LIKE A WEASEL - a pattern occurring
43 generations in the future - because the first several "generations"
(or variant phrases created by the computer) are completely non-
functional . The target phrase guides a process that should have no
guide. If the analogy illustrates anything, therefore, it is the necessity
for an intelligently pre-established goal (in this case, the target
phrase) for the processes of random variation and selection. And that
contradicts the premise of undirected Darwinian evolution. Thus, to
employ this analogy as an illustration of Darwinism, Berlinski
observes, is "an achievement in self-deception." (Origins & Design,
Vol. 17, No. 2, Spring 1996, p12)
Lewontin includes Dawkins among "popularizers" who have "put unsubstantiated
assertions or counterfactual claims at the very center of the stories they
have retailed in the market":
"As to assertions; without adequate evidence, the literature of science
is filled with them especially the literature of popular science writing.
Carl Sagan's list of the "best contemporary science-popularizers"
includes E. O. Wilson, Lewis Thomas, and Richard Dawkins, each of
whom has put unsubstantiated assertions or counterfactual claims at
the very center of the stories they have retailed in the market."
(Lewontin R., "Billions and Billions of Demons", review of "The
Demon-Haunted World: Science as a Candle in the Dark" by Carl
Sagan, New York Review, January 9, 1997, pp30-31)
Lewontin adds that he is worried that sientists may believe what Dawkins
(and Wilson) tell them about evolution:
"Who am I to believe that quantum physics if not Steven Weinberg,
or about the solar system if not Carl Sagan? What worries me is that
they [other scientists] may believe what Dawkins and Wilson tell
them about evolution." (Lewontin R., ibid., pp30-31)
Regards.
Steve
-------------------------------------------------------------------
| Stephen E (Steve) Jones ,--_|\ sejones@ibm.net |
| 3 Hawker Avenue / Oz \ Steve.Jones@health.wa.gov.au |
| Warwick 6024 ->*_,--\_/ Phone +61 8 9448 7439 (These are |
| Perth, West Australia v my opinions, not my employer's) |
-------------------------------------------------------------------