Dawkins’ Information Challenge: an Open Letter to Richard Dawkins after reading parts of a devil’s Chaplain



Yüklə 54,5 Kb.
tarix19.07.2018
ölçüsü54,5 Kb.
#56792

Brig Klyce /Astrobiology Research Trust / [168 Grove Park Circle] / Memphis, TN / [38117] / USA

[901-763-2222] / http://www.panspermia.org/art.htm / bklyce@pansermia.org


4 January 2005

Dr. Richard Dawkins

Charles Simonyi Professor of the Public Understanding of Science

Department of Zoology

University of Oxford

Oxford, UK

Dear Dr. Dawkins;
In an unsent letter to the New York Review of Books, you explain why you plan to ignore any future criticism of Darwinism by creationists (1). You add, “In the unlikely event that a significant argument should ever emerge from the ranks of creationism/’intelligent design’, we will be happy to debate it.” I am not by any measure a creationist, but I think some of their arguments already merit consideration.
In 1997, creationists from Australia challenged you to “give an example of a genetic mutation or an evolutionary process which can be seen to increase the information in the genome.” Although the challenge could have been stated better, it points to a significant issue. Yet in your 1998 response, also reprinted in A Devil’s Chaplain, you restate it trivially as, “to ask whether the information content increases in evolution.” Of course it does; nobody disagrees about that.
The creationists had asked for an example of “a genetic mutation” or “an evolutionary process”. For a process, you focus on the duplication and divergence by which the “dozen or so globin genes inside you” appear to have descended from a single ancient globin gene. But for your featured example, globins-from-globins is awfully weak. (More on this weakness below.)
Ultimately you deal explicitly with mutations, which, you observe, increase uncertainty and decrease information. However, selection removes the uncertainty, thus increasing information, “by definition,” you emphasize. Yes, selection can eliminate harmful mutations. But your logic here establishes that selection increases information only up to its earlier, pre-mutation level – not higher.
New information, or new programs?
In your defense, the creationists did not state the question carefully enough. Most importantly, the term, “information,” is imprecise. Omitting the nonsense, one can also distinguish data versus programs. In evolution, new features require new programs *.

For an example of evolution that adds only data, consider your own computer model named “Biomorphs” (2). It has programs that randomly change the values of a few variables (fewer than twenty). These changes cause the screen images to change. But the variables are merely data. Changing them adds only data, not new programs. And the selection process is best described as “optimization”. Even you do not claim that “Biomorphs” will ever evolve beyond the screen images they originally are.


Now suppose that biological genomes have a similar capability – to change randomly certain specific variables. (Is this phenomenon already known as “directed mutation”?) Natural selection acting on a duplicated gene undergoing such mutations might cause it to evolve rapidly. (Is this what “positive selection” is?) These processes might well account for the divergence of the globin genes you describe. But this manner of divergence adds only data, and the process of selecting the best data is, again, optimization. Of course, optimizing evolution in biology can have tremendous benefits. For example, the optimization of a duplicated gene could even account for the evolution from dichromatic to trichromatic vision (3). But optimization has a strictly limited range which does not extend to writing wholly new genetic programs needed for new features, the big steps in evolution.
So in restating the problem, first I would ask not for increased information, but new genetic programs.
And I would add two other provisions.
Gene transfer
Nowhere in your essay do you mention one family of process that already, clearly “can be seen to increase the information in the genome,” namely, gene transfer **. Among prokaryotes, lateral or horizontal gene transfer is now accepted as the primary means by which, not merely information, but new genetic programs are acquired. Among eukaryotes as well, we now recognize hundreds of examples of transferred programs – by an even greater array of mechanisms than the handful available to prokaryotes. A striking instance is the acquisition of critical components of the vertebrate immune system by gene transfer (4).
No one doubts that gene transfer can add new genetic programs to a genome. But in Darwinian theory, these programs must have been created de novo by another process at some point. That’s the process the creationists and I want to know about. To demonstrate that a given new program was created by a strictly Darwinian process, one must eliminate transfer as the ultimate source for it.
In an evolutionary experiment, gene transfer can be ruled out if the studied genome’s experimental environment is a “closed system” like Richard Lenski’s system for studying cloned bacteria (5). Whereas adaptation by program-switching is often seen in his trials, so far they appear to have produced no new genetic programs. These experiments in biology are certainly worthwhile, although they are time-consuming.
A closed system that solves the time problem is available in computer models. In silico, if the starting instructions are well known and new inputs are not permitted after the experiment begins, the computer version of gene transfer may be safely ignored. Examples of computer models used to study evolution include your own “Biomorphs,” Tom Ray’s “Tierra,” John Koza’s “Genetic Programming,” and several others. These models can demonstrate optimization and puzzle solving, and they can be instructive. But new programs do not emerge from them. Rather, they seem, in the end, merely to execute existing programs. And besides, not everyone accepts their applicability for biological evolution. (More on new programs in computer models below.)
A different way to demonstrate the strictly Darwinian evolution of new genetic programs is to use genomic sequence data to reconstruct the past, as you have done with the globin genes. This method is less straightforward than real-time experiments in biology or computers. The process is inferential, and a closed system cannot be imposed on the past. Nevertheless, as sequence databases have grown it has become the primary method for studying genomic evolution. And with this method a fair case for Darwinism could still be made. If new genetic programs arise in the Darwinian manner, without the benefit of gene transfer, one should be able to reconstruct the process as a series of plausible small steps such as point mutations. If you could show, for example, how something totally unlike a globin gene evolved by point mutations into a globin gene, the latter would count as a new genetic program, and the process would be fairly good evidence for strictly Darwinian evolution.
Whether by a closed system experiment, or a stepwise genomic historical reconstruction, a Darwinian example needs to stand on its own, without benefit of gene transfer. So that provision would be my second amendment to the creationists’ challenge.
Sustainability
A Darwinist could say that computer models like Tierra or Genetic Programming actually do produce “new programs,” because the files they create are executable. Even if so, these models have a very near limit beyond which they have never yet demonstrated the capability to evolve. That’s why to me it appears that the models merely execute existing programs. However, an argument about the definition of “new programs” would probably be unproductive, and a convincing demonstration could easily settle the matter. If computer models can reproduce the evolution we observe in biology, they should be able to compose new programs, not just once, following the rules with which they were created. They should compose them beyond that point, without an obvious limit, in a sustainable manner.
In biology, many studies show that recombinations of exon-sized blocks of DNA have produced new genetic programs. For example, Tre2 in humans is apparently a hybrid of two different genes (6). Such recombinations could be late events in a multistage process that begins with gene transfer. On the other hand, a Darwinist might reasonably see constructive recombination as simply “wholesale mutation” – as capable of writing new genetic programs as point mutations were once said to be.
Therefore, a single instance of a new program that becomes (became) effective after a genetic recombination is ambiguous evidence. It might be said to exemplify evolution by gene transfer or Darwinian wholesale mutation. Even in a closed biological experiment, if one insertion event assembles a new program, gene transfer may have supplied the inserted segment, already composed, into the genome under study before the experiment began.
But if recombination without ongoing gene transfer continues to write new programs in a sustainable manner, that’s different. In that case, Darwinian evolution would appear to have the virtually unlimited capability that it claims.
To avoid arguing about computer models, and to treat the problem of large insertions, therefore, I would also add the phrase, “in a sustainable manner.”
The challenge restated
In sum, I would restate the challenge to Darwinism this way: Give an example of a process besides gene transfer that adds new genetic programs to a genome in a sustainable manner.
If I have studied you well, you will feel that this is still the same old problem as, “How could something as complex as an eye evolve?” Perhaps it is the same. But genomes containing the genetic programs for eyes may well acquired them by gene transfer. Because this account is possible, evidence confirming the strictly Darwinian account for new genetic programs needs to be better than it is now.
“Without stirring from your armchair,” you could claim that the evidence I am requesting is unnecessary because everyone knows that the universe began with a big bang, confining all of life to a very large genetically closed system. If so, you could say that a universal closed experiment has already run, with clear results favoring Darwinism. But surely this response is insufficient. If Darwinian processes can write new genetic programs, there should be much firmer and more immediate evidence than the often-amended big bang theory provides.
And if you know me already, you could say that I, like the creationists, have a hidden agenda. In my case it would be the strong version of panspermia. But no, it is not hidden; it is simply not the subject of this essay. (And unlike creationism, strong panspermia is fully scientific.)
The challenge I am offering here is a legitimate scientific one that a true theory should be able to meet. If it is met, Darwinism will gain much-needed support. But if it cannot be met, perhaps it is time to consider scientific alternatives or major amendments to the current theory of evolution, whether the one I promote, or others. Clearly, if Darwinism does not fully account for evolution, we need scientific options, not supernatural ones.
Sincerely,

Brig Klyce



Astrobiology Research Trust

* “Programs,” as used here, may be broadly construed to include their essential genetic components, analogous to algorithms, subroutines or executable files.

** Gene transfer processes include transformation, conjugation, transduction, hybridization, infection, symbiosis, and others.
References


  1. Richard Dawkins, “Unfinished Correspondence with a Darwinian Heavyweight,” pp 218-222, A Devil’s Chaplain, Mariner Books, 2004.

  2. Richard Dawkins, The Blind Watchmaker, W.W. Norton and Company, 1987.

  3. K.S. Dulai, et al., "The evolution of trichromatic color vision by opsin gene duplication in New World and Old World primates," pp 629-638 v 9 n 7, Genome Research, July 1999.

  4. Alka Agrawal et al., "Transposition mediated by RAG1 and RAG2 and its implications for the evolution of the immune system," pp 744-751 v 394, Nature, 20 August 1998.

  5. Tim F. Cooper et al., "Parallel changes in gene expression after 20,000 generations of evolution in Escherichia coli," pp 1072-1077 v 100 n 3, Proc. Nat. Acad. Sci. USA, 4 February 2003.

  6. Charles A. Paulding et al., "The Tre2 (USP6) oncogene is a hominoid-specific gene," pp 2507-2511 v 100, Proc. Nat. Acad. Sci. USA, 4 March 2003.

Yüklə 54,5 Kb.

Dostları ilə paylaş:




Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©genderi.org 2024
rəhbərliyinə müraciət

    Ana səhifə