More on mathematicians

I seem to have annoyed someone by the moniker of 'dlamming', apparently a graduate student who is interested in yeast. While searching through my junk trackbacks, I discovered his response to my post on the mathematician William Hart (see here), whom he feels is more than adequately qualified to talk about evolution.

'dlamming' makes a number of claims. The first is simply that math is important to many aspects of biology. This is not terribly interesting or, for that matter, controversial and no one - certainly not me - has been claiming that math and mathematicians are not part of evolutionary studies. Math is a tool, a language, that enables us to quantify and model the natural world and it is a relevent to biology as it is to any other science. However, it is dlamming's second point that is deeply flawed. He claims that Hart (who describes himself as 'a mathematician and experienced computer programmer') is qualified to make comments on evolution sensu stricto. He takes at face value Hart's statement that 'I find it disconcerting that genetic and other evolutionary algorithms [EAs] are pointed to as evidence that evolution works. The fundamentally flawed principles that these "experiments" are based upon and the derisible simplicity of the systems "evolved" both have more to say about the failure of evolution to produce meaningful innovations than it does about the robustness of the theory of evolution." From there, he claims that Hart is more qualified than I to comment, because I "mainly study the morphology of bones."1

Now let's pay attention here. 'dlamming' takes Hart's comments on evolutionary algorithims (which may or may not be correct2) and extrapolates that Hart is correct in his critique of Darwinism. Obviously, logically, if EA's are invalid as proofs for evolution, that in no way invalidates evolution in the natural realm. I, for example, have never used EAs as evidence for evolution. Though it is possible to see them as a vindication of selection & mutation, EAs do not directly suuport organismic evolution.

What a mathematican says about EAs is of little relevence to evidence for or against the evolution of organisms. It would be the utmost of arrogance for a mathematican to claim that it was.

Update: 'dlamming' is Dudley Lamming, grad student at Harvard.

--
1 For the record, if I was asked, I would describe my field of interest/expertise as being the application of mathematical and statistical techniques to problems in the evolution of organismal form and diversity. Mathematics and evolutionary biology ... whoda thunk it.

2 Contrary to what 'dlamming' claims, Hart isn't obviously even qualified to talk about EAs. Yes, he is a mathematician, but his publications have been on "Modular Equations and Eta Evaluations," "Evaluation of the Dedekind Eta Function," and "Schlaefli Modular Equations for Generalized Weber Functions". I may be wrong, but this doesn't appear to have anything to do with EAs. Mathematics - like biology - is a broad area, and competency in one realm doesn't mean competency in another.

More like this

Yup, it is.

By John Lynch (not verified) on 01 Apr 2006 #permalink

From his CV - "Studying the regulation of aging in budding yeast and in humans."

Damn common descent and model organisms, or is it happy common designer, common design?

By Bruce Thompson (not verified) on 01 Apr 2006 #permalink

I'm always a bit bewildered at the expectations people who haven't spent a long time mucking around with mathematical models place on mathematical models. Many of these models are sufficiently idealized versions of the real systems of interest that they can really only answer "how possibly?" questions. Even our more "realistic" mathematical models depend a lot on assumptions we make about the target systems -- and our ability to feed in good assumptions may depend quite a lot on what features of the target system are susceptible to accurate and precise measurements.

Mathematics may be the language of nature, but I think nature spends a lot of time mumbling...

I don't claim that Hart is qualified to comment on evolution - I do, however, note that he claims to be (and gives reasons why). If we take his comments at face value, he seems to have put a decent amount of study into the question. You, however, saw fit to dismiss his criticisms out of hand due to his CV.

Now, you state above that your area of expertise is indeed in the area of "application of mathematical and statistical techniques," so you may be well qualified to comment of Hart's complaints about genetic algorithims. My apologies for not asking you personally. :) However, no one would find that out from your CV (maybe it needs some updating?). Maybe, just maybe, Hart's CV similarly misrepresents exactly what he has experience with, and we should concentrate on the substance of his arguments before dismissing them out of hand.

From Hart's letter:

As a mathematician and experienced computer programmer, I find it disconcerting that genetic and other evolutionary algorithms are pointed to as evidence that evolution works. The fundamentally flawed principles that these "experiments" are based upon and the derisible simplicity of the systems "evolved" both have more to say about the failure of evolution to produce meaningful innovations than it does about the robustness of the theory of evolution.

I'd like to see (1) what "fundamentally flawed principles" Hart sees, and (2) what a "meaningful" innovation might be in a theory that depends on descent with modification.

Would someone who actually works with genetic algorithms agree that their work is "derisibly simple"?

The fundamentally flawed principles that these "experiments" are based upon and the derisible simplicity of the systems "evolved" both have more to say about the failure of evolution to produce meaningful innovations than it does about the robustness of the theory of evolution.

1) If Hart can see any way to improve those principles I for one would be interested. Historically speaking, however, people who have claimed that the current generation of genetic algorithms are "fundamentally flawed" have not been able to back this up by proposing a better model

2) Does an evolved solution still count as derisibly simple if it's better than anything that the guy who set up the genetic algorithm could come up with?

Dudley,

Two quick points. Firstly, a CV is something specific not what either Hart or I available online. Secondly, and more importantly, "application of mathematical and statistical techniques to problems in the evolution of organismal form and diversity" does not mention EAs because, guess what, it has nothing to do with EAs and EAs have nothing to do with "the evolution of organismal form and diversity." Why would anyone claim competence in a field they know nothing about and is not relevent to the matter at hand? Oh, never mind.

I could claim to be widely read in, say, nuclear physics, but that doesnt mean that I truly understood what I read and have anything important to say. Hart's claim is solely backed up with "I've read stuff ... lots of stuff" - hardly a reason for you to support him.

By John Lynch (not verified) on 02 Apr 2006 #permalink

Lamming seems to have a problem with people possessing legitimate expertise and saying so (elitist and all that) yet seems to have no problem assigning expertise to people that merely claim to have it.

It is an odd reverse-elitism that smolders in the ranks of the anti-'Darwin' camp. They seem to think that everyone has sufficient 'expertise' to comment authoritatively on evolution-related topics unless, of course, they actually have expertise in the subject. Then, to point out the insufficiency of the claims fo the pseudoexperts is derided as elitism and snobbery and the like. Because after all, anybody's opinion on any topic is equally valid in creationdom.

Prof. Lynch wrote (April 1, 2006 -- footnote 2):

"Contrary to what 'dlamming' claims, Hart isn't obviously even qualified to talk about EAs. Yes, he is a mathematician, but his publications have been on "Modular Equations and Eta Evaluations," "Evaluation of the Dedekind Eta Function," and "Schlaefli Modular Equations for Generalized Weber Functions". I may be wrong, but this doesn't appear to have anything to do with EAs. Mathematics - like biology - is a broad area, and competency in one realm doesn't mean competency in another."

====

Once again, this attempt to dismiss arguments by reference to the writer's CV and the titles of his or her research interests and published articles.

Hmmm ... I have reviewed your ASU website, Prof. Lynch, and I see no evidence that you have an advanced degree in logic and argumentation. You have not published in the fields of informal logic and/or argumentation.

Therefore, using the rubric you apply to anyone who disagrees with you, I can conclude that you have no business making arguments and posting them publicly.

You aren't "qualified."

As it happens, I do have an advanced degree in applied informal logical and argumentation. I daily practice in the field. That permits me to belittle you publicly for your lack of such an advanced degree.

Right? Or have I missed something about your debunking methods?

By Justasking7 (not verified) on 30 Apr 2006 #permalink

>I do have an advanced degree in applied informal logical and argumentation. I daily practice in the field.

In other words, Richard Clayderman, a.k.a. "justasking7", is a lawyer.

Move along folks, nothing to see her.

By John Lynch (not verified) on 30 Apr 2006 #permalink