Joined: May 2002
Here is the Berlinski-relevant bit of my recent t.o. post:
A mistaken popularization, even fairly widely repeated, does not amount to "scientific fraud" as Berlinski alledges. Further, it should be noted that some sites follow the paper closely, e.g. good ol' Don Lindsay's site from 1998 is fine and worth recommending to people.
Nilsson ('s lab?) describes this 1994 paper as "theoretical modelling", with full computer simulations in the "something we will do" category:
| Research in the Nilsson lab|
A long standing question has been how variation and selection can produce an imaging eye. We have previously approached this question by theoretical modelling (Nilsson and Pelger, 1994: A pessimistic estimate of the time required for an eye to evolve. Proc R Soc Lond B 256: 53-58) which demonstrate that even with rather weak selection, the structures of a focused camera type eye can evolve in less than half a million generations. We now continue this line of research by computer simulations of eye evolution. These simulations are made to accurately mimic a realistic genetic control, and involves selection from populations of partially mutated offspring. The project has three principal aims: 1, to understand the conditions and criteria that select the fundamental optical types of eye (compound and simple etc) during early eye evolution; 2, to better understand the fine tuning of eyes to special visual requirements and habitat conditions; 3, to provide an insight into the mechanisms of genetic control required for evolution in general and eyes in particular. (Dan-Eric Nilsson, Lars Gislén)
...however, I strongly doubt that the calculations that they performed to determine eye acuity (their measure of eye "goodness") were done by hand. The paper repeatedly speaks of functions, curves, etc. which, in order to plot them, would always be done with a computer. Therefore it was a computer model of a sort, although not what I would call "a stochastic computer simulation" which is what Dawkins was implying had occurred (he did not use those exact words, mind you).
I think that people have a kind of mystical notion about computer models, that if the model is "more complex than I can understand" then it is equivalent to "they modelled the world in all necessary detail inside the computer!!". This is particularly true if someone presents a nice animation as the output. This is not the case. All computer models are gross simplifications, and IMO there is no particular hard-and-fast distinction between a theoretical model where calculations are done with a computer and a computer simulation where there is some attempt to perform a simulation with random inputs and explicit time-steps. It is, to plagiarize from Darwin, a difference of degree, not of kind.
"Mere calculations" computer models, e.g. the kind that can be done in a spreadsheet, have a very valid place. E.g. for a class I once simulated atmospheric temperature profiles in various latitudinal zones (a 2-D climate model) in Excel. You can reproduce some important features of the atmosphere using this kind of simple model (and in fact modern Global Climate Models are in part just a repetition of this kind of model in 3D).
So, to sum up my assessment of Berlinski:
|How very remarkable all this is--inasmuch as there are no computer models mentioned, cited, or contained in Nilsson and Pelger's paper;|
|inasmuch as Dan-Erik Nilsson denies having based his work on any computer simulations;|
The DI reports that Nilsson says this, here:
"Evolution" Series Claim on Eye Evolution More Fiction than Fact
...but there is no indication that the complexities of defining "computer model" were discussed.
|inasmuch as Nilsson and Pelger never state that their task was to "set up computer models of evolving eyes" for any reason whatsoever;|
...well, Nilsson says that he is doing just that now.
|inasmuch as Nilsson and Pelger assume but do not prove the existence of "a smooth gradient of change, from flat skin to full camera eye, such that every intermediate is an improvement";|
Here I strongly disagree. This is exactly one of the things that they documented IMO: that at least at the morphological level, just such a continuum exists, given the starting point. Only small quantitative variations in "cell" position, thickness, density, etc. are required.
|and inasmuch as the original light-sensitive patch in Nilsson and Pelger's paper was never allowed to undergo "localized random mutations of its refractive index."|
This is true as far as I can tell; however we *know* that variation + selection results in the kinds of gradual "microevolutionary" changes that Nilsson & Pelger were modelling in their paper. The only changes they invoked in their model were of the finch-beak-size variation type: small and quantitative. This is pretty much the point of their paper.
|And how very remarkable again--inasmuch as there are no computer "screens" mentioned or cited by Nilsson and Pelger, no indication that their illustrations were computer-generated, and no evidence that they ever provided anyone with a real-time simulation of their paper where one could observe, "almost like a conjuring trick," the "swift and decisive" results of a process that they also happen to have designed.|
True, but this gets us back to what a naive view of a "computer model" Berlinski's depiction is. Still, Dawkins shouldn't have implied that such existed.
(But: Didn't the "Evolution" special have a section wherein Nilsson or somebody constructed a *physical* model of their eye evolution scenario, where various stages could be enacted, variations tried, and the resulting "vision" actually seen? I remember someone gradually inflating a lens or something at any rate, and the picture moving from blurry to being focused.)
|And yet again how very remarkable--inasmuch as Nilsson and Pelger's "computer-simulation model" did not home in unerringly on Mattiessen's ratio, Nilsson and Pelger having done all the homing themselves and thus sparing their model the trouble.|
Tis true, there was no "homing" in the form of a search algorithm --however, they did demonstrate progressive advantage for eye shapes approaching Mattieseen's ratio. The relative advantage can be calculated, which is what they did.
|Each and every one of these very remarkable asseverations can be explained as the result of carelessness only if" one first indicts their author for gross incompetence.|
Mistakes happen. I can think of more than a few that Berlinski has made, and yet he is the one declaring the central theory of modern biology to be no better than ID (which he now considers similarly unsupported).
BTW, it appears that someone has kindly put an HTML version of the Nilsson article online here (there is no pdf that I am aware of anywhere):
A pessimistic estimate...
(unfortunately there are no graphics included, and a few typos/format issues)
|FINAL QUESTIONS. Why, in the nine years since their work appeared, have Nilsson and Pelger never dissociated themselves from claims about their work that they know are unfounded?|
They probably did, somewhere. It is not the duty of authors to chase down every representation of their work, particularly *in other countries*. It might surprise Berlinski but not every country has the United States' hangup with evolution.
|This may not exactly be dishonest, but it hardly elicits admiration. More seriously, what of the various masters of indignation, those who are usually so quick to denounce critics of Darwin's theory as carrying out the devil's work? Eugenie Scott, Barbara Forrest, Lawrence Krauss, Robert T. Pennock, Philip Kitcher, Kelly Smith, Daniel Dennett, Paul Gross, Ken Miller, Steven Pinker--they are all warm from combat. Why have they never found reason to bring up the matter of the mammalian eye and the computer simulation that does not exist?|
What, are they obligated to dig through every popular science book, look up the original literature on every single topic and determine the degree of accuracy of representation? This can be done but it is not a small matter. I suspect that Berlinski's evolution-related works would not fare well. For instance:
|Commentary, Vol. 101, June 1996 No. 6 The Deniable Darwin|
Unflagging Success Darwin conceived of evolution in terms of small variations among organisms, variations which by a process of accretion allow one species to change continuously into another. This suggests a view in which living creatures are spread out smoothly over the great manifold of biological possibilities, like colors merging imperceptibly in a color chart.
Life, however, is absolutely nothing like this. Wherever one looks there is singularity, quirkiness, oddness, defiant individuality, and just plain weirdness. The male redback spider (Latrodectus hasselti), for example, is often consumed during copulation. Such is sexual cannibalism -- the result, biologists have long assumed, of "predatory females overcoming the defenses of weaker males." But it now appears that among Latrodectus basselti, the male is complicit in his own consumption. Having achieved intromission, this schnook performs a characteristic somersault, placing his abdomen directly over his partner's mouth. Such is sexual suicide-awfulness taken to a higher power.2
It might seem that sexual suicide confers no advantage on the spider, the male passing from ecstasy to extinction in the course of one and the same act. But spiders willing to pay for love are apparently favored by female spiders (no surprise, there); and female spiders with whom they mate, entomologists claim, are less likely to mate again. The male spider perishes; his preposterous line persists.
This explanation resolves one question only at the cost of inviting another: why such bizarre behavior? In no other Latrodectus species does the male perform that obliging somersault, offering his partner the oblation of his life as well as his love. Are there general principles that specify sexual suicide among this species, but that forbid sexual suicide elsewhere? If so, what are they?
Once asked, such questions tend to multiply like party guests. If evolutionary theory cannot answer them, what, then, is its use? Why is the Pitcher plant carnivorous, but not the thorn bush, and why does the Pacific salmon require fresh water to spawn, but not the Chilean sea bass? Why has the British thrush learned to hammer snails upon rocks, but not the British blackbird, which often starves to death in the midst of plenty? Why did the firefly discover bioluminescence, but not the wasp or the warrior ant; why do the bees do their dance, but not the spider or the flies; and why are women, but not cats, born without the sleek tails that would make them even more alluring than they already are?
Why? Yes, why? The question, simple, clear, intellectually respectable, was put to the Nobel laureate George Wald. "Various organisms try various things," he finally answered, his words functioning as a verbal shrug, "they keep what works and discard the rest."
But suppose the manifold of life were to be given a good solid yank, so that the Chilean sea bass but not the Pacific salmon required fresh water to spawn, or that ants but not fireflies flickered enticingly at twilight, or that women but not cats were born with lush tails. What then? An inversion of life's fundamental facts would, I suspect, present evolutionary biologists with few difficulties. Various organisms try various things. This idea is adapted to any contingency whatsoever, an interesting example of a Darwinian mechanism in the development of Darwinian thought itself.
A comparison with geology is instructive. No geological theory makes it possible to specify precisely a particular mountain's shape; but the underlying process of upthrust and crumbling is well understood, and geologists can specify something like a mountain's generic shape. This provides geological theory with a firm connection to reality. A mountain arranging itself in the shape of the letter "A" is not a physically possible object; it is excluded by geological theory.
The theory of evolution, by contrast, is incapable of ruling anything out of court. That job must be done by nature. But a theory that can confront any contingency with unflagging success cannot be falsified. Its control of the facts is an illusion.
I don't have answers to all of these questions, but just as an amateur I can provide some. Unfortunately Berlinski never saw fit to look any of these up, and none of his DI fellows have bothered to correct him since the article came out in 1996. Roughly in order:
1) Male-eating is a subset of the widespread practice (in arthropods) of males offering up nutrient "gifts" to females for the purposes of (1) obtaining copulation and (2) prolonging copulation and preventing other males from copulating. And it is not true that the males always get eaten -- some of them will run for their lives at the first hint of hostility. This is only my hazy summary, but there is a massive literature on the topic. See e.g.:
Choe, Jae C., Crespi, Bernard J. The Evolution of Mating Systems in Insects and Arachnids. Cambridge, 1997.
...I would like to see a theory other than evolution devote the kind of attention and scientific sophistication that evolutionary theory has brought to the issue. What evolution predicts is that there will be a selective benefit for whatever apparently crazy behavior one sees (assuming it is a species-wide trait that easily could/should be something else). And apparently entomologists have confirmed that this is the case in the spider that Berlinski cites, even though he seems not to appreciate the significance of the fact.
2) "Why is the pitcher plant carnivorous, but not the thorn bush?"
If Berlinski had bothered to get up off his tush and go to the library to consult the literature -- which after all is what libraries are for -- he wouldn't have been so mystified. In a 1989 article by Thomas Givnish (of the Dept. of Botany, University of Wisconsin) entitled "Ecology and Evolution of Carnivorous Plants," Berlinski's question is anticipated and answered. I have included translation of the scientific jargon [in square brackets like this] for non-nerds in the audience.
|Insofar as plants essentially trade carbohydrates [carbon compounds -- aka sugars -- generated from photosynthesis] for nutrients from animals or microbes in nitrogen fixation and ant-fed myrmecophily [ant-plant mutually beneficial relationships], the question naturally arises as to whether the same cost-benefit considerations and expected pattern of distribution apply to species with these associations as to carnivorous plants. With regard to nitrogen fixation, the answer is probably a qualified yes. Sunny, moist, nitrogen-poor conditions are most likely to favor nitrogen-fixing symbioses, as they do carnivores. However, the conditions favoring these two groups should differ in three important respects. First, because highly anaerobic [oxygen-deprived, as in bogs] conditions in the soil are inimical to nitrogen fixation in root nodules (Pate, 1986), nitrogen-fixing symbioses are more likely to occur in well-drained or seasonally arid sites than carnivores. Second, legumes and other nitrogen-fixing plants seem to be most competitive on sites that are relatively rich in other limiting nutrients, especially phosphorus (e.g., see Tilman, 1982); carnivores might be expected to have an advantage in soils that lack any nutrient that is abundant in prey carcasses [nitrogen can be fixed from the atmosphere, but phosphorus and other nutrients cannot be]. Third, nitrogen fixation always entails the use of nitrate reductase, and thus, of molybdenum; nitrogen-fixing symbioses should thus be excluded from molybdenum-poor soils, as they indeed are (Pate, 1986). This same exclusion should apply only to those carnivores that obligately produce nitrate reductase [that is, few to none of them --NT]. (Givnesh 1989, p. 282, emphasis added)|
In other words, the main answer to Berlinski's question is that nitrogen fixation doesn't work in the oxygen-poor muck of stagnant bogs, which gives the advantage to plants with an alternative means of acquiring nitrogen, namely carnivory. The secondary answer to Berlinski's question is that nitrogen fixing plants still need to get their trace nutrients (such as phosphorous) from the soil, and have the special requirement of molybdenum, so that in places where these nutrients are absent carnivorous plants have a further advantage.
Givnesh, T. J. (1989). Ecology and Evolution of Carnivorous Plants. Plant-Animal Interactions. W. G. Abrahamon. New York, McGraw-Hill Book Company: 243-290.
3) "[w]hy does the Pacific salmon require fresh water to spawn, but not the Chilean sea bass?"
I have no idea. Salmon are however related to a number of freshwater fish like trout. Are seabass? Is there an icthyologist in the audience? Did Berlinski ever consult one?
4) "Why has the British thrush learned to hammer snails upon rocks, but not the British blackbird, which often starves to death in the midst of plenty?"
I doubt that this even occurs as a regular matter. Why did Berlinski never provide a citation? I expect that the specialization of bird species on different foods would however be an important thing to consider.
5) "Why did the firefly discover bioluminescence, but not the wasp or the warrior ant."
Last I checked, wasps and ants were colonial social species where the queen is fertilized by one or a few males, IIRC just once before a long period of hive-making. Fireflies are, I expect, totally different. There are however chapters on both groups in Choe & Crespi (1996).
6) "why do the bees do their dance, but not the spider or the flies"
Well, gee, maybe because bees are colonial honey gatherers that live or die as a hive, and need to communicate the good nectar sources, while flies and spiders do nothing of the sort.
7) "why are women, but not cats, born without the sleek tails that would make them even more alluring than they already are?"
Berlinski is speaking for himself regarding attaction to tails, but the topic was recently discussed on t.o.:
|Code Sample |
> >And oh, another "little question". Why did humans lose their tails?
> Now, that's the question. Actually, the tails in that line were
> already lost in the last common ancestors of all apes [gibbons and
> great apes all lack (external) tails].
> Note that among Old World monkeys, there are groups like the
> genus Macaca with both long-tailed and almost tailless species.
> What might be a selective advantage for taillessness? It occurs to me
> that in an animal like an early ape [or a macaque] that doesn't use
> the tail for anything much, it's just something that can get injured,
> such as bitten in a fight and infected.
> >It seems like a good long strong tail could have supplied balance for
> >walking on two legs.
> But it was already too late by then. Our ape ancestors already
> lacked tails.
The pattern may be running on top of branches vs. brachiating (hanging
underneath them). For the former, a tail is handy for balance, for
the latter it's not. Once a primate gets big enough it switches to
brachiation and (several times indepedently, apparently), the tail is
reduced or lost.
Some macaques (see above, I don't know much about them)
Large lemurs -- e.g., the largest extant lemur, Indri indri, has but a
stubby little tail. I believe the larger (up to gorilla-sized)
extinct subfossil lemurs also have no tails.
...maybe others. These aren't primates, but do tree sloths have
...and I can add:
"With an increase in body size, a critical point is reached where a potentially unstable system will inevitably lead to overbalancing (fig. 3.21).
Figure 3.21 Caption: ...with increase in body size the ratio [of branch size to body size] becomes critical and overbalancing can occur; ...suspension provides one solution."
(Napier JR, and Napier, PH. The natural history of the primates, MIT Press, Cambridge, 1997, p. 44, )
...and note that humans are descended from large, brachiating -- and not coincidentally, tailless -- apes.
Berlinski's comparison of geology to evolution is more apt than he knows. He can see how geology gives good overall explanations of mountains, if not every detail -- but for some reason he can't see that evolution does the same with organisms. Evolution predicts correlation of traits with the environment. Berlinski never looks at the environment (physical and ecological) in which his particular traits of interest occur, so he doesn't understand why they occur.[/quote]
Returning to Berlinski's original article in this thread:
|And what should we call such a state of affairs? I suggest that scientific fraud will do as well as any other term.|
If Dawkins' mistake is fraud, then Berlinski's multiple and repeated mistakes amount to a far ranging conspiracy (perhaps, a conspiracy to keep himself from understanding biology). I suggest that pronouncing on biology, without first knowing any, is the greater sin.
This anti-Berlinski rant has been developing in my head for some time. Thanks to Sunday morning for allowing its expression...
Edited by niiicholas on April 05 2003,01:03