Tuesday, August 31, 2004

Hunter in Wonderland, Part One

It seems that Cornelius Hunter has forgotten the first rule of holes: When you're in a hole, stop digging.

As my regular readers are doubtless aware, I devoted three recent entries of this blog (available here, here, and here) to Hunter's essay “Why Evolution Fails the Test of Science” in William Dembski's anthology Uncommon Dissent. Hunter, showing a staggering lack of appreciation for all the time and patience I put into responding to his drivel, has expectorated a poorly-reasoned reply available here.

Over the next two weeks or so I will post three replies to Hunter's recent missive, corresponding to the three parts of Hunter's reply (one dealing with fossils, one dealing with the universal genetic code, and one dealing with anatomical homologies). I do not intend to participate further in this debate beyond that point; if Hunter chooses to reply to my reply, then he can have the last word.

Of course, I realize that by responding at all I'm violating the first rule of debate: Never argue with a fool, for people will forget who's who. But what can you do?

I'd also like to acknowledge the helpful comments of Richard Wein in preparing this reply. Some of the arguments I will be making here were suggested to me by him.

Now, on to the fossils. Hunter begins his reply as follows:

Rosenhouse first addresses my section entitled “The Fossil Record.” The section begins by explaining that the fossil record contains substantial evidence for evolution. I followed this with evidential problems in the fossil record. My point was that while there is positive evidence in the fossil record, that evidence is compromised by a number of important problems.

For instance, in the synapsid reptile to mammal transition there is a plethora of similar fossil species. Included are species with finely graded changes in the jaw anatomy. While this is evidence for evolution, we must also consider problems with this evidence. Ever since Stephen Jay Gould extolled these fossils as providing a “lovely sequence of intermediates,” they have too often been assumed to be without issue. But across the reptile-mammal transition ancestral-descendant relationships are not obvious. Also, there are large gaps between and within the reptiles and mammals. Organisms must have evolved so rapidly that they appear fully formed and diverse in the fossil record. And amidst this spread of fossil species convergent evolution must have occurred many times, as many similar designs, in similar species, could not have derived from a common ancestor.

In my original post I presented the following argument for why the mammal-like reptile fossils provide such strong confirmation of common descent:

The primary skeletal difference between reptiles and mammals is found in the strucutre of their jaws. Reptiles have three bones that connect their lower jaw to their upper jaw, while mammals have only one such bone. Two of the bones found in the reptilian jaw are nearly identical to two of the bones in the mammalian inner ear. If our hypothesis is that mammals evolved from reptiles, then there must at one time have been animals whose jaw structure was transitional between reptiles and mammals. There was a time when this hypothesis was considered so absurd that it was treated as prima facie evidence for the impossibility of evolution (I mean, if jaw bones became ear bones then the intermediates would have disfunctional jaws and disfunctional ears, right?) Yet the fossil record documents that animals having the required features actually existed.

Hunter ignores this argument completely.

Instead, he simply conflates two different questions. Namely, “Does the fossil record provide strong evidence of common descent?” with “Can we infer specific lines of descent from fossils alone?”. This error was present in the original essay as well, and I pointed it out in my reply. Hunter never gets around to replying to this rather obvious point, though he does mock it towards the end of his essay. His failure to take this point seriously does not speak well for his understanding of this subject.

Hunter protests that ancestor-descendant relationships are not obvious. Of course not. They never are from fossils alone. A fossil is a snapshot of a particular animal that existed long ago. It tells us only that an animal having certain combinations of features once existed. In this case the snapshots we have tell us that creatures exhibiting bizarre amalgams of reptilian and mammalian features, amalgams that have no proposed explanation outside of evolutionary theory, actually existed. But pile of bones A does not come with a little tag to tell us how it is related to piles of bones B and C. That's just a fact about fossils, not an indictment of evolutionary theory.

He next protests that that there are gaps “between and within the reptiles and mammals” and concludes from this that “Organisms must have evolved so rapidly that they appear fully formed and diverse in the fossil record.” The implication here is that the only reason a particular intermediate species fails to be represented in the fossil record is that it evolved itself out of existence too quickly to fossilize. That is hardly the only explanation for such a gap, however. The chances that any given species will fossilize are small indeed, even if the species existed for hundreds of thousands of years.

Furthermore, rapid evolution is not especially problematic in this context since we are talking about rapid in the geological sense. That is, we are talking about transitions that took place over periods of tens of thousands of years or so (as opposed to the millions of years in which paleontologists usually traffic). And as was shown by Stephen Jay Gould and Niles Eldredge, the pattern of stasis punctuated by relatively rapid bursts of change is the logical consequence of extrapolating the allopatric model of speciation to the fossil record. This was the basis of their theory of punctuated equilibrium, and we will have more to say about it later.

He closes his paragraph by repeating his bizarre assertion (he made the same point in his original essay) that convergent evolution must have occurred many times in the lineage leading from reptiles to mammals. In my original post I commented that I couldn't imagine what Hunter was getting at here. I still can't. Why could many similar designs not have derived from a common ancestor? Multiple lineages starting from the same ancestor begin with an identical design. Over time, the different lineages evolve differing variants of this design. Where's the mystery here? Nothing Hunter presented in his original essay sheds light on this question.

But he does offer some further thoughts on the subject here:

Examples of these can readily be found in paleontology textbooks [e.g., Benton, Carroll, Kemp, and Romer]. Rosenhouse, however, is apparently unaware of these basics of the fossil record, and expresses disbelief of any such outstanding issues. Regarding the cases of supposed convergent evolution [e.g., Benton, 291; Carroll, 377; Romer, 184], Rosenhouse can only confess that “I can't imagine what he's talking about.”

To explain his strange comment about convergent evolution, Hunter provides three page numbers in three different paleontology textbooks. No details. No quotes. Just the bare citation, made in the hope that people sympathetic to his cause will not check them out. Hunter wishes to persuade us that the various mammal-like reptiles preserved in the fossil record record alarming quantities of convergent evolution. Let's see if his references bear him out.

[Romer, 184] refers to the book Vertebrate Paleontology by Alfred Romer. Page 184 contains the following mention of convergent evolution:

It is certain that such mammalian characters as the reduction in phalanges and the formation of a secondary palate had been independently acquired in several therapsid groups. The cynodonts are sometimes thought to be mammalian ancestors, but there are minor features which debar them, some believe, from such position. Bauria in some features is closer still but may likewise prove to be somewhat off the path. But although the details of the phylogenetic history are still uncertain, the therapsid ancestry of mammals seems established.

Romer gives two examples of convergent evolution in the lineages of the mammal-like reptiles: reduction in phalanges (toes) and the formation of a secondary palate. A reduction in the size of the phalanges is hardly a major morphological change. It is not at all implausible to think that so minor a change in anatomy could be converged upon, especially if there was selection pressure to do so. Rampant convergent evolution would be a problem if we found several lineages evolving major, complex morphological innovations in parallel. Parallel reduction in toe size is not very impressive.

While we're at it, we should note that Romer draws a clear distinction between describing evolutionary trends and inferring specific lines of descent. Glad to see I'm in good company on that one.

Incidentally, page 184 of Romer also contains the following quote:

In the varied therapsid types, we span nearly the entire evolutionary gap between a primitive reptile and a mammal.

As for the development of a secondary palate, Hunter should have read Carroll's book (Vertebrate Paleontology and Evolution) more carefully. The parallel formation of the secondary palate in two different lineages of mammal-like reptiles is the only example of convergent evolution given by Carroll on page 377. Happily, he goes into some detail about the differing paths the different lineages took (please forgive the jargon):

In contrast with the gorgonopsians, the jaw musculature in early therocephalians expanded dorsally over the braincase, leaving only a narrow sagittal crest between the adductor chambers. Both therocephalians and cynodonts developed a secondary palate, but this structure evolved in different ways in the two groups. In primitive therocephalians (Figure 17-28b), the vomer participates along with the premaxillae and maxillae, but the palatine remains dorsal in position. In primitive cynodonts (see Figure 17-30b), the vomer remains dorsal to the secondary palate, which incorporates the palatine in its posterior border. In both groups, the epipterygoid is expanded as a plate of bone lateral to the braincase.

I still can't imagine which part of this is supposed to make me uncomfortable about evolutionary theory. Sounds like we have good fossil documentation for how two different lineages converged on a similar beneficial change.

By the way, Carroll begins his lengthy chapter on the origin of mammals with this statement:

The sequence from the early amniotes to the early mammals is the most fully documented of the major transitions in vertebrate evolution. (P. 361)

Sadly, the library here did not have a copy of Benton's book. Anyone want to guess what I would find if I had it in front of me?

Hunter then provides two quotes to back up his assertions about alarmingly rapid rates of evolution. The first comes from T. S. Kemp's book Fossils and Evolution:

The apparent rate of morphological change in the main lineages of the mammal-like reptiles varies. The sudden appearance of new higher taxa, families and even orders, immediately after a mass extinction, with all the features more or less developed, implies a very rapid evolution. [Kemp, 327]

I'm sure Kemp actually said this somewhere, but it certainly wasn't on page 327 of his 1999 book Fossils and Evolution, since that book only has 284 pages. On page 227 of that book he does begin a fascinating discussion of the reptile to mammal transition, which concludes on page 234, but the quote above is not part of it. Kemp does include a chapter on mass extinctions. I skimmed that chapter without finding the quote, but perhaps it was there. Perhaps Hunter can fill in the details of where, exactly, this quote appears.

But since I can't believe that Hunter would simply concoct a quote I will assume that the quote is accurate. Kemp was quite prolific on the subject of the reptile-mammal transition, so perhaps Hunter got the quote from some other piece of Kemp's writing. It would be nice to see more of the quote's context, since I suspect that Kemp was not presenting this statement as some sort of problem for evolution.

So let's take the quote at face value. Kemp is describing a relatively rapid rate of evolution following a mass extinction. What's the problem? We would expect evolutionary rates to be rapid after a mass extinction. Such extinctions come about as a result of major environmental shocks. After the extinction numerous niches have opened up that were previously closed, and those populations of animals that remain are likely to be far from their maximal fitness. Both of these facts suggest that species will evolve rapidly in the years following the extinction event. Why does Hunter think this is helpful to his cause?

Hunter's second quote comes from Carroll:

The transition between pelycosaurs and therapsids has not been documented. It may have involved an environmental shift, as well as changes in morphology and physiology. The therapsids are already quite diverse when they first appear in the Upper Permian of Russia. [Carroll, 397]

This is an especially odd quote for Hunter to cite. The topic under discussion is the reptile to mammal transition. Pelycosaurs and therapsids are two different orders of reptiles. And there is nothing in this quote to suggest that evolution was unusually rapid.

Here's Hunter's explanation of why this is supposed to be troubling for evolution:

With evolution, we must believe that the process speeds up so as to conveniently leave no trace in the fossil record. This results in new organisms, fully formed and diverse (rather than in a lineage). But Rosenhouse misses the point and wonders out loud what the alternative is. “Partially formed fossils?”

But there is no convenient speeding up of the evolutionary process, not in Carroll's quote and not anywhere else. As Carroll describes on page 362 of his book, Pelycosaurs begin appearing in the fossil record at the base of the Pennsylvanian era, which he dates at 320 million years ago. The Permian era dates from 286 million years ago to 248 million years ago. The Upper Permian would refer to the more recent portion of that range, but let's use the 286 mya date. That means we are discussing a transition that had, at a minimum, 34 million years in which to occur. Where's the conveniently rapid evolution here?

I did not miss Hunter's point because he has no good point to make. He has only fog and confusion to offer. And I stand by my charge that talking about certain fossils being fully formed doesn't make much sense. Every fossil represents an animal which, when it was alive, was fully formed. If Hunter's concern is that certain specific structures appear without a clear fossil history, then he should spell out precisely which structures he has in mind and what sorts of fossilizable intermediates he expects to find.

We continue:

Rosenhouse continues to misread my essay. For instance, I explained how the horse sequence has been touted for years as strong evidence, but that in fact the horse-like fossil species do not align as was once thought. In fact, they persist unchanged and co-exist in the fossil record. Nonetheless, in textbooks and museums a sanitized version of the horse sequence too often continued to be used as an outstanding example of evolution. I used a Niles Eldredge quote to illustrate the situation.

Well, we certainly can't fault Hunter for consistency. His argument here is wrong for the same reason his argument about the reptile-mammal transition was wrong. For the thousandth time, the issue is not whether we can infer specific lines of descent from fossils alone. The horse fossils we have are compelling evidence for common descent because they document that animals having features transitional between small, primitive, vaguely horse-like animals on the one hand, and modern horses on the other, actually existed.

Hunter's criticism about species not aligning right has to do, again, with inferring specific lines of descent. That the fossil record documents a large number of stable horse-like species has no relevance to the question of whether the horse fossils we have provide strong evidence of common descent.

Throughout his essay Hunter seems to harp on two aspects of the fossil record as being harmful to evolution: stasis and relatively rapid change. Since these are exactly the points Gould and Eldredge developed punctuated equilibrium to address, Hunter really ought to explain why he finds PE inadequate. Hunter does make one brief mention of punctuated equilibrium, and we will revisit this topic then.

Next comes this:

Rosenhouse criticizes my use of the Eldredge quote. Rosenhouse says I have misrepresented Eldredge because Eldredge does not question evolution; rather, Eldredge is merely bemoaning the over simplified, speculative presentation of the horse fossils. I agree, and in that passage I did not attribute anything more than this to Eldredge.

In my original essay I said explicitily that I felt that Eldredge's intention was perfectly clear in the quote Hunter cited. But the fact remains that Hunter presented it in a misleading context. Hunter wants to persuade us that the horse fossils we have are not strong evidence of common descent. It was in making that argument that he invoked Eldredge. But Eldredge does not agree with Hunter's assertion, as the further quote from Eldredge that I provided shows.

Hunter presents Eldredge's quote right after Hunter asserts that the horse sequence “...turned out to be more problematic than evolutionists thought” Eldredge's statement is described as an “admission”, as if he were conceding something that was embarrassing to evolutionists. But Eldredge does not see the horse sequence as problematic for evolution and he did not believe he was admitting anything.

Anyone reading Hunter's essay without already being familiar with Eldredge's thoughts on this subject would get the idea that Eldredge supports Hunter's argument. He does not. That is why Hunter's presentation was misleading.

Hunter continues:

The problem is not that I misrepresented Eldredge's point, but that Rosenhouse misrepresented my point. Rosenhouse next quotes Eldredge decrying the erroneous claim that the order of the horse fossils had been manipulated in order to mislead the public. But, of course, I said no such thing. The problem is not that the order was misrepresented, but that the nature of the fossil record was misrepresented. The actual fossil record presents a dizzying array of species with little evidence of gradual change. Instead, what we see are species appearing and persisting in virtually unchanged condition.

But there was no misrepresentation of the nature of the fossil record, as Eldredge states clearly in the quote I provided in my original essay. The museum exhibit Eldredge was discussing shows four horse fossils lined up in their proper chronological order. Lovely. Eldredge was simply observing that some people, who don't know a whole lot of biology, get the misleading impression that those are all the fossils there are and that evolution necessarily follows a linear, progressive course. Regrettable, but hardly insidious.

And I have already addressed the business about stasis and change. Since Hunter is so fond of Niles Eldredge, he really ought to give more consideration to what Eldredge (and Gould) have been saying on this subject since 1972. (More on this below)

Hunter then asserts:

What Huxley failed to understand, and what Darwin knew very well, was that evolution had to be presented as gradual. Allowing for jumps would have opened the door to a non natural explanation. Darwin's arguments for evolution would have had much less traction if, in the end, he was to admit that his new found process just happened to act in quick spurts so that new species appeared abruptly. Saltations would have to wait until a later time, after evolution was firmly in place as the accepted explanation.

A century later, Eldredge and co-worker Stephen Jay Gould could safely propose the notion of punctuated equilibrium. Then, and only then could the nature of the fossils be simultaneously confessed and assimilated. With evolution firmly in place as a fact, any such evidential problems could be relegated to the question of how, not whether, evolution occurred. Hence, Rosenhouse admits that while inferring specific lines of descent among fossil species is a thorny problem, nonetheless he reassures us that this “has nothing to do with whether fossils provide strong evidence for common descent.”

Oh, brother. As soon as you see a critic of evolution mention saltations in the context of explaining rapid evolution, you know you are dealing with a major-league crank. No one, not Gould, not Eldredge, not anyone, is talking about saltations. Punctuated Equilibrium has nothing to do with saltations. In fact, it has nothing to do with mechanisms at all. Gould and Eldredge have never suggested that anything other than classical neo-Darwinian mechanisms were necessary to explain the rapid bursts of evolution they described.

Here Hunter is taking the standard ID approach of presenting PE as a desperate kluge concocted to explain away inconvenient facts. That's nonsense, of course. As I already mentioned, PE is simply the extrapolation of the allopatric model of speciation to the fossil record, and this aspect of PE is uncontroversial among biologists (it was first suggested by the most orthodox of evolutionists, Ernst Mayr, in a 1954 paper). Patterns of stasis and relatively rapid change are irrelevant when the question is the viability of the hypothesis of common descent.

By presenting this cartoon version of a major biological theory, Hunter demonstrates that he is only interested in parroting ID talking points, and not in getting at the truth of anything.

Hunter concludes this portion of his essay by descending even further into silliness.

He quotes me as saying “There is no rival theory that makes the same predictions about the fossil record that evolution does. That is why the fossil record provides such good evidence for evolution.” His reply:

It is true that no other theory makes the same predictions as evolution. In fact, evolution can accommodate everything from species persisting unchanged for eons to the abrupt appearance of new species in the fossil record.

The advanced trilobites, for instance, can appear before the less advanced ones, and placental mammals can predate marsupials in Australia. Major new designs can appear out of nowhere, as though planted there. Furthermore, if on a distant planet we found that nothing but bacteria had lived there for billions of years, that too would become a prediction of evolution. Yes, Rosenhouse is correct that only evolution makes all these predictions. But there is a difference between fulfilled predictions and good evidence.

These concluding paragraphs are an object lesson for anyone choosing to wade into the anti-evolution literature: No matter how reasonable, how dignified, how nuanced the author tries to seem, the cartoons and the caricatures are always there just under the surface. In the last paragraph it was the garbage about saltationism. Now its the old “Evolution can explain anything” dodge.

We have already seen that Hunter has no reply to the specific claims about the fossil record that I made. The patterns of stasis and change that Hunter keeps harping on are nothing more than the consequences of how fossils are formed and standard models of speciation. There's almost no end to the number of fossil discoveries we could potentially make that would provide a severe blow to evolution. Cambrian rabbits, to pick a favorite example. No, evolution can't explain everything we could find. It's just that it happens to explain everything we do find.

Hunter closes by throwing out a few more bare assertions without any details or references. They are throwaway lines, intended only to impress the scientific illiterates who make up his target audience.

And his closing line is especially strange. Among scientists, fulfilled predictions are effectively equivalent with good evidence.

I have not yet read the remaining two parts of Hunter's essay. Why am I not optimistic about finding anything worth pondering there?