News and Commentary
Origins & Design 17:1

Colin Patterson Revisits His Famous Question about Evolution

Authoritative," conveying "an impression of moody rebelliousness," and "habitually pessimistic"--thus writer Tom Bethell described the paleontologist and systematist Colin Patterson, on first meeting him in 1983.1

Patterson, who works at the British Museum of Natural History in London, is one of the leaders of the philosophy of biological systematics known as "transformed cladistics." The public hubbub which surrounded "transformed" or "pattern" cladistics in the 1980s has now generally subsided, although the issues lying at the heart of the controversy have not. Indeed, Patterson has revisited those issues, which were captured in large measure by a famous question he first posed nearly fifteen years ago. "Can you tell me anything about evolution," he asked his listeners, "any one thing, that is true?"

On November 5, 1981, Patterson gave a now infamous talk at the American Museum of Natural History in New York, to the Systematics Discussion Group which met monthly at the museum. An unknown creationist in the audience secretly taped the talk, and a transcript was soon circulating as samizdat among creationists, and shortly thereafter among the scientific community at large.

The uncorrected transcript was plainly flawed--giving "Conbear" for "von Baer," for instance, and omitting the names of well-known biologists--but enough of Patterson's provocative points came through to ignite the firestorm which followed. Patterson soon came in for heavy criticism from the evolutionary community. The talk was much debated: what had he meant; was he simply tweaking noses in New York; what did he really think about evolution?

As creationist writers trumpeted the speech, Patterson retreated, understandably annoyed by the episode and the voluminous correspondence he received in its wake.

In August 1993, at a Systematics Association meeting in London, Patterson revisited his 1981 talk; specifically, the bearing of evolution on the practice and philosophy of systematics: ordering the relationships of organisms. In his recollections (published last year; see the notes on p.7), Patterson describes the background to the talk:

In November 1981, after an invitation from Donn Rosen [a fish systematist at the American Museum, now deceased], I gave a talk to the Systematics Discussion Group in the American Museum of Natural History. Donn asked me to talk on 'Evolutionism and Creationism', and it happened that just one week before my talk Ernst Mayr published a paper on systematics in Science (Mayr 1981). Mayr pointed out the deficiencies (in his view) of cladistics and phenetics, and noted that the 'connection with evolutionary principles is exceedingly tenuous in many recent cladistic writings.' For Mayr, classifications should incorporate such things as 'inferences on selection pressures, shifts of adaptive zones, evolutionary rates, and rates of evolutionary divergence.' Fired up by Mayr's paper, I gave a fairly radical talk in New York, comparing the effect of evolutionary theory on systematics with Gillespie's (1979, p. 8) characterization of pre-Darwinian creationism: 'not a research govering theory (since its power to explain was only verbal) but an antitheory, a void that had the function of knowledge but, as naturalists increasingly came to feel, conveyed none.' Unfortunately, and unknown to me, there was a creationist in my audience with a hidden tape recorder. A transcript of my talk was produced and circulated among creationists, and the talk has since been widely, and often inaccurately, quoted in creationist literature. 2

But despite the inaccuracies, Patterson's central question about evolution came through unmistakably:

But one sentence from the talk was accurately reproduced, and was perhaps quoted more than any other. The sentence was a rhetorical question; I quote it from a creationist source (Johnson 1991, p.10): 'Can you tell me anything about evolution, any one thing that is true?'3

The question still matters, Patterson argues, because evolution is still assumed to be the primary determinant of phylogenetic reasoning. But Patterson's agnosticism about evolution--expressed in 1981 as, "I had been working on this stuff for twenty years, and there was not one thing I knew about it"--continues today.

Patterson describes that agnosticism by looking at patterns in molecular data.

At first, he notes, he thought he had found answers to his own question:

In 1981, I knew of no sensible answer to the question, but in the ensuing decade I came to believe that there were two things I knew about evolution. First, that transitions [purines, adenine (A) and guanine (G), mutating to purines, e.g., A --> G; or pyrimidines, cytosine (C) and thymine (T), mutating to pyrimidines, e.g., T --> C] are more frequently fixed than transversions [where a purine mutates to a pyrimidine, or vice versa] and second, that at the level of DNA, the great majority of substitutions take place despite natural selection rather than because of it. 4

However, as Patterson continues, he came to doubt whether in seeing these patterns he was grasping the process of evolution:

...do transition bias and neutral substitution represent knowledge about evolution, or something else? Further, and more generally, why should I, a morphologist, claim to know something about molecular evolution but nothing of morphological evolution? 5

We must distinguish between patterns to be explained, Patterson urges, and the process theories by which we explain those patterns--a distinction foundational to the "transformed cladistic" perspective on systematics and phylogeny. The molecular patterns he observed, Patterson believes, are thus only data awaiting explanation.

"I therefore believe I was mistaken in thinking that I knew something about molecular evolution," he writes. "Instead, I know (or have learned) something about the properties of molecular data, and those properties are amongst the things that must be explained by evolutionary theory."

Patterson concludes:

...I mentioned a question ('Can you tell me anything you know about evolution?') that I have put to various biologists, and an answer that had been given: 'I know that evolution generates hierarchy.' In the framework of phylogenetic reconstruction and our current problems with it, another answer comes to mind: 'I know that evolution generates homoplasy' [or "convergence," in the older jargon of systematics]. In both cases, the answer is not quite accurate. It would be truer to say, 'I know that evolution explains hierarchy' or 'I know that evolution explains homoplasy.' We must remember the distinction between the cart--the explanation--and the horse--the data. And where models are introduced in phylogenetic reconstruction, we should prefer models dictated by features of the data to models derived from explanatory theories. 6

Among the issues raised by Patterson's revisiting of his 1981 question, probably the most significant is how do we know which patterns are real (and therefore, actually need explaining)? "Transition bias is data," he argues--but only if one assumes the common descent of the gene sequences in question, meaning, in most cases, macroevolution. For some design theorists, who doubt that macroevolution is possible, sequence comparisons between divergent animal phyla (for instance) will not be data showing a historical relationship from an unknown common ancestor--a relationship requiring causal explanation. Rather, those data may reflect any number of causes other than descent.

Data--"facts" requiring explanation--emerge against a background of causal possibility. Because they are highly skeptical of the possibility of macroevolution, some design theorists would deny that the systematic hierarchy is real, in the same sense that the branching pattern of an elm tree or the Hapsburg Dynasty is real, i.e., something extending through space and time.

Thus, they would seek a nominalist explanation for the systematic hierarchy (reflecting, perhaps, our desire to organize data in bifurcating trees), and would defend that by arguing that the transformations required by macroevolution are probabilistically inaccessible. (Other design theorists would accept the hierarchy as actual, of course, a genuine tree of life, but would argue that the tree exists only because of a designer's intervention to make the necessary transitions between forms possible.)

The real crunch between Patterson and design theorists, however, arises from the possibility of design. The presence of cytochrome c in, say, echinoderms, and the same protein in humans, gives evolutionists an historical linkage between the two forms, even in the absence of a theory of macroevolution, simply because high degrees of molecular similarity require a common cause.

For most methodological naturalists, that can only be descent with modification. 7

But suppose, as seems possible, that a designer employed the same cytochrome molecule in two distinct lineages (echinoderms and humans)? The usual naturalist response holds while this is certainly possible, any theory making such predictions is empty. Assertions about the designer's actions are unconstrained, and would fit whatever data turn up.

But that does not follow. A designer who is free to employ the same molecules in constructing diverse organisms is not therefore an agent who acts capriciously. Furthermore, a design theorist would locate the empirical content of his theory elsewhere, in those predictions which distinguish design from naturalistic descent. It is the whole case for design, taken broadly, that gives the theory empirical content. 8

But the step to the possibility of design, it seems, is a far longer stride than the shorter step from certainty to agnosticism about evolution. One hopes, in any case, that Patterson will join the debate, wherever he ends up standing.

A Colin Patterson Sampler

Notes

Copyright © 1996 Paul Nelson. All rights reserved. International copyright secured.
File Date: 6.22.96