The theorem of Organic Evolution is one thing; the problem of deciphering the lines of evolution, the order of phylogeny, the degrees of relationships and consanguinity, is quite another. Among the higher organisms we arrive at conclusions regarding these things by weighing much circumstantial evidence, by dealing with the resultant of many variations, and by considering the probability or improbability of many coincidences of cause and effect...
But in so far as forms can be shown to depend on the play of physical forces, and the variations of form to be directly due to simple quantitative variations in these, just so far are we thrown back on our guard before the biological conception of consanguinity, and compelled to revise the vague canons, which connect classification with phylogeny.
The physicist explains in terms of the properties of matter, and classifies according to a mathematical analysis, all the drops and forms of drops and associations of drops, all the kinds of froth and foam, which he may discover among inanimate things; and his task ends there. But when such forms, such conformations and configurations, occur among living things, then at once the biologist introduces his concepts of heredity, of historical evolution, of succession in time ... if fitness for a function, of adaptation to an environment, of higher and lower, of “better” and “worse.” This is the fundamental difference between the “explanations” of the physicist and those of the biologist.
In the order of physical and mathematical complexity there is no question of [Page 1207] the sequence of historic time. The forces that bring about the sphere, the cylinder or the ellipsoid are the same yesterday and tomorrow. A snow-crystal is the same to-day as when the first snows fell. The physical forces which mould the forms of Orbulina, of Astrorhiza, of Lagena or of Nodosaria to-day were still the same, and for aught we have reason to believe the physical conditions under which they worked were not appreciably different, in that yesterday which we call the Cretaceous epoch; or, for aught we know, throughout all that duration of time which is marked, but not measured, by the geological record.
An epilog to an argument
We can, today, easily identify D'Arcy Thompson's primary error as an expression of the venerable post hoc fallacy (“after this, therefore because of this”). He had correctly noted a strong correlation, throughout organic nature, between the forms of organisms and the shapes that inorganic objects assume under direct molding by physical causes acting upon them. He therefore advocated the simplest hypothesis that these physical causes had directly fashioned the organic forms (just as we would unhesitatingly assert for inorganic objects). Here he made an empirical rather than a logical error. That is, we cannot accuse D'Arcy Thompson of not recognizing the potential fallacy of drawing a causal inference (direct physical production) from the observation of a correlation (between realized organic forms and the idealized optima constructed by these physical forces in the inorganic realm). He understood perfectly well that biologists preferred a different and more complex explanation for the same generality — that the inorganic objects may be directly crafted, but that organisms generally achieve the same result by operation of a different kind of biological force, natural selection working by differential reproductive success and survival of the fittest. In the following passage, for example, D'Arcy Thompson separates the two arguments: first, the false inference of direct organic production, followed by the correct observation that organic forms obey physical laws (p. 10): “We want to see how, in some cases at least, the forms of living things, and of the parts of living things, can be explained by physical considerations, and to realise that, in general, no organic forms exist save such as are in conformity with ordinary physical laws.”
In other words, and using D'Arcy Thompson's favored Aristotelian terminology, he tried to depict the physical laws to which organic shapes conform so well as the actual efficient causes of these shapes. But, in general, Darwinians were right all along. These physical laws are formal causes, or blueprints of optimal adaptive designs for given circumstances of size, materials and ecology. The laws give us insight into the adaptive values, or final causes, of organic designs. But the efficient cause of good organic design is usually natural selection.
Ironically, this great student of Aristotle (D'Arcy Thompson wrote standard translations, still in print, for two of Aristotle's biological treatises) guessed wrong about the category of causes embodied in the correlation of [Page 1208] organic shapes with the optimal forms directly produced by physical forces. He regarded the correlation as a map of the actual efficient cause. This aspect of good organic design does express a final cause in adaptation, but any evolutionary changes must still must be crafted by an efficient cause — and Darwinian natural selection generally acts as the efficient cause we seek for our explanations.
But before we dismiss Growth and Form as a brilliant and wonderfully written disquisition rooted in a central error, we should pause to reflect upon the partial validity of D'Arcy Thompson's theory of direct impress, or “order for free” in current parlance (Kauffman, 1993). D'Arcy Thompson admitted that he could not apply his theory to explain the groundforms of complex creatures, which he then accepted as “givens” in his analysis of transformed coordinates. This admission scuttled more of his hopes for generality than he was ever willing to acknowledge. But D'Arcy Thompson's theory cannot be rejected as entirely, or even generally, wrong. Surely his arguments for the hexagonal forms of crowded corallites, and the conformity of the ends of hive cells to the Maraldi angle, are correct: organisms don't have genes “for” hexagonality per se. Developmental genetics may regulate the types of materials, and their rates and places of production. But hexagonal shapes probably arise, just as for inorganic materials in similar conditions, by direct shaping under laws of closest packing.
I am confident that biologists can trace the lineage of hippos deep into an artiodactyl past in the early Cenozoic — as a genuine historical particular of unique form, requiring a phylogenetic explanation. But when someone tells me that a particular form of bacterium has not changed for 3.5 billion years because the oldest of all fossils displays the same shape as some modern species, then I doubt that this correct observation teaches me anything about filiation in a particular and continuous lineage; whereas I may learn something about basic forms, homoplastically attained again and again during the history of life, and therefore bearing no particular phyletic message. And I may use D'Arcy Thompson's procedures to establish the probable reasons behind these shapes, whether or not my explanations lie in direct shaping by these physical forces (a strong possibility for the bacteria, but not for the hippos), or in the adaptive values of designs actually built by the efficient cause of natural selection.
ORDER FOR FREE AND REALMS OF RELEVANCE FOR
THOMPSONIAN STRUCTURALISM
In the most important modern work in the D'Arcy Thompsonian tradition, Kauffman (1993, p. 443) “invites our attention to the central theme of this [that is, his] book: Order in organisms may largely reflect spontaneous order in complex systems.” Kauffman retains strong feality to D'Arcy Thompson's central principle that the adaptive order of biological systems arises by direct imposition of physical forces — thus advocating an “externalist” form of structuralist thought that has not played a large role in the history of evolutionary biology. [Page 1209] But Kauffman's specific insights and foci of concern differ widely from D'Arcy Thompson's emphasis upon morphometric geometry under laws of classical Newtonian dynamics. In the two major disparities, Kauffman first calls upon a different set of physical principles that has recently inspired both professional activity and public interest. In a second, and welcome, difference, Kauffman wishes to abet selection by supplying “order for free” from the inherent nature of the physical world, whereas D'Arcy Thompson tried to develop a largely substitutional theory of adaptive form that would relegate natural selection of effective insignificance.
In stating his fealty and staking out his differences, K
auffman (1993, p. 643) pays homage to D'Arcy Thompson and acknowledges the “small trickling of intellectual tradition” that this “outlier” species of structuralism has engendered, although Kauffman would surely wish to enlarge the flow to Mississippian proportions:
D'Arcy Thompson's famous and elegant book On Growth and Form stands as one of the best efforts to find aspects of organismic order which can be understood as aspects which we might, on good grounds, expect. His enquiry, which led him to consider minimal energy surfaces, transformations of coordinate systems as a function of differential growth, and a whole beautiful panoply of phenomena, has stood as a persistent spring for a small trickling of intellectual tradition down through contemporary biology. Thompson applied classical physics to biology. It has been said that a weakness of some biologists is persistent physics-envy: the seeking of a deep structure to biology.
Kauffman then extols the virtues of physics-envy, while recommending that biologists redirect their jealousy away from the Newtonian mechanics that D'Arcy Thompson revered (p. 644): “There is a new physics aborning, and it is time to again fall open victim to physics-envy. For want of a better name, the area which is emerging is something like a theory of complex systems ... This book is an effort to continue in Thompson's tradition with the spirit now animating parts of physics. It seeks origins of order in the generic properties of complex systems.”
By pluralizing his title, and by being even more explicit in his subtitle, Kauffman emphasizes his different aim of arranging a marriage between selection and inherent order, with the latter as the older and more experienced partner who encourages a younger spouse to invigorate and direct the united effort emerging from a preexisting substrate: The Origins of Order. Self-Organization and Selection in Evolution. “I have made bold to suggest that much of the order seen in organisms is precisely the spontaneous order in the systems of which we are composed. Such order has beauty and elegance, casting an image of permanence and underlying law over biology. Evolution is not just 'chance caught on the wing.' It is not just a tinkering of the ad hoc, of bricolage, of contraption. It is emergent order honored and honed by selection” (p. 644).
“My own aim,” Kauffman adds (p. 26), “is not so much to challenge as to [Page 1210] broaden the neo-Darwinian tradition. For, despite its resilience, that tradition has surely grown without seriously attempting to integrate the ways in which simple and complex systems may spontaneously exhibit order.” “My aim throughout is to attempt to characterize . . . those aspects which may reflect the self-organized properties of the . . . system and those which reflect selection — and to determine a way of recognizing the marriage between the two” (p. 407).
Kauffman continually invokes two key phrases to epitomize his understanding of direct physical molding in the evolution of adaptive form in anatomy, ontogeny and interacting biological systems in general. First, he seeks to explicate the spontaneous “order for free,” to which systems naturally conform, and which provides natural selection with a rich substrate for fine-tuning and more specific molding. In stating his intentions for ontogeny, and recalling his similar conclusions for ecosystems, Kauffman writes (p. 409). (I like his phrase “gratuitously present” as a description of order for free):
Highly constrained, poised cell types and ordered patterns of gene activity, each able to change to only a few others, are gratuitously present in a vast class of genomic regulatory systems . . . The phase transition from one regime to another is governed by simple parameters of the system, such as richness of coupling among the variables. The order seen in ontogeny, I shall suggest, is just that which arises spontaneously in the powerfully ordered regime found in parallel-processing networks. Selection, I shall further suggest, by achieving genomic systems in the ordered regime near the boundary of chaos, is likely to have optimized the capacity of such systems to perform complex gene-coordination tasks and evolve effectively.
In the other important book from the 1990's on this view of life, Goodwin (1994, p. 186) emphasizes the “generic” nature of order for free: “Much (and perhaps most) of the order that we see in living nature is an expression of properties intrinsic to complex dynamic systems organized by simple rules of interaction among large numbers of elements. This order is generic, and what we see in evolution may be primarily an emergence of states generic to the dynamics of living systems.”
With his second phrase, Kauffman emphasizes evolvability rather than form or organization per se in arguing that biological systems naturally evolve to “adaptation at the edge of chaos.” He holds (p. 645) that “the capacity to evolve is itself subject to evolution and may have its own lawful properties. The construction principles permitting adaptation, too, may emerge as universals. Adaptation to the edge of chaos is just such a candidate construction principle.” Kauffman continually stresses the abstract, general and timeless nature of those aspects of biological order that he would ascribe to “the nature of things” rather than to any distinctively organic mechanism like natural selection (which can then act upon the inherent and generic properties to construct more specific utilities in particular environments).
For these reasons of non-competition at their different scales (generic order [Page 1211] vs. specific design) and the sequential nature of their interaction (generic order as given, selective order as superimposed refinement), the structural order generated by physical necessity meshes well with the functional order built by natural selection (p. xv): “Selection achieves and maintains complex systems poised on the boundary, or edge, between order and chaos. These systems are best able to coordinate complex tasks and evolve in a complex environment. The typical, or generic, properties of such poised systems emerge as potential ahistorical universals in biology.”
I have emphasized throughout this section that the unusual structuralist theme of spontaneous order externally imposed by physical law (and therefore so different from structuralism's conventional focus upon internal channeling set by phyletic history, and then encoded in genetic and developmental programs) enjoys its greatest potential strength in two areas where other styles of explanation either don't apply in principle, or have simply (and so far) failed to yield robust results: the origin of life and its early history up to the construction of a prokaryotic cell; and the explanation of broad, recurrent, and potentially ahistorical, or at least not phylogenetically constrained, patterns (the form of ecological pyramids, rather than the particular top carnivore resident thereupon, the right skewed distribution of life's complexity rather than the occupation of its realized neurological extremity by Homo sapiens). The structure of Kauffman's book affirms these foci (and the resulting promise of this approach through noninterference with the large and legitimate domain of necessarily historical explanation).
Life's origin and precellular history occupies one of three major sections in The Origins of Order. Kauffman stresses his message of physical necessity in designating this set of chapters as “The Crystallization of Life.” He structures his argument by attacking a straw man that, at least in the understanding of most paleontologists, fell from popularity more than 20 years ago when the fossil record yielded cells of bacterial form in the most ancient sediments that could preserve organic structures (3.5 to 3.6 billion years old). But this old and superseded claim of exceedingly low probability for life's origin does serve as a convenient foil for Kauffman's (and virtually the entire profession's) search for different answers rooted in the predictable and generic nature of organic chemistry and the physics of self-organizing systems (p. 285): “The second part of this book . . . explores a heretical possibility. The origin of life, rather than having been vastly improbable, is instead an expected collective property of complex systems of catalytic polymers and the molecules on which they act. Life, in a deep sense, crystallized as a collective self-reproducing metabolism in a space of possible organic reactions. If this is true, then the routes to life are many and its origin is profound, ye
t simple.”
Kauffman then devotes the other two sections of his book to the second theme of broad and timeless structural generalities behind the specific adaptive solutions crafted by the functionalist mechanism of natural selection: Part 1 on models for interaction on rugged fitness landscapes, and Part 3 on the order of ontogeny. As an example of structural generality behind functional specificity, Kauffman's model predicts that the waiting time for successful [Page 1212] long-step jumping to higher, but distant, adaptive peaks (contrasted with different rules for simple expansion to adjacent peaks) doubles after each successful transition, thus forecasting, for example, that if a first success emerges from two tries on average, the tenth will require a mean wait equal to the time needed for more than 1000 attempts. In this manner, and very sensibly in my opinion (see Gould, 1989c), Kauffman tries to account for the origin of all fossilizable metazoan phyla in the Cambrian explosion, followed by more than 500 million years featuring no further origin of body plans of such distinctly different design. But this level of generality offers no insight into the particular historical questions of taxon and time that have always defined the guts and soul of biology: why arthropods, and why then rather than a billion years before or after (with the latter, contingently plausible, scenario precluding my writing and your reading this book, among other differences between our actual world and innumerable sensible, but unrealized, alternatives).
The Structure of Evolutionary Theory Page 192