But lineages in phyletic senility cannot prosper; their decreases in size denote waning viability, and their irregularities of growth signify the last gasp of faltering strength. A wounded or senile individual in a progressive race may uncoil in injury or dotage, whereas the same fate awaits all ordinary adults in regressive stocks. “Retrogressive characteristics . . . could be compared with the pathological conditions, normal or abnormal, of occasional diseased and senile individuals of the progressive series. They [regressive features] are... [Page 379] inherited with ever increasing effect in successive species, occasioning distortions and retrograde metamorphoses, and finally leading to the extinction of the race” (1880, p. 14).
Early in his monograph, Hyatt advances the conventional claim of scientific methodology — that his phylogeny should be judged and accepted on the criterion of objectivity in research: “These series, having been the result of no preconceived plan of arrangement as far as the author could judge, were considered to be approximately natural” (1880, p. 8). Yet Hyatt clearly falls victim to his own admonition. He did not establish his four-lineage scheme by any principle of ordering specimens in a manner that could be called “objective” or even bound by rules independent of his phyletic preferences. Hyatt's scheme of multiple parallel lineages represents a theoretical construction, dictated by his orthogenetic conviction about racial life cycles, not a proclamation of nature.
First of all, Hyatt could not separate his lineages with accuracy or confidence. He uses a method of “eyeballing” and, following the limits of his time, presents no statistical arguments. Such a failure to quantify need not derail a study in principle; we need no measuring rods to sort a mixed pile of sparrows and elephants into two groups. But the Steinheim planorbids interfinger and intergrade in the most complex manner, both spatially and temporally. The shells occupy a grand clump of morphospace, with sub-clumps here and there to be sure, but with no clear or persistent piles. No researcher since Hyatt has been able to specify four distinct lineages, each maintaining integrity through time.
Second and most important, Hyatt admitted that he could not use the standard method for establishing lineages — temporal succession in the fossil record. For he could not identify any clear stratigraphic sequences at all! Following his words about natural series and lack of preconceptions, Hyatt wrote: “These series . . . were assumed to be a reliable basis for working hypotheses, in spite of the fact, that no certain data with regard to succession in time were obtainable” (1880, p. 8). How then can phyletic series be established?
Hyatt then admits that, in the absence of stratigraphy, phyletic sequences must be identified by the expectations of a biological principle — dare we label it a preconception? — namely, the biogenetic law itself.* Hyatt even permitted himself to construct phyletic sequences from specimens found on the same bedding plane if successive stages of a recapitulatory series could be identified: “This assumption rests largely upon well known laws of heredity, such [Page 380] as these, that an animal found to repeat the stages of another animal of a closely allied species in the young, with the addition of new characteristics in the adult, may be considered to be either a lineal descendant of that species, or of some form common to both; that in such cases as these, whether the form or species occur mixed on the same level, or on different levels, there is but one natural arrangement” (1880, p. 8).
If recapitulation entered the argument as an a priori assumption, then so did the cognate notion (in Hyatt's mind at least) of racial life cycles. Hyatt assumed the very phenomenon he hoped to prove when he built his phyletic sequences in the absence of stratigraphic resolution, often from specimens on the same bedding plane. He resolved all violations of stratigraphic order by dictates of the “old age” theory. For example, both Hyatt and Hilgendorf regarded P. oxystomus as a derived branch of the main stock. But Hyatt interpreted this species as the base of an extensive progressive series, built upon variation within a single bedding plane because he could establish no empirical continuity to still higher stratigraphic levels. Hyatt's preferred order helped his argument immensely — for the idea of parallel yet non-synchronous lineages strongly supported his claim for internal necessity, not environmental entrainment, in the unfolding of phyletic stages. But he presented no evidence beyond the ordering power of the preconceived theory itself!
In a second example that engendered bitter debate with the followers of Hilgendorf, Hyatt claimed additional evidence for non-synchronous parallelism by finding the most advanced specimens of P. trochiformis (the acme of the main progressive lineage) in the lowest stratigraphic levels — leading to the assertion that this entire lineage unfolded with great rapidity (and providing more evidence for internal programming since other lineages would then be evolving much more slowly, and Hyatt could therefore identify differential phyletic vigor, rather than variation in environmental pressure, as the cause for disparities in evolutionary rate). But other researchers could only find P. trochiformis at high stratigraphic levels (Hyatt's lower specimens may have eroded from upper levels and washed down) — and therefore interpreted this lineage as evolving much more slowly.
If Hyatt maintained such overweening faith in the validity of his orthogenetic theory of ontogenetic programming, we can scarcely be surprised that he also read the supposed empirics of the Steinheim planorbids as a disproof of Darwinism — hence Darwin's negative response in receiving the monograph (see p. 372). Hyatt attempted to specify both the potential and the limits of natural selection for the Steinheim planorbids — and the restrictions overwhelmed the possibilities. Following the usual formalist critique of selection's creativity, Hyatt allowed that selection might explain why four lineages, rather than fewer or many more, became established, and why they continued to propagate. The origin of four lineages in the Steinheim lake, Hyatt tells us:
... appears therefore to be perfectly well accounted for by Darwin's theory of natural selection. In no other way can we possibly account for the [Page 381] selection of but four out of the varieties of PL levis, and the continuous propagation and increasing intensity of the differences which they exhibit. An examination . . . will show anyone how many variations are lost in each form or species of the series, and how few are continued. This can only be accounted for upon the supposition that those which survived possessed in some way advantages indicated by their peculiar variations, which enable them to propagate those variations, and suppress their less fortunate neighbours (1880, p. 26).
But how shall we explain the far more important issue of causes for actual directions of modification in the lineages — that is, the evolutionary changes themselves? Hyatt asks whether selection could be effective here: “Are these parallelisms adaptations, and can they possibly be attributed to the direct action of the uniform external environment upon the forms of the different series?” (1880, p. 19). Hyatt denies any formative power to the Darwinian mechanism. Selection can do the negative work of weeding and separating, but cannot perform the positive action — the essence of evolution — of changing and progressing.
Hyatt offers two major critiques of selection in the light of ontogenetically programmed orthogenesis. First, how can the functional premise of adaptation be supported when so much change occurs in the regressive mode following phyletic maturity? “Nothing can exceed the confidence with which the strict Darwinist assumes, without any appeal to observation, that all characteristics which are inherited are necessarily advantageous. Exactly the reverse is very often true” (1880, p. 101). Second, he recognizes that Darwin's system entails a crucial assumption of isotropic, undirected variability. Since the conveyor belt of the grand potential ontogeny introduces new characters with a decided bias, creativity resides in the internal directionality of variation. Natural selection can only work as a subsidiary force to the primary agent of directed variation. “Natural selection, in fact, is simply one of the transient conditions of the physical surround
ings, having no value as a cause of origin of characteristics” (1880, p. 102). The ontogenetic conveyor belt feeds new characters and creates evolutionary novelty; selection can only eliminate, separate and impose a little cosmetic shaping upon the internally generated trend:
Thus it may be said that the struggle for existence, and the survival of the fittest, is a secondary law grafted upon laws of growth, and governed by them in all its manifestations. The law of natural selection, as generally understood, assumes in the first place the existence of an animal type, of its descendants, and of a tendency to variation (indefinite and unlimited) in every one and all of these descendants, from which (an indefinite and unlimited) selection may take place during the struggle for existence between competing forms, destroying the weak and permitting only the strongest and fittest of these variations to survive. The truth is, as far as my studies have gone, that there is no such thing as indefinite or unlimited variations in any species... This obvious proposition, if admitted, [Page 382] leads at once to the question, what are the limits within which a species may vary? ... The limits of variation in the species have been found to correspond to the growth changes in an individual (1880, p. 20).
Hyatt viewed his orthogenetic theory as a contribution to the larger vision of 19th century mechanistic science — a hope that when we finally learn the laws of heredity and discover the principles of biased variability, then evolution shall become as predictable as ontogeny: “In all cases the individual and its series must change by growth along certain lines of modification, which it is but reasonable to suppose we shall someday be able to map out beforehand for a series of forms with the same precision that we can now forecast the metamorphoses of any individual in a given species” (1880, p. 18). Charles Darwin, who understood the contingent character of history, could not have disagreed more forcefully.
Hyatt's hardline version of orthogenesis offered no quarter for fruitful interaction with Darwinism. If all structuralist and formalist thinking, and all theories of channeled variation, existed only (and in principle) in this adversarial mode, then the important 19th century debate on orthogenesis could teach us little today. But my account of Eimer's orthogenetic theory has already illustrated the potential for useful interaction between internal orthogenesis and external adaptation — even though Eimer chose Lamarck rather than Darwin for the basis of his functionalist component. I shall next present the even more accommodating version of C. O. Whitman, to illustrate a maximal contrast with Hyatt, and to emphasize the possibility of Darwinian insight from orthogenesis.
Kurt Vonnegut introduced the useful word “karass” to describe groups of people who may not explicitly interact, or even know each other, but whose lives seem tied together by action and circumstance. Hyatt and Whitman (who did, in fact, know each other well) must have belonged to the same karass. Both studied under Louis Agassiz. (Hyatt spent his career in the Boston area, worked primarily on ammonites, and occupied the office that I now inhabit at the Museum of Comparative Zoology.) Both participated in the early days of summer courses in New England natural history and marine biology (Whitman studied with Agassiz on Penikese Island; Hyatt ran a teacher's school of natural history in Annisquam). Both men rank as the two key figures in the early days of the Marine Biological Laboratory at Woods Hole, Hyatt as first president of the board of trustees, Whitman as first director. Both devoted their major research efforts to formulating theories of orthogenesis, Hyatt with ammonites and snails, Whitman with pigeons. But their orthogenetic theories could not have differed more profoundly, particularly in their divergent attitudes toward complementarity with Darwinism. In understanding why the adversarial Hyatt reached a dead end, and in grasping the insight offered by the accommodating Whitman (who, for unfortunate historical reasons, gained very little historical impact for his orthogenetic views), we may better appreciate both the blind alleys and the vital themes of [Page 383] orthogenesis (and of formalism in general) in our struggle to develop a more adequate modern theory of evolution.
C. O. WHITMAN: AN ORTHOGENETIC DOVE IN DARWIN'S
WORLD OF PIGEONS
The old benediction — “may you live in interesting times” — has been regarded as either a blessing or a curse. Charles Otis Whitman certainly merited such an epithet, for his professional life spanned the greatest range of opposites through the grandest transition in ideas that biology has ever experienced. He began by studying with Louis Agassiz, last of the great and legitimate creationists, and ended as the chief American promoter of mechanistic embryology in the German tradition. He made his primary reputation in “cell lineage” studies of the fates and products of the earliest blastomeres. But, unlike most experimentalists of the time, he also pursued other research as a gifted natural historian and evolutionary theorist. His major work in later years, while he served as professor at the University of Chicago, treated a subject that could not have been more canonical for Darwinian evolutionary biology in the naturalistic tradition — heredity, variation, and evolution in Darwin's own chosen organism, the domestic pigeon. Whitman, however, used Darwin's pigeons to support orthogenesis, and to deny selection a primary or formative role in evolution.
C. O. Whitman died in 1910, of pneumonia contracted after working furiously on the first cold day of winter to provide shelter for his birds. (F. R. Lillie, once his assistant and later his successor at Woods Hole, eulogized his old boss: “In his zeal for his pigeons, he forgot himself.”) Whitman had never published an extensive defense of his orthogenetic theories. His diverse and voluminous writings were finally collated and published posthumously as a large three-volume monograph by the Carnegie Institute of Washington in 1919. The evolutionary debates of the early 20th century had been fierce, and finally won by the followers of Darwinian theory. I have often wondered how this history might have differed if this paramount biologist (Kellogg, 1907, p. 288, called Whitman “the Nestor of American zoologists”) had lived to publish what might have been the best empirical defense of orthogenesis. In any case, the posthumous and much delayed 1919 monograph was too disjointed, too incomplete and above all, too late, to win any potential influence.
Whitman accepted Kellogg's classification of evolutionary theories as auxiliary or alternative to Darwinism. He also agreed with Kellogg that three major alternatives fueled the great debate as the century turned: Lamarckism, orthogenesis, and macromutationism. Since Whitman rejected the inheritance of acquired characters with all Weismann's zeal, his own list of viable alternatives included only two theories. Since he also believed in the strict continuity of Darwinian gradualism, de Vries's mutationism held no appeal for him either (though he regarded the theory as a viable contender, while treating Lamarckism as a dead issue). These rejections left only orthogenesis as a [Page 384] potentially valid challenge to Darwinism. The first and most important volume of Whitman's monograph bears the title: “Orthogenetic Evolution in Pigeons.”
Whitman provides our best example for disproving the false equation of orthogenesis with some form of theistic teleology — the main source for current derision, and for our failure to grasp the strengths and serious recommendations of this approach. The link to teleology can be dismissed as not only wrong, but entirely backwards. No case could be clearer than Whitman's, for he spent a maximally distinguished career as one of the great mechanists of American experimental embryology. He did not conceive orthogenetic trends as mystical impulses from outside, but as mechanistic drives from within, based upon admittedly unknown laws of genetics and embryology. Consider the last words of his 1919 monograph (p. 194):
If we are to draw the line sharply between science and all transcendental and telistic mysticism, we must regard the germ-organism as wholly mundane in origin and nature. If the germ is a thing of evolution from purely physical foundations — and any contrary assumption is a denial of the evolution principle, then we may say that it is a self-building within the limits of physical conditions, and just as truly auto
nomic in its form and behavior as is the crystal. In the formation of a crystal self-determination is ever present, and so it must be in the case of the organism.
Orthogenesis therefore emerges as a favored a priori prediction of deterministic science. Whitman's opening words strike the same theme with a note of triumphal optimism:
Progress in science is better indicated by the viewpoints we attain than by massive accumulation of facts. Darwin's perspective made him a prodigy in the assimilation of facts and an easy victor in the greatest conflict science has thus far had to meet. His triumph has won for us a common height from which we see the whole world of living beings as well as all inorganic nature; phenomena of every order we now regard as expressions of natural causes. The supernatural has no longer a standing in science; it has vanished like a dream, and the halls consecrated to its thraldom of the intellect are becoming radiant with a more cheerful faith (1919, p. 3).
Moreover, the particular character and personal history of Whitman's mechanistic outlook suggested the specific form of his orthogenetic argument. His work on cell lineages had mapped the fate of the earliest blastomeres, and had indicated that the source of eventual organs could be specified even in minute and formless clumps of initial cells. If embryos grew so predictably, why should evolutionary change be devoid of similar order? Ontogeny, in other words, should serve both as a model and a source for evolution — a joint vision of directional change from within. Whitman, in fact, argued that ontogeny and phylogeny represent the same essential process: “Development is the one word that seems to me to best circumscribe the [Page 385] more general problems of biology. It is also the one word that best emphasizes the essential unity of ontogeny and phylogeny. These two terms have been used as if they stood for two distinct series of phenomena, when in reality they apply to one and the same series” (1919, p. 177). Phylogeny therefore becomes as determinate as growth itself: “Not only is the direction of the change hitherto discoverable, but its future course is predictable” (1919, p. 38).
The Structure of Evolutionary Theory Page 61