As species are produced and exterminated by slowly acting and still existing causes, and not by miraculous acts of creation and by catastrophes; and as the most important of all causes for organic change is one which is almost independent of altered and perhaps suddenly altered physical conditions, namely, the mutual relation of organism to organism — the improvement of one being entailing the improvement or the extermination of others; it follows, that the amount of organic change in the fossils of consecutive formations probably serves as a fair measure of the lapse of actual time (pp. 487-488).
In summary, Darwin's link of progress to biotic competition completes his argument against evolutionary systems (like Lamarck's) that propose separate forces for progress and adaptation, and that, as an unintended result, fall [Page 479] into the disabling paradox analyzed in Chapter 2: palpable phenomena are unimportant; while important phenomena remain intractable. Lamarck distinguished a lateral process of local adaptation from a linear force of progress. Adaptation, as a local event of potentially rapid occurrence, could be observed, but this diversionary change provided no insight into the orthogonal and more important vector of progress through substantial time.
Darwin wanders on the fringes of the same dilemma. He identifies natural selection as a force of local adaptation. He wishes to escape the Lamarckian paradox of orthogonal causes by arguing for strict uniformity and extrapolation. The palpable and local force of adaptation therefore becomes, by smooth extension, the source of all evolutionary change at all levels. But how then could Darwin render progress — an idea that we might dismiss today as a cultural bias (Gould, 1996a), but that Darwin, as an eminent Victorian, did not wish to abandon (see Richards, 1992; Ruse, 1996)? Natural selection cannot provide the answer all by itself and without auxiliary principles, for this force must work in Lyell`s world of non-directional uniformity. Natural selection, at the “bare bones” of its mechanism, only builds adaptation to changing local environments; the principle includes no statement about inherent directionality of any kind, not to mention progress.
Darwin resolved this tug of war between the logic of his theory and the needs of his century by invoking a particular ecological context as the normal stage for natural selection. If most ecosystems are chock full of life, and if selection usually operates in a regime of biotic competition, then the constant removal of inferior by superior forms will impart a progressive direction to evolutionary change in the long run. In opposition to most of his evolutionary predecessors (Lamarck in particular), who postulated a higher (and impalpable) realm of causality to encompass progress, Darwin stuck with his single level of immediate and testable natural selection — and ensured progress by adding a boundary condition about the state of ecology, rather than by devising an additional and untestable causal apparatus. By this ingenious strategy, Darwin managed to have his cake of unified theory at a single, accessible level, and also to satisfy his culture's hunger for rationalizing progress.
Uniformity on the Geological Stage
LYELL'S VICTORY IN FACT AND RHETORIC
I spoke in Chapter 2 of a “Goldilocks problem” in Darwin's views on the nature of environment and geological change. Since Darwin uses “trial and error” (with the organism proposing and environment disposing) as the chief metaphor in his predominantly externalist theory of change, the outer environment (biotic and abiotic) assumes a more important role in the theory of natural selection than in most other evolutionary accounts of the 19th century. For Darwin, environmental change must be neither too little, lest the external prod fail, nor too great, lest the prod become a determinant in itself, thus demoting the role of the organism. In practice, too little change only [Page 480] emerged as a serious option in a metaphorical sense, as embodied in Kelvin's argument for the restriction of geological time (see pp. 492–502). But too much change characterized the core of a geological system — catastrophism — that, if generally valid, would severely compromise Darwinism by the fundamental criterion of relative frequency.
Darwin's need for a “golden mean” of geological change flows from his extrapolationist premise that observable and small-scale natural selection can provide, by extension, the causal basis for life's history at grand scales of morphological transformation through geological time. Darwin rooted his defense of this premise in the validity of uniformitarianism, as preached by his guru, Charles Lyell. The uniformitarian defense of extrapolationism therefore undergirds the third leg of my proposed tripod for an “essence” of Darwinian theory. This Lyellian assumption buttresses the ordinary operation of natural selection in the immediacy of any ecological moment, but the theme of the first section of this chapter raises the ante by including Darwin's treatment of pattern on a geological stage. The raw mechanism of natural selection provides no direction for organic change, and yields no predictable order for life's history through time. However, by adding a set of distinctive ecological arguments to the bare-bones mechanics — notably the domination of overt biotic competition as a primary mode of struggle within perpetually crowded communities — Darwin could validate the central belief of his surrounding culture, the concept of progress, as a primary signal of life's history.
Thus, the “golden mean” of geological change became doubly important to Darwin, because both the general operation of natural selection, and his particular rationale for progress in macroevolution, require a Lyellian geological world. The specter of catastrophism also became much more potent in the light of Darwin's stipulation that biotic competition acts as the chief agent of direction in life's history. For if mass extinction (and other phenomena of “too much” environmental change) establish patterns in the history of life at too high a relative frequency, then biotic competition will be demoted, if not replaced, by an ordering force of opposite meaning — for mass extinctions introduce a powerfully confusing and potentially confuting new actor: the tumbling, whimsical wheel of fortune rather than the slow and steady wedge of progress.
The norms of science dictate that major works be presented as objective explorations of data, with general conclusions derived from empirical evidence and devised late in the process of discovery. But most seminal books in the history of science can be read as briefs for passionately held, elegantly articulated, brilliantly advocated (and, to be sure, well-defended) views of nature. As a premier example, Charles Lyell, a lawyer by profession, may have presented his epochal Principles of Geology in the conventional style of humble factual documentation. But this great work must be understood as perhaps the most explicit and most able brief ever presented in the guise of a major scientific treatise.
The sources of Lyell's success in promoting his uniformitarian view — which later emerged as such a fitting solution to Darwin's Goldilocks problem by [Page 481] describing an earth with a “just right,” intermediate and dependable level of geological change — have been extensively explored by scholars in the past thirty years, and a general consensus has emerged (Hooykaas, 1963; Simpson, 1963; Porter, 1976; and especially Rudwick, 1969; my own first publication developed some of the same ideas independently and in the midst of some embarrassing juvenilia — see Gould, 1965). Lyell presented a plethora of compelling and well-presented evidence in his favor (gradualism does, after all, maintain at least a respectable relative frequency among patterns of geological change); but he triumphed as much by force of rhetoric, as by strength in documentation. Two features of his rhetoric stand out for effectiveness.
1. He invented a persuasive dichotomy, pitting uniformity and rectitude on one side, against catastrophism and reaction on the other. Catastrophism, Lyell argued, represented everything that had stifled the development of geology in a dismal past — not only for the falsity of claims for worldwide paroxysmal change, but also (and especially) for the sterility of a method that sought to explain the past by causes that do not operate today on our slowly changing earth. In attacking his cardboard version of catastrophism, Lyell penned som
e of the finest polemical lines ever written by a scientist: “Never was there a dogma more calculated to foster indolence, and to blunt the keen edge of curiosity, than this assumption of the discordance between the former and existing causes of change.” Catastrophist geology became “a boundless field for speculation” that could “never rise to the rank of an exact science.” Lyell ended this volley with his most famous metaphor: “We see the ancient spirit of speculation revived, and a desire manifested to cut, rather than patiently to untie, the Gordian knot” (1833, volume 3, p. 6).
2. He took advantage of a “creative confusion” by extending the umbrella of his single term “uniformity” over a variety of concepts with differing status — thereby attempting to win assent for claims of dubious merit by giving them the same name as other arguments that all scientists accept as valid. In particular, Lyell stoutly defended — and defined as “uniformity” — a set of methodological assumptions included within any full and proper definition of science (and embraced with equal vigor by all serious catastrophists as well; see Gould, 1987b): especially the spatiotemporal invariance of natural law and the actualistic principle that hypothetical causes should not be postulated so long as observable modern processes can generate the phenomenon in question, at least in principle. But Lyell also extended the term “uniformity” to a set of empirical claims about the natural world — testable statements that might be true or false, but emphatically cannot be treated as methodological assumptions, necessarily embraced a priori as a license to practice science at all. Two of these “substantive uniformities” influenced Darwin greatly, and have echoed loudly through the 20th century as well: gradualism, or uniformity of rate (especially the production of large-scale phenomena by accumulation of ordinary, daily effects through immense stretches of time); and non-directionalism, or uniformity of state (the empirical pattern of ceaseless, often cycling modifications, without vectors of directional change). Lyell eventually abandoned uniformity of state, when he finally became convinced, [Page 482] in the 1850's, that the fossil record of vertebrates exhibited vectorial change, and when Darwin then argued so persuasively that evolution could serve Lyell as a strategy of minimal retreat, permitting him to retain all other major components of his world view, while moving to acknowledge life's directional history. Gradualism then became — and has remained ever since — the sole surviving cardinal claim of uniformity in the substantive mode.
In his most clever, and devastatingly effective, trope of rhetoric, Lyell argued that the substantive claims of “uniformity” must be valid because the basic practice of science requires that we accept a set of methodological assumptions bearing the same name despite their truly different status (“uniformity” of law and process). In so doing, Lyell managed to elevate a testable claim about gradualism to the status of a received a priori doctrine vital to the successful practice of science itself. This subtle conflation has exerted a profound, and largely negative, influence upon geology ever since, often serving to limit and stifle hypotheses about rates of processes, and to bring derision upon those who advocated even local catastrophes. (Consider the now standard story of J Harlen Bretz and his long reviled, but later vindicated, catastrophic explanation for the channelled scablands of Washington by sudden flooding — Gould, 1980d; Baker and Nummedal, 1978.) In the obvious contemporary example (see Chapter 12 for details), no one can comprehend the emotional vigor of the debate engendered by Alvarez's proposal for catastrophic mass extinction by extraterrestrial impact (Alvarez et al., 1980, and the oral history of Glen, 1994) without understanding the historical legacy of Lyell's successful and tricky rhetorical argument against catastrophism.
Ernst Mach and many others have truly (and famously) noted that, for “big” issues, scientific reform proceeds largely by persuading the next generation. Mach's claim has usually been cited in the somewhat cynical mode: one must wait for the old generation to die because nothing can change their minds. But the same transgenerational theme applies, in an oddly backwards manner, to false characterizations that win assent by force of rhetoric. Such misattributions don't persuade contemporaries who understand the subtleties of the real issues by direct experience. But, since historical memory tends to occupy only a narrow range from nonexistent to short among scientists, false versions begin to prevail as soon as the actual practitioners die, and cardboard can quickly replace flesh. Thus, anyone who knew Cuvier, Elie de Beaumont, or d'Orbigny, recognized their mental power, their scientific integrity, and the considerable empirical support enjoyed by their systems. But when these men died, Lyell's characterization persisted, and “catastrophism” became equated with anti-science and dogmatic theological reaction. The label stuck, and Lyell's rhetorical triumph placed catastrophism beyond the pale of scientific respectability.
The arms of misreason extend across generations. When primary documents disappear from sight,* textbook pap can clone itself, and resulting legends [Page 483] then beget further fantasy with little hope for correction within an established system of belief. Thus Cuvier, one of the greatest intellects of 19th century science, a child of the Enlightenment and a champion of rationality, became a miracle-mongering apologist for ecclesiastical reactionaries who had thrust their fingers into the crumbling dike of superstition in a vain effort to stem the inexorable advance of Lyellian science. Consider just two characterizations of Cuvier from leading geological textbooks of the last generation: Gilluly, Waters and Woodford (1959) on catastrophes: “These, he [Cuvier] believed, destroyed all existing life, and following each a whole new fauna was created: this doctrine, called Catastrophism, was unquestionably inspired by the Biblical story of the Deluge.” Or Stokes (1973, p. 37) on the progress of science: “Cuvier believed that Noah's flood was universal and had prepared the earth for its present inhabitants. The Church was happy to have the support of such an eminent scientist, and there is no doubt that Cuvier's great reputation delayed the acceptance of the more accurate views that ultimately prevailed.”
I don't raise this example in the abstract interest of intellectual justice. The acknowledgment of catastrophism as a viable alternative to Darwin's geological requirements establishes an important theme of this book, and a potent reforming force within modern evolutionary theory. We might pursue this issue only by assessing the validity of modern arguments in the catastrophic mode, thus continuing to ignore earlier history (usually because we accept the cardboard characterization of Cuvier and colleagues, and therefore regard contemporary claims as viable for the first time). I will discuss some modern rationales in Chapter 12; but, in this historical part of the book, I need to demonstrate that catastrophism contained important elements of validity from the start — elements that rebut Darwin's crucial claim for gradual accumulation of changes induced by biotic competition as the predominant vector of life's history. I therefore present the basic argument in the most important of all catastrophist texts, the Discours preliminaire of Georges Cuvier (1812, but in its canonical English translation by Jameson, 1818). I do not, of [Page 484] course, (claim that the demonization of catastrophism arose as a Darwinian plot, for Lyell’s effort long predates the acceptance of evolutionary theory. But I will argue that Darwin needed Lyellian geology to grant natural selection (and biotic competition) a dominant role in setting macroevolutionary pattern. Moreover, I do not claim that Cuvier was right and Lyell wrong — thereby “correcting” Lyell's persisting unfairness with a modern version that would be equally false and one-dimensional in the other direction. Cuvier should not be resurrected as more right than Lyell, but his views must be reassessed as sufficiently valid to revoke the license that Darwin recognized as so crucial for granting a dominant relative frequency to gradual geological change.
CATASTROPHISM AS GOOD SCIENCE: CUVIER'S ESSAY
A central irony pervades the story of Lyell's rhetorical victory over catastrophism. Textbook pap, extending the exaggeration even beyond Lyell's mischaracterization, has lev
eled two major charges against the catastrophists: first, that they downplayed or distorted geological facts to defend their a priori beliefs; second, that they invented theories primarily to support a religious traditionalism linked to a restricted time scale and the defense of Noah's flood and other Biblical stories.
I call this description ironic because all leading catastrophists embraced a general conception of science entirely contrary to this mischaracterization. The catastrophist synthesis, as a working theory, rested upon two pillars, one substantive and the other methodological. Substantively, as the name implies, the theory regarded major geological change as concentrated in infrequent bursts of global paroxysm.* But, in a coordinating theme of equal importance, this sequence of catastrophes imparted a directional history to the earth and life.
Most catastrophists viewed the series of paroxysms as diminishing in intensity through time. They also postulated a geological dynamics to explain the link of directionality with occasional paroxysm. The theory of the French geologist Elie de Beaumont summarized the postulated mechanics of catastrophism. The earth, as a result of “hot” formation under the nebular hypothesis of Kant and Laplace, has cooled continuously through time, thus establishing a primary directional vector of change. This secular cooling engenders a catastrophist dynamics, for the outer crust solidifies into a rigid shell, while the inner matter, still molten, contracts in cooling. The “pulling away” of this inner core from the rigid outer crust creates instability, resolved not by gradual [Page 485] change, but by rare, global paroxysms, when the crust fractures and collapses upon the shrunken core. Life's vector of progress records an increasing adaptation to harsher climates of a cooling earth.
The Structure of Evolutionary Theory Page 77