The second (definitional) positive meaning of causes outside accepted mechanisms
A second, and conceptually quite distinct, sense of positivity for the concept of constraint also arises from a vernacular meaning of the word, but embodies a philosophical position about the general nature of theories and arguments in science, rather than a specific empirical claim about the nature of evolution. [Page 1033]
Consider ordinary linguistic usage for the following scenario: a favored or orthodox theory undergirds the basic research program of a discipline — the usual situation that Kuhn (1962) calls “normal science” as practiced under the influence of a reigning paradigm. All scientists know, of course, that rare episodes of transition between explanatory systems, often occurring with sufficient speed and upset — both structural (to theories) and emotional (to practitioners) — to be deemed revolutionary, mark our most interesting times. Moreover, nearly all scientists, if not utterly devoid of ambition or intellectual verve, regard the development of a new explanatory system as the highest form of achievement in their profession. Nonetheless, the full working careers of most scientists proceed in the usual mode of research within a basic paradigm — a “good life” full of interest and intellectual excitement, as any rich paradigm features forests of unsolved puzzles, and byways (or even substantial roads) of expansion and originality.
Within such a ruling theory, a set of accepted causes and mechanisms operates to yield a range of outcomes specified as permissible. (When too many inexplicable results become well documented outside this permissible range, ruling theories become strained, and an interesting time of theoretical transition may soon be at hand.) Now, as a purely linguistic point, what should we call a set of anomalous results that would not have occurred if our reigning theory held the dominant or exclusive sway usually granted to its precepts? What, for example, would we say about our inability to turn mercury into gold if our causal theory proclaimed the possibility of so doing, or (to choose a case of expansion rather than restriction) what would we call our newfound ability to generate living insects from decaying flesh if our theory dictated that only plants, but not animals, could originate by spontaneous generation?
We might, of course, eventually abandon our old theory for a novel system of explanation. But what if we do not choose to do so, at least not yet, and especially if we know that our theory really does work well, and as specified, for a large range of well documented cases? We would have to acknowledge that the old theory does not enjoy so wide or exclusive a domain of application as we had previously asserted. What would we then call the classes of exceptions — particularly the results of unorthodox causes that forced us to accept limitations upon the old beliefs? We generally label such exceptions as “constraints” because they restrict the range and power of our orthodox explanations.
I regard this conceptual meaning of constraint — the imposition of limits upon the range of orthodox theories by documentation of exceptions and demonstration of unorthodox causes — as undeniably “positive” in the important intellectual and psychological sense that any scientist worth his salt must cherish such upsetting discoveries for the conceptual challenges thus unleashed. Thus, if the Darwinian functionalism of natural selection acts as a reigning theory, then any documented constraint from internal channeling of variation — whether positive or negative in the empirical sense discussed in the last section — must be viewed as intellectually positive for questioning our orthodoxy and documenting something new and interesting that shouldn't [Page 1034] have happened under our usual views. Thus, this second, conceptually positive meaning of “constraint” embodies a relative concept that might be feared by those who enjoy the comforts of power, but should provoke the delight of all scientists.
To cite two recent examples of this relative meaning from recent evolutionary literature, Weiss (1990) wrote an iconoclastic paper arguing that geometric laws of serial repetition, combined with limited structural paths of alteration, dominate directions of phyletic change. (By the way, I disagree with his conclusion and am only discussing his terminology.) In this context, the usual orthodoxy of ordinary natural selection, working towards optimality in local adaptation, will be judged in an opposite manner as an annoying trifle that might falsify the grand pattern and temporarily hide its effects — in other words, as a constraint upon the regularity of geometrically predictable transformation. Weiss (1990, p. 21) acknowledged that natural selection occurs, but he dismissed the process as local distortion: “The pervasiveness of metameric 'duplication with variation' shows that it is a central principle of evolution . . . Despite . . . the pattern-distorting effects of selection and drift, this evolutionary strategy is essentially unidirectional.”
Weiss's taxonomy of concepts, so peculiar to those of us with Darwinian training, makes sense in his system. We would never think uniting selection and drift into the same category, for we view them as opposite processes with respect to our primary interest in adaptation. But, within a theory of predictable linear change enjoined by geometric principles, both drift and selection operate as local oddities that distort a broader and fundamental pattern. In any case, when we note how selection becomes a constraint upon a structuralist theory of geometrically rule-bound transformation, we can understand more easily why internal channels of preferred variation would be labeled as constraints upon a theory that ascribes all evolutionary direction to natural selection. (When a rebel labels one's own central belief as a limiting constraint, the generality of the usage becomes startlingly clear!)
In another example, Jackson and Cheetham (1999) cite punctuated equilibrium as constraining because phylogenetic patterns generated by this theory preclude several classes of results predicted by orthodox selectionist models of gradualistic anagenesis in populations. They write (1999, p. 72): “The realities of punctuation and stasis need to be better incorporated into evolutionary studies. Punctuated speciation does not contradict conventional neodarwinian mechanisms, but it does constrain the range of probable evolutionary scenarios for speciation, evolution of life histories and macro-evolutionary trends.” “Macroevolutionary trends,” they add in explanation (p. 76), “must arise through differential rates of origination and extinction, and not by adaptive evolution within single species.”
Several participants in debates about the evolutionary meaning of constraint (see Gould, 1989a) have explicitly embraced this relative definition. Stearns (1986), for example, properly rejected a usage so overly broad that the term would then lose all meaning — namely, the designation of all cause as “constraint” because any active force must direct change in one way rather [Page 1035] than down other conceivable paths. “The meaning of the word would [then] vanish,” Stearns notes (1986, p. 35). He therefore recommends: “We can preserve it in a relative sense if we recognize that it only has meaning in a local context where one concentrates on the possibilities latent in certain processes and views the limitations on those possibilities as arising from outside that context.” Therefore, in considering revisions and expansions of Darwinian theory, ordinary natural selection becomes the context and any force (like internally channeled variation) limiting its exclusive sway in directing evolutionary change, becomes a constraint.
Antonovics and van Tienderen (1991), in an influential article that cleared away much of the accumulating nonsense in definitional debates about constraint in evolution, also favored this relative concept as a solution. They agreed with my argument (Gould, 1989a) that “it is those factors that influence the process but are external to the favored theory that should be termed constraints” (Antonovics and van Tienderen, 1991, p. 167). But, choosing a terminology that struck them as more consistent with the ethos of scientific neutrality, they preferred the term “null model” to my “favored theory.” (I would reply that we do not usually refer to strong theories — like natural selection — based on particular and well-articulated causes, as “null models.”
I would also argue that nothing negative attends the admission that disciplines operate under favored theories — a “good thing” for science, so long as we retain flexibility for change and do not equate “favored” with “established”; and, especially, so long as we treat the status of “favored” as an impetus for challenge rather than passive acquiescence, as we manifestly do when we invoke constraints to rebut overly strict versions of natural selection.)
In any case, Antonovics and van Tienderen survey the literature and find, in support of the argument developed here, that “the overall null model used by most authors was one of evolution by natural selection (irrespective of the level of selection)” (p. 167), and that nearly all explicit claims for “constraints” upon change within populations and lineages “dealt with evolutionary constraints to adaptation by natural selection” (p. 166). They also noted the “odd” feature of relative definitions that strikes many scientists as paradoxical, but would not be so regarded if we accepted the honorable and inevitable principle, so familiar to philosophers of science and language, but still faced with discomfort by many scientific professionals, that all terminology must be “theory bound” — specifically, in this case, that orthodox results of one theory become constraints in other theories. They write (p. 167): “Given evolution by random drift as a null model, natural selection now becomes a constraint!” Yes, and appropriately so — with no exclamation point needed to register surprise.
Although I disagree with his particular recommendation, Eble (1999) published a thoughtful and conceptually innovative paper rooted in this important principle of the inevitability and appropriateness of theory-bound terminology. Eble notes, and brilliantly analyzes, two entirely distinct, but all too frequently conflated, meanings of “chance” and “randomness” in evolutionary theory. His article, entitled “On the dual nature of chance in evolutionary [Page 1036] biology and paleontology,” distinguishes the conventional statistical meaning from a particular and distinctive sense frequently employed in Darwinian literature — namely, “chance” defined as events occurring for reasons unrelated to the canonical mechanism of natural selection. Eble writes (1999, p. 77): “The gist of the evolutionary notion of chance is that events are independent of an organism's need and of the direction provided by natural selection in the process of adaptation.”
Eble discusses examples ranging widely across all scales of evolution, but we all know (and we all make excuses for the resulting confusion in our lectures to elementary courses) the most troubling and common case — the claim that mutational variation in populations, the fuel of natural selection, is “random. “ Of course, we know perfectly well that such usage does not invoke the usual mathematical concept of randomness, and that we only mean “unrelated to the direction of natural selection” — a point emphasized in Chapter 2 in my discussion of Darwin's need for isotropy in variation (see pp. 144–146). Eble (1999, p. 78) cites the acknowledgment of many biologists, and the analyses of such leading philosophers as Popper and Sober, of this almost “studied” confusion, including a quotation of my own statement (Gould, 1982b, p. 386): “By 'random' in this context, evolutionists mean only that variation is not inherently directed towards adaptation, not that all mutational changes are equally likely.”
(Eble recommends that we retain the words “chance” and “random” for both meanings, and then enforce the separation with the restricting adjectives “statistical” vs. “evolutionary” chance. He argues (p. 75) that “evolutionary studies . . . can benefit from the simultaneous application of statistical and evolutionary notions of chance” — defining the second concept as “independence from adaptation and the directionality imposed by natural selection,” a definition as clearly and explicitly “theory-bound” as any I have ever read. I agree entirely with Eble's analysis. I dissent only from his terminological decision to retain the word “chance” for both concepts, and to rely upon moderating adjectives to enforce the distinction. I would prefer the codification of a different name for the evolutionary meaning both because I don't trust the power of subsidiary adjectives to clarify the vital distinction, and because the statistical meaning represents such an important concept, in both science and practical human life, that exclusivity of usage might aid our uphill battle to educate people about the basic meaning of probability. But my terminological disagreement with Eble does not detract from my admiration for his clear characterization of the distinction, and his rich discussion of the largely unrecognized confusions thus generated.)
In any case, Eble's characterization of evolutionary “chance,” and his documentation of such extensive usage in a sense so contrary to the basic mathematical meaning of a fundamental term in science, only underscores the enormous range and influence of natural selection as our canonical theory. If organismal selection, and its key consequence of adaptation, have become so prototypical in defining how evolution works, and what evolution does, that we usually designate any other result as a “chance” phenomenon — even [Page 1037] though the outcome may have been generated in a deterministic manner by a process that would be called “causal” in any standard scientific usage — then we achieve a better understanding of how subtle, and how extensive, the clutches of convention can become, even among people committed to innovation and the value of novelty.
When language unconsciously promotes orthodox mechanisms, setting barriers against our examination of alternative modes of causality, then we should vigorously analyze our terminological usages to seek a clarity that might open new possibilities. When we understand the relative meaning of constraint as a theory-bound term, expressing the orthodoxy of selection and designating all other causes of change as limitations upon an expectation* — and when we come to view this relative sense of “constraint” as a positive definition that urges us to explore alternatives to standard explanations — then we can stand a terminological bias on its head, for potential use against the same conceptual lock that engendered such a peculiar terminology in the first place.
HETEROCHRONY AND ALLOMETRY AS THE LOCUS CLASSICUS
OF THE FIRST POSITIVE (EMPIRICAL) MEANING: CHANNELED DIRECTIONALITY BY CONSTRAINT
I advocate nothing original in asking evolutionists to focus upon the empirically positive concept of constraint as channels for change, rather than (as in the negative meaning) limits to natural selection imposed by insufficient raw material in variation. The “consensus paper” of Maynard Smith et al. (1985), while stressing a minimalist definition of absent variability for change in certain directions (as a strategy for achieving a “least common denominator” of agreement among authors of very disparate opinions), emphasized both the legitimacy and greater interest of the positive meaning: “Does development merely prevent evolution from following particular paths or does it also serve as a directing force, accounting in part for oriented features of various trends and patterns?” (Maynard Smith et al., 1985, p. 281). Alberch (1982, p. 313) also accentuated the positive by stressing the two great themes — saltations and channels, or speed and directionality — that have always anchored the formalist or structuralist critique of Darwinian functionalism (see Chapters 4–5): “Development does not only define the apportionment of phenotypic [Page 1038] variation upon which selection operates, but it can result in discontinuities and directionality in morphological transformations.” For further discussion of the positive meaning of constraint, see Riedl (1978), Gould (1980c, 1989a), and Wagner (1988). Note also that both statements, cited just above, define the positive meaning in explicit contrast with the more usual negative reading, while emphasizing the far greater evolutionary interest of the positive sense.
The familiar and conceptually conjoined realms of allometry and heterochrony define a locus classicus for positive constraints in providing a sensible link between the two central themes of speed (for ease) and channeling (for direction). If we wish to argue that biased channels of internally-set variation can aid natural select
ion or any other functional theme in evolution, where could we find a better example than ontogeny itself, especially when the course of life features substantial allometry across a broad range of size, and often of environment as well (especially for organisms with distinct phases of larva and adult, for example). After all, this fundamental channel already generates a series of well-adapted stages each time an organism grows to maturity, for all parts of the life cycle must “work” in the Darwinian world of environmental interaction, or else the organism would not exist. (See Chapter 5 for my historical discussion of orthogenesis, as advocated by Eimer, Hyatt, and Whitman, for longstanding recognition of ontogenetic allometry as the primary source of positively channeled constraints.)
If any of these phenotypes would benefit the organism at a different size or stage of life, or if any different combination of characters (reachable by retuning the rates of development among relevant features), might yield increased adaptation, then the existing channel of ordinary ontogeny already holds the raw material in a particularly effective state for evolutionary change. And the more pronounced the allometry, the greater the potential extent of such realizable change.
The Structure of Evolutionary Theory Page 165