Substantial change in any domain usually follows such a scenario, and cannot unfold in smooth and untroubled gradualistic continuity. The venerable Hegelian triad of thesis ® antithesis ® synthesis may not adequately describe all examples of important change (see p. 23 for more on this general concept), but this classic philosophical model of tension and (often episodic) resolution seems more in tune with nature — or, to use the terminology of this book, higher in relative frequency among patterns of change. Human thought, unlike the evolution of life, does include the prospect of meaningful progress as a predictable outcome, especially in science where increasingly better understanding of an external reality can impose a fundamental organizing vector upon a historical process otherwise awash in quirks of individual personalities, and changing fashions of cultural preferences. Surely our views on the nature of taxonomic order have progressed (in the sense of better consonance with the true causes of diversity) from the eclecticism of Aldrovandi, to the coherent creationism of Linnaeus, to Cuvier's addition of a temporal dimension, to Darwin's evolutionary synthesis of space, form, and time.
The Hegelian triad proceeds by confrontation between old and new systems (thesis and antithesis), and by their melding into a novel theory preserving worthy aspects of both — synthesis. But the continuing interplay of confrontation and reconstitution does not spin in an endless circle, preceding nowhere. Useful synthesis builds a transformed structure, and does not merely shuffle an unaltered deck (or raise an unstable house of cards precariously built from unaltered parts). Darwin constructed a powerful antithesis to older evolutionary views rooted in predictable progress and internal drives. Modern versions of the three critiques now present a worthy antithesis to the limitations of strict Darwinism. The second part of this book presents this Hegelian antithesis, written in the hope and expectation of synthesis and improvement. The synthesis that must eventually emerge will build a distinct theoretical architecture, offering renewed pride in Darwin's vision and in the power of persistent critiques — a reconstitution and an improvement, waiting for the next antithesis that must lead us onward to the next of many future syntheses in the wondrous, eternal play of mind and nature.
[Page 592]
[Page 595]
CHAPTER EIGHT
Species as Individuals in the
Hierarchical Theory of Selection
The Evolutionary Definition of Individuality
AN INDIVIDUALISTIC PROLEGOMENON
The perceived excesses of the French Revolution may have sapped English enthusiasm for the tenets of Enlightenment Rationalism — the faith of Darwin's grandfather Erasmus. The subsequent romantic movement stressed opposite themes of emotion vs. logic, and national variety vs. universal reason. Charles Darwin, who revered his grandfather but also loved Wordsworth's poetry, received a firm grounding in both great philosophical and aesthetic traditions. He also — and perhaps as a direct result — maintained strong fascination for a central theme common to both movements, but for different reasons: the role of individuals as agents of change in larger systems. (The Enlightenment focussed on individuals as effective intellectual agents and inherent bearers of rights — “unalienable” in Jefferson's memorable phrase — and therefore as primary causal and moral agents in themselves, not as expendable items of a larger collectivity. The Romantics exalted individual effort as the motive force of social change through the actions of occasional heroes of higher sensibility.)
In any case, and whatever the deeper source, we do know that, as Darwin stitched together his theory of natural selection in 1838, he centered his major intellectual struggle in the few weeks before his “Malthusian” insight (Schweber, 1977) upon the role of individuals as primary causal agents of evolutionary pattern, even at largest scales (see full discussion in Chapter 2). He first studied the economic theory of Adam Smith through the major secondary source then available — Dugald Stewart's On the Life and Writing of Adam Smith. He expressed special fascination for Smith's distinctive notion that the overall optimality of an economy might best (and paradoxically) is fostered by allowing individuals to maximize personal profit without restraint (the doctrine known ever since as laissez faire, or “let do” — more roughly, “leave 'em alone”). He then read an extensive analysis of the work of the great Belgian statistician Adolphe Quetelet — particularly his central notion of I'homme moyen (average man), based on the aggregation of individual attributes into collectivities. [Page 596]
In the context of this conscious and directed search, we should not be surprised that Darwin's theory of natural selection rests upon the same central paradox that fueled Adam Smith's system: postulate a cause based on individuals ruthlessly pursuing their own benefits; an ordered polity will then arise as an incidental side consequence. No dismissal of Paley's omniscient God as the direct creator of general order could possibly have been more incisive, or more radical. We can therefore understand why Darwin insisted so strongly upon a single-level theory of natural selection — with struggle among individual bodies as a virtually exclusive locus of causality (see Chapters 2 and 3 for extended analysis). The downward shift of agency, from a purposively benevolent deity to the amoral self-interest of organisms, embodies the most distinctive and radical aspect of Darwinism.
Given Darwin's intense and conscious desire to restrict causality to competing organisms, I have been particularly struck, in researching and writing this book, by the inability of all-the most diligent, and most thoughtful, early selectionists to make such a system work fully and consistently — despite intense and clearly focussed efforts to “cash out” Darwin's vision. As discussed in Chapters 3 and 5, only two early evolutionists fully grasped the meaning of natural selection, and the logic behind Darwin's restriction to the level of organisms. The first, August Weismann, defended Darwin's system with utmost zeal, as he spoke with pride about the Allmacht (all-might, or omnipotence) of natural selection. He began by advocating rigid adherence to Darwin's level of organisms. But, in fighting the resurgent Lamarckism of late 19th century thought, Weismann had to descend a notch to postulate important “germinal selection” at the level of hereditary units. Late in his career, he recognized the logical generality enjoined by his admission of a second locus — which selection can work on objects with requisite properties at any level of the genealogical hierarchy. Weismann therefore articulated a fully hierarchical model of selection operating at several levels both below and above individual organisms. Moreover, he developed this full theory not in retreat or as a hedge, but as a compelling extension of selection's central logic — and as further testimony to an Allmacht even more inclusive.
The second, Hugo de Vries, postulated a macromutational mechanism that logically precluded the production of new species by gradual selection of intrapopulational variation. In so doing, he committed intellectual parricide against his personal hero, Charles Darwin — and this mental act caused him great psychological distress. He assuaged his feelings of guilt, and showed his understanding of the abstract logic of selection, by insisting that he remained loyal to Darwinism — but at a higher level of selection among species, rather than among organisms (see pp. 446–451).
We must also not neglect the man who had invested the most effort in holding the line at organismal selection, and who had the most to lose if such a restriction could not work — Charles Darwin himself. Darwin struggled mightily to bring all evolutionary phenomena, including a host of apparently exceptional items from hymenopteran colonies to prevention of interspecific hybridization, under the umbrella of organismic selection, often with truly ingenious [Page 597] formulations — and he ultimately failed, as all others had. The logic of his argument led Darwin to postulate higher-level selection at two crucial points (see pp. 127–137) — to explain the evolution of altruism in human societies by interdemic selection, and to encompass multiplication of species under his essential “principle of divergence” by a partial appeal to spe
cies selection. If none of the most rigorous and savvy early Darwinians could render evolution without some appeal to selection at levels higher than individual organisms, shall we not tentatively conclude that both the logic of the theory and nature's empirical record compel such an expansion, and the attendant notion of hierarchy?
THE MEANING OF INDIVIDUALITY AND THE EXPANSION OF THE DARWINIAN RESEARCH PROGRAM
We may agree with the strictest formulation of agency in Darwinian theory: natural selection works by a struggle (actual or metaphorical) among individuals for personal reproductive success. In other words, selection occurs when properties of a relevant individual interact with the environment in a causal way to influence the relative representation of whatever the individual contributes to the hereditary make-up of future generations. If we place the theory's causal focus so squarely upon individuals as agents, then we might suppose that Darwin's unitary perspective must apply: all results at all evolutionary scales must cascade from the causal process of selection among individuals, defined in the conventional vernacular manner as the bodies of organisms.
But, as Hamlet said of the fears that prevent suicide (an adaptation, some would no doubt argue, for keeping humans viable as Darwinian agents), “ay, there's the rub.” What is an individual? Are vernacular bodies the only objects in nature that merit such a designation — especially when discrete “bodiness” doesn't always define an unambiguous individual at the focal level of Darwin's intent (not to mention the difficulties encountered in trying to characterize entities at levels above and below bodies in the genealogical hierarchy of nature)?
For example, biologists spent more than a fruitless century trying to decide whether the parts of siphonophores are “persons” in a colony or organs of an organism — only to recognize that the question cannot be answered because both solutions can justly claim crucial and partial merit (see Gould, 1984d). Are grass blades or bamboo stalks bodies in their own right (as some aspects of functional organization suggest), or parts (called ramets) of a larger evolutionary individual (called a genet)? Do our feelings about definition shift when ramets become spatially discrete and therefore look just like conventional bodies — as in the parthenogenetic offspring of an aphid stem-mother (designated, in their totality, as a single El, or evolutionary individual, by Janzen, 1977)? And what shall we do with discrete bodies that maintain some genetic variation among themselves (and cannot, therefore, form a set of identical ramets), but operate together as differentiated items (analogs of organs) [Page 598] in a larger “totality” like a beehive or ant colony with a single queen? Wilson and Sober (1989) have urged a revival for the old concept of “super-organism” in such circumstances.
As so much uncertainty surrounds the issue of how we define an “individual” at the supposedly unambiguous level of Darwin's own intent, we should not be surprised that attempts to restrict the concept to organic bodies have yielded more confusion than resolution. Perhaps we should try a different and more general approach. Perhaps we should attempt to specify a set of minimal properties required to designate an organic entity as an “individual” — and then ask whether any objects at levels above or below traditional bodies possess these properties, and therefore qualify for inclusion under an expanded concept of individuality. If so, we might obtain a useful definition divorced from the happenstances of scale, and therefore sufficiently general to provide a deeper (and clearer) understanding for this central concept in Darwinism.
This subject has generated an enormous and often confusing literature throughout the history of Darwinian thought — and more so than ever before during the past twenty years. Some colleagues may wish to throw up their hands and brand the entire enterprise with labels usually invoked pejoratively by scientists — merely “semantic” or “philosophical.” Indeed, several of the finest contemporary philosophers of science have devoted considerable attention to this issue — see, for example and in alphabetical order, Brandon (1982), Hull (1980), Lloyd (1988), Sober (1984), and Wimsatt (1981). But I believe that both the volume and the confusion arise for two reasons that compel primary attention to the subject: the issue is both exceedingly difficult and enormously important.* The best scholars tend to gravitate to the most fascinating and portentous questions — and the confluence of extensive consideration by the most prominent philosophers of science (as mentioned above) and the most thoughtful evolutionary biologists from early days (Darwin, Weismann, de Vries, as discussed above) to current times cannot be accidental or wrong-headed. As a testimony to this current concern, and again in alphabetical order, I cite as a small sample: Arnold and Fristrup [Page 599] (1982), Dawkins (1976, 1982), Eldredge (1985a), Fisher (1958), Ghiselin (1974a and b), Leigh (1977), Lewontin (1970), Maynard Smith (1976), Stanley (1975, 1979), Vrba (1980; Gould and Vrba, 1982; Vrba and Gould, 1986), Williams (1966, 1994), D. S. Wilson (1983), and Wright (1980). Collaborations between philosophers and biologists have also added to the interest (for example, Wilson and Sober, 1994; Sober and Wilson, 1998; Lloyd and Gould, 1993; Gould and Lloyd, 1999).
Discussion of this most difficult and most important subject may be organized in a hundred different ways. I have chosen a point of entry that may seem peculiar or indulgent as an abstract philosophical question tenuously related to the “real” biology of organic objects: are species individuals or classes? As a twofold justification for this strategy, I found, first and personally, that I could best organize this material and place all subjects into logical sequence, if I started here and worked systematically outward through a particular net of implications. (Others, no doubt, would choose different beginnings and construct just as sensible and comprehensive a sequence.) Second and collectively, this particular philosophical question has been widely and passionately discussed in the biological literature, and has struck several scientists (e.g. Eldredge, 1995) as a potential centerpiece unwisely relegated to a peculiar periphery by many scholars.
In 1974, Michael T. Ghiselin published an article in Systematic Zoology under a title that I found insufferably self-indulgent at the time (especially since his manuscript directly followed my own densely empirical article on local geographic variation in the land snail Cerion bendalli on the Bahamian island of Abaco), but have since come to view as adequately justified: “A radical solution to the species problem.” In short, Ghiselin argued that many classical problems about species (not primarily or especially related to this chapter's topic of levels in selection) could be instantly resolved if we — in the Pauline manner of “scales falling from the eyes” — reversed our customary definition of species as classes (or universal categories that can “house” objects) and reconceptualized them instead as individuals (or particular things). A species then becomes a singular item — an evolutionary entity defined by both a unique historical genesis and a current particular cohesion.
I will not trace the large and complex trail that Ghiselin's proposal generated in the scientific and philosophical literature (see, for example, Ghiselin, 1987). In my reading and understanding, I do not think that any clean resolution can be stated, or even any consensus described. Perhaps we might best acknowledge, with Mayr (1982a and b), that the term “species,” as conventionally used and understood, includes statements about both classes and individuals. In this sense, the extensive discussion of Ghiselin's proposal sharpened our thinking, but provided no closure.
In another sense, however, and following a common (largely sociological) pathway in science, the explicit airing of such an interesting theme launched, or at least impacted in major ways, a substantial set of theoretical issues, including two of central importance to this book: the nature of evolution as a historical discipline, and the definition of individuality as crucial to the “units [Page 600] of selection” problem. In his initial article, Ghiselin (1974b, p. 543) dimly perceived the key implication for hierarchical selection if species be construed as individuals rather than classes, and selection (by D
arwin's definition) works on individuals — namely, that selection must also operate among species-individuals (and, by extension, potentially at several levels in a hierarchy of units, each properly construed as an “individual”). But Ghiselin did not complete his argument and grant full evolutionary individuality to species. “Species are units, and they have evolutionary importance, but the same may be said of organisms. Doubtless both organisms and species specialize. And probably organisms become adapted but species do not, except in so far as they consist of adapted organisms” (Ghiselin, 1974b, p. 543).
David Hull (1976), in the first major philosophical extension of Ghiselin's proposal, firmly linked the concept of species as individuals to the older issue of units (or levels) of selection, thus properly tying the rationale for a causal theory of hierarchical selection to the generalization of Darwin's key insight that selection can only operate by the differential reproductive success of “individuals”: “Entities at various levels of organization can function as units of selection if they possess the sort of organization most clearly exhibited by organisms: and such units of selection are individuals, not classes” (Hull, 1976, p. 182). In his important later article — the locus classicus of the pivotal distinction between “replicators” and “interactors” (see next section of this chapter) — Hull then added (1980, p. 315): “Individuality wanders from level to level, and as it does, so too does the level at which selection can occur.”
The Structure of Evolutionary Theory Page 95