Gould on Adaptationism

Francis F. Steen
Department of English, UCSB

Gould on Adaptationism and Evolutionary Psychology
A review of Stephen J. Gould, "Evolution: The Pleasures of Pluralism" (New York Review of Books, June 26, 1997).

Introduction
Stephen J. Gould has long been one of my favorite authors, long before I became interested in evolutionary psychology. His ability to blend the discourses of culture and nature, his sustained curiosity and suspended laughter, and his unrestrained knowledge of the most obscure of facts endeared his voice to me, as it has to thousands of others. This eloquent spokesman for biology to the masses is also a major player in a certain disagreement in biology. Gould was intensely opposed to sociobiology, and is now turning his ire on the adaptationists in general and evolutionary psychology in particular.

Let me say at the outset that I do not find much of substance in this most recent debate. The stakes appear to be mainly those of status and reputation, which magnify every straw of difference to a forest of spears. The underlying issue, however, remains one of central interest: how do we integrate our understanding of biology with our understanding of culture? Gould, a Harvard paleontologist with a clear humanistic bent, should be in an excellent position to undertake this work.

The first part of this article, published on July 12, begins by setting up Gould's straw opponents, the "fundamentalists." This initially appears to be little more than a slugger for striking out Maynard Smith's unfortunate remark that Gould is seen by many as "a man whose ideas are so confused as to be hardly worth bothering with" (37), a charge that calls for heavy artillery. The fundamentalists are a prominent group of misguided biologists, a tight band held together by a single credo: that natural selection explains everything about the origin of species, an attitude characterized as adaptationism.

However, one would have to look hard to find a member, since the broad arguments Gould uses to mark himself apart are surprisingly inclusive:

natural selection does not explain or account for all aspects of the modification of species (no debate--no theory explains everything, pluralism is a pragmatic strategy)
evolution is path-dependent: "limitation of possibilities rather than adaptive honing to perfection becomes a dominant theme in evolution" (35). Why "becomes"? This is such a fundamental fact about the theory of descent with modification that it can hardly be characterized as news. However, the opposition is not between "limitation of possibilities" and "adaptive honing"--these clearly must operate together. The contrast is with "adaptive honing to perfection"--a rhetorical flourish rather than an adaptationist concept.
external factors matter: climates change, meteors fall, tectonic plates shift--this is hardly a source of debate
not all changes in the genes are adaptive; many are neutral (no debate)

This leaves one point that is perhaps a real source of disagreements: the model of punctuated equilibrium. What are the theoretical implications of this model? Does it require that "long-term evolutionary trends be explained as the distinctive success of some species versus others"? And if so, is this an argument against adaptationism?

First of all, this is clearly itself an adaptationist argument. The disagreement centers around what can usefully be considered a unit of selection--the allele, the genome, the organism, the species, etc. It is not clear to me that this argument is substantive; it appears that different events may act on different levels of order. However, this is at least a genuinely contentious area.¹

In sum, however, Gould doesn't do a very convincing job in depicting anyone in particular--Richard Dawkins and Daniel Dennett are mentioned--as adaptationist monomaniacs. Whatever it is he's getting at--whether the curve ball thrown him by Maynard Smith or a more serious and more general challenge--the first article fails to get a hit. After such a light aperitif, the pièce de resistance must lie in part two, which I will comment on in some detail.

The Role of Adaptationism in Evolutionary Psychology

In broad strokes, and without any call for evidence, Gould starts out the second article (full text below) by reiterating the notion that "the 'fundamentalists' among evolutionary theorists revel in the belief that one overarching law--Darwin's central principle of natural selection--can render the full complexity of outcomes" (¶2). Dividing the race of biologists in twain isn't the greatest proof of a pluralistic vision, and has little basis in reality:

ultra-Darwinian fundamentalists pursue one true way, while pluralists seek to identify a set of interacting explanatory modes, all fully intelligible, although not reducible to a single grand principle like natural selection. (¶3)

However, there is something at stake, and we're getting to it: he will take up the role of adaptationism in evolutionary psychology. Noted foe of sociobiology, Gould takes a benign view of the emerging discipline:

Evolutionary psychology could, in my view, become a fruitful science by replacing its current penchant for narrow, and often barren, speculation with respect for the pluralistic range of available alternatives that are just as evolutionary in status, more probable in actual occurrence, and not limited to the blinkered view that evolutionary explanations must identify adaptations produced by natural selection. (¶4)

The olive branch has evidently been extended; are we ready to throw narrow speculation overboard and embrace pluralism? The meat is on the table.

Gould's response to Dennett's critique

The paleontological record shows long periods of stasis, followed by rapid speciation. This argues against "the slow and continuous transformation of single populations" (¶6), and suggests that forces acting on whole populations are important. I take this point to be valid, and argue that different units of selection are relevant in different environmental events. It is hard to construe this as an argument against adaptationism, though I think Gould is correct in saying it is an argument against single-minded gene-level adaptationism. This point remains controversial.
That contingency and chance play a vital role in the evolution of life (¶8) is not controversial; it is a cornerstone of Darwinian evolutionary theory. Humility may be appropriate in the face of chance events at all levels, and chance by definition describes the limits to our knowledge. However, when life carries the hallmarks of functional design, invoking chance is tantamount to denying there is anything to explain.

Gould's third point is that many functional features of organisms are by-products of adaptations rather than adaptations in themselves--like the spandrels of San Marco, which are triangular not because the good citizens of Venice wanted triangles in their cathedral ceilings, but because intersecting arches willy-nilly produce triangles. Biologically speaking, the notion of spandrels is that there are dimensions to an adaptation that are initially fortuitous and an effect of orders not relevant to the problem solved by the adaptation. If we grant that reality has an infinite number of dimensions, all adaptations will be replete with spandrels. This, however, is in some sense trivial, in that by definition none of these features serve any function, and cannot be used to explain anything.

The question then becomes, What if one of these features acquire a functional significance? At that point, when the feature is first "co-opted for useful purposes" (¶7), it is a proper spandrel, a by-product rather than an adaptation. However, being functional, it will be subject to selective pressure, and any modification will make it an adaptation. In accounting for this adaptation, will it be important to remember it arose as a spandrel? Perhaps--but what is at stake in the debate is if something is an adaptation at all. The remaining category is spandrels that are functional but have not yet been modified by natural selection. Why would this be a common category?

The spandrels of San Marco, incidentally, clearly serve a vital function from the beginning: they constitute the ceiling of the structure.² While their precise shape is a contingent upon a specific solution to the problem of skeletal structure, other dimensions were crucial to the initial design: they are rigid, cohesive, of a certain structural strength, etc. And once the particular structural solution has been chosen, it is functional for the spandrels to fit precisely in between the arches and the wall. In fact, if the spandrels weren't able to do that, the whole structure would have to be demolished; the arches themselves must be constructed in such a way that the spaces in between can be filled in. Our muscles, to fill in the analogy, could have an infinite number of shapes and still perform their basic contractive function. The fact that they have the shapes they do could perhaps just as usefully be considered a 'fortuitous by-product' of the shape of our bones.

As for Gould's examples: the claim that the "mental machinery" for reading and writing "must have originated as spandrels that were co-opted later, for the brain reached its current size and conformation tens of thousands of years before any human invented reading or writing," fails some simple logical tests. Reading uses the mental machinery of the eyes, phonemic parsing (Shaywitz 1996), and memory; writing draws on fine-motor coordination in the hands. These all have the hallmarks of adaptive features: they are complex, unlikely, and functionally useful in domains that indeed predated reading by tens of thousands of years. Gould is making the elementary error of assuming the cultural application of an adaptation must necessarily coincide with its natural usage--a distinction captured in Dan Sperber's notion of proper and actual domains. While spandrels may provide the raw material for natural selection, logic suggests they soon turn into adaptations.

My position in this discussion--and I cannot speak for Dennett--is that Gould is correct in pointing out that there are several dimensions to evolution, and that an exclusive focus on gene-level adaptationism is myopic. However, Gould may be as guilty as Dennett of "sniping at false targets of his own construction" (¶11), and the arguments don't get him very far. In particular, they don't yet add up to a coherent critique of the project of evolutionary psychology. In the interest of hearing more of his views on this, let us skip forward to section 2.

The Quest of All Curious People

Gould's approach to evolutionary psychology is positive in theory; it is the practice he lambastes. However, so much passes for evolutionary psychology that it is at times hard to catch sight of his target.

The first victim is the theory of memes, espoused by Dennett. Are thoughts subject to Darwinian processes of variation, selection, and evolution? I personally find the analogy largely unhelpful, especially as a model of cultural change, and Gould's arguments strike me as valid ((¶19-20). It therefore comes as a bit of a surprise that he concludes his strictures against the memetic model of cultural change "undermines the self-proclaimed revolutionary pretensions of a much-publicized doctrine--'evolutionary psychology'" (¶21). Something is not adding up here; evolutionary psychology is not making claims about the Darwinian nature of cultural change.

Brushing past the generic charge of "a fatally restrictive view of the meaning and range of evolutionary explanation"--evolutionary psychology has fallen into the "ultra-Darwinian trap" (¶23)--we get the following list of arguments:

There is a lot of good evidence (from neurophysiology and cognitive science) that the brain has a modular design, but evolutionary psychologists tend to "atomize behavior into a priori, subjectively defined, and poorly separated items" (¶26). I suppose there's always some bad science; this level of critique needs to engage with the details in a way Gould doesn't do here.
The focus on human universals is a good thing too, except that it should avoid coming up with differences between men and women (¶27). It's unclear why it would be a good thing for men and women to be identical, and even more unclear what to do with the evidence they are not.
The change from the idea that all human behavior is adaptive to an acknowledgment that human proclivities are not necessarily adaptive today is a good thing, but chance and spandrels should be given a larger share of the action. On this count, I disagree with Gould; as discussed above, his arguments for the explanatory power of spandrels seems weak.

In sum, Gould sees positive features in the project of evolutionary psychology, on all the important counts: modularity, universals, adaptationism. His beefs are appear minor: the methodology should not be sloppy, the results should be more sensitive to his values, and the importance of chance should be acknowledged.

The degree of his anti-adaptationist stance, however, is more extreme than one might expect from this evaluation. In response to Robert Wright's suggestion that "our fondness for sweetness" is an adaptation to fruit-eating, and maladaptive in today's world of candy stores, Gould exclaims,

This ranks as pure guesswork in the cocktail party mode; Wright presents no neurological evidence of a brain module for sweetness, and no paleontological data about ancestral feeding. This "just-so story" therefore cannot stand as a "classic example of an adaptation" in any sense deserving the name of science. (¶29)

This critique is completely off target. Wright never argued there was a distinct "brain module for sweetness"--an instance of the kind of atomistic thinking Gould just upbraided evolutionary psychologists for. As for paleontological data about ancestral feeding, the fact that Wright does not present it doesn't mean there isn't any; this is after all a popular book written by a journalist, not a scholarly presentation of the discipline.

What about the implied contrast between "science" and "just-so stories"? Evolution, as Gould himself is well aware, is a historical theory, and a convenient format for presenting evidence of historical change is provided by the synthesized narratives of episodic memory. A credible story is a powerful way to present a scientific theory; what Gould's contrast implies is that the credibility comes cheap, at the expense of verified facts. However, this is hardly a principled argument against a representational format, a rationale for cognitive apartheid. For one, Gould's notion of spandrels is itself effectively couched in it. If I say our living relatives the primates share a proclivity for fruit, that the adaptive problem of harvesting the fruit at the peak of its nutritional value was solved in our common ancestor through a series of fortuitous but cumulative genetic mutations that led to a preference for certain chemicals, and that these mutations conferred fitness advantages, I have concocted a "just-so story," rich in predictive detail. It is then the job of working scientists to corroborate the many elements of such an explanation. A story that posited a sweet tooth as a by-product of other adaptations would necessarily be considerably more complex.

While Gould applauds evolutionary psychology's critique of the idea that all human behavior is adaptive in the present, he seems very reluctant to grant it may have been adaptive in the past. To this end, he points out that the EEA--the Environment of Evolutionary Adaptedness--cannot be known in detail, and thus any claim for adaptations "is untestable, and therefore unscientific" (¶32). The problem about this argument is that it demolishes too much of the Darwinian project: in fact we know next to nothing about the environments in which organisms evolved; how can we make claims natural selection gave rise to adaptations? A simple criterion is evidence of design--complex, unlikely, and functionally appropriate structures. These can be found both on the physiological and cognitive levels. Since we are lucky enough to have live human beings around us, living relatives, a decent collection of fossils, and converging evidence about our environment from several sources, we are on balance in a privileged position for understanding human adaptations.

I agree with Gould that this program can be pursued with the unjustifiable zeal of single vision. Wright's claim that "[i]f the theory of natural selection is correct, then essentially everything about the human mind should be intelligible in these terms" (¶32) is irrational. Not only can a theory be correct without explaining everything concerning a phenomenon, but no theory has ever explained "essentially everything"--and there is good reason to think none ever will (cf. the discussion of Knowledge as Perception). However, this journalistic presentation is not a suitable basis for indicting the validity of evolutionary psychology, though it is a valid critique of its popular dissemination. Pick your partner and you will win the game.

When Gould claims that "[t]his adaptationist premise is the fatal flaw of evolutionary psychology in its current form" (¶33), he is stacking the cards: the idea that everything can be explained as an adaptation is in no sense the premise of evolutionary psychology. What Gould does, however, is to use these irrational and absolutist claims to throw out the central importance of adaptationist thinking in the understanding of life--an argument that does not hold water.

Surprisingly expressing sympathy for the existence of "some different, and broadly general, emotional propensities of human males and females," Gould objects to the notion that "members of my gender are willing to rear babies only because clever females beguile us" (¶34). Indeed, this is not a claim of evolutionary psychology; a complex set of factors are likely to have affected the evolution of male propensities for parental investment, and Gould's intuitions serve him well. The game would have been more interesting if he had picked better partners; how would Gould like to have his work judged by presentations of it in the popular press?

All in all, Gould's warning about excessive reliance upon a narrowly conceived adaptationist model is well taken, but such a reliance is not characteristic of the discipline. By classifying his opponents as extremists and fundamentalists, Gould obscures the weakness in his own position. His arguments against the adaptationist project--beyond the injunction to pluralism--are weak, speculative, and ill founded. The examples he selects are from the popular literature, and not representative of the best work (cf bibliography). On the other hand, he is clearly open to the possibility of converging lines of evidence, and his skeptical position may form a useful touchstone for generating high-quality work.

Francis Steen
July 1997

In frame view, use the Back key on your browser to return
Full text of Gould's article | Back to CogWeb's Debate page

Notes

¹Darwin was aware of the uneven pace of change, noting that "the periods during which species have been undergoing modification, though very long as measured by years, have probably been short in comparison with the periods which these same species remained without undergoing any change" (On the Origin of Species, fourth edition (1866), pp. 359-360). (Thanks to Mark Leney for the reference.) (back)

²For an architectural demolition of Gould's argument, see Robert Mark (1996). Architecture and evolution. American Scientist 84: 383-389. (Thanks to Warren Searle for the reference.) (back)