Darwin's Dangerous Idea (52 page)

magnificent volumes that the mosaics are the
raison d'etre
of San Marco, and What would be a
good
architectural example of a spandrel
(sensu
Gould)?

hence of many of its architectural details. In other words, there wouldn't be If adaptations are examples of (good, cunning) design, then perhaps a any such pendentives in Venice if the "environmental problem" of how to spandrel is a "no-brainer"—a feature exhibiting no design cunning at all. The display Byzantine mosaic images of Christian iconography had not been existence of a doorway—just a rough opening—in a building might seem to posed and this solution found. If you look closely at the pendentives (this is be an example, since we would not be particularly impressed by the wisdom detectable in figure 10.1, but unmistakable if you look at the actual of the builder who included such a feature in his house. But there is, after all, pendentives, as I did on a recent visit to Venice ), you will see that care has a very good reason why dwellings should have doorways. If spandrels are been taken to round off the transition between the pendentive proper and the just dead-obvious good solutions to design problems that tend therefore to arches it connects, the better to provide a continuous surface for the become part of a relatively unthinking tradition of building, then spandrels application of mosaics.

abound. In that case, however, they would not be alternatives to adaptation, Gould and Lewontin's other example from architecture was also ill-chosen, as but examples
par excellence
of adaptation—either forced moves or, in any it turns out, since we simply don't know whether the King's College bosses event, moves you'd be foolish not to consider. A better sort of example, then, alternating rose and portcullis are the
raison d'etre
of the fan vaulting or vice might be what engineers sometimes call a "don't-care": something that has to be one way or another, but that nothing makes better one way than another. If versa. We do know that fan vaulting was
not
part of the original design of that we put a door in the doorway, the

chapel, but a later revision, a change order introduced years after the construction had begun, for reasons unknown ( Fitchen 1961, p. 248). The very heavy (and heavily carved) keystones at the intersections of the ribs of 2. I am not the first, I have recently discovered, to note these minor errors in Gould's earlier Gothic vaults had been a sort of forced move for builders, as I noted in excursion in art history. Some years ago, two evolutionary biologists were there before chapter 8, since they needed the extra weight of this keystone to counteract me: Alasdair Houston (1990) drew attention to the point about spandrels, pendentives, the rising tendency of the pointed arches, especially during the construction and squinches, and Tim Clutton-Brock, in a lecture at Harvard, questioned Gould's in-phase, when deformation of partially completed structures was a major terpretation of the fan vaulting of King's College Chapel.

problem to be solved. But in late fan vaulting of the King's College type, the It is interesting that these points were overlooked by all the deconstructionists and rhetoricians who contributed essays to a recent book (Selzer 1993) devoted in its purpose of the bosses is probably entirely to provide focal points for entirety to an analysis of the rhetoric of Gould and Lewontin's essay. You might suppose ornament. Did the bosses have to be there anyway? No. From an engineering that someone among this group of sixteen humanists would have noticed the factual point of view, there could have been neat round holes there, "lanterns" letting problems in the fundamental rhetorical device of the essay, but it must be remembered in daylight from above if it weren't for the roof. Maybe fan vaulting
was
that these sophisticates are interested in "deconstructing knowledge"—which means that they have transcended the stodgy, old-fashioned dichotomy between fact and fiction, and chosen by the builders so that the ceiling could carry the Tudor symbols! So hence are not professionally curious about whether what they read is the truth!

the fabled spandrels of San Marco are not spandrels but adaptations
The Spandrel's numb 111

276 BULLY FOR BRONTOSAURUS

Sometimes, however, it does seem that he thinks this is the view to attack.

door will need hinges, but should they go on the left or the right? Perhaps nobody cares, so a coin is flipped, and hinges on the left get installed. If other He characterizes adaptationism as "pure adaptationism" and "panadapta-builders copy the result unthinkingly, establishing a local tradition (reinforced tionism"—apparently the view that every feature of every organism is to be by the latchmakers, who make latches for left-hinged doors only), this might explained as an adaptation selected for. In her recent book,
The Ant and the
be a spandrel masquerading as an adaptation. "Why are all the doors in this
Peacock,
the philosopher of biology Helena Cronin is particularly acute in village hinged on the left?" would be a classic adaptationist question, to diagnosing this view (Cronin, pp. 66-110). She catches Gould in the act of which the answer would be: "No reason. Just historical accident." So is that a sliding into exactly this misconstrual;

good architectural example of a spandrel? Perhaps, but, as the example of the autumn leaves in the preceding chapter showed, it is never a mistake to
ask
Stephen Gould talks about 'what may be the most fundamental question the adaptationist's "why" question, even when the true answer is that there is in evolutionary theory' and then, significantly, spells out not one question no reason. Are there many features in the biosphere that exist for no reason?

but two: 'How
exclusive
is natural selection as an agent of evolutionary It all depends on what counts as a feature. Trivially, there are indefinitely change? Must
all
features of organisms be viewed as adaptations?' (Gould many properties (e.g., the elephant's property of having more legs than eyes, 1980[a], p. 49; my emphasis). But natural selection could be the only true the daisy's property of buoyancy ) that are not themselves adaptations, but no begetter of adaptations without having begot all characteristics; one can adaptationist would deny this. Presumably, there is a more interesting hold that all adaptive characteristics are the result of natural selection doctrine that Gould and Lewontin are urging us to abandon.

without holding that all characteristics are, indeed, adaptive. [Cronin 1991, What is the doctrine of "pervasive adaptation," then, that Gould supposes p. 86.]

such an admission of widespread spandrels would overthrow? Let us consider the most extreme form of Panglossian adaptationism imaginable—the view Natural selection could still be the "exclusive agent" of evolutionary that
every designed thing
is
optimally
designed. A sidelong glance at human change even though many features of organisms were not adaptations. Ad-engineering will show that even this view not only permits but requires the aptationists are—and should be—
always
on the lookout for adaptive ex-existence of plenty of undesigned stuff. Imagine, if you can, some masterpiece planations of whatever feature captures their attention, but this strategy falls of human engineering—the perfectly designed widget-factory, energy-short of committing anybody to the caricature that Gould calls "panadap-efficient, maximally productive, minimally expensive to operate, maximally tationism."

humane to its workers, simply unimprovable in any dimension. The waste-Perhaps what Gould opposes will become clearer if we look at what he paper collection system, for instance, makes recycling by type of wastepaper recommends in its place. What alternatives to adaptationism did Gould and maximally convenient and agreeable to the staff, at minimal energy costs, and Lewontin suggest, as components of their recommended pluralism? Chief so forth. A Panglossian triumph, it seems. But wait—what is the
wastepaper
among them was the idea of a
Bauplan,
a German architectural term that had for? It's not for anything. It's a by-product of the other processes, and the been adopted by certain continental biologists. The term would usually be wastepaper collection system is for dealing with it. You can't give an adap-translated in English as "ground plan" or "floor plan"—the basic outline of the tationist explanation of why the disposal/recycling system is optimal without structure as seen from above. It is curious that an architectural term should be presupposing that the wastepaper itself is just... waste! Of course, you can go highlighted in a counteradaptationist campaign, but it makes a certain daft on and ask whether the clerical operations could be made "paperless" by sense when you see how the original
Bauplan
theorists pushed it. Adaptation, better use of computers, but if that happens not to be the case for one reason they said, could explain
superficial
modifications of the design of organisms or another, there will still be wastepaper to deal with, and other wastes and to fit the environment, but not the fundamental features of living things: "The by-products as well in any case, so there will always be plenty of undesigned important steps in evolution, the construction of the
Bauplan
itself and the features in a system that is maximally well designed. No adaptationist could transition between
Baupldne,
must involve some other unknown, and perhaps be such a "pervasive" adaptationist as to deny it. The thesis that every prop-

'internal' mechanism" (Gould and Lewontin 1979, p. 159). The floor plan is erty of every feature of everything in the living world is an adaptation is not a not designed by evolution, but just somehow given? Sounds a bit fishy, thesis anybody has ever taken seriously, or implied by what anybody has doesn't it? Were Gould and Lewontin buying this radical idea from the taken seriously, so far as I know. If I am wrong, there are some serious loonies continent? Not for a moment. They quickly (p. 159) granted that English out there, but Gould has never shown us one.

biologists had been right "in rejecting this strong form as close to an appeal to mysticism."

278 BULLY FOR BRONTOSAURUS

The Spandrel's Thumb
279

But once the mystical version of
Baupläne
is shunned, what is left? Our Moreover, Gould, in spite of the appeal to pluralism in the co-authored old friend: the claim that good reverse engineering takes the building process paper, has persisted in describing it as laying waste to adaptationism (e.g., into account. As Gould and Lewontin put it (p. 160), their view of matters 1993a), and has held out for a "non-Darwinian" interpretation of its central

"does not deny that change, when it occurs, may be mediated by natural concept, spandrels. It may have occurred to you that I have overlooked an selection, but it holds that constraints restrict possible paths and modes of obvious interpretation of spandrels: perhaps spandrels are just QWERTY

change so strongly that the constraints themselves become much the most phenomena. QWERTY phenomena, you recall, are constraints, but con-interesting aspect of evolution." Whether or not they are the
most
interesting straints with an adaptive history and hence an adaptationist explanation.3

aspect, they are certainly important, as we have seen. Perhaps adaptationists Gould himself briefly considered this alternative (1982a, p. 383): "If the (like art historians) need to have this point repeatedly drawn to their attention.

channels [that constrain current options] are set by past adaptations, then When Dawkins, an arch-adaptationist if there ever was one, says, "There are selection remains preeminent, for all major structures are either expressions some shapes that certain kinds of embryology seem incapable of growing"

of immediate selection, or channeled by a phylogenetic heritage of previous (Dawkins 1989b, p. 216), he is expressing a version of this point about the selection." Nicely put, but he promptly rejected it, calling it Darwinian constraint of the
Bauplan,
and it was something of a revelation to him, he (which it certainly is), and recommending an alternative "non-Darwinian says. It was forcefully brought home to him by his own computer simulations version" which he described as "not widely appreciated but potentially of evolution, not by the Gould and Lewontin paper, but we might let them fundamental." Spandrels, he then suggested (p. 383), aren't the frozen chime in: "We told you so!"

constraints created by earlier adaptations; they are
exaptations.
What contrast Gould and Lewontin also discuss other alternatives to adaptation, and was he trying to draw?

these, too, are themes we have already encountered in orthodox Darwinism: I think he saw the difference between the exploitation of something random fixation of genes (the role of historical accident and its ampli-previously designed, and the exploitation of something originally unde-fication), developmental constraints due to the way genes get expressed, and signed, and was claiming that it was an important difference. Perhaps. Here the problems of getting around in a fitness landscape with "multiple adaptive is some indirect textual evidence for that reading. A recent article in the peaks." These are all real phenomena; as usual, the debate among
Boston Globe
quotes the linguist Samuel Jay Keyser of MIT: evolutionists is not about whether they exist, but about how important they are. Theories that incorporate them have indeed played a significant role

"Language may well be a spandrel of the mind," Keyser says, and then waits within the growing sophistication of the neo-Darwinian synthesis, but they patiently while his questioner looks "spandrel" up in the dictionary— The first are reforms or complications, not revolutions.

builder who supported domes with arches created spandrels
by accident
So some evolutionists have accepted Gould and Lewontin's pluralism in an

[emphasis added), and at first builders paid no attention to spandrels and irenic spirit, as a call not to abandon but, rather, to improve adaptation-ism.

decorated only the arches, Keyser says. But after a couple of centuries, builders As Maynard Smith (1991, p. 6) has put it, "The effect of the Gould-Lewontin began focussing on and decorating the spandrels. In the paper has been considerable, and on the whole welcome. I doubt if many people have stopped trying to tell adaptive stories. Certainly I have not done so myself." Gould and Lewontin's paper has had a welcome effect, then, but one of its by-products has not been so welcome. The inflammatory rhetoric 3. In his own discussion of the original QWERTY phenomenon (1991a), Gould makes a suggesting that these somewhat neglected themes constituted a major useful point ( 1991a, p. 71), but does not develop it further, so far as I know: because of alternative to adaptationism opened the floodgates to a lot of wishful thinking the curious historical sequence of events that led to the general adoption of the standard QWERTY typewriter keyboard, "An array of competitions that would have tested by Darwin-dreaders who would prefer that there
not
be an adap-tationist QWERTY were never held." That is, it is simply
irrelevant
to ask whether QWERTY is a explanation of one precious phenomenon or another. What would their dimly better design than alternatives
X, Y,
and
Z,
since those alternatives were never pitted imagined alternative be? Either the "internal necessity" that Gould and against QWERTY in the marketplace or the design workshop. They just never came up at Lewontin themselves dismiss as an appeal to mysticism, or utter cosmic a time when, it seems, they could have made a difference. Adaptationists should be alert coincidence—an equally mystical nonstarter. Neither Gould nor Lewontin to the fact that, even though whatever we see in nature has been "tested against all comers" and not found wanting, only a Vanishingly small (and biased) subset of all the explicitly endorsed either wild alternative to adaptation, but this was imaginable competitions has ever been held. The inevitable parochiality of all actual overlooked by those who wanted to be dazzled by the authority of these tournaments means that one must be cautious in characterizing the virtues of the win-eminent Darwin-doubters.