Darwin's Dangerous Idea (58 page)

we have no idea how to identify or define them), but I suspect that [the Lady Luck hypothesis] grasps a central truth about evolution. The Burgess Really? Consider a parallel. There you sit, on a rock in Wyoming, watch-Shale, in making this ... interpretation intelligible by the hypothetical ing a hole in the ground. Nothing much happens for ten, twenty, thirty experiment of the tape, promotes a radical view of evolutionary pathways minutes, and then, suddenly—whoosh!—a stream of boiling water shoots and predictability.

more than thirty meters into the air. It's all over in a few seconds, and then nothing much happens—just like before, apparently—and you wait for an Gould's suggestion, then, is not that he can prove the Lady Luck hypothesis, hour, and still nothing much happens. This, then, was your experience: a but that the Burgess Shale makes it at least intelligible. As Darwin insisted single, amazing explosion lasting but a few seconds out of an hour and a half from the beginning, however, all it takes is "some groups" with an "edge" to of tedium. Perhaps you would be tempted to think, "Surely this must be a put the wedge of competition into action. So is Gould just saying that
most
of unique and unrepeatable event!"

the competition (or the competition with the largest, most important effects) So why do they call it Old Faithful? In fact, this geyser repeats itself once was a true lottery? That is what he "suspects."

every sixty-five minutes, on average, year in and year out. The "shape" of the What is his evidence for this suspicion? He offers none at all. What he Cambrian Explosion—its "sudden" onset and equally "sudden" termination—

offers is the fact that he, looking at these amazing creatures, can't imagine is no evidence
at all
for the thesis of "radical contingency." But Gould seems why some would be better designed than others. They all seem about equally to think that it is.11 He seems to think that, if we replayed the tape of life, we bizarre and ungainly to him. That is not good evidence that they didn't in fact couldn't get another "Cambrian" Explosion the next time, or ever. But differ dramatically in engineering quality, given their respective although that might be true, he has not yet offered us a single bit of evidence.

predicaments. If you don't even try to engage in reverse engineering, you are Where might such evidence come from? It might come from the computer not in a good position to conclude that there
is no
reverse-engineering simulations of Artificial Life, for instance, which do permit us to rewind the explanation to be discovered. He does offer a wager (p. 188): "I challenge any tape again and again. It is surprising that Gould has overlooked the paleontologist to argue that he could have gone back to the Burgess seas and, possibility that he might find some evidence for (or against) his main without the benefit of hindsight, picked out
Naroia, Can-adaspis, Aysheaia,
and
Sanctaris
for success, while identifying
Marrella, Odaraia, Sidneyia,
and
Leonchoilia
as ripe for the grim reaper." That's a pretty safe sucker bet, since 11. Gould says (1989a, p. 230), in response to Conway Morris's objections: "The Cam-all such a paleontologist would have to go on is the outlines of organs visible brian explosion was too big, too different, and too exclusive." See also the remarks on the in fossil traces. But it could be lost. Some really ingenious reverse engineer unpredictability of "zigzag" trajectories ( Gould 1989b).

might someday be able to tell an awfully

Tinker to Evers to Chance
307

306 BULLY FOR BRONTOSAURUS

We have reached second base. Just what
is
Gould's claim about contin-conclusion by looking at the field of Artificial Life, but he never mentions the gency? He says (1990, p. 3) that "the most common misunderstanding of prospect. Why not? I don't know, but I do know Gould is not fond of evolution, at least in lay culture," is the idea that "our eventual appearance" is computers, and to this day does not even use a computer for word-processing;

"somehow intrinsically inevitable and predictable within the confines of the that might have something to do with it.

theory."
Our
appearance? What does that mean? There is a sliding scale on A much more important clue, surely, is the fact that when you do rerun the which Gould neglects to locate his claim about rewinding the tape. If by "us"

tape of life, you find all sorts of evidence of repetition. We already knew that, he meant something very particular—Steve Gould and Dan Dennett, let's of course, because convergent evolution is nature's own way of replaying the say—then we wouldn't need the hypothesis of mass extinction to persuade us tape. As Maynard Smith says:

how lucky
we
are to be alive; if our two moms had never met our respective dads, that would suffice to consign us both to Neverland, and of course the In Gould's "replay from the Cambrian" experiment, I would predict that same counterfactual holds true of every human being alive today. Had such a many animals would evolve eyes, because eyes have in fact evolved many sad misfortune befallen us, this would not mean, however, that our respective times, in many kinds of animal. I would bet that some would evolve pow-offices at Harvard and Tufts would be unoccupied. It would be astonishing if ered flight, because flight has evolved four times, in two different phyla; but the Harvard occupant's name in this counterfactual circumstance was I would not be certain, because animals might never get out on the land.

"Gould," and I wouldn't bet that its occupant would be a habitue of bowling But I agree with Gould that one could not predict which phyla would alleys and Fenway Park, but I
would
bet that its occupant would know a lot survive and inherit the earth. [Maynard Smith 1992, p. 34]

about paleontology, would give lectures and publish articles and spend thousands of hours studying fauna ( not flora—Gould's office is in the Maynard Smith's last point is a sly one: if convergent evolution reigns, it Museum of Comparative Zoology ). If, at the other extreme, by "us" Gould doesn't make any difference
which phyla
inherit the earth, because of bait-meant something very general, such as "air-breathing, land-inhabiting verte-and-switch! Combining bait-and-switch with convergent evolution, we get brates," he would probably be wrong, for the reasons Maynard Smith men-the orthodox conclusion that
whichever
lineage happens to survive will tions. So we may well suppose he meant something intermediate, such as gravitate towards the Good Moves in Design Space, and the result will be

"intelligent, language-using, technology-inventing, culture-creating beings."

hard to tell from the winner that would have been there if some different This is an interesting hypothesis. If it is true, then contrary to what many lineage had carried on. Consider the kiwi, for instance. It has evolved in New thinkers routinely suppose, the search for extra-terrestrial intelligence is as Zealand, where it didn't have any mammals to compete with, and it has quixotic as the search for extra-terrestrial kangaroos—it happened once, here, converged on an amazing number of mammalian features—basically, it's a but would probably never happen again. But
Wonderful Life
offers no bird that pretends it's a mammal. Gould himself has written about the kiwi evidence in its favor (Wright 1990 ); even if the decimations of the Burgess and its remarkably large egg (in 1991b), but as Conway Morris points out in Shale fauna were random, whatever lineages happened to survive would, achis review (1991, p. 6):

cording to standard neo-Darwinian theory, proceed to grope towards the Good Tricks in Design Space.

... there is something else about the kiwi that receives only passing men-We have answered our second question. We are finally ready to tackle first tion, and that is the extraordinary convergence between kiwis and mam-base: why would this thesis be of great importance, whichever way it came mals. ... I am sure Gould would be the last to deny convergence, but surely it undermines much of his thesis of contingency.

out? Gould thinks that the hypothesis of "radical contingency" will upset our equanimity, but why?

Gould does not deny convergence—how could he?—but he does tend to We talk about the "march from monad to man" (old-style language again) ignore it. Why? Perhaps because, as Conway Morris says, it is the fatal as though evolution followed continuous pathways of progress along weakness in his case for contingency. (See also Maynard Smith 1992, Daw-unbroken lineages. Nothing could be furtiier from reality. [Gould 1989b, kins 1990, Bickerton 1993)

P-14.]

So now we have an answer to our third question. The Burgess Shale fauna inspire Gould because he mistakenly thinks that they provide evidence for his
What
could not be further from reality? At first it might appear as if Gould thesis of "radical contingency." They
might
illustrate the thesis—but we was saying here that there is no continuous, unbroken lineage between the won't know until we do the sort of research that Gould himself has ignored.

308 BULLY FOR BRONTOSAURUS

Tinker to Evers to Chance
309

"monads" and us, but surely there is. There is no more secure implication of to consider, and rule out, alternative hypotheses. As I said at the outset, I am Darwin's great idea than that. As I put it in chapter 8, it is not controversial more interested in the reasons that have held the myth together than I am in that we are all direct descendants of macros—or monads—simple precellular the actual motives of the actual man, but it might seem disingenuous for me replicators under one name or another. So what can Gould be saying here?

not even to mention the obvious "rival" explanations crying to be considered: Perhaps we are meant to put the emphasis on "pathways
of progress"
—it is politics and religion. (It could well be that there is a political or religious the belief in progress that is so far from the truth. The pathways are motivation behind the yearning for skyhooks I impute to him, but those continuous, unbroken lineages all right, but not lineages of global progress.

would not be rival hypotheses; they would be elaborations of my This is true, but so what?

interpretation, postponable to another occasion. Here I must briefly consider There aren't
global
pathways of progress, but there is incessant
local
whether one of these—politics or religion—might offer a simpler, more improvement. This improvement seeks out the best designs with such great straightforward interpretation of his campaigns, obviating my analysis. Many reliability that it can often be predicted by adaptationist reasoning. Replay the of Gould's critics have thought so; I think they are missing the more tape a thousand times, and the Good Tricks will be found again and again, by interesting possibility.)

one lineage or another. Convergent evolution is not evidence of global Gould has never made a secret of his politics. He learned his Marxism from progress, but it is overwhelmingly good evidence of the power of processes his father, he tells us, and until quite recently he was very vocal and active in of natural selection. This is the power of the underlying algorithms, mindless left-wing politics. Many of his campaigns against specific scientists and all the way down, but, thanks to the cranes it has built along the way, specific schools of thought within science have been conducted in explicitly wonderfully capable of discovery, recognition, and wise decision. There is no political—indeed, explicitly Marxist—terms, and have often had right-wing room, and no need, for skyhooks.

thinkers as their targets. Not surprisingly, his opponents and critics have often Can it be that Gould thinks his thesis of radical contingency would refute supposed, for instance, that his punctuationism was just his Marxist antipathy the core Darwinian idea that evolution is an algorithmic process? That is my for reform playing itself out in biology. Reformers are the worst enemies of tentative conclusion. Algorithms, in the popular imagination, are algorithms revolutionaries, as we all know. But that, I think, is only a superficially
for
producing a particular result. As I said in chapter 2, evolution can be an plausible reading of Gould's reasons. After all, John Maynard Smith, his polar algorithm, and evolution can have produced us by an algorithmic process, opposite in the evolution controversies, has a Marxist background as rich and without its being true that evolution is an algorithm
for
producing us. But if active as Gould's, and there are others with left-wing sympathies against whom you didn't understand that point, you might think:

Gould has directed attacks. (And then there are all the ACLU liberals like myself, though I doubt if he knows or cares.) Following his return from a visit
If
we are not the predictable result of evolution, evolution cannot be an to Russia, Gould (1992b) drew attention, as often before, to the difference algorithmic process.

between the gradualness of reform and the suddenness of revolution. In this interesting piece, Gould (p. 14) reflects on his experiences in Russia, and the And then you would be strongly motivated to prove "radical contingency" if failure of Marxism there— 'Yes, the Russian reality does discredit a specific you wanted to show that evolution wasn't just an algorithmic process. It Marxist economics"—but goes on to say that Marx has been proven right about might not have recognizable skyhooks in it, but at least we'd know it wasn't

"the validity of the larger model of punctuational change." That does not mean all done with nothing but cranes.

that, for Gould, Marx's economic and social theory was never the point, but it Is it likely that Gould could be so confused about the nature of algorithms?

is not hard to believe that Gould would keep his attitudes about evolution on As we shall see in chapter 15, Roger Penrose, one of the world's most board while jettisoning some political baggage that had outlasted its welcome.

distinguished mathematicians, wrote a major book ( 1989) on Turing As for religion, my own interpretation is, in one important sense, a hy-machines, algorithms, and the impossibility of Artificial Intelligence, and his pothesis about Gould's religious yearnings. I see his antipathy to Darwin's whole book is based on that confusion. This is not really such an implausible dangerous idea as fundamentally a desire to protect or restore the Mind-first, error, on either thinker's part. A person who really doesn't like Darwin's top-down vision of John Locke—at the very least to secure
our
place in the dangerous idea often finds it hard to get the idea in focus.