Darwin's Dangerous Idea (18 page)

the jokes we hear and pass on have evolved from earlier stories, picking up To bring out just how unspecial Mitochondrial Eve—that is, Amy—prob-revisions and updates as they are passed along. A joke typically has no one ably was, suppose that tomorrow, thousands of generations later, a virulent author; its authorship is distributed over dozens or hundreds or thousands of new virus were to spread around the Earth, wiping out 99 percent of the tellers, solidifying for a while in some particularly topical and currently human race in a few years. The survivors, fortunate to have some innate amusing version, before going dormant, like the ancestors from which it resistance to the virus, would probably all be quite closely related.
Their
grew. Speciation is equally hard to witness, and for the same reason.

closest common direct female ancestor—call her Betty—would be some When has speciation occurred? In many cases (perhaps most, perhaps woman who lived hundreds or thousands of generations later than Amy, and almost all—biologists disagree about how important the exceptions are), the the crown of Mitochondrial Eve would pass to her, retroactively. She may speciation depends on a geographical split in which a small group— maybe a have been the source of the mutation that centuries later came into its own as single mating pair—wander off and start a lineage that becomes a species-saver, but it didn't do
her
any good, since the virus against which
it
reproductively isolated. This is
allopatric
speciation, in contrast to
sym-is
to
triumph didn't exist then. The point is that Mitochondrial Eve can only
patric
speciation, which does not involve any geographic barriers. Suppose be
retrospectively
crowned. This historically pivotal role is determined not we watch the departure and resettlement of the founding group. Time passes, just by the accidents of Amy's own time, but by the accidents of later times and several generations come and go. Has speciation occurred? Not yet, as well. Talk about massive contingency! If Amy's uncle hadn't saved her certainly. We won't know until many generations later whether or not these from drowning when she was three, none of
us
(with our particular individuals should be crowned as species-initiators.

mitochondrial DNA, thanks ultimately to Amy) would ever have There is not
and could not be
anything internal or intrinsic to the individuals—or even to the individuals-as-they-fit-into-their-environment—from 6. Philosophers have often discussed strange examples of individuals known to us only via definite descriptions, but they have usually coniined their attention to such boring—if 7. There are, of course, the writers who make their living writing funny lines for televi-real—individuals as the shortest spy. (There has to be one, doesn't there?) I suggest that sion comedians, and the comedians themselves, who create much of their own material, Mitochondrial Eve is a much more delicious example, all the more so for being of some but, with negligible exceptions, these people are not the creators of the joke stories genuine theoretical interest in evolutionary biology.

("Did you hear the one about the guy who...?") that get passed around.

100 THE TREE OF LIFE

Patterns, Oversimplification, and Explanation
101

which it followed that they were—as they later turn out to be—the founders trends, forces, principles—or historical events—have influenced these of a new species. We can imagine, if we want, an extreme (and improbable) shapes or made them possible? Eyes have evolved independently in dozens case in which a single mutation guarantees reproductive isolation in a single of lineages, but feathers probably only once. As John Maynard Smith ob-generation, but, of course, whether or not the individual who has that serves, mammals go in for horns but birds do not. "Why should the pattern mutation counts as a species-founder or simply as a freak of nature depends of variation be limited in this way? The short answer is that we do not know"

on nothing in its individual makeup or biography, but on what happens to (Maynard Smith 1986, p. 41).

subsequent generations—if any—of its offspring.

We
can't
rewind the tope of life and replay it to see what happens next Darwin was not able to present a single instance of speciation by natural time, alas, so the only way to answer questions about such huge and ex-selection in
Origin of Species.
His strategy in that book was to develop in perimentally inaccessible patterns is to leap boldly into the void with the detail the evidence that artificial selection by dog- and pigeon-breeders could risky tactic of deliberate oversimplification. This tactic has a long and dis-build up large differences by a series of gradual changes. He then pointed out tinguished history in science, but it tends to provoke controversy, since that
deliberate
choice by title animals' keepers was inessential; the runts of scientists have different thresholds at which they get nervous about playing the litter tended not to be valued, and hence tended not to reproduce as much fast and loose with the recalcitrant details. Newtonian physics was overas their more valued siblings, so, without any conscious policy of breeding, thrown by Einstein, but it is still a good approximation for almost all pur-human animal-keepers presided unwittingly over a steady process of design poses. No physicist objects when NASA uses Newtonian physics to calculate revision. He offered the nice example of the King Charles spaniel, "which the forces at liftoff and the orbital trajectory of the space shuttle, but, strictly has been unconsciously modified
to
a large extent since the time of that speaking, this is a deliberate use of a false theory in order to make calculation monarch"
(Origin,
p. 35)—as can be confirmed by a careful examination of feasible. In the same spirit, physiologists studying, say, mechanisms for the dogs in various portraits of King Charles. He called such cases changing the rate of metabolism try in general to avoid the bizarre com-

"unconscious selection" by human domesticators, and he used it as a plexities of subatomic quantum physics, hoping that any quantum effects will persuasive bridge to get his readers to the hypothesis of even more cancel out or in other ways be beneath the threshold of their models. In unconscious selection by the impersonal environment. But he had to admit, general, the tactic pays off handsomely, but one can never be sure when one when challenged, that he could provide no cases of animal-breeders'

scientist's grubby complication will be elevated into another scientist's Key to producing a new species. Such breeding had definitely produced different the Mystery. And it can just as well work the other way around: the Key is
varieties,
but not a single new species. Dachshund and St. Bernard were not often discovered by climbing out of the trenches and going for the panoramic different species, however different in appearance. Darwin admitted as much, view.

but he might quite correctly have gone on to point out that it was simply too I once got in a debate with Francis Crick about the virtues and vices of early to tell whether he had given any examples of speciation accomplished Connectionism—the movement in cognitive science that models psycho-by artificial selection. Any lady's lapdog could at some future date be logical phenomena by building up patterns in the connection-strengths discovered
to have been
the founding member of a species that split off from between the nodes in
very
unrealistic and oversimplified "neural nets" sim-Canis familiaris.

ulated on computers. "These people may be good engineers," Crick averred The same moral applies to the creation of new genera, families, and even (as best I recall), "but what they are doing is terrible science! These people kingdoms, of course. The major branching that we would retrospectively willfully turn their backs on what we
already
know about how neurons crown as the parting of the plants from the animals began as a segregation of interact, so their models are utterly useless as models of brain function." This two gene pools every bit as inscrutable and unremarkable at the time as any criticism somewhat surprised me, for Crick is famous for his own brilliant other temporary drifting apart of members of a single population.

opportunism in uncovering the structure of DNA; while others struggled up the straight and narrow path of strict construction from the evidence, he and Watson took a few daring and optimistic sidesteps, with gratifying results.

4. PATTERNS, OVERSIMPLIFICATION, AND EXPLANATION

But in any case, I was curious to know how widely he would cast his denunciation. Would he say the same thing about population geneticists? The Much more interesting than the question of how to draw the species bound-derogatory term for some of their models is "bean-bag genetics," for they ary are all the questions about the shapes of the branches—and even more pretend that genes for this and that are like so many color-coded beads on a interesting, the shapes of the empty spaces between the branches. What string. What they call a gene (or an
allele
at a
locus)
102 THE TREE OF LIFE

Patterns, Oversimplification, and Explanation
103

bears only a passing resemblance to the intricate machinery of the codon even have necks. So that's how come giraffes have long necks. End of ex-sequences on DNA molecules. But thanks to these deliberate simplifications, planation. (And if that doesn't satisfy you, note that you will be even less their models are computationally tractable, enabling them to discover and satisfied if the answer throws in all the details about the individual devel-confirm many large-scale patterns in gene flow that would otherwise be opmental and nutritional history of each giraffe in the lineage.) utterly invisible. Adding complications would tend to bring their research to Any acceptable explanation of the patterns we observe in the Tree of Life a grinding halt. But is their research good science? Crick replied that he had must be contrastive: why do we see this actual pattern rattier than that one—

himself thought about the comparison, and had to say that population or no pattern at all? What are the nonactualized alternatives that need to be genetics wasn't science either!

considered, and how are they organized? To answer such questions, we need My tastes in science are more indulgent, as perhaps you would expect to be able to talk about what is possible in addition to what is actual.

from a philosopher, but I do have my reasons: I think the case is strong that not only do "over"-simplified models often actually
explain
just what needs explaining, but no more complicated model could do the job. When what CHAPTER 4: There are
patterns in the unimaginably detailed Tree of Life,
provokes our curiosity are the
large patterns
in phenomena, we need an
highlighting crucial events that made the later flourishing of the Tree pos-explanation at the right level. In many instances this is obvious. If you want
sible. The eukaryotic revolution and the multicellular revolution are the most
to know why traffic jams tend to happen at a certain hour every day, you will
important, followed by the speciation events, invisible at the time, but later
still be baffled after you have painstakingly reconstructed the steering, brak-seen to mark even such major divisions as those between plants and animals.

ing, and accelerating processes of the thousands of drivers whose various
If science is to explain the patterns discernible in all this complexity, it must
trajectories have summed to create those traffic jams.

rise above the microscopic view to other levels, taking on idealizations when
Or imagine tracing all the electrons through a hand calculator as it mul-necessary so we can see the woods for die trees.

tiplies two numbers together and gets the correct answer. You could be 100

percent sure you understood each of the millions of causal microsteps in the CHAPTER 5:
The contrast between the actual and the possible is fundamental
process and yet still be utterly baffled about
why
or even
how
it always got
to all explanation in biology. It seems we need to distinguish different grades
the
right
answer to the questions you posed it. If this is not obvious, imagine
of possibility, and Darwin provides a framework for a unified treatment of
that somebody made—as a sort of expensive prank—a hand calculator that
biological possibility in terms of accessibility in "the Library of Mendel," the
usually gave the wrong answers! It would obey exactly the same physical
space of all genomes. In order to construct this useful idealization, we must
laws as the good calculator, and would cycle through the same sorts of
acknowledge and then set aside certain complications in the relations
microprocesses. You could have
perfect
explanations of how both calculators
between a genome and a viable organism.

worked at the electronic level, and still be utterly unable to explain the intensely interesting fact that one of them got the answers right and the other got them wrong. This is the sort of case that shows what would be silly about the preposterous forms of reductionism;
of course
you can't explain all the patterns that interest us at the level of physics (or chemistry, or any one low level). This is undeniably true of such mundane and unperplexing phenomena as traffic jams and pocket calculators; we should expect it to be true of biological phenomena as well. (For more on this topic, see Dennett 1991b.)

Now consider a parallel question in biology, a textbook standard: why do giraffes have long necks? There is one answer that could in principle be

"read off" the total Tree of Life, if we had it to look at: Each giraffe has a neck of the length it has because its parents had necks of the lengths they had, and so forth back through the generations. If you check them off one by one, you will see that the long neck of each living giraffe has been traced back through long-necked ancestors all the way back... to ancestors who didn't
Grades of Possibility?
105