chance variation and evolutionary contingency: darwin...
TRANSCRIPT
Chance Variation and Evolutionary Contingency:
Darwin, Simpson (The Simpsons) and Gould
John BeattyDepartment of Philosophy
University of British ColumbiaVancouver, BC V6T 1Z1, Canada
1. Introduction
The unpredictability of evolution by natural selection was championed most famously by
Stephen Gould in terms of the following thought experiment:
I call this experiment “replaying life’s tape.” You press the rewind button and, making
sure you thoroughly erase everything that actually happened, go back to any time and
place in the past. . . . Then let the tape run again and see if the repetition looks at all like
the original. (Gould 1989, p. 48)
His expectation was that “any replay of the tape would lead evolution down a pathway radically
different from the road actually taken” (p. 51).
The basic idea—sans the hyperbole (and the high-tech metaphor)—goes back to Darwin.
But in reasserting it, Gould was hardly stating the obvious. It had fallen out of favor with
evolutionary biologists on more than one occasion. In this chapter, I will discuss the origins and
fate of the idea, which is closely tied to considerations of the meaning and significance of chance
variation.
2
2. Darwin on “The Details Left to Chance”
We associate chance variation with Darwin. But the notion appears earlier, for example
in William Paley’s Natural Theology (1809). Which is somewhat surprising, because throughout
the book Paley argued that living things and most everything about them are the products of
design, not chance.
What does chance ever do for us? In the human body, for instance, chance, i.e. the
operation of causes without design, may produce a wen, a wart, a mole, a pimple, but
never an eye. [Never was] an organized body of any kind, answering a valuable purpose
by a complicated mechanism, the effect of chance. In no assignable instance hath such a
thing existed without intention somewhere. (Paley, 1809, pp. 62-63)
Toward the end of the book, however, Paley acknowledged that God had left some important
things to chance, and for (supposedly) good reason. Take for example the variety of occupational
abilities that we see in every society, and that every society relies on: some people are cut out to
be laborers, some are cut out to be land or factory owners, etc. Such differences are associated
with differences in fortune, and therefore it is only fair that they be distributed by lot. So God
designed the world with this sort of chance variation built-in (p. 520).
Darwin, an admirer of Paley, also figured that variation was part of God’s plan. Toward
the beginning of his “Transmutation of Species” notebooks, Darwin privately wondered why
organisms reproduce and die rather than living forever. The clue to the puzzle, he decided, was
variation. The “final cause” or purpose of reproduction and death, together with the “tendency to
vary,” was to allow species to adapt to changing environmental circumstances (see, e.g., Darwin
3
1987 [1837, Notebook B, pp. 1-5, 64], pp. 170-171, 187; Kohn 1980; Ospovat 1981, pp. 39-86;
Hodge and Kohn 1985). The natural theological roots of Darwin’s evolutionary theorizing, and
his understanding of variation in particular, are important for reasons that I will address shortly.
At first Darwin assumed that environmental changes cause adaptively appropriate
variations to arise in the offspring of affected organisms. But he soon came to the conclusion that
most variations arise by “accident” or “chance,” i.e., by causes that are 1) complex and unknown
and 2) in no way related to what would be useful for surviving and reproducing. As Darwin
expressed the first point, “No doubt each slight variation must have its efficient cause; but it is as
hopeless an attempt to discover the cause of each, as to say why a chill or a poison affects one
man differently from another” (1875, vol. 2, p. 282). The causes of variation were, to Darwin, so
complex that even God—as omniscient as He is—might not foresee, or at least might not take
note of the variations that would arise in a particular species in a particular environment (I will
return to this point).
Concerning the absence of any relationship between what variations arise and what
variations would be useful, Darwin expressed himself by way of analogy:
[Evolution by natural selection] absolutely depends on what we in our ignorance call
spontaneous or accidental variation. Let an architect be compelled to build an edifice with
uncut stones, fallen from a precipice. The shape of each fragment may be called
accidental; yet the shape of each has been determined by the force of gravity, the nature
of the rock, and the slope of the precipice,—events and circumstances, all of which
depend on natural laws; but there is no relation between these laws and the purpose for
which each fragment is used by the builder. In the same manner the variations of each
creature are determined by fixed and immutable laws; but these bear no relation to the
4
living structure which is slowly built up through the power of selection, whether this be
natural or artificial selection. (1875, vol. 2, p. 236)
Many of Darwin’s readers could not see past the role he attributed to chance variation.
Some accused him of explaining organic form in terms of chance alone. For example, Karl von
Baer complained that the Origin reminded him of Gulliver’s Travels, especially Gulliver’s
account of the Laputans and their amazing discovery machine. Laputan scholars had inscribed all
the forms of all their words on the sides of dice, which were then placed in a mechanical grid. By
cranking the handles of the device, they could spin each die independently, and generate
different word combinations. They recorded the sensible word phrases and then cranked the
handles again. Etc., etc. And this is how they planned to advance knowledge. It all sounded, to
von Baer, vaguely Darwinian: chance begetting usefulness.
For a long time the author of these reports was taken to be joking, because it is self-
evident that nothing useful and significant could ever result from chance events. . . . [But]
now we must acknowledge this philosopher as a deep thinker because he foresaw the
present triumphs of science! (von Baer [1873] 1973, p. 419)
Of course, in construing Darwinian evolution as a matter of chance variation alone, von
Baer left out natural selection. In this regard, he did not make sufficient use of his own analogy.
The Laputans relied on chance only to generate word combinations; then they carefully selected
the phrases to be passed down to future generations. Similarly, in the case of Darwinian
evolution, variations appear by chance, but it is not entirely a matter of chance which variations
are perpetuated; it is a matter of natural selection.
Nonetheless, even correcting for von Baer’s misconstrual, it is still the case that evolution
by natural selection is a matter of chance. It may not be a matter of chance that the fitter
5
variations are selected and that the outcomes of evolution by natural selection are adaptive. But it
is a matter of chance which variations arise, and in this sense also a matter of chance which
variations will be selected and hence which adaptive outcomes will obtain.
In principle, evolution by natural selection could result in very different outcomes, even
starting with closely related and in all important respects identical species, inhabiting identical
environments, depending on what variations happen to arise in each lineage, and in what order.
Darwin set out to demonstrate the application of this principle in his book, On the Various
Contrivances by which British and Foreign Orchids are Fertilised by Insects (1862). There he
argued that, as diverse as orchid flowers are, they all serve basically the same function, namely
to enlist flying insects in cross-pollination, and thus to avoid inbreeding. These otherwise very
different “contrivances” for intercrossing had evolved, Darwin believed, under virtually the same
environmental circumstances, e.g., the same range of available insects. Sometimes one part of
the flower had been modified to entice insects in the vicinity, by mimicry or by scent; sometimes
another part had been modified to do the same job. Once the insects had arrived, the pollen had
to be attached. Some flowers were so constructed as to catapult pollen at the visiting insects;
some catapult the insects against the pollen; some simply induce the visitors to travel past and
brush-up against the pollen. Etc., etc. Thus cross-pollination is accomplished in very different
ways, the different outcomes being due in large measure—Darwin argued—to natural selection
acting on chance differences in variation among different lineages.
Darwin might have explained the different outcomes mainly in terms of differences in
environmental conditions—e.g., differences in the insects available for conscription as
pollinators. Thus, he might have presumed that different lineages of orchids experience the same
variations, in the same order, and then he might have argued that different variations are selected
6
in different lineages depending on which pollinators are in the vicinity. But he did not. His
emphasis was on chance differences in variation among lineages.
An example that particularly struck Darwin involved the position of the so-called
“labellum” petal, which in most fully formed orchid flowers is the lowermost of the three petals.
In that position, it often serves as a landing pad for pollinators. But interestingly, the labellum
actually arrives at that position through a 180-degree twisting of the flower’s stem (usually
including the ovarium) as the flower develops. Darwin reckoned that the position of the labellum
in the ancestral orchid had been uppermost, presumably on the grounds that this is also the
original position in development, and assuming more generally that the order of development
reflects the order of ancestry. He understood the now-typical, lowermost position of the labellum
to be an outcome of evolution by natural selection of the more twisted variations that had
happened, by chance, to arise (1877, p. 284).
But Darwin was especially intrigued by cases where the labellum had resumed its
uppermost position, which in some cases had resulted from the selection of less and less twisted
variations, and in other cases had come about as the result of selection for more and more twisted
forms. Flowers of the latter sort twist a full 360 degrees to resume their starting position (see Fig.
1)!
7
Figure 1. The twisted stalk of Malaxis paludosa (from Darwin 1877, p. 130). See the
text.
As Darwin described the situation,
. . . in many Orchids the ovarium (but sometimes the foot-stalk) becomes for a period
twisted, causing the labellum to assume the position of a lower petal, so that insects can
easily visit the flower; but . . . it might be advantageous to the plant that the labellum
should resume its normal position on the upper side of the flower, as is actually the case
with Malaxis paludosa, and some species of Catasetum, &c. This change, it is obvious,
might be simply effected by the continued selection of varieties which had their ovaria
less and less twisted, but if the plant [for whatever reason, by chance] only afforded
variations with the ovarium more twisted, the same end could be attained by the selection
of such variations, until the flower was turned completely on its axis. This seems to have
8
actually occurred with Malaxis paludosa, for the labellum has acquired its present upward
position by the ovarium being twisted twice as much as is usual. (pp. 284-285)
So it had apparently become advantageous for Malaxis and some species of Catasetum to have
their labellae uppermost. But due to differences in the variations that chanced to occur in the two
lineages, evolution by natural selection had resulted in very different means of serving this end: a
360-degree twist in the first case, and no twist in the second (see also Lennox 1993).
Darwin invoked basically the same process—evolution by natural selection of chance
differences in variation—to account for the endless diversity of floral morphologies among
orchids. To quote him, summarizing the process leading to the diversity of orchid flowers:
In my examination of Orchids, hardly any fact has struck me so much as the
endless diversities of structure,—the prodigality of resources,—for gaining the very same
end, namely, the fertilization of one flower by pollen from another plant. This fact is to a
large extent intelligible on the principle of natural selection. As all the parts of a flower
are co-ordinated, if slight variations in any one part were preserved from being beneficial
to the plant, then the other parts would generally have to be modified in some
corresponding manner. But these latter parts might not vary at all [who knows, it is a
matter of chance], or they might not vary in a fitting manner [again, who knows], and
these other variations, whatever their nature might be [ditto], which tended to bring all
the parts into more harmonious action with one another, would be preserved by natural
selection. (p. 284; my emphasis)
Darwin came to believe that much of the diversity of life could be understood in this way:
as different means to the same ends, reached by natural selection of whatever variations happen
to arise. He hammered the point home in later editions of the Origin:
9
[I]t is a common rule throughout nature that the same end should be gained, even
sometimes in the case of closely-related beings, by the most diversified means. (1872, p.
153)
[T]his subject of the same end being gained by the most diversified means well deserves
attention. (p. 154)
How, it may be asked . . . can we understand the graduated scale of complexity and the
multifarious means for gaining the same end? (p. 156)
[T]he common rule throughout nature is infinite diversity of structure for gaining the
same end. (p. 165)
The sometimes amusing (e.g., in the case of Malaxis paludosa), and at other times
disturbing contingency of evolutionary outcomes had, for Darwin, profound theological
implications that he first discussed privately in correspondence with Asa Gray, and then publicly
as well, all in response to Gray’s review of the Origin. Gray had intended to defend Darin
against charges that the theory of evolution by natural selection was atheistic. Just because
Darwin had remained silent concerning God’s role in evolution, Gray argued, that is no reason to
think that Darwin denied God a role. As for what that role might be, Gray suggested that it would
surely be guiding variation “along certain beneficial lines,” like a stream “along definite and
useful lines of irrigation” (1860, pp. 413-414). In other words, God either directly causes
particular variations to appear at particular times in order to be perpetuated by natural selection,
or else indirectly provides for their occurrence.
Indeed, Darwin had not intended to write God out of creation. God had, he initially
believed, designed the sorts of laws—including laws of variation, heredity, population growth,
etc.—that would lead to the production of well adapted species. But Darwin balked at Gray’s
10
suggestion that God also arranged the sequence of variations that were subsequently selected. To
what end would God have done so? Presumably to guarantee the existence of humans (“in His
own image,” Genesis 1:27). But then did He also directly or indirectly cause just the right
variations to occur at just the right times to ensure the evolution of every other species that has
ever existed? Did He, for example, arrange the sequence of variations that led to the existence of
the ichneumonids, parasitic wasps that lay their eggs in living caterpillars so that the larvae can
devour their hosts from the inside out? Given the complexity of variation, God would have had
to go to an awful lot of trouble to produce such mean beasts. Better to think that He did not
attend to that! Thus, Darwin counter-suggested that the living world is the result of “designed
laws, with the details, whether good or bad, left to the working out of what we may call chance.”
But if God did not set things up in such a way as to get ichneumonids, then why suppose that he
arranged for humans? Getting out of one theological jam got Darwin into another, but he seems
to have preferred the latter conundrum to the former. As he wrote to Gray,
With respect to the theological view of the question; this is always painful to
me.—I am bewildered.—I had no intention to write atheistically. But I own that I cannot
see, as plainly as others do, & as I shd wish to do, evidence of design & beneficence on
all sides of us. There seems to me too much misery in the world. I cannot persuade
myself that a beneficent & omnipotent God would have designedly created the
Ichneumonidae with the express intention of their feeding within the living bodies of
caterpillars, or that a cat should play with mice. Not believing this, I see no necessity in
the belief that the eye was expressly designed. On the other hand I cannot anyhow be
contented to view this wonderful universe & especially the nature of man, & to conclude
that everything is the result of brute force. I am inclined to look at everything as resulting
11
from designed laws, with the details, whether good or bad, left to the working out of what
we may call chance. Not that this notion at all satisfies me. (Darwin to Gray, 22 May
1860, in Darwin 1993, p. 224)
He later chose to elaborate the point—about God leaving “the details” to chance—in the
closing pages of Variation in Plants and Animals under Domestication. He began with the
analogy about the architect that constructs a building out of stones fallen from a nearby cliff.
Does anyone really think that God, foreseeing the architect’s situation, would have somehow
ordained the coming-into-being of stones in just the right sizes and shapes, and at just the right
time and place, for the builder’s convenience? If so, then one might also suppose that God,
foreseeing the particular predilections of future breeders, ordained the existence of variations that
they would artificially select. But now we have a theological problem even worse than the willful
provision for ichneumonids—namely, the existence of pit bulls! As Darwin continued his
reductio ad absurdum line of questioning, “Did He [God] cause the frame and mental qualities of
the dog to vary in order that a breed might be formed of indomitable ferocity, with jaws fitted to
pin down the bull for man's brutal sport” (Darwin, 1875, vol. 2, pp. 430-431)? (Pit bulls bring
down their much larger opponents by clamping their jaws over the bull’s snout and not letting go
until the bull suffocates.) This is partly (largely) God’s handiwork?!
The alternative is to suppose that God leaves variation to chance, which raises other
theological problems, especially concerning humans, but at least avoids a malicious conception
of the creator.
But if we give up the principle in one case [i.e., if we deny that there is any divine
ordering of variation in the domesticated world],—no shadow of reason can be assigned
for the belief that variations, alike in nature and the result of the same general laws,
12
which have been the groundwork through natural selection of the formation of the most
perfectly adapted animals in the world, man included, were intentionally and specially
guided. (1875, vol 2, pp. 431-432)
3. Simpson on “Evolutionary Opportunism”
However, Darwin’s position on chance variation and evolutionary contingency did not
win the day. Gray’s position—directed variation with predetermined evolutionary
outcomes—became more and more popular in the late nineteenth and early twentieth centuries,
at least in a somewhat more secular form known as “directed evolution” or “orthogenesis.” This
view was especially popular among paleontologists; as Vernon Kellogg reported in his state-of-
the-art, Darwinism To-Day (1907):
Paleontologists believe, practically as a united body, that variation has followed fixed
lines through the ages; that there has been no such unrestricted and utterly free play of
variational vagary as the Darwinian natural selection theory presupposes. (p. 33)
Proponents of this view believed that directed variation was the only way to explain (what they
perceived to be) the morphological directionality of evolution, together with the (supposed) fact
that while the morphological changes were ultimately adaptive, the requisite initial stages were
not (e.g., Osborn 1925, p. 750; Osborn 1926). One example worked out in detail by Henry
Fairfield Osborn concerned the horns of the titanotheres (an extinct family of an order whose
only three extant familes are the horses/zebras/asses, rhinoceroses, and tapirs). In lineage after
lineage of titanothere, the same morphological progression was preserved in the fossil record:
from no horns, to minimal protuberances from the snout, to somewhat larger protuberances, to
massive horns. The larger horns, Osborn reasoned, were increasingly useful for fighting, but the
13
initial bumps were quite useless. Such a directional evolutionary change could not, therefore, be
the result of natural selection. It must be due instead to some direction in the process of variation
itself, which guaranteed the same evolutionary outcome time after time (Osborn 1929).
Osborn was generally silent on the process by which variation is directed; he felt that it
might possibly be “beyond human solution” (1926, p, 341), which suggests that he may, like
Gray, have attributed it to Providence (who else, besides humans, did he think might know the
answer?). The most determined critic of orthogenesis, George Simpson, ridiculed it for its
mysteriousness:
Most theories of this school . . . involve an element of predestination, of a goal, a
perfecting principle, whether as a vitalistic urge, or a metaphysical necessity, or a frankly
theological explanation of evolution according to which it is under divine or otherwise
spiritual guidance. (1944, p. 152).
He acknowledged that prominent proponents of directed evolution, like Osborn, did not
explicitly invoke supernatural causes (1964, p. 200), but to Simpson, the reasons for its
popularity were evident from the frankly theological versions of it defended by others like Pierre
Lecomte de Noüy, Teilhard de Chardin, and Edmund Sinnott.
Simpson was not only concerned about the mysteriousness of orthogenesis, and the
possibly religious motives of its proponents, however. His critiques of directed evolution were
more pointed than that. For each purported case of directional evolution, he either denied the
phenomenon, or offered an alternative explanation that did not rely on any unknown causal
processes (Simpson 1944, pp. 149-179; 1949, pp. 130-159). For example, in the case of the
titanotheres, he argued that even the most rudimentary of horns would have been better than
nothing for fighting, so that the directionality in question might simply be the result of evolution
14
by natural selection of chance variations, leading to larger and larger horns in each lineage
(1949, pp. 154-157).
Mostly, though, he denied the supposed directionality of evolution. Closely related
lineages do not follow the same trajectories to the same endpoints, he argued. “History does not
repeat itself” in this way; it is not so predictable. His reasoning was similar to Darwin’s (e.g., in
the case of orchids), though somewhat more elaborate. There were, according to Simpson, two
main reasons that, together, explain why evolution is unpredictable. The first is that evolution is
driven principally by natural selection of chance variations in changing environments, and is
therefore—as Simpson put it—“opportunistic” (Simpson 1949, pp. 160-186). Consider that
environmental changes and variational changes both present opportunities to adapt. Natural
selection exploits these changes by perpetuating the most favorable of the existing or new
variants in the existing or new environment. This is “opportunistic” in the sense that there is no
plan or provision by which organisms automatically vary in ways that are adaptive (the OED
definition of “opportunism” is “the practice or policy of exploiting circumstances or
opportunities to gain immediate advantage, rather than following a predetermined plan”).
The second reason why evolution is unpredictable is that, just as there are multiple means
to any end, there are “multiple solutions” to any environmental “problem,” or multiple ways of
exploiting any environmental opportunity.
Thus, closely related and very similar (or even initially identical) lineages, inhabiting a
similar or even an identical series of environments, may nonetheless generate different
variations. And natural selection, opportunistically perpetuating whichever of those variations
are most advantageous, may lead to different outcomes representing multiple solutions to the
15
environments faced. The outcomes are only as predictable as the variation on which natural
selection acts.
[T]he results of mutations do not tend to correspond at all closely with the needs or
opportunities of the mutating organisms. It is a rather astonishing observation that the
supply of this basic material for evolution seems to have no particular relationship to the
demand. This accounts for much of the opportunism in evolution, and the nature of that
opportunism in turn attests the random nature of mutation.
Evolution works on the materials at hand: the groups of organisms as they exist at
any given time and the mutations that happen to arise in them. The materials are the
results of earlier adaptations plus random additions and the orienting factor in change is
adaptation to new opportunities. If this view of evolution is correct, then we must expect
to find similar opportunities exploited in different ways. The problems involved in
performing certain functions should have multiple solutions. (1949, pp. 164-165)
16
Figure 2. Antelopes of the (then) Belgian Congo (from Simpson 1949, p. 166). See the
text.
Simpson illustrated the point with the existence of a wide variety of horns among
antelopes in the Belgian Congo (see Fig. 2). Should these be seen as adaptations to different
circumstances—e.g., different predators—within that region, or as alternative adaptations to the
same or very similar circumstances? Don’t the doubly curved horns of the impala (11) and kob
(15) serve basically the same function as the singly curved horns of the two reedbuck species (12
and 14)? Aren’t these are just two different means to the same end of having forward-pointing
horns (much as a 360-degree twist and no twist at all are different means to the same end of
17
having the labellum petal on top)? In which case the difference is due to chance differences in
the order in which variations appeared and were subsequently selected. Similarly, must we
imagine that the antelopes with forward-pointing horns face a different range of threats than
those with backward-pointing horns? Or are these both effective ways of dealing with the same
or similar range of predators? In which case, again, the difference is due to chance differences in
the order in which variations appeared and were subsequently selected.
As Simpson summarized this case, and by extension many others,
If evolution were really operating according to a fixed plan, surely these radical
discrepancies would not arise. Nor would they arise if evolution were basically
orthogenetic or if a rigid orientation were everywhere the rule. It looks as if there had
been an orienting factor, indeed—all these animals do have horns that serve them well
enough even if not perfectly—but as if what it had to orient was not uniform or
completely responsive. As horns developed in the different lines different sorts of
mutations affected them. As long as the mutations increased the development of a
functional horn, however various in other respects, they served the adaptive end and were
retained and promoted in the evolution of each line. The mutations that would have been
mechanically the best, that an engineer would have chosen, simply did not happen to
occur in all, or perhaps in any, of these lines. There are two aspects of opportunism: to
seize such diverse opportunities as occur, and when a single opportunity or need occurs,
to meet it with what is available, even if this is not the best possible. The antelopes, and
many other groups of animals, well illustrate this second sort of evolutionary
opportunism. (1949, pp. 167-168)
18
So it is not the case that evolution is directed toward particular adaptive outcomes by
some preordained or otherwise built-in order of variation. But Simpson went further with this
line of reasoning, critiquing not only directed evolution, but also every attempt to generalize
about evolutionary trajectories, other than to say that they are adaptive. The opportunism of
natural selection, together with the possibility of multiple solutions to any environmental
problem, render it highly unlikely that evolution will ever repeat itself. So we should not expect
to find any generalizations, and definitely not any “laws” about the course of evolution, again
with the exception that evolution is generally adaptive ([1950] 1964, pp. 176-189). There is just
one other exception to the rule that there are no such rules: a rule that, like the general
adaptiveness of evolutionary outcomes, makes sense in light of the opportunism of natural
selection and the possibility of multiple solutions. This is the rule that “evolution is irreversible.”
Simpson understood this to be just a special case of the principle that evolutionary history does
not repeat itself. In other words, it is just as unlikely that evolution would exactly reverse itself
given the same series of environmental circumstances, except in reverse, as that evolution would
exactly repeat itself under the same series of environments (Simpson [1950], 1964, pp. 186-187).
Simpson’s concerns about directed evolution and “laws” of evolution, however, were not
entirely biological. It was fairly common at the time to argue not only that evolution is
directional, but that the direction must be good. And not only should we not interfere to divert
evolution from its natural course, but we should even intervene to hasten it along his proper
trajectory. This line of reasoning was often used to argue for the inevitability and desirability of a
highly collectivist organization of society. Simpson (1941) singled-out the physiologist Ralph
Gerard, who had argued that evolution proceeded in the direction of greater and greater
“integration” on all levels, from the organs of an organism, to the organisms that compose a
19
species. This involved decreased totipotency and self-sufficiency, and increased specialization
and reliance on the other parts of the whole. And this was all for the good. To be sure, it involved
a loss of freedom. But Gerard argued that his fellow Americans had perhaps oversold the
importance of autonomy and freedom:
Men born into a caste, apprenticed into a guild, trained in a craft are proud to do their
jobs well and do not feel forlorn at the impossibility of becoming king. The relative
fluidity of our own early culture and the “freedom” that American youth has to “rise to
the top” has certainly produced little contentment—European visitors and local
psychiatrists are impressed with the restless striving which is almost a national psychosis.
No, it is possible for men to be part of a highly integrated society and yet feel, as
individuals, more free, actually to have more avenues open for satisfying self-expression,
than when they are epiorganisms of their own, like single-celled organisms. Which of us
would exchange our present state for the privilege of roaming the woods naked and
unarmed, without language or fire? (Gerard, 1940, pp. 411-412)
Simpson could also have referred to Conrad Waddington, who had argued that one could
and should look to see if there is an overall direction to evolution, and then do what one can to
promote that outcome. And not surprisingly for the time, the alternatives that most interested him
were whether evolution tends toward greater autonomy among species members, or toward a
more highly organized form of interaction. Having discovered which way evolution (including
human evolution) is headed, we should then act on that knowledge.
Once we have decided from evolutionary data whether the continuation of man's progress
demands a high degree of respect for the individual rather than his thorough subjection to
a group organization, for instance, we can work out the implications of that directive in
20
present political terms. Only when we have found our Ten Commandments in general
evolution, can we discover our Deuteronomy in political analysis. (Waddington 1942, p.
125)
That is as openly as Waddington ever posed the question. Elsewhere, he expressed his
view that evolution leads to greater control over environmental circumstances, which in turn
favors more highly integrated social relations, and in humans to more collectivist political
economies. He concluded that “totalitarianism . . . does seem to be inevitable,” and that “the
application of conscious control to the economic functioning of society is actually a step along
the path of man's evolutionary advance” (1948, pp. 22, 171). He also wrote in carefully chosen
Marxist terms of the importance of “assisting” the “birth pangs” of the new economic order.
In explaining his concerns about this use of evolutionary thinking, Simpson switched
from talking about directed evolution and laws of evolution, to “evolutionary fatalism.”
Biological justification for the totalitarian development of society has also been
sought in the doctrines of evolutionary fatalism. Regardless of such labels as “right,”
“wrong,” “good,” or “bad,” it is argued, this is the inevitable future. Mankind is going
this way just as horsekind was going toward Equus throughout the Tertiary. Opposition is
as futile and foolish as if the little Eohippus had said, “I am going to be a dinosaur,”
instead of, “—a horse.”
Even aside from the fact that this is another false use of analogy, it has been
shown that a fatalistic view of evolution has little scientific support. (Simpson 1941, p.
20)
21
Simpson complained that biology, in the form of evolutionary fatalism, had become more
dangerous than physics—with all its firepower—because of the way it had been “used to provide
the more insidious and still more menacing moral implementation of totalitarianism.”
If this use is wrong, scientifically, and if free biologists support it or even tacitly permit
it, then they will deserve an accusation stronger than any that can be brought against
physical science, and they will be contributing to their own destruction. (1941, p. 15)
Simpson proceeded to apply his own expertise in issues of macroevolution and long-term
evolutionary trends to undermine the view that evolution is directional and fatalistic. Proponents
of such a deterministic conception of evolution, he argued, overlooked the extent to which “the
products of evolution are the results of a sequence of accidents” (1941, p. 8). Evolution may
sometimes lead to the extreme integration of individual organisms, so that they become more like
parts of a whole, as for example in the case of colonial organisms like corals. But, he argued, this
is rare.
In considering Simpson’s worries, it is also worth recalling Karl Popper’s initial concerns
about “Darwinism,” which were based mainly on historically deterministic (mis)renderings of
Darwinian evolutionary thought, and connections drawn to Marxist and fascist theories of social
and racial history. When Popper’s first extended discussion of Darwinism appeared in book
form, as Poverty of Historicism (1957), it was dedicated to “the countless men and women of all
creeds or nations or races who fell victim to the fascist and communist belief in Inexorable Laws
of Historical Destiny.”
Simpson argued that, in light of the contingency of evolutionary outcomes, it was up to
us to take control of our own fates:
22
Man has risen, not fallen. He can choose to develop his capacities as the highest
animal and to try to rise still farther, or he can choose otherwise. The choice is his
responsibility, and his alone. There is no automatism that will carry him upward without
choice or effort and there is no trend solely in the right direction. Evolution has no
purpose; man must supply this for himself. (ME, 310)
4. Gould on “Replaying Life’s Tape”
Gould read Simpson’s Meaning of Evolution at the age of ten, and later in life could not
recall what impression it had on him as a boy, though he had in the interim come to regard it as
“the finest introduction to deeper issues of life's meaning, insofar as biology can approach these
questions” (Gould 1985, p. 230). He was also deeply influenced by Simpson’s collection of
essays, This View of Life, the title of which was derived from the last passage of the
Origin—where Darwin reflected that “There is grandeur in this view of life . . .”—and which in
turn served as the title of Gould’s own long-running column in Natural History. Either book
could have started Gould thinking about what would happen if evolutionary history could be
rerun. In Meaning, Simpson elaborated his views of evolutionary opportunism. In This View,
Simpson had defended the indeterminism of evolution by natural selection, and had posed the
question, what would become of the Cambrian trilobites if they could be put back in place, and
the environmental conditions of the Cambrian and subsequent Ordovician periods could be
repeated (Simpson, [1950] 1964, p. 187)? But alas, evolution is not repeatable.
Gould would rewind the videotape of life back to the Cambrian, and then replay it for his
generation. It got people thinking, in a way. In one of the most popular episodes of The Simpsons
(no. 606, 1994), Homer tried to fix his toaster and accidentally turned it into a time machine. It
23
transported him back to the age of the dinosaurs, over and over, with a different evolutionary
outcome each time. Homer rejected each world in turn (the first one he rejected was,
interestingly, a totalitarian world in which Ned Flanders was the dictator and molded everyone
into his personality—very scary for Homer; equally scary was a world in which everything was
better than Homer could imagine, except that no one had ever heard of donuts), before finally
settling on a world in which everyone had reptilian tongues and could get their food from plate to
mouth without the need of a fork. (Gould later appeared as himself in a segment—no. 908,
1997—in which he was consulted about a fossilized, winged protohuman—allegedly an angel;
you have to see it!).
Gould never, to my knowledge, discussed the replay scenario without also discussing the
Darwin-Gray correspondence and Darwin’s views on ichneumonids and “the details left to
chance” (see, e.g., 1989, pp. 290-291; 1990, p. 21; 1995; 1999, pp. 173-207). He liked to
confront readers who might find the contingency of evolutionary outcomes—especially the
contingency of human existence—“depressing.” He always countered that he found the thought
“exhilarating,” for reasons similar to those that Simpson raised:
Homo sapiens, I fear, is a “thing so small” in a vast universe, a wildly improbable
evolutionary event well within the realm of contingency. Make of such a conclusion what
you will. Some will find the prospect depressing; I have always regarded it as
exhilarating, and a source of both freedom and consequent moral responsibility. (1989, p.
291)
And elaborating on the last point:
We are the offspring of history, and must establish our own paths in this most diverse and
interesting of conceivable universes—one indifferent to our suffering, and therefore
24
offering us maximal freedom to thrive, or to fail, in our own chosen way. (1989, p. 323;
see also Gould 1999, pp. 191-207)
Of course, Gould did not promote evolutionary contingency for the sole reason that he
found it morally satisfying, nor defend it solely by arguing that it was not as morally repugnant
as one might think. He felt that it needed to be promoted/defended on other grounds. While
Simpson had championed evolutionary contingency in response to proponents of directed
variation, Gould aimed his case at proponents of the “all importance” of natural selection (see
also Sterelny and Griffiths 1999, pp. 296-302). Such “pan-selectionists” assume, at least as a
working hypothesis, that the traits that predominate in a population are optimal for the
environment inhabited by the population, and hence if different traits prevail in different
populations, this must reflect environmental differences. I am not aware of any pan-selectionist
arguments explicitly to the effect that chance variation is irrelevant to evolutionary outcomes.
But it is as if chance variation can safely be ignored—the optimal traits will always arise.
Evolution is predictable on the basis of selective considerations alone.
It was during the late forties, fifties, and sixties that natural selection came to be regarded
as all-important. Gould referred to this development as the “hardening of the evolutionary
synthesis,” meaning that earlier evolutionary biologists had been more pluralistic with regard to
the agents and factors of evolution (e.g., Gould 1983). The increasing importance attributed to
natural selection had to do with the reinterpretation of some phenomena that had previously been
thought to be inexplicable in selectionist terms. It is important to note that these phenomena had
been accounted for in terms of “random drift” in particular, rather than natural selection.
Proponents of the importance of random drift had argued that many or most traits do not confer
any selective advantage or disadvantage, and hence do not increase or decrease in frequency as a
25
result of selection for or against them. Rather, their frequencies fluctuate by “chance,” not in the
sense of chance origin of variation but rather in the sense of chance perpetuation of
variation—e.g., when a parent has two different genes for a trait, but just happens to pass down
the same one to all of its offspring, or one organism is struck by lightning while another is spared
and lives to reproduce, etc. This position was quite popular in the nineteen thirties and forties. It
seemed to be the only way to account for many otherwise senseless differences—e.g., that
different blood groups prevail in different populations of humans, different shell colors and
patterns prevail in different populations of snails, and different chromosomal inversions prevail
in different populations of fruitflies. There was no conceivable selective basis for such
differences. But as time went on, the traits that had been singled-out as paradigm cases of
selective neutrality—whose frequencies were supposedly entirely a matter of chance—were one-
after-another shown to confer selective advantages or disadvantages in particular environments,
and to have frequencies that reflected the degree of selection for or against them in those
environments. It was especially in the context of this reversal of fortune that natural selection
came to be regarded, by Gould’s elders and many of his contemporaries, as the paramount, and
even all-important agent of evolution (Beatty 1984).
Somehow in the course of these debates, random drift came to be seen as the, rather than
a source of the unpredictability of evolutionary outcomes, and evolution by natural selection
came to be seen, in contrast, as eminently predictable. But to identify evolution by natural
selection with evolutionary predictability was to miss the point that Darwin and Simpson had
made earlier. As long as natural selection acts on chance variation, and as long as there are
multiple possible solutions to any adaptive problem, then the outcomes of evolution by natural
selection are going to depend on which variations happen to arise. Gould and Richard Lewontin
26
felt the need to remind enthusiasts of natural selection that when different traits prevail in
different populations or species, this need not reflect differences in their environments. Evolution
by natural selection can lead to different outcomes even from the same starting point and in the
same environments:
When “multiple adaptive peaks” are occupied, we usually have no basis for asserting that
one solution is better than another. The solution followed in any spot is a result of
history; the first steps went in one direction, though others would have led to adequate
prosperity as well. Every naturalist has his favourite illustration. In the West Indian land
snail Cerion, for example, populations living on rocky and windy coasts almost always
develop white, thick and relatively squat shells for conventional adaptive reasons. We can
identify at least two different developmental pathways to whiteness from the mottling of
early whorls in all Cerion, two paths to thickened shells and three styles of allometry
leading to squat shells. All 12 combinations can be identified in Bahamian populations,
but would it be fruitful to ask why—in the sense of optimal design rather than historical
contingency—Cerion from eastern Long Island evolved one solution, and Cerion from
Acklins Island another? (Gould and Lewontin 1979, p. 593)
As we have seen, the unpredictability of evolutionary outcomes has been stressed from
time to time ever since Darwin. But Gould’s particular version of the thesis struck a chord and
was followed-up by a flurry articles purporting to demonstrate or refute his position. These
included macroevolutionary “natural experiments;” laboratory-controlled, microevolutionary
experiments; and artificial-life simulation-experiments (see Beatty 2006, n. 4, p. 337). The
results have been mixed. For example, Jonathan Losos et al. argued that the same four
ecomorphs (adaptive types) of lizard had evolved independently at least four times on islands of
27
the Caribbean, a seeming devastating blow to Gould’s view that “any replay would lead
evolution down a pathway radically different from the road actually taken.” While Michael
Travisano et al. (1995a) and Travisano et al. (1995b) cultivated twelve genetically identical
clones of E. coli in identical environments for two thousand generations, and found that the
twelve evolutionary outcomes were similar in terms of their fitness, but were quite different
genetically.
Gould intended for the replay metaphor to cover numerous sources of evolutionary
contingency, not just chance variation. But his initial articulation of the metaphor coincided with
and reinforced a burst of interest in the evolutionary significance of “mutational order,” as
opposed to random drift. For example, Mani and Clarke (1990) emphasized that random drift
was not the only source of evolutionary stochasticity:
It is common to consider only a single stochastic process in evolution, random
genetic drift, yet there are at least three others that have often been neglected because of a
failure to recognize their distinctive properties, and a tendency to confuse them with each
other and with drift itself.
The three processes are:
1. Random fluctuations in selective values.
2. The random consequences of gene conversion and other mechanisms
‘homogenizing’ the DNA.
3. The random order of mutations. (1990, p. 29)
It was the latter that they focused on. In order to distinguish the effects of mutational
order from those of random drift, they simulated evolution in two groups of identical
populations, inhabiting identical environments. In the first group, mutations were introduced in
28
the same order—i.e., the same mutations arose in each population at the same time. Any
divergence between these populations must be due to random drift alone. In the second group of
populations, mutations were introduced randomly. In both groups, there was selection, and the
same selection regimes held. The investigators measured the evolutionary divergence within
each group in terms of genetic distance. Under every scenario that they studied, the mean genetic
distance of the random-mutation group was considerably greater than that of the ordered-
mutation group. And not too surprisingly, the difference was greater, the larger the population
size (see Fig. 3). This is partly because the larger the population size, the less important are the
sorts of fortuities associated with random drift.
Figure 3. Mean genetic distance (y-axis) vs. 10-2 x number of generations (x-axis) for
populations in which mutations are ordered (continuous line) vs. random (dotted line);
population sizes (a) 5000, (b) 2000, (c) 500, (d) 100 (from Mani and Clarke 1990, p. 32)
29
Others pointed out that large population size in combination with high mutation rates
would increase the chance of coincident mutations, and hence reduce the impact of mutational
order. In introducing their discussion of this issue, Johnson et al. (1995; see also Wahl and
Krakauer 2000) invoked Gould’s replay metaphor, and the body of literature that had recently
addressed the role of chance variation in the unpredictability of evolutionary outcomes:
How reproducible is evolution, and in particular adaptive evolution? Gould
(1989) proposed a thought experiment of ‘replaying life’s tape’ to address this
question. . . . Rigorous experiments on the predictability of evolution using organisms
with short generations, such as fruit flies and bacteria, demonstrate that: (i) the random
origin of mutations, either alone or in concert with genetic drift, can cause
unpredictability; and (ii) substantial divergence of (initially identical) replicate
populations in identical environments can occur even for traits that are subject to strong
selection. (p. 125)
The importance of mutational order vs. random drift was also demonstrated in a
continuation of the study previously mentioned, in which twelve identical (cloned) populations
of E. coli were grown in identical media for two thousand generations, and had similar fitnesses
at the end. Lenski and Travisano (1994) extended the experiment from two thousand to ten
thousand generations. Somewhat surprisingly, the fitness values of the twelve populations were
still only similar after ten thousand generations. Why after so long in exactly the same medium
had not all of the populations evolved the same ability to compete for resources within that
environment? Lenski and Travisano argued that the residual differences in fitness reflected
underlying genetic differences, which had in turn resulted largely from the different order of
mutations occurring in the twelve populations:
30
Evolutionary biologists usually regard diversification as being caused by either (i)
adaptation to different environments, which often produces conspicuous phenotypic
variation, or (ii) random genetic drift, which is usually seen in molecular genetic
variation. Yet our experiments demonstrate diversification, in identical environments and
with very large populations, of no less selected a trait than fitness itself. Someone
confronted with the variability among our derived populations (and unaware of the
experimental design) might attribute this diversity to environmental heterogeneity or
phylogenetic constraints , but any such “just-so story” would clearly be misguided in this
case. Instead, or experiment demonstrates the crucial role of chance events (historical
accidents) in adaptive evolution. (Lenski and Travisano 1994, p. 6813)
Demonstrating the actual vs. the theoretical importance of evolutionary contingency (due
to whatever source of unpredictability) is difficult for a number of reasons discussed by Kim
Sterelny and Paul Griffiths (1999, pp. 301-302) and by Sterelny (2005). Take for instance
“convergence,” which is generally regarded as the strongest form of evidence against the thesis
that evolution is intrinsically contingent (and which is often cited in critiques of Gould—e.g., by
Dennett 1995, pp. 305-308; Conway Morris 2003). Convergence occurs when distantly related
lineages face the same or similar adaptive problems, and reach the same or similar solutions.
However, inasmuch as lineages can be more or less related, convergence is a matter of degree,
and hence there is no straightforward way to tally cases of convergence.
An even more important problem, I think, arises in spades in the case of the twelve
cloned populations inhabiting identical environments and evolving toward similar overall
fitnesses by genetically diverse means, has to do with what Sterelny (2005) calls the
“courseness” of description of the trait being studied. What bears more on Gould’s thesis? The
31
fact that the twelve populations were, after however many generations, similar with regard to a
course-grained trait like fitness? Or that they had diverged with regard to the more fine-grained,
underlying genetics? Is the case of the twelve cloned populations a victory for evolutionary
contingency, or evolutionary predictability? Again, it is very difficult to count evidence for and
against.
And then there is the usual problem involved in adjudicating relative significance
controversies in biology—namely, extrapolating to overall frequencies when the number of cases
studied is so small. What does one study, or one hundred, or even one thousand tell us about the
overall importance of evolutionary contingency?
By way of conclusion, I would like to return briefly to the moral message that Gould
drew from the contingency of evolutionary outcomes: that it is “a source of both freedom and
consequent moral responsibility”—that it offers us “maximal freedom to thrive, or to fail, in our
own chosen way.” Given that Gould also argued strongly against attempts to derive ethical
norms from nature it may seem, as Daniel Dennett has suggested, that Gould was “trying to deny
to others what he allow[ed] himself” (1995, p. 311). It does look that way. But there is a different
way to make sense of Gould’s point, and in order to see that, it is important first of all to take his
intended audiences into account.
Dennett remarks that “Gould . . . seems to think that the view he is combating so
vigorously is deterministic and ahistorical, in conflict with this creed of freedom” (p. 311).
Gould was indeed responding to those who believe that the fixed direction of evolution provides
moral direction. As we have seen, that was not a made-up audience (and there is much more that
could be said about these sorts of believers; see Ruse 1996). The other equally real (and
overlapping) audience that he was addressing consisted of those who believe that God built the
32
world in such a way as to guarantee our existence, along with our rational faculties and moral
sense. According to Gould, evolutionary contingency undermines both positions. There is no
direction to evolution, hence no moral guidance to be had there. And the kinds of processes that
led to us hardly guaranteed us. We, along with our rational faculties and moral sense, were no
more intended than parasitic wasps (or pit bulls).
When Gould argued that evolutionary contingency was a source of “freedom and
consequent moral responsibility,” he did not mean “freedom” per se. He only meant that we are
free to look elsewhere—beyond the fossil record and creation stories—for the distinction
between right and wrong. And if we are free to look elsewhere, then we are also morally
obligated to do so. That’s all he meant.
33
References
Baer, Karl Ernst von. [1873] 1973. “The Controversy over Darwinism.” Trans. by David L. Hull
in Hull, ed., Darwin and his Critics: Cambridge: Harvard University Press.
Beatty, John. 1984. “Pluralism and Panselectionism.” PSA 1984, Proceedings of the Philosophy
of Science Association 2: 113-128.
Beatty, John. 2006. “Replaying Life’s Tape.” Journal of Philosophy 103: 336-362.
Conway Morris, Simon. 2003. Life’s Solution: Inevitable Humans in a Lonely Universe.
Cambridge: Cambridge University Press.
Darwin, Charles. 1872. On the Origin of Species by Means of Natural Selection, 6th ed. London:
John Murray.
Darwin, Charles. 1877. On the Various Contrivances by Which Orchids Are Fertilised by
Insects, 2nd ed., revised. London: John Murray.
Darwin, Charles. 1875. The Variation of Animals and Plants under Domestication, 2 vols., 2nd
ed. London: John Murray.
Darwin, Charles. 1987. Charles Darwin’s Notebooks, 1836-1844, Paul H. Barrett et al., eds.
Ithaca: Cornell University Press.
Darwin, Charles. 1993. The Correspondence of Charles Darwin, vol. 8, Frederick Burkhardt et
al., eds. Cambridge: Cambridge University Press.
Dennett, Daniel. 1995. Darwin’s Dangerous Idea: Evolution and the Meaning of Life. New
York: Simon & Schuster.
Gerard, Ralph W. 1940. “Organism, Society and Science, I and II.” Scientific Monthly 50: 340-
350, 403-412.
34
Gould, Stephen Jay. 1983. “The Hardening of the Synthesis.” In Marjorie Grene, ed.,
Dimensions of Darwinism. Cambridge: Cambridge University Press.
Gould, Stephen Jay. 1985. “Recording Marvels: The Life and Work of George Gaylord Simpson
(1902-1984).” Evolution 39: 229-232.
Gould, Stephen Jay and Richard C. Lewontin. 1979. “The Spandrels of San Marco and the
Panglossian Paradigm: A Critique of the Adaptationist Programme.” Proceedings of the Royal
Society of London B 205: 581-198.
Gould, Stephen Jay. 1989. Wonderful Life: The Burgess Shale and the Nature of History. New
York: Norton.
Gould, Stephen Jay. 1990. The Individual in Darwin’s World. Edinburgh: Edinburgh University
Press.
Gould, Stephen Jay. 1995. “Boyle’s Law and Darwin’s Details.” Natural History 104: 8-11, 68-
71.
Gould, Stephen Jay. 1999. Rocks of Ages: Science and Religion in the Fullness of Life. New
York: Ballantine.
Gray, Asa.1860. “Darwin and his Reviewers.” Atlantic Monthly 6 (1860 October): 406-425.
Hodge, M.J.S. and David Kohn. 1985. “The Immediate Origins of Natural Selection.” In Kohn,
ed., The Darwinian Heritage. Princeton: Princeton University Press.
Johnson, Paul A. et al. 1995. “Theoretical Analysis of Divergence in Mean Fitness Between
Initially Identical Populations.” Proceedings of the Royal Society of London B 259: 25-130.
Kellogg, Vernon L. 1907. Darwinism To-Day. New York: Henry Holt.
Kohn, David. 1980. “Theories to Work By: Rejected Theories, Reproduction, and Darwin’s Path
to Natural Selection.” Studies in the History of Biology 4: 67-170.
35
Lennox, James. 1993. “Darwin was a Teleologist.” Biology and Philosophy 8: 409-421.
Lenski, Richard E. and Michael Travisano. 1994. “Dynamics of Adaptation and Diversification:
A 10,000-Generation Experiment with Bacterial Populations.” Proceedings of the National
Academy of Sciences, USA 91: 6808-6814.
Losos, Jonathan, 1998. “Contingency and Determinism in Replicated Adaptive Radiations of
Island Lizards.” Science 279: 2115-2118.
Mani, G.S. and B.C. Clarke. 1990. “Mutational Order: A Major Stochastic Process in
Evolution.” Proceedings of the Royal Society of London B 240: 29-37.
Osborn, Henry Fairfield. 1925. “The Origin of Species. II” Proceedings of the National Academy
of Sciences, USA 12: 749-752.
Osborn, Henry Fairfield. 1926. “The Problem of the Origin of Species as It Appeared to Darwin
in 1859 and as It Appears to Us To-Day.” Science 64: 337-341.
Osborn, Henry Fairfield. 1929. The Titanotheres of Ancient Wyoming, Dakota, and Nebraska.
U.S. Geological Survey, Monograph 55, 2 vols.
Paley, William. 1809. Natural Theology; or, Evidences of the Existence and Attributes of the
Deity, Collected from the Appearances of Nature, 12th ed. London: J. Faulder.
Popper, Karl. 1957. The Poverty of Historicism. New York: Harper.
Ruse, Michael. 1996. Monad to Man: The Concept of Progress in Evolutionary Biology.
Cambridge: Harvard University Press.
Simpson, George Gaylord. 1941. “The Role of the Individual in Evolution.” Journal of the
Washington Academy of Sciences 31: 1-20.
Simpson, George Gaylord. 1941. Tempo and Mode in Evolution. New York: Columbia
University Press.
36
Simpson, George Gaylord. 1949. The Meaning of Evolution: A Study of the History of Life and of
its Significance for Man. New Haven: Yale University Press.
Simpson, George Gaylord. 1950. “Evolutionary Determinism and the Fossil Record.” Scientific
Monthly 71: 262-267. This essay was reprinted in Simpson 1964, pp. 176-189.
Simpson, George Gaylord. 1964. This View of Life: The World of an Evolutionist. New York:
Harcourt, Brace & World.
Sterelny, Kim. 2005. “Another View of Life” (a review of Simon Conway Morris, Life’s
Solution: Inevitable Humans in a Lonely Universe). Studies in History and Philosophy of
Biological and Biomedical Sciences 36: 585-593.
Sterelny, Kim and Paul E. Griffiths. 1999. Sex and Death: An Introduction to Philosophy of
Biology. Chicago: University of Chicago Press.
Travisano, Michael et al. 1995a. “Experimental Tests of the Roles of Adaptation, Chance, and
History in Evolution.” Science 267: 87-90.
Travisano, Michael et al. 1995b. “Long-Term Experimental Evolution in Escherichia coli. III.
Variation Among Replicate Populations in Correlated Responses to Novel Environments.”
Evolution 49: 189-200.
Waddington, Conrad H. 1948. The Scientific Attitude, 2nd ed. West Drayton: Penguin
Waddington, Conrad H. 1942. Science and Ethics. London: Allen and Unwin.
Wahl, Lindi M. and David C. Krakauer. 2000. “Models of Experimental Evolution: The Role of
Genetic Chance and Selective Necessity.” Genetics 156: 1437-1448.