Why Are Scientists
Confident that Complex Biological Systems Evolved Gradually?
By Jason Rosenhouse
Posted August 9, 2005
intelligent design (ID) assert that certain complex biological systems could
not emerge from a gradual evolutionary process. They argue instead that such structures are best explained
via the deliberate action of an unspecified intelligent designer.
endorse this conclusion, and they have good reasons for being skeptical. They understand that the prolonged
action of natural selection can be expected to leave traces behind in the
structure of modern organisms. And
when scientists go looking for those traces they invariably find them in
natural selection operates by preserving small, favorable variations that occur
naturally in any population of organisms.
Over time these variations accumulate to the point that large-scale
change is the result. This implies
that natural selection works by modifying structures already present in the
organism. It does not craft new,
complex systems from scratch.
is crucial in distinguishing between those systems that could have been crafted
by selection and those that could not have been. If we find that a particular organism possesses a complex
system made from parts wholly distinct from anything to be found in the
organism's closest evolutionary cousins, it will be difficult to explain that
system via selection. But if we
find that the system appears to be cobbled together from parts that were
readily available, then natural selection remains a strong candidate.
employed this principle in his studies of the complex systems used by orchids
to attract pollinating insects. He
discovered that these contrivances, as he called them, were indeed fashioned
out of modified versions of parts present in closely related flowers. Stephen Jay Gould famously used the
panda's "thumb" to illustrate the same principle. The panda possesses a sixth digit on its front paws that it
uses to strip the leaves off of bamboo.
This digit is not a true opposable thumb like that possessed by apes and
humans. If it were, we would have
a strong argument against natural selection in this case, since the panda's
closest relatives have nothing like such a thumb. In reality, however, the panda's thumb is cobbled together
from alterations in the bones found in the paws of other bears. Since examples
like these are ubiquitous in nature we see that natural selection passes its
first big test.
extends to the biochemical realm as well.
In his book Finding Darwin's God, cell biologist Kenneth Miller
offers the following thoughts on the structure of the vertebrate blood clotting
The striking thing about this particular Rube
Goldberg machine is how similar most of its parts are. Nearly all of the regulatory molecules
belong to a single class of protein-cutting enzymes known as "serine
proteases," and that...is the clue to understanding the system's evolution
beginning with organisms that lacked a protein-based clotting system.
From this starting point Miller
develops a scenario for the evolution of the modern blood clotting system from
simple precursors found in invertebrates.
An important step in his scenario involves a series of gene duplications,
followed by the divergence of the duplicate copies. This explains why the individual proteins in the clotting
systems are so similar. Miller
goes on to describe ways of testing his scenario:
If the clotting cascade really evolved the way I
have suggested, the clotting enzymes would have to be near-duplicates of a
pancreatic enzyme and of each other.
As it turns out, they are.
Not only is thrombin homologous to trypsin, a pancreatic serine
protease, but the six clotting proteases...share extensive homology as well. This is consistent with the notion that
they were formed by gene duplication, just as suggested. But there is more to it than that. We could take one organism-humans for
example -- and construct a branching tree based on the relative degrees of
similarity and difference between each of the clotting proteases. Now, if the gene duplications that
produced the clotting cascade occurred long ago in an ancestral vertebrate, we
should be able to take any other vertebrate and construct a similar tree in
which the relationships between the clotting proteases match the relationships
between the human proteases. This
is a powerful test for our scheme because it requires that sequences still
undiscovered should match a particular pattern. And...it is also a test that evolution passes in one organism
Many other tests and predictions can be imposed on
the scheme as well...If the modern fibrinogen gene really was recruited from a
duplicated ancestral gene, one that had nothing to do with blood clotting, then
we ought to be able to find a fibrinogen-like gene in an animal that does not
possess the vertebrate clotting pathway.
Miller goes on to describe the
discovery of such a gene in a sea cucumber, thereby producing another piece of
evidence for his proposed scenario.
description a second principle emerges.
Not only does every complex system studied in detail give the appearance
of a Rube Goldberg machine cobbled together from available parts, but also
evolutionary scenarios for their formation invariably lead to further testable
Contrary to the protestations of
anti-evolutionists, insight into the evolutionary histories of complex systems
seems to come out of laboratories on almost a daily basis. For example, the March 18, 2005 issue
of the professional journal Science contained a research article on the
evolution of swim bladders in fish.
The accompanying commentary described the article's findings as follows:
Scuba divers wear air-filled dive vests to move up and down
in the water column. Researchers have now used the fish family tree to piece
together how the piscine equivalent, an internal air sac called a swim bladder,
evolved a complex capillary network and special hemoglobin molecule to inflate
it with oxygen. Moreover, according to the proposal presented ... by Michael
Berenbrink of the University of Liverpool, United Kingdom, and his colleagues,
these innovations helped fish expand their species diversity. "The scenario
developed presents a fascinating picture of the evolution and radiation of
fish," says Bernd Pelster, an animal physiologist at the University of
Shortly thereafter the March 2005
issue of the journal Genome Research published the results of recent
research into the evolution of snake venom. In both cases the scientists applied logic similar to what
Miller described, and they were rewarded with success in finding plausible
evolutionary scenarios that were consistent with the copious data they had
collected. Surely such consistent
success in applying the logic of natural selection to modern complex systems
counts as evidence in selection's favor.
the evidence in these cases is inevitably circumstantial. It could hardly be otherwise
considering that the systems whose formation we are trying to explain evolved
long ago. With that in mind we
might ask whether there is some "in principle" reason for rejecting natural
selection as a plausible explanation.
In other words, can we find some theoretical reason why natural selection
is fundamentally inadequate to explain complex biological systems?
argument made by ID proponents in this regard is based on the idea of
irreducible complexity. Michael
Behe coined this term in his 1996 book Darwin's Black Box. He defined a system to be irreducibly
complex if it consisted of several, well-matched parts each of which was
essential for the system to function properly. It was his assertion that such a system could not evolve by
gradual accretion, because any intermediate structures would have to be
nonfunctional. Since there are
plenty of biological systems that fit Behe's definition, the conclusion is that
there are complex biological systems whose formation simply can not be attributed
to prolonged selection, regardless of any other evidence.
If Behe were
right, the observation of irreducible complexity would instantly trump whatever
circumstantial evidence I could provide in favor of natural selection. But he is not right. Immediately after Behe's book hit the
stores, scientists took up the thankless task of stating the obvious:
irreducible complexity in the present tells us nothing about functional
precursors in the past. This has
been demonstrated in two ways: (1) by describing general schemes, based solely
on known biological processes, whereby an irreducibly complex system could
arise gradually (for example, irreducible complexity could result from the
reduction in redundancy that occurs when subsequent mutations cause the two
copies of a duplicated gene to diverge); (2) by using these schemes to produce
scenarios for explaining specific biochemical machines.
In response to
these observations ID proponents generally respond that the various schemes
referred to in point (1) above are mere guesses, while the scenarios in point
two (2) invariably lack sufficient detail to be considered definitive. Both of these objections miss the
point. It is the ID proponents who
are making sweeping assertions about what is possible and what is not. Scientists are simply offering an "in
principle" response to an "in principle" argument. And since scientists base their scenarios solely on familiar
processes, it is the ID folks who have to explain why irreducible complexity is
something people should get worked up over.
make other "in principle" arguments against natural selection, variously based
on probability theory or on selection's lack of foresight, but all such
arguments are completely without merit.
The determination that a given system could not have evolved gradually
can only be based on a detailed understanding of the structure and function of
that system, not on any abstract, armchair reasoning.
further lines of evidence we could cite in support of natural selection's
importance, but they will have to wait for future columns. Let us instead consider a different
question: Circumstantial evidence
notwithstanding, does ID provide a better explanation than natural selection
for complex biological systems?
that possibility we should begin with the observation that ID arguments are
always indirect. ID proponents
never argue, "We observe X.
Therefore, ID." Instead
they argue, "We observe X. X
cannot plausibly be explained naturalistically. Therefore, ID."
Since we have already seen that prolonged natural selection is capable
both in theory and practice of explaining complex systems, the design
hypothesis receives a serious blow right from the start.
But there is a
graver objection. Design
proponents try to present ID as a simple extrapolation from the actions of
known intelligent agents. William
Dembski expresses the argument this way in his book The Design Revolution:
It is well known that intelligence produces
irreducibly complex systems. (For
example, humans regularly produce machines that exhibit irreducible
complexity.) Intelligence is thus
known to be causally adequate to bring about irreducible complexity.
This is rather like claiming that
mountains are evidence for the existence of giant moles. After all, molehills are something that
moles are known to produce, and what is a mountain if not a giant
fact is that the feats performed by the designer in ID are orders of magnitude
beyond anything known intelligent agents are capable of. Human beings may possess the highest
level of intelligence in the known universe, but we have no idea how to jigger
with an organism's genome to bring a blood clotting cascade or a swim bladder
into existence. And those are among
the simpler things our hypothetical designer is called upon to do. Ask a scientist to create life,
manipulate fundamental constants of the universe, or bring whole worlds into
being, and he will stare at you helplessly. The fact is, if we were only extrapolating from known causes
we would have to conclude that intelligence is fundamentally incapable of
accomplishing what is being asked of it.
the situation is this: On the one
had we find that there is no reason in theory why evolution cannot account for
complex systems. Furthermore,
every one of the numerous complex systems studied in detail has just the
structure it ought to have if it originated via known mechanisms. Scientists use this fact to formulate
useful hypotheses about the history of these systems, and can claim one
explanatory success after another as a result. For scientists the hypothesis that a system evolved by
natural selection is the beginning, not the end, of their investigation.
the other hand we find that ID proponents find it more plausible to invent out
of whole cloth a designer capable of performing feats that can only be
described as magical. They have
not a single explanatory success to their credit, and have given no reason to
believe their hypotheses can ever lead to anything useful. For ID proponents the assertion of
design is the end of the investigation.
explanation do you think scientists should embrace?