Darwin's Black Box
Michael Behe
Reviewed by Brian Rhees
I'm going to make this as brief as I can (which unfortunately isn't very), in part because there are any number of substantive and detailed reviews already available and because I have very little to add to this body of critiques.
First off, in terms of style, clarity, etc. the book is well written. Behe does a good job of marching out the relevant details, showcasing them, and marching them off stage with minimal discomfort to the audience. Given that his point has to be made in the realm of biochemistry, this is rather impressive.
As for content, well, I'd originally intended to write something
more along the lines of a point for point refutation because I
found the book to be exceptionally problematic. After about 8
pages of text and only about 2 chapters of coverage, I abandoned
the approach. Here are 5 general comments.
1. Those of you with the relevant background will probably notice
as you read the book that the further along you go, the more uncomfortable
his arguments become. Details which early on you might dismiss
as naive but trivial become more troubling as you see where he
is headed. For instance, Behe spends quite a bit of time simplifying
the protein world via analogy so that the audience can understand
the mechanistic properties of these molecules. All fine and good
until you realize that he relies on these analogies later on to
help the audience infer how they could interact and evolve. All
of a sudden, some of the details he brushed over earlier become
less trivial (another example would be his portrayal early on
of gradualism as the only flavor of evolutionary theory which
merits debate). I would caution that, when you read Behe's book,
you read the early parts carefully because they are not just collections
of causally strewn details.
2. Behe is not above a little good natured rhetoric. One of
his favored techniques is to stand up an evolutionary straw dummy,
raise his favored standard (irreducible complexity) and declare
the straw man inadequate and vanquished. Then he will separately
prop up the more obvious objections to his critique and show how
they are flawed on some unrelated ground. (The first such example
can be found roughly between pages 59-66 where he tackles the
problem of cilia evolution. Note that even though his aim is to
show that there is no possible naturalistic explanation for cilia
evolution, he
never actually does this. When he gets around to rejecting such
he does so, not because of its impossibility, but because
there is no proven explanation, two very different criteria.)
Another habit of his is to chastises evolutionists for focusing
on icing and ignoring cake, such as when it comes to his intellectual
progenitor, Paley -- this after seriously skewing Russ Doolittle's
work on clotting factors and followed shortly by a one page rebuttal
of David Hume (which constitutes a new height in hubris, IMO)
and a disappointing and naive portrayal of Richard Dickerson's
view on what constitutes "science" (see p. 239, 241
and 243).
3. On to the more substantive, I disagree with Behe on his
more fundamental points dealing with "irreducible complexity"
and its use to infer intelligent design. His points on irreducible
complexity can be simply summarized: 1) We know, from a mechanistic
and human world, that irreducibly complex structures can certainly
be generated by intelligent
designer. Paley's watch, for instance, certainly has a creator--it
is simply inconceivable that the pieces of a watch could ever
have randomly assembled. 2) Nature (and thereby evolution) is
capable of creating complexity but never irreducible complexity.
Irreducible complexity requires design and design implies forethought.
This is essentially Behe's
position.
The problem is Behe is certainly over generalizing -- we know
that nature is capable of producing irreducible complexity (or
other flavors of complexity) through non-intelligent design. Take
for instance the creationists argument of the watch and their
pregnant question: Who then is man's creator? Behe feels that
evolutionists evade the obvious answer by focusing on trivial
flaws in Paley's arguments, so let's eat the cake for a moment.
Who, for instance was my creator? Easy enough, my mother and father
or, more specifically, a pair of gametes encoding the unfolding
of my phenotypic
structure donated by Ma and Pa Rhees. I could obviously give a
more detailed explanation of who designed me by describing maternal
effects, development, etc., but the point is made. In terms of
proximal causes, I am the product of a genetic code derived
from my parents, a design which, apart from the when and where,
they had no intelligent control over. Now clearly we could examine
more distal causes, say who designed my parents and my parents'
parents, etc. in an effort to infer ultimate causes but this quickly
takes us into a realm that the watch analogy fails to illuminate
since watches don't quasi-auto-replicate--when we've discovered
the designer of the watch we've reached the end of that chain.
Now, by itself this is a philosophically weak example because
it does not address the possibility that, in terms of ultimate
causes, man was intelligently designed. I start here, however,
to make (IMO) a critical point. Irreducible complexity can easily
be shown to be the product of design, but showing that such design
was intelligently specified is a separate issue. Perhaps a less
argumentative example, where for instance the entire causal chain
is observed, is to be preferred. One of my favorites is the story
of Kwang Jeon (retold in From Gaia to Selfish Genes, ed.
Connie Barlow) who studied amoebas and discovered a rod shaped
bacteria which infected his population.Many amoebas died from
the infection, but some lived on, carrying the bacteria inside
of them. For 5 years, Jeon selected on the population until he
had a new population at least as resistant to environmental stresses
as the original in spite of their harbored bacteria.
Interestingly, this new population was also in some unspecified way dependant on the bacteria -- removal of the bacteria led to death of both amoebae and bacteria. In case the point is not clear, the achieved novel state of complexity had become irreducible through no apparent invocation of intelligent design (I suppose we could quibble about the fact that Jeon selected for viability, but I think it is clear that nature can certainly do the same). However, if your tastes run to the more complex morphological structures, there are other examples in the literature, the clearest, in my opinion, in another review of Behe (Kenneth R. Miller, Creation/Evolution,16(2):36-40)
"As factual examples we could choose any of the systems whose evolution is documented by the fossil record, a source apparently acceptable to Behe. The three smallest bones in the human body-the malleus, incus, and stapes-carry sound vibrations across the middle ear, from the membrane-like eardrum to the oval window. This five-component system fits Behe's test of irreducible complexity perfectly-if any one of its parts is taken away or modified, hearing would be lost. This is the kind of system that evolution supposedly cannot produce. Unfortunately for "intelligent design," the fossil record elegantly and precisely documents exactly how this system formed. During the evolution of mammals, hones that originally formed the rear portion of the reptilian lower jaw were gradually pushed backwards and reduced in size until they migrated into the middle ear, forming the bony connections that carry vibrations into the inner ears of present-day mammals. A system of perfectly-formed, interlocking components, specified by multiple genes, was gradually refashioned and adapted for another purpose altogether - something that this book claims to be impossible. As the well-informed reader may know, creationist critics of this interpretation of fossils in the reptile to mammal transition once charged that this could not have taken place. What would happen, they joked, to the unfortunate reptile while he was waiting for two of his jaw bones to migrate into the middle ear? The poor creature could neither hear nor eat! As students of evolution may know, Fuzz Crompton of Harvard University brought this laughter to a deafening halt when he unearthed a fossil with a double articulation of the jaw joint-an adaptation that would allow the animal both to eat and hear during the transition,enabling natural selection to favor each of the intermediate stages. "
Now, key point: Certainly none of these examples are at the biochemical level which is where Behe frames his argument, but they should give us pause to consider. We know that irreducible complexity can be produced by biology on many levels. Is there any reason, apart from Behe's incredulity, that we should simply assume that it is not possible at the molecular as well? Alternatively, could we frame hypothesis of natural causes which might generate biochemical pathways and can we then validate our hypothesis? Here we run into a key problem: lack of information. We can frame hypotheses, even plausible ones (Behe even helps us here), but validating them is difficult. Why? Because biochemical evolution is, for the most part, the oldest, most obscured piece of our biological history. Much of what Behe talks about involves hundreds of millions of years (and perhaps the occasional billions of years) of random noise to sort through. (All of Behe's touting the literary silence on this subject only serves to underscore this point, IMO.) It is one thing to track an animal through new fallen snow, and another thing entirely to track the same animal a year later through an urban intersection. Should we fail to track in the former case we are incompetent; if we fail in the latter, what does it mean? That the animal went a different way? Of course not. We simply find that we now have no grounds to prefer one hypothesis over the other, but Behe doesn't like this sort of situation (which is perhaps why he dispenses so offhandedly with David Hume's strict empiricism). He wants us to prefer a hypothesis, specifically his hypothesis of intelligent design, in the absence of empirical evidence. This may sound like an extreme claim on my part but you may find the relevant piece on page 219 where (quoting Eliott Sober) he dispenses with reasoning by induction and reclassifies intelligent design as "an inference to the best explanation." To which I wish to append " . . . in the absence of useful information."
One other thought occurred to me, along these lines, as I read the book -- the whole presentation of data is of course a carefully crafted tour, and one which doesn't take any unnecessary turns. Certainly, part of this is just good showmanship, but there were times when I questioned the strict honesty of the tour. For instance, when Behe describes cilia does he tell you how tubulin and dynein are otherwise employed in the cell? No, of course not. Instead, he gives you: The Cilia, A Closed System. This is very much like presenting a human arm, removed from the body, to an alien audience which is then asked to theorize about its purpose functionality and origin. In many cases, Behe's strength is simply our ignorance. When you take Behe's tour, you have to take it on faith that he is presenting all the possibly relevant information that may bear on our judgment.
Enough said here. There are those that will accept Behe's position because it appeals to them for various reasons. Personally, I don't think science offers any justification for preferring Behe's explanations, but it is also abundantly clear that science isn't the only criteria for evaluating the appeal of an argument.
4. There has been some discussion on this group of whether or not Behe's reasoning constitutes a "God of the Gaps" mentality. After reading his views, I'd say it is difficult to conclude otherwise. For reasons I've argued above it may be premature to concede that there is a one to one correspondence between irreducible complexity and intelligent design.
5. Even if we allow that biochemical irreducible complexity is only producible via intelligent design, this really doesn't help us much in identifying intelligently designed structures and it doesn't take us a step closer to understanding how things were constructed (which, by the way, was Behe's first and key criticism against naturalistic explanations) since Behe is quick to concede that, intelligent though the creator may be, we shouldn't expect Him/Her/It to conform to our sense of rationality. Behe offers all sorts of criteria for determining irreducible complexity but it should be exceedingly clear that intelligent design can yield structures that are pretty darn simple. How we are supposed to distinguish these from those naturally evolved is not explained, and it seems that Behe is not concerned if this sort of misidentification occurs in spite of his irrepressible search for Truth, regardless of where it might take him. As for demonstrating irreducible complexity, in essence, he puts the burden on the researcher to show they've exhausted all relevant alternative hypothesis. His belief is that scientists are not so naive as to need much assistance in this area in discriminating between the truly irreducibly complex and the merely apparently irreducibly complex. At this point it should be clear that Behe's disaffection goes well beyond evolution itself and into the realm of scientific method. In fact it is probably also worth mentioning that Behe's fundamental characterization of naturalistic science is that, "it is a prescription for timidity."
In closing, I found Behe to be articulate but uninspiring. Usually when I invest effort into a book I walk away with some kernel of value. Behe's portrayal of science and philosophy has left me with something more akin to an allergic rash. While I would not suggest that anyone avoid the book, I would not recommend it as a first book on the subject of evolution and design either.
Note: There was a discussion on Eyring-l concerning this book and it's claim. It is available from the list archives. Search for Behe.