Koons, Part II
In yesterday's posting I began discussing Robert Koons' essay from the recent anthology Uncommon Dissent
. Koons pretends that natural selection is an abstract principle used by scientists to avoid confronting the complexity of living organisms. In reality scientists use natural selection as a starting point for generating hypotheses about how various structures evolved. Such scenarios must successfully account for all that is known about the operation of a system. Having cleared that explanatory hurdle, such scenarios inevtiably have consequences for what should be found in other, related organisms. When those consequences are seen to obtain, our confidence in the hypothesis grows. This method of reasoning has been so successful in biology that we are justified in saying that natural selection lies behind much of the complexity we see around us.
Koons, good philosopher that he is, completely ignores all of the actual, empirical work biologists have done over the years. Instead he offers his own list of “Five Stages in the Confirmation of Darwinism” (P. 9-11).
- Stage 0. The original, pre-Darwinin situation, in which the complex functionality of life triggered a natural human disposition to recognize intelligent agency, creating a strong presumption in favor of such agency as the cause of life.
- Stage 1. An alternative mechanism is proposed, random variation culled by natural selection, and preliminary evidence in favor of the new hypothesis is gathered and systematized.
- Stage 2. In several paradigmatic cases, hypothetical Darwinistic pathways leading to actual adaptive forms are described in sufficient detail and with sufficient understanding of the underlying genetic and developmental processes that it seems virtually certain that these pathways represent genuine possibilities. These pathways must be possible, not only in the sense of involving no violation of physical or chemical laws, but also in the sense that every step in the path can be assigned an estimated probability that is sufficiently high for the joint probability of the entire pathway to be consistent with a reasonable belief that such a thing might really have happened.
- Stage 3. For a significant number of hypothetical pathways of the kind described in stage two, we are able to verify that the pathway was probably actualized in history. New evidence from fossils and homologies is found that conforms to our specific expectations, based on the hypothetical pathways, and few if any instances of evidence are found that cannot readily be explained in terms of these pathways. Each hypothetical pathway describes a large number of intermediate steps, leading from some known ancestral form lacking the adaptation in question to some known form possessing it. Each step should be fully described at both the genetic and the morphological level: that is we should be specific about what mutations, lateral gene transfers, or other processes have occurred, and how the new genotype is expressed in morphology. For each step a hypothetical environment needs to be specified, and the tools of population genetics employed to show that the hypothetical new genotype would in fact be selected over its rivals in the hypothetical environment.
- Stage 4. If nearly every case of apparent design has been successfully explained in Darwinian terms, and in each case we have found an overwhelming body of specific, confirming evidence, we are justified in treating Darwinism as established beyond a reasonable doubt.
Want to guess the stage at which modern evolutionary theory resides?
Where are we today? Leading biologists assure us that we are at stage four. In fact, however, I believe that we are still in stage one.
As is stated in Koons' explanation of Stage 0, his starting point is that some vague notion of ID is the default position. In other words, if scientists can not meet his explanatory hurdles then design should be accepted as the correct explanation. It is a starting point so asinine that it is enough, all by itself, to disqualify anything else he says from serious consideration. Nonetheless, let us consider whether Koons' five stages have any merit.
We begin with a simple observation: Population genetics has no tools to tell us whether a hypothetical new genotype will be selected over its rivals in a hypothetical environment. The determination of whether a particular allele will be selected over its rivals will depend on the effect it has on the organism's phenotype. The selective advantage (or disadvantage) conferred by a new mutant appears as a variable in the equations of population genetics; some determination of its value must be made before these tools can be applied to any real world situation.
I do not claim that only biologists should be allowed to comment on evolution (after all, my training is in mathematics). But I do say that if a non-biologist is going to argue that the community of biologists is all mixed up about the epistemic status of their discipline, then he really must be extra careful to learn the basics of the subject. Do you get the impression that Koons has really made much of an effort to educate himself on the nature of population genetics? Do you think he could give a coherent description of what lateral gene transfer is, or on the various types of mutations and their effects on organisms (other terms he throws around rather loosely)? If he has not mastered these basic elements of biology, then why should we take seriously his assessments of evolution?
Now, back to his list. Koons charges evolution with being mired at Stage One. This is unsurprising, considering that he has rigged Stage Two to make it impossible to reach. Koons expects us to be able to describe specific genetic events that happened millions of freakin' years ago
in animals that are currently extinct
! He then expects us to perform elaborate probability calculations to determine whether our hypothetical genetic events might actually have happened. Leaving aside the fact that such a calculation would depend on far more variables than we could realisticly estimate, there is also the fact that low probability by itself tells us nothing about whether an event might actually have happened. Even William Dembski understands that point.
What we do know is that every bit of an organism's phenotype is under genetic control, and that changes in genes can, in principle, affect any bit of that phenotype. We also have a large catalog of mutations in modern organisms and the effect those mutations have on the phenotype of the organism. We know enough to assert that any sufficiently small genetic event will almost certainly happen in a large population given enough time. Finally, we can assert that any sufficiently small change in phenotype can be attained by a correspondingly small genetic change.
Ken Miller's scenario for the evolution of blood clotting
, mentioned yesterday, provides a good example of these ideas in action. Recall that Miller's scenario does not invoke any sort of genetic mechanism that is not known to occur in modern organisms, he discusses the effect these changes would have on our nascent blood clotting system, and ultimately uses his scenario to derive verifiable (and verified!) predictions about blood clotting in other organisms. Of course, Miller did not specify specific genetic loci that mutated, nor did he isolate specific species in which these changes occurred, because that level of detail is simply impossible to obtain. If Koons believes that Miller's scenario is too vague to be taken seriously, he will have to tell us what more he wants.
To see how Koons treats actual biology, let's consider the only example he discusses: the paper “A Pessimistic Estimate of the Time Required For an Eye to Evolve”, by Nilsson and Pelger. First, here is what Koons says:
Take, for example, Richard Dawkins's attempts to prove that Darwinism is able to explain the emergence of the vertebrate eye. Dawkins refers to a computer simulation by Nilsson and Pelger, showing that one can gardually improve a light-sensitive spot and reach, in 1800 steps or so, a fully-functional, lens-bearing eye. This might be impressive, except that the computer simulation (like every single simulation referred to by Dawkins in the book) entirely omits the two crucial details about real biology: the genotype/phenotype distinction (the distinction between the genetic constitution of an individual or group as opposed to the properties produced by interaction with its environment) and the processes of embryological development. The steps in Nilsson and Pelger's program are phenotypical (that is, they concern changes in gross, morphological feaures in the fully formed adult). We are not given a model in which the successive forms of the eye are determined by successive trajectories of embryological development, nor are we given a model of how these successive trajectories are determined by successive, feasible mutations. Given these limitations, it is, of course, impossible to estimate the probabilities of the mutations required for each of the 1800 steps in the creation of the vertebrate eye. (P. 9)
It is amusing that Koons repeats Dawkins' error in describing what Nilsson and Pelger did as a computer simulation, since his colleague David Berlinski published a much ballyhooed article in Commentary
in which he made a big fuss over the fact that there was no computer simulation in the paper. This point is not relevant to discussing the arguments of either Dawkins or Koons, however.
The paper Koons mentions contained two major accomplishments. The authors described a sequence of small steps bridging a light-sensitive spot of the sort possessed by various worms and microorganisms with a vertebrate eye. They then performed a calculation in which they showed that such an eye could evolve in a remarkably short amount of time.
The sequence Nilsson and Pelger described began with a light sensitive spot. At each step, small changes were made in the size or shape of the spot in such a way that the visual acuity of the resulting proto-eye would be an improvment over what came before. Here is a brief quote from the paper:
We let the evolutionary sequence start with a patch of light-sensitive cells which is backed and surrounded by dark pigment, and we expose this structure to selection favouring spatial resolution. We assume that the patch is circular, and that selection does not alter the total width of the structure. The latter assumption is necessary to isolate the design changes from general alterations of the size of the organ. There are two ways by which spatial resolution can be gradually introduced: (i) by forming a central depression in the light-sensitive patch; amd (ii) by a constriction of the surrounding pigment epithelium. Both these morphological changes reduce the angle through which the individual light-sensitive cells receive light. The relative effects that depression and constriction have on the eye's optical resolution are compared in figure 1a. Initially, deepening of the pit is by far the most efficient strategy, but when the pit depth equals the width, aperture constriction becomes more efficient than continued deepening of the pit.
It was a significant accomplishment for Nilsson and Pelger to exhibit a sequence of small steps leading from a light-sensitive spot to a lens-bearing eye. Prior to their work, it was not obvious that such a sequence existed. Their scenario gains strength from the fact that every proto-eye in their sequence can be found in modern organisms. This proves that the eyes they described really would be functional and useful.
Koons asks us to dismiss their work out of hand, on the grounds that they described morphological changes to the eye without specifying which gene mutated when to produce those changes. Is Koons really suggesting that it is too implausible to suggest that a small genetic mutation changed a light-sensitive spot into a slightly invaginated light-sensitive spot? Do we have to describe elaborate processes of embryological development before we accept the contention that natural selection could cause an invaginated spot to constrict itself into a camera eye? Known mutations in known organisms cause precisely those sorts of changes to occur.
The fact is that Koons does not raise these issues because they actually cast doubt on the validity of Nilsson and Pelger's work. He raises them because they allow him to present himself to his scientifically-illiterate readers as the hard-nosed skeptic facing off against ideology-blinded scientists willing to wave their hands to prop up their worldview.
In case after case biologists have gotten their hands dirty understanding the minutiae of complex systems in biology and have emerged with useful scenarios for how these systems evolved. These scenarios have passed every test we can possibly put them to. But Koons prefers to fold his arms and shake his head. Don't bother him with evidence, he has religious assumptions to defend.
Before closing, let me also point out that paleontology has provided us with precisely the sort of new fossils that he asks for in Stage Three. Surely he is aware of, among many other examples, the reptile to mammal transition. This is a sequence of fossils which shows in great detail how the mammalian inner ear bones evolved from similar bones in the reptilian jaw. This transition was once considered so inherently implausible (after all, the intermediates would be unable either to eat or to hear, right?) that creationists used it as exhibit A in their case against Darwin. Yet we now have detailed fossil and ebryological evidence to tell us how it happened. Meanwhile, each new fossil dug up is another opportunity to score a devastating blow against evolution. But no fossil has turned up that is out of place from an evoltuionary standpoint.
As for new evidence from homologies, the various genome projects have been producing them in droves of late. Do you think Koons cares?
Koons also completely ignores the fact that we can have indirect evidence for selection's action, as I discussed yesterday.
The simple fact is that Koons demands a level of detail that is flatly impossible to obtain. Those requirements on his list that actually can be met, have been. In droves. Koons only mentions a single paper from the scientific literature, and his discussion of it is laughably inadequate. It should be evident that Koons hasn't the slightest interest in giving serious thought to the accomplishments of modern biolgy.
Sadly, there is still more to critique in Koons' paper. We will turn to that tomorrow.