Letters to the Editor
I am pleased that your correspondents Fezer and Greene (Creation/Evolution XX) were favorably impressed by my pro-evolution argument based on the observation of the same genetic "errors" shared between different species (Creation/Evolution XIX), and I am happy to address the "loose ends" they question.
1. The legal cases that I cited (in which shared errors were taken as evidence of plagiarism) are: Colonial Book Co., Inc. v. Amsco School Publications, Inc., District Court, S.D. New York, September 9, 1941 (the chemistry textbook case), and SubContractors Register, Inc. v. McGovern's Contractors & Builders Manual, Inc., et al. District Court, S.D. New York, August 2, 1946 (the construction industry directory case). These and several other cases making essentially the same point are cited in Nimmer on Copyright by Melville B. Nimmer (1985, volume 3, pp. 13-44 to 13-45).
2. The quantitative comparisons between the two human epsilon pseudogenes and the functional human epsilon gene show 98 percent nucleotide identity for the truncated classical pseudogene and 86 percent identity for the processed pseudogene (comparing the corresponding coding sequences of each pseudogene with the functional gene). The probability of these sequence similarities occurring by chance is extremely small. Though I have not made a formal calculation of the odds, the statistical remoteness of such an event may be compared with an example often cited by the creationists: the likelihood of a hurricane blowing through a junkyard and assembling a 747 airliner. The reason that this picturesque analogy applies to the formation of the pseudogenes and not to the formation of functional genes is critical but is commonly overlooked by creationists.
Their often heard argument—about the statistical improbability of assembling the one hundred amino acid sequence of a typical protein by a random selection of amino acids—simply suggests that such single step selection is not a reasonable model for the evolution of protein sequences, a conclusion that evolutionists can readily accept. Instead, functional genes evolved by multiple successive mutations, each one of which was not highly improbable, and each successive mutation was selected by its improved function compared with that of the predecessor gene or other competitors. As beautifully presented in Richard Dawkin's recent book, The Blind Watchmaker, this model can readily explain the evolution of functional genetic sequences of high complexity. In contrast, functionless genetic sequences are not subject to this mechanism because they are not subject to selective pressure. Thus, if one wants to discard the obvious explanation of pseudogenes—namely, that they derived from functional genes and consider instead that their similarities to functional genes are due to chance coincidence, then the likelihood of the observed sequence similarities arising by chance can only be estimated by the sort of single-step model that the creationists convincingly argue is negligibly likely.
3. The sequence discrepancies between gene and pseudogene can be classified according to their position in the triplet-nucleotide codon as suggested by Fezer. For the record, the discrepancies for the case of the truncated classical epsilon pseudogene break down as follows: four in the first position of the codon; two in the second; and five in the third. A more sophisticated analysis would consider which mutations, regardless of their position in the codon, actually cause a change in the amino acid encoded (Perler et al., Cell, 1980, 20:555). Additional evidence to test the nonfunctional nature of pseudogenes will come when the sequences of the corresponding human and nonhuman pseudogenes are available for comparison. To the extent that the pseudogenes were nonfunctional throughout their history, the frequency of human versus nonhuman sequence discrepancies within the pseudogenes should equal the frequency of discrepancies in the surrounding nonfunctional DNA matrix.
4. It is true that most sequence comparisons between human, chimpanzee, and gorilla suggest that human-gorilla species divergence occurred before the human-chimpanzee divergence—a conclusion opposite to that suggested by the presence of the classical epsilon pseudogene in human and gorilla but not in chimpanzee. As Fezer suggests (and as was pointed out both in our original paper noting the absence of this sequence in chimpanzees and in the paper from Honjo's laboratory), the absence of this pseudogene in chimpanzees could be explained by a deletion of this sequence from the chimpanzee DNA. If such a deletion occurred, it should be demonstrable by a detailed comparison of the immunoglobulin gene locus of chimpanzee and human; such an analysis has not yet been reported.
In the meantime, it is worth considering that current estimates of the dates of divergence of these three species are largely based on quantitative estimates of species differences. These estimates suggest that the gorilla-human and chimp-human splits occurred closer to each other in time than the interval between either of these and the present, so it is not surprising that a relatively small error in either divergence date could obscure the true sequence of events. For example, a small systematic increase in the rate of accepted point mutations in gorilla could have increased the human-gorilla differences and led to the present observed sequence comparisons that would thus overestimate the time since the human-gorilla split. As pointed out by Honjo's laboratory (Ueda et al., PNAS, 1985, 82:3712), the absence of the classical pseudogene in chimpanzees represents a qualitative, yes-or-no distinction between that species on one hand and human and gorilla on the other hand that may reflect the true sequence of species divergence unobscured by possible variations in the mutation acceptance rate. Additional evidence for the close human-gorilla relationship may be forthcoming if other qualitative discrepancies support the distinction between chimpanzees versus gorilla and human.
5. Finally, I would like to comment on a creationist response to the shared pseudogene argument that was communicated to me in private correspondence. How can we rule out the possibility that the sequences we call pseudogenes are not genetic errors but functional genetic elements designed by a creator for purposes that we do not presently understand? According to this view, the presence of the same pseudogene in different species could be explained by the usual creationist explanation for species similarities: that the creator designed similar species to function similarly on all levels, including the function—as yet unknown—of pseudogene sequences.
It should be stated that it is virtually impossible to prove with mathematical certainty that any biological structure is completely functionless. However, several facts we have learned about genetics in recent years strongly predict the existence of pseudogenes, at least of the "classical" type. Thus:
- It is clear that genetic sequences can be duplicated. Evidence for such duplication comes from genetic comparisons between individuals of the same species showing duplications in one individual that are absent in another. Such duplications even can be observed in comparisons between individual organisms from successive generations.
- Mutations are known to occur, altering the genetic information from one generation to the next, sometimes destroying the function of genes.
- It is known that, after genes are duplicated, mutations that destroy the function of one copy do not generally cause a serious problem for the organism because of the presence of the other functional copies.
- Finally, it is clear that there is no mechanism for the cell to recognize functionless genetic sequences and eliminate them from its DNA. Higher organisms do not appear to be under much selective pressure to minimize the amount of genetic information they carry.
These principles predict that functionless genetic sequences derived from duplicate gene copies—that is, pseudogenes—should occur at some frequency in higher organisms. Now we find genetic sequences that appear to fit exactly what these principles predict: crippled duplicates with no identifiable function. Clearly, it makes most sense to tentatively consider these pseudogenes to be examples of the functionless sequences expected from the genetic facts listed above.
The alternative—to consider these sequences serve a physiological role seems to violate Occam's razor: we would have to postulate that unexpected novel functional elements have been discovered, elements whose function is unknown, unsupported by any experimental evidence, and would have to violate many principles we have learned from experimental study of normal functional genes. For examples like the epsilon classical pseudogene—shared by some but not by other very similar species—if we postulated an important function for this sequence in humans and gorillas, we would have to explain how chimpanzees survive without it. Finally, we would have to explain why we have not observed the functionless pseudogenes expected from the facts listed above. This view of pseudogenes is clearly opposed by common sense. Furthermore, it would seem even more preposterous to suppose that retroviral sequences embedded at the same position in the DNA of humans and apes represent functional sequences. The most reasonable conclusion is that these useless DNA segments represent the products of genetic accidents that occurred in a common ancestor of ape and humans and that, despite their uselessness, these sequences were copied into the DNA of the modern descendants of that ancestor (human and ape).
—Dr. Edward E. Max
After reading the article by Francis J. Arduini in Creation/Evolution XX, I felt compelled to write—not because the article was poorly done (in fact, it was quite well written) but because the target of the article was wrong. That is to say, from what I read I believe that the point which was made and the point which was supposed to be made are two different things.
As I understand the article, the object was to refute the "argument from design." However, I missed exactly how this was done. Without a doubt, the article demonstrated a potential weakness in the "if watch, then watchmaker" analogy, but this does not in any way weaken the true argument.
It seems that Mr. Arduini mistook an illustration of the argument for the argument itself. No argument ever relies upon the validity of its illustrations to prove its truth. If the analogy is inaccurate, then it might be best not to use it, but it does not reflect in any way upon the accuracy of the argument. . . .
The purpose of the "if watch, then watchmaker" analogy is not to define the argument from design but, instead, to show that it is not unnatural or illogical to assume a creator when confronted with nature. The analogy points out the fact that, whenever we are confronted with an organized product, the initial and natural reaction is to assume that an organizer exists. . . .
The analogy does not say that creation was made in the same way that the watch was made. The analogy says that if we find organization we should look for an organizer. The argument from design declares that reason should dictate the assumption of a creator when confronted with nature in the same way the existence of a watch causes reason to conclude the existence of a watchmaker.
. . . Mr. Arduini's article also misses the point by not really understanding the meaning of the word design. . . . To define design from the article would have one believe that the products of design would always be acceptable to the American commercial society. There is much more involved in design than what was presented. Rather than proving that there is no real design in nature, Mr. Arduini simply showed that he would not hire the creator of the universe as a design engineer. . . .
Based upon an understanding of design, the argument from design poses two questions: (1) can random
events produce an ordered product, and (2) does the presence of an ordered product imply design? According to the argument, the answer to the first question is "no," and the answer to the second question is "yes." At this point, to conclude that a designer exists is nothing more than common sense. . . .
I am a reasonably intelligent human being. Being interested in the creation-evolution debate and being a Christian, I sought as many books on both sides of the issue as I could. Mostly, I did not believe the creationist ideas, but I heard them out. I listened to their radio shows and subscribed to the ICR newsletter.
Soon, I found myself in an odd position. As I listened to their case, I had two feelings. On one hand, my stomach seemed to knot up because I sensed that I was hearing something that was essentially wrong, but I could not put my finger on why it was wrong. On the other hand, my uneducated (in science) layperson's view could not argue successfully against what seemed to be impressive facts presented by mature scientists. After a while, I found myself gradually saying, "I think this creationist stuff might—just might—have something to it."
I decided to trust my stomach.
Boy, talk about being vulnerable!
If anybody ever asked me why I opposed creationism, my only answer would have been, "I don't know. My stomach gets upset when I hear about it." I would have been made into mincemeat by a committed creationist's intellectual attack. I had no real scientific arguments [to throw] back to him.
Then I discovered you guys and a whole lot of confidence came flooding back. You offer current, specific responses to the creationist arguments. I feel like I was "had" by those guys and their "evidence." I feel like a light has been lit inside me. Thank you. Thank you.
I have tried to keep my wits about me during all this searching of the creation-evolution issue. Those guys have radio, churches, subscriptions (for free!), and are networked with many other groups. (You guys are not very well known, it seems.) They state their case over and over and over. They almost won me—almost. I fought. I can understand how people who are sympathetic to their cause can become loyalists. Many folks don't appreciate and don't care about the ambiguities of mature faith and mature science. They genuinely want to clarify their faith with uncluttered "science" that has all the answers. Some really do push away their doubts about creationism because they've been warned that acceptance of evolution is being a traitor to God. I am helping to spread your news letter around the science department at my school, the library, and my friends.
—Paul R. Blundin