ICONS OF EVOLUTION? WHY MUCH OF WHAT JONATHAN WELLS WRITES ABOUT EVOLUTION IS WRONG ALAN D. GISHLICK NATIONAL CENTER FOR SCIENCE EDUCATION

INTRODUCTION that are commonly used to help to teach evolu- tion. Wells calls these the “icons,” and brands THE PARADIGM OF EVOLUTION them as false, out of date, and misleading. volution is the unifying paradigm, the Wells then evaluates ten “widely used” high organizing principle of biology. school and college biology textbooks for seven EParadigms are accepted for their overall of these “icons” with a grading scheme that he explanatory power, their “best fit” with all the constructed. Based on this, he claims that their available data in their fields. A paradigm func- treatments of these icons are so rife with inac- tions as the glue that holds an entire discipline curacies, out-of-date information, and down- together, connecting disparate subfields and right falsehoods that their discussions of the relating them to one another. A paradigm is icons should be discarded, supplemented, or also important because it fosters a research amended with “warning labels” (which he pro- program, creating a series of questions that vides). give researchers new directions to explore in According to Wells, the “icons” are the order to better understand the phenomena Miller-Urey experiment, Darwin’s tree of life, being studied. For example, the unifying para- the homology of the vertebrate limbs, digm of geology is plate tectonics; although Haeckel’s embryos, Archaeopteryx, the pep- not all geologists work on it, it connects the pered moths, and “Darwin’s” finches. entire field and organizes the various disci- (Although he discusses three other “icons” — plines of geology, providing them with their four-winged fruit flies, horse evolution, and research programs. A paradigm does not stand human evolution — he does not evaluate text- or fall on a single piece of evidence; rather, it books’ treatments of them.) Wells is right is justified by its success in overall explanato- about at least one thing: these seven examples ry power and the fostering of research ques- do appear in nearly all biology textbooks. Yet tions. A paradigm is important for the ques- no textbook presents the “icons” as a list of our tions it leads to, rather than the answers it “best evidence” for evolution, as Wells gives. Therefore, the health of a scientific field implies. The “icons” that Wells singles out are is based on how well its central theory explains discussed in different parts of the textbooks for all the available data and how many new different pedagogical reasons. The Miller- research directions it is spawning. By these Urey experiment isn’t considered “evidence criteria, evolution is a very healthy paradigm for evolution”; it is considered part of the for the field of biology. experimental research about the origin of life In his book Icons of Evolution (2000), and is discussed in chapters and sections on the Jonathan Wells attempts to overthrow the par- “history of life.” Likewise, Darwin’s finches adigm of evolution by attacking how we teach are used as examples of an evolutionary it. In this book, Wells identifies ten examples process (natural selection), not as evidence for 1 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education evolution. Archaeopteryx is frequently pre- 2000: xii). This is a serious charge; to support sented in discussions of the origin of birds, not it demands the highest level of scholarship on as evidence for evolution itself. Finally, text- the part of the author. books do not present a single “tree of life”; Does Wells display this level of scholar- rather, they present numerous topic-specific ship? Is Wells right? Are the “icons” out-of- phylogenetic trees to show how relevant date and in need of removal? And more impor- organisms are related. Wells’s entire discus- tantly, is there something wrong with the theo- sion assumes that the evidence for evolution is ry of evolution? a list of facts stored somewhere, rather than the In the following sections, each textbook predictive value of the theory in explaining the “icon” is reexamined in light of Wells’s criti- patterns of the past and present biological cism. The textbooks covered by Wells are world. examined as well, along with the grading cri- TEXTBOOK “ICONS”: teria (given in the appendix of Icons [Wells, WHY DO WE HAVE THEM? 2000] and on the Discovery Institute’s web- Paradigms and all their components are not site) that he used to assess their accuracy. What necessarily simple. To understand the depth of was found is that although the textbooks could any scientific field fully requires many years always benefit from improvement, they do not of study. It is the goal of elementary and sec- mislead, much less “systematically misin- ondary education to give students a basic form,” students about the theory of biological understanding of the “world as we know it,” evolution or the evidence for it. Further, the which includes teaching students the para- grading criteria Wells applied are vague and at digms of a number of fields of science. In times appear to have been manipulated to give order to do this, teaching examples must be poor grades. Many of the grades given are not found. It is this need to find simple, easy-to- in agreement with the stated criteria or an explain, dynamic, and visual examples to accurate reading of the evaluated text. Beyond introduce a complex topic to students that has that, Icons of Evolution offers little in the way led to the common use of a few examples — of suggestions for improvement of, or changes the “icons.” Yet, with our knowledge of the in, the standard biology curriculum. When natural world expanding at near-exponential Wells says that textbooks are in need of cor- rates, the volume of new information facing a rection, he apparently means the removal of textbook writer is daunting. The aim of a text- the subject of evolution entirely or the teaching book is not necessarily to report the “state of of “evidence against” evolution, rather than the art” as much as it is to offer an introduction the fixing of some minor errors in the presen- to the basic principles and ideas of a certain tation of the putative “icons.” This makes field. Therefore, it should not be surprising Icons of Evolution useful at most for those that introductory textbooks are frequently sim- with a certain political and religious agenda, plified and may be somewhat out-of-date. In but of little value to educators. Icons of Evolution, however, Wells makes an even stronger accusation. Wells says: References “Students and the public are being systemati- Wells, J. 2000. Icons of evolution: science or myth?: cally misinformed about the evidence for evo- why much of what we teach about evolution is wrong. lution” through biology textbooks (Wells, Regnery, Washington DC, 338p. 2 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education THE MILLER-UREY how the early atmosphere was probably differ- EXPERIMENT ent from the atmosphere hypothesized in the original experiment. Wells then claims that the THE EXPERIMENT ITSELF actual atmosphere of the early earth makes the he understanding of the origin of life Miller–Urey type of chemical synthesis was largely speculative until the 1920s, impossible, and asserts that the experiment Twhen Oparin and Haldane, working does not work when an updated atmosphere is independently, proposed a theoretical model used. Therefore, textbooks should either dis- for “chemical evolution.” The Oparin– cuss the experiment as an historically interest- Haldane model suggested that under the ing yet flawed exercise, or not discuss it at all. strongly reducing conditions theorized to have Wells concludes by saying that textbooks been present in the atmosphere of the early should replace their discussions of the Miller– earth (between 4.0 and 3.5 billion years ago), Urey experiment with an “extensive discus- inorganic molecules would spontaneously sion” of all the problems facing research into form organic molecules (simple sugars and the origin of life. amino acids). In 1953, Stanley Miller, along These allegations might seem serious; how- with his graduate advisor Harold Urey, tested ever, Wells’s knowledge of prebiotic chemistry this hypothesis by constructing an apparatus is seriously flawed. First, Wells’s claim that that simulated the Oparin-Haldane “early researchers are ignoring the new atmospheric earth.” When a gas mixture based on predic- data, and that experiments like the Miller– tions of the early atmosphere was heated and Urey experiment fail when the atmospheric given an electrical charge, organic compounds composition reflects current theories, is simply were formed (Miller, 1953; Miller and Urey, false. The current literature shows that scien- 1959). Thus, the Miller-Urey experiment tists working on the origin and early evolution demonstrated how some biological molecules, of life are well aware of the current theories of such as simple amino acids, could have arisen the earth’s early atmosphere and have found abiotically, that is through non-biological that the revisions have little effect on the processes, under conditions thought to be sim- results of various experiments in biochemical ilar to those of the early earth. This experiment synthesis. Despite Wells’s claims to the con- provided the structure for later research into trary, new experiments since the Miller–Urey the origin of life. Despite many revisions and ones have achieved similar results using vari- additions, the Oparin–Haldane scenario ous corrected atmospheric compositions remains part of the model in use today. The (Figure 1; Rode, 1999; Hanic et al., 2000). Miller–Urey experiment is simply a part of the Further, although some authors have argued experimental program produced by this para- that electrical energy might not have efficient- digm. ly produced organic molecules in the earth’s early atmosphere, other energy sources such as WELLS BOILS OFF cosmic radiation (e.g., Kobayashi et al., 1998), ells says that the Miller–Urey exper- high temperature impact events (e.g., iment should not be taught because Miyakawa et al., 2000), and even the action of Wthe experiment used an atmospheric waves on a beach (Commeyras et al., 2002) composition that is now known to be incorrect. would have been quite effective. Wells contends that textbooks don’t discuss Even if Wells had been correct about the 3 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Researcher(s) Year Reactants Energy source Results Probability Miller 1953 CH4, NH3, H2O, H2 Electric discharge Simple amino acids, unlikely organic compounds Abelson 1956 CO, CO2, N2, NH3, H2, Electric discharge Simple amino acids, unlikely H2O HCN Groth and Weyssenhoff 1957 CH4, NH3, H2O Ultraviolet light Simple amino acids (low under special conditions (1470–1294 ?) yields) Bahadur, et al. 1958 Formaldehyde, Sunlight Simple amino acids possible molybdenum oxide (photosynthesis) Pavolvskaya and 1959 Formaldehyde, nitrates High pressure Hg lamp Simple amino acids possible Pasynskii (photolysis) Palm and Calvin 1962 CH4, NH3, H2O Electron irradiation Glycine, alanine, aspartic under special conditions acid Harada and Fox 1964 CH4, NH3, H2O Thermal energy 14 of the “essential” under special conditions (900–1200º C) amino acids of proteins Oró 1968 CH4, NH3, H2OPlasma jet Simple amino acids unlikely Bar-Nun et al. 1970 CH4, NH3, H2O Shock wave Simple amino acids under special conditions Sagan and Khare 1971 CH4, C2H6, NH3, H2O, Ultraviolet light (>2000 Simple amino acids (low under special conditions H2S ?) yields) Yoshino et al. 1971 H2, CO, NH3, Temperature of 700°C Glycine, alanine, unlikely montmorillonite glutamic acid, serine, aspartic acid, leucine, lysine, arginine Lawless and Boynton 1973 CH4, NH3, H2O Thermal energy Glycine, alanine, aspartic under special conditions acid, ?-alanine, N-methyl-?-alanine, ?-amino-n-butyric acid. Yanagawa et al. 1980 Various sugars, Temperature of 105°C Glycine, alanine, serine, under special conditions hydroxylamine, aspartic acid, glutamic inorganic salts, acid Kobayashi et al. 1992 CO, N2, H2O Proton irradiation Glycine, alanine, aspartic possible acid, ?-alanine, glutamic acid, threonine, ?-aminobutyric acid, serine Hanic, et al. 1998 CO2, N2, H2O Electric discharge Several amino acids possible Figure 1. A table of some amino acid synthesis experiments since Miller–Urey. The “probabili- ty” column reflects the likelihood of the environmental conditions used in the experiment. Modified from Rode, 1999. Miller–Urey experiment, he does not explain since 1961 (see Oró, 1961; Whittet, 1997; that our theories about the origin of organic Irvine, 1998). Wells apparently missed the vast “building blocks” do not depend on that exper- body of literature on organic compounds in iment alone (Orgel, 1998a). There are other comets (e.g. Oró, 1961; Anders, 1989; Irvine, sources for organic “building blocks,” such as 1998), carbonaceous meteorites (e.g., Kaplan meteorites, comets, and hydrothermal vents. et al., 1963; Hayes, 1967; Chang, 1994; All of these alternate sources for organic mate- Maurette, 1998; Cooper et al., 2001), and con- rials and their synthesis are extensively dis- ditions conducive to the formation of organic cussed in the literature about the origin of life, compounds that exist in interstellar dust clouds a literature that Wells does not acknowledge. (Whittet, 1997). In fact, what is most striking about Wells’s Wells also fails to cite the scientific litera- extensive reference list is the literature that he ture on other terrestrial conditions under which has left out. Wells does not mention extrater- organic compounds could have formed. These restrial sources of organic molecules, which non-atmospheric sources include the synthesis have been widely discussed in the literature of organic compounds in a reducing ocean 4 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education (e.g., Chang, 1994), at hydrothermal vents troversy is really over why it took so long for (e.g., Andersson, 1999; Ogata et al., 2000), and oxygen levels to start to rise. Current data in volcanic aquifers (Washington, 2000). A show that oxygen levels did not start to rise cursory review of the literature finds more than significantly until nearly 1.5 billion years after 40 papers on terrestrial prebiotic chemical syn- life originated (Rye and Holland, 1998; thesis published since 1997 in the journal Copley, 2001). Wells strategically fails to clar- Origins of life and the evolution of the bios- ify what he means by “early” when he discuss- phere alone. Contrary to Wells’s presentation, es the amount of oxygen in the “early” atmos- there appears to be no shortage of potential phere. In his discussion, he cites research sources for organic “building blocks” on the about the chemistry of the atmosphere without early earth. distinguishing whether the authors are refer- Instead of discussing this literature, Wells ring to times before, during, or after the period raises a false “controversy” about the low when life is thought to have originated. Nearly amount of free oxygen in the early atmos- all of the papers he cites deal with oxygen lev- phere. Claiming that this precludes the sponta- els after 3.0 billion years ago. They are irrele- neous origin of life, he concludes that vant, as chemical data suggest that life arose “[d]ogma had taken the place of empirical sci- 3.8 billion years ago (Chang, 1994; Orgel, ence” (Wells, 2000:18). In truth, nearly all 1998b), well before there was enough free researchers who work on the early atmosphere oxygen in the earth’s atmosphere to prevent hold that oxygen was essentially absent during Miller–Urey-type chemical synthesis. the period in which life originated (Copley, Finally, the Miller–Urey experiment tells us 2001) and therefore oxygen could not have nothing about the other stages in the origin of played a role in preventing chemical synthesis. life, including the formation of a simple genet- This conclusion is based on many sources of ic code (PNA or “peptide”-based codes and data, not “dogma.” Sources of data include RNA-based codes) or the origin of cellular fluvial uraninite sand deposits (Rasmussen and membranes (liposomes), some of which are Buick, 1999) and banded iron formations discussed in all the textbooks that Wells (Nunn, 1998; Copley, 2001), which could not reviewed. The Miller–Urey experiment only have been deposited under oxidizing condi- showed one possible route by which the basic tions. Wells also neglects the data from pale- components necessary for the origin of life osols (ancient soils) which, because they form could have been created, not how life came to at the atmosphere–ground interface, are an be. Other theories have been proposed to excellent source to determine atmospheric bridge the gap between the organic “building composition (Holland, 1994). Reduced pale- blocks” and life. The “liposome” theory deals osols suggest that oxygen levels were very low with the origin of cellular membranes, the before 2.1 billion years ago (Rye and Holland, RNA-world hypothesis deals with the origin of 1998). There are also data from mantle chem- a simple genetic code, and the PNA (peptide- istry that suggest that oxygen was essentially based genetics) theory proposes an even sim- absent from the earliest atmosphere (Kump et pler potential genetic code (Rode, 1999). Wells al., 2001). Wells misrepresents the debate as doesn’t really mention any of this except to over whether oxygen levels were 5/100 of 1%, suggest that the “RNA world” hypothesis was which Wells calls “low,” or 45/100 of 1%, proposed to “rescue” the Miller–Urey experi- which Wells calls “significant.” But the con- ment. No one familiar with the field or the evi- 5 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education dence could make such a fatuous and inaccu- deep-sea hydrothermal vents (Figure 2). No rate statement. The Miller–Urey experiment is textbook claims that these experiments conclu- not relevant to the RNA world, because RNA sively show how life originated; and all text- was constructed from organic “building books state that the results of these experi- blocks” irrespective of how those compounds ments are tentative. came into existence (Zubay and Mui, 2001). It is true that some textbooks do not mention The evolution of RNA is a wholly different that our knowledge of the composition of the chapter in the story of the origin of life, one to atmosphere has changed. However, this does which the validity of the Miller–Urey experi- not mean that textbooks are “misleading” stu- ment is irrelevant. dents, because there is more to the origin of WHAT THE TEXTBOOKS SAY life than just the Miller–Urey experiment. Most textbooks already discuss this fact. The ll of the textbooks reviewed contain a textbooks reviewed treat the origin of life with section on the Miller–Urey experi- varying levels of detail and length in “Origin Ament. This is not surprising given the of life” or “History of life” chapters. These experiment’s historic role in the understanding chapters are from 6 to 24 pages in length. In of the origin of life. The experiment is usually this relatively short space, it is hard for a text- discussed over a couple of paragraphs (see book, particularly for an introductory class like Figure 2), a small proportion (roughly 20%) of high school biology, to address all of the the total discussion of the origin and early evo- details and intricacies of origin-of-life research lution of life. Commonly, the first paragraph that Wells seems to demand. Nearly all texts discusses the Oparin-Haldane scenario, and begin their origin of life sections with a brief then a second outlines the Miller–Urey test of description of the origin of the universe and that scenario. All textbooks contain either a the solar system; a couple of books use a dis- drawing or a picture of the experimental appa- cussion of Pasteur and spontaneous generation ratus and state that it was used to demonstrate instead (and one discusses both). Two text- that some complex organic molecules (e.g., books discuss how life might be defined. simple sugars and amino acids, frequently Nearly all textbooks open their discussion of called “building blocks”) could have formed the origin of life with qualifications about how spontaneously in the atmosphere of the early the study of the origin of life is largely hypo- earth. Textbooks vary in their descriptions of thetical and that there is much about it that we the atmospheric composition of the early earth. do not know. Five books present the strongly reducing atmosphere of the Miller–Urey experiment, WELLS’SEVALUATION whereas the other five mention that the current s we will see in his treatment of the geochemical evidence points to a slightly other “icons,” Wells’s criteria for judg- reducing atmosphere. All textbooks state that Aing textbooks stack the deck against oxygen was essentially absent during the peri- them, ensuring failure. No textbook receives od in which life arose. Four textbooks mention better than a D for this “icon” in Wells’s eval- that the experiment has been repeated success- uation, and 6 of the 10 receive an F. This is fully under updated conditions. Three text- largely a result of the construction of the grad- books also mention the possibility of organic ing criteria. Under Wells’s criteria (Wells, molecules arriving from space or forming at 2000:251–252), any textbook containing a pic- 6 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 2. Textbook treatments of the Miller–Urey experiment. Textbooks are listed in order of increasing detail (AP/College textbooks highlighted; note that Futuyma is an upper-level col- lege/graduate textbook). ture of the Miller–Urey apparatus could pictures. Wells’s criteria would require that receive no better than a C, unless the caption of even the intelligent design “textbook” Of the picture explicitly says that the experiment Pandas and People would receive a C for its is irrelevant, in which case the book would treatment of the Miller–Urey experiment. receive a B. Therefore, the use of a picture is In order to receive an A, a textbook must the major deciding factor on which Wells eval- first omit the picture of the Miller–Urey appa- uated the books, for it decides the grade irre- ratus (or state explicitly in the caption that it spective of the information contained in the was a failure), discuss the experiment, but then text! A grade of D is given even if the text state that it is irrelevant to the origin of life. explicitly points out that the experiment used This type of textbook would be not only scien- an incorrect atmosphere, as long as it shows a tifically inaccurate but pedagogically defi- picture. Wells pillories Miller and Levine for cient. exactly that, complaining that they bury the correction in the text. This is absurd: almost all textbooks contain pictures of experimental apparatus for any experiment they discuss. It is WHY WE SHOULD STILL TEACH the text that is important pedagogically, not the MILLER–UREY 7 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education he Miller–Urey experiment represents of scientific understanding of the origin of life. one of the research programs spawned This is the kind of “good science” that we want Tby the Oparin-Haldane hypothesis. to teach students. Even though details of the model for the origin Finally, the Miller–Urey experiment should of life have changed, this has not affected the still be taught because the basic results are still basic scenario of Oparin–Haldane. The first valid. The experiments show that organic mol- stage in the origin of life was chemical evolu- ecules can form under abiotic conditions. Later tion. This involves the formation of organic experiments have used more accurate atmos- compounds from inorganic molecules already pheric compositions and achieved similar present in the atmosphere and in the water of results. Even though origin-of-life research has the early earth. This spontaneous organization moved beyond Miller and Urey, their experi- of chemicals was spawned by some external ments should be taught. We still teach Newton energy source. Lightning (as Oparin and even though we have moved beyond his work Haldane thought), proton radiation, ultraviolet in our knowledge of planetary mechanics. radiation, and geothermal or impact-generated Regardless of whether any of our current theo- heat are all possibilities. ries about the origin of life turn out to be com- The Miller–Urey experiment represents a pletely accurate, we currently have models for major advance in the study of the origin of life. the processes and a research program that In fact, it marks the beginning of experimental works at testing the models. research into the origin of life. Before Miller– Urey, the study of the origin of life was mere- HOW TEXTBOOKS COULD IMPROVE THEIR PRESENTATIONS OF ly theoretical. With the advent of “spark exper- THE ORIGIN OF LIFE iments” such as Miller conducted, our under- standing of the origin of life gained its first extbooks can always improve discus- experimental program. Therefore, the Miller– sions of their topics with more up-to- Urey experiment is important from an histori- Tdate information. Textbooks that have cal perspective alone. Presenting history is not already done so should explicitly correct good pedagogy because students understand the estimate of atmospheric composition, and scientific theories better through narratives. accompany the Miller–Urey experiment with a The importance of the experiment is more than clarification of the fact that the corrected just historical, however. The apparatus Miller atmospheres yield similar results. Further, the and Urey designed became the basis for many wealth of new data on extraterrestrial and subsequent “spark experiments” and laid a hydrothermal sources of biological material groundwork that is still in use today. Thus it is should be discussed. Finally, textbooks ideally also a good teaching example because it shows should expand their discussions of other stages how experimental science works. It teaches in the origin of life to include PNA and some students how scientists use experiments to test of the newer research on self-replicating pro- ideas about prehistoric, unobserved events teins. Wells, however, does not suggest that such as the origin of life. It is also an interest- textbooks should correct the presentation of ing experiment that is simple enough for most the origin of life. Rather, he wants textbooks to students to grasp. It tested a hypothesis, was present this “icon” and then denigrate it, in reproduced by other researchers, and provided order to reduce the confidence of students in new information that led to the advancement the possibility that scientific research can ever 8 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education establish a plausible explanation for the origin Organic compounds in stony meteorites. Geochimica et of life or anything else for that matter. If Cosmochimica Acta. 27:805–834. Wells’s recommendations are followed, stu- Kobayashi, K., T. Kaneko, T. Saito, and T. Oshima. dents will be taught that because one experi- 1998. Amino acid formation in gas mixtures by high energy particle irradiation. Origins of Life and the ment is not completely accurate (albeit in hind- Evolution of the Biosphere 28:155–165. sight), everything else is wrong as well. This is Kump, L. R., J. F. Kasting, M. E. Barley. 2001. Rise of not good science or science teaching. atmospheric oxygen and the “upside-down” Archean mantle. Geochemistry, Geophysics, Geosystems –G3, 2, References paper number 2000GC000114. Anders, E. 1989. Pre-biotic organic matter from comets Maurette, M. 1998. Carbonaceous micrometeorites and and asteroids. Nature 342:255–257. the origin of life. Origins of Life and the Evolution of the Andersson, E. and N. G. Holm. 2000. The stability of Biosphere 28: 385–412. some selected amino acids under attempted redox con- Miller, S. 1953. A production of amino acids under pos- strained hydrothermal conditions. Origins of Life and sible primitive earth conditions. Science 117:528–529. the Evolution of the Biosphere 30: 9–23. Miller, S. and H. Urey. 1959. Organic compound syn- Chang, S. 1994. The planetary setting of prebiotic evo- thesis on the primitive earth. Science 130:245–251. lution. In S. Bengston, ed. Early Life on Earth. Nobel Miyakawa, S., K-I. Murasawa, K. Kobayashi, and A. B. Symposium no. 84. Columbia University Press, New Sawaoka. 2000. Abiotic synthesis of guanine with high- York. p.10–23. temperature plasma. Origins of Life and Evolution of the Commeyras, A., H. Collet, L. Bioteau, J. Taillades, O. Biosphere 30: 557–566. Vandenabeele-Trambouze, H. Cottet, J-P. Biron, R. Nunn, J. F. 1998. Evolution of the atmosphere. Plasson, L. Mion, O. Lagrille, H. Martin, F. Selsis, and Proceedings of the Geologists’ Association 109:1–13. M. Dobrijevic. 2002. Prebiotic synthesis of sequential peptides on the Hadean beach by a molecular engine Ogata, Y., E-I. Imai, H. Honda, K. Hatori, and K. working with nitrogen oxides as energy sources. Matsuno. 2000. Hydrothermal circulation of seawater Polymer International 51:661–665. through hot vents and contribution of interface chem- istry to prebiotic synthesis. Origins of Life and the Cooper, G., N. Kimmich, W. Belisle, J. Sarinana, K. Evolution of the Biosphere 30: 527-–537. Brabham, and L. Garrel. 2001. Carbonaceous meteorites as a source of sugar-related organic compounds for the Orgel, L. E. 1998a. The origin of life – a review of facts early Earth. Nature 414:879–882. and speculations. Trends in Biochemical Sciences 23:491–495. Copley, J. 2001. The story of O. Nature 410:862-864. Orgel, L. E., 1998b. The origin of life — how long did Hanic, F., M. Morvová and I. Morva. 2000. it take? Origins of Life and the Evolution of the Thermochemical aspects of the conversion of the Biosphere 28: 91–96. gaseous system CO2—N2—H2O into a solid mixture of amino acids. Journal of Thermal Analysis and Oró, J. 1961. Comets and the formation of biochemical Calorimetry 60: 1111–1121. compounds on the primitive Earth. Nature 190:389-390. Hayes, J. M. 1967. Organic constituents of meteorites, a Rasmussen, B., and R. Buick. 1999. Redox state of the review. Geochimica et Cosmochimica Acta 31:1395– Archean atmosphere; evidence from detrital heavy min- 1440. erals in ca. 3250-2750 Ma sandstones from the Pilbara Holland, H. D. 1994. Early Proterozoic atmosphere Craton, Australia. Geology 27: 115–118. change. In S. Bengston, ed. Early Life on Earth. Nobel Rode, B. M., 1999. Peptides and the origin of life. Symposium no. 84. Columbia University Press, New Peptides 20: 773–786. York. p. 237–244. Rye, R., and H. D. Holland. 1998. Paleosols and the Irvine, W. M., 1998. Extraterrestrial organic matter: a evolution of atmospheric oxygen: a critical review. review. Origins of Life and the Evolution of the American Journal of Science 298:621–672. Biosphere 28:365–383. Washington, J. 2000. The possible role of volcanic Kaplan, I. R., E. T. Degens, and J. H. Reuter. 1963. aquifers in prebiologic genesis of organic compounds 9 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education and RNA. Origins of Life and the Evolution of the Biosphere 30: 53–79. Wells, J. 2000. Icons of evolution: science or myth?: why much of what we teach about evolution is wrong. Regnery, Washington DC, 338p. Whittet, D. C. B. 1997. Is extraterrestrial organic matter relevant to the origin of life on earth? Origins of Life and the Evolution of the Biosphere 27: 249–262. Zubay, G. and T. Mui. 2001. Prebiotic synthesis of nucleotides. Origins of Life and Evolution of the Biosphere 31:87–102.

10 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education DARWIN’S “TREE OF LIFE” mon descent. Finally, he demands that text- books treat universal common ancestry as PHYLOGENETIC TREES unproven and refrain from illustrating that n biology, a phylogenetic tree, or phyloge- “theory” with misleading phylogenies. ny, is used to show the genealogic relation- Therefore, according to Wells, textbooks Iships of living things. A phylogeny is not should state that there is no evidence for com- so much evidence for evolution as much as it mon descent and that the most recent research is a codification of data about evolutionary his- refutes the concept entirely. Wells is complete- tory. According to biological evolution, organ- ly wrong on all counts, and his argument is isms share common ancestors; a phylogeny entirely based on misdirection and confusion. shows how organisms are related. The tree of He mixes up these various topics in order to life shows the path evolution took to get to the confuse the reader into thinking that when current diversity of life. It also shows that we combined, they show an endemic failure of can ascertain the genealogy of disparate living evolutionary theory. In effect, Wells plays the organisms. This is evidence for evolution only equivalent of an intellectual shell game, put- in that we can construct such trees at all. If ting so many topics into play that the “ball” of evolution had not happened or common ances- evolution gets lost. try were false, we would not be able to discov- THE CAMBRIAN EXPLOSION er hierarchical branching genealogies for ells claims that the Cambrian organisms (although textbooks do not general- Explosion “presents a serious chal- ly explain this well). Referring to any phylo- lenge to Darwinian evolution” genetic tree as “Darwin’s tree of life” is some- W (Wells, 2000:41) and the validity of phyloge- what of a misnomer. Darwin graphically pre- netic trees. The gist of Wells’s argument is that sented no phylogenies in the Origin of Species; the Cambrian Explosion happened too fast to the only figure there depicts differential rates allow large-scale morphological evolution to of speciation. If anyone deserves credit for occur by natural selection (“Darwinism”), and giving us “trees of life,” it is Ernst Haeckel, that the Cambrian Explosion shows “top- who drew phylogenies for many of the living down” origination of taxa (“major” “phyla” groups of animals literally as trees, as well as level differences appear early in the fossil coining the term itself. record rather than develop gradually), which WELLS’S SHELL GAME he claims is the opposite of what evolution ells uses phylogenetic trees to attack predicts. He asserts that phylogenetic trees the very core of evolution — com- predict a different pattern for evolution than Wmon descent. Wells claims that text- what we see in the Cambrian Explosion. These books mislead students about common descent arguments are spurious and show his lack of in three ways. First, Wells claims that text- understanding of basic aspects of both paleon- books do not cover the “Cambrian Explosion” tology and evolution. and fail to point out how this “top-down” pat- Wells mistakenly presents the Cambrian tern poses a serious challenge to common Explosion as if it were a single event. The descent and evolution. Second, he asserts that Cambrian Explosion is, rather, the preserva- the occasional disparity between morphologi- tion of a series of faunas that occur over a 15– cal and molecular phylogenies disproves com- 20 million year period starting around 535 mil- 11 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education lion years ago (MA). A fauna is a group of claiming that this proves that the fossil record organisms that live together and interact as an is complete enough to show that there were no ecosystem; in paleontology, “fauna” refers to a precursors for the Cambrian Explosion ani- group of organisms that are fossilized together mals. This claim is false. His evidence for this because they lived together. The first fauna “well documented” Precambrian fossil record that shows extensive body plan diversity is the is a selective quote from the final sentence in Sirius Passet fauna of Greenland, which is an article by Benton et al. (2000). While the dated at around 535 MA (Conway Morris, paper’s final sentence does literally say that 2000). The organisms preserved become more the “early” parts of the fossil record are ade- diverse by around 530 MA, as the Chenjiang quate for studying the patterns of life, Wells fauna of China illustrates (Conway Morris, leaves out a critical detail: the sentence refers 2000). Wells erroneously claims that the not to the Precambrian, but to the Cambrian Chenjiang fauna predates the Sirius Passet and later times. Even more ironic is the fact (Wells, 2000:39). The diversification contin- that the conclusion of the paper directly refutes ues through the Burgess shale fauna of Canada Wells’s claim that the fossil record does not at around 520 MA, when the Cambrian faunas support the “tree of life.” Benton et al. (2000) are at their peak (Conway Morris, 2000). Wells assessed the completeness of the fossil record makes an even more important paleontological using both molecular and morphological error when he does not explain that the “explo- analyses of phylogeny. They showed that the sion” of the late Early and Middle Cambrian is sequence of appearance of major taxa in the preceded by the less diverse “small shelly” fossil record is consistent with the pattern of metazoan faunas, which appear at the begin- phylogenetic relationships of the same taxa. ning of the Cambrian (545 MA). These faunas Thus they concluded that the fossil record is are dated to the early Cambrian, not the consistent with the tree of life, entirely oppo- Precambrian as stated by Wells (Wells, site to how Wells uses their paper. 2000:38). This enables Wells to omit the Wells further asserts that there is no evi- steady rise in fossil diversity over the ten mil- dence for metazoan life until “just before” the lion years between the beginning of the Cambrian explosion, thereby denying the nec- Cambrian and the Cambrian Explosion (Knoll essary time for evolution to occur. Yet Wells is and Carroll, 1999). evasive about what counts as “just before” the In his attempt to make the Cambrian Cambrian. Cnidarian and possible arthropod Explosion seem instantaneous, Wells also embryos are present 30 million years “just grossly mischaracterizes the Precambrian fos- before” the Cambrian (Xiao et al., 1998). sil record. In order to argue that there was not There is also a mollusc, Kimberella, from the enough time for the necessary evolution to White Sea of Russia (Fedonkin and Waggoner, occur, Wells implies that there are no fossils in 1997) dated approximately 555 million years the Precambrian record that suggest the com- ago, or 10 million years “just before” the ing diversity or provide evidence of more Cambrian (Martin et al., 2000). This primitive primitive multicellular animals than those seen animal has an uncalcified “shell,” a muscular in the Cambrian Explosion (Wells, 2000:42– foot (Fedonkin and Waggoner, 1997), and a 45). He does this not by producing original radula inferred from “mat-scratching” feeding research, but by selectively quoting paleonto- patterns surrounding fossilized individuals logical literature on the fossil record and (personal observation; Seilacher, pers. 12 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education comm.). These features enable us to recognize Cambrian Explosion, for example, is the first it as a primitive relative of molluscs, even time we are able to distinguish a chordate from though it lacks a calcified shell. There are also an arthropod. This does not mean that the chor- Precambrian sponges (Gehling and Rigby, date or arthropod lineages evolved then, only 1996) as well as numerous trace fossils indi- that they then became recognizable as such. cating burrowing by wormlike metazoans For a simple example, consider the turtle. How beneath the surface of the ocean’s floor do you know a turtle is a turtle? By the shell. (Seilacher, 1994; Fedonkin, 1994). Trace fos- How would you recognize the ancestors of the sils demonstrate the presence of at least one living turtle, before they evolved the shell? ancestral lineage of bilateral animals nearly 60 That is more complicated. Because its ances- million years “just” before the Cambrian tors would have lacked the diagnostic feature (Valentine et al., 1999). Sixty million years is of a shell, ancestral turtles may be hard to rec- approximately the same amount of time that ognize (Lee, 1993). In order to locate the has elapsed since the extinction of non-avian remote ancestors of turtles, other, more subtle, dinosaurs, providing plenty of time for evolu- features must be found. tion. In treating the Cambrian Explosion as a Similarly, before the Cambrian Explosion, single event preceded by nothing, Wells mis- there were lots of “worms,” now preserved as represents fact — the Cambrian explosion is trace fossils (i.e., there is evidence of burrow- not a single event, nor is it instantaneous and ing in the sediments). However, we cannot dis- lacking in any precursors. tinguish the chordate “worms” from the mol- Continuing to move the shells, Wells lusc “worms” from the arthropod “worms” invokes a semantic sleight of hand in resur- from the worm “worms.” Evolution predicts recting a “top-down” explanation for the diver- that the ancestor of all these groups was worm- sity of the Cambrian faunas, implying that like, but which worm evolved the notochord, phyla appear first in the fossil record, before and which the jointed appendages? In his argu- lower categories. However, his argument is an ment, Wells confuses the identity of the indi- artifact of taxonomic practice, not real mor- vidual with how we diagnose that identity, a phology. In traditional taxonomy, the recogni- failure of logic that dogs his discussion of tion of a species implies a phylum. This is due homology in the following chapter. If the ani- to the rules of the taxonomy, which state that if mal does not have the typical diagnostic fea- you find a new organism, you have to assign it tures of a known phylum, then we would be to all the necessary taxonomic ranks. Thus unable to place it and (by the rules of taxono- when a new organism is found, either it has to my) we would probably have to erect a new be placed into an existing phylum or a new one phylum for it. When paleontologists talk about has to be erected for it. Cambrian organisms the “sudden” origin of major animal “body are either assigned to existing “phyla” or new plans,” what is “sudden” is not the appearance ones are erected for them, thereby creating the of animals with a particular body plan, but the effect of a “top-down” emergence of taxa. appearance of animals that we can recognize as Another reason why the “higher” taxonom- having a particular body plan. Overall, howev- ic groups appear at the Cambrian Explosion is er, the fossil record fits the pattern of evolu- because the Cambrian Explosion organisms tion: we see evidence for worm-like bodies are often the first to show features that allow first, followed by variations on the worm us to relate them to living groups. The theme. Wells seems to ignore a growing body 13 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 3. Stepwise evolution of vertebrate features as illustrated by living and fossil animals. of literature showing that there are indeed cally worms with a stiff rod (the notochord) in organisms of intermediate morphology present them. The amount of change between a worm in the Cambrian record and that the classic and a worm with a stiff rod is relatively small, “phyla” distinctions are becoming blurred by but the presence of a notochord is a major fossil evidence (Budd, 1998, 1999; Budd and “body-plan” distinction of a chordate. Further, Jensen, 2000). it is just another small step from a worm with Finally, the “top-down” appearance of a stiff rod to a worm with a stiff rod and a head body-plans is, contrary to Wells, compatible (e.g., Haikouella; Chen et al., 1999) or a worm with the predictions of evolution. The issue to with a segmented stiff rod (vertebrae), a head, be considered is the practical one that “large- and fin folds (e.g., Haikouichthyes; Shu et al., scale” body-plan change would of course 1999). Finally add a fusiform body, fin differ- evolve before minor ones. (How can you vary entiation, and scales: the result is something the lengths of the beaks before you have a resembling a “fish” (Figure 3). But, as soon as head?) The difference is that, many of the the stiff rod evolved, the animal was suddenly “major changes” in the Cambrian were initial- no longer just a worm but a chordate — repre- ly minor ones. Through time they became sentative of a whole new phylum! Thus these highly significant and the basis for “body- “major” changes are really minor in the begin- plans.” For example, the most primitive living ning, which is the Precambrian–Cambrian chordate Amphioxus is very similar to the period with which we are concerned. Cambrian fossil chordate Pikia. Both are basi- 14 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

CONGRUENCE OF PHYLOGENIES BASED ON DIFFERENT SOURCES OF DATA ells also points to the occasional lack of congruence between molec- Wular- and morphology-based phylo- genies as evidence against common descent. (Molecular phylogenies are based on compar- isons of the genes of organisms.) Wells omits the fact that the discrepancies are frequently small, and their causes are largely understood (Patterson et al., 1993; Novacek, 1994). Although not all of these discrepancies can yet be corrected for, most genetic and morpholog- ical phylogenies are congruent for 90% of the Figure 4. Amniote relationships based on taxa included. For example, all phylogenies, different sources of data. Note that the only whether morphological or molecular, consider group whose position varies is turtles. all animals bearing amniotic eggs to be more closely related to one another than to amphib- a few possibilities (Rieppel and deBraga, ians. Within this group, all reptiles and birds 1996; Lee, 1997; deBraga and Rieppel, 1997; are more closely related to each other than they Zardoya and Meyer, 1998; Rieppel and Reiz, are to mammals. Finally, birds and crocodiles 1999; Rieppel, 2000; Figure 4), and none of are more closely related to each other than to these claim turtles are mammals. The uncer- lizards, snakes, and the tuatara (Gauthier et al., tainty over the precise placement of turtles 1988; Gauthier, 1994). The only group whose with respect to other groups, however, does not placement varies for both molecular and mor- mean that they did not evolve. Unfortunately, phology data sets is turtles. This is due to a genes can never be totally compared to mor- phenomenon called “long branch attraction” or phology since genetic trees cannot take fossil the “Felsenstein Zone” (Huelsenbeck and taxa into account: genes don’t fossilize. No Hillis, 1993). Long branch attraction is caused diagnostic tool of science is perfect. The when a organism has had so much evolution- imperfections in phylogenetic reconstruction ary change that it cannot be easily compared to do not make common ancestry false. Besides, other organisms, and due to the nature of the are these extremely technical topics really methodology used to evaluate phylogeny, it appropriate for introductory textbooks? can appear to be related to many possible Instead of clearly discussing these actual organisms (Felsenstein, 1978; Huelsenbeck phylogenetic issues, Wells invents one that and Hillis, 1993). This is the case for turtles. isn’t even real. He cites a 1998 paper that Turtles are so morphologically and genetically placed cows phylogenetically closer to whales different from the rest of the reptiles that they than to horses, calling that finding “bizarre” are hard to place phylogenetically (Zardoya (Wells, 2000:51). Yet this is not “bizarre” at and Meyer, 2001). Still, researchers have nar- all; it was expected. All the paleontological rowed down the possible turtle relationships to and molecular evidence points to a whale 15 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education orgin within artiodactyls, and further to the THE UNIVERSAL COMMON fact that artiodactyls (cows, deer, antelopes, ANCESTOR pigs, etc.) are not more closely related to peris- inally, Wells cites the “failure” of molec- sodactyls (horses, rhinos, and the tapir) than ular phylogeny to clarify the position of they are to whales (Novacek, 1992, 2001). Fthe Universal Common Ancestor as Wells makes this statement smugly, as if to proof that there is no common ancestry for any suggest that everyone should think that this of life. Here, Wells mixes up the different sounds silly. Unfortunately, it is Wells’s criti- scales of descent in order to tangle the reader cism that is silly. in a thicket of phylogenetic branches. He is

Figure 5. The traditional view of phylogenetic relationships for the three “domains” of life com- pared to Woese’s view. Note that the only difference lies in whether there is a single “root” at the base of the tree. Regardless, eukaryotes, archaeans, and bacteria all share a common ancestor on both, although Woese does posit a greater degree of lateral trasfer for single-celled organisms. 16 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education attacking the notion that life originated with pods, or angiosperms (Figure 5). That is still a one population, and that all life can trace its lot of evolution that Wells’s inaccurate attack ancestry back to that population, the Universal on the idea of a UCA does nothing to dispel. Common Ancestor (UCA). The problem has WHAT THE TEXTBOOKS SAY been that it is hard to determine relationships when there is nothing to compare to. How do he concept of common ancestry is at the you compare “not life” to “life”? We have no core of evolution. The very idea that fossils of the earliest forms of life, and the high Tdifferent species arise from previous degree of genetic change that has occurred in forms via descent implies that all living things the 3.8 billion years since the early stages of share a common ancestral population at some life make it nearly impossible to reconstruct point in their history. This concept is support- the “original” genetic code. This does not ed by the fossil record, which shows a history invalidate the concept of common ancestry; it of lineages changing through time. Because just makes it difficult and potentially impossi- evolution is the basis for biology, it would be ble to untangle the lineages. And this does not surprising if any textbook teaching contempo- mean that there is not one real lineage: the rary biology would portray common descent inability to determine the actual arrangement other than matter-of-factly. of “domains” at the base of the tree or to char- Textbooks treat the concept of common acterize the UCA does not make the UCA any descent in basically the same way as do scien- less real than the inability to characterize light tists; they accept common ancestry of living unambiguously as either a wave or a particle things as a starting point, and proceed from makes light unreal. there. Phylogenies thus appear in many places Some authors (e.g., Woese, 1998) go further in a text, which makes it very hard to evaluate and suggest that there is no “UCA”; rather, exactly how textbooks “misrepresent” biologi- they suggest, life arose in a soup of competing cal evolution using trees. Most texts show a genomes. These genomes were constantly phylogeny in chapters discussing systematics exchanging and mixing, and thus cellular life and taxonomy. In this section there is usually a may have arisen multiple times. Wells misrep- tree of “kingdom” or “domain” relationships, resents the statements of those scientists to which may be what Wells considers a tree make it look as if they are questioning the showing “universal” common ancestry; unfor- entirety of common ancestry, when what they tunately, his discussion is too vague for a read- question is just the idea of a single common er to be sure whether that is what he is refer- ancestor at the base of life. Further, when some ring to. Many textbooks show additional, more suggest that we should abandon the search for detailed trees in their discussions of different the UCA, they do not mean that they don’t taxonomic groups. In terms of textbook pre- think it existed. They mean only that it may be sentations, then, there is no single “Darwin’s a waste of time to try to find it given the cur- tree of life” presented in some iconic state, but rent technology and methods at our disposal. many various phylogenies shown in the appro- Regardless of the status of a UCA, which is at priate sections of most books. Textbooks also the base of the tree of life, the entire debate has present trees in the chapters on processes and nothing to do with the branches of the tree — mechanisms of evolution, in the “Origin of the shared descent of eukaryotes, of animals, life” or “History of life” chapters, and in chap- or common descent among vertebrates, arthro- ters dealing with individual taxonomic groups. 17 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 6. Evaluation of Wells’s grading of Textbook Icon #2, “Darwin’s Tree of Life.” Parenthetical notations indicate the number of phylogenetic trees shown in the book.

18 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education This is because phylogenetic trees are not part “explosion.” These discussions usually men- of the “evidence for evolution,” but rather tion that it was a “rapid” origin of animal graphical representations of the history, groups. Does Wells actually require that the genealogy, and taxonomy of life. No textbook book explicitly mention the “Cambrian misrepresents the methods that are used to Explosion” by name? If so, it should have been construct trees or the trees themselves, specified in the criteria. Or is it that it he only although some trees contain out-of-date rela- looked for “Cambrian Explosion” in the noto- tionships and occasionally incorrect identifica- riously spotty indexes of the textbooks? A tions of organisms pictured in them. When reevaluation suggests that five of the books to textbooks cover the Cambrian period, the rapid which he gives an F should receive, even by appearance of many body plans is discussed his criteria, a D. Finally, one text (Miller and not as a “paradox” for evolutionary theory, but Levine’s) even mentions the confusing nature as an interesting event in the history of life — of the basal divergence of life caused by later- which is how paleontologists and evolutionary al transfer, yet this discussion can receive no biologists consider it. credit in the grading. This is because although WELLS’SEVALUATION Wells considers the “phylogenetic thicket” to be extremely important to reject universal verall, Wells’s grading system for this common ancestry, he apparently does not con- “icon” is so nebulous that it is hard to sider it important enough to account for it in Ofigure out exactly how he evaluated his grading scheme. All of this calls into ques- the textbooks at all. The “Universal Common tion how well Wells actually reviewed the texts Ancestor” is far different from the “Cambrian he graded as well as whether his grades have Explosion.” These deal with very different any utility at all. places in the “tree of life” as well as very dif- ferent issues in evolution. Wells’s grades seem WHY WE SHOULD CONTINUE TO largely based on presentations of “common TEACH COMMON DESCENT ancestry.” For example, according to Wells, if here is no reason for textbooks to sig- the textbook treats common ancestry as “fact,” nificantly alter their presentations of then it can do no better than a D. In order to get Tcommon descent or phylogenetic trees. a C or better, a book must also discuss the As long as biological evolution is the paradigm “top-down” nature of the Cambrian explosion of biology, common descent should be taught. as a “problem” for evolution; if a book only All living organisms that reproduce have off- mentions the Cambrian Explosion, it gets a D. spring that appear similar to, but not exactly Here Wells does not even apply his grading like, their parents. We can observe descent scheme consistently (Figure 6). For example, with modification every day, and like Darwin, Wells chastises textbooks (Miller and Levine’s we can confidently extrapolate that it has gone in particular) for not discussing the Cambrian on throughout the history of life. Through this Explosion, yet most of the textbooks he process, small differences would accumulate reviews actually mention it (Figure 6) and to larger differences and result in the evolution Miller and Levine devote an entire page (p. of diversity that we see today and throughout 601) to it. Many of the reviewed textbooks dis- the history of life. cuss the Cambrian period in the history of life The concept of descent allows us to make sections, but do not specifically call it an testable predictions about the fossil record and 19 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education the genetics of organisms. For example, we Budd, G. E. 1998. Arthropod body-plan evolution in the predict that all animals sharing a common Cambrian with an example from anomalocarid muscle. ancestor would have a similar genetic code, Lethaia 31:197–210. use the same cellular processes, and so on; Budd, G. E., 1999. Does evolution in body patterning genes drive morphological change –– or vice versa? these predictions are confirmed by biochem- Bioessays 21:326–332. istry and genetics. In terms of fossils, we Budd, G. E. and S. Jensen. 2000. A critical reappraisal would expect to see animals with transitional of the fossil record of the bilaterian phyla. Biological morphologies in the past, as well as animals Reviews 75:253–295. that appear similar, but not identical, to those Chen, J.-Y., D.-Y. Huang, and C.-W. Li. 1999. An early living today. We also predict that these organ- Cambrian craniate-like chordate. Nature 402:518–522. isms, both past and present, can be arranged Conway Morris, S. 2000. The Cambrian “explosion”: into a branching hierarchy of forms, which slow fuse or megatonnage? Proceedings of the National appears much like a genealogy. This is what Academy of Science 97:4429–4439. the biological community considers science; DeBraga, M., and O. Rieppel. 1997. Reptile phylogeny and the interrelationships of turtles. Zoological Journal this is what we should teach. of the Linnean Society 120:281–354. HOW TEXTBOOKS COULD IMPROVE Eernisse, and A. G. Kluge. 1993. Taxonomic congru- THEIR PRESENTATIONS ence versus total evidence, and amniote phylogeny OF PHYLOGENY inferred from fossils, molecules, and morphology. Molecular Biology and Evolution 10:1170–1195. here is always room for improvement in Fedonkin, M. A., 1994. Vendian body fossils and trace the presentations of the concept of com- fossils. In S. Bengston, ed. Early Life on Earth. Nobel Tmon descent. Textbooks could improve Symposium no. 84. Columbia University Press, New by updating the phylogenies, many of which York. p. 370–388. are now out-of-date. They should also remove Fedonkin, M. A., and B. M. Waggoner. 1997. The Late discussions of “phenetics” (an outdated form Precambrian fossil Kimberella is a mollusc-like bilater- of phylogenetic reconstruction and classifica- ian organism. Nature 388:868–871. tion) from the phylogenetic reconstruction sec- Felsenstein, J. 1978. Cases in which parsimony or com- patibility methods will be positively misleading. tions, and expand discussions of cladistics and Systematic Zoology 27:401–410. more modern descent-based taxonomies. Gehling, J. G., and J. K. Rigby. 1996. Long expected Finally, textbooks should make a clear distinc- sponges from the Neoproterozoic Ediacara fauna of tion between molecular clocks and genetic South Australia. Journal of Paleontology 70:185–195. phylogenies, something many fail to do clear- Gauthier, J. A., 1994. The diversification of the ly. However, to make textbooks conform to amniotes. In D. R. Prothero and R. M. Schoch eds. Wells’s criteria would be to misrepresent the Major Features of Vertebrate Evolution. Paleontological entire life sciences and to deprive students of Society Short Courses in Paleontology 7:129–159. pedagogically useful visual representations of Gauthier, J. A., A. G. Kluge, and T. Rowe. 1988. Amniote phylogeny and the importance of fossils. the unity of life. Cladistics 4:105–209. Huelsenbeck, J. P., and D. M. Hillis, 1993. Success of References phylogenetic methods in the four-taxon case. Systematic Benton, M. J., M. A. Wills, and R. Hitchin. 2000. Biology 42:247–264. Quality of the fossil record through time. Nature Knoll, A. H., and S. B. Carroll. 1999. Early animal evo- 403:534–537. lution: emerging views from comparative biology and geology. Science 284:2129–2137. 20 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Lee, M. S. Y., 1993. The origin of the turtle body Woese, C. R. 1998. The universal common ancestor. plan:bridging a famous morphological gap. Science Proceedings of the National Academy of Science 261:1716–1720. 95:6854–6859. Lee, M. S. Y. 1997. Pareiasaur phylogeny and the origin Xiao, S., Y. Zhang, and A. H. Knoll. 1998. Three-dimen- of turtles. Zoological Journal of the Linnean Society sional preservation of algae and animal embryos in a 120:197–280. Neoproterozoic phosphorite. Nature 391:553–558. Martin, M. W., D. V. Grazhdankin, S. A. Bowring, D. A. Zardoya, R. and A. Meyer. 1998. Complete mitochondr- D. Evans, M. A. Fedonkin, and J. L. Kirschvink. 2000. ial genome suggests diapsid affinities of turtles. Age of Neoproterozoic Bilatarian body and trace fossils, Proceedings of the National Academy of Science White Sea, Russia: implications for metazoan evolution. 95:14226–14231. Science 288:841–845. Zardoya, R., and A. Meyer. 2001. The evolutionary Novacek, M. J. 1992. Mammalian phylogeny: shaking position of turtles revised. Naturwissenschaften 88:193– the tree. Nature 356:121–125. 200. Novacek, M. J. 1994. Morphological and molecular inroads to phylogeny. In L. Grande and O. Rieppel, eds. Interpreting the Hierarchy of Nature: from Systematic Patterns to Evolutionary Process Theories. Academic Press, New York, p. 85–132. Novacek, M. J. 2001. Mammalian phylogeny: Genes and supertrees. Current Biology 11:R573–575. Patterson, C., D. M. Williams, and C. J. Humphries. 1993. Congruence between molecular and morphologi- cal phylogenies. Annual Review of Ecology and Systematics 24:153–188. Rieppel, O. 2000. Turtles as diapsid reptiles. Zoologica Scripta 29:199–212. Rieppel, O., and M. deBraga. 1996. Turtles as diapsid reptiles. Nature 384:453–455. Rieppel, O., and R. R. Reiz. 1999. The origin and early evolution of turtles. Annual Review of Ecology and Systematics 30:1–22. Seilacher, A. 1994. Early multicellular life: Late Proterozoic fossils and the Cambrian explosion. In S. Bengston, ed. Early Life on Earth. Nobel Symposium no. 84. Columbia University Press, New York. p. 389–400. Shu, D.-G., H.-L. Luo, S. Conway Morris, X.-L. Zhang, S.-X. Hu, L. Chen, J. Han, M. Zhu, Y. Li, and L.-Z. Chen. 1999. Lower Cambrian vertebrates from south China. Nature 402:42–46. Valentine, J. W., D. Jablonski, and D. H. Erwin. 1999. Fossils, molecules and embryos: new perspectives on the Cambrian explosion. Development 126:851–859. Wells, J. 2000. Icons of evolution: science or myth?: why much of what we teach about evolution is wrong. Regnery, Washington DC, 338p.

21 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education HOMOLOGY ...BACK TO THE FUTURE efore 1859, the year Darwin published HOMOLOGY the Origin of Species, homology was omology is a specific explanation of Bdefined as similarity of structure and similarity of form seen in the biologi- position, and distinguished (although inconsis- Hcal world. Similarities can often be tently) from “analogy,” which was defined as explained by common descent; features are similarity of function but not necessarily of considered homologous if they are shown to be structure and position. An example of homolo- inherited from a common ancestor. For exam- gy and analogy are the wings of birds and bats. ple, although the arms of four-limbed verte- The arms of birds and bats would be consid- brates externally appear quite different, all ered homologs because they have the same have the same basic underlying skeletal and structure and position in both animals. Their muscular pattern. Such shared patterns are best wings, however, are analogs. Both wings have explained by the inference that they were the same function (flight), yet the bird’s wing inherited from a common ancestor that also is made of feathers, and the bat’s is made of had this pattern. Proposed homologies are skin. They are different structures. evaluated using comparative anatomy, genet- The pre-Darwinian basis for similarity was ics, development, and behavior. the idea of an “archetype.” The archetype, WELLS RIDES THE HOMOLOGY however, was never clearly defined. The idea MERRY-GO-ROUND... belongs to a morphological theory that came ells claims that homology is used in from the German transcendentalist philoso- a circular fashion by biologists phers of the late 1700s and early 1800s. It was because textbooks define homology largely out of fashion by the 1840s, but W Richard Owen, who codified this distinction, as similarity inherited from a common ances- tor, and then state that homology is evidence was dedicated to a philosophy of transcenden- for common ancestry. Wells is correct: this tal causes, as many historians of science have simplified reading of homology is indeed cir- noted (e.g. Russell, 1916; Desmond, 1982; cular. But Wells oversimplifies a complex sys- Rupke, 1993; Padian, 1995a, 1997). tem into absurdity instead of trying to explain Yet the pattern of the biological world more it properly. Wells, like a few biologists and many resembles a genealogy than a gallery of cook- textbooks, makes the classic error of confusing ie-cutter “archetypes.” Darwin accounted for the definition of homology with the diagnosis the similarities in structure and position among of a homologous structure, the biological basis very different animals as being the result of of homology with a procedure for discovering natural selection working on shared ancestral homology. In his discussion, he confuses not patterns. The concept of homology shifted only the nature of the concept but also its his- from reflecting a vague “archetype” to reflect- tory; the result is a discussion that would con- ing descent with modification. fuse anyone. What is truly important here is Today, biologists still diagnose homologous not whether textbooks describe homology cir- structures by first searching for structures of cularly, but whether homology is used circu- similar form and position, just as pre- larly in biology. When homology is properly Darwinian biologists did. (They also search for understood and applied, it is not circular at all. genetic, histological, developmental, and 22 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education behavioral similarities.) However, in our post- animal groups is tested against the pattern it Darwin period, biologists define a homologous makes with these as well as other groups. structure as an anatomical, developmental, Sometimes, though not always, the pattern behavioral, or genetic feature shared between reflects a genealogical relationship among the two different organisms because they inherited organisms — at which point the inference of it from a common ancestor. Because not all common ancestry is made. Today, the testing features that are similar in two organisms are process is carried out by a method called “phy- necessarily inherited from a common ancestor, logenetic systematics” or “cladistics,” which and not all features inherited from a common can be done without assuming an evolutionary ancestor are similar, it is necessary to test relationship among the groups — but descent structures before they can be declared homol- with modification is the best explanation for ogous. To answer the question, “could this fea- the patterns the comparisons of features it ture in these groups be inherited from a com- reveals. mon ancestor?” scientists compare the feature Evolution and homology are closely related across many groups, looking for patterns of concepts but they are not circular: homology form, function, development, biochemistry, of a structure is diagnosed and tested by out- and presence and absence. Many features are side elements: structure, position, etc., and tested simultaneously against genealogy whether or not the pattern of distribution of the through a process that Kluge (1997; see also trait is genealogical. If the pattern of relation- Kluge, 1998, 1999 for discussions of inde- ships looks like a genealogy, it would be per- pendent homology tests) termed testing “mul- verse to deny that the trait reflects common tiple ad hoc hypotheses of homology.” ancestry or that an evolutionary relationship If, considering all the available evidence, exists between the groups. Similarly, the close- the distribution of characteristics across many ness of the relationship between two groups of different groups resembles a genealogical pat- organisms is determined by the extent of tern, it is very likely that the feature reflects homologous features; the more homologous common ancestry. Future tests based on more features two organisms share, the more recent features and more groups could change those their common ancestor. Contrary to Wells’s assessments, however — which is normal in contention, neither the definition nor the appli- the building of scientific understanding. cation of homology to biology is circular. Nevertheless, when a very large amount of As mentioned, new evidence from various information from several different areas fields of biology has expanded our understand- (anatomy, biochemistry, genetics, etc.) indi- ing of homology beyond just anatomical struc- cates that a set of organisms is genealogically tures. Anatomical homologies, behavioral related, then scientists feel confident in declar- homologies, developmental homologies, and ing the features that they share are homolo- genetic homologies can be independently gous. Finally, while judgments of homology diagnosed and tested. are in principle revisable, there are many cases Behavioral homology recognizes features of in which there is no realistic expectation that animal behavior that can be traced to common they will be overturned. ancestry. For example, consider the nesting So Wells is wrong when he says that homol- practices of birds and crocodilians. Both of ogy assumes common ancestry. Whether a fea- these groups share the behaviors of nest-build- ture reflects common ancestry of two or more ing, parental care of young, and “singing” to 23 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education defend territory and attract mates. Most people of the eye. Biologists hypothesize that the gene know birds do these things, but fewer know is inherited from a common ancestor not only that their cousins the alligators and crocodiles because of its biochemical similarity but also do these things as well. They inherited these because of its distribution in numerous taxa. behaviors from a shared ancestor. Because of However, the actual eyes that the gene forms homology, we infer these behaviors for their are not a result of common ancestry — their extinct ancestors as well; thus it came as no shared ancestor most likely lacked eyes, surprise when fossils of many non-avian although it may have had light-sensing ability. dinosaurs were found nesting with their young The eyes of flies, mice, and many other crea- (Horner and Makela, 1979; Horner, 1982; tures are of different structure and position and Clark et al., 1999). are not historically continuous, yet the Pax6 Developmental homologies are features in gene is historically continuous and responsible the developmental programs of organisms. An for them all. This homologous gene functions example of this is the “pharyngeal pouches” as a “switch” that triggers the development of that nearly all vertebrates acquire to some light-sensing organs (Gilbert, 2000), but the degree during their development, but which “downstream” genes that they trigger are no become very different structures in the adults. longer the same: they govern different devel- For example, the embryological pharyngeal opmental programs and thus build structurally pouches of jawless chordates (e.g., different eyes in flies, mice, and other organ- Amphioxus, hagfishes, and lampreys) develop isms. The relatively new field of evolutionary into pharyngeal arches and slits, which support developmental biology (evo-devo for short) the gill structure and allow water to exit the deals with these processes. The discoveries pharynx after passing over the gills. In jawed made in just the last 10 years in this field have vertebrates, such as sharks and fish, the pha- greatly increased our understanding of homol- ryngeal pouches develop into gill supports and ogy, and have made the picture more complex. portions of the jaw skeleton. In land verte- Wells nearly ignores this important new field brates (tetrapods), these arches and pouches in his discussion, a surprising omission for one develop into jaw skeleton and musculature, but whose background includes a degree in biolo- other pouches/arches, which in gill bearing gy. vertebrates developed into gill structure, now develop into ear bones and cavities, and thy- HOMOLOGY, EVOLUTION, AND roid and tracheal cartilages (Gilbert, 2000). THE NATURE OF SCIENCE The evolution of the different adult pharyngeal ome formulations of the concept of morphologies of vertebrates are the results of homology appear to be circular, but as alterations of these embryonic structures and Sdiscussed above, because there is an their components through the developmental external referent (the pattern that characteris- program (Graham, 2001). tics take across groups) that serves as an inde- Today we also recognize genetic homolo- pendent test, the concept, properly defined and gies. There are similar genes that control the understood, is not. Wells’s claim that homolo- development of non-homologous features. For gy is circular reveals a mistaken idea of how example, there is a gene, named “Pax6,” pos- science works. In science, ideas frequently are sessed by fruit flies, mice, and many other formulated by moving back and forth between organisms, which influences the development data and theory, and scientists regularly distin- 24 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education guish between the definition of a concept and homologies to reconstruct relationships the evidence used to diagnose and test it. (Figure 7). All textbooks include diagrams of Homology is in fact no more circular than the forelimbs of various vertebrates, and all the methods used in geology to determine but one color-code homologous elements for paleogeography and plate tectonics. For exam- easy identification. Guttman includes a second ple, in the 1920s, Alfred Wegener used the figure showing homologous bones in a number shape of the continents, the correlation of rock of tetrapods and one fish skull, clearly illus- strata, the correlation of fossil organisms, and trating how skulls have been reshaped. the position of glacial striations as evidence for Futuyma, Guttman, and Campbell, et al. his proposal that the continents were once include the best discussions and illustrations of joined in one supercontinent and have subse- homology, but nevertheless earn a D from quently “drifted” to their current locations. Wells. Today, geologists can estimate where a certain Most textbooks include discussions of anal- section of a continent used to be by looking at ogy and vestigial structures along with discus- polar wander, paleomagnetism, glacial stria- sions of homology. Analogous features are fea- tions, correlation of strata and fossils, and tures with similar functions (but not necessari- shape. Is this any more circular than the rea- ly similar structures) that are not inherited soning for homology? No. Evidence was used from a common ancestor but evolved conver- to infer that continents had moved, and then gently, whereas vestigial features are remnant the concept of plate tectonics was applied to structures that have been retained from previ- different data to determine the positions of ous forms. Wells notably leaves out any men- continents at different times. The analogy to tion of analogous features or vestigial struc- plate tectonics is also relevant to Wells’s impli- tures from his evaluation (such as the limb gir- cation (Wells, 2000:77) that we don’t fully dles of snakes or the limb girdles of whales understand the mechanisms of homology: the cited by most textbooks). mechanism of sea floor spreading may not yet WELLS’S TEXTBOOK EVALUATION be fully understood, but the continents still move. ccording to Wells, textbooks should explain that homologies are similari- WHAT THE TEXTBOOKS SAY Aties of structure and function due not he presentations of homology in the to common ancestry but to a common “arche- textbooks reviewed by Wells differ only type” or basic plan on which all forms were Tin the lengths of their discussions. based (Wells is remarkably cagey as to what he Overall, textbooks give homology (usually means by “archetype”). When Wells proposes including discussions of analogy and vestigial that textbooks revert to a pre-Darwinian view features) 2–10 paragraphs (Figure 7). Because of homology, he doesn’t explain what that the shorter introductory textbooks have little would mean for biology or biology teaching. space to devote to the complexities of how He doesn’t explain that it would replace a homology is defined, diagnosed, and applied, testable model (descent) with a non-existent, their explanations verge on the circular. The untestable, transcendentalist construct. Wells longer upper-level textbooks make a clearer is vague because he merely wants to advance distinction between the explanation for homol- his position and the archetype is consistent ogy (common ancestry) and using sets of with some notion of special creation, as 25 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 7. Examination of grades applied by Wells for Icon #3. Parentheticals refer to addition- al coverage. Plus (+) numbers reflect secondary treatments of homology and convergence in phylogenetic recontstruction sections of the text. favored by proponents of “intelligent design” concept of an “archetype.” Further, they creationism and their allies. should state that an “archetype” could mean Wells’s grades (he gives only Ds and Fs) many things, not just common ancestry. He appear to correlate with the length of the text- also wants textbooks to state — inaccurately book’s coverage (Figure 7). For example, all — that mechanisms such as genetics and books given Ds devote well over 200 words to developmental programs do not account for the discussion of homology, whereas the three homology. Finally, he wants textbooks to state books given an F devote fewer than 200 that the concept of homology is “controver- words. This is because the difference between sial.” This scheme is rigged for failure because a D and an F for Wells is whether the book contemporary biology does not consider defines homology “circularly.” Therefore, the homology to be either controversial or based ability to treat homology “well” (meaning a D) on archetypes. There is certainly no reason to depends largely on how much space is devoted accept these grading criteria. to the discussion of it. Wells does, however, allow the book to have a picture. In order to receive a B or better, textbooks must define homology as similarity of structure and posi- tion and state that homology is based on the 26 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

WHY WE SHOULD STILL TEACH THAT would help to remove the appearance of circu- HOMOLOGY IS A RESULT larity caused by oversimplified descriptions. OF COMMON ANCESTRY Describing how homology is used as a tool to s our current knowledge of biology discover evolutionary relationships would suggests, there is no reason to doubt make it a much more pedagogically useful Athe fact that the patterns of structures, concept for students because it would show behavior, genes, and developmental programs them how evolutionary biologists use anatom- fit best with the hypothesis that all organisms ical observations to discover evolutionary rela- share common ancestors. Many of the similar- tionships. Finally, adding the notions of multi- ities among these widely divergent groups are ple layers of homology from genetics and a result of that ancestry. The questions current- developmental biology would better show stu- ly being debated in biology are not whether dents just how different lines of evidence con- homology is real, but rather what structures are verge to support homologies and phylogenies. homologous and how we may best determine Textbooks should not follow Wells’s sugges- homology (because our diagnostic approaches tion to say that homology is merely similarity are fallible). This type of discussion of relia- in structure and position, nor should they state bility of methodology is typical for science in that there are “other” reasons for homologies all fields, not just biology. Descent is the basis beyond inheritance from a common ancestor. for homology; similar genes, acting through To revert to Wells’s 19th-century notion of development, convey homology between gen- homology would leave students unprepared to erations. Genes build structures through their participate in 21st-century science. interactions in the developmental program. Therefore genes, development, and similarity References of structure and position are discovery proce- dures for homology; they help biologists to Clark, J. M., M. A. Norell, and L. M. Chiappe. 1999. An determine evolutionary relationships. This fits oviraptorid skeleton from the Late Cretaceous of Ukhaa the patterns and processes we observe in the Tolgod, Mongolia, preserved in an avianlike brooding natural world; this is what we should teach. position over an oviraptorid nest. American Museum Novitates 3265:1–36. OW TEXTBOOKS COULD IMPROVE H Desmond, A. 1982. Archetypes and Ancestors: THEIR DISCUSSIONS OF HOMOLOGY Palaeontology in Victorian London 1850-1875, Blond he biggest flaw in textbook descriptions & Briggs, London 287p. of homology is that they, like Wells, Gilbert, S. F. 2000. Developmental Biology, Sixth Ttend to confuse the definition of homol- Edition. Sinauer Associates, Sunderland 749p. ogy with the diagnosis of homologous fea- Graham, A. 2001. The development and evolution of tures. Textbooks need to state explicitly that pharyngeal arches. Journal of Anatomy 199:133–141. homologies are similarities seen in the biolog- Horner, J. R. 1982. Evidence of colonial nesting and ical world that are best explained as being due ‘site fidelity’ among ornithischian dinosaurs. Nature 297: 675–676. to common descent. They should then explain Horner, J. R. and Makela, R. 1979. Nest of juveniles how homologous structures are diagnosed by provides evidence of family structure among dinosaurs. their similar structure and position, biochemi- Nature 282: 296–298. cal basis, developmental path, and so on. A Kluge, A. G. 1997. Testability and the refutation and cor- more detailed and lengthened discussion roboration of cladistic hypotheses. Cladistics 13:81–96. 27 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Kluge, A. G. 1998. Total evidence or taxonomic con- gruence: cladistics or consensus classification. Cladistics 14:151–158. Kluge, A. G. 1999. The science of phylogenetic system- atics: explanation, prediction, and test. Cladistics 15:429–436. Padian, K. 1995a. Pterosaurs and typology: archetypal physiology in the Owen-Seeley dispute of 1870. In W.A.S. Sarjeant, ed. Vertebrate Fossils and the Evolution of Scientific Concepts. Gordon and Breach, Yverdon, Switzerland p. 285–298. Padian, K. 1997. The rehabilitation of Sir Richard Owen. BioScience 47: 446–453. Rupke, N. A. 1993. Richard Owen’s vertebrate arche- type. Isis 84:231–254. Russell, E. S. 1916. Form and function: a contribution to the history of animal morphology. John Murray, London. Reprint of 1982, University of Chicago Press, Chicago 383p. Wells, J. 2000. Icons of evolution: science or myth?: why much of what we teach about evolution is wrong. Regnery, Washington DC, 338p.

28 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education HAECKEL’S EMBRYOS to be central to our understanding of evolution. Comparative embryology shows how different ERNST HAECKEL AND adult structures of many animals have the COMPARATIVE EMBRYOLOGY same embryonic precursors. These shared rnst Haeckel (1834–1919) is both a hero developmental features suggest that many ani- and a villain in the biological communi- mals have ancestors in common. Further com- Ety. He was a prominent figure in the late parative embryology shows that closely relat- nineteenth-century comparative anatomy com- ed animals show a unity of developmental pat- munity and is famous for his phylogenetic tern, particularly in earlier stages, and have trees, anatomical illustrations, support for evo- more developmental features in common than lution, and strong personality. He is perhaps as do more distantly related organisms. The fact well known, and considerably misunderstood, that certain incipient structures such as pha- for his studies in embryology and his dictum ryngeal pouches or arches exist in all verte- that “ontogeny recapitulates phylogeny,” brate embryos yet develop into very different called the Biogenetic Law. Haeckel espoused adult structures suggests that they all share a the view that evolution generally proceeds by common ancestor whose embryo had pharyn- placing each innovation on top of a previous geal pouches (at least at some stage in devel- opment). In this way, developmental similari- one, like adding layers on a cake. Therefore, ties that are inherited from a common ancestor the embryo of an “advanced” organism should are homologous, just like the patterns of bones pass through (“recapitulate”) the adult stages in adult limbs. of more “primitive” forms as it develops. However, repeated observations of develop- DEVELOPING AN ARGUMENT ment by other workers (e.g., Wilhelm His, ells’s entire chapter on embryology Walter Garstang, Wilhelm Roux, Adam amounts to little more than a mis- Sedgwick, Gavin de Beer, and others; see Wreading of Darwin, Haeckel, and Gilbert, 1991, or Gould, 1977 for a detailed others, combined with a general failure to history) clearly showed that embryos do not go acknowledge recent work on Haeckel and his through adult stages of lower forms; rather, embryos by Gould, Richardson, and others. In they share many common features in develop- it, he conflates ideas in history of developmen- ment. No biologist has accepted the biogenetic tal biology with ideas of contemporary devel- law for many decades and it may have been a opmental biology. He also fails to recognize caricature of Haeckel’s actual views anyway. close to 60 years of work in developmental Much of Haeckel’s developmental work is biology and thus completely omits any discus- now considered invalid, and some historians of sion of the real developmental evidence for science have provided reasonable evidence to evolution. It almost seems that Wells’s goal is suggest that he manipulated his drawings to fit to discredit the entire field of comparative his preconceived views about development embryology by proxy, employing a bait-and- and evolution. Haeckel’s views about the pro- switch between Haeckel and Darwin. Wells’s gressive nature of evolution are no longer ploy is reminiscent of a child’s false logic accepted. proof. It goes like this: Darwin relied on Regardless of Haeckel’s accuracy or precon- Haeckel, Haeckel was a fraud, therefore ceptions, comparative embryology continues Darwin is a fraud. 29 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education The charge that Ernst Haeckel intentionally go through the adult stages of their ancestors “faked” his drawings is irrelevant. Regardless — with the idea that shared features of of his intent, the drawings that Haeckel made embryos give insight into their phylogenetic are incorrect, especially in what he labeled as relationships. Failing to distinguish these the “first stage.” But it really does not matter allows Wells to avoid dealing with the actual what Haeckel thought or whether his drawings evidence for shared developmental features in are accurate: modern comparative embryology various embryos and to dismiss the entire field does not stand or fall on the accuracy of as based on an outdated and outright refuted Haeckel any more than modern physics stands claim, one that embryologists know to be false or falls on the accuracy of Kepler or Newton. but cling to anyway because of an ideological Historically, Wells actively ignores the accu- commitment to evolution. Wells should know rate work of many of Haeckel’s predecessors better, as the holder of a Ph.D. in cell and and contemporaries (such as William and developmental biology. Jeffrey Parker, Hans Gadow, Hans Selenka, REWRITING HISTORY FOR Heinrich Rathke, Virgil Leighton, Hugo THE GREATER GLORY OF Schauinsland, and Alfred Voeltzkow, to name THE REV. MOON a few). Haeckel and von Baer were not the only embryologists in nineteenth-century sci- n the introduction to Icons, Wells states ence, but you wouldn’t know that from reading that he first became aware of the problems Wells. Worse, Wells speciously extends his cri- Iin evolutionary theory when he was “fin- tique of Haeckel to the present day. Wells ishing his Ph.D. in cell and developmental implies that textbooks misrepresent the study biology” (Wells, 2000:xi). He claims that he of developmental programs as evidence for knew that the drawings of embryos presented evolution by accusing them of using Haeckel’s in textbooks were false because he was a inaccurate drawings, in effect accusing text- developmental biologist. Shortly thereafter, he books that show any embryos of “mindlessly claims, his observation was confirmed by repeating” Haeckel. The important question is other scientists. Before that seminal event, he whether textbooks, and more importantly says, “I believed almost everything I read in developmental biologists, still rely on my textbooks” (Wells, 2000:xi). This state- Haeckel’s work. The answer is no, but that ment is inconsistent with other claims of doesn’t stop Wells from acting as if they do. Wells’s. According to statements made by Wells sets up a straw man in his bait-and- Wells in a sermon on a Unification Church switch, starting with Darwin’s famous asser- website (http://www.tparents.org/library/unifi- tion that embryology represented the “single cation/talks/wells/DARWIN.htm), he went to strongest class of facts” in favor of his theory. graduate school with the specific intent of Here Wells misrepresents both early embryol- attacking evolution: “Father’s words, my stud- ogy and Darwin’s own words. When quoting ies, and my prayers convinced me that I should both Darwin and other historical figures, he devote my life to destroying Darwinism” and quotes them out of context, leaves out impor- he believed that its weakest point was devel- tant parts of quotes, and even changes the opmental biology. “I was convinced that order of their appearance, all to misrepresent embryology is the Achilles’ heel of their real meaning and intent. Wells also con- Darwinism; one cannot understand how organ- flates “recapitulation” — that is, that embryos isms evolve unless one understands how they 30 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education develop. In 1989, I entered a second Ph.D. pro- ed to developmental ones) to generate classifi- gram, this time in biology, at the University of cations for organisms. Darwin is praising the California at Berkeley. While there, I studied application of his theory by Haeckel. embryology and evolution.” So it was not so Although Darwin did not use Haeckel on much a “revelation” as it was a plan. If Wells embryology, he did use von Baer. Recognizing is so revisionist about his own history, how can Darwin’s use of von Baer, Wells then accuses we trust him with the history of science? Darwin of “misusing” von Baer’s work, twist- ing the data to fit his views. But Darwin does DEVELOPMENTAL ANATOMY, not. Wells claims that von Baer’s embryologi- DARWIN, AND EVOLUTION cal laws are incompatible with Darwin’s con- ells opens the chapter by telling us clusions, but they are not. Von Baer may have what Darwin thought about devel- disagreed with Darwin about his conclusions, Wopment and evolution. Wells uses but his laws do not prohibit development elu- about 5 different quotes from the Origin in an cidating common ancestry. Darwin came to a attempt to show that Darwin was advocating different conclusion from the same body of recapitulation in spite of what the data showed. evidence — this is not “distorting” the evi- To do this, he distorts the history. Wells tries to dence. Darwin was making a general inductive connect Darwin to Haeckel so that he can use argument and searched for data that could test that to dismiss Darwin. Wells says that Darwin the general proposition of common descent; he was not an embyrologist and thus he relied on argued that von Baer’s data could be reinter- Haeckel (Wells, 2000:81). Anyone familiar preted in terms of common ancestry. This was with the history of biology knows that this is no more a “misuse” of von Baer than was impossible. Haeckel did not publish his Alfred Wegener’s reinterpretations of the data Anthropogenie until 1874 (where the much- of geology in light of mobile continents. New maligned embryo drawings first appear), 15 scientific theories always use previous data. Is years after the publication of the Origin. (It Wells implying that evolutionary biology can- should also be noted that the drawings referred not cite any research that predates 1859? Is to by Wells [2000] are not from Haeckel but Wells implying that developmental sequences redrawn from the first edition of such as those illustrated by von Baer and oth- Anthropogenie in a textbook by Romanes ers are not data? [1892; see figure 10a]. In later editions of That Darwin and all modern evolutionists Anthropogenie, Haeckel corrected some of the advocate some form of the “Biogenetic Law” errors of the first edition drawings [Richardson is the central falsehood of this chapter; in fact and Keuck, 2002; personal observation].) the entire “resurrecting recapitulation” section Wells quotes Darwin’s praise of Haeckel in his does nothing but assert this. But Wells fails to sixth and final edition of the Origin in such a explain fully what recapitulation means. There way as to obscure the fact that Darwin lauds are a number of meanings for “recapitulation” Haeckel for his phylogenies, not his embryol- that Wells conflates in order to tar the entire ogy. The quote is not even from the embryolo- field of embryology with a biogenetic brush. gy section of the book; rather it comes from As he says in a footnote, a “plain reading” of the classification section, in the final sentence Darwin shows that Darwin was advocating of which Darwin praises Haeckel for using recapitulation — but just what kind? (1) An homologous features (including but not limit- embryo of an “advanced” form goes through 31 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education all the adult stages of all its ancestors. This is a tion from von Baer. Does “earliest” reflect caricature of Haeckelian recapitulation, which Darwin’s belief, or is he merely reporting von is false, and few scientists ever believed it any- Baer’s? This is important because numerous way. (2) Evolution proceeds as an “add-on” scholars have made the mistake of confusing process so that there is a general progression of Darwin’s reporting of what others thought with embryological stages from “primitive” to his expression of his own views (Padian, “advanced” forms. This more traditional read- 1999). So apparently has Wells. But it really ing of recapitulation is also false and has not does not matter what Darwin thought: just as been accepted for nearly a century. (3) All modern embryology does not rely on Haeckel, closely related organisms go through all the neither does modern evolutionary biology same stages of development and always look slavishly follow Darwin’s beliefs. similar. This is vague: how closely related is It is also important to understand what nine- closely related? How are the stages individuat- teenth-century scientific workers may have ed? But however these questions are answered, meant by the use of “embryo” and “early this reading of recapitulation would generally stage.” For many workers in the nineteenth be agreed to make too sweeping a claim. (4) century, developing organisms weren’t called Some parts of developmental sequences (and embryos until they reached the tailbud (phylo- some specific characters of them) in closely typic) stage. During earlier stages, they were related animals share more specific similarities called “developing ovum” or “developing egg” (in pattern, sequence, position, etc. of develop- (see Barry, 1839, or just about any embryolo- mental features) with each other than with gy work from 1820 to 1900). What this means those of more distantly related animals. That’s is that Haeckel, von Baer, and others, have a basically true. All modern biologists recognize different meaning for “early embryo.” Yet that all stages of development are open to Wells interprets them using modern defini- modification. This is generally the type of tions. “recapitulation” accepted by the post- Wells also criticizes the field of comparative Haeckelian embryologists (such as Frank embryology for the way it chooses its data and Lillie) cited by Wells, as well as by current for its names for embryonic structures. First, embryologists, but Wells treats it derisively as Wells emphasizes the disparity of “earliest” if it were exactly what Haeckel thought. developmental stages, accusing biologists of Finally, a “plain reading” of Darwin shows “choosing” taxa (animals) that look most sim- that he was suggesting something between (2) ilar for illustrations in textbooks and else- and (3); even though he was not an embryolo- where. He criticizes Haeckel for not using ani- gist, he had a more sophisticated notion of mals such as monotremes in his work. But embryology and development than does Wells. developmental sequences for monotremes Wells chides Darwin and nineteenth-century were not available in 1874. Monotreme devel- embryologists for saying that the “earliest” opmental sequences were not known or stages of development are similar when in fact described until 1884 (Caldwell, 1887; Hughes they are not. However, “earliest” is Wells’s and Hall, 1998), and it was the developmental word, not Darwin’s. It does not appear in any features monotremes shared with marsupials of the quotes that Wells uses. Indeed, in the that led Caldwell to conclude that monotremes entire section on embryology in the Origin, the were indeed mammals (Caldwell, 1887). Was word “earliest” only appears once, in a quota- the sample of organisms available to Haeckel 32 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 8. Developmental sequences of various vertebrates shown in phylogenetic context. Note the shared similarities of some closely related taxa, particularly the amniotes (modified from Richardson et al. 1998). biased? Yes, but only in the sense that early sexual maturity, rapid reproduction, ability for embryologists worked with the animals that development to occur in the laboratory, and were available to them. Most specimens of ability to live indoors for several generations “exotic” animals were shipped to researchers (Bolker, 1995). They were not chosen to “sup- by explorers and received in varying states of port evolution” as Wells implies. In fact, the decay (Caldwell, 1887). Most nineteenth-cen- model organism that is the subject of Wells’s tury embryologists loved to describe the devel- own dissertation, Xenopus, was not the origi- opment of any animal they could. And Haeckel nal “model” amphibian. The discovery that was continually updating and adding new Xenopus does not need a breeding season was organisms to his embryonic series as they a boon to embryologists and led to its came available. Contrary to what Wells serendipitous adoption as a model organism implies, there was no attempt to limit the data, (Gurdon and Hopwood, 2000). How Wells and the sample was not “chosen” for any par- knows that “model organisms” were chosen to ticular reason. mislead is unclear, especially given his own Today, embryologists work mainly with use of model organisms later in his chapter. “model organisms,” which were largely cho- Wells doesn’t show developmental sequences sen for practical reasons such as ready avail- for any of the organisms he complains others ability, small body size, large litter size, rapid don’t show. Why not? Because there is no evi- 33 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education dence for his insinuation that developmental Second, they are not all scaled the same. In the biologists treat their data selectively in order to figure showing the neural crest infolding, the hide something. The fact that embryologists turtle and chicken are shown at a large scale, tend to present, at least in textbooks, develop- neglecting the large yolk they sit on, while the mental sequences for which there is good data human is shown as part of the whole develop- does not refute the idea that closely related ing ovum, so that the germinal disc and primi- taxa, should, and do, have more shared simi- tive streak formation are shown differently, larities in developmental programs than more even though it is shared by all amniotes distantly related taxa (Figures 8, 9). Wells tries (Schaunislaund, 1903; Nelson, 1953; Cruz, to support his claim by using a quote by 1997; Schoenwolf, 1997; Figure 9). Also pic- Darwin in which he states that embryos of the tured is a frog embryo, despite its indirect same “class” are most similar in their earliest development, which is very different from that stages. Wells then says that the quote is false, of the other vertebrates pictured. Many of the and cites how the different “classes” of verte- general “differences” in early embryo develop- brates are very different in their “earliest” ment that Wells mentions are a result of organ- stages. This is merely a semantic sleight-of- ization due to the yolk size rather than being hand, a bait-and-switch. Darwin is not talking specific differences in the basic body-plan of about different “classes.” Wells leaves out the embryo (Arendt and Nübler-Jung, 1999). important information, as usual. Embryos do reveal phylogenetic informa- In the figures of embryos (Wells, 2000:95, tion in terms of specific shared features, shared especially stage 4, “gastrulation”), Wells’s early developmental features such as the for- illustrator resorts to a number of graphic tricks mation of a germinal disc and primitive streak in order to make the embryos appear more dif- in all amniotes or the neural crest cells of all ferent than they are. First, the embryos are not vertebrates. The presence, and sequence of shown from the same rotational angles. The development, of eyes, ears, somites, limbs, chicken is shown in a different position than guts, nerve cords, tails, organs, etc. are indi- the other “Haeckel’s first stage” embryos. vidual features that no one would deny are

Figure 9. Embryos of various amniotes shown during somite stage. All amniotes go through the same sequence of development: primitive streak–neural tube–somite formation.

34 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education present in vertebrates and not present in the the letter and spirit of the concepts and authors same way in other animals. These are individ- he presents do little to inspire confidence in ual characters whose developmental features what he says about Haeckel or embryology in are treated as shared features in reconstructing general. Even if Wells were right about vertebrate evolution; these features do not Haeckel’s work and Darwin’s use of it, what always have to be in agreement, and some ani- Haeckel and Darwin thought doesn’t matter; mals can show unusual derived features early embryology has moved beyond them. Wells in development, such as the snake’s tail needs to show a lack of specific similarities to (Figure 8). support his case. Is Wells actually claiming Wells’s treatment of comparative embryolo- that there are no shared features in develop- gy is remarkably limited; for example, he ment at all? That a chicken gets a planula while never discusses invertebrate development. Yet the duck gets a naupius? If so, he needs to there are plenty of shared developmental pat- show it, but Wells never gets to specifics — terns there as well. Despite the very different apparently because the specifics aren’t there. appearance of echinoderm, hemichordate, and Innuendo and accusations of fraud do not cut it chordate embryos, they all share the deuteros- in science. tome condition, in which the first cell opening WHAT TEXTBOOKS SAY becomes the anus, before they diverge to their adult body plans. Or what about the tro- or any textbook to show Haeckel’s chophore larvae of most protostomes and spi- drawings themselves as unqualified ral cleavage shared by annelids, arthropods, Fstatements of developmental anatomy or mollusks (Nielsen, 1995; Fell, 1997)? The to advocate “recapitulation” in a Haeckelian nauplius larvae of crustaceans (Gilbert, 1997) sense would be inexcusable, but none of the or the verliger larvae and development of gas- textbooks reviewed by Wells appear to do so. tropods, which go through flexure, torsion, and Wells gleefully excoriates Futuyma for using degeneration of muscles on one side of the Haeckel’s drawings (Figure 10a), but appar- body, suggestive of their evolutionary history ently in his fit of righteous indignation, he for- (Nielsen, 1995; Collier, 1997)? These are just a few of the specific similarities of the kind that Wells implies do not exist. Similarities in embryonic sequences are data — characters by which we can discover shared similarities among organisms that can be used to recon- struct their relationships. Using such data in phylogeny is not the same as using those char- acters in any “recapitulationist” way. Finally, Wells concludes by attacking prominent biologists such as Gould and Futuyma for supposedly not knowing the truth about Haeckel, saying that this is “a confession of ignorance not likely to inspire confidence in the quality of our biology textbooks” (Wells, Figure 10a. Romanes (1892) embryo draw- 2000:107). Wells’s own misrepresentations of ings reproduced in Futuyma (1998:653). 35 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 10d. Embryos redrawn and some- what corrected in Raven and Johnson (1999:288). accurate drawings (Miller and Levine, Johnson, Biggs, Kapicka and Lundgren; Figure 10b. Romanes (1892) embryos Figures 10f,g,h); some use photos (Campbell, reproduced and placed in historical con- Reese and Mitchell, Mader; Figures 10i,j); text in Guttman (1999:718) only Starr and Taggart (Figure 10c), Raven and Johnson in their development chapter along got to read the text, in which the drawings are with accurate drawings and photos; (Figure discussed in a historical context — stating 10d), and Schraer and Stolze (but redrawn and why Haeckel is wrong — and Futuyma has an corrected; Figure 10e) use what could be con- entire chapter devoted to development and sidered embryos “redrawn” from Haeckel. No evolution. Guttman (Figure 10b) uses them in textbook discusses embryology in any way an explicitly historical context as well. Wells states that books use “Haeckel’s drawings, or redrawn versions of them” (Wells, 2000:255), but this is not true. Figures 10a–j show Haeckel’s drawings compared to the drawings in the textbooks reviewed by Wells. It can be clearly seen that a majority of the drawings are not “redrawn.” Some textbooks show more

Figure 10c. Romanes (1892) embryos Figure 10e. Embryos redrawn and corrected redrawn in Starr and Taggart (1998:317). in Schraer and Stolze (1999:582)

36 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education that could be considered strongly “recapitula- tionist.” In most textbooks, embryology is pre- sented in just one or two paragraphs, making it hard to discuss all the complexities of devel- opment. At a high school level, the aim of the book is to convey some basic concepts of biol- Figure 10h. Embryo drawings in Biggs et al. ogy, not to confuse students with the complex- (1998:433). Identical drawings appear in the ity of a subject. evolution chapter (p.416) of Raven and Johnson (1999).

Figure 10i. Embryo photos in Campbell, Reese, and Mitchell (1999:424).

Figure 10f. Original embryo drawings in Miller and Levine (2000:283).

Figure 10j. Embryo photos in Mader Figure 10g. Original embryo drawings in (1998:298). Johnson (1998:179). 37 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

WELLS’S “WELL-DEVELOPED” more accurate pictures only earns a book a D. GRADING SCHEME In order to earn a C or higher, a book must not he grading scheme employed by Wells use “misleading drawings or photos.” This is designed for failure. This is because amounts to complaining that textbooks should- Wells assumes all drawings to be n’t allow students to be misled by reality! T Wells does not specify what kind of drawings “redrawn” from Haeckel and gives any book with a drawing an F (Figure 11). Wells does or photos would not be misleading. Thus Wells not explain how one would determine whether apparently thinks that all visual presentations they are simply redrawn from Haeckel; in any of embryos are misleading, whether they are case none of the books appear to contain mind- accurate or not. Wasn’t Wells the one com- lessly redrawn figures (Figure 10a–j). Using plaining about selective use of data? He actu-

Figure 11. Wells’s grades for the embryology sections of textbooks.

38 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education ally attacks Mader and Campbell, Reese, and WHY WE SHOULD STILL TEACH Mitchell, for using “misleading photos” COMPARATIVE EMBRYOLOGY because they show embryos of a chick and a espite changes in how we view the role human, which he says “just happen” to have a of developmental programs as reflec- stronger resemblance than would embryos Dtions of evolutionary history, we can from any other “classes” of vertebrate. Wells is still see how the same embryonic structures wrong: a chick embryo at that stage looks develop into different adult structures. We much more like an alligator embryo than a observe the unity of developmental plan in all mammal embryo (comparisons made from vertebrates. This is what we see, and no Nelson, 1953, Schaunisland, 1903, and Reese, amount of wishful thinking on the part of evo- 1915). This is in accordance with the predic- lution detractors can change that. There is no tions of evolutionary theory, because an alliga- reason to let their baseless complaints and tor and a chicken share a more recent ancestor character assassination dissuade biology with each other than they do with a mammal, teachers from presenting the evidence to stu- and thus should have more similar a develop- dents. mental program. Wells also chides Mader for HOW TEXTBOOKS COULD IMPROVE saying that embryos “have many features in THEIR PRESENTATIONS OF common” (Wells, 2000:103–104). Does Wells COMPARATIVE EMBRYOLOGY assert that they have no features in common? If extbooks could largely improve the pre- so, he should document it. Having failed to do sentations of embryology by lengthen- this, Wells merely labels anything he does not ing their discussions of it, and by using like “misleading.” Wells also takes exception T photos rather than cartoonish drawings. They to the colloquial term “gill slits,” which is a could also be more explicit about how embry- commonly used non-technical term for pha- onic precursors develop into different adult ryngeal pouches. Wells implies that by using structures. Finally, adding discussions of Hox this term, biologists and textbooks are saying gene complexes (master developmental con- that all animals’ embryos have gills. This is trol genes) and evolutionary developmental patently false. No textbook reviewed even biology would help bring the books up-to-date implies the presence of gills in embryos. The in their treatment of developmental biology. question is what these structures are and what We are learning more about the evolutionary they become, not what they are called. Using history and underpinnings of developmental the terms “gill slits” automatically results in a programs every day. We are learning how C even if the textbook contains no images, and developmental programs are the source of regardless of its content. Campbell, Reese, and much of the evolutionary novelty that natural Mitchell, and Guttman both contain entire selection shaped. Wells ignores all this. To fol- chapters devoted to developmental biology in low Wells’s advice would arrest the develop- which they do discuss some of the “early stage ment of students’ knowledge. differences” that Wells suggests they do not. They receive no credit for these extensive treatments (Figure 11).

39 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

References tion in theory and in practice. Systematic Biology 48:352–364 Arendt, D. and K. Nübler-Jung. 1999. Rearranging gas- Reese, A. M. 1915. The Alligator and its Allies. G. P. trulation in the name of yolk: evolution of gastrulation Putnam and Sons, New York 358p. in yolk-rich amniote eggs. Mechanisms of Development Richardson, M. K., J. Hanken, M. L. Gooneratne, C. 81:3–22. Pieau, A. Raymond, L. Selwood, and G.M. Wright. Barry, M. 1839. Researches in Embryology. — Second 1997. There is no highly conserved embryonic stage in Series. Philosophical Transactions of the Royal Society the vertebrates: implications for current theories of evo- of London 129:307–380 lution and development. Anatomy and Embryology Bolker, J. A. 1995. Model systems in developmental 196:91–106. biology. BioEssays 17:451–455. Richardson, M. K., J. Hanken, L. Selwood, G. M. Caldwell, M. A. 1887. The embryology of Monotremata Wright, R. J. Richards, and C. Pieau. 1998. Haeckel, and Marsupialia — Part I. Philosophical Transactions embryos, and evolution. Science 280:983–984. of the Royal Society of London, B 178:463–486. Richardson, M. K. and G. Keuck. 2002. Haeckel’s ABC Collier, J. R. 1997. Gastropods, the snails. In S. F. of evolution and development. Biological Reviews, Gilbert and A. M. Raunio, eds. Embryology: construct- 77:495–28. ing the organism. Sinauer and Associates, Sunderland p. Romanes, G. J. 1892. Darwin and After Darwin.Volume 189–217. 1: The Darwinian Theory. Open Court, Chicago. Cruz, Y. P. 1997. Mammals. In S. F. Gilbert and A. M. Schauinsland, H. 1903. Beitrage zur entwicklungs- Raunio, eds. Embryology: constructing the organism. geschichte und anatomie der wirbeltiere. Zoologica her- Sinauer and Associates, Sunderland p. 459–489. ausgeg, C. Chun, Stuttgart. Fell, P. E. 1997. The concept of larvae. In S. F. Gilbert Schoenwolf, G. C. 1997. Reptiles and birds. In S. F. and A. M. Raunio, eds. Embryology: constructing the Gilbert and A. M. Raunio, eds. Embryology: construct- organism. Sinauer and Associates, Sunderland p. 21–28. ing the organism. Sinauer and Associates, Sunderland p. Gilbert, S. J., ed. 1991. A conceptual history of modern 437–458. embryology. Johns Hopkins Press, Baltimore, 266p. Wells, J. 2000. Icons of evolution: science or myth?: Gilbert, S. F. 1997. Arthropods: the crustaceans, spiders why much of what we teach about evolution is wrong. and myriapods. In S. F. Gilbert and A. M. Raunio, eds. Regnery, Washington DC, 338p. Embryology: constructing the organism. Sinauer and Associates, Sunderland p. 237–257. Gould, S. J. 1977. Ontogeny and Phylogeny. The Belknap Press, Cambridge, 501p. Gurdon, J. D., and N. Hopwood. 2000. The introduction of Xenopus laevis into developmental biology: of empire, pregnancy testing and ribosomal genes. International Journal of Developmental Biology 44:43–50. Hughes, R. L., and L. S. Hall. 1998. Early development and embryology of the platypus. Philosophical Transactions of the Royal Society of London, B 353:1101–1114. Nelson, O. E. 1953. Comparative embryology of the vertebrates. Blackiston, New York, 982p. Nielsen, C. 1995. Animal evolution: interrelationships of the living phyla. Oxford University Press, New York, 467p. Padian, K. 1999. Charles Darwin’s views of classifica- 40 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education ARCHAEOPTERYX ple of how fossils are important for showing transitional features of evolution, and how the ARCHAEOPTERYX: THE FOSSIL fossil record is good evidence that evolution ontrary to Wells’s subtitle, has occurred. Archaeopteryx is not a “missing link.” WELLS MISSES MORE CThe term “missing link” is an outdated THAN THE LINKS term that does not accurately reflect the way ells objects to textbook treatments biologists and paleontologists think about fos- of Archaeopteryx as a transitional sils. We prefer not to talk about “missing Wform or as an “ancestor” of birds. links” or “intermediate forms,” but rather Wells wants textbooks to say that intermediate features. Archaeopteryx has fea- Archaeopteryx was not an “ancestor” because tures intermediate between those of living modern birds are not descended from it and birds and ancient reptiles; along with many that its transitional status is “controversial.” other fossils, it preserves ancestral features Wells claims that Archaeopteryx has been while it shows descendant novelties. “quietly shelved” by paleontologists and that Archaeopteryx retains the ancestral “reptilian” the search for a “missing link” between features of a long bony tail, clawed hands, dinosaurs and birds goes hopelessly on “as teeth, and many others. It also has the derived though Archaeopteryx had never been found” “avian” features of feathers and powered (Wells, 2000:138). Paleontologists would find flight. Archaeopteryx, along with other this surprising. By making such claims, Wells dinosaur fossils, shows the evolution of avian exposes the depths of his ignorance of phylo- features and flight. These fossils show that genetic methodology, paleontology, and avian many features thought of as unique to a certain evolution. group of animals were also shared by some of Wells is clearly confused by Archaeopteryx, their ancestors; this helps paleontologists to “transitional forms,” and ancestors. First of all, reconstruct the evolutionary history of living Wells asserts that Archaeopteryx is no longer animals. When many fossils are looked at in considered a transitional form or an “ances- their genealogical context, they blur the lines tor.” Wells is correct, but only in a specialized between the normally recognized taxonomic sense, not appropriate in the context of his groups (most of which were based originally generalized discussion. We cannot — and do only on living forms). Archaeopteryx is fre- not — say for certain that the animal that we quently used for pedagogical purposes because call Archaeopteryx was actually genetically it is easy to recognize its mixture of “bird” and transitional to living birds, or that it was a “reptile” features and because it played an his- direct genetic ancestor of living birds. torical role in helping to cement Darwin’s the- However, in a less strict sense (that appropriate ory (it was discovered 2 years after publication to Wells’s discussion), Archaeopteryx has a of the Origin). Textbook authors like great many transitional features between living Archaeopteryx for these reasons and often birds and Mesozoic dinosaurs: if it was not a illustrate their discussions with pictures of the direct ancestor, it was surely a close collateral Berlin specimen, one of the most beautiful fos- ancestor (see below). sils ever discovered, and remarkably complete. Second, there is no such thing as a “missing Textbooks also use Archaeopteryx as an exam- link,” and paleontologists are not looking for 41 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education them. Paleontologists collect, survey, and rect. Archaeopteryx has no features that would reconstruct past forms of life. Some of these actually disbar it from being a direct ancestor fossil organisms have features that illustrate of living birds. Whether it was a direct ances- the path evolution took to reach the forms we tor of today’s birds or not is irrelevant: see today. We can think of these organisms as Archaeopteryx exhibits unique features of the showing transitional or ancestral features. last ancestor it shared with birds, so, regardless Paleontologists are also not looking for ances- whether it is a lineal ancestor, it still preserves tors, but rather features of ancestors. features that indicate what the last ancestor of Paleontologists distinguish between lineal and Archaeopteryx and birds may have been like. collateral ancestors. Lineal ancestors are those In other words, Archaeopteryx has many fea- that are directly ancestral to living organisms: tures intermediate between those of its your lineal ancestors are your father and moth- dinosaurian ancestors and its avian descen- er, grandfathers and grandmothers, and so on. dants, which is exactly what would be predict- Collateral ancestors are those organisms that ed by evolution. No amount of stridency on share an ancestor with living organisms: your Wells’s part can change that. collateral ancestors are your uncles, great- When paleontologists reconstruct relation- uncles, cousins, second cousins, and so on. ships of living and fossil organisms, they use Paleontologists do not claim to be able to iden- the features of both living and fossil organ- tify lineal ancestors. Without observational or isms. This allows them to reconstruct the fea- genetic evidence, how could you ever know tures of the ancestors and get a pretty good pic- that a fossil organism left any offspring? It is ture of what the ancestors were like. not the ancestry that is important to paleontol- Phylogenetic systematics, commonly called ogists, but rather the ability to reconstruct the “cladistics,” is the method that nearly all biol- features of those ancestors. This is a powerful ogists use to determine relationships, whether and important concept, one completely lost on they work on dinosaurs or dinoflagellates, and Wells. whether they use molecules or morphology. Its To illustrate this powerful approach, let’s simplicity, objectivity, testability, repeatability, say you wanted to know something about your utility, and firm rooting in the principle of own ancestors. If you knew your ancestors descent has led to its near-universal applica- came from a certain small village in France in tion. Contrary to Wells’s characterization, the 1600s, you could return to that village and, cladistics is not a search for “missing links” or even if you can’t locate their graves, you might direct ancestors, but for shared evolutionary find those of many of their contemporaries in features. The basic idea behind cladistics is the churchyard. A collection of artifacts from that when novel features arise, they are passed any of those people would give you a perfect- on to descendents. Therefore, these “derived ly adequate idea of the characteristics, culture, features” should be more informative in recon- possessions, and daily life of your direct structing relationships than those that are pres- ancestors (Padian and Angielczyk, 1999). ent across a larger group. For example, if a Using similar methods for similar reasons, population of animals evolve stripes on their paleontologists try to uncover features of backs and all their descendants continue to ancestors, not the ancestors themselves. sport stripes, then all the members of that Even Wells’s claim that paleontologists do species that have stripes are probably more not think Archaeopteryx is “ancestral” is incor- closely related to each other than they are to 42 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education those without stripes. It is that simple, yet ing” the evidence, stating that the supposed Wells’s discussion of cladistics reveals that he “ancestors” of Archaeopteryx are millions of either does not grasp the method or has no years younger. First of all, none of these more interest in explaining it properly. “recent” avian-like dinosaurs thought to be In the nearly two pages devoted by Wells to closely related to Archaeopteryx (e.g., a discussion of cladistics (Wells, 2000:118– troodontids and dromaeosaurs) are considered 119), he states that cladistics is based on over- “ancestors”; rather, they retain ancestral fea- all similarity. Yet as stated above, cladistics is tures that show us what the ancestors of not based on mere similarity, but instead Archaeopteryx were like. Here again Wells focuses on a special kind of similarity — fea- mistakes lineal for collateral ancestry. Second, tures that are derived, or evolutionary novel- the statement that there are no fossils of these ties. Evolutionary novelties help to show rela- close cousins of Archaeopteryx until “millions tionships and thus are “phylogenetically of years” later is false. Fossils of non-avian informative.” In contrast, similarities that are maniraptor dinosaurs, which are closely relat- not evolutionary novelties are “ancestral” fea- ed to the ancestors of Archaeopteryx, have tures and are not phylogenetically informative. been found in rocks dating to the same age as For example, a derived feature of primates is those in which Archaeopteryx has been found an opposable thumb; this feature is phyloge- (Jensen and Padian, 1989); this discovery was netically informative because it allows us to reported over 10 years ago. Wells apparently group all primates together to the exclusion of has not done his homework very well. other mammals. On the other hand, a five-fin- Despite Wells’s claims to the contrary, gered hand is an ancestral feature and not phy- Archaeopteryx is still an important contributor logenetically informative because we would to our knowledge of transitional features, and not group all animals possessing a five-fin- it clearly shows the dinosaurian ancestry of gered hand together to the exclusion of those birds (Figure 12). To confirm this, all one has that do not. For example, we do not propose to do is peruse any piece of literature on the that all five-fingered mammals are more close- origin of birds. Papers on Archaeopteryx and ly related to each other than they are to three- bird evolution appear in many journals each fingered, two-fingered, or one-fingered mam- year, and there is even an entire journal (called mals. So cladistics is not based on mere simi- Archaeopteryx) devoted to the study of larity. Further, paleontologists who apply Archaeopteryx and its environment. Rather cladistic methods to the problem of avian evo- than consult the vast body of literature on the lution do not think that how flight evolved is origin of birds, Wells appears to base much of “irrelevant”; in fact we specifically use clado- his discussion on two popular works, one tech- grams to inform our models of how flight (and nical — The Mistaken Extinction by Lowell other things) evolved. By rooting our explana- Dingus and Tim Rowe (1998) — and the sec- tions in phylogenies, we can move beyond ond non-technical — Taking Wing, by Pat subjective models, and constrain our hypothe- Shipman (1998). Both are excellent books. ses (e.g., Witmer, 1995; Padian, 1995). These However, during the same period when Wells explanations can also serve as yet another apparently wrote Icons (1998–1999), well over independent test of our phylogenies (Padian, 50 papers were published that in some way 2001a, 2001b). dealt with Archaeopteryx and the dinosaurian Wells then accuses cladistics of “rearrang- origin of birds. A number of these were very 43 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 12. Archaeopteryx (bold) shown in evolutionary context with respect to crocodilians, non- avian dinosaurs, and birds. Some relevant features plotted. 44 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education important (e.g., Britt et al., 1998; Padian and well done and can fool an untrained eye, which Chiappe, 1998; Forster et al., 1998; Ji et al., is more or less what happened with 1998; Burgers and Chiappe 1999; Chiappe et “Archaeoraptor.” The first paleontologists to al. 1999; Clark et al., 1999; Garner et al., 1999; see the specimen were immediately suspicious Norell and Makovicky, 1999; Ostrom et al., because the prevalence of composite speci- 1999; Wagner and Gauthier, 1999; Xu et al., mens was already known, and its distribution 1999), yet Wells cites none of them. of features were not what would be expected in Wells also ignores the many fossil discover- an avian-like dinosaur. We would not expect it ies of feathered non-avian dinosaurs from to have the arms of a primitive bird and the Liaoning, China (see Figure 13), which should legs of a non-avian theropod. Even though a play an important role in any discussion of number of paleontologists were skeptical, avian origins, save for one notable exception. National Geographic went ahead with an arti- In an attempt to discredit the entire field, Wells cle that featured this specimen along with two brings up “Archaeoraptor,” which he regards others. This became an embarrassment for as a “hoax” and indicative of the sloppy sci- National Geographic when, at nearly the same ence that paleontologists do. In fact Wells time it ran its article, computerized axial spends the remaining third of the chapter try- tomography (CAT) scanning of the specimen ing to use “Archaeoraptor” in an attempt to showed it to be a composite. As it turns out, the slander the field of paleontology. Here too, he legs of the specimen belong to the counterslab gets most of the facts wrong. of a tiny non-avian theropod called “Archaeoraptor” was a fossil bought at the Microraptor (Xu et al., 2001); a full descrip- Tucson Gem and Mineral Show for Steve tion of the composite was published by Rowe Czerkas, a knowledgeable dinosaur enthusiast et al. (2001). To view the scans of the compos- and skilled sculptor and artist. Its remains ite, visit the UT Austin CT lab website came from the Liaoning area of China, which (www.ctlab.geo.utexas.edu/pubs/nature2000). has produced numerous beautifully preserved Wells concludes that this sorry episode fossils of fish, mammals, lizards, and both occurred because of “the cladists’ desire to avian and non-avian dinosaurs. Many of these prove their theory. Just as the need for a miss- were preserved with their body coverings, ing link between apes and humans led to such as fur or feathers, intact. So it was not Piltdown man, so the need for a missing link unexpected to see an allegedly new find from between dinosaurs and birds paved the way for there that combined features of fossil birds and the ‘Piltdown bird.’” (Wells, 2000:125). Not closely related dromaeosaurid dinosaurs, espe- so. The people who bought and promoted the cially given the large body of evidence sug- specimen weren’t cladists, and they never per- gesting that birds evolved from these formed a cladistic analysis or attempted to dinosaurs. The fossils of Liaoning are collect- place the specimen in a phylogeny. Piltdown ed by local villagers and farmers who know man was an intentional hoax played on scien- that “complete” specimens, particularly those tists, and the hoax was revealed by scientists with feathers, are preferred by scientists and when the specimen was studied. The forgery of collectors. Therefore, a cottage industry has “Archaeoraptor” was discovered by scientific sprung up around using parts to enhance or investigation as well, and it was cladists Tim make “whole” specimens (Chiappe et al., Rowe, Xu Xing, and Phil Currie who uncov- 1999). These constructed specimens are very ered it. The name “Archaeoraptor” was never 45 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education formally published as a scientific name, and tematists don’t use the term “fish” except in a has no scientific standing — the animal never restricted sense referring either to a subgroup existed. This doesn’t prevent Wells from itali- that is monophyletic such as Actinopterygia or cizing the name as if it were a real species. to “rayfins” (things like goldfish, trout, sword- Further, the specimen was never considered fish, etc.) — the vast majority of living “fish- important to our understanding of avian evolu- es.” Humans are vertebrates; so are fishes. tion. This doesn’t stop Wells from pretending Birds, by phylogenetic relationship, are otherwise, as if it were somehow important, dinosaurs. Just as dogs are canids, and also even crucial, to the idea that birds are descend- mammals, and also tetrapods and vertebrates. ed from dinosaurs. Consider a mailing address: just because you Returning to Archaeopteryx, Wells then live on 1010 Main Street does not mean that resorts to a classic creationist taxonomy game. you don’t live in Peoria or in Illinois, or that In this game, the creationist says that scientists someone living on 411 South Street doesn’t have to choose whether a fossil belongs to one live in the same town or state. taxonomic group or another. So, in the case of Wells’s most ridiculous treatment of “sci- Archaeopteryx, it has to be a bird or a reptile. ence” in this chapter is when he takes childish Then the creationist says that because it has shots at paleontologists. This is another popu- feathers it is a bird, and therefore because it is lar creationist tactic: attacking the character of a “bird” it cannot be a transitional form. In a prominent scientist or scientific field. In fact, effect the transitional features of the fossil are he devotes six pages to making fun of paleon- defined out of existence. This is a classic cre- tologists at a Florida symposium without ationist ploy, and nothing new; it is what we appearing to understand what they were say- have seen for decades from Duane Gish and ing. Worse yet, Wells completely misrepre- Henry Morris. Wells uses a slightly different sents the proceedings. For example, he claims approach, claiming that if Archaeopteryx and that a “cladistic analysis” showed a specimen birds are just dinosaurs, then humans are just presented there, called “Bambiraptor,” to be fish, which — he implies — is absurd. But this an ancestor of Archaeopteryx, yet no “cladistic is another case of Wells trying to use semantics analysis” was mentioned in either the descrip- to negate the evidence of evolution, just as he tion (Burnham et al., 2000) or the conference did with the Cambrian Explosion. proceedings. To my knowledge, no cladistic Here Wells exploits the systematic practice analysis has ever been performed on that spec- by which all groups of organisms must be imen. Wells then claims to be appalled that in “monophyletic,” that is, consist of an ancestor the reconstruction, “Bambiraptor” was shown and all of its descendants. In Wells’s rather covered in feathers even though none were naïve example, “fish” must be taken to include found fossilized with it. But other fossilized hagfishes, lampreys, sharks, goldfish and other dromaeosaurid dinosaurs are found covered in rayfins, coelocanths, and lungfishes. If “fish” feathers (e.g. Xu et al., 1999; Ji et al., 2001; were defined that way, then tetrapods (all ani- Norell et al., 2002) and so are the more basal mals that have four limbs) would indeed be Oviraptorids (Ji et al., 1998). The even more “fish” and “fish” would become another name basal Compsognathids are found with down- for “vertebrate.” But “fish” is not a taxonomic like feathers as well (Chen et al., 1998; see name; it is a colloquial term, and as a Ph.D. Figures 12, and 13). So it is conservative to biologist, Wells should know that. Real sys- reconstruct “Bambiraptor” with a covering of 46 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 13. Some examples of feathered dinosaurs discovered in Laioning, China.

47 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education feathers. Besides, the reconstruction is a pic- has no training or expertise in the field. ture, not scientific evidence — a confusion of Instead, he relies on caricatures of paleontol- Wells’s revealed further in the peppered moths ogy and paleontologists, and lampoons the chapter. By Wells’s logic, we shouldn’t accept entire field, treating scientists as if they were a the likelihood of fur on a fossil sabertooth cat. bunch of dinosaur-loving buffoons who are Is this the kind of “critical thinking” Wells easily fooled and misled. This is not science or wants us to teach our students? scholarship; this is tabloid journalism. Finally, Wells charicatures the conference WHAT THE TEXTBOOKS SAY presentation of Kevin Padian, who not only is a respected paleontologist but also happens to extbooks cover Archaeopteryx with be the president of NCSE. Padian’s talk was a varying degrees of brevity, frequently critique of the hypothesis that birds evolved Tgiving only a paragraph to from something other than dinosaurs. Wells Archaeopteryx, usually in the section on rep- likens Padian’s talk to an “old lawyers’ joke” tiles or birds or in the history of life section. about a “cracked kettle.” Wells even says that The lengths of the paragraphs vary from 54 Padian was not trying to be funny, and that it words to well over 500 (Figure 14), and the would be unkind to compare his talk to the average length falls around 200 words. joke, yet he continues the ad hominem attack Archaeopteryx is frequently used as an exam- summarizing Padian’s talk as a joke. Wells’s ple of a transitional form between reptiles summary, however, looks nothing like either (dinosaurs) and birds. Eight of the books treat the abstract that Padian submitted, which it as showing a dinosaurian ancestry for birds, Wells (as a conference attendee) received, or while two state that the ancestry is simply rep- the text of the talk he gave. In particular, tilian (Figure 14). Few of these books treat Padian never called the critics of the dinosauri- Archaeopteryx well and most of these discus- an origin of birds “unscientific,” just their crit- sions are garbled and contain factual errors icisms. He never accused them of “selective about Archaeopteryx. For example, Guttman interpretation” of the evidence; he just said contains numerous errors, even suggesting that that they did not use accepted methodologies it could not fly. Wells apparently does not even to evaluate the evidence. He never said that know enough about the topic to point this out. scientists reject their methodology regardless Wells only singles out the two books that use of the evidence; he said that we cannot evalu- the word “link” in their descriptions, Mader ate their methodology because they do not pro- and Schraer and Stolze. The most accurate dis- vide one. Finally, Padian’s conclusion was not cussions can be found in Raven and Johnson, that there was no controversy, but that the con- Campbell et al., and Johnson. Archaeopteryx is troversy over bird origins was journalistic, not sometimes used as an example of how fossils scientific (Padian, pers. comm.). If Wells was can elucidate evolutionary relationships. Few taking notes at the conference, he didn’t do a books use Archaeopteryx as direct evidence very good job. for evolution; some books (e.g., Johnson) Although Wells smugly chides paleontolo- instead use the origin of whales as the princi- gists for their supposed views about bird evo- pal example of a transitional sequence. lution, he has not attended any meetings of the Society of Vertebrate Paleontology or the Ostrom Symposium on the origin of birds. He 48 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 14. Textbooks’ treatment of and Wells’s grades for Archaeopteryx. WELLS’SEVALUATION ent scientifically incorrect data. What is most puzzling is that some books are given rather n grading textbooks on Archaeopteryx, the high grades compared to those given for other grading scheme, as usual, seems skewed to “icons.” Close examination of these books fail the books. Any book that does not I suggests that Wells misgraded them (Figure describe the transitional status of 14). For example, Wells gives Campbell, Archaeopteryx between reptiles and birds as Reese, and Mitchell a B, yet they clearly state “controversial” gets a D. As mentioned above, that Archaeopteryx is a transitional form there is no controversy about whether it is between dinosaurs and birds, for which a C or transitional, i.e, possesses structural features D would have been a more accurate grade both of its reptilian ancestors and of birds. To given Wells’s criteria. This negligent applica- get better than a D, a book would have to pres- tion of his own criteria calls into question the 49 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education rigor of Wells’s evaluation and the value of his about Archaeopteryx are simply inaccurate. To grades whether or not one accepts his idiosyn- follow his lead would mislead students into cratic criteria. thinking that fossils tell us nothing about evo- lutionary relationships. Considering the fact WHY ARCHAEOPTERYX STILL FLIES that Wells doesn’t understand ancestry or phy- IN TEXTBOOKS logenetic reconstruction, and he isn’t even f anything, the value of Archaeopteryx as a aware of Archaeopteryx’s status in paleontol- pedagogical tool is increasing with all the ogy, should we really be inclined to trust any- Inew discoveries of feathered dinosaurs thing he says on these topics? from China. Literally every new fossil discov- ery has added to the utility of Archaeopteryx. Archaeopteryx is still one of our best examples References of a fossil that preserves ancestral features while showing descendant novelties. Britt, B. B., P. J. Makovickey, J. A. Gauthier, and N. Archaeopteryx is but one of many fossils Bonde. 1998. Postcranial pneumaticity in showing a clear genealogical connection Archaeopteryx. Nature 395:374–376. between dinosaurs and birds (Figure 12). Burnham D. A., K. L. Derstler, P. J. Currie, R. T. Bakker, Z. Zhou and J. H. Ostrom. 2000. Remarkable new bird- Much like Mark Twain’s, the reports of its like dinosaur (Theropoda: Maniraptora) from the Upper death are greatly exaggerated. Cretaceous of Montana. University of Kansas HOW TEXTBOOKS COULD IMPROVE Paleontological Contributions 13:1–14. THE USE OF ARCHAEOPTERYX Burgers, P., and L. M. Chiappe 1999. The wing of Archaeopteryx as a primary thrust generator. Nature AND TRANSITIONAL FORMS 399:60–62. extbooks could improve their explana- Chen, P.-J., Z.-M. Dong, and S.-M. Zhen. 1998. An tions of transitions in evolution by exceptionally preserved theropod dinosaur from the Tfocusing on transitional features (not Yixian formation of China. Nature 391:147–152. forms or individual animals) that are borne by Chiappe, L. M., S.-A. Ji, Q. Ji, and M. A. Norell. 1999. Anatomy and systematics of the Confuciusornithidae a series of closely related organisms. Further, (Theropoda: Aves) from the late Mesozoic of northeast- textbooks should be clear in presenting the ern China. Bulletin of the American Museum of Natural idea that in general fossils are not considered History. 242:1–86. to be direct ancestors, but as records of ances- Clark, J. M., M. A. Norell, and L. M. Chiappe. 1999. An tral features. Finally, in discussions of oviraptorid skeleton from the Late Cretaceous of Ukhaa Archaeopteryx, textbooks need to tighten up Tolgod, Mongolia, preserved in an avianlike brooding position over an oviraptorid nest. American Museum their descriptions and check their facts about Novitates 3265:1–36. the history of both Archaeopteryx and the Dingus, L. and T. Rowe. 1998. The mistaken extinction: dinosaur–bird relationship. Textbooks should dinosaur evolution and the origin of birds. W. H. be clear that birds are descendants of dinosaurs Freeman and Company, New York 332p. and that there are no other credible potential Forster, C. A., S. D. Sampson, L. M. Chiappe, and D. W. ancestral groups; they should also augment Krause. 1998. The theropod ancestry of birds: new evi- their rather short discussions of avian evolu- dence from the Late Cretaceous of Madagascar. Science 279:1915–1922. tion with some of the new fossil evidence from Garner, J. P., G. K. Taylor, A. L. R. Thomas. 1999. On China where non-avian dinosaurs have been the origins of birds: The sequence of character acquisi- found with feathers (Figure 13). Wells’s claims tion in the evolution of avian flight. Proceedings of the 50 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Royal Society of London, Biological Sciences Series B Rowe, T., R. A. Ketcham, C. Denison, M. Colbert, X. 266:1259–1266. Xu, and P. J. Currie. 2001. Forensic paleontology: the Jensen, J. A., and Padian, K. 1989. Small pterosaurs and Archaeoraptor forgery. Narure 410:539–540. dinosaurs from the Uncompahgre fauna (Brushy Basin Shipman, P. 1998. Taking wing: Archaeopteryx and the Member, Morrison Formation: ?Tithonian), Late evolution of bird flight. Simon and Schuster, New York Jurassic, western Colorado. Journal of Paleontology. 336p. 63:364–373. Wagner, G. P., and J. A. Gauthier. 1999. 1,2,3 = 2,3,4: a Ji, Q., P. J. Currie, M. A. Norell, and S.-A. Ji. 1998. Two solution to the problem of the homology of the digits in feathered dinosaurs from northeastern China. Nature the avian hand. Proceedings of the National Academy of 393:753–761. Science 96:5111–5116. Ji, Q., M. A. Norell, K.-Q. Gao, S.-A. Ji, and D. Ren. Wells, J. 2000. Icons of evolution: science or myth?: 2001. The distribution of integumentary structures in a why much of what we teach about evolution is wrong. feathered dinosaur. Nature 410:1084–1088. Regnery, Washington DC, 338p. Norell, M., Q. Ji, K. Gao, C. Yuan, Y. Zhao, L. Wang. Witmer, L. M. 1995. The extant phylogenetic bracket 2002. ‘Modern’ feathers on a non-avian dinosaur. and the importance of reconstructing soft issues in fos- Nature 416:36–37. sils In J.J. Thomason ed. Functional Morphology and Norell, M. A., and P. J. Makovicky. 1999. Important fea- Vertebrate Paleontology Cambridge University Press p. tures of the dromaeosaurid skeleton II: information from 19–33. newly collected specimens of Velociraptor mongolien- Xu, X., X.-L. Wang, and X.-C. Wu. 1999. A dro- sis. American Museum Novitates 3282:1–45. maeosaurid dinosaur with a filamentous integument Ostrom, J. H., S. O. Poore, and G.E. Goslow. 1999. from the Yixian Formation of China. Nature 401: 262– Humeral rotation and wrist supination: Important func- 266. tional complex for the evolution of powered flight in Xu, X., Z. Zhou, and X. Wang. 2001. The smallest birds? Smithsonian Contributions in Paleobiology known non-avian theropod dinosaur. Nature 408:705– 89:301–309. 708. Padian, K. 1995b. Form and Function: The evolution of a dialectic. In J.J. Thomason ed. Functional Morphology and Vertebrate Paleontology Cambridge University Press p. 264–277. Padian, K. 2001a. Cross-testing adaptive hypotheses: phylogenetic analysis and the origin of bird flight. American Zoologist 41:598–607. Padian, K. 2001b. Stages in the origin of bird flight: beyond the arboreal–cursorial dichotomy. In J.A. Gauthier and L. M. Gall eds. New Perspectives on the Origin and Early Evolution of Birds. Yale University Press, New Haven p.255–273. Padian, K., and K. D. Angielczyk. 1999. Are there tran- sitional forms in the fossil record? In P.H. Kelley, J.R. Bryan, and T.A. Hansen eds. The Evolution–Creation Controversy II: Perspectives on science, religion, and geological education. Paleontological Society Papers 5:47–82. Padian, K. and L. M. Chiappe. 1998. The origin and early evolution of birds. Biological Reviews of the Cambridge Philosophical Society 73:1–42.

51 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education PEPPERED MOTHS HOW MANY MOTHS CAN DANCE ON THE TRUNK OF A TREE? THE STORY OF THE PEPPERED MOTH DISTRACTION BY IRRELEVANT DATA ndustrial melanism in peppered moths is ells disagrees with the results of the one of the most frequently used examples research on industrial melanism in Iof natural selection in action. This is large- Wthe peppered moth, and manipulates ly because of its pedagogical simplicity — it is the literature and the data to fit his views. He a straightforward example that is visual and points out that the “problem” of the peppered dynamic — and its copious documentation. moths is far from simple. His discussion cen- Industrial melanism refers to the darkening of ters on three points where he believes text- color that occurred in a number of species of books are in error, alleging that (1) the daytime insects following the Industrial Revolution. resting places of peppered moths invalidates This change appears to be related to the Kettlewell’s experimental results; (2) the pho- increase in pollutants in the environment. tos of the moths are “staged”; and (3) the Before the Industrial Revolution, individuals recovery patterns of populations dominated by of the moth species Biston betularia (com- light moths after the levels of pollution were monly called the “peppered moth”) were pre- reduced do not fit the “model,” although he is dominantly white with black speckles. By the unclear as to what the “model” is. All three of end of the 1800s, they were predominantly these objections are spurious. They are distrac- charcoal grey. This change was well docu- tions from the general accuracy of the story mented and led Tutt (1896) to hypothesize that and its value in showing the effects of natural this change was a result of pollution-stained selection on genetic variability in natural pop- trees’ affecting the camouflage potential of the ulations. moths. This change was termed “industrial First, Wells argues that the story is seriously melanism.” In the 1950s, Bernard Kettlewell flawed because “peppered moths in the wild decided to test the hypothesis that natural don’t even rest on tree trunks” (Wells, selection was working on the differential cam- 2000:138). He repeats this point throughout ouflage of the moths. In order to do this, he the chapter. However, it is both false and irrel- released marked light and dark moths into pol- evant, and only serves as a distraction to lead luted and non-polluted forests. He found that the reader away from the actual story of the birds appear to prey selectively on light moths moths. Contrary to Wells’s assertions, data in polluted forests and on dark moths in non- given by Majerus (1998:123) indicate that the polluted forests and so documented the idea of moths do indeed rest on the trunks of trees natural selection of these color patterns in 25% of the time. The rest of the time moths moths by birds. After anti-pollution laws took rest in branches (25%) or at branch-trunk junc- effect and the bark lightened, the moth popula- tions (50%). The facts have been pointed out tions in formerly polluted areas returned to repeatedly to Wells; his response has been previous color distributions. mostly to claim that moths don’t rest on “exposed” tree trunks (Wells, 2002 www.dis- covery.org/viewDB/index.php3?program=CR SC&command=view&id=1144). But this is not what he said in the text of Icons, which 52 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education remains flatly wrong. Moths are found all over words, birds hunt the prey they can see and trees, which is not a surprise (Clarke et al., hunt it where it is, not where it isn’t. Therefore, 1994) and it is mentioned in the references that no matter where the moth rests in the tree, it is Wells cites. visible to predatory birds, and thus its differen- To clear up any confusion, no researcher tial camouflage is important. doubts that the peppered moth rests in trees The purpose of Wells’s distraction is to put (Clarke et al. 1994; Majerus 1998), which the actual experiments into question and make means that the resting substrate is bark. Entire it sound as if the textbook authors are either trees are stained by pollution — the leaves, mistaken, or intentionally trying to fool stu- twigs, branches, trunks, and the surrounding dents. The insinuation is that because ground (Kettlewell, 1973) — and so the colors Kettlewell released the moths during the day, of the moths are relevant no matter where on they did not find “normal” resting places. the tree they rest — trunks, trunk-branch junc- Whether or not this is so, the release and cap- tions, branches, twigs, and even the leaves. ture experiments took place over a number of Wells’s argument implies that predatory birds days, so the moths were able to take up posi- can only see moths that are on exposed trunks. tions of their choosing, even if the first day By making this argument, however, Wells was not perfectly “natural” (Kettlewell, 1955, shows an apparent ignorance of the ecology of 1956, 1973). Kettlewell’s experiments were birds and woodland ecosystems. If you walk not perfect — few field experiments are — and into any forest, you can see that the birds fly they may have magnified the degree of selec- from tree to tree, branch to branch, and hunt at tion, but all serious researchers in the field all levels of the forest. Woodland species of agree that they were certainly not so flawed as birds that prey on moths and other insects live to invalidate his conclusion. and hunt in the canopy (the leafy part of the In his second objection, Wells ties the trees). These birds are not hunting from out- Kettlewell experiments to textbooks by con- side, soaring above the trees like hawks, as stantly repeating the statement that the illustra- Wells’s argument would require. tive photos were “staged” (Wells, 2000:150); In the scientific literature, there is extensive the important issue here is not how the photos discussion of the hunting behavior of birds, were made, but rather their intent. Wells including those that hunt peppered moths. implies that the photos purport to show a “life- Ornithologists have shown the woodland like” condition to prove that moths rest on ecosystem to be vertically stratified by compe- trunks. This is not the case. The photos are tition between different bird species. This meant to demonstrate the visibility of the dif- zonation means that there are skilled predators ferent forms of the moth on polluted and patrolling all levels of the forest: the trunks, unpolluted trees. It is absurd to expect a pho- trunk-branch joints, branches, and higher tographer to just sit around and wait until two canopy (Colquhoun and Morley, 1943; differently colored moths happen to alight side Hartley, 1953). Further, birds learn to distin- by side. Further, how the photos were pro- guish their prey against various backgrounds duced does not change the actual data. Birds and preferentially hunt prey in locations where eat moths and they eat the ones that they see they have found it in the past and that birds more easily first. The textbook photos never selectively prey on the more visible moths claim to depict a real-life situation, and it is (Pietrewitz and Kamil, 1977, 1981). In other improper to imply otherwise. 53 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

The third criticism, and the only scientific GREY AREA GRADES one that Wells levels, deals with the recovery ike the grading schemes for the other of the light form of the moth following the “icons,” this one is stacked against the institution of pollution control laws. The main textbooks as well. Even books that have thrust of his argument is that because the L more extensive discussions of the problems recovery of light-colored lichens does not cor- and details (such as Miller and Levine) can at relate with the recovery of the light form of the best earn a D. Like the grading schemes for moths, the entire story is incorrect. Wells Miller-Urey and Haeckel’s embryos, it is based exploits the fact that the original researchers largely on the presence or absence of pictures thought that the camouflage of the light moths (Figure 15). Explaining the peppered moth depended on the presence of lichen. However, story without photos (as in Biggs et al.), gar- the light forms recovered before the lichens ners a peculiar X grade. In order to get an A or did; therefore, Wells concludes, natural selec- B, a book must contain pictures of moths in tion has nothing to do with the story. Although “natural” resting places. Given Wells’s expla- it is true that the moths are well-camouflaged nation that these are unknown, presenting against lichens, and lichens are destroyed by those would be impossible. How can textbooks pollution, nevertheless the camouflage of the be expected to do that? A C would be awarded moths ultimately depends upon the color of the to a textbook that (1) used “misleading” pic- trees, which reflect the amount of soot staining tures, but (2) referred to them as “staged,” and the trees. Although lichens play a role in cam- (3) stated that the results of the experiment are ouflage, they are not necessary. This is what in doubt. Any standard textbook discussion of happened: pollution was reduced, the trees got the issue, even if it mentions that the story is lighter, then the moths got lighter. Further, in more complicated, is given a D. So as usual, all areas, the light moths have recovered, as this is a “no win” situation. This falls into predicted by the hypothesis. This is clearly Wells’s pattern of requiring the books to “crit- stated in the literature (e.g., Grant et al., 1998), icize” their examples, although the criticisms but it does not fit Wells’s story, and he just he insists on are largely fallacious. ignores it. WHY WE CAN STILL TEACH TEXTBOOK TREATMENT OF PEPPERED MOTHS AS AN EXAMPLE THE PEPPERED MOTHS OF NATURAL SELECTION ll but one of the textbooks (Campbell, lthough there will always be details of Reese, and Mitchell) reviewed in the peppered moth story that we do not AIcons cover the peppered moths and Afully understand, its status as an exam- present the basic story correctly. Again, how- ple of natural selection is not even remotely in ever, the coverage is limited to only a couple doubt. There is a clear correlation between pol- of paragraphs (Figure 15), varying from 117 to lution levels and moth color. Even if bird pre- over 500 words. Miller and Levine devote dation may not be the only factor involved in more than a page to the story, and even discuss the selection of one color over another, obser- some of its complexity, suggesting that the vations show that bird predation and substrate story is not as simple as it seems. color play the major roles in natural selection of the color of peppered moths. There are many areas in science where our knowledge is 54 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 15. Wells’s grades for textbooks’treatments of peppered moths. incomplete, but that does not mean we should ing on genetic variation. Some books already not teach about them. There are things we still cover sickle-cell anemia in humans (Figure do not know about gravity, but no one is 15). Other possible examples include antibiot- demanding that examples of gravity in action ic resistance in bacteria and myxomatosis virus be removed from textbooks. in rabbits in Australia. The key here is to expand the exposure of students to the many HOW TEXTBOOKS COULD IMPROVE examples of natural selection-driven evolu- THEIR PRESENTATIONS OF tionary change (e.g., Endler, 1986). What is PEPPERED MOTHS curious about Wells’s criticism of the peppered or the most part, textbook coverage of moth is that he says in Icons that he accepts the peppered moth story is adequate. As “microevolution.” The peppered moths are an Falways, expanding the discussion would example of “microevolution,” so why does he improve the coverage in the textbooks that have a problem with teaching it? cover it briefly. Textbooks could qualify the captions with a statement that the pictures illustrate differential camouflage in order to clear up any misunderstanding (however unlikely) as to the meaning of the photos. A better way for books to improve the topic is by adding other examples of natural selection act- 55 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

References

Clarke, C. A., B. S. Grant, F. M. M. Clarke, and T. Asami. 1994. A long term assessment of Biston betular- ia (L.) in one UK locality (Caldy Common near West Kirby, Wirral), 1959-1993, and glimpses elsewhere. Linnean 10:18–26. Colquohoun, M. K, and A. Morley. 1943. Vertical zona- tion in woodland bird communities. Journal of Animal Ecology 12:75–81. Endler, J. AL. 1986. Natural Selection in the Wild. Monographs in Population Biology 21. Princeton University Press, Princeton, 336p. Grant, B.S., A.D. Cook, C.A. Clarke, and D.F. Owen. 1998. Geographic and temporal variation in the inci- dence of melanism in peppered moth populations in America and Britain. Journal of Heredity 89:465–471. Hartley, P. H. T. 1953. An ecological study of the feed- ing habits of the English titmice. Journal of Animal Ecology 22:261–288. Kettlewell, H. B. D. 1955. Selection experiments on industrial melanism in the Lepidoptera. Heredity 9:323–342. Kettlewell, H. B. D. 1956. Further selection experiments on industrial melanism in the Lepidoptera. Heredity 10:287–301. Kettlewell, H. B. D. 1973. The evolution of melanism: the study of a recurring necessity; with special reference to industrial melanism in the Lepidoptera. Clarendon Press, Oxford. 423p. Majerus, M. E. N. 1998. Melanism: evolution in action. Oxford University Press, New York. 338p. Pietrewicz, A. T. and A. C. Kamil. 1977. Visual detec- tion of cryptic prey by blue jays Cyanocitta cristata. Science 195:580–582. Pietrewitz, A. T., and A. C. Kamil. 1981. Search image and the detecting of cryptic prey: an operant approach. In A. C. Kamil and T. D. Sargent eds. Foraging behav- ior. Ecological, ethological and psychological approaches. Garland Press, New York, p. 311–331. Tutt, J. W. 1896. British Moths. George Routledge, London 368p. Wells, J. 2000. Icons of evolution: science or myth?: why much of what we teach about evolution is wrong. Regnery, Washington DC, 338p.

56 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education “DARWIN’S FINCHES” evidence for evolution and natural selection, they have been dubbed “Darwin’s finches.” THE STORY OF This has led many people to conclude (mistak- “DARWIN’S” FINCHES enly) that Darwin’s theory of evolution was specifically inspired by the finches. arwin’s finches,” along with Hawaiian A LEGEND IN HIS OWN MIND honeycreepers and African cichlids, ells apparently feels the need to Dare frequently used as examples of attack the finches largely because adaptive radiation. In an adaptive radiation, a Wthey are an “icon” in need of “founder” species enters a new environment destruction; the chapter on the finches is per- with many unoccupied niches. This species haps the most poorly conceived section in the expands (radiates) and evolves adaptations to book. Wells initially focuses on the “biological fit these niches better. The process of becom- urban legend” that the finches inspired Darwin ing adapted to these different niches may lead to compose his theory of evolution. Of course to, and in these cases has led to, the formation this has nothing to do with whether or not the of new species. All the species of finches on finches are a good example of an adaptive the Galápagos Islands appear morphologically radiation. Therefore, his “requirement” that very similar, varying mostly in terms of beak textbooks specifically mention that the finches size and behavior; they all look very much like “played no role” in Darwin’s formulation of a species of finch from the mainland of South natural selection is irrelevant, only serving America. This suggests that all the finches on Wells’s efforts to portray evolutionary biolo- the Galápagos are descended from one original gists as people who just “make things up.” colonist species that went through an adaptive This is like saying that because Betsy Ross did radiation. Because of the small, isolated envi- not really sew the U.S. flag, the flag does not ronment of the Galápagos, the finches have actually exist. Wells even goes so far as to become the topic of extensive study into natu- brand the finches a “legend” — what is he try- ral selection. The studies that have been con- ing to imply? Finally, Wells’s assertion that ducted on the finches show strong selection for Darwin was not inspired by the finches is not larger beaks during droughts. These data show exactly correct. Although Darwin did not real- that climatic changes can have profound ize the significance of the finches until after effects on the morphology of a species and Gould pointed it out to him in 1837, he then potentially lead to the formation of new noted that the different species of finches were species. When Darwin visited the Galápagos, island-specific like the other Galápagos ani- he observed and collected some of the finch mals and suggested that they too were descen- species, believing that they represented a very dants of a mainland ancestor. Darwin made diverse set of birds that were not closely relat- extensive notes about the finches in his diaries ed. Their significance was not recognized until (Desmond and Moore, 1991). The finches, later, when ornithologist John Gould pointed then, did play a role in the formulation of out that the birds were all closely related finch- Darwin’s theory and they became an important es (Desmond and Moore, 1991). But because part of his evidence for the role of natural Darwin originally collected some of the speci- selection in evolution; they were not a “specu- mens and because the finches showed so much lative afterthought” as Wells claims. 57 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education After branding the finches a “legend,” Wells ued, larger beak sizes would continue to be switches gears and discusses the finches them- selected for, but the droughts did not. selves, acknowledging the strength of the evi- Evolutionary theory would predict that if cli- dence for an adaptive radiation, given the sim- mate oscillates, morphology would oscillate as ilarities of the different species. Wells almost well. The finches fit the predicted pattern. seems to accept that the finches are descended Speciation would require selection to be more from a common ancestor; at least, he does not constant than a couple of years here or there. It argue explicitly against it. But he demands that is not unreasonable to extrapolate that if just a there be direct evidence for speciation by nat- couple of years of drought can have that sig- ural selection; in his attempt to explain how nificant an effect on beak size, then extended this demand could be met, the remainder of the droughts could cause such variations to chapter degenerates into a series of non become fixed in a population, and lead to spe- sequiturs. This is particularly apparent in ciation. This is no different than extrapolations Wells’s discussion of what would constitute of unknown orbits. When a new comet is dis- “direct” evidence. covered, its orbit is calculated based on a few Suggesting that the work of Grant and Grant short-term observations. We assume that the claimed to be that direct evidence, he discuss- forces acting on the comet are constant and es their experimental work on finch beak vari- thus we can predict its position in 10, 20, 100, ation. The most detailed selection work on the etc. years. If gravity varied, then these extrap- finches was done by the husband and wife olations would be in doubt. In the case of the team of Peter and Rosemary Grant. For over finches, climate varied and the extrapolations two decades, the Grants and their students changed. Does Wells not allow scientists to have monitored the sizes of the beaks of some make reasonable extrapolations based on data of the finches on one small island (Grant, and observations? If so, physicists must be up 1999). They have documented that the size of next for Wells’s scorn. Perhaps what is most the finch beaks is correlated to the relative interesting about Wells’s discussion of this rainfall on the island, and thus to the abun- “icon,” however, is that in chapter 7 on the dance and hardness of the food. During dry peppered moths, he denies natural selection years larger beak size is selected for, while entirely, when he could have made the same during wet years the beak size is more varied. argument — that “no net evolution occurred” Wells acknowledges that the beaks vary and because the distribution of dark and light that this shows natural selection. He seems to forms of the moths returned to pre-industrial accept that the changes in beak shape are levels just as the finch beaks return to pre- caused by natural selection in reaction to drought levels. For finches he accepts natural drought-caused changes in the food supply. selection, but for the peppered moths he does These data are some of the most compelling not. for natural selection in the wild — something Wells goes on to complain about the extrap- that even Wells has a hard time denying. olations of speciation rates based on the However, he then contends that because the Grants’ data, complaining that the finches beak shape returns to a pre-drought size distri- aren’t an example of natural selection-driven bution, that no “net” evolution has occurred. speciation because no new species of finches But this is a mysterious contention. Natural arose during the duration of the Grants’ study. selection occurred. If the droughts had contin- However, no one would expect speciation to 58 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education occur on that scale, and the Grants never TEXTBOOK COVERAGE OF claimed to expect it either. And how would THE FINCHES you recognize a new species had formed? extbooks use the finches to illustrate a More importantly, one wonders how Wells wide variety of concepts, from the his- would recognize new species based on his gar- Ttory of evolutionary theory to adaptive bled discussion of species concepts (Wells, radiation, natural selection, taxonomy, phy- 2000:172-173), where he claims that one logeny, and niche partitioning. Textbooks that should “expect” “true” species to be separated discuss the finches in an historical context gen- by more than “just” beak shape and song pat- erally devote a paragraph or two to the finch- tern. This is important because in order to doc- es, sometimes in the discussion of how Darwin ument speciation, you need a model by which constructed his theory. Finches also frequently to recognize species. Wells provides none, and appear in sections dealing with patterns of evo- cannot even manage to explain the currently lution as an example of natural selection accepted models properly. and/or adaptive radiation. Only the upper-level Wells makes much of how the species of books discuss the Grants’ work specifically. finches are freely hybridizing and may in fact Space allotted to the finches vary from a few be merging. He claims that in order to be words to a few pages (Figure 16). In terms of “true” species, they should be separated by the historical discussion, most books discuss “more than beak shape and song pattern” the finches in connection with Darwin’s visit (Wells, 2000:172). However, such a separation to the Galápagos Islands. Few books explicitly is a perfectly acceptable definition of species credit the finches as Darwin’s inspiration, based on Recognition Concept (Paterson, however. Most do discuss the fact that they 1985), according to which species are separat- were part of his overall evidence that he col- ed by behaviors that lead animals to recognize lected on his voyage. Many books treat the potential mates. This species definition is finches as an example of an adaptive radiation. widely accepted amongst animal workers, Some books discuss the finches as examples of which Wells should know, having a Ph.D. in natural selection and niche splitting instead; biology. If Wells does not, one would expect these discussions occur in the chapters on evo- him to learn it as minimum required research lutionary processes or patterns. In Raven and before critiquing others’ diagnosis of species. Johnson, the finches are treated in detail; the Whether the species are merging or diverging discussion includes an accurate summary of is unimportant because both divergence and the historical story and the work of the Grants. merging are forms of long-term evolutionary This book mentions the finches as an example change. If indeed selection favors hybrids, as of adaptive radiation along with the African Wells appears to think, then the separate cichlids. species will merge. That’s still evolution and BIRD-BRAINED GRADING speciation by natural selection because the ue to the diversity of treatment of the new hybridized form will be a new species finches in textbooks, it is hard to eval- favored by natural selection. Duate the textbook coverage under Wells’s grading scheme. The grading scheme employed for the finch icon is perhaps the strangest of all Wells’s schemes. Like others, 59 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 16. Wells’s grades of textbook treatments of Darwin’s finches. Plus numbers refer to addi- tional treatments of finches. this grading scheme appears constructed teria. Adaptive radiation is a description of a specifically for failure. First, Wells objects to pattern and makes no statement as to the textbooks using the finches as an example of process — which the “origin of speciation by adaptive radiation, and he incorrectly equates natural selection” does. This is important an “adaptive radiation” with the “origin of because one can document an adaptive radia- species by natural selection” in his grading cri- tion without knowing the process by which it 60 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education occurred. He also wants textbooks explicitly to the more of those that we teach to students, the state that the beak shape oscillates. Further, in better they will understand evolution. order for a book to get an A, a B, or even a C, Comically, Wells never really objects to the the book must explicitly point out that the finches as an example of natural selection, finches had nothing to do with Darwin’s for- even concluding that “In this limited sense, the mulation of the theory of evolution. While finches provide evidence for Darwin’s theory” books should not suggest that the finches were (Wells, 2000:173). If that is the case, what’s more important in the formulation than they the big deal? were, it is interesting that in order to get a good grade, Wells insists that the books assert the References negative. In his grading scheme, this means Desmond, A., and J. Moore. 1991. Darwin: the life of a that any treatment of the finches that does not tormented evolutionist. Warner Books, New York, 808p. explicitly say that the finches did not inspire Grant, P. R. 1999. The ecology and evolution of Darwin automatically gets a D, even if it men- Darwin’s finches. Princeton University Press, Princeton, tions the beak size oscillation or evidence of 492p. merging. Thus the only criterion for the books’ Paterson, H. E. H. 1985. The recognition concept of grade is the statement of an unnecessary piece species. Transvaal Museum Monograph 4:21–29. of information — that Darwin was not inspired Wells, J. 2000. Icons of evolution: science or myth?: by finches. This has no pedagogical value and why much of what we teach about evolution is wrong. isn’t even wholly true; even if it were wholly Regnery, Washington DC, 338p. true, it has no bearing on the theory of evolu- tion one way or the other. This brings up the question of Wells’s real intent. His true goals are made apparent by the grades themselves. Wells grades many of the books needlessly low. When reevaluated on Wells’s own criteria, many of the books given a D or F should hav been given a C (Figure 16). Is Wells simply looking for any excuse to damage textbooks’ reputations? WHY WE CAN STILL USE THE GALÁPAGOS FINCHES AS A TEACHING EXAMPLE he finches clearly show adaptive radia- tion and were important to Darwin’s Tresearch. Their inclusion in texbooks is perfectly legitimate and should not change. The best way textbooks could improve their presentations of adaptive radiation is to include other examples such as Hawaiian hon- eycreepers or African cichlids as well. There are numerous examples of adaptive radiation; 61 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education CONCLUSION to support them with science. In order to get a “good” grade from Wells, ICONS — that is to portray a piece of evidence for evolu- SHOULD WE KEEP THEM? tion “accurately” (in Wells’s opinion), one must mention it and then proceed to criticize it. he role of primary and secondary edu- This is not standard pedagogical practice; if an cation is to pass on a certain body of example is that bad, it should be removed from Taccepted knowledge and basic concepts the biology curriculum, rather than introduced and then criticized. What we see is a pattern of to students in order to prepare them to learn grading to create bias rather than accuracy. more. The question is whether the criticisms Rewriting textbooks to criticize evolution leveled by the author of Icons would aid us in serves no teaching purpose (teaching is a pos- that goal; the resounding answer is no. The itive endeavor, not negative), yet it is clear educational program that would result from from the grading that this is the goal of the implementing the suggestions contained in author. What’s worse is that the grading crite- Icons would have just the opposite result: It ria are not even consistently applied. There is would seriously hamper science education and no pedagogical or factual basis for these leave students unprepared for the future in grades, and they should not be taken seriously. which science, and biology in particular, will To follow Wells’s advice would not only result play an increasing role. in mis-education about evolution, but about all Figure 17 shows the grades that Wells gives of biology and other sciences as well. Good each textbook, the number of pages each text- teaching may value critical thinking, but it book specifically devotes to evolutionary the- does not value wanton criticism for the sake of ory compared to the total number of pages that criticism. contain evolutionary content (such as phyloge- Finally, in his zeal to attack the textbooks’ netic relationships), and when evolution is first treatments of evolution, Wells misses a chance mentioned in the text. Although the amount of to provide a good listing of actual errors in text devoted to evolution varies widely in text- textbooks. A study of the textbooks that Wells books, the coverage as a percentage of the total evaluated uncovered factual errors, inexact text that evolution is given is very small — wordings, and garbled explanations of biolog- usually less than 10% of the total book. Wells ical phenomena that Wells either did not notice evaluates only four high school textbooks in or considers unimportant. This lack of docu- the review, and those have a far smaller treat- mentation of real textbook errors is yet anoth- ment of evolution than do the college texts. er failure of Wells’s effort. Far from being It is clear from Wells’s treatment of the tracts of “evolution propaganda,” as Wells “icons” and his grading scheme that his inter- implies, many biology textbooks devote too est is not to improve the teaching of evolution, little space to evolution, especially in early but rather to teach anti-evolutionism. Under chapters. Most of evolution is reserved for the Wells’s scheme, teachers would be hostile to middle of textbooks; it is frequently given less evolution as part of biology instruction. Wells coverage than ATP cycles or photosynthesis. and his allies hope that this would open the The topics of which Wells is so critical amount door to alternatives to evolution (such as to only a small fraction of any given textbook. “intelligent design”) without actually having In fact, evolutionary biologists consider the 62 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education

Figure 17. Grades given to textbooks in comparison to the coverage of evolution given in the text. AP and College level texts shaded. Parenthetical numbers under the heading “number of pages devoted to evolution” refer to the number of pages in the “evolution” chapter. lack of coverage of evolution, and the failure out the work of Richardson et al. (1997, 1998), to interweave it throughout the entire book, to or the “peppered moths” chapter without be the greatest deficit of textbooks. Coyne (1998) and Majerus (1998), or the In conclusion, the scholarship of Icons is “Archaeopteryx” chapter without Shipman substandard and the conclusions of the book (1998). Even then, Wells’s discussions are rife are unsupported. In fact, despite his touted sci- with inaccuracies and out-of-date information. entific credentials, Wells doesn’t produce a Wells seems to think that scientific theories are single piece of original research to support his supported by certain “keystone” pieces of evi- position. Instead, Wells parasitizes on other dence, removal of which causes the theory to scientists’ legitimate work. He could not have collapse. Paradigms in science work when written the “Haeckel’s embryos” chapter with- they provide solutions and further research; 63 Icons of Evolution? Why Much of What Jonathan Wells Writes about Evolution is Wrong Alan D. Gishlick, National Center for Science Education their health is not tied to single examples. The References paradigm of evolution is not tied to a single piece of evidence. Coyne, J.A. 1998. Not black and white. Nature 396:35– If that is the case, why “defend” the “icons” 36. at all? If evolution doesn’t need them, why not Majerus, M. E. N. 1998. Melanism: evolution in action. just replace them? The answer is simple: There Oxford University Press, New York, 338p. is no reason to throw out good teaching exam- Richardson, M. K., J. Hanken, M. L. Gooneratne, C. ples unless the criticisms leveled against them Pieau, A. Raymond, L. Selwood, and G.M. Wright. 1997. There is no highly conserved embryonic stage in are valid. We should not just acquiesce to the vertebrates: implications for current theories of evo- Wells’s arguments unless they have merit. Just lution and development. Anatomy and Embryology as no piece of evidence becomes a teaching 196:91–106. example without extensive testing, no example Richardson, M. K., J. Hanken, L. Selwood, G. M. should be removed on the basis of one poorly Wright, R. J. Richards, and C. Pieau. 1998. Haeckel, argued, inaccurate, and tendentious book. In embryos, and evolution. Science 280:983–984. each case, it is Wells’s arguments that are Shipman, P. 1998. Taking wing: Archaeopteryx and the evolution of bird flight. Simon and Schuster, New York, wanting, not the “icon.” 336p. When Alfred Wegener first proposed his Wells, J. 2000. Icons of evolution: science or myth?: theory of continental drift, he was laughed at why much of what we teach about evolution is wrong. and ridiculed. What did he do? Did he form a Regnery, Washington DC, 338p. non-profit advocacy group and lobby state school boards and lawmakers to force teaching of “evidence against” geosynclinal theory? Write a book called Icons of Uniformitarianism? Evaluate and grade earth science textbooks and demand that they be rewritten to remove examples of “border- lands”? No. He went back and did more research. He found like-minded colleagues and they produced research. He fought in the peer-reviewed literature. He produced original research, not polemical popular tracts or poli- tics. Eventually his ideas were adopted by the whole of geology — not through politics but because of their overall explanatory power. If Wells and his colleagues want “intelligent design” to succeed, they need to produce that research. Until they do, evolution remains the reigning paradigm and the “icons” are perfect- ly acceptable teaching aids.

64