Swatting through Luke Barnes’ Review of “Why Evolution is True”

Luke Barnes is a postdoctoral researcher in the field of astronomy at the University of Sydney. He is probably a very competent astronomer. However, he seems to have some issues with modern evolutionary biology but dislikes being labeled a creationist. Despite his statements being very carefully engineered, he repeats many classic tactics and tropes of creationists.

Barnes wrote a three-part book review of Jerry Coyne’s book “Why Evolution is True” a while back that I will take pleasure in disentangling. I’m not someone who would defend Coyne no matter what, as I have strongly criticized his anti-psychiatry stance a couple of times before on this blog.

Barnes alludes to the stereotype that physicists tend to march into a field not closely related to physics and make sweeping proclamations about conclusions and problems in that field, especially if this field is perceived as being less stringent than physics. There are a few notable examples of where this has not turned out that good such as Freeman Dyson and climate change, Roger Penrose and consciousness as well as Linus Pauling (quantum chemist) and high doses of Vitamin C.

Generally speaking, the three parts roughly corresponds to criticizing the positive case for evolution, opposing the positive and negative case against intelligent design creationism and the supposedly negative effects of evolutionary biology on society. However, I will do my review of the review in one single post because I can focus on the core claims and misunderstandings.

No flies were actually harmed during the production of this post.

0. Why read a popular science book rather than textbooks or the scientific literature?

Barnes admits that he doesn’t really know that much about biology. That is fine. I do not know a lot about damped Lyman Alpha systems. However, it seems reasonable to suppose that he understands the value of textbooks and the scientific literature and realizes that you can’t really go into a lot of details in a popular-level book on a scientific topic. So if we wants to learn more about e. g. evolution, why not pick up a textbook on the subject like Stearns and Hoekstra (2005)? Or why not Davies, Krebs and West (2012) for a more research-based approach? Even certain Internet resources like Theobald (2012) offers a detailed overview of the evidence for evolution. This argument is even stronger in the field of physics. A good introductory textbook is probably several orders of magnitude better than the average popular science book on physics.

1. The genetic relationship between humans and chimps

Barnes starts off by butchering a quote from Coyne about the genetic relatedness of humans and chimps as well as lifting it out of context. Here is how Barnes portrays it.

A good example of Coyne’s clarity is in explaining to the oft-heard claim that chimps and humans are 98.5% alike:

[…] recent work shows that our genetic resemblance to our evolutionary cousins [chimps] is not quite as close as we thought […] to consider an analogy, if you change only one percent of the letters of this page, you will alter far more than 1 percent of the sentences […] more than 80% of all the proteins shared by the two species differ by at least one amino acid.” (pg 230)

A common sense reading would conclude that Coyne is arguing against the position that chimps and humans share 98.5% of their protein primary sequence. However, this is not the case when we understand the context of the quote.

Our Genetic Heritage

If we don’t yet understand why selection made us different from other apes, can we at least find out how many and what sort of genes differentiate us? “Humanness” genes have become almost a Holy Grail of evolutionary biology, with many laboratories engaged in the search. The first attempt to find them was made by 1975 by Allan Wilson and Mary-Claire King at the University of California. Their results were surprising. Looking at protein sequences taken from humans and chimps, they found that they differed on average by only about 1 percent (more recent work hasn’t changed this figure much, the difference has risen to about 1.5 percent). King and Wilson concluded that there was a remarkable genetic similarity between us and our closest relatives. They speculated that perhaps changes in just a very few genes produced the striking evolutionary differences between humans and chimps. This result garnered tremendous publicity in both the popular and scientific press, for it seemed to imply that “humanness” rested on just a handful of key mutations.

But recent works shows that our genetic resemblance to our evolutionary close as we thought. Consider this. A 1.5 percent difference in protein sequence means that when we line up the same protein (say, hemoglobin) of humans and chimps, on average we’ll see a difference at just one out of every 100 amino acids, but proteins are typically composed of several hundred amino acids. So a 1.5 percent difference in a protein 300 amino acids long translates into about four differences in the total protein sequence (to use an analogy, if you change only 1 percent of the letters on this page, you will alter far more than 1 percent of the sentences). That oft-quoted 1.5 percent difference between ourselves and chimps, then, is really larger than it looks: a lot more than 1.5 percent of our proteins will differ by at least one amino acid from the sequence in chimps. And since proteins are essential for building and maintaining our bodies, a single difference can have substantial effects.

Now that we’ve finally sequenced the genomes of both chimp and human, we can see directly that more than 80% of all the proteins shared by the two species differ in at least one amino acid. Since our genomes have about 25000 protein-making genes, that translates to a difference in the sequence of more than 20000 of them. That’s not a trivial divergence. Obviously, more than a few genes distinguish us. And molecular evolutionists have recently found that humans and chimps differ not only in the sequence of genes, but also in the presence of genes. More than 6% of genes found in humans simply aren’t found in any form in chimpanzees. There are over 1400 novel genes expressed in humans but not in chimps. We also differ from chimps in the number of copies of many genes that we do share. The salivary enzyme amylase, for example, acts in the mouth to break down starch into digestible sugar. Chimps have but a single copy of the gene, while individual humans have between two and sixteen, with an average of six copies. This difference probably resulted from natural selection to help us digest our food, as the ancestral human diet was probably much richer in starch than that of fruit-eating apes.

Putting this together, we see that the genetic divergence between ourselves and chimpanzees come in several forms — changes not only in the proteins produced by genes, but also in the presence of absence of genes, the number of gene copies, and when and where genes are expressed during development. We can no longer claim that “humanness” rests on only one type of mutation, or changes in only a few key genes.

Coyne affirms that the figure of 1.5% differences in amino acid is correct. His quibble is with the previous view that this means that this 1.5% merely represents a few mutations in key genes. Instead, Coyne takes the position that because the size of a protein and the size of the proteome is so large, 1.5% will represent quite a few differences in amino acid sequences on an absolute scale and that out of this difference, as well as copy number variation and other factors, emergences the phenotypic difference between human and chimps. This does not in any shape or form overturn the fact that humans and chimps are very closely related genetically and the evolutionary conclusions that stem from this.

I need not point out that quoting out of context is a classic tactic by creationists. In fact, creationists have produced entire books with nothing but quotes from biologists and other scientists out of context.

2. Why not mock young-earth creationists with whale atavisms?

Barnes expresses his amazement with the existence of atavistic hind legs in humpback whales and asks why someone have not taken a skeleton with these and parked it outside the Creation Museum in Kentucky. There is a very simple explanation for this: after more than half a century of merciless human predation, the International Whaling Commission gave it protection status worldwide way back in 1966. That means that a biologist would probably not get approved by an ethical review board and if he or she still tried to go out there and catch specimens, I’m fairly certain the scientific career of that person would be over and may, depending on the country, face legal consequences. The most cited photographs show the bones of the legs themselves detached from the whale skeleton, so it is rather uninformative for the causal reader. Moreover, most existing skeletons are in the property by natural history museums that probably would not support such a bold venture. Finally, creationists tend to distort the evidence from atavisms, so it is probably not such an effective tool.

3. An incautious analogy out of context

I agree with Barnes that Coyne made a careless analogy comparing certain aspects of evolutionary biology to astronomy and that he gets some of the specific cosmological details wrong, but this is really nitpicking from Barnes. Nothing wrong with nitpicking in itself (I have a habit to do that as well), but it does not address to cumulative case for modern evolutionary biology made in the book. It is also slightly out of context. Coyne does not paint a picture of inductive reasoning. The surrounding context makes clear that evolutionary biology makes use of both inductive and deductive reasoning.

The way we have discovered how species arise resembles the way astronomers discovered how stars “evolve” over time. Both processes occur too slowly for us to see them happening over our lifetime. But we can still understand how they work by finding snapshots of the process at different evolutionary stages and putting these snapshots together into a conceptual movie. For stars, astronomers saw dispersed clouds of matter (“star nurseries”) in galaxies. Elsewhere they saw these clouds condensing into protostars. And in other places they saw protostars becoming full stars, condensing further and then generating light as their core temperature became high enough to fuse hydrogen atoms into helium. Other starts were large “red giants” like Betelgeuse; some showed signs of throwing off their outer layers into space; and others still were small, dense white dwarfs. By assembling all these stages into a logical sequence, based on what we know of their physical and chemical structure and behaviour, we’ve been able to piece together how stars form, persist, and die. From this picture of stellar evolution, we can make predictions. We know, for example, that starts about the size of our Sun shine steadily for about ten billion years before bulging out to form red giants. Since the Sun is about 4.6 billion years old, we know that we’re roughly halfway through our tenure as a planet before we’ll finally be swallowed up by the Sun’s expansion.

And so it is with speciation. We see geographically isolated populations running the gamut from those showing no reproductive isolation, through those having increased degrees of reproductive isolation (as the populations become isolated for longer periods), and, finally, complete speciation. We see young species, descended from a common ancestor, on either side of geographic barriers like rivers or the Isthmus of Panama, and on different islands of an archipelago. Putting all this together, we can conclude that isolated populations diverge, and that when that divergence has gone on for a sufficiently long time, reproductive barriers develop as a by-product of evolution.

Creationists often claim that if we can’t see a new species evolve during our lifetime, then speciation doesn’t occur. but this argument is fatuous: it’s like saying that because we haven’t seen a single start go through its complete life cycle, starts don’t evolve. historical reconstruction of a process is a perfectly valid way to study that process, and can produce testable predictions.

In fact, the quote Barnes posted is inside a section specifically detailing the predictions by allopatric speciation. The general message of the passage is that both stellar and biological evolution are conclusions based on a convergence of independent lines of evidence. It should also be understood as an argument against the false characterizations of the evidence for speciation made by many creationists.

4. Evolution and prediction

Barnes thinks that Coyne’s discussion of the patterns of the fossil record (Barnes selects a single example, the number of whorls in a certain unicellular eukaryote) does not cut it. I think it is important to make the distinction between two different things: (1) Predictions of common descent / natural selection / other evolutionary mechanisms and (2) model selection. Coyne does not really separate these either, but when we do, the situation will be clearer.

Coyne decided to compare how we would except to find the fossil record under the common creationist notion that different kinds of organisms were formed roughly at the same time with their basic features already existing with that of gradualist common descent by means of natural selection.

Most forms of creationism allow changes within “kinds” (a taxonomic category that is not defined), but one “kind” of animal cannot turn into another “kind” and two different “kinds” do not share a common ancestor. So if this form of creationism was true, he would expect to see fossils of mammals, reptiles, birds, bacteria, flowers and so on, all start to exist at roughly the same time. This is not what we see. There are no mammals before reptiles, no flowers before bacteria, no angiosperms before woody plants and so on. They also do not fall into ecological zoonation: flying dinosaurs are below the first humans, although they would live at a higher altitude than humans and certain slow organisms, like turtles, are more ancient than birds, although obviously a turtle would be buried further down than birds i they were surprised by a global flood. If creationism is true, we would not find any fossils with intermediate features, such as mosaics of features from reptiles and birds, or fish and amphibians. On the other hand, if evolutionary common descent is true, on the other hand, we would expect to see precisely that which we do see. I wrote about this in the comment section of another entry on Barnes’ blog, so I will settle for reposting it.

Phylogenetic trees indicating common descent makes specific predictions about the morphology of fossils with intermediate features because each node between any two branches predicts a common ancestor. These morphological features are explicitly specified based on an analysis of the most common derived characters (more common means higher probability that the common ancestor had them) and what transitions would have occurred at the tie. Common descent also predicts the chronological order, not just existence, of these fossils with intermediate features. Classic cases where almost a full spectrum of fossils with few or no morphological gaps have been found includes fossils with intermediate features of dinosaur and bird, reptile and mammal, ape and human and several others.Finding any fossil with intermediate features forbidden by common descent, such as a fossil with intermediate features form birds and mammals (the most recent common ancestor of birds and mammals where not a bird or a mammal so it could not be explained by the common ancestor having those features) would falsify common descent.

I also discussed three other cases of predictions from common descent (some are even quantitative!), namely nested hierarchies (which we will come back to), consilience of independent phylogenies and endogenous retroviruses (ERVs).

If common descent is correct, we should observe nested hierarchies or groups-within-groups (which we do). To take plants as an example, monocotyledons and dicotyledons are nested within the angiosperm (plants with enclosed and protected seeds) group, and the angiosperm and gymnosperms (seeds that are not enclosed) are nested in the seed plants and seed plants and non-seed plants are nested in the vascular group. The vascular group and non-vascular group is nested within plants. Also, the standard phylogenetic tree have statistical and practically significant high values of hierarchical structure. Now you can classify almost anything, such as cars, into hierarchies if you want, but there is no requirement for, say, a blue car to have four doors in the same way that nonvascular plants cannot have seeds or flowers. So hierarchical classifications of cars would be subjective, but that of organisms would not. Take languages as an other example. They share common ancestors and derived by the same decent with modification and can be classified in objective nested hierarchies. No reasonable person could claim that Swedish should be grouped with Mandarin rather than with Norwegian. On the other hand, it would have been profoundly problematic if it had turned out that many different species had combined characteristics from different nested groups. What if we found nonvascular plants with seeds? Conifers with flowers? Ferns with woody stems? Birds with mammary glands or hair? Fish with differentiated teeth? If there existed a mix and match of characters, it would make it hard to create objective nested hierarchies. In fact, this would be an eminent way to falsify common descent. On the other hand, cars do mix and match characters, which is why they can only be subjectively grouped.

If common descent is correct, we should expect that phylogenetic trees constructed from independent lines of evidence (say, genetics and morphology etc.) should seamlessly converge with a high degree of statistical and practical significance. Now, as the number of taxa analyzed increases, the number of theoretically possible trees increase very fast. There are over 10 to the power of 38 different ways to order 30 major taxa. Yet trees constructed from morphological characters are congruent with trees determined independently from e. g. molecular studies on cytochrome c. Even partly incongruent trees match with a high degree of significance, simply by the vast number of theoretically possible trees. A very loose analogy is trying to measure a physical variable and coming up with results around, say 10 (your test series could be 9.82, 9.94, 10.04, 10.12 etc.). These are strictly speaking not the same, but it is easy to see that they center around something that is close to 10. If it was the case that there was no consilience of independent phylogenetic trees derived from morphology and molecular data, it would have falsified common descent. Indeed, based on the extremely large number of theoretically possible trees, a lack of consilience of independent phylogenetic trees is the most likely result. Not only that, we can now test the notion that independent phylogenetic trees are consilience by checking things like chromosome number, length and the chromosomal position of genes. These are roughly independent of morphology and DNA sequence of specific genes. If the results show that these too are consistent, then this is further evidence for common descent. If they do not, we have a problem.

Endogenous retroviruses (ERVs) are remnants of retroviruses that have been incorporated into the genome of the infected cell and for some reason failed to complete its cycle. This can be due to mutation within the integrated viral DNA or unfavorable circumstances whatever. If this occurs in germ cells, descendants will also carry this broken retroviral integration. Now, this type of integration is uncommon and mostly random, so finding similarities of endogenous retroviruses (sequence, number, types etc.) indicate closer evolutionary relationship. Humans and chimps share more of their ERVs that either compared to other primates. These two and gorillas share more of their ERVs compared to other primates. These three share more of their ERVs that either of them do with orangutangs, which in turn share more of their ERVs than either of them do with gibbons, which in turn share more of their ERVs than either do with old world monkeys, which together share more of their ERVs than either do with new world moneys and so on. This supports common descent. Now, if other mammals, say cows, had exactly the same ERVs in the exact same positions as humans. It would be extremely unlikely for dogs to carry the three HERV-K ERVs that we know unique to humans since no other primate have them. This would be a falsification of common descent or at the very least seriously damage the model.

5. Darwin and data for evolution

Barnes quotes Coyne stating that while there were other evolutionary models before Darwin, Darwin was the first person to use data for his conclusion. This should be understood as saying that Darwin was the first person to gather, organize and present scientific evidence for common descent and natural selection. Vestiges of the Natural History of Creation (1844) did present an evolutionary scenario, but did not really use that much evidence or data to support such a position. Vestiges also entirely lacked a plausible mechanism for evolutionary change.

6. The scientific contribution of saints

Barnes chastise Coyne for making fun of catholic sainthood. It is true, but nitpicking. The counterexample of Euler does not fly since he is not a catholic saint. There are people who made contributions to since and was later made saints (e. g. Albertus Magnus, the probable discoverer of arsenic), although Coyne may escape on the technicality that it had to be an “important” contribution. It is also to point out that Steno was not actually the first person to discover the geological principle of superposition. It was the arabic scholar Avicenna. This was one of the errors that Coyne did that Barnes missed.

7. Different types of speciation, the founder effect and continental drift

Barnes think he has found a contradiction between two statements on speciation in relation to oceanic islands, yet this is yet another case of quoting out of context. Barnes also fails to appreciate the the differences between allopatric and sympatric speciation. He writes that:

“… accidental colonists were able to form many species because oceanic islands offer lots of empty habitats that lack competitors and predators.” (pg 116) But on page 201 we’re told that oceanic islands don’t facilitate speciation because there are too few chances for geographic isolation to produce non-interbreeding populations. I thought that evolution would be slow in the absence of predators and competitors, because the weak would be able to survive and keep their genes in the genepool.

Let’s define our terms: continental islands are islands that was connected to the mainland in the past. It is no longer connected due to rise in sea level or continental drift. Examples Coyne gives of continental islands are the British isles, Madagascar and Sri Lanka, but there are others. Oceanic islands are those islands, like Hawaii, that arose from the sea floor and was never attached to the continents. Now, the empirical observation is that while the mainland and continental islands have roughly the same species, oceanic islands tend to lack certain organisms like fresh water fish, mammals and reptiles. The groups actually present on oceanic islands tend to contain very similar species, such as finches on the islands of Galapagos., even thought they are ecologically diverse. In other words, oceanic life is unbalanced. It primarily contain organism like insects, birds and plants that can disperse easily, while mammals and reptiles have it harder.

So what is the rate of evolution among species arriving to an oceanic island? Barnes maintains that Coyne contradicts himself on the topic. This is not the case. On page 116, Coyne explains that these organisms will have few predators and few competitors. This is true, but he is talking about inter-species competitors, not intra-species competitors. Of course members of the same species will compete for resources. This kind of situation will tend to benefit specialists over generalists (less competition for your particular food), so diversification occurs. Generalists will have it harder to compete for food, so they will be selected against. This is a form of disruptive selection.

So far so good. But what about page 201? Before we get to that, we need to discuss different types of speciation. The two types relevant for us is allopatric and sympatric speciation. Allopatric speciation is the one you always think of: geographical separation promotes genetic divergence. Sympatric speciation, on the other hand, occurs in the same general geographical region and the genetic isolation is not complete. This means that speciation is harder, but can nevertheless sometimes occur.

On page 201, Coyne is talking about small patches of multiple oceanic islands, where allopatric speciation is rarer and sympatric speciation would be the dominant mode of speciation. Coyne argues that speciation on these patches of multiple oceanic islands would be rare, because there would not be sufficient geographical isolation between close oceanic islands.

Thus, the two statements are entirely compatible. The important factor missing for Barnes is that the two situations differ: a single oceanic island versus a small patch of multiple oceanic islands.

8. Molding multiple traits simultaneously?

Barnes’ general argument is that some or many beneficial mutations may arise in an individual that might not reproduce do to some other facts. This is true, and it offers one of many constraint on evolution. This have been discussed many decades ago and the general reply is that it is true, but we can think of organisms as average statistical optimizers. The general point is also that it is not as easy as “one gene, one trait”, but think of genes and traits as an immense network of interaction components. Most genes affect multiple traits and most traits are influenced by many genes. Genes can enhance or mask the effects of other genes. The status of “beneficial” or “detrimental” will also depend on both the genetic context and surroundings of that particular mutation, and also the external environment. Some mutations that are harmful in one environment, may be beneficial in another (the classic example here is sickle cell anemia, which confers malaria resistance).

The biologist P. Z. Myers describes the problem with this naive view of evolution that Barnes portraits in the article It’s more than genes, it’s networks and systems.

9. What exactly is a just-so-story?

To answer this question, we need to go back to a very influential article written by Gould and Lewontin in the late 1970s called The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme. Simply put, the paper argues that (1) biologists should not become extremist adaptationists and believe that every biological trait is a direct adaptation, (2) that other factors, such as evolutionary time lag, historical constraints etc. may be important partial explanations, (3) that some biological features may be by-products of selection on something else and (4) that it is important to consider alternative explanations.

In this context, a just-so-story is an adaptationist account of the evolutionary origin of a feature relying merely on plausibility, rather than arguments and evidence. As they put it: “Often, evolutionists use consistency with natural selection as the sole criterion and consider their work done when they concoct a plausible story. But plausible stories can always be told. The key to historical research lies in devising criteria to identify proper explanations among the substantial set of plausible pathways to any modern result.”

This is the problem Coyne sees in certain adaptationist explanations for human behavior. He considers some of these to be without evidence and merely relying on plausibility. Sometimes, he even scolds then for misunderstanding evolutionary processes, such as those psychologists who has proposed adaptation by group selection as a partial explanation for depression, which is a marginalized viewpoint among modern evolutionary biologists.

It is perfectly alright to appeal to adaptation by natural selection, but biologists should not go overboard and ignore the biological context.

This concludes the first post of Barnes book review. In the next part, he discusses Coyne’s criticisms of intelligent design creationism.

10. Definitions

Barnes complains that Coyne did not define intelligent design creationism. Definitions are important, but it is equally important to understand that the intelligent design movement is a loose association of various types of creationists from those who believe in a young earth like Paul Nelson or Philip E. Johnson to those who accept an old earth, like Michael Behe and William Dembski. Some accept common descent, some do not. Some accepts natural selection, some do not. And so on. This is because the originator of the movement, Johnson, wanted to set up a big tent of different types of Christians to present a unified front in the form of a wedge against the supposedly stifling materialism of Darwinian evolution (the log) and to put aside internal conflicts, like the age of the earth or whether a global flood occurred. After convincing the establishment about intelligent design, he thought, the internal debates could be settled. This was in some ways an interesting tactic, but ultimately it fell apart because of internal differences. Liberal and moderate theological seminaries and universities did not buy into intelligent design creationism.

Now, proponents of intelligent design creationism obviously do not call themselves creationists. Then they would face the same constitutional issues as so called creation since did in the 70s and 80s. So they invented the phrase “intelligent design” and removed references to a young earth and a global flood. But the general arguments against evolution, and the rhetorical tactics remained. This has lead mainstream scholars to conclude that everything that you can find in intelligent design can be found in previously existing creationism, making it one of many different subsets.

So we can think of intelligent design as a politically orientated slimmed down version of creationism. An internal document from the late 90s, called the Wedge Strategy (alluding to Johnson’s wedge of unified Christians splitting the foundations of materialism), lead out on the Internet. In it, we can read the real goals of the intelligent design movement, which includes “To defeat scientific materialism and its destructive moral, cultural, and political legacies”, “To replace materialistic explanations with the theistic understanding that nature and human beings are created by God” and “alongside a focus on the influential opinion-makers, we also seek to build up a popular base of support among our natural constituency, namely, Christians. We will do this primarily through apologetics seminars. We intend these to encourage and equip believers with new scientific evidences that support the faith, as well as to popularize our ideas in the broader culture”.

The evidence presented at the Dover Trial convincingly showed that intelligent design was nothing more than creationism when the science side showed that the intelligent design textbook Of Pandas and People contained almost identical wording of definitions of intelligent design as previous creationist textbooks defined creationism. After the Supreme Court verdict in Edwards v. Aguillard (1987) banning the teaching of creation science in the classroom, the publisher known as Foundation for Thought and Ethics, simply did a search-and-replace on their central book on the topic, replacing “creation*” with “intelligent design”. Unfortunately, this produced a transitional fossil: the phrase “cdesign proponentists” could be found in the documents.

The definitions of intelligent design provided by members of the Discovery Institute, such as “certain features of the universe and of living things are best explained by an intelligent cause, not an undirected process such as natural selection.” tend too be vague, lack mechanism and be too overlapping with a belief in a god.

So for all intents and purposes, we can consider intelligent design a subset of creationism, but that this is not the same as a general belief in a god. After all, it is entirely possible to fully accept modern evolutionary biology and believe in god.

11. Predictions and model selection, again

Coyne explains the evidence for evolution form nested hierarchies. He then makes the related argument that evolution fits better with the existence of nested hierarchies than intelligent design creationism. Barnes complains hat, surely, Coyne cannot know what an intelligent designer would or would not have done, so therefor nested hierarchies are poor evidence for evolution. However, this confuses the hypothetico-deductive argument that evolution necessarily predicts nested hierarchies with the abductive argument concerned with model selection between evolution and creationism. The potential weakness in the latter does not affect the former. Imagine a criminal defending himself against an evidentialist attack from a forensic scientist by saying that the easter bunny killed the victim and framed him. Clearly, it is hard for the forensic scientist to argue on what the easter bunny would or would not do, but that has no bearing on the fact that the evidence suggests that the criminal is guilty.

The analogy with technology has to do with technology evolving in the sense that you keep what works and build from that. It is an analogy with natural selection, not with common descent, as you can group e. g. cars any way you want, but there are no non-vertebrate mammals or seedless flowering plants. See the discussion of nested hierarchies above.

12. Level of Optimization and Rhetorical Questions

This is yet another case of the above confusion between abductive and hypothetico-deductive arguments for evolution. First, few creationists believe that the creator optimizes ecosystems rather than organisms. Also, the existence of large-scale global extinction events suggests this is not the case either. The argument from bad design have one hypothetico-deductive formulation (evolution predicts it) and one abductive formulation (evolution fits better with suboptimal designs than creationism). Potential problems with the latter does not affect the former.

The same problem occurs in the discussion of the pseudogene for synthesizing vitamin C in humans. It is a strong piece of evidence for evolution, and it is irrelevant what a designer would or would not do. Furthermore, it is entirely possible to accept evolution, but think that this was the way the designer did it. But Coyne’s general argument against creationists is this: it makes more sense on evolutionary common descent, than creationism (which would require independently inactivating mutations without any clear design intent).

13. False and unfalsifiable

Barnes complains that Coyne simultaneously claims that intelligent design creationism is false and unfalsifiable and thinks this cannot be the case: “A theory cannot be both unfalsifiable and falsified”. The problem here is two-fold. Intelligent design creationism is not a theory. In science, a theory is a strongly evidence-based explanation to some part of the natural world, that can include facts, laws, inferences and tested hypotheses. The reason that intelligent design is not a scientific theory is that it does not provide a testable mechanism (explanation). So intelligent design is unfalsifiable precisely because it does not provide an explanation. The negative attacks on evolution that many intelligent design creationists do, however, are false. They are false, because the evidence supports evolution.

The positive case for intelligent design does not exist. It does not provide any mechanism or explanation and it does not offer testable predictions from their assumptions (because who knows what a superior cosmic intellect would do?). So the positive case is unfalsifiable. The negative case against evolution repeats many of the same old creationist canards of previous forms of creationism and it is false. These two parts do not form a coherent scientific theory, so there is no contradiction.

Think of it like this: some beliefs of intelligent design creationists are unfalsifiable (e. g. “the designer did it”), whereas others are falsifiable and have been falsified (standard creationists “criticisms” of evolution).

14. Does Coyne correctly characterize creationism?

No, as Coyne is trying to convey that most creationists reject common descent. That is what Coyne means when he says that creationism entails “the instantaneous creation of forms designed de novo to fit their environment” and “Young earth creationists believe … that life has no evolutionary history at all”. Obviously creationists think that there is variation within biblical “kinds”. They do have eyes.

15. Pointless complaining of epic proportions

Barnes quotes Coyne saying “we easily accept that the Grand Canyon resulted from millions of years of slow, imperceptible carving by the Colorado river, even though we can’t see the canyon getting deeper over our lifetime” and asks what Coyne means by “we” as most creationists do not accept an ancient earth. This may be the case, but yet again the context will set you free.

Coyne is clearly talking about the types of creationists that accepts common descent but does not accept natural selection as the main mechanism when he just before that section states that “But many people would like more: they’d like to see natural selection in action, and witness evolutionary change in their lifetime. It’s not hard to accept the idea that natural selection could cause, say, the evolution of whales from land animals over millions of years, but somehow the idea of selection becomes more compelling when we see the process before our eyes. This demand to see selection and evolution in real time, while understandable, is curious”.

Another quotation out of context.

16. The RM & NS strawman

Barnes brings out another classic creationist straw man, namely the notion that evolution is nothing but random mutation and natural selection. However, there are now over 50 different mechanisms other than “random mutation” that generates biological variation. These include gene duplication, exon shuffling, horizontal gene transfer, changes in promotor regions, changes in substrates or products of biochemical pathways, combination of pathways, various forms of genetic recombination etc.

Of course it will sound difficult for just RM and NS to be “sufficiently powerful to produce the tree of life”, but it is just a straw man.

17. Confusing “if” with “how”

Barnes attempts to refute Coyne’s discussion of the unsolved puzzles of evolutionary biology. The important thing here is to not that Coyne is not denying the existence of unsolved problems within evolution, merely that these problems are of such a nature that they are about how common descent occurred, not whether. In other words, it is a difference in scope. Even though there will be future modifications to the answers to the “how” questions, it is unlikely that there will be any large-scale modifications to the answers to the “if” question.

Let’s take cosmology as an analogy. It is very improbable that cosmologists will disprove the Big Bang theory in the future. To be sure, there will always be a genuine scientific debate over the details (some of which Barnes mentioned). But it is very improbable that scientists will reject the Big Bang (or evolution) in the near future as kaput.

The case of Newtonian mechanics is also illuminating. It is true that Einstein’s theories of relativity superseded Newton’s ideas, but it is still a good approximation to every day objects. So the process was one of determining the interval for which Newton’s equations were good approximations, rather than a whole-sale falsification or rejection. This is why undergraduates in physics are taught how to apply Newtonian physics in undergraduate physics, but medical students are not taught that bloodletting is a viable treatment for most medical conditions.

There are plenty of unsolved “how” questions in evolutionary biology, such as the evolutionary origin of sex and aging or the precise mechanisms behind the Cambrian radiation, just to name a few. But these are not “if” questions. They are also not unimportant questions.

This concludes the second part of Barnes’ review. In the final part, he talks about materialism and morality. I will make it short and sweet.

18. Materialism

Barnes writes that

This section is full of holes. It is a simple fact that naturalism and materialism are not part of science. They are worldviews which go beyond science. Take materialism – I agree with Coyne’s definition. If science is the study of the material world, then it cannot conclude that there is no reality beyond the material world – just as if your trap can only catches lobsters, you cannot conclude that there are only lobsters in the lake. Materialism may be true, but it logically cannot be a deduction from science.

It is important to separate philosophical naturalism from methodological naturalism. Philosophical naturalism is the philosophical world-view the position that the natural world is all that exists. Methodological naturalism is a pragmatic principle that states that science has been very successful with limiting its scope to studying ideas that make testable predictions about the natural world and thus we might as well go with the successful method, although of course we cannot know absolutely what the future brings. Science may not be or imply philosophical naturalism, but science is a discipline where methodological naturalism reigns supreme. Methodological naturalism is not a complete rejection of supernaturalism, but rather the humble admission that science is not omnipotent.

The notion that, say, demons cause clogged toilets is a pretty useless idea, because it does not generally solve the problems that, uhm, evidence-based toilet cleaning does. This does not mean that plumbers need to be philosophical naturalists. So science does not a priori exclude supernatural claims, only those claims that do not make any testable predictions. For instance, scientists have been testing (and arguably disproved) the supernatural ideas of remote viewing, dowsing and astrology without needing to say “well, that is false by definition”.

It is true that Coyne does not differentiate between philosophical and methodological naturalism in the book, but he does so in other places.

In this case, a science-stopper refers to ideas that do not make testable predictions (the scientists cannot test the idea further). Physics, even fundamental physics, do make testable predictions. This makes the analogy between supernatural claims and fundamental physics as “inquiry-stoppers” a false one. Even if you do develop the ultimate physical theory, you can still test it in currently untested areas, investigate applications etc. Finally, the difference between “that‘s just what the laws of nature are” and “that’s what God chose the laws of nature to be” is that we have evidence for the existence of the things that the final description of physics describe, but not for the deity in question. We just do not need to postulate the additional entity.

19. Morality

Coyne says that European countries with relatively little religious belief and a fairly high acceptance of evolution have it relatively good. Barnes tries to counter this by pointing to the Finnish school shooter who believed he was carrying out natural selection to eliminate the unfit for the “survival of the fittest”.

However, this is a flawed argument for several reasons: (1) Individuals who go on to commit murder tend to find rationalization for their behavior and these can be related to secularism, religion, politics or science, (2) evolution is a description of what occurs in nature, not a moral theory of human behavior, (3) survival of the fittest has nothing to do with evolutionary biology (the relevant thing is differential reproduction, not survival) and was coined by the philosopher Herbert Spencer, (4) portraying himself as “the natural selector” the man took on the role of a god who got to decide who lives and dies. But of course, there are no intentional entities in evolution that select organisms to reproduce, it is just the differential reproduction and finally (5) a single counter-example cannot overturn the claim that most of Europeans accept evolution, yet are civilized.

Barnes continue to incorrectly characterize Coyne and at the same time confuse many different things:

Coyne argues as follows: because we do not know which (if any) human behaviours are hardwired by evolution, we are free to act morally. This is both incorrect and irrelevant. Just because we have not worked out which human behaviours are hardwired, it does not mean that none are. If we are genetically determined to be selfish, then our ignorance of this fact does not change the fact itself. However, Coyne rejects genetic determinism. I believe he is correct, but he does not give any argument to support this rejection.

Barnes is jumbling issues regarding naive adaptationism, genetic determinism, gene-centric view of evolution and the evolution of moral behavior without separating these issues.

Coyne’s general argument is that evolution is descriptive, not a moral theory and that evolutionary processes have favored the evolution of moral behavior. In other words, “survival of the fittest” is a bad metaphor. He also notes that he should wait before accepting what behavior evolution has selected for before evidence is in (this goes back to his rejection of naive adaptationism above).

Genetic determinism is the view that genetic influences reign supreme and that no amount of environmental influences an change so called “hardwired traits”. Coyne is a good student of Lewontin, a strong critic of genetic determinism, so of course he does not believe in it. All it takes is an easy counterexample, say, the disease known as phenylketonuria.

Barnes also confuses the level and unit of selection. Modern evolutionary biologists do not consider organisms selfish, but rather genes as “selfish” (in the metaphorical sense that genes that tend to produce effects that lead to the increase of reproducing copies, in direct offspring and other related individuals, will increase in frequency over generations). So “selfish” genes can produce organisms that are altruistic.

When biologists talk about the evolution of moral behavior, they do not mean that “X is an evolved behavior” implies “X is morally good”, but rather than evolution can explain the existence of moral behavior and emotions. On top of this we can make methodologically naturalistic moral theories, perhaps something like those of Sam Harris and Richard Carrier.

But it is imperative that we understand that “X is an evolved moral behavior” does not mean that there is no such thing as moral realism. These are just different levels of analysis covering different aspects of morality: “What is the evolutionary origin of moral behavior?” is not the same question as “why should be behave morally?” or “what is moral behavior?”. The existence of an answer to the first one does not mean the latter two questions are irrelevant. A “nothing but” statement has secretly sneaked into the discussion.

“Moral behavior is an evolutionary adaptation” is of course not the same as “moral behavior is nothing but an evolutionary adaptation”. This is no different from the fact that Newton did not really destroy the beauty and wonder of the rainbow as the poet Keats felt. In my opinion, the awesomeness of rainbows becomes even better when we understand the physical principles at work. As Feynman so deliciously put it:

I have a friend who’s an artist and he’s some times taken a view which I don’t agree with very well. He’ll hold up a flower and say, “look how beautiful it is,” and I’ll agree, I think. And he says, “you see, I as an artist can see how beautiful this is, but you as a scientist, oh, take this all apart and it becomes a dull thing.” And I think he’s kind of nutty.

First of all, the beauty that he sees is available to other people and to me, too, I believe, although I might not be quite as refined aesthetically as he is. But I can appreciate the beauty of a flower.

At the same time, I see much more about the flower that he sees. I could imagine the cells in there, the complicated actions inside which also have a beauty. I mean, it’s not just beauty at this dimension of one centimeter: there is also beauty at a smaller dimension, the inner structure…also the processes.

The fact that the colors in the flower are evolved in order to attract insects to pollinate it is interesting — it means that insects can see the color.

It adds a question — does this aesthetic sense also exist in the lower forms that are…why is it aesthetic, all kinds of interesting questions which a science knowledge only adds to the excitement and mystery and the awe of a flower.

It only adds. I don’t understand how it subtracts.

So it does not follow that moral realism is necessarily false on philosophical naturalism. It is only false if you, in secret, adds the “nothing but” clause, but you do not have to.

20. Conclusion

I found many aspects of Barnes review of Coyne’s book to be seriously flawed. Others were nothing more than quotations out of context and basic misunderstandings of the biology. The important distinction between philosophical and methodological naturalism was left out of the discussion and Barnes mentally added “nothing but” to the evolutionary discussion of moral behavior.

Knowledge only adds.