Evolution under the microscope

Why evolution is true
by Jerry Coyne

Why evolution is true

As you’d expect of a book [1] with this title, Jerry Coyne presents a variety of evidence that is consistent with evolution. Some of the evidence he presents is valid, but he gets the following wrong:

I expand on these below (and I apologise for the length of this page!), but first ...

Show page contents.


Towards the end of Chapter 1 'What is evolution' he acknowledges the principle of falsification – that no matter how much evidence supports a theory, there is always the possibility that some new evidence could show it to be false.

A theory becomes a fact when so much evidence has accumulated in its favor – and there is no decisive evidence against it – that virtually all reasonable people will accept it. This does not mean that a "true" theory will never be falsified. ... As we'll see, it is possible that despite thousands of observations that support Darwinism, new data might show it to be wrong. I think this is unlikely, but scientists, unlike zealots, can't afford to become arrogant about what they accept as true. (p16)

This is important because many evolutionists point to the substantial evidence that is consistent with evolution and seem to think that this more-or-less proves it is true. So Coyne's endorsement of the principle of falsification is to be welcomed.

It is also clear from the above quote that he believes scientists are open-minded and more committed to the truth than to a particular theory; so, of course, if any counter-evidence were to come to light then scientists would embrace it and reject the theory of evolution. But he appears confident that such counter-evidence does not exist...

Show page contents.

Evidence contrary to evolution

It is to be expected that throughout a book which sets out to demonstrate that evolution is true Coyne will emphasise the evidence that supports evolution and downplay any contrary evidence. However, downplaying is one thing, to deny that any contrary evidence exists is another. In the final chapter he asserts plainly:

And every fact that has something to do with evolution confirms its truth. (p222)

Despite innumerable possible observations that could prove evolution untrue, we don’t have a single one. (p223)

There is no point in beating about the bush: for Coyne to write this, he must be either woefully uninformed about his subject, or disingenuous. So, this is where I’ll start – with some plain scientific evidence that is clearly contrary to evolution. Using an example from his book, but including contrary evidence which he conveniently overlooks.

Some plain contrary evidence: the non-homologous embryonic development of vertebrae

In chapter 1 Coyne writes:

Let’s examine one evolutionary tree, that of vertebrates (figure 2). On this tree I’ve put some of the features that biologists use to deduce evolutionary relationships. For a start, fish, amphibians, mammals and reptiles all have a backbone – they are "vertebrates" – so they must have descended from a common ancestor that also had vertebrae. (p8, his figure 2 not included here)

This seems a reasonable inference, especially in view of the conservative nature of embryonic development / processes, which he affirms in chapter 3 'Remnants: Vestiges, embryos and bad design':

First, why do different vertebrates, which wind up looking very different from one another, all begin development looking like fish embryos? (p75)

The probable answer – and it’s a good one – involves recognizing that as one species evolves into another, the descendant inherits the developmental program of its ancestor: that is, all the genes that form ancestral structures. And development is a very conservative process. Many structures that form later in development require biochemical 'cues' from features that appear earlier. If for example, you try to tinker with the circulatory system by remodeling it from the very onset of development, you might produce all sorts of adverse side effects in the formation of other structures, like bones, that mustn’t be changed. To avoid these deleterious side effects, it’s usually easier to simply tack some less drastic changes onto what is already a robust and basic developmental plan. It is best for things that evolved later to be programmed to develop later in the embryo.

The 'adding new stuff onto old' principle is just a hypothesis – an explanation for the facts of embryology. It’s hard to prove that it was easier for a developmental program to evolve one way rather than another. But the facts of embryology remain, and make sense only in light of evolution. All vertebrates begin development looking like embryonic fish because we all descended from a fishlike ancestor with a fishlike embryo. (p77-8, original italics, underlining added)

He is right that embryonic processes are conservative, and in the reason he gives as to why they cannot be tinkered with. And this is reinforced the more we find out about how these processes operate at the genetic and molecular level. (Given the number of genes acting in coordinated ways to effect embryonic processes, it would be totally unrealistic to think that they could be modified constructively by means of random mutations.)

However, the embryological facts are that the vertebrae of different classes of vertebrate form embryonically in different ways, even from different embryonic tissues (which is the cardinal criterion for homology). In particular, the vertebrae of tetrapods (e.g. amphibians, reptiles, birds, mammals) form in a radically different way from those of teleosts (most fish). Which clearly shows that tetrapods have not descended from teleosts. For more on this see non-homology refutes a common ancestor of vertebrates.

Hence, just as Coyne argues that similarity of embryonic development indicates common ancestry, the corollary is that different embryonic development militates against common ancestry. To put it another way (to use Coyne’s words), just as similar embryonic development ‘makes sense in light of evolution’, evolution does not make sense in the light of different embryonic development. So, at the very least, this shows that Coyne’s statement ‘every fact that has something to do with evolution confirms its truth’ is clearly untrue! And whilst he says that ‘science continues to accumulate intriguing features about development that support evolution’ (p79), at best that is only one side of the coin, because it is certainly true that science continues to discover features of embryonic development that contradict evolution.

Further, note that this is not an ‘intelligent design’ argument. It is not saying that embryonic development is too complex to have arisen in an evolutionary way (although that is the case). Neither is it arguing that any particular feature appears designed or not. And it certainly isn’t a creationist or God-of-the-gaps argument.

Rather, it is simply observing that non-homologous embryonic development of vertebrae undermines the superficial appearance of vertebrate homology at the morphological level. (And this contrary evidence cannot be dismissed as just a question of not yet knowing how the evolution occurred.) And a direct consequence of this observation is that it shows that vertebrates do not have vertebrae because they have inherited them from a common vertebrate ancestor; i.e. contrary to Coyne’s inference (p8), vertebrates have not descended from a common vertebrate ancestor.

In the quote (p16) I gave earlier, Coyne says that scientists are not zealots, not arrogant about what they accept is true. Presumably he considers himself a good scientist. So, I wonder, in view of this plain evidence against evolution, is he prepared to change his mind and recognise that evolution is false?

Cambrian explosion

Throughout his book Coyne repeatedly tells his readers about the predictions from evolutionary theory that have proved right. But here's another example, mentioned by him, where this is not the case. In chapter 2 ‘Written in the rocks’ Coyne writes:

The [fossil] record confirms several predictions of evolutionary theory: gradual change within lineage, splitting of lineages, and the existence of transitional forms between very different kinds of organisms. There is no getting around this evidence, no waving it away. Evolution happened, and in many cases we see how. (p57)

And in this chapter he gives examples of gradual evolution in the fossil record (pp  29-33):

However, all of these are minor changes which can readily be accounted for by changes in gene frequencies and/or segregation of genes; i.e. without new genetic material arising. And this applies to horse evolution (p52) as well.

But then in his closing chapter he asks:

What caused the Cambrian "explosion" of life, in which many new types of animals appeared within only a few years? (p223)

Why does he ask the question? Because the theory of evolution predicts gradual change: radically new forms of life (with substanitally different body plans) should emerge progressively through the gradual diversification of earlier forms. But the sudden appearance of diverse types of animals in the Cambrian is not gradual change. Neither does it illustrate splitting of lineages; and the absence of transitional or intermediate forms is part of the problem of the Cambrian explosion (from an evolutionary point of view). So, far from confirming the truth of evolution, the Cambrian explosion is a clear challenge. His response is of course that it’s just one of those examples to which we do not (yet) have an evolutionary explanation. Nevertheless, here is further evidence that certainly is not consistent with evolution. And yet, on the same page (can he not see the plain contradiction?) he asserts yet again that

Despite a million chances to be wrong, evolution always comes up right. (p223)[!]

This seems to be something of a mantra for Coyne, perhaps hoping that by repetition his readers will believe it, despite it being evidently false.

Show page contents.

The genetic basis of morphological novelties

A common failing of contemporary evolutionary biologists is that they do not take seriously the inescapable fact that morphological novelties require specific new genetic material:

I expand on this elsewhere, but the three main reasons for this failing are:

It seems that these factors have lulled biologists into thinking that appropriate new genes can and will arise as required (for their proposed evolutionary scenarios), and the actions of these new genes blend seamlessly with those of previously existing genes; to the extent that they can completely ignore these essential genetic requirements and propose evolutionary scenarios solely in terms of morphological variations. (All that we now know about how embryonic development is effected through the orchestrated action of many interdependent genes is completely set aside.)

It is clear that Coyne fully subscribes to this pervasive but deeply flawed Neo-Darwinian belief.

Natural selection

But I’ll start with something he gets right – the action of natural selection.

Selection … is a process, a description of how genes that produce better adaptations become more frequent over time. (p117, emphasis in original)

He emphasises that it is not a chance process:

True, the raw materials of evolution – the variations between individuals – are indeed produced by chance mutations. … But it is the filtering effect of that variation by natural selection that produces adaptations, and natural selection is manifestly not random. (p119, emphasis in original)

And I agree with him that:

Adaptation to the environment is inevitable if a species has the right kind of genetic variation. (p117)

But that is a crucial condition: IF a species has the right kind of genetic variation.

Source of genetic variation

So what is the source of genetic variation? He’s right about this as well:

Where does this genetic variation come from? Mutations – accidental changes in the sequences of DNA that usually occur as errors when the molecule is copied during cell division. Genetic variation generated by mutation is widespread: mutant forms of genes, for example, explain variation in human eye color, blood type, and much of our – and other species' – variation in height, weight, biochemistry, and innumerable other traits.

mutations occur regardless of whether they would be useful to the individual. ... Most are harmful or neutral, but a few can turn out to be useful. The useful ones are the raw material for evolution. But there is no known biological way to jack up the probability that a mutation will meet the current adaptive needs of the organism. (p118, italics in original, underlining added)

However, he seems to overlook this crucial aspect once he starts to consider the evolution of morphological novelties.

It is a key question: Does mutation produce the genetic variation required for evolutionary progress? Yet it seems contemporary biologists, including Coyne, don’t even bother to consider this question. They just assume the answer is Yes, perhaps for the reasons I mentioned above; or perhaps they know that if they really take the question seriously, then it is clear the answer is an emphatic NO (see obstacles to new genes).

Invalid extrapolation

One of the ways Coyne evades the question is to extrapolate invalidly from the success of selection. He gives several examples of selection, both artificial and natural. One is dog breeding.[a]

dog breeds

... there was ample variation in color, size, shape and behaviour in the ancestral lineage of dogs to make possible the creation of all breeds. (p126)

And he comments:

What is most astonishing about dog breeding is how fast it got results. All those breeds have been selected in less than ten thousand years, only 0.1 percent of the time that it took wild dog species to diversify from their common ancestor in nature. If artificial selection can produce such canine diversity so quickly, it becomes easier to accept that the lesser diversity of wild dogs arose by natural selection acting over a period a thousand times longer. (p126)

This is the sort of extrapolation that Darwin made, and is why he used domestic breeding to introduce the concept of natural selection. What is astonishing is that a professional biologist in the 21st century should make what we now know to be a clearly invalid extrapolation! We know very well that diversification can happen readily and rapidly through the segregation of genetic variation; not just artificially, but naturally, such as the Galapagos finches, Hawaiian snails, and cichlid fish. Coyne himself comments that dog breeding has simply made use of the inherent variation in the wild dogs from which they are derived. But how did the genetic variation arise? Not primarily through selection, but primarily through mutation (and remember selection cannot guide to a useful gene, it can only recognise one when it occurs). That is, he is trying to extrapolate from one type of process (selection) to a completely different type of process (mutation) – which is clearly invalid. So no matter how successful the subsequent selection to derive the different breeds, it is completely misguided to extrapolate from that process of segregating existing genetic variation to the process of producing genetic variation through mutation. Regardless of how much evidence there is for selection based on existing genes, it has no bearing on – does not inform at all – the question of whether mutation can produce useful genes.

As well as this invalid extrapolation, Coyne illustrates the lack of clear thinking that seems to be so common among evolutionary biologists. He refers to the lesser diversity of wild dogs; but that is primarily, if not only, in terms of morphological diversity. They must have (had) the wide genetic diversity that Coyne has already mentioned (first quote from p126), otherwise the selective breeding – whether slow or rapid – would not have worked. (For further comment on comparing genetic and morphological diversity see the evolution of horses.) So it is not just a matter of producing some limited genetic variation, but the whole spectrum that has been utilised in dog breeding.

This invalid extrapolation pervades Coyne’s ‘reasoning’.

To really see the power of selection, we must extrapolate the small changes that selection crafts in our lifetime over millions of years that it has really had to work in nature. (p143)

And, in his section about ‘animal and plant breeding’ he suggests:

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. (p135)

It’s not hard to accept, provided you ignore the problems associated with originating useful genes, and gloss over them as just a matter of selection. And he concludes his section on the evolution of whales with:

All of its benefits [moving from land to sea] were only a few mutations away. (p52)

This is what I mean by not taking seriously the genetic implications of morphological novelty! When Coyne makes a statement as ridiculous as that, it’s hard to take him seriously!

Can selection build complexity?

This is one of Coyne's headings (p136). Yet again he's reduced the question to one of just the action of natural selection, which we know works; but the essential prerequisite for the action of natural selection – the availability of appropriate genetic variation – is completely ignored. And it's clear that Coyne knows he’s on shaky ground because of the excuses he starts off with:

... complex features take a long time to evolve, and most of them did so in the distant past … (p136)

But first we must ask: What's the alternative theory? We know of no other natural process that can build a complex adaptation. (p136)

So, unfortunately, having correctly recognised that the action of natural selection is dependent on the availability of appropriate genetic variation, he now seems to forget this and proceeds to discuss the possible evolution of complex organs only in terms of natural selection acting on morphological variations. In effect, he simply assumes that the appropriate morphological variations will arise as desired, with absolutely no consideration given as to whether the necessary genetic bases for such morphological variations could arise. As I said at the beginning of this section, this is a common / pervasive oversight by evolutionary biologists.

This flawed approach is amply illustrated in his description of the evolution of eyes.

The evolution of eyes

eye sequence

This is the sort of progression
envisaged by Darwin, and proposed
by Nilsson and Pelger in 1994. [b]

The eye is of course the classical example of a complex organ, and Coyne asks,

… is natural selection sufficient to explain a really complex organ, such as the eye? (p141, emphasis in original)

After first mentioning how Darwin envisaged an eye might have evolved, he cites the well-known model by Nilsson & Pelger. [2]

… starting with a patch of light-sensitive cells backed by a pigment layer (a retina). They then allowed the tissues around this structure to deform themselves randomly, limiting the amount of change to only 1 percent of size or thickness at each step. To mimic natural selection, the model accepted only “mutations” that improved the visual acuity, and rejected those that degraded it. (p142)

The phrase ‘deform themselves randomly’ illustrates so clearly that Nillson & Pelger, and all contemporary biologists such as Coyne who accept their model, have not moved on from the 19th-century thinking that biological tissues are innately plastic, able to deform in practically any /all ways. Darwin can be excused for holding such a view because there was no knowledge then of the genetic and biochemical mechanisms behind embryonic development of tissues. But we now know that for models such as Nilsson & Pelger’s to be valid would require the right genes (and combinations of cooperating genes) to arise at the right time to effect the modelled changes. And, as Coyne emphasised previously (p118), there is no way an organism can anticipate or promote the production of new useful genes; they must arise by chance, and will only by retained if they actually (not potentially) confer an advantage. (And a part of this requirement is that part-formed genes are of no use; genes will be of use only when enough of their sequence is right so that they are functional, which is what makes it prohibitively improbable that they will arise by chance. See obstacles to new genes.)

It’s not even as if biologists recognise this requirement and have reasonable grounds for thinking that appropriate genes would arise (the only reason offered by Coyne is that we know evolution is true!). In most if not all cases the requirement for appropriate and specific genetic material is not considered at all. There just seems to be a widespread naïve belief in the power of random mutations to produce the specific genes required to effect the specific morphological changes implied by these models. (At least Richard Dawkins is clear about this assumption when he says that evolution can be effected through very small changes to embryonic development. It’s just that this can’t work in practice because of the discrete nature of genes, compounded by the obligatory interdependent action of many genes in the course of embryonic development.)

The failure to take seriously the need for new genes is the fundamental failing of Nilsson & Pelger’s model. But their other serious failing is to use an equation that describes the effect of artificial selection to predict the effect of natural selection! Not only that, but their constant proportional improvement per generation means that the improvement will increase geometrically (resulting in very large changes in the later stages of their model) – which surely should have alerted professional biologists reading their paper that something was seriously wrong. But the disturbing fact is that biologists such as Coyne have swallowed this evidently flawed model hook, line and sinker:

And mathematical models show that natural selection can produce complex features easily and quickly. (p143)

Show page contents.


Just as Richard Dawkins starts The Blind Watchmaker by saying that nature looks designed, so does Jerry Coyne:

The more one learns about plants and animals, the more one marvels at how all their designs fit their way of life. What could be more natural than inferring that this fit reflects conscious design? (p3)

But, of course, also like Dawkins, he says that nature has the appearance of design, not because it was actually fashioned by a designer, but because its various features have been moulded mindlessly by the guiding hand of natural selection acting on randomly generated variations. I agree with Coyne (and Dawkins) that we should not accept the prima facie appearance of design, and that we need to examine whether or not the apparently designed objects could have arisen through natural processes. We just disagree about the efficacy of his proposed solution. We disagree not about the action of natural selection – whether it can favour advantageous variations – but whether natural processes can produce the advantageous variations upon which natural selection can act (mutations can produce innumerable variations, but not the required advantageous ones). As the renowned early geneticist Hugo de Vries said:

Natural selection may explain the survival of the fittest, but it cannot explain the arrival of the fittest. [3]

However, the point I want to make here is a bit different.

On one hand Coyne says that it is not the appearance of design that matters, but whether natural processes can produce the apparently designed objects (which I agree with). But on the other, his book is littered with examples of what he says appear to be ‘bad design’ – not only in his chapter 3 'Remnants: Vestiges, Embryos, and Bad design' – which he sees as evidence against intentional design. But he should be consistent. If we should not infer the existence of a designer from the mere appearance of good design, then neither should we infer the absence of a designer from apparent bad design. The important question is not whether something appears to us to be designed or not (or whether our subjective perception of good/bad design is reliable or not), but whether natural processes could produce the relevant object, even if it appears badly designed. As Paley noted, even a watch that didn’t keep good time, would nevertheless require a designer and watchmaker. [4] So Coyne's umpteen references to apparent bad design carry no weight at all.

Intelligent design

And here is Coyne’s convincing argument for dismissing intelligent design:


Show page contents.

Rationality and Naturalism

The main purpose of this website, and in writing Evolution under the microscope, is to expose the theory of evolution to better scientific scrutiny. This is because contrary evidence is usually glossed over (e.g. with ad hoc explanations for the sudden appearance of new phyla in the fossil record), or ignored altogether (e.g. the obstacles to evolving new genes, or non-homologous embryonic development). That is, my primary aim is not to challenge naturalism. However, if evolution is false then it probably does challenge naturalism, because there does not seem to be a naturalistic alternative – as Dawkins says in The Greatest Show on Earth, evolution is 'the only game in town'.

On the other hand, it is clear that one of Coyne’s objectives is to advocate naturalism.

Naturalism is the view that the only way to understand our universe is through the scientific method. Materialism is the idea that the only reality is the physical matter of the universe, and that everything else, including thoughts, will, emotions, come from physical laws acting on matter. The message of evolution, and all of science, is one of naturalistic materialism. … Now, science cannot completely exclude the possibility of supernatural explanation. It is possible – though very unlikely – that our whole world is controlled by elves. But supernatural explanations like these are simply never needed: we manage to understand the natural world just fine using reason and materialism. Furthermore, supernatural explanations always mean the end of enquiry: that’s the way God wants it, end of story. Science on the other hand is never satisfied: our studies of the universe will continue until humans go extinct. (pp224-5, my underlining)

The crux of the issue is whether or not natural processes are sufficient to account for the origin of biological organisms. Whilst Coyne claims they are, it will be apparent from the foregoing examples that in fact there is substantial evidence that is difficult to reconcile with this. Some evidence poses a formidable challenge, and some is plain contrary evidence which prima facie falsifies evolution. Indeed, it seems the only reason for not rejecting evolution in the light of such evidence is commitment to naturalism rather than an objective assessment of the evidence. Coyne may try to dismiss questioning naturalism as superstition e.g.

The battle for evolution seems never-ending. And the battle is part of a wider war, a war between rationality and superstition. (p xiii)

or as fanciful as believing in elves (above quote, p225). But such comments do not get rid of the evidence against evolution.

Coyne appeals for rationality; but, I suggest, he should consider whether his approach is rational.

Rationality is the quality or state of being rational – that is, being based on or agreeable to reason. [5]

Reason is the capacity for consciously making sense of things, establishing and verifying facts, applying logic, and changing or justifying practices, institutions, and beliefs based on new or existing information. [6]

Naturalism is a presumption of science. Indeed, I cannot see how a scientist can even plan an experiment without presuming naturalism; which is why I accept methodological naturalism (unlike many ID proponents), provided it means what it says – it is the method of doing science. However, the presumption of ideological naturalism should be subject to the evidence. Whilst naturalism may constrain the methods of science, it should not constrain permissible conclusions.

It is the stubborn refusal to reject evolution despite the contrary evidence – allowing ideology to cloud scientific judgement (a clear example is the widespread acceptance of Nilsson and Pelger's model for eye evolution, despite it being so seriously and plainly flawed) – that is doing the sort of harm to science that Coyne is so quick to accuse others of doing.

Although Coyne says it is a war between rationality and superstition (p xiii), perhaps in reality it is more a conflict between rationality and ideological naturalism. And the question to Coyne is: is he actually prepared to embrace rationality and be willing to change his beliefs, including about naturalism, in the light of the substantial evidence against evolution?



Notes display in the main text when the cursor is on the Note number.

1. Jerry Coyne. Why evolution is true. Viking Penguin, 2009.

2. Dan-E. Nilsson and Susanne Pelger, A pessimistic estimate of the time required for an eye to evolve, Proceedings: Biological Sciences, Vol. 256, No. 1345 (Apr. 22. 1994), 53-58.

3. Hugo de Vries. Species and Varieties: Their Origin by Mutation, based on lectures that he delivered in 1904 at the University of California, Berkeley.

4. William Paley. Natural Theology, Chapter 1: State of the argument. 1802.

5. Wikipedia: 'Rationality', based on definitions from mirriam-webster.com. Accessed 14 February 2018.

6. Wikipedia: 'Reason', citing Nikolas Kompridis, 'So we need something else for reason to mean', International Journal of Philosophical Studies, Vol 8 (3), 2010. Accessed 14 February 2018.

Image credits

a. Image © Junojess | Dreamstime.com

b. Image from Dan-E. Nilsson and Susanne Pelger, A pessimistic estimate of the time required for an eye to evolve, Proceedings: Biological Sciences, Vol. 256, No. 1345 (Apr. 22. 1994), 53-58.

Page created March 2018.