A complex adaptation: the eye



Alan R. Rogers wrote a book The Evidence For Evolution. In this book he has a section on the evolution of the eye, a chapter on the evidence for common descent in support of the evolution of the eye, and some scattered comments about what we know about eye evolution and how we know it. I take up the discussion in Chapter 4. Can Evolution Explain Design in the section A complex adaptation: the eye on page 37.

Rogers begins by describing some of the intricate inter-working of various elements required for proper function and concludes the opening paragraph with the statement that “If any one of these components had failed, you would not be reading.” He opens the next paragraph with a question “How could all this have evolved by natural selection.” He states that “…the eye has always been central to the argument against his [Darwin’s] theory.”

I want to pause here and say that this is the question: “How could all this have evolved by natural selection?” This is important because I want to ask whether Rogers actually ever answers this question. It certainly appears from other statements that he makes in the book that he thinks the matter has been settled. Take for example the following:

“…we understand in considerable detail how eyes were assembled by natural selection.” (p. 49)

“We saw on p. 42 that a complex eye can evolve from an eye spot in about 360,000 years.” (p. 80)

Rogers references two “critics” of Darwin’s theory who objected to Darwin’s claim that natural selection could create the eye, Charles Pritchard and Joseph John Murphy. He characterizes their arguments as being about plausibility. And of such arguments he says:

The nice thing about such arguments is that one can attack them without any data at all, merely by inventing a plausible story about how complex eyes might evolve. The story only needs to be plausible; it does not need to be true. Thus, the first piece of Darwin’s argument was a made-up story about how eyes might have evolved. I often tell my own version to introductory classes.

Rogers then shares his own story, of which he later writes:

This story is of course a fabrication. The cartoons in Figure 4.3 are just that–cartoons. By what right do I claim the story is plausible? Could creatures with eyes like those ever live? The answer to this question is an unambiguous “yes,” for all these sorts of eye can be found in creatures alive today.

Notice the subtle shift from “how could all this have evolved by natural selection” to “could creatures with eyes like those ever live.” Presumably we are being asked to reason that if creatures with eyes like these can live, then it must be the case that all this could have evolved by natural selection. That, of course, is a non-sequitur. The question is not whether creatures like this can ever live, the question is whether evolution took that path from one sort of “eye” to the other and whether that can be attributed to natural selection.

I also note the potential conflation of two distinct arguments. One is a response to a plausibility argument and the other is a claim about what actually happened. I take from this that when examining these sorts of arguments we need to attempt to discern just what it is that the argument is intended to show. Countering the “implausibilty” argument does not answer the question of what actually happened. Rogers acknowledges this:

So far I have argued more or less as Darwin did, and I hope I have convinced you that eyes might plausibly evolve. This is enough to demolish the arguments of Pritchard and Murphy, but it does not tell us whether eyes really did evolve. We cannot answer that question by making up stories. We need real evidence - evidence that Darwin did not have. If eyes did evolve, then closely related species should have similar eyes. Their eyes, in other words, should show traces of common descent.

Notice that we have yet another subtle shift. A shift from the mechanism of evolution to the subject of common descent, as if the truth of common descent can establish the mechanism. Even if common descent is true that does not compel the conclusion that “this all evolved by natural selection.”

I hate posts that are too long and this one is already too long, so I will end here.


Rogers admits he argues “more or less as Darwin did.” How is that not “Darwinian” and why is that not evidence against the proposition that “Darwinism is dead”? What could be more Darwinian than arguing as Darwin did?


I have stated previously that evolution by natural selection is the only game in town if one seeks to explain “the appearance of design” in complex biological adaptations such as the eye.

Has evolutionary theory managed to divorce itself from trying to address “the appearance of design” or is that still a major component of the theory?

Would anyone take seriously a claim that the vertebrate eye is explained by Kimura’s theory?

(S. Joshua Swamidass) #4

It is not a component of the theory. Science is silent about divine design.

Certain features of the eye are certainly explained by Kimura’s theory, but not everything. A good test of comprehension is to enumerate what will be explained by it, and how we can test it.

(Bill Cole) #5

What features do you think are explained by neutral theory?

(Vincent Torley) #6

Hi @colewd, @Mung and @swamidass,

Some years ago, I wrote an article on Uncommon Descent that was a take-down of Nilsson and Pelger’s 1994 paper on the evolution of the eye. Here it is:

Did the eye really evolve in a geological blink?

In 2013, however, Nilsson authored a much better paper, titled, Eye evolution and its functional basis, which traced the evolution of opsins over time, and which contained a detailed model of how the eye evolved in all animal phyla.

Here’s another paper, by Ramirez et al., which I think might also interest you, as it’s recent (2016) and it deals with opsin evolution:

The Last Common Ancestor of Most Bilaterian Animals Possessed at Least Nine Opsins


(S. Joshua Swamidass) #7

Great to see you again @vjtorley. I hope to see you more.

(Robert Byers) #8

Indeed does the “truth” of common descent demonstrate the mechanisn of selection/mutations to create the eye?! No! It must show reasonable steps from a-b.
I don’t know if the eye is more complex then loads of things in our bodies!
The point about it , as stated, is how does a incomplete eye get selected to get better for helping a population survive.
it sure seems unlikely a positive mutation added to the complexity of a half-way there eye and this was selected on and the process continuing. it makes evolutionary biology look like a fantascy and not just a error.

Yes i have read about light sinsitive things that seem halfway to great eyes.
yet even they are fully working.
No way around it. Evolutionism must always have light/sensitive things fully working as eye like.
This means a trail of evolution must follow a working light sensitive thing.
no sudden mutations can be invoked.

(Alan R Rogers) #9

Thanks for the interest in my book. I think the thrust of Mung’s argument is in this quote:

Notice the subtle shift from “how could all this have evolved by natural selection” to “could creatures with eyes like those ever live.” Presumably we are being asked to reason that if creatures with eyes like these can live, then it must be the case that all this could have evolved by natural selection. That, of course, is a non-sequitur.

This conflates two arguments: one about whether it’s plausible that eyes might evolve and the other about whether they actually did. Many have claimed that the vertebrate eye could not plausibly evolve. To refute this argument, I discussed a sequence of intermediate stages through which eyes might plausibly have evolved. To support my claim that this sequence was plausible, I pointed out that these intermediate forms all exist in the world today.

To this point, the argument was about plausibility. I did not (as Mung suggests) claim that if creatures with eyes like this can live, then they must have evolved by natural selection.

The second part of my chapter moves beyond the question of plausibility to test the hypothesis that eyes evolved. We test hypotheses by asking what they imply about things we can observe. If eyes evolved, then related organisms should have similar eyes. And indeed they do. All evidence is consistent with the view that eyes evolved.

Yet this does not constitute a logical proof that eyes did in fact evolve. To claim otherwise would be to commit what philosophers call “the fallacy of affirming the consequent.” We can prove that hypotheses are false but never that one is unambiguously true. It is always possible that some other hypothesis might account for the data. Nonetheless, the hypothesis that eyes evolved has withstood many tests and is as well established as any scientific hypothesis.

There is one more piece to the argument, which didn’t make it into my book. If the vertebrate eye did evolve, then it must have evolved by natural selection, because no other evolutionary force is capable of assembling a complex adaptation.


@AlanRogers, thank you for your comment. I absolutely agree that this is the missing piece from your argument.

I agree that eyes evolved. The book makes a good case that eyes evolved. But it did not show that they evolved by natural selection. The conclusion that they evolved by natural selection does not follow from what you wrote in the book. We have a thread here now on eye evolution and perhaps you could find some time to contribute?

(Bill Cole) #11

Has natural selection been established as a capable mechanism for complex adaptions?

(Timothy Horton) #12

Yes, it has. You can read about it in pretty much any decent Biology or Genetics textbook.

(Alan R Rogers) #13

I agree with Timothy Horton. It’s easy to understand how selection can favor an increase in complexity. The “irreducible complexity” argument is a red herring that was refuted by Darwin and more recently by Ken Miller and others.

The harder issue is about evolving from one complex adaptation to another. These are called “peak shifts” in the literature. Some evolutionists (e.g. Dawkins) would argue that these never happen–that the adaptations we see are those that can evolve via a sequence of individually-adaptive changes. Others (e.g. Sewall Wright) have argued that peak shifts do happen. This is still a subject of debate within evolutionary biology.

(Alan R Rogers) #14


Rogers admits he argues “more or less as Darwin did.” How is that not “Darwinian” and why is that not evidence against the proposition that “Darwinism is dead”? What could be more Darwinian than arguing as Darwin did?

I’m in full agreement with this post. But I have a feeling that I’ve missed the beginning of the conversation. Is someone claiming that Darwinism is dead?

(Bill Cole) #15

Whats not easy to see is the origin of the functional information that is required for the complexity that can be selected for. How many organized DNA nucleotides does it take to build an eye. Conservatively at least 1 million. How many organized nucleotides does it take to build a light sensitive spot. Conservatively at least 100k. How you get to the points of selection is not very clear. You have to explain the origin of the light sensitive spot just to get your argument started.

(Bill Cole) #16

You didn’t get the memo :grinning:

(Alan R Rogers) #17

I’m no expert on nerves, but I read that most nerves are sensitive to light. If so, then it’s easy to understand how that light-sensitive spot started. All you need is nerves near the surface.

The argument about information is more involved, and I will save it for later.

(Timothy Horton) #18

If you start with a rudimentary nerve connected to a surface cell with the limited ability to detect temperature differences (hot and cold i.e. infrared radiation) evolution just needs to widen the wavelengths of energy which can be detected into the visible light range.

Get ready for the goal post move and the demand for how any nerve cells at all evolved.:slightly_smiling_face:

(Timothy Horton) #19

You mean it’s not easy for you, personally, to see since you’ve never studied the topic. Evolutionary biologists and geneticists can see the process which produces and amasses new functional information quite clearly.

(Bill Cole) #20

Great. I think functional information is important in understanding Behe and other design arguments. Kirk Durston and Joshua have been discussing it on this blog for several weeks.Durston: Functional Information