The Edge question: two bad answers about evolution

January 23, 2014 • 6:52 am

This year’s Edge question, which my agent John Brockman distributes to a stable of scientists and other scholars, is given below, and you could do worse than scanning the 174 answers and reading the ones that intrigue you (the short titles of the posts are the answers to the question):


Ideas change, and the times we live in change. Perhaps the biggest change today is the rate of change. What established scientific idea is ready to be moved aside so that science can advance?

Some answers I really like, others I disagree with but still find provocative, while still others are, I think, misguided. Two in the last class are about evolution.

The first is by Roger Highfield, editor of New Scientist from 2008-2011 and now “Director of External Affairs of the Science Museum Group.” While he was editor, New Scientist got a reputation for overblowing new findings about evolution and using them to cast doubt on the entire neo-Darwinian paradigm. This was most prominent in New Scientist‘s “Evolution is Wrong” cover, which I posted about here, and to which several of us responded by sending a letter to the journal. “Darwin was Wrong” was simply wrong.

At any rate, I see that Highfield, who has provided an answer to the Edge question, is still on his crusade against the widespread acceptance of neo-Darwinism, for the idea he wants to retire—as given in the title of his post—is “Evolution is true.” In other words, he’s suggesting that modern ideas about evolution are wrong. I previously made a comment about Highfield’s piece on this site, but let me reprise and expand it (excerpts from Highfield’s piece are indented):


I had a severe attaack of gastritis this morning when I saw Roger Highfield’s answer to the Edge Question. What he wants to discard is the idea that “evolution is true”.

Here’s part of what he says:

If evolutionary biologists are really Seekers of the Truth, they need to focus more on finding the mathematical regularities of biology, following in the giant footsteps of Sewall Wright, JBS Haldane, Ronald Fisher and so on.

The messiness of biology has made it relatively hard to discern the mathematical fundamentals of evolution. Perhaps the laws of biology are deductive consequences of the laws of physics and chemistry. Perhaps natural selection is not a statistical consequence of physics, but a new and fundamental physical law. Whatever the case, those universal truths-‘laws’-that physicists and chemists all rely upon appear relatively absent from biology.

Little seems to have changed from a decade ago when the late and great John Maynard Smith wrote a chapter on evolutionary game theory for a book on the most powerful equations of science: his contribution did not include a single equation.

Yet there are already many mathematical formulations of biological processes and evolutionary biology will truly have arrived the day that high school students learn the Equations of Life in addition to Newton’s Laws of Motion.

Moreover, if physics is an example of what a mature scientific discipline should look like, one that does not waste time and energy combating the agenda of science-rejecting creationists, we also need to abandon the blind adherence to the idea that the mechanisms of evolution are Truths that lie beyond discussion.

Highfield’s problem seems to be that he wants mathematical laws that describe how evolution proceeds as accurately as mathematical laws now characterize physical processes. Well, that won’t happen because evolution is contingent on unpredictable (and unknowable) things like new mutations, environmental change, and the invasion of predators and parasites. While in many ways evolution is just as deterministic as physics (and could even be somewhat indeterministic if new mutations involve quantum effects), we simply won’t be able to know or measure the things that can be enfolded into general “laws” of evolutionary biology. While I applaud the mathematization of my field, I contend that the important principles of that field will largely remain nonmathematical, and can be expressed in a few sentences:

1. Evolution happens: populations change genetically over time.

2. That change is gradual and transformative rather than instantaneous.

3. Lineages also split, creating the diversity of life on Earth today from the first Ur-organism (and yes, there is, rarely in eukaryotes, horizontal movement of genes, which was the basis of New Scientist’s claim that “Darwin was wrong”).

4. That splitting of lineages is what creates common ancestry, so that any pair of species on Earth had a common ancestor at some time in the past.

5. The “designoid” features of organisms arose through the process of natural selection, although random processes like genetic drift can cause evolution (but not the appearance of design).

These are things about evolution that are true.

What we need to abandon is not the idea that “evolution is true”, but Highfield’s claim that a “mature” scientific discipline of evolutionary biology will rest on the same type of predictive equations as does physics. It would be futile, for instance, to try to use “evolutionary laws” to predict what will happen to the gene pool of, say, the African elephant or the white oak. Evolution, like nearly all fields of biology, will inevitably become more mathematical, but to say that it’s “not true” because it doesn’t comprise a series of inviolable laws shows a deep misunderstanding of both evolution and biology


Kevin Kelly is a smart guy who’s had an interesting career. He was editor of the Whole Earth Catalogue (remember that?), founder of Wired Magazine, and someone who writes extensively on technology and other subjects. A bit over three years ago, I reviewed his book What Technology Wants for the New York Times, and found it less than satisfactory.

But what’s even less satisfactory is his response to the Edge question, for what Kelly wants tossed into the dustbin of science is the idea of “Fully random mutations.” Here’s part of what he says:

What is commonly called “random mutation” does not in fact occur in a mathematically random pattern. The process of genetic mutation is extremely complex, with multiple pathways, involving more than one system. Current research suggests most spontaneous mutations occur as errors in the repair process for damaged DNA. Neither the damage nor the errors in repair have been shown to be random in where they occur, how they occur, or when they occur. Rather, the idea that mutations are random is simply a widely held assumption by non-specialists and even many teachers of biology. There is no direct evidence for it.

On the contrary, there’s much evidence that genetic mutation vary in patterns. For instance it is pretty much accepted that mutation rates increase or decrease as stress on the cells increases or decreases. These variable rates of mutation include mutations induced by stress from an organism’s predators and competition, and as well as increased mutations brought on by environmental and epigenetic factors. Mutations have also been shown to have a higher chance of occurring near a place in DNA where mutations have already occurred, creating mutation hotspot clusters—a non-random pattern.

While we can’t say mutations are random, we can say there is a large chaotic component, just as there is in the throw of a loaded dice. But loaded dice should not be confused with randomness because over the long run—which is the time frame of evolution—the weighted bias will have noticeable consequences. So to be clear: the evidence shows that chance plays a primary role in mutations, and there would be no natural selection without chance. But it is not random chance. It is loaded chance, with multiple constraints, multi-point biases, numerous clustering effects, and skewed distributions.

So why does the idea of random mutations persist? The assumption of “random mutation” was a philosophical necessity to combat the erroneous earlier idea of inherited acquired traits, or what is commonly called Lamarckian evolution. As a rough first-order approximation, random mutation works pretty well as an intellectual and experimental framework. But the lack of direct evidence for actual random mutations has now reached a stage where the idea needs to be retired.

This is, as the physicists say, “not even wrong.” For Kelly gets the meaning of “random mutations” wrong from the outset, and that makes the rest of his essay irrelevant. What biologists mean when they say “mutations are random” is not that they don’t occur more often at some genes than others, nor that some errors in DNA replication aren’t more frequent than other errors.  We’ve known for years that some genes are more “mutable” than others and that some types of lesions are far more frequent than others. Nor do we mean that mutation rates are impervious to environmental factors, which is also wrong (you can jack them up, for instance, by feeding mutagens like ethidium bromide to organisms, or dosing them with X-rays).

What we mean by random mutations is simply this:

“The chance of a single mutation being “adaptive” for the organism (i.e., promoting the replication of the gene in which it occurs) does not depend on the environment in which it finds itself.”

In other words, the genome does not somehow “know” that a mutation would be adaptive, so that it can produce a higher proportion of the good ones when the environment changes in a certain way. When it gets colder, for example, we don’t see more mutations in mammals that give them longer fur or shorter ears. Most mutations are deleterious in all environments, and a few are useful, but the proportion of the useful ones doesn’t increase when the environment changes in such a way that they become more useful.

(If you don’t like the term “random” because it’s confusing, as one commenter feels below, then you might use an alternative term suggested by my friend and colleague Paul Sniegowski at Penn, who studies “adaptive mutation.” He suggests that we replace the sentence “mutations are random” with “mutations are indifferent.” What he means is that the proportion of adaptive mutations is indifferent to the environment encountered by the organism. It means the same thing as “random” as I’ve defined it above, but may be less confusing.)

And there are a lot of data that testify to this, particularly in microorganisms. Futher, we know of no genetic or biochemical mechanism for jacking up the proportion of useful mutations when the environment changes.

There are some controversial data in microorganisms that the overall mutation rate can rise in times of stress, and you can show via population-genetic theory that—in principle—such an increase, by promoting an increased number (not proportion) of adaptive mutations, will keep the mutation rate higher than it would be otherwise. (To use technical jargon, the new “adaptive” mutations will be physically linked on the chromosome to those mutations that simply cause an overall increase in mutation rate. The adaptive mutations, when they rise in frequency, will then drag along, by genetic “hitchhiking,” the mutations that jack up the mutation rates of all genes, so what we get at the end is an evolved increase in the overall mutation rate in response to environmental stress. This is called “adaptive mutation.”)

But that is a very special case, is still (as I said) controversial, and at any rate hasn’t been found in eukaryotes, where the data still say that mutations are random in the way I defined above.

And Kelly’s claim that we cling to an erroneous idea about mutations as a bulwark against Lamarckism is simply wrong. Many scientists who believe in truly “nonrandom” mutations are trying to overthrow the “randomness” paradigm, and aren’t clinging to any dogma! I suspect Kelly got things wrong here because he isn’t trained as a biologist (the idea of “random mutations” is not intuitive) and didn’t ask any evolutionists for feedback before he submitted his answer.

There is no “peer review” of the answers to Edge questions, and I guess I’m the first respondent who has decided to criticize the answers of other respondents.  But, as always, the laws of physics have determined that I must be a pain in the tuchus by criticizing folks who spread misconceptions about evolution.

87 thoughts on “The Edge question: two bad answers about evolution

  1. Like microwave popcorn (or radioactive decay), we can’t say which kernel will pop when, but we can say that 90% will pop within 3 minutes. When physicists can say which atomic nucleus will decay when, let us know.

  2. Biologist should then abandon the term “random mutation”. It’s not wise to use a term that has a completely different meaning for 99% of people.

      1. I like the term “indifferent” and think it could be useful even if it’s only added to “random” (mutations are both random and indifferent) as opposed to substituted for it. It gets to the heart of the confusion about evolution, which is that it’s somehow, some way, directed towards a beneficial end. The Great Chain of Being constantly pops up among the general populace, even the educated part which ought to know better. Something cares all the way down. The universe can’t be indifferent.

        I don’t know how many seemingly pro-evolution advocates I’ve met who — upon discussion — come out with the claim that the theory of evolution supports spiritual progression but, well … it’s a lot. Or seems like it to me.

    1. Just as we abandoned the term “species” after Darwin showed that those were not things separately created by God and immutable but constantly changing, interrelated, and (evolutionarily speaking) with no concrete boundaries.

      Oh, wait.

    2. I think what biologists mean by random is actually much closer to the colloquial usage than the very technical and stringent defintion Kelly is using.

      Few everyday uses of the word random would meet Kelly’s criteria either.

      FWIW, I think random is a better viewed as question of degree rather than absolute black and white. Loaded dice still produce random rolls, the distribution of such rolls is just skewed from the uniformly random distribution give by fair dice.

      1. From Kelly:
        “… increased mutations brought on by environmental and epigenetic factors.”

        Can anybody here confirm that epigenetics correlates with increased mutations? I don’t see how turning some genes on or off could have any effect on mutation rates.

    3. In that case the physicists will have to stop using quantum as well, because almost no layperson knows what it really means. They seam to all think it means ‘big’, eg. ‘quantum leap’ in understanding.

  3. Highfield’s problem seems to be that he wants mathematical laws that describe how evolution proceeds as accurately as mathematical laws now characterize physical processes. Well, that won’t happen because evolution is contingent on unpredictable (and unknowable) things like new mutations, environmental change, and the invasion of predators and parasites.

    Speaking as a layman, I think biologists do a very good job integrating mathematics into evolutionary biology. While I only have a passing familiarity with it, I would cite your field’s studies of genetic drift as a good example: biologists looking at patterns of interited mutation and using math to assess whether those patterns can be explained probabilistically, without a need for any external force (such as natural selection).

    While I agree with you about Kelly getting the entire meaning of ‘random mutation’ wrong, I have less heartburn about his essay. I think his misconception is probably very common and may even be representative of the general public’s understanding. Because of that, he is partially right in that this misconceived definition of random mutation needs to be retired. His essay says to me that we have more work to do in science communication and education.

  4. … shows a deep misunderstanding of both evolution and biology

    Highfield also shows a lack of understanding of physics. Indeed his piece as a whole seems confused enough that he isn’t really sure what he is trying to say.

    The last two paragraphs suggest that he is accusing biologists of having too much confidence in evolution. His argument seems to be that because Newton’s gravity was superseded by Einstein, and because Einstein’s gravity may also not be the last word, therefore we should doubt our understanding of evolution.

    But that misunderstands science: just because some things might be over-turned doesn’t mean that other things cannot be secure. Afterall, no revision of the theory of gravity is going to overturn the basic point that gravity makes the Earth round and not flat.

    1. Agreed.

      I made a similar point a couple of days ago. Einstein will never be wrong….

      And evolution is a fact, just like gravity. And we obviously do not know everything about gravity.

      Here is another way to say it: Evolution will never be wrong, in the same way that General Relatively will never be wrong.

    2. Asimov is worth quoting here too: “John, when people thought the earth was flat, they were wrong. When people thought the earth was spherical, they were wrong. But if you think that thinking the earth is spherical is just as wrong as thinking the earth is flat, then your view is wronger than both of them put together.” Error (and hence truth) comes in degrees!

  5. Highfield should know that all partial differential equations in physics are deterministic, but their measurements are not. This goes back to a fundamental misunderstanding of determinism and its “discontents.”

  6. The bad responses weren’t limited to biology. I noticed a couple of bad ones in the physics category as well.

    Kai Krause, a philosopher, apparently doesn’t understand that the name of a concept is not the concept itself, and seemed to think that since “uncertainty principle” was not the original name in German, that physicists have just been befuddled about the concept all these years.

    And Bruce Parker (also not a physicist) wants to do away with the concept of entropy because he took the very-simplistic description (that it vaguely means “disorder”) too seriously and appears to think that the second law of thermo is broken all the time. He says, “complexity is synonymous with low entropy” (which it’s not) and that we see complexity forming all the time, so clearly we need to replace the law of entropy, right? No.

    1. With reference to Phil Moriarty’s video in a later post: quantum “spin” is not spin (originally, “two-valued quantum degree of freedom” — but in German); quantum “colour” is not colour; usw.


  7. The bad responses weren’t limited to biology. I noticed a couple of bad ones in the physics category as well.

    Kai Krause, a philosopher, apparently doesn’t understand that the name of a concept is not the concept itself, and seemed to think that since “uncertainty principle” was not the original name in German, that physicists have just been befuddled about the concept all these years.

    And Bruce Parker (also not a physicist) wants to do away with the concept of entropy because he took the very-simplistic description (that it vaguely means “disorder”) too seriously and appears to think that the second law of thermo is broken all the time. He says, “complexity is synonymous with low entropy” (which it’s not) and that we see complexity forming all the time, so clearly we need to replace the law of entropy, right? No.

  8. I am curious about your point #4:

    4. That splitting of lineages is what creates common ancestry, so that any pair of species on Earth had a common ancestor at some time in the past.

    When I was studying Taxonomy (almost 40 years ago- I went on to become a layman, so have not kept up) the idea that there might be 7 kingdoms (instead of 5) was news; since then the concept of Kingdoms almost seems quaint as the various classification schemes have been continually revised. Most of the revisions seem to be splitting off groups from “Monera” and “Prokaryota” as we develop technology to investigate these artificial groupings. Most notably new (to me) is the Archaea.

    All of the cladistic trees I have seen recently show only one trunk, typically labeled “life”. Is there any evidence that life only evolved once?

    1. I’ve always been curious about a so-called ‘second genesis’. If the prebiotic conditions were good enough for life to arise once it would make sense that it could happen somewhere else (Earth is a pretty big place), of course only while those conditions persisted. It would also speak to the predictability of abiogenesis if this ‘other’ ‘life’ ended up using the same mechanisms as the assumed ancestor. Who knows, it could have even started using RNA and moved to DNA and been indistinguishable from any other type of life that arose. Cue Gould’s rewind the tape metaphor. Going way back in time, at a certain point, biology turns into chemistry. I’d like to think the inevitability of chemistry under the certain conditions would lead to heavy convergence among prebiotic self-replicators.

      That said, I think most sequences of homologous genes involved in cellular processes of in DNA replication (thought of as the most ancient of genes) are highly conserved among all life, leading to an assumption of a single instance of life arising. Correct me if I’m wrong though, I don’t know too much about microorganism genomes let alone archea and monera replication genes. If you assume a multiple-genesis scenario, you could defend that by invoking sequence convergence as a form of refinement of the gene – they are under pretty heavy selection pressure and have been around a very long time, after all. Seems a bit of a long shot. Maybe someone more well versed in bacterial genetics could expand on this?

      1. And the universality of the mechanisms used to translate and replicate that code, as well as other stuff like the use of L amino acids. It would be remarkable if that stuff represented convergent evolution from multiple origins of life.

        1. That’s my main reason for remaining with the single event idea. Although entertaining the multiple event idea does raise interesting questions about the power and reach of convergence (and inevitability) among far simpler prebiotic replicators!

          1. The population math also argues that the first good replicator would have very quickly dominated all other contenders. As a selective advantage, being able to replicate is huge. How fast could the first good replicator spread in a ripe but pristine world with no competition?

      2. I don’t think the “genetic code” applies to bacteria in the same way as for multicellular organisms. They don’t reproduce sexually, and don’t have DNA, as I recall. Our mitochondria are believed to be ancient bacteria-like organisms that we have incorporated into our cells- this is base on RNA evidence, among other things. Maybe it is possible to do RNA analysis similar to what is becoming commonplace with DNA now?

        1. my mistake- bacteria do have their own DNA, as do mitochondria. Thus, we have two separate genetic histories to be analyzed.

      3. Now if all these “kinds” the creationists keep talking about each had unique genetic material, THAT would be something.

  9. I think your response Kelly’s essay gets a bit imprecise in it’s terminology. I think your point is that the *proportion* of useful mutations does not go up when such mutations are “needed”, but sometimes you say the *frequency* of useful mutations does not go up. The latter would not be true (i.e., the frequency of useful mutations would go up) if environmental stress caused an overall increase in the frequency of mutations, as long as the proportion of useful mutations remains the same. (“Frequency” = # per time; “Proportion” = % of the total)

  10. Do other fields in science have physics envy? There seems to be a growing interest in evoking math to prove everything; interestingly, physicists know that just because it works in a mathematical model, that doesn’t mean that’s how it works out there in the crazy mixed up world.

    1. It is possible to focus too much on the equations and miss what’s going on in the real world. Observation comes first, unless maybe you’re working at the outer limits of physics where observation is difficult or impossible.

    2. And physics gets a fair share from the creationists. And with all their equations can not predict complex events, i.e. the field has not yet been put to rest.

      As for biology, there are many equations. Yet it general discussions Hard-Weinberg never comes up and most students have a hard time with a dihybrid Punnett square. The world is not ready to teach biology from a mathematical perspective.

    3. As I’ve noted hereabouts before, when I was doing my Ph.D., I was quite facile doing the mathematics and creating the computer models of collider experiments to yield results that we could test against the experimental data from CERN… 


      It wasn’t until years later, outside of academia, when I read David Deutsch’s The Fabric of Reality, that I felt I had any glimmer of an understanding of the physics that the maths described.


  11. With respect, Professor Ceiling-Cat is NOT the first to criticise the submissions to the Edge Question. On a previous letter, I talked of my disappointment with many of the Edge responses, saying that a quick mind could destroy about half of those submissions, and I gave the whole exercise 29% for effort.

    But to the Professor’s great stuff on mutations, above. Absolutely fascinating. And remember, those of us at school and college before the seventies, missed-out on the great age of genetics, and are running to catch-up. I like observation and examples; after all Darwin’s ‘Origins…’ is all observation and anecdote.

    First up is the puzzle on why organisms, animals and plants in stress, may increase the rate of mutation. The mechanism sounds exciting. I would have thought that various stresses such as heat, drought, hunger, being crushed, … would be received by the organism through evolution, as one singular exceptional stimulus – and here’s the clever part- but how could that in practical terms, possibly give rise to an increased rate of mutation? We know it to be an evolved reaction, but I am at a loss to explain how that reaction takes the form of increased rate of mutation.

    On a lighter note, we know that those in war-zones, and while under threat of sudden death, as I have been many times, – perceive the opposite sex at mightily attractive; which lasts until you get you war-bride home!! Hope not to be sexist about this!

    I host several swallow’s nests at my farm each spring, and each pair have two or three broods. On occasion the female takes a new male, who will often tip her earlier chicks out of the nest to die on the floor. But it is my observation that as many as ten percent of chicks have a problem flying, and die on the floor of the barn. I have always assumed it to be an unfortunate mutation in the growth of the brain, rather than simply an accident of chick-rearing.

    1. I also found this round disappointing. Many of the responses seemed to be less about “an idea ready for retirement”, and more “an idea that is inconvenient for my current work”. For example, Sean Carroll’s been thinking a lot about the many worlds model of quantum mechanics latelu. This model (and competing models) have not yet come up with experimental tests to determine which model is a better explanation for what the universe is really up to, or if, in fact, any of them are.

      So, surprise, Sean is convinced falsifiability has to go. One suspects that if he had found an experimental test that could disprove or support many-worlds, he would be busy designing that experiment and much less worried about how falsifiability might be constraining progress in physics.

      1. Strict falsifiability is already dead: auxillary assumptions are needed to test (and hence potentially confirm, to some degree) any large scale theory. This applies to theories in physics (e.g. general relativity) as much as biology (e.g. the various evolution theories). This is not a problem since in a way it is unavoidable (the only alternative, if is one, is to just make stuff up and not care).

        1. That seems to me to be a critique of empiricism in general, not distinctly a critique of falsifiability. And yes, we do pretty much live with it.

          But staying away from “strict falsifiability” for the moment, it’s still a useful idea that it’s very desirable for a theory to make predictions that can disprove the theory through observation or experiment, especially if those predictions differ from predictions of competing theories.

          General relativity and the neoDarwinian synthesis have no difficulty with this aspect of falsifiability.

          Many-worlds quantum mechanics does. None of the theories that incorporate quantum mechanics make testable prediictions different from any of the others. This is a problem with those theories that can’t just be wished away.

          1. That’s not actually the case. MWI is falsifiable. Search for “Everett FAQ” (the link is toxic in WordPress; I’ve never been able to embed it In a comment).


          2. Testable = falsifiable + confirmable, the way I take it. We’d like it everything was testable in that sense, but sometimes we have to settle for one or the other component.

  12. But, as always, the laws of physics have determined that I must be a pain in the tuchus by criticizing folks who spread misconceptions about evolution.

    Oh, come now — you’re not being a ‘pain in the tuchus.’ Someone actually said that the established scientific idea which is ready to be moved aside so that science can advance is “evolution is true” … and you have a very well-known BOOK and popular WEBSITE both titled “Evolution is True.

    Where I come from, they call that a dare. Highfield was asking for it, basically smacking you in the face with a glove and sticking his tongue out. No apologies required.

    If evolutionary biologists are really Seekers of the Truth, they need to focus more on finding the mathematical regularities of biology …

    I interpreted Highfield’s point as a very simple one: empirical truths are not absolute certainties in the way analytical truths are. You can capitalize an analytical truth as a “Truth” because we define the systems and can’t be wrong when conditions are controlled like that. But if it’s based on evidence and reason, then one can never be 100% certain. Science searches for extremely high probabilities, not metaphysical certainties like Truth -with-a Capital-T.

    Well, yes. Duh. But this is usually the part where some smart ass comes out and says therefore atheists can’t KNOW there is no God because their “truths” are contingent but the Truths of God are NOT contingent but completely reliable and absolute blah blah blah. If math and physics can be counted on because they deal with the analytical then “Seekers of Truth” need to use them — AND metaphysics, which is just like physics only undemonstrable to skeptics.

    Highfield may not be going there. But maybe he is. I don’t know.

  13. If anything [from Highfield’s perspective] I would suggest that the mathematics behind evolutionary biology will be best described under fractal geometry because of its randomness.

  14. It always saddens and/or irritates me when people equate the concept of “random” with “uniformly at random.” The prevalence of this error is maybe an indictment of general science and mathematics education. But I have a hard time excusing such an error when it comes from people who should know better; i.e. people who work in the sciences (even if the work is in a journalistic capacity). It’s such an easy misconception to fix. Yet this type of misinformed argument keeps popping up.

    Kelly said, “Mutations have also been shown to have a higher chance of occurring near a place in DNA where mutations have already occurred, creating mutation hotspot clusters — a non-random pattern.” Yet he doesn’t even seem to realize that he’s using the words “higher chance” in this sentence, and that if some event has a higher chance of happening under some condition, then the underlying process is random by definition. He just needs to change the last part of the sentence to “a non-uniformly-at-random pattern.” Of course, no one is claiming that random mutations occur uniformly at random, so he really has nothing to say, as has already been pointed out.

    This type of thing just drives me nuts though. I always thought, I guess naively, that the “common” understanding of the term “random” was something that is “not deterministic.” Yet we see Kelly’s error over and over again. Why does this keep happening?

    1. I don’t know, but it is trivial to demonstrate to anyone who has ever played with dice or cards.

      “Roll a six sided die. If it is 1-4, step in that line. If it is 5-6, go there.”

      Large enough group produces a random nice distribution of people, with roughly the 2:1 ratio, etc.

      We all know this, but this is the sort of demonstration that works well, I think. (To be fair I’ve only used it a few times.) Others can try it out and let me know …

  15. By the way, one of the other Edge responses which drew well-deserved criticism was Alternative Medicine advocate Dean Ornish and his complaint that we need to get rid of “Large Randomized Controlled Trials.” Oh really? A very convenient argument — if RCTs aren’t showing that your favorite bit of woo works.

    It was nicely ripped into by Science-Based-Medicine advocate Orac over at Respectful Insolence.

  16. “The messiness of biology has made it relatively hard to discern the mathematical fundamentals of evolution. ”


    Take a population genetics course. You’ll have all the ‘mathiness’, regularity, and predictiveness you can ask for, let alone discern.

    Or just contemplate Phi at one’s leisure.

  17. physics is an example of what a mature scientific discipline should look like, one that does not waste time and energy combating the agenda of science-rejecting creationists,…

    What a dopey comment. If creationists spent their time trying to sell the concept of a geocentric solar system as and alternative to the heliocentric view, and polluting primary and secondary education with anti-physical nonsense, then physicists would have to waste their time combating the agenda of science-rejecting creationists. And does he really think that the mathematization of biology would shut the creationists up?

  18. This is an excellent post.

    It’s odd that as an example of a mutagen Jerry Coyne picked Ethidium Bromide.

    “by feeding mutagens like ethidium bromide to organisms…”

    And yet EtBr is traditionally used to visualize DNA under UV light…another mutagen. However, both are somewhat weak mutagens, otherwise, they’re utility in molecular biology would be short lived.

  19. I think that Highland’s comments show a person who talks and writes about science but never actually gets his hands dirty doing any. Sure those equations are pretty, but just take a first year college physics lab course and see how much trouble you have to go through to get messy reality to actually conform to those nice simple equations. Then maybe he can try his hand at some engineering. Ohm’s Law sure is nice, but does he have any idea how hard people have to work to try to ensure that the color coded resistor that you put in your circuit really does have voltage and current vary linearly (and only over a small range of voltage and current at that)?

    Biology deals with very complex systems. It’s tough to describe it with simple mathematical models. Sure, there are some physical situations that can be approximated with simple equations. Newton’s laws of motion and law of gravity can give you a pretty good solution for the orbits of the planets in the solar system. But I would love to see Highland give a precise analytical solution for the motion of the hundred billion stars in the milky way under their mutual gravitational attraction. Please show all your work!

  20. Richard Dawkins has a good discussion of what we do and don’t mean by mutations being “random” in Chapter 11 of The Blind Watchmaker.

  21. I’m about halfway through the list so far, there are some good ones but there are some baffling ones too.

    Though I was curious what you though of Martin Nowak’s idea of retiring inclusive fitness. Is it really as dead as he thinks, or does it still have practical utility?

  22. I’ve had a course in population genetics, so I am fairly comfortable with it. However, there was a period when the journal Evolution, publication of the Society for the Study of Evolution, was overrode with population genetics articles, page after page of equations. I got irritated enough that I dropped my membership in the society.

    I think the universality of the genetic code is the best evidence of a single common ancestor for all living things. Reason is that it doesn’t have to be that way. In principle, one could construct a functional organism with a genetic code reading differently that the standard model. That yeasts for example,have several code letters (11 or 8, I forget which) which transcribe differently than the standard model suggests that this could be done.

    The change in number of kingdoms, whatever, is the result of more information on which to make relationship decisions, and a shift from a classification which intermixed considerations of grade level and branching pattern, to a strictly cladistic (in one form or another) analysis of relationships.

  23. Jerry, can I suggest a change to your second point about the truth of evolution, for rhetorical symmetry and completeness?

    2. That change is gradual and transformative rather than instantaneous and discontinuous.

    Something like that…

    (One might also take issue with “transformative”, given the dictionary [NOAD] definition of “transform”: “make a thorough or dramatic change in the form, appearance, or character of” [my emphasis].)


  24. As one who views the idea of “facilitated variation’ quite promising, I found your article very enlightening. Thank you.

  25. I like to point out that Highfield is ALSO very wrong when he states that biology lacks strong “mathematical regularity”. Darwinian mechanics are encapsulated in a number of key equations that shed great light on fundamental biological processes:
    The Replicator Equation which expresses the process of competition, natural selection, fitness and population dynamics and sets out the basis of Evolutionary Game Theory
    The Price Equation that establishes the link between biological characteristics and fitness and the separate influences of selection and heritability/mutation in these characteristics and establishes the resulting population dynamics
    The Lotka–Volterra equation that lay out the dynamics of population when multiple species are competing especially for inter-dependent predator prey situations
    The Quasispecies Equation which models the evolution of replicators related to overall genetic sequences, all having attributes of rates of mutation, growth, and decay and tracks associated dynamics.
    Furthermore all these equations fall into an overall general framework that allows for much inter-derivability; something that physics has not yet achieved.
    Yes- biological systems are very complex- having far more than some first order effects in play. But still, much can be understood by reference to these few general equations.
    Sadly though, Highfield is right in one way- the significance of these equations is often very underplayed in the field of biology.

    1. I’d also point out the complete irony that Martin Nowak , who posted the totally misguided idea in Edge that Kin Selection is an invalid measure, is one of the main authors of the “Unifying Evolutionary Dynamics” paper ( ) which best describes the interrelationship of these equations. In that paper he quite rightly states “the “Price equation” was used by Hamilton in his seminal work on kin selection”. Which all goes to show that sometimes mathematicians can’t follow the implications of their very own calculations.
      And finally I’d like to point out that these very same equations have their expression in the area of Applied Mathematics (my own interest area), where in the study and use of Genetic Algorithms engineers can harness them to seek out effective design adaptations (as indeed, Evolution itself does in Biology).

          1. No. The logic of kin selection is that the effects of a behavioural trait on close relatives can matter for whether that trait is selected, and so, for example, altruistic behaviour can be selected for if it affects close relatives sufficiently more often than random members of the population. This logic, which is not disputable, is *independent* of mathematical formalizations of it, such as inclusive fitness. These formalizations are disputable: one can argue against the generality of the mathematics of inclusive fitness, while still accepting the logic of kin selection. You’ll note that the phrase “kin selection” does not appear once in Nowak’s Edge response

          2. In general usage Kin Selection is the by-product of inclusive fitness e.g. Wikipedia on Kin Selection – “Kin selection is an instance of inclusive fitness”
            You are perhaps right in one sense, there really are two different mathematical formulations describing Kin Selection- Inclusive fitness and Direct fitness. But before you congratulate yourself on making this distinction I would point out they are mathematically equivalent in results produced.
            Reference: Journal of Evolutionary Biology.
            2007 Jan;20(1):301-9.”Direct fitness or inclusive fitness: how shall we model kin selection?” Taylor PD, Wild G, Gardner. Quote: “In particular, direct and inclusive fitness always give the same answer”.

            As for Nowak he uses both terms, and I just followed his example by quoting his exact terminology from the paper which I quoted . In his controversial Nature paper on Eusociality, co-authored with Tarnita and Wilson he states: “For the past four decades kin selection theory, based on the concept of inclusive fitness, has been the major theoretical attempt to explain the evolution of eusociality.” Nowak is certainly correct on this one particular point.

  26. ‘I guess I’m the first respondent who has decided to criticize the answers of other respondents.’ I think I beat you to the punch: last week I wrote a constructive criticism of Tania Lombrozo’s response to the same question in her piece ‘The Mind Just the Brain’. Like evolution, mind-brain reductionism is still widely misconceived and the target of straw man arguments.

      1. Hi Ant,

        In her piece for Edge, Lombrozo targets mind-brain reductionism as a general theory, not individual theorists who propose and defend the theory. Thanks.


    1. What I meant was that I was the first person who answered that Edge question on the site who criticized other people’s answers. Obviously there were critics, but none that I know of who were in Brockman’s stable for this one.

    1. Nowak’s theistic account of evolution asserts a predictable trajectory to the creation of humans in God’s image. Nothing in science warrants such an assertion.
      Cooperative behaviour, notwithstanding. What does warrant it? Nothing but predetermined and absolute religious dogma.

      1. “Cooperative behaviour, notwithstanding. What does warrant it?”

        Perhaps Nowak has recognised that there is a teleological link between writing nonsense like this and getting a $10.5 million grant from Templeton


    Dark Matter. It’s Modified Newtonian Dynamics, people!

    (Only mildly kidding. But the evidence for DM is not conclusive, and building many houses on its foundation feels a bit like String Theory to me.)

  28. Kelly also seems to be making a much more common error than the one Jerry suggests, and one that I often encounter in college students. It is the identification of “random” with “uniformly random”, i.e., all possibilities are equally likely. Of course, that isn’t what “random” means at all.

    He compounds his error by using a loaded dice as an example of something non-random (it would be non-random only if it always showed up with one outcome), and finally takes the cake when he uses the term “chaotic” to refer to the loaded dice! He even admits that “chance” plays a role, and that there is a “skewed distribution”, but that is still not “random”! Next he’ll tell us that a biased coin is also “chaotic” and “non-random”, though “chance plays a role”, at which point all textbooks on probability will need a major revision.

  29. There is instantaneous speciation in plants through polyploidy, as well as in a few animals (one case in treefrogs for sure, maybe another in butterflies). In plants there is also speciation through hybridization where a sterile hybrid continues to reproduce vegetatively. Polyploidy may also occur in the sterile hybrid, creating a new sexually reproducing species. I’d also include the formation of an all female species by hybridization, as in some fishes and lizards.

  30. If you are going to replace the word “random” with “indifferent”, I think you will actually end up increasing the confusion about what “random” means. I say stick with your guns.

  31. The chance of a single mutation being adaptive for an organism does typically depend on the environment in which the organism finds itself. Some organisms are on fitness peaks – where the surrounding fitness landscape all lies below them and there are no nearby mutations that give a fitness boost. Some organisms are climbing “fitness ridges” – where there’s some scope for going uphill and others are on “fitness slopes”. In these latter cases, some nearby mutations do result in adaptive, uphill progress. The the chance of finding an adaptive mutation fairly trivially does depend on where you are on the fitness landscape. So: mutations are not typically ‘random’ – in the proposed technical sense given in this post.

Leave a Comment

Your email address will not be published. Required fields are marked *