Monday, January 27, 2014

Seven things you need to know about evolution

John Hawks is a pretty cool guy. He studies anthropology with a special emphasis on the genetics of human evolution over the past million years or so. Right now he's in Ethiopia looking at fossils. I visited him a year ago at his lab in Madison, Wisconsin.

He written a summary of seven things about evolution. You should read what he has to say to see if you know enough about evolution. Here's a list of the seven things but each one requires a bit of explanation.
  1. Evolution is change in a population
  2. Evolution is genetic change
  3. Many kinds of genetic changes are important to evolution
  4. Evolution can be non-random
  5. Evolution can be random, too
  6. Populations evolve all the time
  7. Evolutionary theory has changed a lot since Darwin's day


  1. Did you notice that #1 denies that there can be macroevolutionary processes such as species selection? Did Hawks notice?

    1. John Harshman: Evolution is indeed genetic change within populations. That may result in speciation (as two parts of the population become increasingly different). It may result in macroevolution (when the population has changed a LOT over many generations. #1 is not a complete description of evolution, but I don't see why it would be considered problematic.

    2. It's problematic if you think that some evolution isn't genetic change within populations, species selection being the most common example.

    3. I agree with John Harshman. John Hawks is describing the minimal definition of evolution just as I do in my post on "What Is Evolution." According to Hierarchical Theory, however, you can get evolution above the population level.

    4. It's the minimal definition of evolution if you think the minimal definition of a house is "foundation and walls", and leave the roof for someone else to mention.

    5. To be fair, John Hawks has discussed speciation and macroevolution on his blog:

    6. Hmmm. Well. I'm of two or three minds about this. Yes, subjects grouped under the term "macroevolution (so often misused by YEC's), such as speciation, adaptive radiation, and extinction are important patterns well worth studying and understanding, and part of the big picture of evolution. On the other hand, speciation and adaptive radiation result from those within-population changes in allele frequency. (Extinction, though? Does the change from a certain pattern of allele frequency to no allele frequency count?) On the other other hand, by arguing that evolution is all changes in allele frequencies, I tend to become like those annoying people who argue that all biology can be understood as biochemistry or as physics; true enough, but often not relevant since there are many different levels of biological organization and we can understand a lot of biology without starting with atoms.

      So, while I'd argue that technically the "changes of allele frequencies" doesn't rule macroevolution out, it doesn't clearly rule it in, either, and thus certainly isn't complete.

      Sorry, John Harshman.

    7. Barbara,

      What exactly are you sorry about? Let me correct a possible misunderstanding or two. First, speciation itself isn't generally considered macroevolution. It's differences in speciation and extinction rates that we're talking about here. When thinking about macroevolutionary processes, speciation can be thought of as analogous to reproduction and extinction as analogous to death. Speciation certainly can (arguably always does) result from changes in allele frequencies in populations, but it's best understood by considering at least two populations at a time, which renders it less easy to encompass by the usual definition of evolution. Extinction is certainly not a change of allele frequencies at all; no, a change to no alleles at all doesn't count. So we can have processes analogous to selection and drift happening at this higher level, in which species are the individuals and clades are the populations. It's another question whether these processes are important in evolution, but I think they are, though not perhaps as much as within-population processes.

      So Hawks' statement #1 needs at least a footnote, which I have kindly supplied.

    8. John Harshman says,

      It's the minimal definition of evolution if you think the minimal definition of a house is "foundation and walls", and leave the roof for someone else to mention.

      I understand where you're coming from but your analogy is bad. If the definition of house only requires foundation and walls then "roof" is irrelevant. By this definition, you could live in a house that doesn't have a roof and it would still be a house.

      So, we could easily construct a minimal definition of a house as "foundation + walls." As long as every house has to have at least foundations and walls then it works.

      Similarly the minimal definition of evolution applies to everything we want to call evolution including species sorting. Changes in the allele frequencies in a population are necessary, but not sufficient, to describe evolution by species sorting.

      Having a house with a foundation and walls is necessary, but not sufficient, to describe an enclosed house with a roof. (Note that it's actually the walls that aren't necessary but I hope you get the point.)

    9. You have a different definition of "minimal definition" than I do. I think a minimal definition of a term has to encompass everything that people refer to by it. If all you require is that some arbitrary label has an arbitrary meaning, then the minimal definition of evolution could be "narf!". To me, a house isn't a house without a roof, and I expect most people would agree. But we may be descending to semantic quibbling.

      I'd also say that changes in allele frequencies in a population have absolutely nothing to do with species sorting. They do set up the initial conditions (state of the populations when sorting begins), but if you take that as part of the process, then you will have to agree that mutation is natural selection, since mutations set up the initial conditions for selection.

    10. No one has yet persuaded me that species sorting and species selection should be considered as evolutionary mechanisms. My inclination would be to classify species sorting as an ecological fact, related to community composition. I am skeptical of strict-sense species selection, which proposes that factors like geographic range and population size are "evolved" features of species.

      That being said, I work on a group for which such macroevolutionary mechanisms have no utility. If I were more deeply engaged in other questions, I might change my mind.

    11. I don't particularly see the point that species selection must involve species-level, emergent characters. It seems to me that if a species speciates, or becomes extinct, and another doesn't, on the basis of characters for which each species possesses no selectable variation, that ought to count as species selection. It certainly isn't allele frequency change within populations.

      Whether species selection is evolution would be an interesting question to debate. I can't see why not. It's directly analogous to within-species selection, which we agree is evolution. Sure, it's an ecological fact, but so is evolution within species. And I'd say that competitive exclusion is a potentially important evolutionary mechanism that usually involves no within-species selection.

      Whether any of this applies to hominids is not a question I can comment on.

    12. Thanks, John -- as I say, I don't have very strong feelings about it. Seems to me that species selection is similar to natural selection only by analogy, as it involves different units and a different concept of heritability. I understand why paleontologists are interested in the concept. It seems to me like a "land grab" meant to ensure that "evolution" is the only explanation for the diversity and history of life, including all biogeographic and ecological factors as well as population genetic ones.

      I like the cleanliness of the population genetic concept of evolution, as it means all macroevolutionary processes really reduce to microevolutionary ones. But that is a preference that many people don't share!

    13. @John Hawks,

      Let me give you a possible example of species sorting. Consider Europe about 50,000 years ago. There were two subspecies of Homo sapiens; Neanderthals and the so-called modern humans. After they came into contact, only one of them survived.

      Same with Denisovans and Homo erectus. That's evolution at a level above that of individuals within a population. As Gould would put it, the "individuals" in this case are the subspecies and the "population" is the clade.

      Now, the reason this works is only because the subspecies are genetically distinct and that's due to traditional change within a population.

      I think you might know a thing or two about the genetic distinctiveness of the Neanderthal and Denisovan populations. :-)

    14. I don't disagree that species sorting is important to the history of life, I just am not persuaded that we gain anything by calling it an macroevolutionary mechanism. (Though I'm not sure I agree about that specific example as species sorting, seems like simple vicariance.)

      Is species sorting evolution? It is true that the survival of individuals in a population depends on the presence or absence of other species in the population's environment. But that is true of all manner of abiotic factors also. We usually call those relationships "ecology". We don't talk about the macroevolutionary mechanism of altitude.

      The best argument for considering species sorting as an evolutionary mechanism is that it may lead to a certain degree of regularity -- maybe in the same way that genetic drift is a stochastic process with some clear regular outcomes. But examples of this seem to be hard to come by.

    15. When the survival of individuals is dependent on their genotypes in relation to either biotic or abiotic factors we call that natural selection, and we call it evolution. Why should we refrain from calling it evolution when the entities are species rather than individuals?

      Are mass extinctions evolution? (I'd say yes.) Is interspecific competition a cause of evolution? (I'd say yes. You would say yes only if it results in a change in allele frequencies in the affected population, but I would consider the extinction of the affected population too.)

      And why not call it species selection rather than species sorting? I've never liked Vrba's distinction.

    16. Did you notice that #1 denies that there can be macroevolutionary processes such as species selection?

      It rather depends how you delineate a 'population'. Diplocentrists tend to go for sexually interbreeding groups (perhaps with asexual offshoots as honorary members), but it is hardly the only way to distinguish an evolutionary group of individuals. They are close ecological competitors, and if sex is obligate, types obliterate each other within that 'bubble' that ignores the rest of ecology. But for 2 billion years of life history, that wouldn't work at all.

  2. John Harshman: John Hawks' categories don't seem to be exclusive, see, e.g., categories 4 and 5.

    1. Note the presence of "can be" in 4 and 5, but "is" in 1. Big difference. If he had wanted to say "evolution can be change in a population" I expect he would have. Now of course what he wants to emphasize is that evolution isn't something individuals do, and he probably isn't thinking about macroevolutionary processes at all. I just want to point out that such processes, if they exist (and I think they do) contradict 1, though in a direction opposite from the one Hawks is thinking of.

    2. Yes because clearly speciation has to be a single event.

    3. After reading the origional post, I think his point in 1 is that evolution does not occur on the individual, but on the population. I really don't think he is implying what you are saying, John.

    4. Hm, how so? Macroevolutionay processes still happen at the level of populations afaik(?)

    5. jmwallach: I don't know what you're trying to say.

      TheOtherJim: Yes, that was his point. But if he was intending to allow for processes acting above the population level, he forgot to.

      Fukuda: No. Macroevolutionary processes, if there are any, are by definition acting above the within-population level. Species selection is an excellent example. Differential speciation and extinction of species do not require any changes in allele frequencies within populations. Now, you may think that macroevolution is just summed microevolution, but in that case there's no such thing as a macroevolutionary process.

    6. Yes - I agree he did not directly mention higher-level processes. He just left a big caveat;

      "Here's a list of seven things about evolution. It's not comprehensive but it hits on several important issues that help to understand how evolutionary biologists think about the process of evolutionary change."

  3. Hawks has a Coursera course on human evolution that just started last Tuesday. Still plenty of time to start it for those interested.

  4. I thought it funny his thing is about genetics but he's looking at fossils.
    I say there is no scientific genetic evidence for the evolution of biology outside of kinds etc.
    Could this man demonstrate some three top evidences for genetic evidence backing up evolutionary biology?

    1. There are a lot of funny things in it, but you are not one of them... Do you get it?
      Well. I have my doubts....

    2. Could this man demonstrate some three top evidences for genetic evidence backing up evolutionary biology?

      He's had a blog for years. Reading it, or better yet, taking his Coursera course (thanks, NG Carter!) might answer your question and increase your knowledge if you're open to that sort of thing.

  5. You forgot the eighth thing you need to know about evolution, which is a synopsis of the first seven:

    Evolutionary theory is such gibberish that it can be expressed in banalities that would be laughed at if applied to any other science.

    "The seven things you need to know about quantum mechanics:

    Quantum mechanics is change in nature
    Quantum mechanics is molecular change
    Many kinds of molecular changes are important to quantum mechanics
    Quantum mechanics can be non-random
    Quantum mechanics can be random, too
    Nature changes all the time
    Quantum mechanics has changed a lot since Schrodinger's day"

    Evolutionary biology is a parody of itself.

    1. Your "parody" does seem stupid, but don't blame evolutionary biology or quantum mechanics for your inability to erect even a convincing-looking straw man.

    2. Mregnor, the reason he feels the need to spell things out so simply is because however simple the point being made is, there are people like yourself who are determined to not understand the point.

      And if we're going to condemn people for explaining big things in simple, aphoristic ways ... well, let he who is without sin cast the first stone. Do unto others as you would have them do unto you. In stories about Jesus, it's claimed that's the way he spoke.

    3. 1.Astrophysics is change in the cosmos
      2.Astrophysics is change in stars and galaxies
      3.Many kinds of stars and galaxies are important to astrophysics
      4.Astrophysics can be non-random
      5.Astrophysics can be random, too
      6.Stars and galaxies change all the time
      7.Astrophysics has changed a lot since Hubble's day

      This is such a knee-slapper.


    4. Re Egmpr

      Quantum mechanics has changed a lot since Schrodinger's day

      Actually, not-relativistic quantum mechanics has evolved very little since Schrodinger's day. What has evolved is the rise of relativistic quantum field theory which combines quantum mechanics with special relativity. Once again, Egnor pontificates on subjects of which he is entirely ignorant.

      Astrophysics is change in stars and galaxies

      Yes indeed, stars and galaxies evolve in time. Not a startling revelation.

      Astrophysics has changed a lot since Hubble's day

      Yes indeed, the discovery of dark matter and dark energy has certainly changed cosmology since Hubble's day.

      Of curse, none of this has anything to do with biological evolution, except in the mind of IDiots like Egnor.

    5. No, it's only a sad testimony to your scientific illiteracy. Evolution (as described by Hawks) is a process. Quantum mechanics is a theory. Astrophysics is a branch of science. They are not analogous to each other. Evolution is (among other things) change in the genetic makeup of populations, but -- for example -- QM is neither "change in nature" nor "molecular change" (whatever that's supposed to mean). John Hawks uses the items in his list as paragraph headings: he explains rather lucidly what he means. You offer a silly parody of the style of the headings but ignore the explanation, which makes me doubt if you've bothered to read it at all.

    6. [Evolution is (among other things) change in the genetic makeup of populations]

      Evolution is seven things, according to Hawks. I added an eighth.

      What are the "other things"?

      And your dissembling about "evolution"-- here it's change in gene frequency, there it's Darwin's theory, over there it's a fact like heliocentrism, around the corner it's the best idea anyone ever had, behind you it's proof God doesn't exist-- evolution is a chameleon. It becomes what you want it to be, for rhetorical expediency.

      It's scientific heft is summmed up nicely in the list, particularly #8.

    7. Hawks doesn't say that "evolution is seven things". He gives a list of "seven things about evolution" which he thinks are worth keeping in mind. As John Harshman has pointed out, Hawks actually describes microevolution in this particular post on his blog -- a process that even most creationists accept as fact. Not that he's unaware of macroevolutionary processes -- he discusses them elsewhere, if you are interested (yeah, I realise you aren't likely to be).

      Evolution is not chameleonic -- it's just a complex process, taking place at several levels and in different time scales, like many things in this universe. There are people who specialise in different aspects of evolution and view it from different perspectives -- so what? Their perspectives are complementary, not contradictory.

      Darwin's theory is now of historical importance. It certainly can't be identified with modern evolutionary theory. Do you only like theories that are set in stone, canonised and protected from modification as we learn new things? Then stick to religion and forget science.

      Whether biological evolution is the greatest idea ever is a matter of personal taste, not an alternative definition. Evolutionary theory surely isn't proof God (or the Tooth Fairy, or Rumpelstiltskin) doesn't exist. The existence or nonexistence of gods (or fairies, or imps) doesn't matter in science.

    8. It's good that you think these topics are so obviously true as to be trivial, mregnor. That means you can enlighten your IDiot friends. For instance, every time they make an argument against "Darwinism" and think they are arguing against evolution, they show that they don't understand at least two of these topics.

    9. Michael Egnor says,

      And your dissembling about "evolution"-- here it's change in gene frequency, there it's Darwin's theory, over there it's a fact like heliocentrism,

      Do you mean to tell me that you been engaged in this debate for a decade and you still don't know the difference between the definition of evolution, evolutionary theory, and evolution facts?

      Read today's post on Uncommon Descent [“Intelligent Design is NOT Anti-Evolution” — a guest post]. The IDiots themselves will explain it to you.

    10. Smegnor uses the word "trivial" to describe evolution.

      When Smegnor's DI colleague Michael Behe called universal common descent "trivial", he meant that it was trivially easy to show with evidence that it had happened.

      Which of course means that it is "trivially" easy to show that humans and apes are descended from a common ancestor, according to Smegnor's DI colleague, Michael Behe.

      Perhaps this is what Smegnor means when he calls evolutionary theory "trivial"-- trivial in the same sense as common descent: easily proven with all the evidence we've got.

      If Smegnor disagrees with Behe, why then doesn't he challenge Behe to a debate, or demand he be imprisoned or have his funding cut off? Smegnor has demanded that climate scientists be imprisoned, and everyone who knows a climate scientist be imprisoned, and biologists who present evidence that evolution happened should have all their funding cut off because that kind of evidence is inconvenient for the Right and its corporate sponsors.

      So if Smegnor disagrees with Behe over universal common descent, why doesn't Smeggie demand that Behe be defunded by the DI, or be imprisoned by the Department of Homeland Security? That's what Smeggie demands of every other scientist. Go get ''im, Mike!

    11. Diogenes, Behe's view on Common Descent is a little more complex than you what you state above. In an interview on his book "Edge of Evolution" he says: "In my view it is certainly not “beyond the pale” for a scientist to question anything. Questioning and skepticism are healthy for science. I have no solutions to the difficult problems pointed to by scientists who are skeptical of universal common descent: ORFan genes, nonstandard genetic codes, different routes of embryogenesis by similar organisms, and so on. Nonetheless, as I see it, if, rather than Darwinian evolution, one is talking about "intelligently designed" descent, then those problems, while still there, seem much less insuperable. I certainly agree that random, unintelligent processes could not account for them, but an intelligent agent may have ways around apparent difficulties. So in judging the likelihood of common descent, I discount problems that could be classified as "how did that get here?" Instead, I give much more weight to the "mistakes" or "useless features" arguments. If some peculiar feature is shared between two species which, as far as we can tell, has no particular function, and which in other contexts we would likely call a genetic accident, then I count that as rather strong evidence for common descent. So, if one looks at the data in the way that I do, then one can say simultaneously that: 1) CD is very well supported; 2) grand Darwinian claims are falsified; 3) ID is confirmed; 4) design extends very deeply into biology."

    12. Andy Wilberforce,

      Why the long post, just to confirm that Diogenes was right?

    13. Here's a good one for yourself, Mike:

      Brain surgery is change in a brain
      Brain surgery is structural change
      Many kinds of tools are important to brain surgery
      Brain surgery can be non-random
      Brain surgery can be random, too
      Brain surgery techniques evolve all the time
      Brain surgery has changed a lot since Mike Eggy's day

      this is actually pretty fun :B

    14. But Ken Ham frowns at those deviations from the straight an narrow:

      Somebody call the Spanish Inquisition!

    15. Wow, mregnor really doesn't seem to know very much about quantum mechanics or astrophysics!

    16. We have yet to locate the subject he does know very much about. So no surprise.