Saturday, July 30, 2005

Splenetic ranting (feel free to ignore)

It's way early Saturday morning and I'm banging around my apartment dealing with a bout of insomnia, so I might as well post this. Fair warning --- I'm going to be writing more about my reactions to this subject than the subject itself. Navel-gazing ahoy.

Nomadic Thoughts has the link rundown for a recent trans-blog dustup, about Jared Diamond's Guns, Germs, & Steel, between Crooked Timber, Brad DeLong, and a bunch of anthropologist types at Savage Minds. I wrote some slightly lengthy comments on the first post in the controversy, though I now realize that this was stupid of me. My summary of the debate:

  • Ozma of Savage Minds writes: "Jared Diamond is wrong, and the reason people like Guns, Germs, & Steel is that they're racists."
  • Brad DeLong writes:: "No, you're confused about Jared Diamond's ideas. And we like the book because it's well-written, and makes a striking argument about an important question."
  • Kerim of Savage Minds writes: "Well, it's stupid and maybe even morally wrong, even to ask that question."
  • Henry of Crooked Timber writes: "No, it's just a different question than the one you'd investigate."
  • Ozma repeats, in comment threads on the above posts, over and over again, that "Jared Diamond is wrong, and the reason people like Guns, Germs, & Steel is that they are racists." She doesn't address any of the objections, evidently subscribing to the theory that vociferous repetition is the key to winning arguments.
  • Kerim repeats, in comment threads on the above posts, that either Diamond's question is immoral, or that he's not really asking a different question, or that the reason people like Diamond is that they mistakenly believe he's asking a different question. His position keeps changing, so that as one position is refuted he moves to another one and claims he held that other position all along; in this fashion, he cannot be proven wrong, nor can anybody engage in productive debate with him.
  • Throughout all of the above, several anthro types insist that (1) Diamond's theories are old hat, and (2) they have been thoroughly debunked. They fail to present evidence or argument for these two points, insisting that they are common knowledge among anthropologists.[0] (Yeah, OK, you supercilious jackasses; we have crackpots in computer science too, but when somebody claims that, for example, they've invented an algorithm that can compress all strings, we can concisely explain exactly why that claim is a crock, in terms that anybody with a little math background can understand; and you should be able to do the same.)

Reading all this got me pretty mad. As those who've known me for a long time can attest, there's almost nothing in the world that makes me madder than bald-faced nonsense. My reaction to nonsensical statements is usually much more intense than, for example, my reaction to immoral or insulting statements. Allow me to elaborate slightly. When I read an argument that's completely unsound (as opposed to one that's wrong, but defensible), it strikes me like a blow to the head. I literally feel a bolt of dull pain pulse between my temples (and I do mean literally, not "metaphorically, but with such intensity that I feel obligated to describe it as literal"). I get a surge of adrenaline-fueled fight-or-flight response, and often feel the need to get up and pace around the room. I find myself wishing that I were omnipotent, solely so that I could obliterate that argument from the face of the Earth, because the very fact that a human being has the effrontery to make such an argument, and to expect other people to believe it, is an insult to all of creation. In short, nonsensical arguments send me completely off the deep end. I'm not claiming that this is a sane reaction, but it's how I react.

Ozma and Kerim's original posts were bad enough, but as I read their replies, and the replies of some others, I could not avoid smacking myself on the forehead, just to make the pain stop. Ozma, especially, seems to embody a perfect storm of rhetorical incompetence: she's unable to understand Diamond's argument, she's unable to coherently explain her objections to (her mangled understanding of) Diamond's argument, and she's unable to handle critical responses to her own arguments. She's also a new professor at the U. of Alberta, and I sincerely pity her students.

In retrospect, it's pretty obvious to me that Ozma and Kerim are fractally wrong, i.e. so wrong that arguing with them will lead to an infinitely recursive regress of wrongness on every sub-point of debate. I should have recognized this before I invested time in commenting, but somehow I mistakenly thought that since (a) I liked Diamond's book, and (b) I wanted to understand the objections to it, that there was some value in sinking my time into this debate. Stupid me.

p.s. Lots of anthro types seem to have some pretty absurd notions about biology. Two examples:

  1. I compared the difference between Diamond-style hypotheses of ultimate causation, and anthropological hypotheses of proximate causation, to the distinction between evolutionary and developmental biology; and I suggested that, as with "evo-devo", there might be gains to merging the two. Ozma followed with the absurd statement that "evo" was "in retreat", which is absurdly incorrect --- evolution is stronger than ever. She then had the amazing temerity to refer to this bizarre misconception in the Crooked Timber thread, evidently believing that it was a point in her favor rather than evidence of her disconnection from reality and willingness to spout off about subjects that she's completely ignorant of.
  2. "J Thomas" wrote, in a comment, that Darwin "got evolution all wrong but at least said evolution was going on". Ahem. Biology has advanced a great deal in the past century and a half, but Darwin was fundamentally right. It would be accurate to say, "We know a great deal more than Darwin; in their daily work, modern biologists rely on modern theories that are not directly attributable to him." But it is a complete distortion to say that Darwin "got evolution all wrong", and only somebody who was supremely misinformed could say otherwise. (UPDATE 31 July: J Thomas has clarified, so I retract this; see comments.)

This, combined with some disparaging comments by Kerim about "biological/deterministic" reasoning, leads me to believe that anthropology in general (or at least a certain sub-community thereof) has some real problems with the theory of evolution. Given that the theory of evolution is simultaneously one of the greatest intellectual achievements of humankind, and the engine driving one of the most productive and exciting scientific disciplines today, this doesn't speak well of these anthropologists.

UPDATE 31 July: I was too hasty in inferring too much about anthropologists and evolution from what I read. See comments for further discussion.

[0] OK, they did present two references. However, closer examination by commenters --- including myself --- revealed that one is totally irrelevant to Diamond's thesis, and the other seems to be based on significant misreadings of the conditions necessary for Diamond's thesis to be valid. Also, the anthro types never explain in the first place how these citations supposedly demonstrate that Diamond's arguments are either old hat or thoroughly debunked --- they simply point to the footnote and say "See?" This might be acceptable if the citations themselves were not so weak, but as it is...


  1. I don't think it's that anthropologists ahve aproblem with evolution per se -- step over to the Anthro-L listserv and you'll see dozens of anthropologists vociferously defending evolution, Darwinism, and natural selection (and bemoaning the Creationism and ID). What many of us have a problem with is what might be called "vulgar evolutionism", the argument popular with the ev-psych types that all human (and other, I suppose) traits must necessarily be adaptive and that the answer for human distinctiveness can best and often only be found in a biology that evolved tens of thousands or more years ago. Most anthropologists will tend to emphasize the cultural flexibility of humans with regard to both social and individual behavior -- while recognizing that this felxibility itself is made possible by the evolution of certain traits in the brain and body.

    The other objection is that people who argue that trait x has something to do with our evolution as hunting animals generally tend to have very little awareness of what a foraging lifestyle is like -- the tendency is to picture it as more or less like modern, industrialized socity, only with spears and animal hides, which lends such arguments a circularity that is difficult to reason with.

  2. Sure. I have a problem with "vulgar evolutionism" and naive, evidence-free arguments about human evolution too. It doesn't lead me to say astonishingly dumb stuff like Darwin "got evolution all wrong" or that "evo[lution] is in retreat". These speech acts say something very unflattering about the people who utter them, roughly: (1) "I am incredibly ignorant about biology" and (2) "In spite of my ignorance, I am willing to shoot my mouth off about it in a disparaging fashion".

    Clearly, not all anthropologists share these characteristics. (I should have made the parenthetical clause "or at least a certain sub-community thereof", in my original post, into the main clause.) But in my brief exposure to the Savage Minds folks, I've observed multiple people make really silly remarks about biology or evolution. Not only did they say these things, but they did so with an offhand certainty that seemed to presume the respectability of their position, not just its correctness. They seem to believe it's a well-established fact that Darwin was wrong, or that evolution is a theory in decline.

    I can't figure out where they could have gotten this idea, except from a community of intellectual peers (I doubt that left-wing anthropologists picked it up from the Discovery Institute et al.). I'm using a small sample size, so I should be wary of generalizing with too much confidence. However, it seems that this contempt for evolution characterizes a significant community of thought within anthropology, rather than a couple of crackpot individuals. Perhaps justifiable contempt for "vulgar evolutionism" has bled over into bad misconceptions about the evolution itself, and the community's ideological predispositions permit individuals with these misconceptions to "fly under the radar" and go uncorrected. Regardless, if someone can get a tenure-track faculty position in anthropology while suffering from such intellectual maladaptation, then anthropology's threshing room is allowing a great deal of chaff into the wheat.

  3. Well, the "Darwin got it wrong" comment was, iirc, from a non-SM commenter. I understood Ozma's comment about "evo in retreat" as making a statement about ev-psych approaches, which almost always can be replaced by developmental approaches -- witness recent data that suggests a growing disparity, at the genetic-molecular level, between even identical twins over the course of their lifetimes.

    But be careful, too -- what Ozma posts or Kerim posts or I post at SM is not representative of all of us. WE have no oversight committee, and very little discussion about who should say what. Check out the response of some SM'ers to my earlier posts on morality.

    As an anthropology teacher deeply committed to evolution and natural selection as explanatory principles, and as a fellow SM'er, I can without qualms say that SM has no "position" on evolution, though I'd be very surprised, very surprised indeed, if the others were not as committed as I am.

  4. In other words, upon encountering the unfamiliar term "evo-devo", Ozma felt comfortable holding forth on the alleged decline of evolution, without first, say, checking out Wikipedia to see what "evo-devo" actually meant. A lesser grade of intellectual failure than denigrating evolution per se, but still a shameful failure, especially from someone who feels comfortable excoriating laypeople for not reading obscure references in the anthropology literature.

    Also interesting is the conflation of evolutionary biology and evolutionary psychology, the latter of which is an incredibly controversial nascent subfield that many biologists hold in thoroughgoing contempt. I suppose anthropologists, by virtue of their subject matter, run into a lot of itinerant evo-psych types peddling their wares, but blaming evolution for the current state of evo-psych is a little like blaming scientists for Scientology.

    On the other hand, if you claim that the overwhelming majority of anthropologists share your commitment to evolution, then I retract my hasty generalizations about anthropology, or any subcommunities thereof.

  5. To be quite honest, I can't recall the exact context that Ozma's statements were made in, but that's ok -- I'm not really her spokesperson, just conveying what my (half-remembered, at that) impressions are. As to anthropology and evolution -- we are a discipline that recognizes the evolution of H. Sapiens as part of our field of study, after all! As I said, a poke through the archives of Anthro-L (accessible through should more than suffice to get an idea of where anthropologists in general stand on the topic.

  6. Cog,

    Jenny and I were watching the PBS special and getting annoyed. Jenny has training in anthropology, so I asked her what could be identified as "wrong".

    She didn't want to level much criticism at the book, since she hadn't read it. But the tv show was superficial, repetitive. The segment in the African hospital is especially crass and manipulative.

    I'd think that most of the complaints are because of the TV show.

    ~ Patrick

  7. Cog,

    Jenny and I discussed this a little more:

    * We agree with the idea that geography is a factor in the expansion of a society
    * Even though, on a globe, Eurasia is sideways, there are also geographical barriers. See the hoopla over Marco Polo

    The biggest hole we saw in Diamond's thesis is the lack of discussion of the sea and the formation of trade and navies. Several of the European empires are defined by their navies (British, Spanish, Dutch, Norse). Perhaps the defining geography of Europe was not the climate, but the amount of landmass adjacent to navigable waters.

  8. My background is in genetics and not anthropology, and I explained at length about Darwin. I believe that summarising that as "Darwin got evolution all wrong" is misrepresenting what I said.

    I'll say it again -- through no fault of his own, Darwin had no understanding of genetics. His explanation of how we get the variation that natural selection selects, was mostly handwaving and was completely wrong. This lack made his arguments more difficult. Evolution that worked the way Darwin suggested would not work very well. Evolution that worked by the neodarwinian synthesis would have worked much better, and evolution that uses the new genetic mechanisms that are being discovered today would work better still.

    The specific mechanisms that biological populations have evolved to let them evolve faster are still not well-understood, so this is an exciting time in evolutionary biology. It's universally accepted that Darwin had no understanding at all of the topic.

    I really don't understand what you thought you read from me, that you disagree with.

  9. J: That Darwin "got evolution all wrong" is a verbatim quote from your comment here. And of course Darwin had to handwave around heredity, since he had no knowledge of genetics, but that's a big leap from saying he got "evolution" all wrong.

    Furthermore, it seems to me that evolution itself doesn't care too much about the precise mechanisms of heredity anyway. Bacteria, protists, and metazoans all have pretty different mechanisms for reproducing (or otherwise sharing genetic information), but evolution operates on them all.

    Sorry for my misconception regarding your background. When you said that Darwin got evolution all wrong and then cited a bunch of anthropologists, I leaped to conclusions and assumed you were not a biologist.

  10. Pat: I haven't seen the TV series, and as time goes by I am growing more convinced that it must be really terrible compared to the book. The book states pretty clearly that Diamond doesn't consider Europe to be that different from the rest of Eurasia prior to the colonial era, and he's not really trying to explain why Europe (as opposed to the rest of Eurasia) conquered the world. If the TV series gives the impression that he is explaining this, then that's bad.

    The geographical barriers within Eurasia are an interesting point, and maybe one of the biggest potential problems with Diamond's theory. I would like to see a rigorous analysis of the difficulty prehistoric people would have faced in traveling around Asia versus the Americas.

    However, note that if Diamond's right that Eurasia had a lead in easily domesticable plant and animal life, then Eurasians would have had a lead in developing seaworthy vessels that would connect the Eurasian continent as well. I think it's important to consider the feedback loops and multiplicative effects between the factors that Diamond describes.

    BTW China had an impressive navy, before they scrapped it. Most of the Chinese population lives relatively close to the coast, or major rivers (which are connected to the ocean). The question of why Europe, and not India, China, or the Middle East, is pretty puzzling.

  11. Cog, you got me on that out-of-context quote. I didn't even remember that it was an exact quote.

    When you say that evolution doesn't care about the mechanism of heredity, isn't that like saying that when you implement genetic algorithms you don't care about the algorithms?

    At any rate, Creation Scientists read Darwin looking for ways to contradict him, but biologists don't read Darwin unless they think it's fun or inspiring or something. Darwin was one of the first to get some of the important concepts, but he isn't a good source for learning about evolution.

  12. J: That phrase appeared with no further explanatory text in your original comment, so I don't think my quotation is really "out-of-context", which implies some irresponsible elision on my part. When you post a random comment on the Internet, you can't assume that people will see your words in the context of some history of previous utterances. Since then, you've clarified what you meant, and I've updated my post to reflect that. Look, if you're a geneticist then you clearly know much more about biology than I do, so it seems to me that your original comment merely used some poor word choices. It happens to the best of us. I think we should just acknowledge that and move on.

    However, if you still really want to discuss how wrong, exactly, Darwin was, I'd suggest that our disagreement stems from a difference of perspective. You see evolution up close, and therefore you see all the myriad details where modern evolutionary biology differs from Darwin's original ideas. I see evolution from afar, at a different level of abstraction, and at that level, modern evolution seems to conserve the most essential features of Darwinian evolution: descent with modification plus selective pressure producing biological diversity, the common ancestry of all terrestrial life, etc.

    Even you admit that Darwin was "one of the first to get some of the important concepts". It may be hard to realize, but to a nonspecialist these may actually be the most important concepts, compared to which all the others are variations on the main theme. The fact that, for example, mutations in gene regulation may be a more important factor in evolution than mutations in genes themselves, is certainly exciting for biologists, but to me this looks like just another kind of descent with modification. Founder effects and allopatric speciation may be cool, but to me they look like just another kind of selection pressure. Etc.

    By the way, when you implement a genetic algorithm, you can generally factor out the mutation and replication strategies as parameters. Using the right parameters may make a big difference in whether the algorithm works well for your particular problem, but changing them won't make it a fundamentally different sort of algorithm. Once again, it's a matter of the level of abstraction.

  13. OK, different orders of abstraction.

    Darwin got the most abstract level basicly correct. Later workers put it into mathematical form.

    The basic ideas are:

    1. In each generation there is a surplus of individuals. They can't all reproduce successfully and be fully represented in the next generation, except in rare circumstances when the population is increasing fast. (And even then some will be able to reproduce quicker than others.)

    2. Offspring tend to resemble their parents, though not exactly.

    3. The individuals in the population vary, and vary in ways that affect their chance to reproduce successfully.

    These are enough to get a population change, given sufficient time.

    But in practice some things evolve a lot faster than other. My limited experience with genetic algorithms gave me the impression that the particular mutation function was vital.

    And if I could adjust the selection function that could make a big difference too. The selection was like -- using a bad selection function is like inverting a poorly-conditioned matrix. You get a lot of extra noise in the system that doesn't have to be there. You could make easy progress up a steep ridgeline, and then it would take a specialised mutation function to stay on the ridgeline while you look for a peak. Slow.

    And specific types of problems demand special mutation functions. There's no one mutation function that works well for all problems. Like, to look for a minimal steiner tree, a mutation function that picks a random link and breaks it (converting the tree into two trees) and then picks a random pair of points, one in each tree, and connects them, will be very inefficient. For early progress it's better to add a link (biased to short links), and then throw away the longest unshared length on the circuit the new link creates.

    Darwin did state the conditions needed for evolution to occur. And he showed plausibly that those conditions might be met by biological populations. At that level of abstraction he was right. But his explanation of biological evolution was fundamentally wrong. That's fine because science evolves. And there are creationists who study Darwin at length because they think of Darwin as their chosen enemy's scripture.

  14. J: So far, everything you've written suggests that Darwin's account of evolution was incomplete, not wrong.

    Are you implying that Darwin claimed mutation occurred at a constant rate, or that a single selection function solved every problem? In Origin of Species he specifically states the opposite of both --- see, for example, the excerpts in part 5 of the punctuated equilibrium FAQ (note: you can mostly ignore the discussion of Gould & Eldredge; I'm citing this because of the quotes from Darwin).

    I agree that even if Darwin were completely disproved, that would do no harm to modern biology. But I would like to understand better why you think Darwin was "fundamentally wrong". What claim did Darwin make that was fundamentally incorrect?

    Finally, yes, I already said that the mutation and replication/selection strategies of a genetic algorithm matter a great deal, for any individual problem. However, you can explain the idea of a genetic algorithm, and why genetic algorithms work in general, without specifying these strategies.

  15. Darwin's book is very long and I haven't read much of it. At one poine Darwin claimed that inheritance went by "blending", that offspring tended to be the averages of their parents. But he'd talked to animal breeders and he recognised that sometimes old traits popped out that had been observed generations previously and had been hidden for a long time.

    If you believe in blending inheritance, you'll see species as tending to be all the same barring geographic variation etc. The differences get blended together. And that's what we do see in most populatoins in nature, you have to look close to see any differences in proportion.

    To get changes he had to assume either a giant amount of new variation -- which he did assume -- or else assume extremely slow change -- which he also did assume.

    Darwin correctly stated conditions that allow evolution to happen. But he was incorrect about how biological evolution does happen.

    I agree with you that just like the genetic algorithm case you can explain the general idea simply and easily, and why it works in general, without dealing with any of the details that could make it work adequately in any particular class of problems.

    I wouldn't insist that the only correct way to state it is that Darwin was wrong, rather than incomplete. I feel like I'm already being more pedantic than I intended. Sorry about that.