Sunday, March 01, 2009

Against intuition   posted by Razib @ 3/01/2009 01:49:00 PM

John Hawks points to an article by James F. Crow, Mayr, mathematics and the study of evolution. As John stated this is Open Access assuming you take the time to register. Here is a taste:
In 1959 Ernst Mayr...flung down the the feet of the three great population geneticists RA Fisher, Sewall Wright and JBS Haldane..."But what, precisely," he said, "has been the contribution of this mathematical school to the evolutionary theory, if I may be permitted to ask such a provocative question?" His skepticism arose in part from the fact that the mathematical theory at the time had little to say about speciation, Mayr's major interest. But his criticism was more broadly addressed to the utility of the entire approach. A particular focus was the simplification that he called "beanbag genetics", in which "Evolutionary genetics was essentially presented as an input or output of genes, as the adding of certain beans to a beanbag and the withdrawing of others."
Recent mathematical work has gone well beyond that of the three pioneers. Partly this is due to skilled mathematicians entering the field and bringing new techniques with them; especially noteworthy are stochastic processes. Second, and perhaps more important, is the extensive use of computers. Often you can use a computer to get by without deep mathematical knowledge. An additional influence is the explosive growth of molecular data, which lend themselves to mathematical treatment. In the first half of the twentieth century, population genetics and evolution had a beautiful theory, but there were very limited opportunities to apply it. Now the situation is reversed. Molecular data accumulate too fast to be assimilated.

Crow referred to some of these questions 3 years ago when I interviewed him. Though much of the essay is a restatement of ideas floated elsewhere, it's still awesome that Crow is publishing at the age of 92. Judging by how quickly he replied when I sent him an email he is also still actively corresponding.

As for the general thesis outlined in the article, of course I tend to agree with Crow. From what I know Ernst Mayr's viewpoint in Systematics was overturned by the cladist revolution, which introduced a rigorous hypothetico-deductive framework into the field. It is perhaps just part of a trend of a marginalization of more philosophical biologists who rely on intuition in the realm of theory, and serves as a specific case study of Mayr's own philosophy of science and how it is ceding ground to more moral analytic techniques. Nevertheless, we can thank Mayr for his mentoring of someone like Robert Trivers. I remember talking to a friend of mine who was at OEB in the early 2000s, and she mentioned getting stuck in the elevator with Ernst Mayr, and my first reaction was, "Dude is still alive?!?!"

Labels: ,

Wednesday, July 09, 2008

Have multiple intelligence theories really been disproven?   posted by birch barlow @ 7/09/2008 04:46:00 PM

[this is a slightly edited version of what was originally a haloscan comment]

I have come to believe that it is crucial to realize that there are other factors in intelligence besides g and its subfactors (e.g. math/performance, visuospatial, verbal, short-term memory). This is important not only factually and scientifically, but politically as well; a less g/traditional IQ-based theory of intelligence and human biodiversity is probably both more accurate and more politically palatable than a heavily g-centered one. The main drawback is that such a theory is also unfortunately quite complicated and difficult to test.

Many apparently non-g factors are almost certainly correlated with g, but they are not the same thing. In terms of higher-visibility phenomena, this would mean factors like creativity, motivation/drive, consistency, effective planning. In terms of lower-visibility phenomena, this would mean factors such as (neocortical) left brain/right brain ability, efficiency, and interconnectedness, as well as the interaction of such entities with the paleomammalian/limbic/midbrain and reptilian/lower brain/brain stem.

The main problem, in my opinion, is that these factors other than g are much harder to measure, and virtually impossible to measure on a 3-hour test, much less a brain scan (given current technology). Of course there are self-report tests for personality, creativity, motivation, and the like, but self-report tests are not, in general, terribly reliable. Also, insightful multivariate data analysis and effective experimental design for such analysis is hard to come by because these tasks are extremely difficult for even an intelligent person (in both the g and non-g sense) to carry out. Thus the failure of "multiple intelligence" theories in spite of the fact that it is clear that there are multiple intelligences.

If you are still unconvinced, how else would one explain a 25-year old with a 150 IQ, but also with Asperger's Syndrome/autism spectrum disorder, living on the streets while a 90-IQ illegal immigrant is living reasonably comfortably, and an intellectually uncurious and largely vacuous (outside of the classroom/lab/workplace) individual with only a 115 IQ is living large?(if you want a picture of the latter individual, think of Julia from (Orwell's) 1984 transplanted to the real America ca. 2008, or the devoted Asian [1] college student who seems to always be studying (high motivation/drive/tolerance for long, boring tasks) but has no intellectual interests and spends most of her/his [2] free time with sleazy entertainment, sleeping around, smoking pot, drinking, and popping pills [3]). Of course social factors such as biases against more autistic personalities may be partially at work, but most stereotypes and social biases have *some* basis in reality, even if they all too often facilitate cruelty and inefficiency. It is also important to remember that biological phenomena can lead to social phenomena (e.g. autistics, due to their biology, are repelled by (and repel) others, leading to a negative social reputation for autistics) just as environmental/social phenomena can lead to biological phenomena (e.g. autistics, due to their negative social reputation, increasingly have their biology wired for being hermits).

[1] and [2] Being intelligent but uncurious seems to be substantially more common amongst Asians (and perhaps amongst high-IQ blacks and Hispanics/Amerinds as well) than whites, and more common amongst females than males. This is only my personal observation, and it may be an entirely sociocultural phenomenon even if it is real.

[3] Nothing against sleeping around, smoking pot, drinking, or popping pills, but these don't tend to be the most intellectual activities in the world, in spite of common protestations to the contrary by horny drug users (such as, admittedly, myself) to the contrary. Also, I realize that immigrant (and particularly Asian) cultures are strongly biased against such hedonistic behaviors, but this bias tends to quickly fade amongst the children of immigrants, and even more so their grandchildren, as they become more modernized, Westernized, and Americanized.

Labels: , , ,

Sunday, October 07, 2007

Why people don't care about statistics   posted by Razib @ 10/07/2007 08:51:00 PM

RPM points to a Language Log post which discusses and highlights the human ignorance of, and disinterest in, statistical thinking. Assman put up an interesting comment. I will say this:

1) There are some basic tools which many intelligent people are just plain ignorant of. On this weblog one reader, a medical doctor, was surprised by Bayes' rule, but immediately understood its relevance. This happens to all of us, there are amazing tools out there which we haven't encountered for a variety of reasons. And most of us are happy to pick up the tools if we see clear utility.

2) But, there's another problem, and that is the fact that statistical and probabilistic thinking is a real damper on "intellectual" conversation. By this, I mean that there are many individuals who wish to make inferences about the world based on data which they observe, or offer up general typologies to frame a subsequent analysis. These individuals tend to be intelligent and have college degrees. Their discussion ranges over topics such as politics, culture and philosophy. But, introduction of questions about the moments about the distribution, or skepticism as to the representativeness of their sample, and so on, tends to have a chilling affect on the regular flow of discussion. While the average human being engages mostly in gossip and interpersonal conversation of some sort, the self-consciously intellectual interject a bit of data and abstraction (usually in the form of jargon or pithy quotations) into the mix. But the raison d'etre of the intellectual discussion is basically signaling and cuing; in other words, social display. No one really cares about the details and attempting to generate a rigorous model is really beside the point. Trying to push the N much beyond 2 or 3 (what you would see in a college essay format) will only elicit eye-rolling and irritation.

Note: And as Assman noted, the above isn't really relevant for most people who are probably too light on the g.


Sunday, February 25, 2007

The tactics of deceit   posted by Razib @ 2/25/2007 04:56:00 PM

Over at The Edge the philosopher Daniel Dennett and evolutionary geneticist H. Allen Orr just had an exchange over Richard Dawkins' The God Delusion. I am not particularly interested in the details of this debate, rather, on the front page this selection from Dennett's letter caught my attention:
When I explained then in a private letter to you what I had meant, you conceded to me in your private response that you had not seen my point in the light I intended, and that my claim was not in fact the blunder you had said it was....

What could Dennett mean? I immediately thought back to this:

I did indeed misspeak (p. 126), but the result was ambiguity, not error. The issue is complicated: it depends on whether you're measuring the (average) speed of departure from a starting point in genetic space, or the speed of attainment of some particular evolutionary product. I meant the former. [Dennett]

Now I've been in the population genetics business for some time and, frankly, I have no idea what Dennett is talking about. And-I can find no polite way of putting this-it's hard to escape the conclusion that Dennett has no idea what he's talking about, either. Even the most charitable interpretation I can come up with is just plain wrong. [Orr's response]

Dennett's phrasing was awkward and peculiar to say the least. Orr is an evolutionary geneticist who "was awarded the Dobzhansky Prize by the Society for the Study of Evolution and the Young Investigator Prize by the American Society of Naturalists." In regards to a technical understanding of evolution Dennett was way out of his league. And only an analytic philosopher could confuse population genetic theory with such lexical opacity. That being said, I did not believe that Orr did not understand what Dennett was trying to say, unpolished though he was. This is how I unpacked it:

1) Because most mutations are deleterious, selection is generally a force for constraint, maintaining the ancestral state.

2) In contrast, random genetic drift operates upon evolutionarily neutral variation, so it exhibits (approximately) equal effect on mutants and ancestral alleles over the long term.

3) These insights lead to the contention of Kimura et. al. that most evolutionary change is driven by the substitution of neutral mutations, with random genetic drift a major engine. Even though the vast majority of neutral mutations go extinct, a small number perpetually substitute themselves over ancestral variants at a constant rate (i.e., the rate of substitution is proportional to the rate of mutation).

4) But, when there is a positively selected mutant, natural selection operates far faster in regards to fixing the variant than random genetic drift and stochastic processes might in the typical population size.

It is more complicated than that...but my point is that though Dennett wasn't exactly clear, I think a reasonable observer could understand what he was getting at, or at least the somewhat confusing waters into which he had ventured. Orr did not acknowledge this. As I said, Orr is a world renowned evolutionary geneticist (as he implies in his response), so I was skeptical that he was as clueless as he'd let on.

So, when I read Dennett's full letter, I was not surprised. Here is what he says:
You leveled very serious charges of error and incomprehension in that review, and when I challenged them, you responded with a haughty dismissal of my objections (in an exchange in the Boston Review). Quoting an example, dealing with the speed of evolution: "Now I've been in the population genetics business for some time and, frankly, I have no idea what Dennett is talking about. And-I can find no polite way of putting this-it's hard to escape the conclusion that Dennett has no idea what he's talking about either."

Dennett is bringing up exactly what I remembered in The Boston Review! This was 10 years ago, but Dennett obviously still resents Orr mocking him. In private correspondence Dennett states that Orr admitted that he wasn't really wrong, and Orr's response to this new salvo from Dennett does not contradict that characterization.

I writing about this for one primary reason: I thought, at the time, that H. Allen Orr was striking a low blow dishonestly when he could have stood on firm ground, though with a weaker impact. I've thought about this exchange on & off for several years now, and have considered blogging it, because I think it is perhaps a reflection of H. Allen Orr's character. Most of the readers of The Boston Review don't spend their free time reading The Genetical Theory of Natural Selection or Motoo Kimura's papers, they wouldn't be able to parse from Dennett's somewhat confused phrasing what he was trying to get at. In other words, assuming that Dennett wasn't a moron they couldn't really map his verbal exposition onto a range of evolutionary models, because it isn't exactly as if a broadly educated person is familiar with the great Neutralist vs. Selectionist debate of the 1970s. They were relying on H. Allen Orr's expertise. I just can't believe that H. Allen Orr didn't understand what Dennett was trying to say, or, that he couldn't have seen what he was getting at if he hadn't prejudged Dennett and his ideas as a whole. If I had been him I would have pointed out that Dennett's graceless exposition of the relationship between neutral and selective forces in evolution should be a clue that the man doesn't have the technical competency to engage in such an ambitious meta-project as the one he laid out in Darwin's Dangerous Idea. Instead, Orr is implying that Dennett simply doesn't know what he's talking about, and making him out to be the fool. This is rhetorically far more powerful, and I am sure it persuaded most readers of The Boston Review, who are not going to be conversant in the details of evolutionary genetics and trust H. Allen Orr. After all, who are you going to listen to on this topic? A philosopher or an evolutionary geneticist?

To my mind Orr did a disservice to science and intellectual discourse. He went for the knockout, calling an intellectual ignorant is basically undermining their worth. Saying that they are a bit confused lacks a similar punch. But the readers of The Boston Review, or the lay audience in general, is not looking for a legal case where you are an advocate for your position at all costs, because science and intellectual discourse is more than one battle, it is a long war against our moral and personal failings, against pride, against ego, against self-interest and self-aggrandizement. Most of the battles are lost, but slowly the war grinds on and the trenches keep moving inch by inch. Ph.D. scientists make considerably less in income than their intellectual inferiors in law or medicine. But the field in which they operate is one of great prestige, of civilizational significance. Presumably they wish to engage in the adventure of the ages, at the cost of financial status. Scientists are human, as the politics which suffuses any university department would confirm, but, like monks meditating upon the nature of God, abjuring themselves of worldly pleasures and satisfactions, scientists have to err on the side of truth as they see it, and not a short term rhetorical victory. For what shall it profit a man, if he shall gain the whole world, and lose his own soul?

Addendum: Readers can follow the links and judge for themselves if Dennett wasn't clear in his intent, even if he was being muddled about it. Perhaps Orr didn't understand what Dennett was trying to say. But if Orr could comprehend the opaque prose in The Genetical Theory, I can't see why he couldn't parse what Dennett was trying to get at. And just to be clear, though I point a finger at Orr on this occasion, rest assured that I understand that we are all guilty of this particular sin, and that includes myself. I simply want to emphasize Orr's transgression here because I think it is important for us to remember that no matter the satisfaction that victory in one battle gives us, we are not fighting without a long term purpose, and dishonorable victories are fundamentally Pyrrhic.

Addendum II: Also, I understand that in practice science between the bounds of accepted & rejected consensus is quite the bloodsport, and an adversarial & amoral dynamic is common. Though this injects quite a bit of "noise" into the system, I trust over the long haul that the scientific culture will beat expectation in modeling reality. That being said, the issue I am pointing to here is slight of hand by gatekeepers. Orr writes reviews books as an evolutionary geneticist, with all the expertise that that entails. His behavior in this case was repulsive to me because he sacrificed the chance to elucidate the nature of evolutionary change to a broad audience so that he could mock (in, I suspect, a dishonest manner) someone and so win a point in an exchange of letters. I will refrain from an evolutionary psychological analysis of how this was truly "rational" from Orr's individual perspective....