Re: Not Exactly Topical, But I've Got Nothing...

1

It does so. All the people over there talk weird because their PARENTS talk weird, LB. Duh.


Posted by: Cala | Link to this comment | 07-13-07 9:30 AM
horizontal rule
2

Hm, I've argued with you, but I think I haven't made that specific argument. The first part of the argument I've made: the idea that average differences in brain structure would not lead to average differences in behavior I find sort of crazy. That said, obviously heritability studies (including twin studies) are a terrible way to determine if something is genetic or environmental. It's also crazy to say that brain development is not influenced by genetics (see: Williams syndrome, among others). Given those two facts, it seems pretty much certain to me than genetics does influence brain development in "normal" people in ways that we'll probably be able to figure out iin detail at some point. However, saying that there's an obvious connection between X area in your brain being this size or that size or connected this way or that way and your cognitive performance in any kind of real world scenario is pretty much horseshit. If there is a connection, it's incredibly non-obvious and (as far as I can tell) impossible to test for until we know a hell of a lot more about how the brain works.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 9:34 AM
horizontal rule
3

Why are twin studies so bad? Because the twins land in similar environments?


Posted by: heebie-geebie | Link to this comment | 07-13-07 9:37 AM
horizontal rule
4

unlikely to change significantly after adolescence

Unless you're Madonna.


Posted by: ogged | Link to this comment | 07-13-07 9:37 AM
horizontal rule
5

"The quality, for those who haven't clicked through yet, is accent, which clearly has no genetic component whatsoever. ..."

Why is this clear?


Posted by: James B. Shearer | Link to this comment | 07-13-07 9:38 AM
horizontal rule
6

Aaaaaaalso I should mention that there is obviously nothing genetically "determined" by the brain. Brain development happens environmentally, and while brain development is not 100% plastic (the consistency of brain organization across species has to mean something), the level of plasticity is pretty astonishing, as seen in hemispherectomy patients.

So, yeah, comity.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 9:38 AM
horizontal rule
7

3: twins spend several months in the same environment even if they're seperated at birth, among other things.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 9:39 AM
horizontal rule
8

re: 5

Because it doesn't?


Posted by: nattarGcM ttaM | Link to this comment | 07-13-07 9:39 AM
horizontal rule
9

...and it's obvious?


Posted by: heebie-geebie | Link to this comment | 07-13-07 9:40 AM
horizontal rule
10

it's incredibly non-obvious and (as far as I can tell) impossible to test for until we know a hell of a lot more about how the brain works.

I think this is the real point. People want to make policy decisions on the basis of something that is fixed and rigorous, and make claims--which, in my limited experience, some set of practitioners are happy to have them make--that are much stronger than the state of the art allows.


Posted by: SomeCallMeTim | Link to this comment | 07-13-07 9:40 AM
horizontal rule
11

10: well, I think this is the world of scientific rainmaking intersecting with the world of political manipulation of science. Steven Pinker has every incentive to say he's figured out how the brain works, because insofar as people believe him that will mean more money and prestige will flow to him and those who follow his theories. At the same time, politicians who want to use his conclusions to their own ends have every incentive to believe him when he says this. Simultaneously, the field is young enough (and wide open enough) that it really isn't possible to say in a rigorous way that he's wrong, so there are no strong voices to say "this ridiculous book is ridiculous, and here's why".


Posted by: Sifu Tweety | Link to this comment | 07-13-07 9:44 AM
horizontal rule
12

Why is this clear?

James, if you actually find this difficult, let me assure you that in Britain, where a great many profoundly distinct accents are crowded into an area as big as a medium sized state, the experience of people with one regional accent moving a hundred miles and bringing up children who, from contact with their friends, develop a totally different one, is so common that nobody would regard it as worthy of comment.


Posted by: OneFatEnglishman | Link to this comment | 07-13-07 9:45 AM
horizontal rule
13

2: Yeah, I'm taking the blogger's privilege of arguing with composite straw-men here. This:

Given those two facts, it seems pretty much certain to me than genetics does influence brain development in "normal" people in ways that we'll probably be able to figure out iin detail at some point.

I can agree with just fine. I'm arguing against anyone who thinks that common-sense observation is going to get us anywhere with what those connections are.

Really, more than our recent conversation, I was thinking of a couple of interchanges I've had with baa, in which he asked why the non-existence of innate differences in mental capacities/behaviors between the sexes should be privileged as the null hypothesis -- that is, maybe there's no proof either way, but it's just as wrong to take the lack of proof as evidence that the sexes are 'equal' in innate mental capacities as it would be to take it that they aren't.

And I suppose my response to that is that I'm not sure what it means to 'privilege something as the null hypothesis' outside of the context of a particular research project. But that once we accept that there really isn't a lot of scientific evidence drawing causual connections between measureable brain-structure average gender differences and behavioral differences between the genders, we're down to common-sense 'what seems likely to me.' And common-sense doesn't work terribly well in this area.


Posted by: LizardBreath | Link to this comment | 07-13-07 9:46 AM
horizontal rule
14

13: It seems like you're being too nice. It's even worse, because taking common-sense as your null hypothesis means you're starting from a basis which has been poisoned by cultural norms. Right? Right? I'm so right.


Posted by: heebie-geebie | Link to this comment | 07-13-07 9:49 AM
horizontal rule
15

so there are no strong voices to say "this ridiculous book is ridiculous, and here's why".

I think partly this is because no one wants to say "Yeah, we just don't know yet," because it has status implication for the field. Also, we've been screwed up by our consumer engineers: they keep doing things that seem extraordinary to the lay population (me) but that aren't (to the best of my knowledge) based on huge advances in an underlying scientific field. All of which is to say that not only shouldn't tom get a grant, but that he must be stopped for the good of the Republic.


Posted by: SomeCallMeTim | Link to this comment | 07-13-07 9:51 AM
horizontal rule
16

there are no strong voices to say "this ridiculous book is ridiculous, and here's why".

Philosophers do a fair bit of that, however it's by no means clear how much they are listened too.


Posted by: nattarGcM ttaM | Link to this comment | 07-13-07 9:54 AM
horizontal rule
17

14: Well, yeah. Common-sense beliefs about group differences have been so often wrong in the past that I take them as negative evidence, rather than positive evidence. But it's hard to argue with someone who believes "But that was in the past! We're all enlightened now, and our common sense is reliable!" or some reason why, e.g., early 20th century beliefs that eastern European Jews were intellectually inferior don't tend to discredit modern beliefs about group differences.

(By the way, baa, my apologies if I've misparaphrased anything you've said in my comment 13. Please correct me if I've misrepresented you.)


Posted by: LizardBreath | Link to this comment | 07-13-07 9:57 AM
horizontal rule
18

12

This just shows a substantial environmental component to accent. It does not show there is no genetic component "whatsoever". Is it really the case that the speech of adoptive children of different races is identical to native speakers? That there are no similarities between the speech of adopted children and that of their natural parents? I would like some citations. It certainly is not "obvious".


Posted by: James B. Shearer | Link to this comment | 07-13-07 9:59 AM
horizontal rule
19

I always like it when there's some scientist or book from 1915 or 1850 or whenever who proclaims that science is finished, and there's nothing left to discover.


Posted by: heebie-geebie | Link to this comment | 07-13-07 9:59 AM
horizontal rule
20

SCTM is making uniformly good points. But surely there's some compelling evidence for meaningful, biologically-based mental differences between the sexes: we know that we're issued various sex hormones in varying amounts, and that changes in those amounts can affect our minds, right?

I agree that drawing specific conclusions from that is a bad idea, and drawing policy conclusions is a worse one. But the most likely answer seems to me to be that genetically-determined brain differences between the sexes are probably real, although obviously slight.


Posted by: Tom | Link to this comment | 07-13-07 10:01 AM
horizontal rule
21

re: 18

I think you are thinking of something other than 'accent'. There are probably aspects of someone's voice that resemble one's parents even if raised with a different accent. In the sense that the size of your vocal tract, the length of your vocal chords, the size and shape of the various chambers in which the voice resonates, and so on, are going to be (partly) genetically determined. I wouldn't describe those things as forming part of an accent.


Posted by: nattarGcM ttaM | Link to this comment | 07-13-07 10:01 AM
horizontal rule
22

That there are no similarities between the speech of adopted children and that of their natural parents?

Only insofar as the adopted children and biological parents share a common intelligence.


Posted by: heebie-geebie | Link to this comment | 07-13-07 10:01 AM
horizontal rule
23

Accent is genetic, so is language generally. But it's a genetic capactity that requires social context and interaction to be activated. The early interaction context where the language capacity is first used determines how the individual will realize their capacities. The problem is that the whole model which separates "genetic" and "environmental" is not a good way to think about things. Genetic capacities require environmental settings to be realized.


Posted by: marcus | Link to this comment | 07-13-07 10:04 AM
horizontal rule
24

Incidentally, a lot of the things we think of as fairly fixed about our voices -- the 'set' of our larynx, for example -- can be altered. Once you've learned how to raise and lower it at will it can be quite fun playing with it.


Posted by: nattarGcM ttaM | Link to this comment | 07-13-07 10:04 AM
horizontal rule
25

So I should read the LL post before commenting, but my first response would be that the "accent" case has biases that we have the ability to adjust for in other (gender behavior, IQ) cases.

1. We have cross-cultural comparisons for differences in gender behavior. That's not the case for accent, right?
2. We have twin studies, which, though not perfect for the reasons Sifu Tweety notes are nonetheless pretty darn suggestive. If you find that separated identical twins are more similar on some category X than are fraternal twins raised together, doesn't that give you reason to believe that category X has a really strong genetic component? I can think of ways where this would not be true, but in general, yes. I would bet a zillion dollars that fraternal twins raised together have very, very similar accents, while identical twins raised apart don't. That's a kill shot for the genetic hypothesis, right?


Posted by: baa | Link to this comment | 07-13-07 10:04 AM
horizontal rule
26

And I suppose my response to that is that I'm not sure what it means to 'privilege something as the null hypothesis' outside of the context of a particular research project. But that once we accept that there really isn't a lot of scientific evidence drawing causual connections between measureable brain-structure average gender differences and behavioral differences between the genders, we're down to common-sense 'what seems likely to me.'

I don't think it's common sense, I think it's Occam's Razor, which again has it's flaws, but at least has the veneer of scientific rationality. Given the facts that we do know (gendered behavioral differences in other animals, sexual dimorphism, measured average differences in brain structure, differential performance on laboratory cognitive tests), it seems like the simplest answer is that yes, of course there are average cognitive & behavioral differences between the sexes. However, to say that we have any idea what those differences might be in practice seems, to me, laughable. For all we know women are dramatically better at math and science tasks because they had to do all the measuring for food (or whatever). We simply don't know enough about how the brain actually works.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 10:06 AM
horizontal rule
27

FUCK YOU, APOSTROPHE!


Posted by: Sifu Tweety | Link to this comment | 07-13-07 10:07 AM
horizontal rule
28

There's no convincing research-based evidence causually connecting average behavioral differences between the sexes to genetically determined sex physical sex differences in brain structure.

Is there convincing research-based evidence connecting any average behavioral differences at all to physical differences in brain structure? Outside of a few rare brain diseases and injuries, I mean? Don't we just not know very much about how the brain actually works? (As Tweety points out above). It seems to me you're sort of taking refuge in the limits of current science here. Which I'll admit is tempting given this silly tendency to talk as though we have this grand theory that outlines deterministic differences between men and women, low and high-IQ, etc.


Posted by: marcus | Link to this comment | 07-13-07 10:13 AM
horizontal rule
29

18: You've clearly never met any of the SoCal girls who were adopted from China or South Korea at an age of a couple months and then raised in the whitest of whitebread communities.


Posted by: Po-Mo Polymath | Link to this comment | 07-13-07 10:15 AM
horizontal rule
30

22

What are you saying? Only differences attributable to the differences in speech between high IQ and low IQ people?


Posted by: James B. Shearer | Link to this comment | 07-13-07 10:15 AM
horizontal rule
31

Don't we just not know very much about how the brain actually works? (As Tweety points out above). It seems to me you're sort of taking refuge in the limits of current science here.

Well, exactly. The limits of current science means that we don't have science-based answers for this sort of question; we're all bullshitting about 'what seems likely to me.'


Posted by: LizardBreath | Link to this comment | 07-13-07 10:16 AM
horizontal rule
32

It's nice to see that Cosma is writing casually again!
One thing that never ceases to amaze me about sex determination is how variable it is; chickens and other birds have sex determined in a way completely differently from the primates, many fish undergo sex reversal in response to environmental cues, with temperature being a common one, and many plants can self-fertilize.
More relevantly to humans and the policy discussions that matter for us, sexes equal above the neck is wrong, but the evidence to support a particular convoluted causal chain connecting anything physiological with behavior is very weak, so none can say what's unalterably unequal between sexes in any useful way. It's contentious politicized research working from weak bases, to me no more fun to read than the bogus power lines cause cancer stuff and its tedious rebuttal.


Posted by: lw | Link to this comment | 07-13-07 10:16 AM
horizontal rule
33

18, I think the empirical question of whether accent has any genetic component whatsoever is sort of beside Shalizi's point. We certainly don't feel compelled to conclude that it has a genetic component, despite the fact that it is "hereditable", strongly correlated with race/ethnicity, etc. Thus we should not feel compelled to conclude that such correlations in other characteristics, standing on their own, entail a genetic component.


Posted by: suds | Link to this comment | 07-13-07 10:17 AM
horizontal rule
34

25

"... I would bet a zillion dollars that fraternal twins raised together have very, very similar accents, while identical twins raised apart don't. That's a kill shot for the genetic hypothesis, right?"

It would just show the genetic influence is not dominant not that it is nonexistent.


Posted by: James B. Shearer | Link to this comment | 07-13-07 10:18 AM
horizontal rule
35

What are you saying? Only differences attributable to the differences in speech between high IQ and low IQ people?

I was just trying to be circular.


Posted by: heebie-geebie | Link to this comment | 07-13-07 10:18 AM
horizontal rule
36

23: Accent is genetic, so is language generally.

This is misleading. The ability to learn language, and to learn to speak with an accent, is something humans are genetically predisposed to be able to do. The fact that any particular human being speaks with a particular accent has nothing to do with whether their ancestry is European, Chinese, African, etc.


Posted by: My Alter Ego | Link to this comment | 07-13-07 10:19 AM
horizontal rule
37

And again, I don't want to come off as advocating a "genes determine all" view. Even for characteristics like height, where a naive view might have been genetic determinism, it's clear that environmental factors are enormously important. I do reject the ideas that:

a. when we observe differences between two human population groups, a genetic explanation has a higher standard of proof than a environmental explanation
b. we have no reliable methods of getting a sense of the relative importance of genetics/environment to an observed variation in some characteristic


Posted by: baa | Link to this comment | 07-13-07 10:20 AM
horizontal rule
38

28: well, sure, there's lots of evidence when you look at people with brain injuries. There's also a lot of evidence from fMRI studies. We know extremely well that (for instance) a reduction in size of the orbitofrontal cortex correlates to a reduction in social ability, or that loss or damage to Broca's area causes a specific kind of language dysfunction. What we don't have is a useful overall model of how the brain works, how these areas talk to each other, what the actual processing is that is happening when you do a certain task. More than that, we don't have a particularly solid connection between neuroscience and the framework used by cognitive psychology, which means that the theoretical models underlying all of our current models of behavior could turn out to be completely wrong.

I say this, obviously, from the neuroscience point of view.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 10:20 AM
horizontal rule
39

a. when we observe differences between two human population groups, a genetic explanation has a higher standard of proof than a environmental explanation

I think that's the bone of contention. I'm on the other side. Except for the "men must spread their seed, women must tend the hearth," which is obviously true.


Posted by: SomeCallMeTim | Link to this comment | 07-13-07 10:22 AM
horizontal rule
40

Tim (in 39). Is that because of

1. the practical results of attributing responsibility to genetic factors is more likely to be negative? (worse results when you are wrong, horrible history of evil resulting from attributing "innate" goodness/badness to favored/disfavored groups)

2. as a fact of the matter you believe that most differences are environmental, and therefore that should be the default position?


Posted by: baa | Link to this comment | 07-13-07 10:27 AM
horizontal rule
41

a. when we observe differences between two human population groups, a genetic explanation has a higher standard of proof than a environmental explanation

This is setting up a false dichotomy, in light of your b. if nothing else. There is no such thing as a strictly genetic or strictly environmental trait. Genes influence the way the environment influences the individual. The issue is whether you can determine what this influence is, and what it's causal relationship to various specific traits is.

b. we have no reliable methods of getting a sense of the relative importance of genetics/environment to an observed variation in some characteristic

We have somewhat reliable methods, similar to how we have somewhat reliable methods for making economic predictions. In both cases, the reliability is predicated on the assumptions of the underlying model which, in the case of genetics (and cognitive science), could well be massively flawed.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 10:27 AM
horizontal rule
42

33

I don't understand your point. People don't believe IQ has a strong genetic component just because there is a similarity between parents and children. Heritability estimates depend on things like comparing fraternal twins to identical twins or adoptive children to their natural and adoptive parents. These estimates indicate a strong genetic component to IQ but not to accent.


Posted by: James B. Shearer | Link to this comment | 07-13-07 10:29 AM
horizontal rule
43

Does JBS have the slightest bit of evidence that there is any genetically heritable aspect to accent (as opposed to voice timbre)? Or is he just saying that overwhelming anecdotal evidence doesn't count anything against, er, non-evidence?

That said, it'd be easy to conduct an experiment once and for all. Take a longish script from a popular novel. Give it to a group of 50 girls from America who are white and raised by their natural parents, and 50 girls from China who were raised by white American adoptive parents. Blindfold a listener. Have girl 1 (Chinese adopted American) read the script, then girl 2 (white American) read the script, then randomly choose girl 3 from the group and have her read the script. Listener decides who she sounds most like -- girl 1 or girl 2. Process results.


Posted by: Epoch | Link to this comment | 07-13-07 10:31 AM
horizontal rule
44

43: he's just baiting. Note 42.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 10:32 AM
horizontal rule
45

Sifu Tweety:

Agree with your first comment in 41 completely -- sloppy wording on my part. Genes/environment aren't either/or, and the feedback loops you describe can in principle be enormously important. (Flynn has written on this)

Agree likewise with the second point, which is why having multiple sources of data is important. On a gender difference, e.g., when cross-cultural comparisons and historical comparisons, and animal behavior, and a physiologic hypothesis all come together, you've got something.


Posted by: baa | Link to this comment | 07-13-07 10:33 AM
horizontal rule
46

45.2: but none of those is a particularly strong source of data. Weak evidence plus weak evidence plus weak evidence adds up to weak evidence. I don't disagree that there are almost certainly behavioral difference, but to say that the science is in place such that we know what they are seems to me a vast overreach, especially when so many of the sources of data are based on statistical methods that are subject to observer bias (to say the least).


Posted by: Sifu Tweety | Link to this comment | 07-13-07 10:36 AM
horizontal rule
47

'privilege something as the null hypothesis'

It's interesting to me that the "null hypothesis" in discussions of racial inequality is that black people are inherently inferior to white people. Why not the reverse?

Jared Diamond doesn't talk much about evo psych, but when he does, he's just as dopey as the rest of them. It amuses me, though, that he proposes inherent intellectual superiority New Guineans - and he makes his case, too, in the same evo psych terms that are used to justify white supremacy.


Posted by: politicalfootball | Link to this comment | 07-13-07 10:37 AM
horizontal rule
48

baa humbug.


Posted by: heebie-geebie | Link to this comment | 07-13-07 10:37 AM
horizontal rule
49

44: Well, yeah. But I kind of liked my experimental design, so I thought I'd throw it out there.


Posted by: Epoch | Link to this comment | 07-13-07 10:37 AM
horizontal rule
50

I would like some citations. It certainly is not "obvious".

At this time I would like to present this remark with the Asperger's Award of the Year. Thank you, remark, for making all our lives a little more OCD.


Posted by: Populuxe | Link to this comment | 07-13-07 10:39 AM
horizontal rule
51

Look, the problem is that no one clearly defines their terms when talking about what is meant by "genetic" vs. "environmental". Almost all human characteristics (from things like height or skin color to language and accent) are the result of the interaction between genes and and environment, not one alone. The question people are usually interested in is how such characteristics respond to specific kinds of variance in environment and genes.

Three-toed sloth is taking issue with a political, not a scientific, point concerning heritability and lack of plasticity. But it's obvious that heritability and lack of plasticity after a certain age are true of not just accent, but most important characteristics that vary with culture (like what food and music you like). In fact, I would argue that it's quite often the case that genetic characteristics are more easily modified than cultural ones.

The problem is that there are two totally separable debates, one about how a characteristic changes as both genetic inheritance and environment are jointly modified (or one is held constant and the other is modified, etc.). The answer to that will be different depending on the exact types of modification, the range of variation, etc. The other is about the plasticity of human characteristics to policy interventions. The two debates are very different and in some ways have little to do with each other.


Posted by: marcus | Link to this comment | 07-13-07 10:39 AM
horizontal rule
52

40: Both, really. Primarily, it's number one, but I've been astonished in my life by how often I've been wrong about some set of assumptions I had that were usually based in some genetic-like claim. I know I have a bias in that direction, I guess. I don't know if we're much more plastic than I thought, but I think I located the places of plasticity very badly. (Really, this follows I assume from Sifu's point about the lack of a well-described, rigorous consensus theory.)


Posted by: SomeCallMeTim | Link to this comment | 07-13-07 10:39 AM
horizontal rule
53

But of course those things don't come together for any specific claim of innate gender differences. "Animal behavior" can get you to 'gender differences aren't overwhelmingly unlikely', but not to any generalizable prediction of that they're likely to be -- the analogy between, e.g. forming mostly female family groups that take in and support one or two males at a time, and anything humans do universally, is going to be pretty weak. 'Physiologic hypotheses', that is, something you could test for a causual connection between physiological gender differences and behavioral gender differences, are, as I understand it, still pretty thin on the ground. And the cross-cultural comparisons get you right back to 'common sense'.


Posted by: LizardBreath | Link to this comment | 07-13-07 10:41 AM
horizontal rule
54

The other is about the plasticity of human characteristics to policy interventions. The two debates are very different and in some ways have little to do with each other.

I think I agree with that entirely.


Posted by: SomeCallMeTim | Link to this comment | 07-13-07 10:44 AM
horizontal rule
55

53 to 45. To 47:

It amuses me, though, that [Jared Diamond] proposes inherent intellectual superiority New Guineans - and he makes his case, too, in the same evo psych terms that are used to justify white supremacy.

I've seen people say this before, and maybe he does it elsewhere than in Guns Germs and Steel. But in GGS, all I understood him to say was that New Guinean tribespeople are patently, if you get to know them, not intellectually inferior to Joe EuropeanOrigin, and are in fact likely to be able to perform intellectual feats that he can't (e.g., accurately identifying hundreds of plant species) because they need to to survive. That's not a claim that they're genetically superior, it's a claim that their environment places survival demands on them that they've responded to culturally by learning this stuff.


Posted by: LizardBreath | Link to this comment | 07-13-07 10:46 AM
horizontal rule
56

53: testing of physiological differences is subject to environment too, of course, because where are you getting your sample set from, are they infants, etc.

This is why I'm more interested in brain disorders. Who gives a crap how normal healthy people act? Too much noise.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 10:47 AM
horizontal rule
57

43

And you think it is obvious such an experiment would find no difference between the native and adopted groups?


Posted by: James B. Shearer | Link to this comment | 07-13-07 10:48 AM
horizontal rule
58

Weak evidence plus weak evidence plus weak evidence adds up to weak evidence.

I don't think that's true. If it were, repeating a study 45 times and getting the same weak correlation wouldn't make you feel any better than seeing a weak correlation in one study. If that were true getting the same weak correlation in studies performed 5 different ways wouldn't give you more confidence than seeing the same result in a study conducted in a single way.

But of course those things don't come together for any specific claim of innate gender differences.

Predisposition towards violence/aggressive behavior, perhaps?


Posted by: baa | Link to this comment | 07-13-07 10:48 AM
horizontal rule
59

45.2, 46: Don't forget to factor in observer bias. People find what they're looking to find, and we're all programmed to differentiate between men and women; trying not to requires conscious and deliberate effort.


Posted by: bitchphd | Link to this comment | 07-13-07 10:51 AM
horizontal rule
60

58.1: well, that's a different issue. When you repeat the same study with the same conditions each time, you increase your confidence. When you take various, wildly diverging methodologies, each of them with weak correlation, and you combine them, you don't actually get more confidence. You just get a heterogenous set of weak correlations. They might reinforce each other to some degree, but it's not the same statistical animal.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 10:51 AM
horizontal rule
61

59: that's why I mentioned it.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 10:51 AM
horizontal rule
62

This is why I'm more interested in brain disorders.

Lucky you, to have so many subjects right here to observe in their native habitat.


Posted by: bitchphd | Link to this comment | 07-13-07 10:52 AM
horizontal rule
63

18 I'm not sure if you are being deliberately obtuse here, but I will rise to the bait anyway. A single data point from my experience suffices as proof for me that there is no genetic component to accent. Some family friends, in Southwest Germany, adopted two infants: one from Korea, and one from Vietnam (who is pretty obviously the offspring of a Vietnamese woman and an African-American soldier). The two children, despite their heterogenous genetic background, speak with a schwaebisch accent as thick as a board--exactly like all their neighbors. (And yes, my ear for German regional accents is good enough to distinguish them.)


Posted by: Knecht Ruprecht | Link to this comment | 07-13-07 10:53 AM
horizontal rule
64

#57 Define "difference."


Posted by: Populuxe | Link to this comment | 07-13-07 10:53 AM
horizontal rule
65

58: You think 'animal behavior' leads you to think that females are generally non-violent? There are a lot of eviscerated antelopes out there that would disagree; and they're more likely to have been eviscerated by a lioness than a lion.

You've got to start making much narrower claims about aggression before you can get anywhere with 'animal behavior'. And for the human evidence, I'd argue that you'd have to control for body-size/strength and the cultures produced by varying body-size/strength, before you could make any claims about innately differing purely mental tendencies toward aggression. I don't think that control can be done, but I think that it would be necessary to get useful data.


Posted by: LizardBreath | Link to this comment | 07-13-07 10:54 AM
horizontal rule
66

Actually, I read all of three-toed sloth's post now, and it is quite good. Especially toward the end, where he (she?) defines IQ as a kind of learned skill in organized, bureaucratized societies and cultures which emphasize abstraction. Quite true.

But questions remain. Say we've decided we want people in our culture to learn the IQ game, so they can operate more effectively in big bureaucracies. The question is how effectively can everyone learn to play this game, so they don't get left behind. Perhaps IQ is kind of like basketball -- there are lots of ways for all kinds of people to effectively play it, but it probably helps to be genetically tall.

Of course, one thing people never mention is that IQ alone accounts for a relatively small minority of variation in adult income once other factors are controlled for (I think something like 10% (?) once education, parental income, etc. are entered in a multiple regression equation). So clearly there are numerous other factors operating, and freaking out about IQ alone is probably a sign that you are looking for a political excuse to leave people behind socially.


Posted by: marcus | Link to this comment | 07-13-07 10:54 AM
horizontal rule
67

61: Ah, so you did. Still, it deserves emphasis. My bad, though, especially since I really think I should just stay out of this thread altogether.


Posted by: bitchphd | Link to this comment | 07-13-07 10:54 AM
horizontal rule
68

Yeah, so far I haven't wanted to disagree with Tweety about anything but his past bicycling habits.


Posted by: LizardBreath | Link to this comment | 07-13-07 10:55 AM
horizontal rule
69

I really think I should just stay out of this thread altogether

. . . in order to retain my feminine propensity towards non-violence.


Posted by: bitchphd | Link to this comment | 07-13-07 10:55 AM
horizontal rule
70

19 makes a good point. One sure-fire way to be quoted a hundred years from now is to make a prediction so wrong-headed that people will hold it up for ridicule.

"Science is done. We know everything we're going to know. We can shut the labs down. Everyone can go home."

-- Walt Someguy, Unfogged comment board thread, 2007.


Posted by: Walt Someguy | Link to this comment | 07-13-07 11:02 AM
horizontal rule
71

63

This proves nothing. They could have the same blood type also but that would not prove there is no genetic component to blood type.


Posted by: James B. Shearer | Link to this comment | 07-13-07 11:03 AM
horizontal rule
72

The issue is whether you can determine what this influence is, and what it's causal relationship to various specific traits is.

Yup yup yup. Murray's problem is not that he's wrong, it's that he has no evidence that he's right.



Posted by: politicalfootball | Link to this comment | 07-13-07 11:03 AM
horizontal rule
73

re: 71

Are you being deliberately dumb?

The analogous case with blood-type would be if they both turned out to have the same blood type as their adoptive parents and not their birth parents. And that would be prima facie evidence that blood type wasn't strongly heritable.


Posted by: nattarGcM ttaM | Link to this comment | 07-13-07 11:07 AM
horizontal rule
74

57: Yup, my money would be pretty strongly on the null hypothesis there.


Posted by: Epoch | Link to this comment | 07-13-07 11:07 AM
horizontal rule
75

The two children [...] speak with a schwaebisch accent as thick as a board

Oh man, poor kids.


Posted by: Jackmormon | Link to this comment | 07-13-07 11:10 AM
horizontal rule
76

64

A greater than 50% probability that the third subject will be matched with the subject (from the first two) of the same race. Similar experiments are done to determine whether people can tell the difference between Coke or Pepsi etc.


Posted by: James B. Shearer | Link to this comment | 07-13-07 11:10 AM
horizontal rule
77

ttaM: are you new here?


Posted by: Walt Someguy | Link to this comment | 07-13-07 11:10 AM
horizontal rule
78

re: 77

Good point.


Posted by: nattarGcM ttaM | Link to this comment | 07-13-07 11:12 AM
horizontal rule
79

poor kids

Actually, they are both adults now. My phrasing was a little ambiguous on that point.


Posted by: Knecht Ruprecht | Link to this comment | 07-13-07 11:12 AM
horizontal rule
80

That's not a claim that they're genetically superior, it's a claim that their environment places survival demands on them that they've responded to culturally by learning this stuff.

I haven't got GG&S handy, but I think we're talking about the same passages, which, as I recall, were primarily in the foreward. I think I can justify my claim - though I acknowledge that he doesn't dwell on the idea or really develop it, and it's pretty incidental to his purpose. It just amused me, for once, to see brown people come out on top, and the fact that nobody seems to be researching this sort of theory tells you a lot about the motivation of evo psych types.

Because GG&S provides such a fine rebuttal to a lot of evo psych arguments, it's hard to imagine Diamond being infected by this bullshit, but he is.
In a ludicrous chapter in Third Chimpanzee he tells us the purpose of substance abuse among males is to impress chicks.


Posted by: politicalfootball | Link to this comment | 07-13-07 11:16 AM
horizontal rule
81

73

"... And that would be prima facie evidence that blood type wasn't strongly heritable."

The claim I was objecting to was that accent "clearly has no genetic component whatsoever" not a claim that accent isn't strongly heritable.


Posted by: James B. Shearer | Link to this comment | 07-13-07 11:18 AM
horizontal rule
82

81: Can you offer a scintilla of evidence, even anecdotal, that there is any component of accent that can't be explained by environmental factors? Otherwise Occam's razor obtains.


Posted by: Knecht Ruprecht | Link to this comment | 07-13-07 11:25 AM
horizontal rule
83

Has anyone responded to 6 yet? Something like "brain development happens environmentally," while wholly plausible, is by no means necessarily true. It's plausible that a lot more of the micro-structure of the brain is 'genetically' determined than you think.

The example I've seen someone give of this is a comparison of the bronchial tube structure in lungs vs. (say) the duct structure in breasts.

Both are complex, branching structures in a surrounding tissue. Superficially, both seem to be 'random'... but (apparently, and I'm no expert on this) the structure of the lungs is much more determined, down to some fairly high level of detail, than the structure of the ducts in breasts, which are much more the result of a general random process (fingerprints, no two alike, etc).

And I don't think that we have a handle on how the meso- or micro-structure of the brain develops at all, or at least enough to be able to say with certainty that it's more random than not.

Which is not to say that anything that LB or Cosma are writing is wrong at all... it's just that "obviously brain structure is all environmental" isn't quite so obvious, at least yet.

Or is there a neurobiologist lurking around here who can correct me?


Posted by: arthegall | Link to this comment | 07-13-07 11:32 AM
horizontal rule
84

82: A lack of evidence doesn't PROVE that accent isn't genetically determined! Also, can you offer any evidence that JBS lacks such evidence? If not, you can prove nothing!


Posted by: bitchphd | Link to this comment | 07-13-07 11:32 AM
horizontal rule
85

Relatedly?


Posted by: Amber | Link to this comment | 07-13-07 11:33 AM
horizontal rule
86

Link fixed.


Posted by: Amber | Link to this comment | 07-13-07 11:34 AM
horizontal rule
87

84 - Is it irresponsible to speculate? It would be irresponsible not to.


Posted by: snarkout | Link to this comment | 07-13-07 11:38 AM
horizontal rule
88

86: I think this is the key passage:

"We believe that this correlation may reflect some sort of predisposition or cognitive bias induced by the two genes in question. We don't have any detailed idea of what this bias might consist of, but we assume it is very small and would only manifest itself in language change over many generations. We know, of course, that any normal human infant can learn the language of any human community that it's brought up in - genes don't play any role at the individual level. "


Posted by: | Link to this comment | 07-13-07 11:38 AM
horizontal rule
89

I have less than skimmed this thread, and I haven't read the article but I do want to say that I am sympathetic to arthegall's point in 83.

Many African languages, I'm told, have clicking sounds which aren't very common in other regions. Many African-Americans and other Black people pronounce 'ask' as 'aks'. I have often wondered whether these aren't related, and--since most of these people have never been to Africa--I've often wondered whether there wasn't some genetic component to this.

(It could just be that people of African descent have tended to stick together which might suggest that it's just one long environmental influence. I did meet a Ghanaian Etonian/ Cambridge graduate in college on a one year fellowship, and I don't think that I ever heard him say anything other than ask. That might support the environmental thesis, but I'm not sure.)

I'm just thinking aloud here.


Posted by: Bostoniangirl | Link to this comment | 07-13-07 12:10 PM
horizontal rule
90

89: Pittsburghers, white and black alike, call it Piksburgh. I'm having trouble with your !Kung thesis.

Interesting, barely related note: if children are exposed to music education early enough, virtually all of them (85% is the number I heard) can develop enough skill to play professionally (like, symphony orchestra, if not a distinguished one). The window closes around age 5-7 (not unlike foreign languages).

What's fascinating to me about this is that, while the analogy to the well-known language aptitude of toddlers is obvious, we mostly think of musical talent as innate. But even the tone-deaf could have learned, if they'd been taught early enough.


Posted by: JRoth | Link to this comment | 07-13-07 12:20 PM
horizontal rule
91

66

"Of course, one thing people never mention is that IQ alone accounts for a relatively small minority of variation in adult income once other factors are controlled for (I think something like 10% (?) once education, parental income, etc. are entered in a multiple regression equation) ..."

This is potentially misleading. If for example IQ influences how much education you get then controlling for education understates the importance of IQ.


Posted by: James B. Shearer | Link to this comment | 07-13-07 12:20 PM
horizontal rule
92

89: People who say "ask" use the same number of Khoisan clicks as those who say "aks." That number is zero.

But even if the "k" sound and the click were somehow related, my first sentence would still be true.


Posted by: politicalfootball | Link to this comment | 07-13-07 12:33 PM
horizontal rule
93

92: As I said, I was just thinking aloud.

90: Damn, my linguistic aptitude is pretty good. I can pick up languages pretty easily, and my accent is generally good. (When I'm tired or away from native speakers, I speak any foreign language from the back of my throat. I think that this is because my first foreign language was French, and I just default to speaking foreign languages with a French accent.) My musical skill is not great. I'm not totally tone deaf, but my capacities vary a lot from day to day, and my sense of rhythm stinks.


Posted by: Bostoniangirl | Link to this comment | 07-13-07 12:42 PM
horizontal rule
94

90.---Hence all of those posh music schools in Manhattan for TWO AND THREE YEAR-OLDS MY GOD PEOPLE THIS IS INSANITY


Posted by: Jackmormon | Link to this comment | 07-13-07 12:51 PM
horizontal rule
95

The "ask"/"aks" alternation goes back to Old English and the prevalence of the "aks" pronunciation in AAVE is unlikely to have anything to do with clicks, which as pf points out are very different sounds and are also not found in the languages of the areas where the vast majority of American slaves came from.


Posted by: teofilo | Link to this comment | 07-13-07 12:51 PM
horizontal rule
96

This is potentially misleading. If for example IQ influences how much education you get then controlling for education understates the importance of IQ.

The point of multiple regression is precisely to try to handle this statistically. So long as you have people of different education levels and the same IQ, or different IQ levels and the same education, you can isolate separate effects statistically for both. Not controlling for education (one of the things that was done in the Bell Curve) massively overstates the impact of IQ.


Posted by: marcus | Link to this comment | 07-13-07 12:54 PM
horizontal rule
97

93: Oh, I don't think it's the same skillset (maybe neurally related, maybe not), but similar elasticity in the young: everyone can learn foreign languages when very young, only a few have BG skillz as an adult (my MIL is astonishing that way). Same deal with music.

My wife's ex-BF spent 2 years doing Peace Corps in Ghana, and developed a terrible (subconscious) habit of speaking Pidgin (or at least Ghana-accented English) to anyone with an obvious accent - of any nationality. Probably actually worse than speaking SLOWLY and LOUDLY to the cabana owner down at the corner.


Posted by: JRoth | Link to this comment | 07-13-07 12:55 PM
horizontal rule
98

re: 65 Are you fighting the point on human male/female differences in aggressive/violent behavior having a genetic component, LB? Leave aside things we can't see (mental tendencies, etc) for the moment.

re:60
Sifu Tweety, I think it's just the case that aggregating data from different types of studies all of which points in the same direction makes you feel better. Sure, all the studies may be weak and have biases, but that's why you feel good if multiple different analyses give a similar result.

And none of this goes to the general point that the case of accent is very poor grounds for "total skepticism on common-sense arguments about innate qualities of human nature." I am a big believer that there are feedback effects between environment and genetics on all sorts of attributes. But from the twin studies we have awfully good evidence that genes exert a significant influence on IQ. For accent we don't. We also note that people raised in the same environment almost always have the same accent, but they don't always have the same IQ. Indeed, this is a case where "common sense" is a pretty good guide.


Posted by: baa | Link to this comment | 07-13-07 12:59 PM
horizontal rule
99

The click sounds...are also not found in the languages of the areas where the vast majority of American slaves came from

Indeed, West Africans are genetically closer to Europeans than they are to the Khoisan peoples (the ones with the klick sounds).


Posted by: Knecht Ruprecht | Link to this comment | 07-13-07 1:05 PM
horizontal rule
100

83: I said that brain development happens environmentally because it does. That is, the most important stage in the development of cognitive function (synaptic pruning) happens well after a child has left the womb. To (really) approximate what happens, the early brain creates a huge overabundance of synaptic connections, and the way you learn things, at least in early childhood, is that the connections that aren't used in the course of you responding to your environment atrophy and eventually disappear. There is a great of literature on how the micro structure of the brain develops, and it is not random at all: that's why there's commonality to the topography of functional specialization across (most) everyone's brain, and why people can say, e.g., the amygdala regulates the fear response (itself an oversimplification). What I meant was that genetics, insofar as they are involved, regulate the way the brain responds both to chemical cues in the womb, and physiological and environmental cues outside the womb. The actual growth and maturation of the brain happens in a rich, diverse environment, which is why you can't say that the pattern of brain development is explicitly genetic.

98.2: being comfortable is not the same as being accurate. Twin studies have big problems methodologically, for the environmental reasons mentioned above, as well as reasons of small sample size and relative homogeneity of post-natal environment.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 1:13 PM
horizontal rule
101

I don' think we are disagreeing, Sifu Tweety. 4 bad studies with lots of biases are in no way superior to 1 good study without those biases. I just think it's superior to 1 bad study with lots of biases.

I can't speak to the specifics of any twin studies, but in general when I see a similarity comparison on any attribute X that goes: identical twins raised apart > fraternal raised together > non-twin siblings raised together > genetically unrelated raised together (adopted), that seems to me like darn good evidence for the role of genes on attribute X.


Posted by: baa | Link to this comment | 07-13-07 1:36 PM
horizontal rule
102

brain development happens environmentally because it does

Links? Evidence? I bet you're just making this up because you don't want to admit that whites are naturally better at pronouncing English than blacks and that women are naturally better at talking to babies than men.


Posted by: bitchphd | Link to this comment | 07-13-07 1:37 PM
horizontal rule
103

96

"The point of multiple regression is precisely to try to handle this statistically. So long as you have people of different education levels and the same IQ, or different IQ levels and the same education, you can isolate separate effects statistically for both. Not controlling for education (one of the things that was done in the Bell Curve) massively overstates the impact of IQ. "

If you do this and education and IQ are correlated you will get three terms, the effect due to education, the effect due to IQ and the effect due to the correlation between IQ and education. The effect to due to the correlation between IQ and education can not be partitioned between IQ and education in a well defined way. If the correlation is caused by education raising your IQ then it should be assigned to education, if it is caused by higher IQ people seeking more education it should be assigned to IQ, if it is caused by some relationship with a third variable it should be left as a correlation effect. Ignoring eduction overstates IQ only to the extent that education is operating independently of IQ.


Posted by: James B. Shearer | Link to this comment | 07-13-07 1:39 PM
horizontal rule
104

101.2: well, you still have the problems of small sample size, potential homogeniety of adoptive households, dates of adoption. I would say you have some evidence that genes play a role in X, but it's really not very strong. There just aren't that many pairs of identical twins adopted as newborns and raised apart. Plus you'd want to know if identical twins raised together are more or less similar than those raised apart. You might also have different environmental influences (on average) in women who put their children up for adoption (e.g. higher incidence of fetal alcohol consumption). You might have different prenatal environmental influences on twins generally compared to non-twin siblings. You might have different in-womb dynamics between identical and fraternal twins (in terms of competition for maternal resources). Personally I'm much more comfortable when you actually have an underlying genetic explanation for why this might be so, and that certainly isn't there in the case of e.g. intelligence or spatial skills or anything like that.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 1:45 PM
horizontal rule
105

102: sheesh. Here. Search for "post-natal brain development". Part II here.

I ain't funnin' y'all ladies.


Posted by: Sifu Tweety | Link to this comment | 07-13-07 1:49 PM
horizontal rule
106

105: You ain't funnin at all, dude. You're so unfun you don't even recognize the fun when it's handed to you.

I bet your lack of fun is genetically determined.


Posted by: bitchphd | Link to this comment | 07-13-07 1:53 PM
horizontal rule
107

103 - James, people with both high IQ and high education know how to control for both effects simultaneously. We call those people "statisticians".


Posted by: Walt Someguy | Link to this comment | 07-13-07 1:55 PM
horizontal rule
108

Here's a useful discussion of the ask/aks phenomenon that backs up Teo's claim in 95 that it isn't something that developed in African-American English. Nut graf:

Old English had two forms, acsian and ascian, the former being the literary standard until about 1600, when the latter gained the imprimatur of being the high style variant. Specialists in AAVE agree that current African Americans who use /aks/ have inherited this as the base form of the verb.


Posted by: bza | Link to this comment | 07-13-07 1:58 PM
horizontal rule
109

106: just because I knew it was a joke doesn't mean citations aren't totally awesome!


Posted by: Sifu Tweety | Link to this comment | 07-13-07 2:01 PM
horizontal rule
110

Your desire for citations is genetically determined.


Posted by: bitchphd | Link to this comment | 07-13-07 2:12 PM
horizontal rule
111

My wife's ex-BF spent 2 years doing Peace Corps in Ghana, and developed a terrible (subconscious) habit of speaking Pidgin (or at least Ghana-accented English) to anyone with an obvious accent - of any nationality.

I feel his pain. It's worn off by now, a dozen years later, but any kind of a foreign accent used to send me into Samoan accented English. I hope it merely sounded psychotic rather than condescending.


Posted by: LizardBreath | Link to this comment | 07-13-07 2:15 PM
horizontal rule
112

"Your desire for citations is genetically determined"

Probably a touch of OCD, which is at least somewhat heritable.


Posted by: rilkefan | Link to this comment | 07-13-07 2:31 PM
horizontal rule
113

107 - we call the people who ask more from the data than is available "bad statisticians".


Posted by: rilkefan | Link to this comment | 07-13-07 2:34 PM
horizontal rule
114

113 "bad" s/b "Bayesian".

As for 100, fair enough, but I don't understand what would be "explicitly genetic," in this context. Pretty much all of development, at least in the early stages we're talking about, is a choreographed response to internal and external cues (gradients of X, Y, and Z in various directions, etc.)

And that choreography is, for the most part, written in genes. That is to say, genetic.

Everything happens in a "rich, diverse environment," but that doesn't make it any less pre-determined (under normal conditions) by genetic control architectures, or whatever. Maybe what I'm saying is, I don't see how the environmental/genetic distinction does much good, in this context. One cell's environment is another (nearby) cell's genetics, right?

I hope this doesn't sound too snarky. I'm not trying to argue (sorry).


Posted by: arthegall | Link to this comment | 07-13-07 2:52 PM
horizontal rule
115

And that choreography is, for the most part, written in genes. That is to say, genetic.

Right, but what is being choreographed depends on the environment. So yeah, genes say "this axon should follow this gradient," but the intensity (and even existence) of the gradient is dependent on whatever nutrients being available from the mother. You are choreographing with your own individual cast, to invoke the analogy ban, and that cast is your environment.

I don't think we disagree?


Posted by: Sifu Tweety | Link to this comment | 07-13-07 2:58 PM
horizontal rule
116

I thought this article was interesting:

The qualification within a job class is also important. There are quite high correlations between the socioeconomic status of a job and the mean IQ of the jobholders. Truck drivers average slightly under 100, while high-paid professionals, such as doctors and lawyers, have averages of 125 or above. It is sometimes asserted that this is because general intelligence is needed to obtain the educational certification required to qualify for a job, but is less important to on-the-job performance. There is evidence for this. Military and civilian studies have found that IQ tests are better predictors of performance when people are in training programs than when they are on the job itself. After people are on the job, correlations are higher between IQ and tests of job knowledge than between IQ and on-the-job observations of performance. However, none of the correlations vanish. IQ does not predict all aspects of job performance. In an extensive study of enlisted personnel (Campbell, McHenry and Wise 1990), the Army found that it was useful to distinguish between what might be called ability aspects of performance, which includes such things as knowledge of one's job requirements and the ability to operate machinery required in the job, and motivational aspects, which include cooperating with colleagues, showing initiative and leadership. The ASVAB did a good job of predicting the ability aspects but had almost no relation to the motivational aspects. This is not surprising, but it does make any focus on a unitary index of job competence seem simplistic. In summary, it appears that IQ is an important factor in getting into a job or profession, but is less important (although not negligible) once you have learned to do the job. Further improvement is then achieved by acquiring experience, rather than improving upon an abstract knowledge of what the job requires.


Posted by: joeo | Link to this comment | 07-13-07 3:03 PM
horizontal rule
117

115: or, to turn it around, sometimes the environment follows from teh choreography. Like, for instance, when the gradients are being set up not by the mother, but by the other neighboring cells in the developing organism. Interacting genetic pathways, some activating, some repressing, plus diffusion.

But the upshot is: yeah, you're right, I don't think we're disagreeing. Thank you for writing 6.


Posted by: arthegall | Link to this comment | 07-13-07 3:20 PM
horizontal rule
118

116: I do think there's been somewhat of a shift over the past couple of decades toward jobs being more cognitively demanding. More creativity and information processing required, more rewards for such skills, etc. There's an econ literature on this.

Also, I checked out the rest of that sloth guy (or gal's) blog, and s/he's sort of brilliant. Did anyone see this followup post to the one LB linked, it addresses a lot of things discussed in this thread. Except, you know, more brilliantly:

http://bactra.org/weblog/495.html


Posted by: marcus | Link to this comment | 07-13-07 3:27 PM
horizontal rule
119

I do think there's been somewhat of a shift over the past couple of decades toward jobs being more cognitively demanding. More creativity and information processing required, more rewards for such skills, etc.

There's an interesting question about what 'cognitively demanding' means. Like, for example, my reading of Jared Diamond's take on hunter-gatherers -- that's a pretty cognitively demanding lifestyle, but it doesn't look much like an office job.

Or to keep it industrial, which auto mechanic has a more cognitively demanding job; Mr. Goodwrench 1955, listen-to-the-funny-noise and adjust something manually to make it go away, or Mr. Goodwrench 2005, read the diagnostic code off the readout and replace the part? Mr. Goodwrench 2005 has a job that looks more like an office job than Mr. Goodwrench 1955, but I think it's less cognitively demanding.

Does it take more or less brains to type an attractive letter on a computer or a typewriter?


Posted by: LizardBreath | Link to this comment | 07-13-07 3:34 PM
horizontal rule
120

119: Mr Goodwrench 2005 has a vastly less complicated job, and can get by with less training.

But this is true for another reason, all across the trades. I was taught how to solder by an old guy who had been a tin basher most of his life (a trade that is gone). Anyway, his father (think our grandfathers generation) had been a master carpenter. At that time being a carpenter meant a lengthy unpaid (or nearly so) apprenticeship where you would learn a little of everything, followed by a length of time as a journyman. To be a master though, essentially you learned how to do anything that was built with wood. You weren't a specialist in everything, but you knew how to build cabinets and furniture as well as houses, barrels, etc. Hell, you even knew how to build boats.

This is economically impossible now. Apprentices are paid a living wage, and have to work on things that make the company some money. Trades have been subdivided to the point you have things like `framers'. It keeps training costs down, and mobility down, but there are very few people who know anything much outside a very narrow job description.

This was happening in auto mechanics before the computers took over. You had cheap transmission shops that kept their costs down by mostly employing `transmission specialists' who really weren't mechanics. Same with brake shops.


Posted by: soubzriquet | Link to this comment | 07-13-07 3:42 PM
horizontal rule
121

Trades have been subdivided to the point you have things like `framers'.

In fact I know someone who makes his living as a framer.


Posted by: DS | Link to this comment | 07-13-07 3:46 PM
horizontal rule
122

I should say that I'm really asking about the auto mechanic and the secretary there -- I'm not sure. And I'd listen to an argument (or, really I suppose I'm alluding to such an argument, although I don't think I could make it) that jobs have gotten more likely to call for a narrow set of cognitive tasks; processing abstract symbols, specifically.

But tasks other than abstract symbol processing are still cognitive tasks, and I think they're less emphasized. Memory is much less useful than it ever was -- if I were a lawyer even thirty years ago, I'd put a lot more effort into remembering cases than I do now. With the vaguest of recollections of a case, I can find it with a computer, so I don't need to remember it. Spatial skills are also probably down -- Dr. Oops has talked about older surgeons complaining than young surgeons are stupid with their hands; a smart kid growing up now reads and plays with their computer, but does a lot less manipulating actual objects than they might have thirty years or more ago, and it shows when they need to do intelligent object manipulation, like in someone's guts.


Posted by: LizardBreath | Link to this comment | 07-13-07 3:46 PM
horizontal rule
123

121: Sure, lots of people do. My point was that many trades are getting chopped up into little pieces. These skills used to be part of a larger package of skills , i.e. a `carpenter'. And framers don't have the market weight that carpenters do, so it makes sense for developer and it is easier to get up an running on (but it doesn't pay as well and you have less flexibility).

Of course there are new skills too --- people didn't used to have air tools to build houses with, and mechanics didn't used to use computers. On the whole though, much more is lost, these trades are being subdivided skill wise. Pretty much nobody these days has the depth of knowledge that old master carpenter did --- it's just not economically feasible.


Posted by: | Link to this comment | 07-13-07 3:54 PM
horizontal rule
124

123 wuz me


Posted by: soubzriquet | Link to this comment | 07-13-07 3:55 PM
horizontal rule
125

Yeah. I'm wondering if the real pattern isn't something more like "Jobs other than those calling for abstract symbol manipulation have gotten stupider, so the competition for one of the symbol-manipulating jobs has gotten more intense." That's not exactly right, but something like that.


Posted by: LizardBreath | Link to this comment | 07-13-07 3:59 PM
horizontal rule
126

Yeah, using a single broad term like "cognitive" is wrong. But two points: first, we're not talking about jobs per se, but job quality (pay, etc.). The labor demand question is where the money goes. So if Mr. Goodwrench's job is deskilled in the way you describe, he also makes less money and the job is not as "good" a job. Some jobs do get completely transformed -- the fact that letters are a lot easier to type on computers mean that managers now do most of what traditional "secretaries" once did, while the new role of "administrative assistant" or whatever might demand more decisionmaking and a greater range of skills than the old secretary did.

Second, when people try to measure this stuff, they usually end up pointing to increases in skill demand in abstract quantitative /analytic reasoning, and managerial / interactive tasks. So manufacturing employees might not need spatial mechanical skills quite as much, but they might have to reprogram computer-controlled machine tools, which is a more abstract task. Which is more "skilled' is an open question, but they are different. I'd say the abstract quantitative skills are probably more correlated with formal schooling and IQ, and the managerial interpersonal skills are also probably more associated with schooling.


Posted by: marcus | Link to this comment | 07-13-07 4:00 PM
horizontal rule
127

pwned by LB.

Here's an article discussing the shift:

http://econ-www.mit.edu/faculty/download_pdf.php?id=179


Posted by: marcus | Link to this comment | 07-13-07 4:01 PM
horizontal rule
128

I'd say the abstract quantitative skills are probably more correlated with formal schooling and IQ,

This implies that IQ isn't a general measure of cognitive potential, but more tightly connected to a particular mental skillset? That accords with my prejudices.


Posted by: LizardBreath | Link to this comment | 07-13-07 4:02 PM
horizontal rule
129

119. Three generations ago, most people were farmers; those that were not farmers did not usually have the chance to travel professionally, or realistic prospects for improvement. Mcjobs suck, but good people now are not usually stuck in them for long, especially if they're willing to relocate to take better work. The place where this is happening most quickly is maybe China, where many people are leaving primitive farming for initally bad work in cities. Nobody migrates from antlike specialization to cognitively demanding general skills until they're rich. On the other hand, constant danger is extremely demanding; it's easy to forget how grateful we should be for peace and physical safety purchased with boring specialized labor. Maybe we get to compensate by moving from job to job, rarer in the past.


Posted by: lw | Link to this comment | 07-13-07 4:04 PM
horizontal rule
130

This implies that IQ isn't a general measure of cognitive potential, but more tightly connected to a particular mental skillset?

What's the distinction?


Posted by: SomeCallMeTim | Link to this comment | 07-13-07 4:08 PM
horizontal rule
131

That's one of the big IQ battles -- does it actually measure 'g', a real quality that represents a global measure of one's cognitive capacity, or is it a composite of tests of different cognitive skills, which while they are often correlated with each other still measure different things.


Posted by: LizardBreath | Link to this comment | 07-13-07 4:12 PM
horizontal rule
132

I'm deeply honored, but just a little disappointed that this thread didn't produce more cock jokes.


Posted by: Cosma | Link to this comment | 07-13-07 4:51 PM
horizontal rule
133

It's early yet.


Posted by: teofilo | Link to this comment | 07-13-07 4:54 PM
horizontal rule
134

On the one hand, cock size is genetically influenced. On the other, it tends to shift with environmental conditions.

Also, differences in cock size between the sexes are generally considered innate, immutable and genetically based. Except for strap-ons.


Posted by: marcus | Link to this comment | 07-13-07 5:36 PM
horizontal rule
135

Now my happiness is complete.


Posted by: Cosma | Link to this comment | 07-13-07 5:49 PM
horizontal rule
136

We at Unfogged always stand ready to please with our cocks...er, cock jokes.


Posted by: marcus | Link to this comment | 07-13-07 6:11 PM
horizontal rule
137

Our cock.. er... jokes stand at the ready.


Posted by: TJ | Link to this comment | 07-13-07 6:22 PM
horizontal rule
138

I expected to see a lot more cock jokes after Cosma's post. Is everyone suffering from cock-joke performance anxiety? I hear they have drugs for that now.


Posted by: Walt Someguy | Link to this comment | 07-13-07 7:32 PM
horizontal rule
139

LB: I was curious if I was being fair to Diamond on the matter of New Guinean genetic superiority, but after searching my house for GG&S, it finally dawned on me I loaned it to a friend.

Moreover, Amazon's book search, for some reason, doesn't include the Prologue. But I googled up this quote from some random web site:

Intelligent people are likelier than less intelligent ones to escape those causes of high mortality (murder, chronic tribal warfare, accidents, problems procuring food..) in traditional New Guinean societies. However, the differential mortality from epidemic diseases in traditional European societies had little to do with intelligence, and instead involved genetic resistance dependent on details of body chemistry. For example people with blood type B or O have a greater genetic resistance to smallpox than do people with blood group A. That is, natural selection promoting genes for intelligence has probably been far more ruthless in New Guinea than in more densely populated, politically complex societies, where natural selection for body chemistry was more potent. (Diamond, p21)

Am I a hypocrite if I say that, while I find this mode of argument racist and absurd, Diamond's particular deployment of this argument delights me?


Posted by: politicalfootball | Link to this comment | 07-13-07 7:42 PM
horizontal rule
140

Christ, pf, couldn't you have worked one teeny little cock joke into that?

When I read GG&S, I half-suspected that Diamond didn't really mean that argument, that it just functioned to knock off-center anyone who was inclined to assume a genetic explanation for the course of history.


Posted by: Walt Someguy | Link to this comment | 07-13-07 9:32 PM
horizontal rule
141

When I read GG&S, I half-suspected that Diamond didn't really mean that argument, that it just functioned to knock off-center anyone who was inclined to assume a genetic explanation for the course of history.

140: I read it the exact way that you did. I changed my opinion long after I read it, and the reason I went to so much trouble to track down that passage was that I realized I reinterpreted it on memory, and not on re-reading.

What prompted my reinterpretation was reading the Third Chimpanzee, as I mentioned before. Apparently he really does buy into this evo psych crap.

I guess he's inclined toward determinism. GG&S is so great because it makes a strong case for a particular kind of determinism - a kind of determinism that highlights the foolishness of evo psych determinism. Diamond makes it possible to suppose that, once we really understand evo psych, it may be the New Guineans who come out on top.

And yes, I'm not very facile with the cock jokes. Neither is my mom, so I assume it's genetic.


Posted by: politicalfootball | Link to this comment | 07-13-07 10:08 PM
horizontal rule
142

Genetics itself depends on some facility with cocks.


Posted by: ogged | Link to this comment | 07-13-07 10:30 PM
horizontal rule
143

And that's why Tori Amos is so perfect for you.


Posted by: Walt Someguy | Link to this comment | 07-13-07 10:33 PM
horizontal rule
144

Deutsch, Henthorn, Marvin, Xu
Absolute pitch, popularly known as "perfect pitch," is the ability to name or produce a musical note of particular pitch without benefit of a reference note. It is extremely rare in the U.S. and Europe, with an estimated prevalence of less than one in 10,000. This rarity has so far been unexplained. We here report the first large-scale study comparing the prevalence of absolute pitch in two normal populations by means of a direct test. Our findings suggest that the potential for acquiring absolute pitch may be universal at birth. It raises the possibility that parents may be able to encourage the development of absolute pitch in their children during the "critical period" when infants are learning the main features of their native language.


Posted by: bemused | Link to this comment | 07-14-07 11:22 AM
horizontal rule
145

...and then this. It is easy to conflate ethnic correlation with genetic causality.


Posted by: bemused | Link to this comment | 07-14-07 11:46 AM
horizontal rule
146

140, 141: Huh. I'm still reading that quote and remembering it in context as 'If we're going to talk about relative evolutionary pressures imposed by environment, New Guineans have more pressure on them to be intelligent.' That's not a claim that they are genetically more intelligent, just that if that's the type of argument you're going to rely on, you can't use it to explain why New Guineans don't rule the world.

But I suppose he could have been making the arguement straight, rather than to debunk other arguments of the same sort. In which case clever, but you still can't demonstrate differences between populations by theorizing about differential evolutionary pressures, unless you can reliably demonstrate that the genetic differences actually exist.


Posted by: LizardBreath | Link to this comment | 07-14-07 11:52 AM
horizontal rule
147

139

Once you accept that IQ has a different relative value in different environments it follows that population groups should be expected to have different mean IQs. So Diamond should stay away from this argument if he wants to claim that mean IQ is the same for all populations.


Posted by: James B. Shearer | Link to this comment | 07-14-07 11:54 AM
horizontal rule
148

146

"But I suppose he could have been making the arguement straight, rather than to debunk other arguments of the same sort. In which case clever, but you still can't demonstrate differences between populations by theorizing about differential evolutionary pressures, unless you can reliably demonstrate that the genetic differences actually exist."

If IQ has greater relative adaptive value for population A than population B evolutionary models predict this to produce over time genetic differences between A and B leading A to have higher mean IQ than B. Of course confirmation by actual measurement is nice but a failure would be unexpected. In the same way models predict adding CO2 to the atmosphere will cause increased temperatures. This may not be proof absent confirmation by measurement but it is not nothing either.


Posted by: James B. Shearer | Link to this comment | 07-14-07 12:06 PM
horizontal rule
149

Oh god, people, seriously. PLEASE????


Posted by: bitchphd | Link to this comment | 07-14-07 1:26 PM
horizontal rule
150

You'll only encourage them, B. Try to turn away.


Posted by: Jesus McQueen | Link to this comment | 07-14-07 1:46 PM
horizontal rule
151

Two identical twins could have completely different sized cocks depending on what each is looking at. Also, cock size cannot be inherited, since it's totally gross to even think about the size of your relatives' genitals. Cocks are totally socially created.


Posted by: marcus | Link to this comment | 07-14-07 8:48 PM
horizontal rule