ps418

Echidna,

Having done a little bit more reading on the subject, I would like to point out that Bouchard et al's twin studies are just one of many which demonstrate the high heritability of IQ. I was unable to find any articles in the professional journals seriously questioning the high heritability of IQ, but found many providing evidence for the high heritability of IQ. The debate now seems to be largely focused on whether the heritability of IQ is closer to 0.6 or 0.8. I am only bringing this up again because you were pretty harsh on the Bell Curve authors for accepting a large genetic component of IQ, a position you called "laughably lame and self-contradictory." Actually, it seems to me that they have the facts pretty firmly on their side.

For instance, a recent editorial on the genetics of brain function and cognition in the journal <a href="http://www.kluweronline.com/issn/0001-8244" target="_blank">Behavior Genetics</a> states that "[t]here is overwhelming evidence for the existence of substantial genetic influences on individual differences in general and specific cognitive abilities, especially in adults" (de Geus et al., 2001. Behavior Genetics 31(6), p. 489).

A recent review (1998) of twin studies of the genetics of human intelligence by Bouchard shows a convergence of heritability estimates of 0.6-0.8 (Genetic and environmental influences on adult intelligence and special mental abilities.
Human Biology 70(2), pp. 257-79).

Wright et al (2001) report intitial results of large sample twin study of the genetics of cognition, using twins from several countries. They also report a correlation between reaction time and working memory and IQ.

Wright et al., 2001. Genetics of cognition: outline of a collaborative twin study. Twin Research 2001 4(1), pp. 48-56.

Patrick

Amen-Moses

What does "close proximity" mean?

How about the identical twins that were raised in different countries or even on different continents?

Two pairs always stick out to me;

Two women separated at birth who were re-united in their 50s; one was raised in Wales the other in England over 100 miles away. When they met for the first time they had the same build, same hairstyles, same clothes, similar looking husbands with the same profession, the same jobs with the same company, same religion, same politics, heck they even had similar wallpaper in their living rooms!

Two men separated as young children, one raised in Jamaica by his Jewish mother who married a West Indian and the other raised in Nazi Germany by his Catholic father. When they met, again in their fifties, they both smoked the same brand of cigarettes drove the same colour cars had almost identical tastes in clothing, were both right wing politically and again had similar taste in women (although one was married and the other divorced). Again they had similar professions, the same build, both drank beer and both had the same stubborn streak allied with a bad temper. They hated each other on sight and after the study never met up again!

And yes the all tested the same as their twin on IQ tests.

Amen-Moses

ps418

Amen-Moses,

There is also the 'flip-side' to monozygotic twins reared apart as well -- adopted siblings raised together. Siblings adopted at birth and reared in the same environment are on average no more similar in IQ than total strangers, whereas monozygotic twins reared apart are on average as similar as a single individual tested on two different occasions.

Patrick

[ October 20, 2002: Message edited by: ps418 ]</p>

echidna

A-M, I think I’d prefer the more statistical scientific approach rather than anecdotal. The trouble with anecdotal evidence is that it becomes very susceptible to selective subjectivity. Phrases such as “similar politics”, “similar hairstyles”, “same clothes” are suspiciously open to interpretation, much akin to someone seeking to show that all Geminis are the same.

ps418, your post is valid & I’m still contemplating a reply, but I’m sure you’re in no hurry.

Amen-Moses

All documentaries on TV (where I got the above from) are anecdotal as are studies done by other people, the only way you can avoid it is to do your own study.

Amen-Moses

ohwilleke

Still, I think there is a place for anecdotal evidence. Assuming the the examples given by A-M are true (admittedly not corroborated), the point of the example is that twins have more in common than random chance would indicate on so many different matters not commonly understood to be genetic, that the hypothesis that genetics play an extremely strong role in adult behavoir is stronger than standard social scientific means are designed to measure.

In the same way, the famous study in inheritance of hemophelia as a recessive gene (in the royal Romanov family, IIRC), was just an anecdote when first observed (prior to genetic testing and the like), but because it could be so well explained by a theoretical model, was worth pursuing and being confirmed. Other good examples of this are the anecdotal experience of high numbers of deaf people at Martha's Vinyard, and high numbers of color blind people in a Pacific Island, that in each case revealed a genetic bais for a particular disability. Another good recent example is the apparent link between being red headed and requiring large doseages of anethestics, something anecdotally reported for years in the anethesiology trade, before it was studied and confirm within the past few months.

Anecdote, called ethnography by anthropologists, also has the advantage over statistical summary in the social sciences of offering a deeper analysis than a survey can. A survey or poll or paper and pencil test can only ask the questions posed by a researcher before getting into the field and encountering particular sets of subjects. In contrast, an ethnographer can go out and observe relationships that a question formulator might never have thought to ask.

An anecdote, for example, might tell you that not only did two people have simliar IQ scores, but that they tended to get the same questions right and wrong, something a mere IQ score could never tell you.

[ October 21, 2002: Message edited by: ohwilleke ]</p>

echidna

ps418, on Bouchard I find his overt politicisation as concerning as Murray’s. Now not reason in itself and I can hear you saying I should play the ball not the man, but I would prefer to hear from more objective sources. Note the criticism in SciAm that Bouchard would not release his data for third party scrutiny.

Breaking the Last Taboo
<a href="http://www.lrainc.com/swtaboo/taboos/tbc01.html" target="_blank">http://www.lrainc.com/swtaboo/taboos/tbc01.html</a>

If you read this piece by Bouchard himself you may notice how politicised his commentary is, as unfortunately the entire debate seems to be. I find ample circumstantial evidence to suspect that his conclusion were derived before his research was made.

It would seem that many (if not most) and I include myself, presuppose the nature / nurture element of human psychology & then find data to suit. Similarly I note that Murray holds his doctorate in Political Science rather than the field he released his research for.

If I anticipate your thoughts correctly, yes, guilt on both sides.

For the counter pre-supposition to Bouchard’s, here’s another summary, quite likely equally preconceived with even more extreme estimates in the opposite direction.

<a href="http://www.indiana.edu/~edpsych/p540/assign/intel6.html" target="_blank">http://www.indiana.edu/~edpsych/p540/assign/intel6.html</a>

On the 53 signatories to the petition, 15 are recipients of Pioneer Fund grants & as such I would suggest that they represent the genetics side of the debate rather than being genuinely bipartisan.

echidna

ps418, I readily concede that I am wiser on the debate than I was 2 weeks ago, needless to say the goal of testing my convictions in this thread. But I find that part of the problem is the polarisation which seems to exist between 2 competing fields of science which also represent 2 competing political directions. Political libertarians and geneticists lean to the higher estimates of hereditability while leftwingers and behaviouralists lean to the lower. Of course gone are the days when behaviouralists can claim that nurture accounts for all, so the debate seems to be languishing on where that break can be, somewhere between 0.4 to 0.6 seems common but impartiality seems hard to find.

While “laughably lame and self-contradictory” is somewhat hyperbolic , it still seems to me that HM’s assessment is at the top end of the estimates, although I also concede that some estimates range from 0.4 to 0.8 . Certainly Bouchard’s 0.7 is high in the scale.

I still find that the Flynn Effect dramatically contradicts the BC’s primary model. To date I cannot find much explanation for this inconsistency from Bouchard, Murray, et al.

Most likely to me, is that HM may be correct in their finding of a significant element of heritability in intelligence, but at the same time incorrect that they underestimate the malleability of intelligence as we measure it today. HW’s point about being able to train for IQ tests is also very valid & as such I’ve had the same experience.

I think that given the history of this debate I would prefer to remain somewhat agnostic on the issue. Given the similar confidence with which early social scientists made their conclusions which today are considered clearly racist and wrong, I cannot but retain the possibility that another 50 years may regard this as the same.

Inadequate ? Maybe, but scientific agnosticism remains a cornerstone & it seems difficult to make any claims with too much confidence at this point.

echidna

Ohwilleke, as you describe I see anecdote as the precursor to scientific knowledge. It offers the hint of something which needs to be rigorously and objectively studied. In this instance we have the ability to perform both case studies (anecdote) and statistical research (scientific study). Give both I trust the latter more.

echidna

There are some passing thoughts which really shouldn’t be posted, yet nonetheless, for an abstract (and likely irrelevant) tangent, I see that the raw computational power required to pass many IQ tests is relatively simple, patterns, colours, numbers, vocabulary. In fact very simple algorithms very rapidly become impossible for the human brain to calculate. In contrast, to recognise the human emotion of a simple facial expression is a feat which even our best supercomputers are incapable of mastering.

I find it highly counter-intuitive that a device which can readily master such an complex task, in turn should be functionally judged on such simple operation which would be comparable to asking a supercomputer for its 8 times tables.