‹ First  < 2 3 4 5 6 > 
 
   
 

Sam Harris wrong on race

 
Abel Dean
 
Avatar
 
 
Abel Dean
Total Posts:  427
Joined  03-11-2017
 
 
 
14 January 2019 06:42
 

I know that the equation is not Feldman’s and Lewontin’s, knowledge which I expect would be clear since I wrote that Jensen used the same equation two years earlier. Feldman and Lewontin don’t seem to entirely dismiss the equation, but they deny that it is applicable to the study of human groups. Their claim that it is tautological seems intended to follow from their point that an unknown variable exists within the equation. Otherwise, there is nothing remotely tautological about it.

 
burt
 
Avatar
 
 
burt
Total Posts:  15809
Joined  17-12-2006
 
 
 
14 January 2019 08:03
 
TheAnal_lyticPhilosopher - 14 January 2019 06:25 AM

burt, don’t do it!  The Force beckons you!!  Come to the Dark Side here as well!!!

Mr. Dean lacks the elementary skills for intelligent discussion, despite having polished its forms. 

Yet another example: the equation he cites as Lewontin and Feldman’s (at least he seems to think it’s one they endorse) for the relationship between within-group and between-group heritability is nothing more than Lush’s equation for when n is large, which they point out is “definitional tautology, not a causal relationship,” one that leads to the “spurious” conclusion that between-group heritability can be derived from within-group heritability—a criticism that has nothing to do with it being “impossible to estimate the intraclass genetic correlation” in human populations, therefore the equation cannot be used (they don’t even make that criticism).  The context and criticism is right there in the text.  So is their conclusion, after criticizing Lush:

: ...“we are unable to make any inferences from between-group differences and within-group statistics about the degree of genetic determination of the between-group differences. In other words, the concept of heritability is of no value for the study of differences in measures of human behavioral characters between groups.” (emphasis added).

If he can’t even get these elementary things right, what’s the point?

Just the tip of the iceberg of a literate ignorance defending immoral views….

Done with him. Definition of a troll. Couldn’t resist quoting Shakespeare in an apt way. I guess he also thinks women are only about 90% as intelligent as men.

 
Jan_CAN
 
Avatar
 
 
Jan_CAN
Total Posts:  3301
Joined  21-10-2016
 
 
 
14 January 2019 08:39
 
TheAnal_lyticPhilosopher - 14 January 2019 06:25 AM

...
Mr. Dean lacks the elementary skills for intelligent discussion, despite having polished its forms. 

Yet another example: the equation he cites as Lewontin and Feldman’s (at least he seems to think it’s one they endorse) for the relationship between within-group and between-group heritability is nothing more than Lush’s equation for when n is large, which they point out is “definitional tautology, not a causal relationship,” one that leads to the “spurious” conclusion that between-group heritability can be derived from within-group heritability—a criticism that has nothing to do with it being “impossible to estimate the intraclass genetic correlation” in human populations, therefore the equation cannot be used (they don’t even make that criticism).  The context and criticism is right there in the text.  So is their conclusion, after criticizing Lush:

: ...“we are unable to make any inferences from between-group differences and within-group statistics about the degree of genetic determination of the between-group differences. In other words, the concept of heritability is of no value for the study of differences in measures of human behavioral characters between groups.” (emphasis added).

If he can’t even get these elementary things right, what’s the point?

Just the tip of the iceberg of a literate ignorance defending immoral views….

[bolding mine]
Yes, Mr. Dean seems to epitomize the old adage that ‘a little knowledge is a dangerous thing’, i.e. a small amount of knowledge can mislead people into thinking that they are more expert than they really are.  (Without a certain amount of humility and an open mind, that is.)

[ Edited: 14 January 2019 08:48 by Jan_CAN]
 
 
Abel Dean
 
Avatar
 
 
Abel Dean
Total Posts:  427
Joined  03-11-2017
 
 
 
14 January 2019 10:01
 

Do you believe I have lacked humility?

 
Jan_CAN
 
Avatar
 
 
Jan_CAN
Total Posts:  3301
Joined  21-10-2016
 
 
 
14 January 2019 10:33
 
Abel Dean - 14 January 2019 10:01 AM

Do you believe I have lacked humility?

I think it is very important when one is espousing unpopular/controversial views, especially those with moral implications, to self-examine for biases.  To ask oneself carefully and honestly where these come from and if they are justified.  And not to ignore one’s conscience either.  One can argue any point of view by citing statistics and studies, but integrity also requires us to look at the big picture and the implications of what we say.

 

 
 
TheAnal_lyticPhilosopher
 
Avatar
 
 
TheAnal_lyticPhilosopher
Total Posts:  865
Joined  13-02-2017
 
 
 
14 January 2019 10:35
 
Abel Dean - 14 January 2019 10:01 AM

Do you believe I have lacked humility?

In all fairness to you, no.

P.S. My bad on thinking you thought the equation was FL’s.

[ Edited: 14 January 2019 10:42 by TheAnal_lyticPhilosopher]
 
TheAnal_lyticPhilosopher
 
Avatar
 
 
TheAnal_lyticPhilosopher
Total Posts:  865
Joined  13-02-2017
 
 
 
14 January 2019 10:37
 
burt - 14 January 2019 08:03 AM
TheAnal_lyticPhilosopher - 14 January 2019 06:25 AM

burt, don’t do it!  The Force beckons you!!  Come to the Dark Side here as well!!!

Mr. Dean lacks the elementary skills for intelligent discussion, despite having polished its forms. 

Yet another example: the equation he cites as Lewontin and Feldman’s (at least he seems to think it’s one they endorse) for the relationship between within-group and between-group heritability is nothing more than Lush’s equation for when n is large, which they point out is “definitional tautology, not a causal relationship,” one that leads to the “spurious” conclusion that between-group heritability can be derived from within-group heritability—a criticism that has nothing to do with it being “impossible to estimate the intraclass genetic correlation” in human populations, therefore the equation cannot be used (they don’t even make that criticism).  The context and criticism is right there in the text.  So is their conclusion, after criticizing Lush:

: ...“we are unable to make any inferences from between-group differences and within-group statistics about the degree of genetic determination of the between-group differences. In other words, the concept of heritability is of no value for the study of differences in measures of human behavioral characters between groups.” (emphasis added).

If he can’t even get these elementary things right, what’s the point?

Just the tip of the iceberg of a literate ignorance defending immoral views….

Done with him. Definition of a troll. Couldn’t resist quoting Shakespeare in an apt way. I guess he also thinks women are only about 90% as intelligent as men.

Good man!  I mean, how many times does it need pointing out that between-group heritability calculations (h) are ultimately tautological and circular; that they require the coefficient (r), which depends for its value on the values computed from (h), which depends on the presumed coefficient (r).  What else is to be expected from defining r as the relation of the genetic variance within groups to the genetic variance between groups, then using r to compute an estimate of the genetic variance between groups from the genetic variance within groups?

Then Jensen comes along and says “Hey everybody, the equation works just fine if you don’t divide by 0!” 

Yeesh.

 

[ Edited: 14 January 2019 11:15 by TheAnal_lyticPhilosopher]
 
burt
 
Avatar
 
 
burt
Total Posts:  15809
Joined  17-12-2006
 
 
 
14 January 2019 14:11
 
TheAnal_lyticPhilosopher - 14 January 2019 10:37 AM
burt - 14 January 2019 08:03 AM
TheAnal_lyticPhilosopher - 14 January 2019 06:25 AM

burt, don’t do it!  The Force beckons you!!  Come to the Dark Side here as well!!!

Mr. Dean lacks the elementary skills for intelligent discussion, despite having polished its forms. 

Yet another example: the equation he cites as Lewontin and Feldman’s (at least he seems to think it’s one they endorse) for the relationship between within-group and between-group heritability is nothing more than Lush’s equation for when n is large, which they point out is “definitional tautology, not a causal relationship,” one that leads to the “spurious” conclusion that between-group heritability can be derived from within-group heritability—a criticism that has nothing to do with it being “impossible to estimate the intraclass genetic correlation” in human populations, therefore the equation cannot be used (they don’t even make that criticism).  The context and criticism is right there in the text.  So is their conclusion, after criticizing Lush:

: ...“we are unable to make any inferences from between-group differences and within-group statistics about the degree of genetic determination of the between-group differences. In other words, the concept of heritability is of no value for the study of differences in measures of human behavioral characters between groups.” (emphasis added).

If he can’t even get these elementary things right, what’s the point?

Just the tip of the iceberg of a literate ignorance defending immoral views….

Done with him. Definition of a troll. Couldn’t resist quoting Shakespeare in an apt way. I guess he also thinks women are only about 90% as intelligent as men.

Good man!  I mean, how many times does it need pointing out that between-group heritability calculations (h) are ultimately tautological and circular; that they require the coefficient (r), which depends for its value on the values computed from (h), which depends on the presumed coefficient (r).  What else is to be expected from defining r as the relation of the genetic variance within groups to the genetic variance between groups, then using r to compute an estimate of the genetic variance between groups from the genetic variance within groups?

Then Jensen comes along and says “Hey everybody, the equation works just fine if you don’t divide by 0!” 

Yeesh.


The list of troll traits.
http://www.anthropology-news.org/index.php/2018/12/20/government-of-by-and-for-the-trolls/?fbclid=IwAR0bK7tcpSM_aHtVAZMdjiSF_kpRqeF4gFpJkhMvyfSjFoDnYJGp8RgLq2c

 
LadyJane
 
Avatar
 
 
LadyJane
Total Posts:  3292
Joined  26-03-2013
 
 
 
14 January 2019 14:32
 

Are there any malfunctioning ignore buttons to report?

 
 
TheAnal_lyticPhilosopher
 
Avatar
 
 
TheAnal_lyticPhilosopher
Total Posts:  865
Joined  13-02-2017
 
 
 
14 January 2019 15:01
 
burt - 14 January 2019 02:11 PM
TheAnal_lyticPhilosopher - 14 January 2019 10:37 AM
burt - 14 January 2019 08:03 AM
TheAnal_lyticPhilosopher - 14 January 2019 06:25 AM

burt, don’t do it!  The Force beckons you!!  Come to the Dark Side here as well!!!

Mr. Dean lacks the elementary skills for intelligent discussion, despite having polished its forms. 

Yet another example: the equation he cites as Lewontin and Feldman’s (at least he seems to think it’s one they endorse) for the relationship between within-group and between-group heritability is nothing more than Lush’s equation for when n is large, which they point out is “definitional tautology, not a causal relationship,” one that leads to the “spurious” conclusion that between-group heritability can be derived from within-group heritability—a criticism that has nothing to do with it being “impossible to estimate the intraclass genetic correlation” in human populations, therefore the equation cannot be used (they don’t even make that criticism).  The context and criticism is right there in the text.  So is their conclusion, after criticizing Lush:

: ...“we are unable to make any inferences from between-group differences and within-group statistics about the degree of genetic determination of the between-group differences. In other words, the concept of heritability is of no value for the study of differences in measures of human behavioral characters between groups.” (emphasis added).

If he can’t even get these elementary things right, what’s the point?

Just the tip of the iceberg of a literate ignorance defending immoral views….

Done with him. Definition of a troll. Couldn’t resist quoting Shakespeare in an apt way. I guess he also thinks women are only about 90% as intelligent as men.

Good man!  I mean, how many times does it need pointing out that between-group heritability calculations (h) are ultimately tautological and circular; that they require the coefficient (r), which depends for its value on the values computed from (h), which depends on the presumed coefficient (r).  What else is to be expected from defining r as the relation of the genetic variance within groups to the genetic variance between groups, then using r to compute an estimate of the genetic variance between groups from the genetic variance within groups?

Then Jensen comes along and says “Hey everybody, the equation works just fine if you don’t divide by 0!” 

Yeesh.


The list of troll traits.
http://www.anthropology-news.org/index.php/2018/12/20/government-of-by-and-for-the-trolls/?fbclid=IwAR0bK7tcpSM_aHtVAZMdjiSF_kpRqeF4gFpJkhMvyfSjFoDnYJGp8RgLq2c

Nice article.  I especially like: “…when trolls make discursive moves that obfuscate, confuse, deny and distort, those engaging with the troll are often at a loss on how to respond because the troll’s conversational contributions are so far removed from the norms.”

If by “norms” he means ‘fidelity to the basic arguments and ideas,’ then right on.  I also like where he says: “we can respond by not falling for the distractions” and “consistently working to keep the discourse focused on the key issues at hand.” 

But I say, what’s the point when the key issue at hand is the same old basic mistake made all over again?  Best just to hit it once, right where it matters, then let it go.  You can’t get bogged down on a single outlying world when there’s a whole galaxy to rule…

[ Edited: 14 January 2019 15:08 by TheAnal_lyticPhilosopher]
 
burt
 
Avatar
 
 
burt
Total Posts:  15809
Joined  17-12-2006
 
 
 
14 January 2019 15:23
 
TheAnal_lyticPhilosopher - 14 January 2019 03:01 PM

You can’t get bogged down on a single outlying world when there’s a whole galaxy to rule…

Sometimes an outlying world is important. Recommend Doris Lessing’s book The Sentimental Agents in the Volyan Empire, sharpest satire I’ve ever read.

 
TheAnal_lyticPhilosopher
 
Avatar
 
 
TheAnal_lyticPhilosopher
Total Posts:  865
Joined  13-02-2017
 
 
 
14 January 2019 15:46
 
burt - 14 January 2019 03:23 PM
TheAnal_lyticPhilosopher - 14 January 2019 03:01 PM

You can’t get bogged down on a single outlying world when there’s a whole galaxy to rule…

Sometimes an outlying world is important.

It’s true.  It’s so hard to tell the difference sometimes.

I’m enjoying the Omnes book you recommended, so I’ll check the Lessing out. 

 

 

 
lynmc
 
Avatar
 
 
lynmc
Total Posts:  462
Joined  03-08-2014
 
 
 
14 January 2019 20:03
 
Abel Dean - 10 January 2019 09:36 PM

“I get that he’s coming across as judicious and critically careful in his assertions, but they are, as you’ve intuited, bullshit.  A trait can have a heritability of .1 or .9, and either estimate has exactly the same relationship to the probability that a phenotypic difference between groups is based on genetic markers because both scores have the same impact on that probability, whatever it might be—zero.  Period.”

I have paid much attention to the academic arguments against Arthur Jensen’s positions. A common argument against Jensen’s claim that within-group heritability reflects between-group heritability is the farmer argument. Maybe you have seen it already. Two fields are planted sourcing from the same randomly-mixed bag of seeds. You water one field and you leave the other field mostly dry. The stalks in the watered field grows tall, whereas the corn in the dry field grows short. The trait of stalk height may have a very high heritability value, and yet you can have two drastically heights between the two groups.

That seems to be the extent of the argument. It is a purely rhetorical argument from analogy, no math involved. But, Arthur Jensen responded with a mathematical argument, which he laid out fully in one of his books, and I can relay it this weekend when I return home if you like. His claim was that, while within-group heritability does not equal between-group heritability, there is a proved positive relationship. Increasing either the within-group heritability or the phenotypic group gap increases the probability of a greater between-group heritability. I expect that his argument would also make intuitive common sense. A central reason we take it for granted that racial bodily height differences are probably largely genetic is because we all know that height has such a strong within-group heritability. I have seen only one of Jensen’s critics confront the argument, and he admitted that the argument was in Jensen’s favor (James Flynn’s 1980 book, “Race, IQ and Jensen”).


Heritability is a proportion of variance.  Except its use in the calculation (in which average is subtracted from every measure so in a sense “drops out”), average (or mean) has no effect on the variance.  And variance has no bearing on average.  That’s just the mathematics.  So I am rather curious how any proportion of variance says anything meaningful about the mean, or the means of any measures of two subpopulations?

 
Abel Dean
 
Avatar
 
 
Abel Dean
Total Posts:  427
Joined  03-11-2017
 
 
 
14 January 2019 20:39
 
lynmc - 14 January 2019 08:03 PM
Abel Dean - 10 January 2019 09:36 PM

“I get that he’s coming across as judicious and critically careful in his assertions, but they are, as you’ve intuited, bullshit.  A trait can have a heritability of .1 or .9, and either estimate has exactly the same relationship to the probability that a phenotypic difference between groups is based on genetic markers because both scores have the same impact on that probability, whatever it might be—zero.  Period.”

I have paid much attention to the academic arguments against Arthur Jensen’s positions. A common argument against Jensen’s claim that within-group heritability reflects between-group heritability is the farmer argument. Maybe you have seen it already. Two fields are planted sourcing from the same randomly-mixed bag of seeds. You water one field and you leave the other field mostly dry. The stalks in the watered field grows tall, whereas the corn in the dry field grows short. The trait of stalk height may have a very high heritability value, and yet you can have two drastically heights between the two groups.

That seems to be the extent of the argument. It is a purely rhetorical argument from analogy, no math involved. But, Arthur Jensen responded with a mathematical argument, which he laid out fully in one of his books, and I can relay it this weekend when I return home if you like. His claim was that, while within-group heritability does not equal between-group heritability, there is a proved positive relationship. Increasing either the within-group heritability or the phenotypic group gap increases the probability of a greater between-group heritability. I expect that his argument would also make intuitive common sense. A central reason we take it for granted that racial bodily height differences are probably largely genetic is because we all know that height has such a strong within-group heritability. I have seen only one of Jensen’s critics confront the argument, and he admitted that the argument was in Jensen’s favor (James Flynn’s 1980 book, “Race, IQ and Jensen”).


Heritability is a proportion of variance.  Except its use in the calculation (in which average is subtracted from every measure so in a sense “drops out”), average (or mean) has no effect on the variance.  And variance has no bearing on average.  That’s just the mathematics.  So I am rather curious how any proportion of variance says anything meaningful about the mean, or the means of any measures of two subpopulations?

In this context, it is not about trying to find the means. The phenotypic means are the average IQs of blacks (85) and whites (100). It is not so well known among the public, because even those foundational facts are racist, but it is established and well-known among academics who study the issue. The disagreement is about what explains the difference between the two groups (15 points). Does it follow more from genetic differences between the two groups? Or does it follow more from environmental differences between the two groups? Arthur Jensen claimed it is probably largely a matter of genetic differences, whereas Feldman and Lewontin seemed to favor the environmentalist hypothesis. The portion of the group gap that follows from genetic differences is the value of between-group heritability. Arthur Jensen claimed that increasing either the within-group heritability or the phenotypic group gap increases the probability of a greater between-group heritability. For black and white IQ, the within-group heritability is 0.74 on a scale of 0 to 1, and the phenotypic group gap is a standard deviation. Both values are large, so a between-group heritability significantly greater than 0 should be a default hypothesis, though 0 is still possible. It would be plainly wrong to claim that within-group heritability has nothing to do with between-group heritability. I wonder if Feldman and Lewontin would claim that a within-group heritability of 0.9999 has no effect on the probability of a higher between-group heritability.

 
burt
 
Avatar
 
 
burt
Total Posts:  15809
Joined  17-12-2006
 
 
 
14 January 2019 21:53
 

For those interested in digging into details and references: https://en.wikipedia.org/wiki/Heritability_of_IQ

 
‹ First  < 2 3 4 5 6 >