Race and Intelligence: The Diversity Argument

February 9, 2012

In a recent exchange, I actually ran into an argument on the race and intelligence question which I had never heard before. I call it the “diversity argument,” and it basically goes like this:

1. Blacks are more genetically diverse than other racial groups; therefore

2. One would expect that there exist black subgroups at the extreme end of any human attribute which (a) one can measure; (b) is primarily genetically driven; and (c) is not informed by blacks’ ancestral environment.

3. One does not observe any black subgroup with extremely high intelligence; therefore

4. The weak intelligence of blacks must be primarily the result of something besides genetics.

This argument definitely deserves points for creativity but I think it has two, perhaps 3 fatal flaws.

First, and most importantly, it does not appear to be true in practice. For example, there does not appear to be any evidence of any black subgroup which is extremely good at weightlifting. If you look at the list of olympic weightlifting champions over the years; or winners of the world’s strongest man competition, judging by their names and countries of origin, there appear to be very few, perhaps no blacks at all.

Second, it’s entirely possible that blacks’ weak intelligence is informed by their ancestral environment. Presumably being smarter requires that some parts of your brain be larger or more active, which would necessarily require more heat production. Surely heat stroke is a bigger risk in parts of the world where it helps to have dark skin.

Last, the argument assume a certain equilibrium is in effect, i.e. that whatever genes make you smart have had a reasonable opportunity to come into the world’s black population and become part of the black genetic diversity. By why should one assume that? Presumably “smart genes” originated in some subset of the world and if that place was not in Africa it would take time for them to find their way into the black population.


Is Lesswrong More Wrong? (Updated)

November 24, 2011

I am a frequent poster on the website known as “lesswrong.”  It purports to be a site for discussing human rationality.  There are a lot of interesting posts there, particularly by a man named Eliezer Yudkowsky.

After a recent discussion regarding the possible guilt or innocence of Amanda Knox, I have come to the belief that such a site actually makes some people less rational.   How might this be?  Well for one thing, learning about techniques of rationality might make it easier for you to construct rationalizations for your irrational beliefs.   For another, a frequent poster to the site might start to take a lot of pride in his self-perceived rationality, so it might be more difficult psychologically difficult to confront evidence that his thinking is irrational.

Does this actually happen?  It seems to me that the answer is “yes.”  I could give a lot of examples, but here is one I found striking.  The following is from a regular poster known as “komponisto” who has been arguing quite a bit for the innocence of Amanda Knox:

More generally, you are of the opinion that circumstantial and psychological evidence of the sort produced in this case is powerful enough to overcome not only the incredibly low prior probability of guilt for Knox and Sollecito, but also the total lack of any (significant, non-discredited) physical evidence tying them to the crime, the failure of Guede to implicate them until he knew they were suspects, and the fact that Kercher had both a full stomach and an empty duodenum at death, and thus almost certainly died before 9:30 pm, while Sollecito’s computer was in use at his apartment.

Needless to say, I think this is sheer madness. In my view, you are vastly, hugely overconfident in your model of human behavior. However, in the unlikely event that you can actually produce a compelling argument for why I should (for example) regard the presence of glass on top of clothes as 30-decibel evidence in favor of the hypothesis that Knox and Sollecito staged the burglary over the hypothesis that there was an authentic burglary committed by known burglar Rudy Guede, I am all ears.

Really, however, I’m long past the point where I’m ready to write you off as an incorrigible clacker.

It would appear that komponisto believes that a very important piece of evidence in in this matter is the state of the murder victim’s digestive system combined with the use of Sollecito’s computer.   He thinks that my assessment of Knox’s probable guilt is “sheer madness.”  He has pretty much concluded that I am the type of person who resists obviously correct conclusions.

Here was my response:

Where are you getting this from?

According to my research

(1) Sollecito’s computer stopped being used at 9:10pm

(2) Sollecito’s own expert witness — Professor Introna — testified that based on the stomach and duodenum contents, the time of death was between 9:30 pm and 10:30pm.

(3) Professor Bacci, the prosecution’s expert, testified (based on the same digestive issues) that the time of death was between 9:00 to 9:30 pm and 11:00pm to midnight.

Do you disagree with any of this? Because it looks to me like you are suffering from a massive case of confirmation bias. To be sure, I got items (2) and (3) from the sentencing report. But I have a really hard time believing that the report would flat out lie about peoples’ testimony.

Komponisto did not reply at all to my questions.  Now, I’m reasonably confident that what I said in the post is true, i.e. that according to the evidence Sollecito’s computer was last used at 9:10; that Sollecito’s witness testified that the murder happened before 10:30; and so on.  But perhaps more importantly, there is a reasonable basis to believe these things and yet komponisto was apparently not interested in such evidence at all.  Instead, his preference would appear to be to accept pro-Knox statements with little or no scrutiny.  In short, as alluded to in my post, he appears to be suffering from a massive case of confirmation bias.

How could it be that a prominent poster on a message board devoted to rationality could behave so irrationally?  I’m just speculating, but I would guess that lesswrong has made him more wrong.   By arrogantly announcing, in effect, that the Amanda Knox case is a “test” of rationality; by devoting many posts and use of the “tools of rationality” in arguing for her innocence; by investing himself so thoroughly on this issue, komponisto has, in effect mind-killed himself.

P.S.  Of course I am open to arguments or evidence on the duodenum / computer use issue.  I don’t like to eat humble pie, but I am happy to consider evidence on this point in good faith.


Update:  What’s also interesting is how people responding to this argument have a strong tendency to completely miss the point.  For example, lesswrong poster TimS said the following:

Your marshaling of facts about the time of death vs. Sollecito’s computer usage is excellent, in that it provides a real basis for discussion. But is it your true rejection? You pointed out it was a side issue, and even if the prosecution expert was entirely correct, it would only show the evidence was consistent with guilt and would not be [ETA: very strong] evidence of guilt.

And in case anyone things my point was buried or hidden, here’s what I had said:

I was a bit surprised to see such a convincing-sounding argument which I had not heard before, so I decided to look into it. What I found is this:

  1. According to various sources, Sollecito’s computer use ended at 9:10 not 9:30.
  2. According to the sentencing report, Sollecito’s expert testified (based on the state of Kercher’s digestive system) that the time of death was between 9:30 and 10:30.
  3. According to the sentencing report, the prosecution expert testified (based on the state of Kercher’s digestive system) that the time of death was as late as midnight.

It seems unlikely to me that the sentencing report would flat out lie about the expert testimony. It also seems unlikely that Sollecito’s expert would testify that the time of death could have been as late as 10:30 if there was solid evidence that the time of death was almost certainly before 9:30pm.

My conclusion is that komponisto is likely the victim of serious confirmation bias.

(emphasis added).

Similarly, in response to this blog post, commenter “Katie” stated the following:

You should have linked to the discussion on the site – I had to google it to find it. Having read the whole thing – I’m afraid you are the one being irrational.Looking at other evidence, it becomes more and more obvious that she’s innocent, since none of the evidence lines up with the prosecution theory.

Of course, the point of this post is not that Knox is guilty or innocent; the point is that komponisto (and by extension, possibly others) are seriously biased.  So why the confusion?  I would speculate that there is another bias at work, what Eliezer might call a blue/green bias.  This post is perceived as an attack on komponisto and therefore an attack on the pro-Knox camp.  It is perceived as a soldier in the anti-Knox army and therefore must be met on the battlefield regardless of its rightness or wrongness.

Indeed, one can ask where all of these pro-Knox posters came from who have been posting on this blog.  I certainly do not advertise or promote this blog anywhere.  I’ve linked to it only from a couple obscure message boards.  My guess is that posters like “Katie” are knights-errant in the Amanda Knox controversy.  i.e. they regularly do Google searches to find blog posts, news articles and the like in order to do battle on Knox’s behalf.

Is Amanda Knox Guilty?

November 17, 2011

Probably yes.  Before I explain why I think so, it’s worthwhile to explain what I mean when I assert that she is probably guilty.  What I mean is that she was probably involved in the murder of Meredith Kercher, i.e. she either participated directly or was an accomplice.  It seems that the prosecution argued that she physically stabbed Kercher in the course of the murder.  This is a more narrow claim and it is necessarily less likely to be correct than the claim that Knox was involved.

Also, before I explain why I think Knox is guilty, I think it’s worthwhile to explain what I think happened, i.e. the basic outline of the crime.  I suspect part of the reason people have such a hard time believing Knox is guilty is that it’s difficult to imagine a middle class white girl taking part in the rape and murder of another middle class white girl.  In short, I think it’s worthwhile to show how it’s remotely plausible (even if very unlikely) that Knox took part in the murder.

I think what probably happened is similar to the murder of Janet Chandler in Michigan back in the 70s.  Chandler’s roommate (a white girl) took part in (and perhaps orchestrated) a conspiracy to gang-rape Chandler.  Apparently the roommate disliked Chandler who “thought a lot of herself” and wanted to “teach her a lesson.”  In the course of the rape, Chandler was killed.

It’s interesting that in this case, Sollecito had taken a strong interest in deviant pornography.  I think he would have been really turned on by the idea of setting up a pretty white girl to be raped by a black thug.

For Knox’s part, it seems that Kercher resented her promiscuous lifestyle.  For example, Kercher objected to the fact that Knox kept a vibrator and condoms in their shared bathroom.  Also, Knox had had numerous male visitors and Kercher had not.  There would have been strange men using the shared bathroom and Kercher would have resented this.  I would strongly guess that Kercher displayed her resentment by insinuating to Knox that Knox was a slut (which it seems she was.)  Girls do this to each other all the time.  I would guess that Kercher herself was a narcissistic bitch (did you know that she starred in a music video just a few months before she was murdered?)

It’s actually not super hard to imagine a conversation in which Knox mentioned to Sollecito how stuck up and annoying her roommate was and Sollecito proposed setting her up to get raped.  Much like the rape and murder of Janet Chandler back in the 70s.

I would guess that the intention of Knox and Sollecito was for Kercher to be raped at knife-point.  They expected her to submit just like girls presumably do in porn movies which depict rapes.  When she resisted, Guede started stabbing her and Knox and Sollecito fled in panic shortly thereafter.  Later, Knox and Sollecito decided that to avoid getting in trouble, they needed to pretend that they were not at the house at the time of the attack.  Probably in large part this would have been out of a need to avoid having to explain to the authorities why they did not contact the authorities immediately.

Of course, none of the above is proof or even evidence that Knox was involved, it’s pure speculation and I would write it off if there weren’t such compelling evidence against Knox.  Even with all of the evidence against Knox, I still think there is a decent chance — perhaps 5 or 10% — that she is innocent.  The point is that there are levels of extraordinary.  Claiming that Knox participated in her roommate’s murder is not like claiming that the president is actually an extra-terrestrial from Mars.  It does happen that nice white girls (i.e. girls without a criminal history) murder other white girls.

Anyway, the main important evidence against Knox is that when you look at her statements and actions for the time period in question, they just don’t add up but are instead consistent with her having been involved in the murder.  Even at trial, she was unable to provide a coherent account.  (For his part, Sollecito apparently exercised his right to remain silent.)  And we’re not talking about a few isolated inconsistencies — there are major problems with Knox’s story.  And obviously if she is hiding something important, it’s pretty unlikely that it’s something besides involvement in the murder.

For example, consider Knox and Sollecito’s flurry of phone calls made between 12:50 pm and 12:55 pm on November 2. It started with Knox calling her mother; explaining that there had been a break-in to the apartment; that her roommate was missing; and asking what she should do. Sollecito made a similar call to his sister. And a few minutes later, the pair actually did call the police apparently on the advice of their family members.

The problem is that when these calls were made, the (postal) police were already there at the apartment. They recorded their arrival time as approximately 12:30 and Sollecito later admitted that he had not yet called the police when the (postal) police had arrived. Why would Knox and Sollecito call the police when there were already police at the apartment? Why didn’t Knox mention to her mother that there were police already there? None of this makes sense unless you consider the possibility that Knox and Sollecito were trying to set up some kind of alibi to explain their failure to contact authorities upon seeing blood and signs of a break-in.

Here’s another example: Shortly after the postal police arrived at the apartment on November 2, Knox advised them that Kercher normally kept her bedroom door locked even if she left the room for just a few minutes. This fact is confirmed by the testimony of Knox’s roommate and roommate’s friends who also arrived around that time and as far as I know, Knox never denied it at trial. However, the roommate also testified that Kercher pretty much never locked her bedroom door. So it seems that Knox lied to the authorities right out of the blocks. Again, this makes sense if you consider that Knox was setting up an explanation for her failure to immediately contact the authorities earlier that day.

Here’s another example: When Sollecito initially called the police to report the break-in into the roommate’s bedroom, he stated that nothing had been taken. This was true, but how did he know it? By his account, the roommate had not yet had a chance to go through her bedroom and verify that nothing had been taken. Again, this makes sense if you assume that Sollecito (and Knox) staged the break-in.

Another significant piece of evidence against Knox and Sollecito was the fact that a break-in was apparently staged at the house.  Knox’s roommate — Miss Romanelli — apparently testified that upon returning to her room, she noticed that there was glass from the broken window on top of her strewn belongings.  Assuming the testimony is correct, it seems likely that the window was broken after the room was ransacked.  She also testified that nothing was taken and that whoever ransacked the room did not seem to have looked at obvious places where there might be valuables.  This is decent (but perhaps not compelling) evidence that the break-in was staged.  (Of course Sollecito’s statement to the police that nothing was taken is also decent evidence.)

Of course, it’s possible to explain away each piece of evidence against Knox in isolation.  But the scenario where it’s all explained away is extremely unlikely compared to the admittedly unlikely scenario that she took part in the murder.  My conclusion is that she was probably involved.

P.S.  I am happy to debate this with people, but please keep a few things in mind:

First, I am going by what I have read in published accounts of the crime.  I have tried to be cautious, but I admit there is a chance that my understanding of the facts is incorrect.  So please feel free to correct me.

Second, please try to respond to the arguments I actually make.  I can already imagine some joker saying “Sollecito’s interest in deviant pornography is not evidence that he was a murder!!!”  Dude, that’s not the argument I am making.  I have no tolerance for people who strawman me.

Are Jews Parasites?

May 21, 2011

It has been claimed to me that Jews are inherently parasitic. Is this true? Probably not, let’s look at the evidence.

The main evidence that Jews are inherently parasitic is the fact that Jews are heavily involved in jobs and professions like law and finance where workers seem to be just shuffling paper and getting rich while not actually producing anything.

However, it seems pretty clear to me that these sorts of jobs are actually very important. For example, if you want to build a modern factory or chip fabrication plant, you need a lot of fancy contracts to organize all of the workers and contractors involved. You also need bankers to raise capital. An informal arrangement might work for a small project, like two buddies opening a sandwich shop together but for something involving thousands of people and entities and millions of dollars of investment, you need bankers and lawyers.

I would imagine that a lot of the blacks in Zimbabwe used to think that the White farmers were just parasites; sitting back counting their money while the blacks did all the hard work. And yet without whites, those farms came grinding to a halt. Which just goes to show that just because your hands aren’t getting dirty doesn’t mean you are a parasite.

At the same time, there is one piece of evidence that Jews are not inherently parasitic, which is that we now have a country — Israel — which is overwhelmingly Jewish but which does pretty well economically. Would the Jews starve without the Arabs? Of course not.

Now, you might respond that Israel uses a lot of non-Jewish labor, but it doesn’t follow that Israeli Jews are parasites and in any event, similar things could be said about a lot of industrialized countries. For example, there are a lot of Filipino and Bengali workers in Japan and a lot of Turkish workers in Germany. There are a lot of Black and Hispanic workers in the US. Does this mean that Japanese and Germans are parasites? Does this mean that White Americans are parasites? I don’t think so.

What about the fact that some Jews feel that Jews are superior and Gentiles should be serving Jews?  Well first of all, there are supremacists in most groups and it doesn’t necessarily reflect the views of the group as a whole.  More importantly, one’s view of how things should be doesn’t necessarily reflect how things are.  If I believe that Alessandra Toressoni should be my sex slave, it doesn’t make it so.

Is the US Israel’s Puppet? Continued

May 9, 2011

There are those who argue that the US gets involved in conflicts, particularly conflicts in the Middle East because of the influence of the Israel lobby. This seems rather unlikely for a couple reasons:

First, most of these conflicts are between side which both hate Israel. Second, Israel’s interests in the conflicts are pretty minimal compared to other peoples’ interests.

For example, consider the recent conflict in Libya. Is it really in Israel’s interests to support the Libyan rebels? How? Or consider the Persian Gulf War. If the US was there because of Israel, why didn’t the US force the Kuwaiti government to recognize Israel as a condition of getting the country back for them?

Anyway, what exactly is the evidence for the puppet hypothesis? Pretty much the only evidence is that sometimes the parties to these middle-east conflicts blame their actions on Israel and accuse the other side of being controlled by the Zionists. (Frequently both sides make this accusation against the other side.) For example, Saddham Hussein claimed that the real reason he invaded Kuwait was because he’s against Israel. But why should anyone believe this? If it’s really about Israel, why not just attack Israel in the first place rather than first grab his oil-rich neighbor which also hates Israel? Why should we believe the self-serving claims of someone willing to rob and murder, especially when those claims make little sense?

The Race and IQ Question

March 16, 2010

I started this blog to share my thoughts on the Race and IQ question. 

Bascially, I believe that a signifcant amount of the black/white difference in cognitive abilities is genetic in origin.  I also believe this is obvious to anyone who thinks about the issue critically.  Following is a very brief summary of the evidence and arguments which compel this conclusion.  Later, I plan to respond to some common objections.

In short, the black/white difference in cognitive abilities is both universal and intractable.  In other words, you see it pretty much everywhere in the United States and the rest of the world; further, various attempts to eliminate this gap have failed.  This is exactly what one would expect to happen if the difference were largely genetic in origin.    Every other explanation I have seen (and people are very creative in coming up with alternative explanations) is either inconsistent with large amounts of evidence or is so full of epicycles that it’s just not credible.

What is an epicycle?  It’s an ad hoc adjustment to a hypothesis so that the hypothesis will not be contradicted by reality.  To borrow an example from Eliezer Yudkowsky, suppose somebody claims to have a dragon in his garage.  If you open the garage and don’t see a dragon, you might reasonably conclude that the claim is wrong.  However, the claimant might say “No, there’s a dragon in my garage — it’s just an invisible dragon.”  The claimant has just added an epicycle to his hypothesis.  You might then propose setting up a machine to detect the dragon’s breathing and the claimant might respond “well, the dragon doesn’t breath.”  That’s another epicycle.  Given enough epicycles, at a certain point, reasonable people conclude that they hypothesis is no longer credible.

As a side point, it’s interesting to note that just like in the dragon example, epicycles frequently involve factors which are difficult or impossible to measure.

In any event, the alternative explanations on the race and intelligence question are so epicycle-laden; so ridiculous that anyone with half a brain would dismiss them out of hand if it were not for political correctness.  i.e. western culture believes that it’s very very bad to believe or advocate that genetics are a big part of the reason for the black/white gap in cognitive abilities.


Can you make an effort to state in more detailed terms what it would mean to find that “genetics play a significant role in the black/white IQ difference”, in other words what precise predictions this theory makes?

For one thing, one would predict that anywhere in the US or the world, if you give a test of cognitive ability to a somewhat representative group of blacks, they will do worse than a somewhat representative group of whites.

For another, one would expect that this gap would persist over time despite attempts to eradicate it.

Is that precise enough for you?


Talk to the experts in psychometrics, and they’ll tell you that this is still an open question.

I have no idea if this is true, but there’s such a strong taboo against speaking out against the Egalitarian Hypothesis it wouldn’t mean much.

The experiments that would give huge likelihood ratios just haven’t been done.

So what?  There’s more than enough evidence already.   If we were dealing with a subject that weren’t politically sensitive, for example relative intelligence of dog breeds, everyone would be satisfied that the gap is largely the result of genetics.


From what I can tell of your blog post, you said, “there’s evidence, it’s so obvious, people have alternative explanations but they’re bogus, there’s evidence, I bet whites do better than blacks on tests, there’s tons of evidence.”

Where’s the evidence?

I’m a little confused.  Do you deny that whites, generally speaking, outperform blacks on tests of cognitive ability?


And this is compatible with a non-genetic explanation: environmental in African countries, and from discrimination in rich countries. Attempts at eliminating other kinds of discrimination (e.g. gender) have also been less than successful.

That’s not true.  For example, when discrimination against Jews in academia ended, Jewish people did just fine.  When discrimination against blacks in baseball ended, they did just fine.

Hello world!

March 15, 2010

Welcome to WordPress.com. This is your first post. Edit or delete it and start blogging!