Jump to content
Objectivism Online Forum

Racism or Correlations of Race with IQ / Physical Attributes

Rate this topic


The Wrath

Recommended Posts

just in case there is someone else who is confused about my answer to the river example, let me explain it here.

You are facing the same particular points in the same river that have caused/led to drowning of some boats. We do know the fact that some parts of a river can indeed be more structured for drowning boats than others; and we are told that one part has had a significantly higher record for this. Obviously you have to take this information into account since you are facing that same river.

In the case of a black person i meet, i do not know that there is something in the skin pigmentation that can cause a person to behave a certain way, so i have no connection to make whatsoever between this particular black person and the other black persons in jail. Similarly, we do not know that there is any connection between the gene for tallness and propensity to commit crime for me to take into consideration the statistic about the more tall people in prison when i am faced with a tall person.

But if the person i am meeting has a big record of theft, i will be inclined not to trust him (at least with my money) and this is precisely for the same reason as the river situation: we do know that there can be something in the life/experience/choices (not skin) of a human being that can cause them to be a thief. I must take the probability that he is more inclined to steal from me into consideration.

Now, if we MUST know that there is actually something in the skin of the person that can cause such character, we all know who has the burden of proof for that proposition.

Since you do not appear able/willing to address the post by the last contributor and, since my problems with your arguments are (more or less) the same and, since you indicated in your post just previously that you were primarily addressing me, parhaps you might be encouraged to deal with these issues if I put them to you, since, as I have said, much of the previous contributer's issues with your argument match onto my own at least on this particular fine point.

Before I begin, though, please allow me to apologise for referring to the content of one of your previous post's arguments as being silly. Obviously, I am happy to apologise for this. However, please be note that the apology extends that far and no further. That is to say, I apologise for offending your argument. since I did not offend you personally (I didn't actually realise that arguments had feeling that could be hurt). Unlesss, of course, you are unable to distinguish the difference between an argument and a person. I am assuming you are able in this regard.

In the case of a black person i meet, i do not know that there is something in the skin pigmentation that can cause a person to behave a certain way, so i have no connection to make whatsoever between this particular black person and the other black persons in jail.

In reference to the above, your use of the word "cause" causes me concern. The reason being that probabilites rarely imply cause. Or, at least it is very unwise to infer cause from them. However, what they do provide is evidence of a "correlational" relationship" between variable. However, a correlational relationship does not imply no-cause. It simply tells us that two or more variables are related. That is all. You use of the word "cause" suggests to me that you think unless a cause can be absoloutly established, probabilities are innapropriate. But, that is the point of them. If a cause can be absoloutly established, you will not require the use of probabilities.

Similarly, we do not know that there is any connection between the gene for tallness and propensity to commit crime for me to take into consideration the statistic about the more tall people in prison when i am faced with a tall person.

Well, if we accept the original thought-experiment premise (hopefully I have got this the right way round!) that there were more tall people in prison than short people, there is, by definition, a relationship between the two variables of "presence in prison" and "height". It may or may not be a causal relationship. But that there is a relationship in evidence is logially undeniable.

Thus, when faced with a moral dillema to place trust in a tall person or a small person, in the absence of any other evidence, logic would dictate that you chose to place your trust in a shorter person. Probabilities, being what they are, menas that you would be likely to be correct in your actions X percentage of the time and incorrect Y percentage of the time. This would be the case whatever the nature of the relationship between the variables. Be it correlational or causal.

Link to comment
Share on other sites

  • Replies 377
  • Created
  • Last Reply

Top Posters In This Topic

I think you are mixing up certainty with probability.
What exactly do you think "probability" means? And from #201...
The reason being that probabilites rarely imply cause. Or, at least it is very unwise to infer cause from them.
Then do you think that probabilities are of any use at all in life? Why? What difference would it make if two things were tightly correlated, or not correlated?
Link to comment
Share on other sites

I wonder if Onar has posted this link anywhere on these boards before:

http://www.charlesdarwinresearch.org/Race_...on_Behavior.pdf

I haven't read it myself, though as soon as I look into some of the already standing criticisms of the work (http://en.wikipedia.org/wiki/J._Philippe_Rushton), if I find it at all credible I do believe I'll look into it.

But I wonder if anybody else has a response to this.

Link to comment
Share on other sites

But I wonder if anybody else has a response to this.
I don't find it scientifically credible. One of the problems is that we do know that skin color is inheritable, thus genetic. Some facts about the body are genetically determined, which gives him the opportunity to engage in the fallacy of composition. The question is, what do you want to explain? The basic issue in this thread is whether there are cognitive-capacity limitations on blacks as a race, which are genetic in nature. My contention is that there is no evidence for such limitations, and that Rushton's claims are not solid. For example, an inference drawn from observation of American blacks cannot be validly used to infer a property of the whole population of blacks (African blacks in particular, excluding Pacific and Indian Ocean blacks), since we don't have a sample where all members of the population have an equal chance of being included in the sample. [This is because virtually all American blacks are the descendants of slaves, who were selectively taken from specific areas of Africa].

Rushton fails to consider cultural questions in dealing with cultural matters such as birth out of wedlock rates, crime rates, self-esteem questionaires and so on. One explanation for why John (arbitrary subject under investigation) is in jail for theft and has 14 children with 14 different women is that he never knew his father, his mother was a crack whore, he himself is a crack baby, he grew up in a lousy neighborhood listening to value-destroying hip-hop music all the time. We could do a long exegesis on cultural problems, and discover what we've known all along, that the culture actually does matter. Rushton does not control for cultural variables, and when a study omits a major factor influencing ones decisions, it's not going to be a valid study.

Link to comment
Share on other sites

[This is because virtually all American blacks are the descendants of slaves, who were selectively taken from specific areas of Africa].

David,

Doesn't this imply that it could be genetic? I agree that you cannot extrapolate these facts to all blacks from that(I havent read the study-just going by your interpretation of it)but if it was shown to be consistent in the US controlled for culture, then it could still be a genetic based on the particular tribes and backgrounds. Also, I thought viking had previously provided evidence that they did control for culture by looking at adopted blacks?

What I don't understand, though, is why it matters to a layman if it is genetic or cultural. What is, is. If something is positively correlated, for whatever reason, it is going to be of help making decisions with regard to the subject because of it's predictive properties. That to me, is the chief value of probabilities. The question of the ultimate causation only seems relevent to the treatment of mental disorders or physical ailments.

Link to comment
Share on other sites

Ok, first of all:

Race is a valid concept.

Race is not an arbitrary category like "striped animals" or "red animals." Such categories bring together entitites based on a non-essential similarity. Race, however, is formed in exactly the same way as species, namely by descent. (the scientific name for race is sub-species) A race is a group of individuals that have a common set of ancestors that differ from other such groups. The negroid or black race is the set of people whose ancestors lived in sub-Saharan Africa 90.000 years ago. The caucasoid or white race is the set of people whose ancestors lived in Europe some 60.000 years ago, and the oriental or mongoloid race is the set of people whose ancestors lived in Asia some 60.000 years ago.

Species have the additional property that the groups are separated by reproductive incompatibility. This makes species exclusive categories, whereas sub-species or races are statistical cateogories. We humans use a lot of valid statistical concepts, and one of these is climate -- which is just the average weather over some period. Winter, spring, summer and fall are also such valid statistical concepts. They are valid because they are *useful*. Even though some summer days may be cooler than spring days or even winter days in various parts of the world, there is no question that there is a real physical cause to real statistical differences between the seasons. It is perfectly valid to say that summers are warmer than winters.

Race too is a statistical concept. Since it is a measurement of descent it is also perfectly valid to talk about mixed races. To say that someone is 78% white and 22% oriental means that 78% of one's ancestors lived in Europe 60.000 years ago and 22% lived in Asia 60.000 years ago. Just like winter and summer, black and oriental are useful in making statistical predictions. One of the most important variables that can be usefully predicted by race is intelligence.

intelligence is a valid concept

General intelligence 'g' is a biological metric as valid as height or weight. It can be measured and it can be used to make useful predictions. 'g' is correlated to a range of biological parameters such as brain size, IQ, brain pH-value, brain metabolism and perceptual reaction times. The latter is the key component in constructing a new ABSOLUTE value for 'g', namely the emerging science of mental chronometry. By using simple cognitive tests such as how fast a test person is able to respond to a perceptual signal (e.g. a lamp) one can construct a value of how fast the mind is, and this value -- mental speed -- is equally well correlated with general intelligence as IQ-tests. The advantage of such tests is that they are 100% free of cultural influences. They are so simple that no education is required. For more on this emerging science I recommend "Clocking the Mind" by Arthur Jensen:

http://www.amazon.com/Clocking-Mind-Chrono...ndividualdif-20

Intelligence is real and measurable.

Now, using mental chronometry it is easy to verify that the 'g'-loaded IQ-tests are largely culture independent. It is a fact that using these simple cognitive tests such as reaction times, blacks consistently score lower than whites on average who in turn consistently score lower than orientals on average. Hence, mental chronometry confirms what the myriads of IQ-studies already have shown, namely that there is a significant *average* difference between the races in general intelligence.

While race is not a useful concept for individual assessment, it is most certainly useful in explaining *statistical* data. Why for instance are Ashkenazi-Jews so dramatically overrepresented in all areas of intellectual accomplishment such as chess, mathematics, physics, chemistry, economics, entrepreneurship, leadership, wealth, politics, litterature and philosophy? It can easily be explained by the higher average intelligence of ashkenazi-Jews.

Why are whites on average richer than blacks? Why are there more illiterates among blacks than whites? Why are there more criminals among blacks than whites? Again, all these variables can be largely explained by intelligence. Obviously culture plays an important role in crime, but the numbers don't lie: if we categorize people based on intelligence then lo and behold: most racial differences vanish. Blacks with an average IQ of 100-110 are typically no more or less criminal than whites within the same range. Furthermore, regardless of race, there is an extremely strong correlation between crime (especially violent crime) and *low* intelligence. Obviously most people of low intelligence are not criminals, but the numbers tell us that they are more *susceptible* to crime than high-IQ people, and this is true regardless of race. Low-intelligence people are simply more vulnerable to absorbing really bad ideas than others. This should not come as a shocker to anyone.

The good news is that equipped with this knowledge we have real options about various actions, both interpretative and actions for reducing crime. First, higher crime and out of wedlock pregnancy rates among blacks should NOT automatically be interpreted as bad culture/bad moral. When correcting for IQ we find little differences between the races. Second, one should NOT automatically interpret racial differences in social indicators (such as crime, conviction rates etc) as the result of white racism, but rather as the statistical outcome of racial differences in intelligence. Adjusted for IQ there is no evidence of bias against blacks, quite the contrary.

The Bush-Gore ballot recount debacle in Florida is another instance of wrongly interpreted data. Many argued that black votes were systematically thrown out in order to skew the result in favor of Bush, but again, once adjusted for IQ all differences disappeared. In fact, the ballot recount in Florida can be considered the largest elementary intelligence test in human history. Being able to cast a correct vote actually requires some degree of intelligence. Not much, but sufficient to make the success rate less than 100%. Since blacks on average have a lower intelligence than whites they should be dramatically overrepresented in ballot failures, which was exactly the case. In fact, the ballot failure rate of the various races perfectly matched the difference as measured by myriads of IQ tests for the last 70 years, and provided an extraordinary independent validation of them. Again, had the knowledge of racial differences in intelligence been widely known, no false accusations of racism could have been made.

Furthermore, we know that there is an extremely high correlation between crime and illiteracy, and since there is a high correlation between illiteracy and low IQ it is no wonder that blacks are so dramatically overrepresented in crime. Luckily this also spells a way out of the problem: reading and writing training specially tailored for the weaker minded. By IQ-testing children at an early age one can identify the children that are likely to need special learning methods (much rote learning) to achieve a high litteracy rate, and thereby dramatically reduce crime. Had this knowledge been widely known and accepted without social stigma crime would be much less of a problem in the US than it is today.

Link to comment
Share on other sites

Doesn't this imply that it could be genetic?
In the sense that if we're talking about an inheritable trait, it could be inherited. A claim about the black race is a very broad claim, because to be true it must be a fact about the whole race. If it were true only of one tribe living in one country, which happened to be a dominant source of slaves to the New World, then at best you would identify a trait of that tribe, not the entire race. I've been after Viking to clarify his claim, to see what, exactly, it's supposed to be a generalization about, and I still don't get it. I'm not assuming that the argument is valid, but pointing out one very simple way in which the relationship between data (not that there is data under discussion) and conclusion is utterly skewed.
Also, I thought viking had previously provided evidence that they did control for culture by looking at adopted blacks?
No, you see, the problem is that they didn't actually control for the actual facts of the children's cultural associations. Being raised by middle-class Quaker parents doesn't means that you automatically adopt those values. It only means, at most, that you were exposed to those values, and if you adopt a punk lifestyle, it was your choice. My highly limited experience with transracial adoptions tells me that a black child adopted into a white family will, indeed, eventually be able to determine that he is black, and that he's the only black kid in the neighborhood. But the discussion crashed at the point of him actually making the case from the adoption study.
What I don't understand, though, is why it matters to a layman if it is genetic or cultural. What is, is. If something is positively correlated, for whatever reason, it is going to be of help making decisions with regard to the subject because of it's predictive properties. That to me, is the chief value of probabilities. The question of the ultimate causation only seems relevent to the treatment of mental disorders or physical ailments.
If it did have predictive value, sure. But consider this: you cannot chose your race, but you can chose your culture. That means that if blacks are physically predestined because of their bodies to be criminals, then that would be tragic, but it would mean that because of this lack of free will, they can't be trusted in certain occupations. But this clearly is not the case. If observed tendencies to criminal behavior are caused by something cultural, they you want to test for that cultural factor, and not race, since the actual causal factor is a vastly better predictor. It would, for example, allow you to distinguish white mass murderers like Timothy McVeigh, Joseph Stalin, Eric Harris and Dylan Klebold from black scientists like George Washington Carver and Herman Branson or government leaders like Colin Powell and Clarence Thomas. What is a real situation where contemplation of a person's race is of actual value in making a decision? So far, I have seen not one shred of evidence for that.
Link to comment
Share on other sites

But consider this: you cannot chose your race, but you can chose your culture. That means that if blacks are physically predestined because of their bodies to be criminals, then that would be tragic, but it would mean that because of this lack of free will, they can't be trusted in certain occupations. But this clearly is not the case. If observed tendencies to criminal behavior are caused by something cultural, they you want to test for that cultural factor, and not race, since the actual causal factor is a vastly better predictor. It would, for example, allow you to distinguish white mass murderers like Timothy McVeigh, Joseph Stalin, Eric Harris and Dylan Klebold from black scientists like George Washington Carver and Herman Branson or government leaders like Colin Powell and Clarence Thomas. What is a real situation where contemplation of a person's race is of actual value in making a decision? So far, I have seen not one shred of evidence for that.

It is not a deterministic factor. And I didn't say that it's about the individual judging himself. It is about judging strangers based on appearance when that's all you have. Their culture, while certainly a better predictor, is not immediately obvious, whereas their race is. And understand that I am looking at this from a practical point of view.

So to make this less emotional of an issue, if redheads who wear green underware were 100% likely to be violent, and 98% of redheads wore green underwear, it would be advisable for non-redheads to make a wide arc around them on a sidewalk even if they didn't know the color of the particular redhead's underwear. You have no obligation to cater to the feelings of the 2% of redheads who happen to not wear green underwear, putting yourself at risk.

As far as a real example goes; an employer believes that mexicans work harder then caucasions. He interviews 2 people for the job, 1 white and 1 mexican. In the interview he concludes that they have equal skills and background employment. He is in a standard approach-approach conflict unable to decide based on his available information, so he decides to lean in favor of the mexican based on his previously established generalization. He has every right to do that, and depending on the accuracy of his statistics it may or may not be a very good idea.

Making decions in the real world is based on the information one has which can be accurate or inaccurate, and in various states of completeness. The fact that an objectively right answer exists does not mean that anyone will be able to spend the time necessary to find it in all circumstances. Having a limitated amount of resourcs to devote to any particular thing obligates us to make generalizations and take chances based on probability. Fair and causal or not.

Link to comment
Share on other sites

It is not a deterministic factor. And I didn't say that it's about the individual judging himself. It is about judging strangers based on appearance when that's all you have.
And I, likewise, did not say anything about an individual judging himself. If race does not determine behavior, when why would a rational man make reference to race when making a decision that is based on prediction of future behavior? An irrational man could, but a rational would not. A rational man would understand that if you have no evidence and cannot obtain evidence, then a rational decision is not possible. A rational man would therefore not decide -- would not act irrationally. And a rational man would not need to decide, unless you were dealing with the initiation of force.
Their culture, while certainly a better predictor, is not immediately obvious, whereas their race is. And understand that I am looking at this from a practical point of view.
On the other hand, from a practical point of view, there is no circumstance where you can't ask a few questions and get superior knowledge about the individual, which goes directly to the individual choice and culture questions.
You have no obligation to cater to the feelings of the 2% of redheads who happen to not wear green underwear, putting yourself at risk.
Hell, suppose 100% of whites were murderers and 0% of blacks were murderers, you still have no obligation to deal with either blacks or whites. You have the right to deal exclusively with whites, regardless of the facts. Since the argument is not about other people's feelings, and is about whether one is acting irrationally by acting irrationally (the conclusion should be "yes"), I don't know how the invention of funny but impossible stories about redheads and underwear is relevant.
As far as a real example goes; an employer believes that mexicans work harder then caucasions. He interviews 2 people for the job, 1 white and 1 mexican. In the interview he concludes that they have equal skills and background employment.
I notice that this is simply a random and irrational belief, and if he had a different belief, he would make a different decision. What I don't understand is why it is, if he is concerned about whether an employee would work harder, that he doesn't actually inquire directly into that matter. Is it that he cannot imagine how to ask a question that would reveal a man's attitude towards work?
He has every right to do that, and depending on the accuracy of his statistics it may or may not be a very good idea.
In fact, he has every right to knowingly hire the worst worker. Issues of rights are totally irrelevant to this question. A man has a right to act irrationally, even though he should not. It is only when his irrationality and immorality crosses a certain line, by violating the rights of others, that a man mas no right to act irrationally. Because rights violation is not under discussion here, the rights question is irrelevant. The "depending on the accuracy of his statistics" comment is contradictory, insofar as you started with an unreasoned belief and now want to assume that it is a reasoned belief.

So let me restate your hypothetical, making it consistent and also more streamlined. An employer has two candidates, an Anglo and an Indian (most contemporary Mexican immigrants are more Indian than Spanish), and they both know how to work a nail gun. The employer has been told that Indians are harder workers compared to Anglos, and for some reason he believes this, without questioning why it would be true. Depending on the truth of the claim about Indians this may or may not be a very good idea. However, even without introducing the assumption about Indians as hard workers, the decision will turn out to have been a good or bad idea, depending on actual outcome. Hiring the Anglo is a good idea if, and only if, the Anglo is not a worse worker than the Indian. It actually does not matter if you have been given some type of authoritative information about Indians or Anglos, since the information might be grossly wrong, or generally right but not applicable in the present instance. And it would not matter if the information were authoritative, or just something you heard from a drunk guy at the local bar.

This is not to deny that unthinking men who operate on an emotional basis, and reject the mind as man's proper tool for survival, will act in such a way. But then such men do not need a basis for making decisions, and I don't see that we need to waste time discussing how irrational people behave, except insofar as it's useful to be able to identify them and protect yourself against them.

A thinking man would use his mind to integrate the facts that he has. He knows, in fact, that not all redheads are prone to violence; that not all blacks are ignorant or criminal; that not all Anglos are shiftless parasites. Rather than take the lazy way out and pleading "I had no way of knowing that the Anglo was industrious, the redhead was virtuous, the black was intelligent", the employer could act in his own self-interest and take the initiative to directly inquire into the matter. The inquiries may not be exhaustive, but a rational inquiry into the actual matter of interest, be it moral virtue, intelligence, or perseverence, is vastly more likely to yield the desired outcome. The circumstance simply does not exist where race is the only other information available.

Link to comment
Share on other sites

I believe this discussion may have failed to notice Onar's post, which was posted at nearly the same time DavidOdden posted.
Actually, there was a part of Onar's post that I thought especially deserved comment, namely his first section on the validity of the concept race. I think it's clear that what he's proposing is at once quite valid but also, simply, not the same as "race". Race is way too broad a category, so for example it is a serious genetic mistake to lump Arabs and Scandinavians into one race, or put North Indians, Chinese and Philippinos in another race. Georgians, who are (virtually by definition) Caucasians in fact do not have any ancestors who lived in Europe.

Previously, when we were stuck with chainsaw-precision tools, the valid questions could hardly be asked, because we had no scientific means of inquiring into a person's remote ancestory. DNA studies are the microscopic-precision tool that is changing that, whereupon it turns out that many of the MacDonalds, MacDougalls and MacAlisters had ancestors living on Mongolia. If you're interested in genetic factors, you should look at genes, and sort your sample by objective and relevant criteria.

Link to comment
Share on other sites

One of the TV shows we've imported into our country (as we have slightly developed in recent years) is something called "The Jerry Springer Show". I briefly watched some of it this afternoon and i must say that, as a black man, i did not feel particularly overwhelmed by the greater genetic intellect of the white guests on that show!

Link to comment
Share on other sites

One of the TV shows we've imported into our country (as we have slightly developed in recent years) is something called "The Jerry Springer Show".

Truly a sad day for your country... American TV at it's worst.

Link to comment
Share on other sites

One of the TV shows we've imported into our country (as we have slightly developed in recent years) is something called "The Jerry Springer Show". I briefly watched some of it this afternoon and i must say that, as a black man, i did not feel particularly overwhelmed by the greater genetic intellect of the white guests on that show!

This is why arguing averages is pointless, when the spread (standard deviations) cause the curves to not be that different, and even those in the supposedly superior group still have immense populations of idiots. What RB said. I feel for the damage inflicted on your country by the importation of such garbage.

Link to comment
Share on other sites

If race does not determine behavior, when why would a rational man make reference to race when making a decision that is based on prediction of future behavior? An irrational man could, but a rational would not.

What you are consistently not taking into account in your response to my examples is time constraints and limitations on actual knowledge possessed. If you get a 10 minute interview with a bunch of guys with only certain questions which are allowed to be asked in the interview, it is entirely concievable and factually common that you could have two resumes and two interviews in which all things appeared equal. The decision might come down to knowledge about a culture or a race associated with a culture. It's not ideal, but reality usually isn't.

But really these examples are not going to get us anywhere since it seems that you disagree at the most fundemental level that statistics and probabilities have any value at all in the decision making process. If that is your position, there's not anything I can say to convince you. I generally like your posts and find your oppinions well reasoned and valuable, but this strikes me as an out right denial of obvious reality. I hope and assume that I misunderstand your view.

Because I have a finite mind, and perhaps a small "irrational" one at that, I find general principles and statistics that are accurate at predicting events and behaviors above chance to be valuable in making decisions for which I am unable or unwilling to invest enough time to extrapolate all of the causal connections. I consider it to be a metaphysical necessity to make decisions based on limited knowledge and statistical information. If that's irrational then throw me in the category with people who wear meat hats and juggle geese.

Link to comment
Share on other sites

Race is way too broad a category, so for example it is a serious genetic mistake to lump Arabs and Scandinavians into one race

"Genetic clusters" is perhaps a technically more accurate term. Mapping out these clusters you find that they in turn divide into sub-clusters. I.e. there is a tree structure, just like you have with species. In other words, just like climate can be viewed at different time scales, so can races. Thus, when you speak of whites, you can specify what kind of whites you are talking about and thereby increase granularity, e.g. germanians and hispanics. As long as such increased (or decreased) granularity provides useful information (i.e. predictive power) there is a valid basis for it.

Previously, when we were stuck with chainsaw-precision tools, the valid questions could hardly be asked, because we had no scientific means of inquiring into a person's remote ancestory. DNA studies are the microscopic-precision tool that is changing that, whereupon it turns out that many of the MacDonalds, MacDougalls and MacAlisters had ancestors living on Mongolia. If you're interested in genetic factors, you should look at genes, and sort your sample by objective and relevant criteria.

Indeed, the preferred term is genetic or population clusters, I believe. However, there is no question that there is a very strong correlation between geography (and language/culture) and these genetic clusters. The driving force behind clustering is geographical and cultural isolation, hence it makes perfect sense to name the clusters after their geographical origin.

Sometimes the isolation mechanism is strongly cultural as is the case with Jews, especially European Jews or Ashkenazim, originating from a few individuals some 1000 years ago. The genetic evidence shows that they have a Middle Eastern origin, with some influx of European genes. This influx is so small as to suggest strong cultural isolation and intermarriage. Evidence also shows that this category has strong prediction value, meaning that it is reasonable to name them a separate cluster -- or race if you like -- namely Ashkenazi/European Jews.

Link to comment
Share on other sites

But really these examples are not going to get us anywhere since it seems that you disagree at the most fundemental level that statistics and probabilities have any value at all in the decision making process.
The disagreement is over (1) whether there even exists a relevant statistic in this topic -- there doesn't; and (2) whether evading reality is a virtue in an "emergency" -- it isn't. When you substitute an accidental observed correlation for consideration of the real causal factors, you may avoid the consequences of your error for a while, but ultimately you will pay the price for not paying attention to reality. So if you believe that blacks are intrinsically violent and hire your workers according to that principle, you will at some point pay the price in not hiring a better candidate simply on the grounds that he is black and therefore will engage in violence in the workplace, and you will also pay the price by hiring a violent white man who you believed would not be violent because he is white. A simple check of prison background would have saved you a lot of anguish, had you been aware of the true cause.

I'm not denying that people will use invented "statistics" and will avoid actively using their minds on a long term basis, because they feel that they have an emergency that must be addressed right now. I'm simply denying that this is reasoned behavior for a person who survives by using his mind. If a person is overwhelmed by reality and the difficult of solving a problem using the mind, then I certainly understand why a man would want to have some automatic rule to call upon in order to force a choice. Though I can give you other rules that can be applied as quickly and would be vastly more reliable in judging between two paper-equal candidates:

  • Which one speaks better English?
  • Which one says more sensible things?
  • Which one dresses better?
  • Who has the better posture?
  • Which one engages your eyes better during questioning?

How often do you find that answering these questions don't tell you what you need to know?

Link to comment
Share on other sites

As long as such increased (or decreased) granularity provides useful information (i.e. predictive power) there is a valid basis for it.
Right: the other thing to be sure of in talking about correlations is to not accidentally engage in context-dropping. Even though man is a mammal, it's a mistake to treat "man" and "mammal" as interchangeable. This won't happen very often, since we have clearly separate concepts such as "man", "mammal", "monkey", "ape", "primate" and so on. At some point, as our knowledge expands, we may find it useful to form concepts referring to different nodes on such a cluster-analysis tree, so that there could be a clear definition of the concepts "African-American" vs. "Afro-Carribean" vs. "African", and then we could simply check when a person makes a statement about "Africans", that it does indeed apply to that node and not another node.
The driving force behind clustering is geographical and cultural isolation, hence it makes perfect sense to name the clusters after their geographical origin.
I agree that this used to be the case, but is decreasingly less relevant now, especially in the US where the isolation assumptions are mostly invalid, and entire new genetic clusters are being created.
Link to comment
Share on other sites

[*]Which one speaks better English?

[*]Which one says more sensible things?

[*]Which one dresses better?

[*]Who has the better posture?

[*]Which one engages your eyes better during questioning?

How often do you find that answering these questions don't tell you what you need to know?

When, as I stated in my example, both people come across close to the same. Close enough that one does not stand above the other in any significant way.

I don't know enough about race science to agree or disagree as to whether statistics associated with race are valid or not. It really isn't a subject that interests me all that much and I don't intend to debate that or research it. The only thing that bothers me about this discussion is this idea probabilities should not be used because they are a sign of intellectual laziness or irrationality, or being overwhelmed by reality, or any of these other statements you have made.

I call using statistics a recognition of reality. Particualrly recognizing your own mental imitations. I am not advocating replacing causal knowledge with probabilities, but I do believe that probabilities are good to use as an adjunct to causal knowledge or in the absence of causal knowledge.

I want to pin you down osomething here because there is something we seem to be missing in each others arguments.

Are probabilities a valid and useful form of knowledge, or should they be always disregarded?

Link to comment
Share on other sites

In a similar spirit as aequalsa's post, I would like to know: Suppose there is a causal relationship between a person's genes and his brain size, and a relationship between his brain size and his intelligence, and between his intelligence and his ability to perform at a certain job, and between his genes and his geographic location. Moreover, suppose that genes which compose the DNA of whites causes a smaller brain size than the brain size caused by Asian genes; and that the smaller average brain size causes lower intelligence; and that lower intelligence makes one a less desirable worker; and that whites often live in a particular location and Asians often live in a distinct location. And so a business issues advertising campaigns in predominantly Asian locations (unable to obtain more complicated information like average intelligence in a geographic area) as opposed to white locations, though once any given person applies for a job they will consider that person purely by merit rather than race. Would you call this an irrational business decision? Do you have reason to believe, assuming these premises, that the business will do as well or worse than another business which, all other things being equal, ignores this information and does not advertise heavily in one area or advertises more heavily in white locations? If so, why?

Link to comment
Share on other sites

Mass marketing is a good example where businesses do use stats like that. It comes down to the cost of obtaining better information and the cost of better targeting. Businesses would love to "narrowcast" rather than broadcast. They do try various ways to improve their targeting, but even so, they do realize that they're missing many and wasting their message on many others.

For instance, an investment firm may mail "Wealth Management" brochures to people who recently bought Mercedes-Benz cars, but mail more mundane "Investment Account" brochures to people who bought Honda Civics. The first group may have some who overreached their budgets and have little left to invest, while the second group may have people who can invest precisely because they don't buy expensive cars, or who're buying a car for a kid, or whatever. Still, if experience shows that, on average, the high-end car buyers are more likely to be the type who will respond to "Wealth Management", and if little else is known about these people, then it's perfectly reasonable to use that limited information for a marketing campaign. Another marketer may come along and find that he can reach a better audience by ignoring car-buying and looking at something else, e.g. country-club membership.

I can't readily think of a situation where one would use something like that with respect to an individual from whom more information can be gleaned, in any situation that isn't trivial. Very often, even though the individual shares the essential differentiating characteristic used in the study from which one is drawing a statistic, (e.g. the attribute of "race"), he may be unrepresentative. Often, the particular context (a particular location, a particular job interview, etc.) could easily be such that if one took all the black people who are present in those types of contexts one will find that they measure quite different from the country's population of blacks as a whole. So, even if one does not know more about the person, their presence in a particular context changes the applicability of your pre-known statistical probabilities.

A final point, when it comes to government decisions, there is even more reason to be conservative in applying such statistics.

Link to comment
Share on other sites

In a similar spirit as aequalsa's post, I would like to know: Suppose there is a causal relationship between a person's genes and his brain size, and a relationship between his brain size and his intelligence, and between his intelligence and his ability to perform at a certain job, and between his genes and his geographic location.

By transitivity, you are asking us to assume that there is a causal relationship between an individual's genetic makeup and his ability to perform at a certain, unspecified job. Perhaps this can be true if the job involved winning professional deadlifting competitions. However, for a general job in modern society, this appears to be a spurious assumption. Can you think of some other examples of occupations and inherited characteristics where this would be the case?

I assume that you also mean that there is a strong positive correlation between genetic similarity and geographic location. Not a causation.

Moreover, suppose that genes which compose the DNA of whites causes a smaller brain size than the brain size caused by Asian genes; and that the smaller average brain size causes lower intelligence; and that lower intelligence makes one a less desirable worker; and that whites often live in a particular location and Asians often live in a distinct location. And so a business issues advertising campaigns in predominantly Asian locations (unable to obtain more complicated information like average intelligence in a geographic area) as opposed to white locations, though once any given person applies for a job they will consider that person purely by merit rather than race. Would you call this an irrational business decision?

Well, if we are assuming that individuals from one population are invariably better on average at performing a task T than those from another population then any business that seeks to recruit from the better population to fulfill their need for individuals to complete task T will more than likely be better off than a business who chose not to exploit this knowledge that we have assumed to be corrected. Needless to say, the sweeping assumption we have made here allows this conclusion to follow without any sophisticated arguments. I question the genetic causalities that we have assumed.

As for the question of is it the best business decision to market just towards one population; it depends. Is the company able to market towards a different, population that contains individuals from both races who are generally more qualified? For example, can the company recruit from universities where individuals of either population can be highly trained to suit the business' needs? How about running advertisements in professional societies where members are highly qualified in the desired skill set?

It really would be a suboptimal business decision to misconstrue race to be the important factor when there are more descriptive factors that one can just as easily focus on. This would even be the case if one is looking for experts on algorithms and statistically a higher percentage of one population of a certain race holds more advanced degrees in mathematics and computer science than another population of roughly equal size. Why focus on race or geographic regions when you can focus on programs in mathematics and computer science?

Link to comment
Share on other sites

Are probabilities a valid and useful form of knowledge, or should they be always disregarded?
There was discussion of certain relevant points here. The question itself is invalid since it has multiple contradictory referents, so until you explain what you mean by a "probability", I can't tell you the answer to your question. The sense in which you are using "probability" is basically frequency, and as such, it is only "evidence" of itself -- the rate at which X occurred only tells you the rate at which X occurred. You can combine that with other conditions and get a useful form of predictive knowledge and not just descriptive knowledge, though it may be of limited usefulness (i.e. it may be true only for one year and if you rely on it the next year, you would be in trouble). The point is that if you want to use frequency as a tool for predicting the future (or, simply, another audience), the producer of the statistic has to guarantee that certain conditions hold. Since information is massively compressed when vended as a "statistic", you should care whether the information-reduction techniques are valid. Ultimately, the problem with statistics is not that they are founded on a corrupt metaphysics, but that they are very often technically invalid, because significant information has been irretrievably destroyed.

I want to remind you that statistics are created, and not naturally occurring objects; and we are assuming, to make the context most applicable to your life, that some vendor is offering you a statistic and you have to decide whether the buy the statistic and act on it (literally and

metaphorically). Do you have sufficient reason to trust the source (and why)? Are you willing to listen to, and evaluate, competing statistics that contradict the conclusions of provider A -- even if A's statistics square with your emotional prejudices and B contradict them? That is, are you more interested in knowing reality, or in justifying a decision?

BTW I'm not suggesting some vast evil fabrication racket, but something worse, namely ignorance interacting with bad methodology. If you are unaware of how these statistics are created, then you're pretty much stuck with what you're told, and I don't know how you can decide whether to believe a statistic or not, except by learning something about the topic. If you can't do that, you can't; if you're living according to the "active mind" credo", then you should check your premises. Rand has often urged people to check their premises, and I think that's good advice.

Link to comment
Share on other sites

By transitivity, you are asking us to assume that there is a causal relationship between an individual's genetic makeup and his ability to perform at a certain, unspecified job. Perhaps this can be true if the job involved winning professional deadlifting competitions. However, for a general job in modern society, this appears to be a spurious assumption. Can you think of some other examples of occupations and inherited characteristics where this would be the case?

I assume that you also mean that there is a strong positive correlation between genetic similarity and geographic location. Not a causation.

The whole point, though, is that we seem to have evidence that the relationship between genes and potential physical power on the one hand, and between genes and intellectual capacity on the other, are similar relationships. There can be no doubt that genes stand in some kind of causal relationship here, for a plant's genes will never give rise to a conscious brain. So our specific type of genes cause us to have some intellectual capacity. The question, then, becomes: Is this capacity the same in all of us, or do different genes produce different capacities? I am asking that we provisionally assume that we have statistical evidence for the later, and see what would result--what use this information would have for predictive purposes. Therefore, this example is perfectly adequate. Can we predict that a company which followed racist advertising strategies (i.e. consciously and specifically targeted Asians rather than whites) would--all things being equal--beat out a competitor?

Needless to say, the sweeping assumption we have made here allows this conclusion to follow without any sophisticated arguments. I question the genetic causalities that we have assumed.

That's all I'm trying to establish, since even this much seems contested here. With this much granted, then, I wonder if Onar's statistical evidence actually supports the premise that, on average, one race may be more intellectually powerful than another.

As for the question of is it the best business decision to market just towards one population; it depends. Is the company able to market towards a different, population that contains individuals from both races who are generally more qualified? For example, can the company recruit from universities where individuals of either population can be highly trained to suit the business' needs? How about running advertisements in professional societies where members are highly qualified in the desired skill set?

This is why I added the proviso, "all other things being equal"--i.e., no considerations outside of geographic and genetic considerations can distinguish the activities of the two companies, ex hypothesi.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...