Carryover vs “Far Transfer”

It used to be thought that studying certain subjects ( like Latin)  made you better at learning others, or smarter generally – “They supple the mind, sir; they render it pliant and receptive.” This doesn’t appear to be the case, certainly not for Latin – although it seems to me that math can help you understand other subjects?

A different question: to what extent does being (some flavor of) crazy, or crazy about one subject, or being really painfully wrong about some subject, predict how likely you are to be wrong on other things? We know that someone can be strange, downright crazy, or utterly unsound on some topic and still do good mathematics…  but that is not the same as saying that there is no statistical tendency for people on crazy-train A to be more likely to be wrong about subject B.  What do the data suggest?

I’m mostly talking about cases where A and B have no obvious relationship – we know that a dedicated Marxist is more likely to have wrong notions about genetics, since heritability offends him ( probably gravity and levity do as well).   Probably also worth considering whether the individual in question achieved craziness on his own or was immersed in it from birth.

Lastly, sometimes that correlation must be negative: certain flavors of crazy likely make you less  likely to be wrong about certain other things.

 

 

 

 

Advertisements

Let justice be done, though the heavens fall

If we had operated on that motto at the end of WWII, we would have executed a whole lot more Germans and Japanese than we did.

 

 

Gene genealogies vs population splits

Genes have genealogies, just as populations do. Usually, a gene’s  genealogy is the same as that of its species.  For example, turtles have their own versions of hemoglobin, and the common ancestor of those turtle hemoglobin genes is some ur-turtle  hemoglobin ( or is it turtle ur-hemoglobin ?) a long time ago.  This is a stochastic thing: it happens when only a single version of a gene survives – after that occurs, gene trees and population trees match, at least until the next split.

But it ain’t necessarily so, especially in cases where two species separated rather rapidly, or when the splitting populations are large – even more so when populations haven’t yet speciated at all.  When the gene tree conflicts with the population tree, we call it incomplete lineage sorting.  For example, gorillas split off before humans and chimpanzees separated, but some of the human genome (  ~30% ) is more similar to gorillas than it is to chimpanzees.   So, gene trees don’t match species trees very well  among humans, chimps and gorillas – although they do just fine between, say, humans and bears.

How much lineage sorting would you expect from human populations that have been  separated for, say 100,000 years? 250,000 years?  Close to zero.  Hardly any. So, can you use the fact that  human gene genealogies do not match apparent human population splits to prove those splits never happened – that there  had always been lots of gene flow between far-distant populations, say between sub-Saharan Africa and Siberia?  Presumably using magic carpets?

You could, but only if you were a bird-brain.

 

 

 

 

Gay genes

A recent paper discussed the results of a GWAS study of same -sex behavior, based on data from the UK Biobank. The data wasn’t ideal – it was based on self-reporting, and asked whether the respondent had ever had a same-sex experience, rather than trying to detect orientation… But that’s life.

They found two SNPs that influenced both male and female homosexuality, two that affected males only, one that affected females only. All had very small but statistically significant effects.

If we had lots of SNPs, we could look for trends ( are they predominantly expressed on certain tissues or processes ?)  but with only 4 in males  and 3 in females, not really possible. It does look as if the genetic architecture of homosexuality only partially overlaps between the sexes ( 0.63) – usually, say for things like height or iq, the overlap is close to 100%.  So the genetics, like every observation, suggests that male and female homosexuality are qualitatively different. Yet the degree of shared genetic influence is also interesting – I don’t think it would be predicted by most strategy models.

The fraction of the variance influenced by these few SNPs is low, less than 1%. The contribution of all common SNPs is larger, estimated to be between 8% and 25%.  Still small compared to traits like height and IQ, but then we knew that  the heritability of homosexuality is not terribly high, from twin studies and such – political views are more heritable.

So gene influence homosexuality, but then they influence everything.  Does it look as if the key causal link ( assuming that there is one) is genetic?  No, but then we knew that already, from high discordance for homosexuality in MZ twins.

Most interesting to me were the genetic correlations between same-sex behavior and various other traits.

 

The genes correlated with male homosexuality are also correlated ( at a statistically significant level) with risk-taking, cannabis use,  schizophrenia, ADHD, major depressive disorder, loneliness, and number of sex partners. For female homosexuals, risk-taking, smoking, cannabis use, subjective well-being (-), schizophrenia, bipolar disorder, ADHD, major depressive disorder, loneliness, openness to experience, and number of sex partners.

Generally, the traits genetically correlated with homosexuality are bad things.  As far as I can see,  they look like noise, rather than any kind of genetic strategy.  Mostly, they accord with what we already knew about male and female homosexuals: both are significantly more likely to have psychiatric disorders, far more likely to use drugs.   The mental-illness association maybe looks stronger in lesbians.  The moderately-shared genetic architecture seems compatible with a noise model.

The strategy ideas never made sense ( in terms of a workable kin-selection, sexually antagonistic, or other genetic model) , but I think this study makes that clearer.

Find that homosexuality was genetically correlated with various kinds of unpleasantness was apparently an issue in the preparation and publication of this paper. The authors were at some pains  to avoid hurting the feelings of the gay community, since avoiding hurting  feelings is the royal road to Truth, as shown by Galileo and Darwin.

Non-replicating lawyers

If the replication crisis ever rears its head in the courtroom, there will be hell to pay.  Many kinds of routinely used evidence and  testimony are utter trash.

Wheel in the Sky

I noticed some idiot claiming that, in 1700, China and India were ” more  sophisticated scientifically than europe”.  Nonsense, of course. Hellenistic science was more advanced than that of India and China in 1700 ! Although it makes me wonder the extent to which they’re teaching false history of science and technology in schools today-  there’s apparently demand to blot out white guys from the story, which wouldn’t leave much.

Europe,  back then,  could be ridiculously sophisticated, at the highest levels.  There had been no simple, accurate way of determining longitude – important in navigation, but also in mapmaking.  For ships the best solution was  Harrison’s marine chronometer (insensitive to temperature variations), but before that,  astronomers used the Galilean moons of Jupiter as a natural clock in the sky – observing when the moons entered or exited the shadow of Jupiter.  Comparing this with predicted times in astronomical tables, you could determine the local time, and from that your longitude. Not the most practical method on shipboard, but quite useful on land.

Before this technique, estimates of longitude were crappy, and maps showed it.  Afterwards, maps stop looking like something drawn by a sozzled ten-year-old :

In the course of playing with this technique, the Danish astronomer  Ole Rømer noted some discrepancies in the timing of those eclipses – they were farther apart when Earth and Jupiter were moving away from each other,  closer together when the two planets were approaching each other.  From which he deduced that light had a finite speed, and calculated the approximate value.

 

 

 

 

Indian Love Call

Been a bad few days: apparently I’m in trouble with the IRS, the Feds are suspending my Social Security #, and there’s something wrong with my Microsoft account.

Speaking of, why is this phenom never addressed by Presidential candidates?