Hacker News Reader

Comment by thaumasiotes | Hacker News Reader

> even things like allergies and sexualities which one would expect to be more biological in origin.

It doesn't seem to be well-known, but concordance for homosexuality in identical twins is very low (if memory serves, around 40%) compared to concordance for virtually anything else you can measure.

This is pretty unsurprising looking at the biological fundamentals -- a quick naive approach would be to predict "genes will never code for homosexuality, because it is such a crippling failure for the organism" -- but it does tend to undermine the idea that homosexuality is biological in origin.

View on HN

> but it does tend to undermine the idea that homosexuality is biological in origin.

Not genetic and not biological are two different things. E.g. the differentiator might be hormonal influence during the fetal stage. See https://en.wikipedia.org/wiki/Epigenetics for a long list of things other than DNA that can affect how organisms develop.

Just because something isn’t 100% determined by genetics doesn’t mean genetics doesn’t play a heavy role still. A concordance of 40% is still really high compared to random chance by sampling from the general population, so it arguably supports biological causes for homosexuality.

How? If we dont have evidence for it should we just conclude that it exists because our maths does not say it doesn't.

Lets make witchcraft a part of medicine then. A statistically significant part of the human population belives in it. Lets get that into a 4 semester course in a 6 year medical degree.

But we can see that the role of genetics is miniscule compared to its role in any other trait, since identical twin concordance for homosexuality is so low compared to concordance for anything else.

Genetics affects everything including the environmental influences you are subject to. But your genes are trying as hard as they can to ensure you are heterosexual.

https://www.nature.com/articles/d41586-019-02585-6

25% of the variance in sexuality is genetic or heritable. Compared to approx. 80% for intelligence. A small genetic component, but it seems as if experiences are the major deciding factor for sexuality.

I’m very skeptical about intelligence being as high as 80%, how do they remove environmental factors it seems almost impossible.

By having lots and lots of good data.

80%+ is what the current corpus of scientific knowledge says. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4270739/ Here's a somewhat good primer on the topic, but the wikipedia page on heritability of intelligence (and it's sources) is also somewhat good.

There is no wikipedia page on heritability of intelligence, only heritability of IQ. It makes me consider whether the rest of what you've said is actually incorrect since you made such a conflation.

Intelligence is a notoriously hard thing to define and measure, and I think it's more easily predicted by environmental factors than genetics. The idea that smart people breed and have smart children and that's why some people are smart and others are dumb, is more of a cop-out, since predictors of success that tests like IQ try to explain are more accurately explained by socio-economic factors.

Basically this has no real world implications. I realise this isn't the argument you were trying to make necessarily, you were just pointing out a perceived 'fact', but it lacks a lot of nuance that could lead to dangerous assumptions about the real world, especially since you related it to sexuality.

If people demanded as much rigor for other subfields of psychology as they require for psychometrics of intelligence, the whole word would simply get rid of all psychology departments.

People have no clue how much more rigorous these studies than anything else in the field and adjacent fields.

Of course I demand rigor in a subfield that exists almost entirely to inform policies like segregation, eugenics, welfare, education, etc, and I find studies based on IQ tests and the g factor to be very much inadequate at being socially relevant.

> Did you know that there is also a strong correlation of intelligence between parents and adopted children? Kind of puts the OP's claim to bed a bit.

Only in young children. The correlation vanishes as the children age. You were wise to delete this remark from your comment.

> Don't think I'm not aware of the history and research of this area.

Heh.

You replied only to the parts i deleted, unfortunately. Neither of your responses even dispute my points.

> Neither of your responses even dispute my points.

First you try to insinuate that IQ is not related to intelligence and even if it were intelligence has not practical application in the "real world".

Your first point was disputed by the comments related to how rigorously this subject is studied and documented, due to it being a third-rail area of academia.

I am going to dispute your second point myself. That rigorous study and documentation associates IQ with intelligence and intelligence with 'g' which, although not precisely defined, is associated with (at least) higher problem-solving ability.

So now your argument reduces to "higher problem-solving ability has no practical application in the 'real world'".

And so I will just dispute that statement out of hand because problem-solving ability is what has allowed humans to become the dominant animal on Earth.

My point is that you seem to not understand how rigorous is the field compared to academia as a whole and especially those fields that shape policy nowadays. In fact, intelligence research was always on the forefront of using cutting edge methodology and some tools that are used in psychology were invented specifically to research intelligence. Spearman invented factor analysis to measure cognitive performance, to this day this test is still used in behavior research.

If you think that IQ studies are inadequate, don’t even look at other fields that study humans (aside maybe from double blind trials in medicine), you are going to have a heart attack.

By your standards we need to ignore 90% of academia and solely rely on physics research to guide our policy.

Factor analysis has its own set of problems. Watch this at least past 30 or 40 minutes. https://youtu.be/UBc7qBS1Ujo

I have completed over 30 ECTS credits studying econometrics in graduate school, I think I have enough grasp of statistics and factor analysis to not get schooled by some breadtuber who produces infotainment content for teenagers.

Very rational of you.

That was gratuitously obnoxious.

Presumably the only good data can come from siblings who were raised in environments dramatically different from each other and the ended up with similar IQs? It must be extremely hard to know what factors in the environment might contribute to IQ... saying the data is great sure, but I need examples of how environmental factors are removed.

Go look up eg the sources on the wiki article, and you can see for yourself how they controlled for nurture.

I looked into it and the actual answer as with everything it’s extremely complicated and you’re original pithy 80% of IQ is heritable seems to be dependent on how much culture is involved in the testing methodology... https://blogs.scientificamerican.com/beautiful-minds/the-her...

I think the idea that we can control properly is extremely unlikely and IQ will turn out to be a lot difficult to extract from learning/society/environment than you might think, especially as findings with epigenetics and intelligence start to be shown. https://www.europeanscientist.com/en/research/epigenetics-an...

Your first link is a blogpost about someone's doctoral dissertation, and your second link is even more of a stretch (the title of the paper ends with a question mark). That is fringe research, and not the established position. You get what you search for.

As for cultural loading in intelligence tests, maybe you could expand on that? How do you measure how much culture was involved in the methodology, and how it affected the results of the tests?

You're right to be suspicious. The studies that this figure are drawn from are generally twin studies based in wealthy Western nations. Heritability tends to increase with wealth, economic equality, and ethnic homogeneity, all of which are unfortunately selected for in the formation of these trials. This paradigm is thought to be a gold standard, as it reduces independent variables, but it fails to capture the effects of intergroup environmental disparities in a profound way.

The range of SAT scores at a US prep academy populated with the progeny of relatively closely related families will be substantially determined by heritable differences between individuals. Outside of that, it's disingenuous and irresponsible to throw out that 80% figure as a general measure of intelligence heritability. In fact, not even the studies say that: 80% is the measure for middle-aged and senior adults. For everyone else, even their measures are substantially lower.

>The range of SAT scores at a US prep academy populated with the progeny of relatively closely related families will be substantially determined by heritable differences between individuals.

Can you explain that more in depth? To me that is counterintuitive. If there is a group of closely related people, I would expect their differences to be due to environment, rather than genetics, because they have similar genetics.

The opposite. They're closely related compared to the rest of the world, but still come from different families, while living in substantially similar environments, which in many way have "maxed out" advantageous factors (health, stability, stimulation, etc.).

Remember that the heritable component of intelligence is polygenetic; it'd not determined by the presence of a single SNP, and two people could have similar intelligence originated in very different genetic profiles. People who are relatively closely related will have similar profiles that interact with the similar or shared environments in similar ways, meaning that the effect of small genetic differences will be more pronounced. Someone with the same "potential" for IQ, but from a different set of intelligence-related genes, interacting with the same environment might be said to have had the heritable component of their IQ become less efficacious.

For example, nutrition is a component in IQ. Everyone at this school is able to drink milk and benefit from the nutrition value-add. Someone less related to the other students also attends the school; they're lactose-intolerant, but benefit from eating certain fruits native to their homeland. The school doesn't serve such fruits, but it does serve milk. The student subsequently tests at a lower IQ. I would say at this point that variation between the LP students' intelligence is likely to be more related to genetics than is the variation between their IQ and this lone student's, which has a more substantial environmental component.

There are theories for how a low percentage of homosexuality might be selected for.

I have heard at least of the idea that a non reproducing sibling can help you raise your own offspring, who will have mostly the same genes. This is corroborated if I recall correctly by findings that each new child in a family has more chance to be homosexual than their already born siblings.

There is also the idea that there might be genes that are most of the time favorable to reproduction, but once in a while (through crossing or mutations) result in homosexuality.

> This is corroborated if I recall correctly by findings that each new child in a family has more chance to be homosexual than their already born siblings.

Very radical approach, seemingly against the grain of what we observe in the wild. It's such a leap of logic, you may very well have been talking about how santa claus' reindeer achieve thrust for their impressive levitation.

I'm not sure what you mean. Humans have been social animals for a very long time, so it stands to reason that they would be genetically different from non humans, although i'd expect similar results in other social animals.

An example of similar results in other social animals are worker/soldier ants/bees/termites. They don't reproduce, but by contributing to the survival of their hive of relatives, they still end up spreading their genes.

Are they homosexual or asexual? Let's not confuse the two. They "cant" reproduce is quite a leap away from "can produce but won't" because as far as I am aware homosexuality has nothing to do with sexual reproduction and the required 'machinery'.

Unless you have some solid papers or anecdotal experiences that say otherwise I'm willing to change my stance.

> There are theories for how a low percentage of homosexuality might be selected for.

My initial response was to the above statement which to my eyes presupposes homosexuality to be genetically expressed. Which has nothing to do with social/environmental factors.

Why would subsequent children get different gene mutations? I don't think it's a thing, it would be like saying that every new child tends to have thicker eyebrows or something.

Of course it is possible that statistically each new child in a family has higher probability to be homosexual for environmental reasons, but for genetic reasons? I don't see it (could be wrong, not an expert at all)

It is weird: https://en.m.wikipedia.org/wiki/Fraternal_birth_order_and_ma...

Very interesting, thank you for the link!

Edit after reading: apparently it's well studied and proven that each subsequent _male_ child has more chance to be homosexual and that "the fraternal birth order effect operates through a biological mechanism during prenatal life, not during childhood or adolescence"

>genes will never code for homosexuality, because it is such a crippling failure for the organism

It's not necessarily a failure for the gene just because the organism doesn't reproduce. Genes extend beyond the organism to the family.

It would be more surprising that it persisted in less social species.

> It would be more surprising that it persisted in less social species.

It does.

How about that male birthing order study that they've been trying to disprove for years?

[1] the odds of having a gay son increase from approximately 2% for the first born son, to 3% for the second, 5% for the third...

[1] https://en.m.wikipedia.org/wiki/Fraternal_birth_order_and_ma...

Homosexuality is potentially bad for the perpetuation of an individual's genetic line (sexuality is a spectrum, and a percentage if not a majority of homosexual men and women would be open to procreation given a suitable social framework), but for communities, has the potential to be a net positive for survival; for example, it may reduce competition for procreative mating partners, and provide childless adults who can share rearing responsibilities.

In fact, one could say that the benefits of such an alternate sexuality were so recognized that some societies sought to produce it artificially through the creation of eunuchs.

Well no. I’m stupidly in love with my male son and I assume homo-sexuality is just a behavioral generalization into sexuality of same-sex “aggression de-triggers” and “attachment triggers”. It’s underpinnings are built into successful parenting (don’t murder your pups.)

Besides, close relatives in the animal kingdom (Bonobo) use significant amounts of sexual behavior in their sociality, including same-sex interactions. Unfortunately we evolved from chimpanzees, who have strongly abusive and hierarchical social structure. Whence, nasty homophobic comments.

We didn’t evolve from chimpanzees. Bonobos, humans and chimpanzees have a common ancestor. While chimpanzees are violent psychopaths, bonobos are peace-loving hippies who have sex all day. Humans seem to have a mix of both these traits.

It’s quite astonishing how much bonobos look and act like humans when they walk upright or when they play with their babies.

Sorry, I've only taken statistics 101. What's "concordance" in this context?

Correlation but with categories rather than continuous variables, for example as used in https://en.wikipedia.org/wiki/Kendall_rank_correlation_coeff... -- so "has the same sexuality as their twin" would be concordant, or I think what the parent comment is referring to is more specifically "is homosexual and their twin is too". In the general population, you'd expect a concordance of at most 10% (a high-ish estimate of the prevalence of homosexuality in the population), though of course that's conditional on the first person in the comparison being homosexual, otherwise it'd be much higher.

> such a crippling failure for the organism

This is just a homophobic mischaracterisation of sexuality.

I recognise and don't appreciate how much of an effort you made to fit this homophobia into your point.

Also, 40% heritability is informally considered to be a high genetic component, not "very low", and e.g. heritability of eyebrow thickness is 37%.

I'm very certain the parent meant it in an evolutionary sense that the individual will not pass on their genes to offspring - nothing to do with homophobia.

It seems like a very individual natural selection-based idea.

I’ve personally always wondered if there could be tribe-level selection for tribes with some non-heterosexual members.

But I also have personal experience, and I’ve met others who do as well, with life experiences that make me wonder if they could somehow affect sexuality. I know it might be presumptuous of me to consider the idea, but it’s only just an idea—one I usually keep to myself.

It seems obvious to me that your childhood and especially adolescent experiences would impact what you find sexually attractive as an adult. Is this considered a radical position?

It’s a sensitive subject for some, I think.

The parent could have made the claim (that you believe they meant) without saying "crippling failure of an organism", and I would still think it was mildly homophobic:

- there's literally nothing to stop a homosexual person from biologically reproducing,

- it assumes that the only value of an organism is in its Darwinian fitness (its ability to reproduce), which is pretty gross. I reject this assumption.

The parent appears to be applying this heuristic of value only so far as it implies a moral failure of homosexuality: it's not a typical measure of value. Humans value other animals more highly than this, without regard to whether they have reproduced.

Of course genes don’t code for homosexuality. Just look at the percentage of gay people in prison.

Reply on news.ycombinator.com