An evolutionary theory for the variability hypothesis

An elementary mathematical theory based on “selectivity” is proposed to address a question raised by Charles Darwin, namely, how one gender of a sexually dimorphic species might tend to evolve with greater variability than the other gender. Briefly, the theory says that if one sex is relatively selective then from one generation to the next, more variable subpopulations of the opposite sex will tend to prevail over those with lesser variability; and conversely, if a sex is relatively non-selective, then less variable subpopulations of the opposite sex will tend to prevail over those with greater variability. This theory makes no assumptions about differences in means between the sexes, nor does it presume that one sex is selective and the other non-selective. Two mathematical models are presented: a discrete-time one-step statistical model using normally distributed fitness values; and a continuous-time deterministic model using exponentially distributed fitness levels.

That is from a new paper by Theodore P. Hill, via Derek.  Here is some of the history behind the paper, which ended up being “spiked.”  And here is Andrew Gelman’s take.  Here are relevant emails to the dispute.


The other theory is that we are not the only ones evolving. Entirely separate colonies of mitochondria and organelles evolve quite simultaneously. The difference in male/female chromosomes supports separe host environments for these semi-independent life forms. They are part of the fittest, and the supported environment the fittest..

I saw this as it burst upon Twitter, and it looked very odd from the start. Gelman's is a good explainer of what it was all about.

It's almost like a Monty Python skit, but run more painful and slowly in real life.

Yeah, Gelman's take on it looks solid to me. This could be a case study of how an internet mob gets triggered.

Gelman's take is obviously disingenuous. Next, Gelman will argue that Summers was forced out because his haircut did not meet Harvard's personal grooming standards.

I disagree with Gelman, but ironically for the same reason as him.

The very first thing I noticed in the rebuttals also was that it is completely consistent with Hill. But while Gelman sees this as a hint towards a genuine misunderstanding, I'd read it as trying to reframe what happened, without being stupid and claiming things that'd be easy to disproof. Basically all they say is that yes, everything happened as Hill said, but our intentions were different! But talk is cheap. Whether they wanted it before or not, ofc they'd claim now they didn't. And they certainly didn't mind, which is quite telling in my opinion.

Also note that Wilkinson does not say anything about her ignoring the email from Hill asking about comments, which is no proof, but consistent with Hill being basically right, and his opposition just wanting to reframe the debate to make it not blow up too bad. It's also weird if it's really just a disagreement, so why did she avoid answering?

The important part in this conflict is imo how unusual the entire process ran. Accepting and then de-accepting is odd enough, and especially without either a chance to edit it nor any comments on how the paper can be improved, but actually publishing it and then making it vanish without ANY of the usual retraction process?

These both are highly unusual, and they happened to the SAME paper. This imo makes it being merely a honest disagreement very unlikely. That's almost like doing a sliding tackle in football, twice, against the same person, and then claiming you actually just slipped twice.

Exactly. Gelman explanation leaves us without an explanation for precisely the part of this entire episode which most requires explanation. Everyone, Hill included, agrees it would have been perfectly legit to publish the paper with a sidebar criticism but somehow the decision was made (twice!) to memory-hole the thing instead.

I also note Gelman's confusion as to why anybody would think anything about this paper was politically sensitive in the first place. Bless his little heart; it is such a puzzlement.

Yeah... I think Gelman's take on the facts is balanced, but he does seem a bit naive in his interpretation. He suggests that this is not evidence of tantrum throwing by the academic left (the right certainly throws its own tantrums... see reaction to Nike ad campaign). I wouldn't know how to interpret "half of the editors at a journal threatening to quit if the now-published paper is not unpublished immediately" other than it being a tantrum by the academic left.

This obviously goes well beyond mere tantrum throwing. According to Wilkinson and Farb they merely wanted to publish a critical sidebar (Wilkinson) and subject the paper to more expert vetting (Farb). All well and good, the core of the academic process is review, debate and discussion. On these grounds Gelman absolves them all of any fault.

But, in fact, there wasn't any critical sidebar and the further expert review didn't happen. Instead, the paper was sent down the wrongthought memory hole where it would have stayed but for Hill, Peterson and the others. No review, no debate, no discussion, no academic process. It would have vanished like a 1930s era ex-politburo member, without explanation, except that Hill has emeritus status.

I think what is concerning is that this isn't simply activists or consumers engaging in this kind of behavior, but academics with PhDs who hold themselves out to us as fair- and open-minded truthseekers with deep insight and analytical skills.

Of course, this isn't new in academia, Stephen Jay Gould and Richard Lewontin tried to pull roughly the same thing when E.O. Wilson started his work on Sociobiology, but at least they let him publish his work. I can imagine that in the 1850s, there were academic power struggles in Southern universities to force out professors who argued that black people were not inferior, and that in the early 20th century, when the academy was VERY conservative, anyone with Marxist or socialist leanings or who sought to investigate economic or social inequality of the poor and lower classes was blacklisted. Now, though, attempting to prevent publication is a transparent attempt to delegitimize the research by saying "See, this research couldn't even get published in a respectable journal, it's obviously garbage!" and maintain some Gramsci/Foucault intellectual hegemony as holders of true knowledge, and therefore power/knowledge.

Yep. I am slowly finding out that people who are perfectly fine on both personal and proffessional level quite often start behaving completely differently when they start thinking 'politically'. Moral principles are moved out of sight; because this is politics and in politics you are a (moral) rule-maker, not a rule-follower.

It seems to me that publishing their scientific criticism would be an obvious and (hopefully) effective defense for Wilkinson. The silence is rather unsettling.

Her silence is not complete. She posted this on her web page:

Not much in here, though. And the last sentence is ridiculous "Invoking purely mathematical arguments to explain scientific phenomena without serious engagement with science and data is an offense against both mathematics and science." To me, it is exactly what mathematicians do
"invoking purely mathematical arguments [...] with serious engagement with data" is exactly what mathematicians (and other scientists when they acts as mathematicians -- Einstein and his "thought experiments" for instance) have been doing for several millenia, and this activity has been a great factor in the advancement of all sciences and human knowledge.

On the other hand, I find Hill's paper quite weak (I have not read it entirely yet, though), and I think it fails as explaining the (obvious) fact of greater variance in certain variables in males than in females. But nevertheless, what happened to it is more than worrisome.

When mathematicians invoke purely mathematical arguments without engagement with data, they don't write papers purporting to explain the data? Rather they pursue the mathematical model for the intrinsic aesthetic, basically due to an underlying belief that the deepest aesthetic is usually inspired by nature?

Ah yes, the pearl-clutchers prevail by being outraged. Someone’s feelings might be hurt. Seeing what is going on in the universities, how can anyone be surprised?

That's one bad paper.

Not according to Gelman - '... and if you really tried to apply his model to data so, sure, the paper’s not perfect.'

'Not perfect' is his verdict of a paper that in his opinion likely provides nothing workable. Could make one suspect that Gelman is an economist (he isn't).

'But as a math paper that touches on an interesting topic, it is what it is, and I was surprised there’d be a campaign to suppress it.'

What it is apparently, in the eyes of someone who actually knows statistics (one reasonably assumes), 'I don’t think it would work at all.'

Not saying that 'suppress' is wrong in a broader Internet echo chamber context, but what word is correct to say that a paper which has a model that would not work at all deserves no further attention from anyone?

I only had to skim the paper to notice how bad it is. The model is obvious: picking from the upper side of a distribution will shift the distribution up, but the author then makes the weird assumption that up means more variable instead of fitter. It also ignores that both populations (sexes) should inherit variability and selectivity.

Nothing to see here folks, just another bad paper which was accepted and then rejected and then accepted and "printed" online and then mysteriously disappeared altogether without explanation. Happens all the time. And Summer's left Harvard to spend more time with his family. And these are not the droids you are looking for anyway.

They should have published it after accepting it, despite whatever controversy it caused. But it should have never been accepted, because it is very bad.

"But it should have never been accepted, because it is very bad." So you say. Even Gelman disagrees with you on that point but you are obviously entitled to your opinion. Conditionally.

'Even Gelman disagrees with you on that point but you are obviously entitled to your opinion.'

Did you read what Gelman actually wrote? Including this - 'Sex differences isn’t my area of research so I was just considering Hill’s paper as expressing a simple and somewhat interesting mathematical model. If I saw the same kind of model applied to voting (for example, using the binomial probability model to compute the probability of a decisive vote), I’d be screaming. I wouldn’t want a journal to publish such a model, and if it were published, I’d want to run some article alongside explaining why the model doesn’t make sense. Not that the math is wrong but that it doesn’t apply here. I’m not prepared to make that judgment one way or another for the Hill and Tabachnikov paper, but my guess is that that’s where the critics are coming from. It’s not about suppressing a politically offensive idea; it would be more, from their perspective, about not spreading a mistaken idea.'

This follows Gelman already saying that in his opinion concerning the paper's math, 'I don’t think it would work at all.'

Gelman is being polite, of course, and recognizing his own limits while basically saying he considers the actual paper to be worthless.

Though only conditionally, of course.

Yes, I read Gelman's apology for Orwellian/Maoist tactics in academic publishing. Even the portion of Gelman you quoted does not support a conclusion that Gelman thinks the paper is worthless.

The point, obviously, is that when presented with the opportunity of subjecting the paper and its ideas to discussion and, potentially, correction, Wilkinson and Farb (and Harvard and Google and others) decided that outright suppression was the better choice. And Gelman evidently approves (with appropriate commiseration express for the authors).

'Even the portion of Gelman you quoted does not support a conclusion that Gelman thinks the paper is worthless. '

Both portions, actually. Though fair enough - he thinks the math in the paper is basically worthless, and reserves judgment as this is not his area of research concerning the paper itself. Except for the math, which Gelman finds won't work at all.

Absolutely wrong. He doesn't think the math would work, not that the math is "basically worthless".

No offense, but you clearly didn't even understand the paper.

1. Tell me, what's the difference between selecting only the upper 20% or just the upper 80% ?

By your logic, it should be (almost) the same, since any kind of selectivity moves the distribution towards that

But actually, they are fundamentally different in a way: Variance. If the opposite sex is selecting the upper 20% means that if you're (probably) average, any kind of added variance is good for you, because otherwise you probably won't have a chance to mate at all.

On the other hand, if the other sex is selecting 80%, you want to minimize your variance compared to the rest of the population, because you're almost guaranteed to mate if you're just average.

2. The author doesn't just assume this. You can proof that this is what happens, and the author does so. If you had actually read the paper, this should be obvious.

3. No, it isn't ignored that both sexes should always inherit it because it's plainly false. Humans have sex-specific chromosomes after all. It could only be true if sex was purely enviromentally triggered.

TL;DR: This is one bad comment. I only had to skim it to notice how bad it is.

> 1. Tell me, what's the difference between selecting only the upper 20% or just the upper 80% ?

Any selectivity > 50% will make the male average move right and fatten the right tail, but it would necessarily thin the left tail. The result is a distribution where males are fitter on average than females and there is a higher proportion of very fit males.
If you assume that variance itself is inherit, then we should see that very tall fathers have either very tall or very sort sons.

This comment is just wrong. If some biological process associated with the Y-chromosome triggers variability in male intelligence the right distribution would be more fit in each generation but that does not imply that the succeeding generations would acquire only a favorable expression of the characteristic. The favorable characteristic will not thin the left tail because both tails are produced by the same process or trait. It is the variance which is "inherited", not the positive characteristic (even though the "inheritance" comes (mostly) via the right side of the distribution).

Obviously, sickle cell trait provides resistance to malaria but sickle cell anemia kills. For the population as a whole, the benefits of one counter-act the disadvantages of the other so that distribution of the trait and the anemia reach equilibrium. The anemia does not eliminate the trait and success of the trait implies continued existence of the anemia.

And this has nothing to do with whether "selection occurs" at any particular percentage of the distribution.

For some weird reason, you seem to assume you can only ever pass on one trait. This is wrong. If a trait enhances the mean, it can get passed on, but at the same time, if a trait enhances/suppresses variance, it can get passed on, too. So, the son of a tall man probably is tall on average AND has a higher-than-average variance in terms of tallness, too. And afaik based on our data, this is actually what we see. Note though that for this specific example, the advantages from very high size are quite limited with notable disadvantages, so there is probably even some selection pressure against getting too big. Traits which are strictly good should exhibit this behaviour even more strongly.

Okay, but then why do we still see greater variability amongst males than females? Wasn't that based on observations, rather than just some assumption they made to fit a model to?

Who knows? Empirical findings is not what the paper is about. At all.

You said the model was obvious, but it's not to me.

Agudelo, the greater variability (with same average) pattern for males than for females for a certain variable in a certain species or population has many well-known example, and do not need any more "empirical findings".

The most obvious one concerns the single most important variable for an evolutionist (the only variable that matters, from this (narrow) point of view): the number of kids you have. Consider the human species, for instance. Since everyone has 1 mother and 1 father, if the number of male and female is about the same (which is the natural state of affairs, which happens when you don't kill your daughters at birth), then the average number of kids is about the same for a man or for a women.
But the variance of the number of kids tends to be not the same at all, for biological reasons: men who have fathered mote than 1000 kids are known, while it is extremely extremely rare for a woman to be the
mother of more than 10 of 15 kids. Of course since the average of number of kids is the same for male and female, the fact that some men have lots of children means that many others have very few or none -- think of the master of a harem and all his eunuchs. Of course, genetically selected behaviors (as in many species of birds) or social and cultural institutions (for the humans) can counter this natural state of affair and more or less equalize the variability of the variable "number of kids" among males vs females.

What people are looking for is an evolutionist process to explain this difference in variance between males and females. The biological explanation gives the "how", not the "why". This is perhaps a vain quest -there may be no evolutionary "why".

'the number of kids you have'

No - it is the number of children you have that survive to reproduce themselves. That is not a minor quibble, by the way - as noted by primate studies showing the importance of having a grandmother involved in child raising too. 'According to a study published today in the journal Proceedings of the Royal Society B, the answer is grandmothers. “Grandmothering was the initial step toward making us who we are,” says senior author Kristen Hawkes, an anthropologist at the University of Utah. In 1997 Hawkes proposed the “grandmother hypothesis,” a theory that explains menopause by citing the under-appreciated evolutionary value of grandmothering. Hawkes says that grandmothering helped us to develop “a whole array of social capacities that are then the foundation for the evolution of other distinctly human traits, including pair bonding, bigger brains, learning new skills and our tendency for cooperation.”

The new study, which Hawkes conducted with mathematical biologist Peter Kim of the University of Sydney and Utah anthropologist James Coxworth, uses computer simulations to provide mathematical evidence for the grandmother hypothesis. '

'while it is extremely extremely rare for a woman to be the mother of more than 10 of 15 kids'

I believe you are European - I know a number of people who are the youngest of 11 or 16 children (one an in law, for that matter - none of these people are Mormon, it should be noted). It is considerably less rare in the U.S. than in Europe, though obviously, the term outlier is quite reasonable to use.

There are women who have no children too. It would be interesting to see the relative numbers, which I suspect are within a margin of error for the two sexes. And it isn't the number of kids you have that matters, but the number who survive to adulthood that counts, though we tend to forget that in this era of super low childhood mortality. Men who impregnate random women but don't help provide for their children's survival are nit following an optimum strategy for their genetic survival given the limitations of our species.

> but the author then makes the weird assumption that up means more variable instead of fitter

Higher mean and higher variance both increase the right tail.

> It also ignores that both populations (sexes) should inherit variability and selectivity.

Males and females are not genetically identical.

> Higher mean and higher variance both increase the right tail.

Which makes your population fitter, not more variable. Why is the left tail also increasing? The model is making that assumption that variance itself is inherited, so members of the right tale have children with higher variance. That makes no sense to me when applied to biological systems: sons of very tall parents tend to be taller than average, not very tall or very small.

> Males and females are not genetically identical.

Sure, but males don't inherit exclusively from males. Daughters of a tall father also tend to be tall.

> Which makes your population fitter, not more variable.

Females are selecting based on signals which indicate success in life (which involves a lot of randomness), not on genes directly.

"Better" genes move the distribution (of success in life) to the right, increasing the size of the right tail. But variance also increases the size of the right tail.

> Sure, but males don't inherit exclusively from males

Yes, we're talking about differences right now. The fact that we're mostly the same is totally irrelevant to the discussion.

Let me dummy it down a bit: Evolution favors males with penises and higher variance (depending on species). I presume you don't believe that males can't have penises without females also having them? Then why would you believe the same about higher variance?

> "Better" genes move the distribution (of success in life) to the right, increasing the size of the right tail. But variance also increases the size of the right tail.

Yes, but then you have to assume that variance itself is inheritable and that's not clearly not true. Very tall people don't have very short children.

> Very tall people don't have very short children.

They probably are inheriting gene(s) to increase height, but that's not relevant.

The existence of genes influence the mean of some attribute doesn't validate a claim that no genes influence the variance of any attributes, much less the claim that no genes influence the variance of life outcomes.

"They probably are inheriting gene(s) to increase height, but that's not relevant."

Yes, it is relevant - it is the whole point: that you can inherit a tendency to be tall or a tendency to be short, but not a tendecy be be tall or short.

Yes you can, if by tall or short you mean greater variance around a mean

The existence of X doesn't disprove the existence of Y.

"The model is making that assumption that variance itself is inherited, so members of the right tale have children with higher variance. That makes no sense to me when applied to biological systems: sons of very tall parents tend to be taller than average, not very tall or very small."

This isn't that hard. Yes, if the Y chromosome contains a factor which increases variance it creates both super-men and "less than super" men. Those on the left side of the distribution are not fitter and do not reproduce as well but the super-men are fitter able to reproduce with multiple partners. The sons of the super-men carry the gene for increased variance and, as a result, some are super-men (and some are not) but the super-men have a genetic advantage over BOTH their left distribution brothers and also over males without the increased variance gene. Google "neolithic y-chromosome bottleneck" and think about it.

"Sure, but males don't inherit exclusively from males. Daughters of a tall father also tend to be tall."

This one just begs the question.

Are those tall daughters as tall as their tall fathers? If males and females are not genetically identical perhaps there is something in non-identical portion of the genome which increases height (or variance) in males. To make your comment about tall daughters relevant to this discussion you must show that the male population and female population are indistinguishable in terms of height (after all, daughters of tall fathers also tend to be tall).

I thought the higher variance had something to do with the Y chromosome itself; ie, men get only one copy of a lot of genes, whereas women get two, which kind of get averaged together, phenotypically, which eliminates outlier traits. Maybe that's not right at all. I feel like someone who actually knew something about genetics could settle this in about three sentences.

I agree that the nature of the Y and X chromosomes must tend to increase variance in males of all species and that might be the mechanism or part of the mechanism which produces the variance in humans. I can think of many reasons why high male variance might be especially beneficial among humans in particular.

If it's a bad paper, why spike it? Why not let it go through standard academic review?

Of course, but, in fact, the paper actually did go through standard academic review. Twice. The real question is why not publish it with a sidebar criticism as Wilkinson claims she thought appropriate or submit it to a different (and more "expert") panel of reviewers for a more rigorous review as Farb claims he thought was necessary? It is one of those mysteries wrapped in enigmas thingies. The only thing we know for sure is that it had nothing whatsoever to do with the reasons Damore and Summers lost their jobs.

It definitely should have remained published... this doesn't detract from Javier's point of the paper being bad. The thing is... this is a "just so". I liked this bit from the conclusion: "The precise formal definitions and assumptions made here are clearly not applicable in real-life scenarios." Then why go through the exercise in the appendix to highlight evidence (almost all of which only touches on human psychological traits... if the model is so general, you'd expect the author to include evidence from other species) that is consistent with, inconsistent with, or mixed regarding the model.

No version of this paper (there are nine on the arXiv) went through standard academic review, much less twice. The Mathematical Intelligencer is not a research journal, and Hill himself says it was submitted to the "Viewpoint" section. Then he submitted it to NYJM, which is a pure math journal; an editor there emailed him an invitation to submit it on a Friday night, and sent back an embarrassingly low-quality referee report from an admitted non-expert that same Sunday morning (see pages 23-24 of the linked emails). There is no sense in which this counts as "standard academic review," especially given that NYJM is a pure math journal (i.e. one which does not publish papers claiming to model biological phenomena, regardless of quality) and not one of its editors has any expertise in mathematical biology or applied math.

If you have any reason to believe that the pre-publication reviews which were conducted of this paper were not "standard" or otherwise appropriate for the publications and "sections" at issue, I think you should tell us what those reasons are rather than hiding behind "embarrassingly low-quality" and similar garbage. If you are simply claiming that the reviews which were conducted would not have been appropriate for publication under different circumstances, I think your comment is dishonest or, at a minimum, very deceptive. I have no doubt that the pre-publication review of this paper would have been inadequate for publication in "Nature", but, obviously, it was not to be published in "Nature."

But, this, again, is beside the point. If you or Farb or others felt that the review was inadequate for publication the correct course of action would be to find appropriate expert reviewers and subject the thing to whatever level of expert review which was felt to be necessary. If this had been done we might have a better paper before us or, at least, it would be possible to identify precise shortcomings or errors in analysis which are being alleged against it. Instead, the paper was not rejected for specific, identifiable, reasons but suppressed for reasons which remain obscure. That is not how we are supposed to do things.

I have lots of reason to believe this, because there's a report (which was apparently produced over the course of a single Saturday) included in page 24 of the Hill emails. Here are some problems with it:

- The referee asserted that NYJM was an appropriate journal for this paper, even though the briefest glance at the journal and its editorial board by any competent mathematician would tell you that applied math papers do not belong there.

- A third of the review is about politics, which have no relevance in reviewing an alleged math paper for a pure math journal. If they are somehow relevant, that is again a clear sign that the paper does not belong in this venue.

- Another third is a series of extremely basic questions having only the barest connection to the actual math, and the editor adds that they need not be addressed anyway.

- According to people with actual knowledge of the field (, the paper completely ignores virtually the entire body of prior work on the VH and misquotes some of the few sources it does cite, and a good referee report would insist on fixing these problems but this referee is silent on the matter.

- Presumably this is because the referee does not know about any of this either, or about any real shortcomings of the paper: he or she admits that "this is not my field", and we have no evidence that any actual experts in the field ever produced an informed referee report. You can't argue that the paper was subjected to adequate peer review if this is the case.

As for why Farb or the other editors didn't solicit expert reviews before publication, apparently the NYJM process didn't require input from anyone other than the editor handling the submission. I don't think anyone has claimed that the rest of the board, aside from the editor in chief, was even aware of this paper's acceptance before it was published online. If so, this is a major failure of the journal's editorial process, as was the decision to remove the paper completely rather than publish a statement retracting it. But this seems to explain why the rest of the board didn't solicit any more reports or otherwise do anything to stop its acceptance.

So, in fact, you have absolutely nothing to support your original contention that the paper was not subjected to any academic review. Instead, you simply present your own opinion that since the review which actually appears in the cache of e-mails is flawed (according to you), it should not count as a review. As further proof that no review took place, you cite to a blog post which happens to mentions in passing that the standard two reviews were obtained by the NYJM editor.

Sit down. Think. Does your opinion that that the reviewers did not have appropriate expertise, or that the reviews were too concerned with politics, or that the paper did not adequately cite prior work in the field, or that the paper was flawed in other ways, do any of those things really show that the reviews DID NOT EXIST?

But, again, all of this is beside the point. You cannot defend the unexplained suppression of the paper with complaints about the process which led to the original publication or even with complaints about the quality of the paper. I am not aware of any reason to believe that the Wilkinsons, or Farb, or the NYJM board, or Gelman, or you are any more expert than reviewers selected by the NYJM editor and their is no reason at all to believe that their opinions of paper are less tainted by political considerations. Not at all.

I didn't say there was no review, I said it did not get review from an actual expert on the subject, which is the only sort of review that matters in a reputable journal. I also didn't try to defend the removal of the paper, and I explicitly said it was a failure of the editorial process.

As for the two alleged reviews, I would be surprised if Hill had them and yet left them out of his supporting documents, given the sheer volume of other things he included. If these reviews existed prior to publication, it would be very unusual for him not to have them. On the other hand, Farb has alleged that the rest of the board demanded to see these referee reports once the paper was published, and that it took three months for his fellow editor to produce them. What possible explanation could there be, other than that the editor solicited these reviews after the fact to cover up his malfeasance? (For that matter, Farb has also alleged that the reviewers were not experts, and he knows their identities whereas you and I do not.)

As for political considerations, it would have been just as embarrassing if an editor had snuck a paper on some liberal cause into the journal as well. You do not seem to be a mathematician, nor do you seem to have any idea what a legitimate research paper in pure mathematics looks like or how papers are refereed in math (and very few people outside of pure math do). But I cannot find any mathematician not directly involved in this affair who seriously thinks that this journal was an appropriate place for the Hill paper, or for any paper about biology or politics. It hurts the journal both to have a reputation as a place for cheap political stunts like this and to be known for unpublishing a paper, no matter how inappropriate that paper might be.

"That's one bad paper."

If that's true, then the Left shot themselves in the foot. A bad published paper would help prove their point. A twice rescinded paper helps prove the Rights point.

If the right didn't have bad papers, they wouldn't have any science at all.

Remember when global warming was supposed to be nothing but heat island effect?

LOL, I'm old enough to remember when geo-engineering the planet out of the Ice Age was a dream for the future. Now, we live at a time when the EPA has confirmed that the growing season is 10 weeks longer than it was during the 1970's.

Thank you for that confirmation. Of course you would not know the science, or lack of it, behind the ice age myth.


"Of course you would not know the science, or lack of it, behind the ice age myth."

So, in your mind, the Ice Age was a myth?

Looking at your link, it's specifically about Global cooling in the middle of the long the last century. Which was real, just not a long term trend.

For example, if you were to look at the growing season data from my link, you'll notice a peak in the growing season in the mid 1930's and then a significant decline till about 1970. Around 1970, you'll notice a steeper incline till the present day.

It's not a "Myth" that growing seasons declined for 40 years. That's what the data shows.

I think the main of that link is the groupthink in climate science is older than you think.

That's a good one too!

When is scientific consensus bad, when does it become groupthink? Why are the acceleration of gravity or Einstein's equations not bad?

Why, science becomes groupthink when it implies a response that might ruffle feathers on the left or right.

I jumped into this because JWatts so freely interprets a scientific question as being "right or left."

I will freely admit that people on the far-left and far-right both make that that mistake, filtering science first by where it fits in a political spectrum.

Don't do that. Do your honest science first, and then consider political responses if they do in fact seem necessary.

"I jumped into this because JWatts so freely interprets a scientific question as being "right or left.""

I didn't classify the scientific question as Left or Right in a vacuum, Andrew Gelman specifically brings the issue up in his linked commentary. Perhaps you should read the links before you jump to conclusions.

Another fine example of Twitter rants against the "academic Left" (not sure why the ranters qualify the culprits as "academic" or why they capitalize left, presumably because the ranters wish to give the impression that the "academic Left" is a threat to the "Ranter Mob on the Right"). On the subject of gender "variability", here's a good example of the female gender's willingness to take a fresh view of Amazon from an antitrust perspective rather than following the (male) consensus view:

In English, locative words used as collective nouns are common capitalized: the South, the Left, etc.

But will the Left rise again?

Such is life in Temer's Brazil.

It is sad to see how midern politics have become divisive and mean-spirited with Stormies, collusions, trade wars this and croocked that. I do not blame you if you are disilusiined and do not want to hear about it anymore.

I submit you, however, that some causes are above petty party lines. Our friends in Brazil are suffering Chinese interference in their upcoming presidential elections. Leftist foreign agents have been caught carrying millions of dollars in Brazil. If you think we can nkt allow our most reliable ally fall to Red China, please take a few moments to phone, mail or email your Representative or Senators. Ask them to support a Congressional investigation on foreign interference in Brazil elections and to support the pro-American candidate, Representative Captain Bolsonaro, an experienced centrist politician with a distinguished military carreer, against corrupt leftist candidate Mr. Haddad.
Thank you,
May God bless you all

Hey Mark, I presume you've already contacted your representative and senators. Who are they? How did they respond?

And what's your opinion about the Ford family?

Regardless of everything else, bizarre to see Gelman wanting Jordan Peterson and Steven Pinker to "explain what happened to their followers, and to apologize."

Had a generally positive impression of Gelman, so it's a bit sad to see this.

Gelman is great. No need to be sad, no one is 100% on the mark on everything.

Yes, Gelman is great, and yes, Peterson and Pinker apologizing is dumb, and yes, Gelman gets the wrong end of the stick here.

The idea that Pinker in particular is inciting online hate mobs is ridiculous. His followers are mostly alarmed at how he skirts the Rim of Deplorability these days.

And online hate mobs are composed of pretty much the powerless dregs by definition. I'm sure it's unsettling to be on the receiving end of this crap, but these mobs are ultimately all bark. People with real power avail themselves of other means.

"People with real power avail themselves of other means."

For example, by squelching the publication of research papers to which they object.

Yes, I agree, and good point about what may be systematic (wanton?) overestimation of the power of online mobs.

I understand the 'Greater Male Variability Hypothesis’. What I don't understand is why an hypothesis that my explain why there are more male Nobel laureates my ALSO explain why more there are more males in SV tech jobs.

The discussion on Quillete is based on the assumption Google engineers require the same level of genius of Nobel laureates. This is an interesting hypothesis but no one seems to be trying to prove it. CalTech genius population is 2K while Google employees are 85K without counting freelancers and subcontractors. I think SV developers overestimating something.

Because both are outliers relative to the population at large. Males are underrepresented at 100 IQ and overrepresented at 145 and 160 IQ (leaving aside the different gender preferences as another cause of observed reality).

Yeah, I think Axa's point may be that 110 is good enough to get you a software engineering job. And there may be incidence parity at 110 by sex. So Greater Male Variability might explain a lot but not too much about the Google thing.

I highly doubt that 110 is good enough to get you a software engineering job

Here's how I've thought of it for a long time. From your gene's standpoint:

Female genetic payoff is somewhat like a bond - it decreases with volatility.

Male genetic payoff is somewhat like an option - it increases with volatility.

Your genes control the volatility. They're likely to increase it in the males.

Note that this is a over-simplification and exaggeration. Our species is relatively monogamous and has relatively low sexual dimorphism.

"Briefly, the theory says that if one sex is relatively selective then from one generation to the next, more variable subpopulations of the opposite sex will tend to prevail over those with lesser variability; and conversely, if a sex is relatively non-selective, then less variable subpopulations of the opposite sex will tend to prevail over those with greater variability. "

At the first look, the opposite seems to make more sense: if individuals of the sex A are ready to mate with almost every individual of the sex B, and individuals of the sex B only mate with carefully selected individuals of the sex A, the new generations will descend from a diverse set of B-persons but only from a selected set of A-persons, meaning more diversity in genes inherited from B-ancestors than by A-ancestors; than, in the case of the genes who only are active in people from a specific sex, should be more phenotipical diversity in B-persons than in A-persons.

"why an hypothesis that my explain why there are more male Nobel laureates my ALSO explain why more there are more males in SV tech jobs."

Because it's a curve and not a step function.

Paper caused a frackas in Tao's comments as well:

Thanks. But comments are more interesting on average here than there.

This was in response to Dan above and "there" means "Tao's blog" while "here" means MR.

Why are academics such appalling writers? What in the world would possess an otherwise intelligent person to write "relatively selective" rather than "more selective," and, worse, "relatively non-selective" rather than "less selective?"

There is a major logical flaw in this paper:

It's basic argument is of the form:
1. If A, then B
2. B
3. Therefore A.

Which is an invalid inference regardless of how novel the proof of the "If A, then B" part is.

Isn't this how many not-very-testable theories are started, though?

If we have "A -> B" as in criminal trial - "A - X committed a crime, B - crime was committed, A -> B - a plausible explanation how X could have committed the crime). The 'A -> B' is a mandatory condition; not a sufficient, but still good information to have.

It seems to me that there will be attention shift soon; like it was in James Damore case; he claimed it is plausible that gender disparity is caused by other factors than discrimination, therefore Google shouldn't behave as if it was 100% certain that it is caused solely by discrimination. The critics claimed that he didn't prove that the disparity is not caused by discrimination; yet he never claimed that. They should have shown that it was implausible that it was caused by other factors.

Shouldn't it be GMVO - Greater Male Variability Observation? Darwin observed greater variability of males than females in a host of species and we observe it to this day in people across a host of metrics. Any hypothesis would be to *explain* the observation. The metrics are already there.

What does the Left have against Charles Darwin? The fastest male blue-footed booby flies faster than the fastest female blue-footed booby because the female boobies get discouraged by toxic masculinity?

Comments for this post are closed