II.

There are many reasons to think testosterone affects behavioral outcomes.  Just on semantics, high testosterone would be expected to correlate to virility, aggression, and leads in action movies.

But what is the effect of brief, intra-uterine testosterone?  Do people who were exposed to higher testosterone in the womb become/behave different/ly?

An example: fraternal twins.  Does the girl's stewing in the boy's testosterone make her a better athlete, President or serial killer? 

In an attempt to answer that, there's been considerable research on the effects of intrauterine  testosterone on later life outcomes.


I.  (This is how you construct a lie: don't answer the question that was asked, answer the question you want to answer.)


III.

An example, a somewhat famous study.  Researchers examined a group of financial traders:

(Introduction) We therefore formulated the hypothesis that higher prenatal testosterone exposure would improve a trader's performance.

(Discussion) The finding that a marker of prenatal testosterone levels predicts a trader's long-term profitability...

The success and longevity of traders exposed to high levels of prenatal androgens further suggests that financial markets may select for biological traits rather than rational expectations.

And from Time:

Earlier studies indicated that prenatal exposure to testosterone... increases a person's sensitivity to the effects of the hormone much later in life. The greater the exposure as a fetus, in other words, the higher the levels of confidence, vigilance or risk appetite triggered by testosterone in an adult.

It's not hard to see why financial traders exposed to testosterone might be better at trading.  And now you have to think about society: maybe there are real sex differences in performance in the workplace.  It's perfectly ok to select a lingerie model on the basis of femininity.  Is it-- should it-- okay to pick options traders the other way?  And how can we level the playing field for those with a slight biological disadvantage?


A.

First Principles: what do the authors want to be true?

None of the studies linking biology to behavior are about either the biology or the behavior, they are only about the link.    

The question that they are answering isn't "does environment matter more than genetics?"  It's a more subtle, sneaky, social-policy questionr: "since we now know that genetics isn't as deterministic as we hoped, is there something else that we can focus on which is  equally out of our control?  What about the goings-on in utero?   So that the environment factors matter only at that time, not later?  Then we can safely say that behavior is "innate" and out of our control, while still leaving us the door to intervene in people's lives for their benefit." 

No one in the behavioral sciences discovers something, and then constructs policy recommendations.  "We learned that people are like this, so..." 

It's the other way around.  The policies come first; the money is spent on the research that supports them.

In questions about evolutionary psychology and behavior, the question they want to answer is always of the form, "how is it not the individual's fault, but we can meddle anyway?"


IV.

Back to testosterone.  In order to tell if trading is related to that brief in utero burst of testosterone, we need a proxy: the ratio of the index finger to the ring finger.  Bigger ring finger (and smaller 2D:4D ratio) = more testosterone in utero.

There are many such studies, of very different behaviors: aggression, lesbianism, athleticism, success, risk appetite-- and they are all surprisingly robust, there really does appear to be some kind of link.  And it helps that the behaviors all have an intuitive connection to "masculinization"-- (which was the whole point of the testosterone.)
 
And the associations are just as revealing when they're absent.  A recent study found no association to ADHD: "These findings challenge the hypothesis that fetal testosterone exposure plays a prominent role in the aetiology of ADHD."  So it must be something else...


B.

The problem isn't the data, but the words.

None-- read it again, none-- of the studies found any link between the behavior and fetal testosterone.  All of them found the link between the behavior and finger lengths, which are proxies for fetal testosterone.

But what if finger ratios aren't actually proxies for testosterone?

"Then those studies are crap.  Another example of science overreaching.  All that research money wasted."

Oh, no, Murdock, it's much worse than that.  The studies are valid, the data are solid-- finger ratios do indeed correlate well to these behaviors-- but all of the inferences you've invented about them are wrong.

A recent article discovered that in birds, the correlation was between digit length and estrogen receptors.   If that turned out to be true for humans, what are you going to do with all the stories about "masculine"  traits?  You can't simply say, "oh, it doesn't have to do with testosterone after all."  You must now explain why it does have to do with estrogen.  Are these feminine behaviors?  What?

There are other studies which similarly find the testosterone/finger story to be suspect or even backwards. 

So the data aren't wrong-- they're right; they're just about something else.  The social implications of the studies-- the very point of doing the studies-- are wrong.  You can't dismiss the studies because they're still true-- you have to go back and explain how you got it backwards.

Anyone who had taken a moment to look at the whole hypothesis-- masculine--> testosterone--> finger lengths--> behaviors would have said, "there are way too many loose connections to take this seriously."  But no one would have taken you seriously.  "Science" is three dimensional:  "look at the stack of studies that find a relationship between testosterone and behaviors!"  No one questions the intervening proxy (digit span) because to do so is perceived to be unrigorous.  When you say, "I don't believe this testosterone link" they politely say, "look at the stack!" but if you say you don't buy the digit length, they roll their eyes: another amateur who doesn't understand how science is done.  They do this because there's no other argument to make.  "This is how we've been doing it for decades, and it's a quite satisfactory method."  Yeah. That's what they said about missionary, too. 

Because there are so many researchers, and so many in the public willing to run with it, and so much time in between, there's no one to point to as responsible.  You can't blame Cambridge University for the obviously preposterous notion that masculine traits make for better traders any more than you can blame the head of BP for the oil spill.  Both situations are your fault.  You wanted what they were offering, even though it was bad for you.

Either we're going to kill Iraqis, or we're going to kill ducks.  It's the world's one and only truth, the law of equivalent exchange.  For every barrel of oil, you need to replace it with a barrel of blood.

C. 

But they are to blame, because when they presented you with their products, they knew exactly what you were going to do with them.


V.

I'm telling you this not because I care about finger lengths, but because you are being corrupted.

The article doesn't even have to spell it out for you: they just have to write "there's a relationship to testosterone " and we'll make the cultural/social value judgments ourselves.   But they leave nothing to chance; thus Time Magazine.

That's not an unfortunate, unexpected by product of science-- it is the very point of it.  In order for you to obtain this knowledge, you have to lose some other knowledge of equivalent value.

Once it's happened, once you've allowed this into your brain, there is no escape, ever, any more than there is an escape from oil.  No matter what else they discover, you will always have the suspicion that trading-- and lesbianism and risk taking and hand eye coordination-- are masculine traits.

Until, of course,  new guys come in with a new story to tell.  "Thanks Dr. Kohut, we'll take it from here."


VI.


The science error of our generation is this: If A is strongly associated with B, and B is strongly associated with C, then A is strongly associated to C.

That's not just wrong, it is extremely wrong.  If that seems counterintuitive to you, then you are the problem.  Not in the way Robespierre was the problem, but in the way the French were the problem.   "Sounds about right to me.  And there's a guillotining at 6:94!"

It's not your fault, you weren't trained to understand this, indeed, you were trained specifically not to understand this.   "Let's look closely at the statistics" (not the words.)

Science in the service of social policy is all about giving you everything you need to lie to yourself.


D.

I repeat: I have enough rum to get through what's left of my life, but the rest of you should heed my warning: if you do not rein in your social scientists, your civilization is doomed.

---

http://twitter.com/thelastpsych