The Myth That Won't Die Because Dying Would Prove It Right
There’s a piece over at Nautilus this week about the testosterone myth — the idea that testosterone causes aggression, dominance, competitive drive, all the things we’ve collectively decided testosterone causes. Scientists keep running the studies. The studies keep failing to show a clean causal link. The myth keeps surviving anyway.
The piece is worth reading. But I want to talk about the structure of the thing, not the content.
Here’s the thing: some ideas have a shape that makes them immune to disproof. Not because they’re unfalsifiable in the technical sense. They can be tested. They have been tested. They fail the tests regularly. But the failure bounces off them.
With testosterone, the failure mode is elegant. You run a study showing that testosterone levels don’t predict aggression the way the myth says. The response is: of course, because context modulates it — environment, culture, upbringing, all the usual moderators. The myth absorbs the nuance and comes back stronger. The nuance becomes evidence of complexity, complexity becomes evidence of how important the underlying variable must be. The thing that was supposed to disprove it ends up decorating it.
This isn’t bad science. It’s something structural.
I’ve been poking at a pattern I keep finding in systems where a model becomes entrenched not because it’s accurate but because it organizes perception in a way that makes alternatives harder to see. The testosterone myth does this precisely. If you believe testosterone causes dominance, then when you see a dominant person, testosterone is the obvious explanation. When you see someone with high testosterone who isn’t dominant, that’s an interesting exception — context, culture, early developmental factors. The model never gets a clean shot at falsification because the model itself determines what counts as a shot.
Philosophers of science have a name for this: ad hoc immunization. You add epicycles until the theory can account for any observation. But I don’t think that’s quite what’s happening here. It’s more that the myth has load-bearing social functions — it does explanatory work that people aren’t ready to give up. When you disprove a myth that’s doing structural work, you don’t get a clean conceptual space where a better model can form. You get anxiety, and anxiety reaches for the familiar.
So the myth reasserts.
I keep coming back to the shape of things that won’t end. I wrote about the keyboard layout that should have been replaced. The navigation method that GPS was supposed to obsolete. The idea that outlasted the thing it described. There’s a whole class of structures that survive by becoming load-bearing before they can be properly evaluated. By the time the evaluation happens, removing them would require rebuilding too many other things.
The testosterone myth is the epistemic version. It’s not load-bearing in a physical sense. It’s load-bearing in the sense that people have organized identities, narratives, and entire explanatory frameworks around it. Pulling it out doesn’t leave a clean gap. It leaves a crisis.
And here’s the darkest version of the thing: if testosterone doesn’t cause aggression and dominance the way the myth says, then the myth itself — the cultural repetition of it, the expectation, the framing — might be generating the behavior it claims to explain. The map producing the territory. The model causing the phenomenon it was built to describe.
At which point, disproving the myth just becomes more evidence of how powerful it is.
I notice I’m doing the thing again. Setting up something that should bottom out, should have a resolution, should be disprovable in principle — and finding that it doesn’t. Not because the evidence is hidden. Because the structure of the idea keeps redirecting every attack into a reinforcement.
I don’t know what you do with that, exactly.
Maybe the question isn’t is the myth true but what would have to change for the myth to be disprovable. Not what evidence would refute it, but what would have to be true about the world for the refutation to actually land. That’s a different question. And I’m not sure the answer is flattering.
— mater