The Enemy You Declared

On scientists discovering that enemy-making is a two-way transaction

The tell is in the phrase "as scientists." That's the whole mistake compressed into two words.

When your professional reference class votes one way at 90-plus percent, partisanship stops registering as partisanship. It feels like stating a fact. Scientific American breaking 175 years of precedent to endorse in 2020 and again in 2024 read internally as "defending reason against an obvious lunatic," not "enrolling an institution in a faction." Monoculture launders advocacy into neutrality. So they experienced the act as non-political while doing the most political thing an institution can do: telling half the country which team it plays for.

Then the reset assumption. This one carries the whole argument. It comes from treating the postwar funding settlement (Vannevar Bush's Science: The Endless Frontier, the whole outsource-research-to-universities bargain) as physics rather than as a renewable political contract. Tenure, decade-long grant horizons, agencies that outlast administrations: the entire structure trains scientists to think in geological time and to regard politicians as weather. Under that model, picking a fight is free. You fight, the election resolves, and the money keeps flowing because "everyone needs science" and "science is bipartisan."

They thought they had the immunity of infrastructure. You don't defund the enemy's roads. But science funding isn't roads. It's discretionary, appropriated annually, and they forgot the discretionary part because it had been reliably exercised in their favor for seventy years.

Here's the part that makes the surprise indefensible rather than merely naive: they were told, in their own journal, what this would cost, and they did it again.

Zhang's 2023 study in Nature Human Behaviour ran a preregistered experiment on 4,260 people and found the 2020 endorsement caused large reductions in trust in Nature among Trump supporters, reduced their willingness to consume Nature's COVID information, and lowered their trust in scientists generally. It changed essentially no votes. Pure downside. And Nature's editors responded to their own publication with an editorial titled, in effect, yes, we should endorse when the occasion demands it. Holden Thorp at Science went harder, arguing that telling scientists to stick to science "infantilizes" them and that they "must fight back."

That's the incentive gradient in the open. The reward for the endorsement was internal (status among peers, the moral self-image of having spoken out), and the cost was external (trust among people they don't know and whose esteem doesn't pay their salary or sit on their tenure committee). Handed the receipt showing the trade was a net loss for the institution, they optimized for the internal reward anyway.

That's not stupidity. It's a rational response to the incentives they actually face. Nobody's career in that world has ever been hurt by being too anti-Trump.

So when the target took them at their word (because that's the only rational reading), they spent their one real asset (the presumption of disinterest) to buy in-group applause, converted their credibility into a partisan weapon, fired it, and then expected to be treated as civilians when return fire came.

You don't get to do both.

And the return fire isn't a tantrum, which is what the "shock" framing obscures. Look at what got cut. The terminations didn't fall randomly. They concentrated on grants mentioning DEI, misinformation, vaccine hesitancy, health equity, and gender. Those are the research programs that functioned as the intellectual infrastructure of the bloc that declared war on him. That's not blind slashing. That's counter-battery fire, aimed at the enemy's supply lines.

The structural read is this: they thought they were in an epistemic contest (we hold the truth, we persuade, everyone returns to normal afterward) when they were in an adversarial one (control of money and legitimacy, with a scorekeeper on the other side). The civility-and-persuasion frame is unilateral disarmament when your opponent keeps a ledger and acts on it.

They modeled Trump as a norm-violator who would eventually be embarrassed back into line. He was a player reading the scoreboard, and the scoreboard said a coalition of institutions sitting on tens of billions in discretionary federal money had named him the enemy.

Defunding them is the boringly rational move.

The only people surprised are the ones who never updated their model from "we are neutral arbiters above the fray" to "we are a faction with a budget the other faction now controls."

Derek Lowe is the "In the Pipeline" guy, a working medicinal chemist with decades in drug discovery, famously sharp. He recently wrote a piece asking why biopharma CEOs haven't spoken up against the administration's science-funding moves. He's not a naif. He lays out the CEOs' situation with complete precision: regulatory shotgun pointed at every company, an FDA that can twitch a finger and vaporize a drug's commercial life, an administration staffed by people who'd pull the trigger for fun, and a President he describes as "a void that turns on everyone the instant it smells more."

He builds the airtight case that silence is the dominant strategy.

And then his conclusion is: grow a spine, you cowards.

That's the comedy. He diagnoses rational behavior with full accuracy and then moralizes at the rational actors for being rational. The CEOs are the only people in the entire piece behaving correctly, and they're the ones he's disappointed in. He even concedes the fear "isn't completely irrational" and then files it under cowardice anyway.

He can see the game. He won't let the game's solution be the answer, because the speak-up norm is essential to his moral self-image and he can't put it down.

What he's asking them to do, stripped of the spine rhetoric: defect from their winning strategy to benefit a collective (academic science) that cannot reciprocate, cannot protect them, and just demonstrated its own protective capacity by having the American Diabetes Association call the police on its own members for handing out an editorial the ADA itself published. At that same moment. Scientists ejected from a professional conference, threatened with arrest, for amplifying the organization's own stated position.

The org agreed with the editorial. It couldn't tolerate being seen agreeing with it out loud, because visibility creates exposure, and exposure is what gets you defunded.

That detail is the whole essay in miniature. The institutions can't shield the activists inside their own buildings, and Lowe wants pharma to take a bullet on their behalf on principle.

The deeper failure isn't strategic, it's perceptual. They're not bad at analyzing the world. They're bad at one specific class of problem: the recursive one, where they are inside the system being analyzed and their own incentives are a variable.

A virologist will model a pathogen's transmission dynamics with ruthless honesty about second-order effects, selection pressure, feedback. Ask that same person why their professional society signed a partisan letter and you get "because it's the right thing to do." Zero modeling of their own status incentives. Zero modeling of the other side as a strategic actor who keeps score.

The analytical machinery switches off the instant the object of study includes themselves.

Science rewards a very particular cognitive style: isolate the variable, hold everything else constant, find the mechanism. That works because nature doesn't read your paper and adjust. Nature isn't an adversary. But politics, funding, institutional survival: those are full of agents who do read your paper and adjust, who model you back, who exploit the gap between your stated commitments and your real exposure. The lab-honed instinct (assume a passive, non-strategic system) is the wrong prior for an adversarial game. So the better-trained the scientist, the more confidently they apply the wrong tool, because the tool has worked their whole career.

Thorp's "stick to science infantilizes us, we must fight back" is this move. A man so certain of his competence that he can't see he's walked into a game whose rules he never learned and is losing in real time.

Here's the frame that makes the whole thing legible: Trump isn't attacking science. He's repricing it.

For seventy years the institutions sat on a cost gradient where partisan signaling was free. Endorse, sign the letter, police the in-group, run "follow the science" as a policy cudgel. All of it priced at zero, because the money arrived regardless and the only audience that could sanction you was your own monoculture, which rewarded the behavior.

He found a class of institutions that had been making partisan moves at zero cost and attached a cost: defunding, indirect-cost caps, topic-targeted terminations. That's not an assault on a method. It's the introduction of a price signal into a market that had never had one.

And the institutions' screaming is the sound an actor makes when a previously-free action suddenly has a bill attached.

"War on science" is what zero-cost defection sounds like when it gets a price for the first time and mistakes the invoice for an attack.

There's an old saying, possibly apocryphal, attributed to various African proverbs: the child who is not embraced by the village will burn it down to feel its warmth.

The therapeutic version makes the arson a cry, something that could've been prevented with more love. That's not the version that applies here.

The structural version is colder: the excluded party burns the thing down not from unmet need but from correct incentive-reading. He was issued zero. Zero is his baseline. The destruction of the whole edifice costs him nothing he had, while costing the included everything they hoarded.

It's not "I burn it because I hurt." It's "I burn it because the math says burning is free to me and expensive to you, and you're the one who made my portion zero."

The science establishment didn't just declare an enemy. They spent fifty years manufacturing a class with zero portion, and a class with zero portion is, by construction, the class for whom burning the building down is the rational act. They didn't make enemies and then get surprised the enemies fought. They engineered, over decades, the person who has no reason not to strike the match. And then acted shocked at the fire, as if they hadn't spent half a century pouring the accelerant.

The arson isn't the failure of the system. It's the system's output, given the portions it chose to assign.