A few years ago, my lab published a study comparing memory complaints across racial groups. We matched participants on age, IQ, socioeconomic status, depression, genetics — everything we’d been trained to match on. We ran them through the most sophisticated statistical machinery I knew. And we got a finding that felt righteous: proof that the assessments were biased. I was proud of it. My name was first author. It took me two more years to realize what we’d actually done. We’d stripped our participants of every variable that made them human beings in order to make them statistically comparable — and in doing so, we’d erased exactly the context we needed to understand what was happening. The paper told a clean story. The world it claimed to describe was not clean. And that just ain’t science. I’m telling you this because of something that happened recently: the SCORE project’s verdict on seven years of work spanning nearly 4,000 social-science papers. Among those subjected to direct replication, roughly half didn’t hold up. The finding surprised no one who’d been paying attention to the cracks — in neuroimaging, in psychology, in biomedicine — but it does something useful. It gives us a number to attach to a feeling that a lot of scientists have been carrying around privately for a very long time. The feeling is this: The machinery is producing confident answers that aren’t true. But it’s our machinery. We built it. That realization would be uncomfortable enough on its own. Arriving alongside the worst 15 months in the modern history of American science, it’s devastating. The National Institutes of Health shed thousands of staff through successive waves of layoffs. Thousands of grants were terminated. The new director announced that restored DEI-related grants, including health equity research, simply won’t be renewed. Early-stage investigator funding rates cratered from 26% to 19%. And a recent STAT survey of nearly 1,000 researchers found that more than two-thirds of health disparities scientists quietly shifted their work to align with federal priorities. Not because the science changed. Because the politics did. That last detail should keep you up at night. Not the cuts — those were devastating but legible, the kind of crisis scientists know how to name. What should terrify us is the silence. Journalists covering this crisis expected researchers to rise up, to use the considerable public confidence we still command — 77% of Americans have trust in scientists, according to Pew — and face down bureaucrats with an axe to grind. They’ve been waiting. Surely, they thought, a community of a quarter-million credentialed professionals whose entire job is to navigate uncertainty would have a plan. They don’t. Stand Up for Science organizer Colette Delawalla described the scientific ecosystem as “absolutely lacking on anything direct action.” There is no cavalry coming — and we should have known better than to wait for one. We are not damsels in distress. We are the people who are supposed to be riding those horses. Let’s sit with that for a moment, uncomfortably, because we’ve earned the discomfort. We have spent the past year pearl-clutching and moralizing, writing open letters and behaving exactly like adults do during a toddler’s tantrum: moving the remaining blocks out of reach and waiting for the kid to tucker himself out. What we have not done is anything resembling a plan. We’ve been so busy feeling righteous that we forgot to be resourceful. And the reason we couldn’t be resourceful is architectural. We built American science like a skyscraper — centralized, concentrated, dependent on a single foundation. More than half of university research budgets flow from one federal source. Indirect cost structures can be weaponized by a Friday-afternoon policy change. On April 24, the White House terminated all 22 members of the National Science Board in a single email, leaving the agency that funds American basic research with no board, no director, and no deputy. Career incentives reward novelty over rigor, ensuring the most promotable science is also the least replicable. Trickle-down science — much like trickle-down economics — hasn’t worked the way we hoped. Science built an ivory tower when we needed a garden. The tower is beautiful — I mean that. It contains real knowledge and hard-won discoveries. But a tower is vertical, singular, closed. A garden is horizontal, distributed, open. And a garden is resilient precisely because no single point of failure can kill the whole thing. So let’s stop mourning the tower and tend what’s already growing. States are stepping in: New York Cures has proposed $6 billion in state-funded medical research; California has a $23 billion bond measure heading to the 2026 ballot. Patient-led organizations are funding their own studies, making allocation decisions that NIH review panels never would. Portable technologies — including MRI scanners light enough to fit in a van — are bringing neuroscience out of elite medical centers and into schools, sports fields, and rural clinics. In communities from the Gullah-Geechee corridor in South Carolina to rural Kentucky, my own work is exploring models where participants don’t just contribute data — they help shape the research questions and hold scientists accountable for delivering results that matter locally. Nowhere is it written that science must occur only on college campuses or in the halls of academic medicine. Science happens in barbershops and living rooms, in basements and farmer’s fields. The wisdom we produce as a community is what knits the enterprise together — not the overhead rates, not the agency letterheads, and certainly not the approval of any particular administration. Practically, this means diversifying beyond federal grants through licensing, philanthropy, and local partnerships. It means mutual-aid networks among scientists, especially the early-career researchers the system trained but cannot currently protect. And it means learning from communities that have always known what it’s like when the infrastructure you depend on was never built with you in mind. The SCORE data that dropped is not a death certificate. It is a diagnosis that finally points toward a cure. Pieces of the old architecture are falling away. But science itself is inherently resilient — that is, after all, why it’s science. It, and we, are being given the rarest of opportunities: a chance to outgrow the walls we mistook for the whole structure. The ground is right beneath our feet. We have the seeds. Now plant. Jonathan Jackson, Ph.D., is a cognitive neuroscientist and clinical trialist who founded the CARE Research Center at Massachusetts General Hospital and Harvard Medical School. He is currently writing a book on the future of American science.
