The Scientists Who Wrote 'Not Plausible' Didn't Believe It Themselves — On Paper, First

On March 17, 2020, five virologists published a paper in Nature Medicine with a sentence built to end a debate: "we do not believe that any type of laboratory-based scenario is plausible." Six weeks earlier, in private email, the same authors were nowhere close to that certainty. That's not a conspiracy theory anymore. It's a FOIA production and a sworn congressional transcript.

Start with the timeline, because the timeline is the whole story.

On January 31, 2020, Kristian Andersen — who would become the paper's lead author — emailed Anthony Fauci and Francis Collins that he and three colleagues "find the genome inconsistent with expectations from evolutionary theory," and that "some of the features (potentially) look engineered." That's not hedged speculation from a crank. That's the lead author of "not plausible," in writing, five weeks before publication, saying the opposite.

Two days later, Robert Garry — another eventual co-author — wrote that he "really can't think of a plausible natural scenario" for how the virus's furin cleavage site could have arisen naturally, given that it required inserting "exactly four amino acids, 12 nucleotides, and all have to be added at the exact same time." That line was read into the congressional record on July 11, 2023, at a hearing titled, un-subtly, "Investigating the Proximal Origin of a Cover-Up."

That same week, Jeremy Farrar — director of the Wellcome Trust, not a co-author but a central figure funneling these scientists' views to Fauci and Collins — put a number on it in an email to Collins: "On a spectrum if 0 is nature and 100 is release — I am honestly at 50!" In the same message, Farrar described Edward Holmes, another eventual co-author, as "60:40 lab side." That's Farrar's characterization of Holmes, not Holmes's own words — worth being precise about, because the distinction between what a man says about himself and what a colleague says about him is exactly the kind of thing that gets flattened in a viral thread.

Two and a half weeks after that, on February 20, Andrew Rambaut — the paper's third author — wrote in Slack that debating the paper's conclusions was "doing my brain in": "I literally swivel day by day thinking it is a lab escape or natural."

Then, on February 8, Andersen sent the line that tells you what happened between the doubt and the publication: "Our main work over the last couple of weeks has been focused on trying to disprove any type of lab theory."

I want to be careful about what that sentence does and doesn't establish. "Trying to disprove" a hypothesis is, in isolation, ordinary scientific method — you stress-test the theory you're worried is true. Andersen has made exactly that argument publicly, and it's not an unreasonable one on its face. What makes it worth a second look is the gap between the internal register (50-50, 60-40, "swivel day by day," "can't think of a plausible natural scenario") and the external register six weeks later (unqualified, no percentages, no daily swivel — "not plausible"). A paper is allowed to reach a firmer conclusion than its authors held in week one. It's harder to square five separate authors landing on identical, unhedged public certainty when their private math, weeks earlier, was closer to a coin flip.

Here's what the documents don't establish, and I'm not going to pretend otherwise. They don't show anyone was paid to reach a conclusion. FactCheck.org ran that specific claim down in March 2023 — the theory that NIH grant money to Andersen's lab was compensation for writing "not plausible" — and rated it false, because the funding decisions that mattered predate the February 1, 2020 call by months. Grant timelines and doubt timelines are two different documents, and only one of them supports the story people want to tell.

What the documents do establish, cleanly, is that "the science was settled" was never true of the scientists who settled it. Five people who privately rated lab origin anywhere from even odds to 60-40 published a paper six weeks later that offered readers none of that uncertainty. Whether that's normal scientific caution hardening into consensus, or something else, isn't something an email chain can resolve by itself. But the next time someone tells you the lab-leak question was closed by actual virologists who never doubted it — that's the sentence that doesn't survive the FOIA production.