The Public Record

Host: Picture the largest, best-funded, most secretive machine for finding things out that has ever existed. Many agencies. Satellites. Code-breakers. Sources in places you will never read about. Vast and well-funded. That machine is the United States Intelligence Community, and in the early days of the pandemic it was handed one question. Where did COVID-19 come from?

Host: One question. The whole apparatus pointed at it. Three years of work. And the answer that came back was not one answer. It was four.

Host: Sit with that, because it really is strange. This was the single most consequential question of a generation, a question that emptied cities, closed schools, divided families, reshaped how billions of people lived. The kind of question a country pours everything it has into answering. And it did. What came back was not "we found it." It was not even "we think we found it." It was: we, the government, cannot agree among ourselves, and here is exactly how we are divided.

Host: Now, that part, that they disagreed, is not the strange part. Intelligence is rarely unanimous; if it were, you would not need a whole community of agencies. Smart, careful people read the same file and walk away with different readings every day. The strange thing is what they did next. They wrote the disagreement down. And they handed it to the public.

Host: Think about how rare that is. Most governments work very hard to resolve their arguments before they speak. They go behind a closed door, they fight, and they come out with one face and one voice. That is what we are trained to expect on a question this big: a verdict, delivered with the full weight of the seal behind it. This time, the most powerful intelligence community on earth came out and said, in effect: we looked at the same classified file, and we could not agree, and we are going to tell you that.

Host: And it did it in the middle of the loudest argument the country was having with itself. People had buried family members. Whole industries had been shut and reopened and shut again. The question of where the virus came from had stopped being a scientific question and become a tribal one, a thing you could be on a side about, a thing that told a stranger which team you were on before you'd finished the sentence. Into that, the people whose entire job is to know things first looked at the classified file and did the one thing nobody on either side wanted. They refused to take a side. They said: we looked, and we couldn't agree.

Host: It is one thing for a government to be unsure. It is another thing entirely for it to publish the fact that it could not agree with itself.

Host: Here is how they said it, in the unclassified summary they released in August of 2021.

Document voice: After examining all available intelligence reporting and other information, though, the IC remains divided on the most likely origin of COVID-19.

Host: Divided. Their word, in a document they handed to the public. Not "still investigating." Not "we'll get back to you." Divided.

Host: That one sentence is the whole episode. This is the story of a government that looked at one classified record and reached four different answers, and then said so, out loud, in print.

Host: From Fauci's Lies: The Public Record. Episode five. The Split.

Host: Before we can read this record honestly, we have to learn one word. Because without it, the whole story gets misheard, and it got misheard, badly, by a lot of people. The word is confidence.

Host: When an intelligence agency reaches a judgment, it doesn't just state it. It grades it. It attaches a confidence level, high, moderate, or low. And here is the thing a newcomer has to understand, because it is the opposite of how the word sounds. A confidence level is not a measure of how strongly the analysts feel. It is a measure of how good the evidence is.

Host: "High confidence" means the sourcing is solid, corroborated, multiple independent streams pointing the same way. "Low confidence" does not mean someone is guessing, and it does not mean they're timid. It means the underlying evidence is thin. Fragmentary. Contested. Open to more than one honest reading.

Host: Here's an everyday way to feel the difference. Imagine two experienced doctors looking at the same blurry, incomplete scan. One says, "I think it's probably this." The other says, "I think it's probably that." And both add, in the same breath, "but the image isn't clear enough to be sure." They're not fighting. They're not incompetent. They're being honest about a bad picture, and still doing their job, which is to give you their best read with the warning attached. That warning is the confidence level.

Host: So follow this, because it's the hinge of the entire episode. Imagine two agencies looking at the exact same record. One concludes natural exposure. The other concludes a lab-associated incident. Opposite answers. And both, honestly, mark their answer "low confidence." That is not a contradiction, and it is not a failure. Both are telling you the same true thing about the evidence, that it is too thin to carry a firm conclusion, and then making the best call they can on top of it. The low-confidence label is the honesty.

Host: And this matters enormously for this particular story, because out in public, "low confidence" got translated. It got turned into "the agencies secretly think it was a lab leak but won't say it," or "the agencies know it was natural and the lab people are cranks." Both translations make the same mistake. They treat a statement about the evidence as if it were a wink, as if the analysts had a real answer they were holding back behind a cautious word. They didn't. The cautious word was the answer.

Host: Let me hand you the rest of the alphabet quickly, so the names don't blur when they arrive. The Intelligence Community, the IC, is the group of U.S. federal agencies that collect and analyze secret information for the government. Sitting on top of it is ODNI, the Office of the Director of National Intelligence, the office that speaks for the whole community. Both documents at the center of this episode were released in ODNI's name. And when the IC counts its own members, it uses a neutral word: an IC element, just one of the member agencies. You'll hear "four IC elements" before you ever hear a single name, and there's a reason for that, which we'll get to.

Host: First, what they all agreed on before they split. Back to that 2021 summary.

Document voice: All agencies assess that two hypotheses are plausible: natural exposure to an infected animal and a laboratory-associated incident.

Host: All agencies. Both live, both on the table. The argument was never "is one of these crazy." It was about which of two serious, plausible explanations the thin evidence pointed to. And the summary was unusually candid about why they differed.

Document voice: Variations in analytic views largely stem from differences in how agencies weigh intelligence reporting and scientific publications, and intelligence and scientific gaps.

Host: Read that slowly. The agencies were not looking at different facts. They were looking at the same facts and weighing them differently. Same intercepts, same scientific papers, same maddening blank spaces where the decisive data should be. The split was not a fight about what the evidence was. It was a disagreement about what the evidence was worth.

Host: So the question this episode follows is not which side was right. It's the one the documents themselves left open: why did the split hold, for years, across two presidencies, with the full machine working on it?

Host: August 2021. The IC drew its split for the public in a very particular way. It told you exactly how the room divided, the count, but it would not tell you who sat where. Numbers, no names. Listen to the map.

Document voice: Four IC elements and the National Intelligence Council assess with low confidence that the initial SARS-CoV-2 infection was most likely caused by natural exposure to an animal infected with it or a close progenitor virus.

Host: That's the larger group, four agencies plus the National Intelligence Council, ODNI's senior in-house analytic body, the panel that produces the government's big-picture judgments. Five voices. Their read: natural exposure, most likely. Their confidence: low. Now the other pole.

Document voice: One IC element assesses with moderate confidence that the first human infection with SARS-CoV-2 most likely was the result of a laboratory-associated incident, probably involving experimentation, animal handling, or sampling by the Wuhan Institute of Virology.

Host: One element. Moderate confidence. A laboratory-associated incident, and notice the careful phrasing: experimentation, animal handling, or sampling. An accident around ordinary research. Hold that phrase; it is going to matter later, because it is a narrower and a different claim than the ones that filled the airwaves. And then, in the middle, the agencies that would not pick.

Document voice: Analysts at three IC elements remain unable to coalesce around either explanation.

Host: Three more, unable to land at all. Not absent. Not silent. Present in the room, looking at the same record, declining to plant a flag.

Host: So picture that as a room. A long table, the same briefing book open in front of every seat. Same pages, same gaps. Four agencies and a council leaning one way at low confidence. One leaning the other way at moderate. Three who looked at the identical file and could not converge. That is the image to hold for the rest of this hour: not a leak, not a smoking gun, but a table of careful people who read one document and split.

Host: Now here's where the vocabulary from a minute ago earns its keep. The lone lab element marked itself moderate. The larger natural group marked itself low. And it's tempting to do arithmetic with that, moderate sounds bigger than low, so doesn't the lab side win on points? Resist that. Hard. "Low" and "moderate" are not scores in a contest. They are each agency's own warning label about the quality of the evidence under its own conclusion. One agency saying "moderate" is not beating five saying "low." It's one agency telling you its slice of the record was a little firmer, and still warning you not to over-read it. Two honest reads of one thin file.

Host: And notice what the documents have not done yet. They have not told you which agency holds the lab view. In 2021, that one element is un-named. Keep that parked exactly there, in 2021, with no name on it. The next document is going to fill in the names. But for that, you have to wait two years, and open a different document.

Host: June 2023. Almost two years later. A new document, and it matters that it's a different document. Congress had passed a law, the COVID-19 Origin Act, ordering the Director of National Intelligence to declassify what the government knew about possible links between the Wuhan Institute of Virology and the pandemic. That law is the reason this report exists, and is public.

Host: And it matters what the report is for. It is scoped to that one institute, and it says of itself, in plain text, that it "does not address the merits of the two most likely pandemic origins hypotheses." It is not a fresh verdict. It is the standing assessment, re-stated, and this time, with the names filled in.

Host: One caution before we read it, and it is the most important caution in the episode. This 2023 report names the agencies. It does not restate each agency's confidence level in that passage. So the "low" and the "moderate" from 2021 do not travel across to these names. Two different documents, two years apart, doing two different jobs: one gives you confidence without names, the other gives you names without confidence. We keep them apart, because the documents do. Here is the named split.

Document voice: The Department of Energy and the Federal Bureau of Investigation assess that a laboratory-associated incident was the most likely cause of the first human infection with SARS-CoV-2, although for different reasons.

Document voice: The National Intelligence Council and four other IC agencies assess that the initial human infection with SARS-CoV-2 most likely was caused by natural exposure to an infected animal that carried SARS-CoV-2 or a close progenitor.

Host: There it is, with names on it. On the lab side: the Department of Energy and the FBI. The Department of Energy, which most people know for nuclear and energy work, but which also runs the national laboratories and a small intelligence arm with real scientific depth. That's why an "energy" department gets a vote on the origin of a virus: it has scientists who understand the work. And the FBI, the bureau most people know as domestic law enforcement, which also does intelligence work, and tends to reason like an investigator. Two agencies, lab-associated incident, most likely.

Host: And listen to the four words the report is careful to include, "although for different reasons." The Department of Energy and the FBI landed on the same side, but they did not get there the same way, and the document goes out of its way to say so. That's not a footnote. That's the IC refusing to manufacture more agreement than it actually had.

Host: On the natural side: the National Intelligence Council and four other agencies. Natural exposure, most likely. And then the agency the public most associated with this entire question, and what it did is, in its own way, the most striking move in the document.

Document voice: The Central Intelligence Agency and another agency remain unable to determine the precise origin of the COVID-19 pandemic, as both hypotheses rely on significant assumptions or face challenges with conflicting reporting.

Host: The CIA, and one more agency, declined to land. Sit with how unusual that is. This is the agency whose entire reputation is built on knowing the things no one else can know, the name the public reaches for first on a question like this, the one whose verdict would have ended a thousand arguments. It looked at the record and said: we cannot determine the precise origin, because both explanations rest on too many assumptions, or run into reporting that points two ways.

Host: Think about the pressure to do the opposite. There were people, loudly, on every side, demanding the CIA produce the answer, the real one underneath all the others. It would have been the easiest thing in the world to land somewhere, to give the country the certainty it was hungry for. Instead it wrote down that it couldn't, and explained why. That is an agency refusing to claim more than its evidence could carry.

Host: So count it with me, the way a map gets drawn. By 2023: two agencies on the lab side. The Council and four others on the natural side. The CIA and one more, undecided. Three positions on the origin question, then, natural, lab, undecided, and beneath them a fourth answer they nearly all shared: not engineered, not a weapon. One question, four different answers, from a single government. We are here to show you that the table existed, and that the government drew it for us itself.

Host: There is one movement between the two snapshots, and I want to name it carefully, because it's easy to turn into a story it isn't. In 2021, one element held the lab view. By 2023, two named agencies hold it, the Department of Energy and the FBI. The lab side grew from one to two. The assessment did not stand still; it shifted, the way assessments shift when analysts keep working a hard problem. The documents record that shift, plainly, without dramatizing it, and neither will we. A view moving from one agency to two is the system doing its job. It is not the lab answer "winning," and it is not vindication. That's all the record says, so that's all we'll say.

Host: The split is the headline. But underneath every headline there is a floor, a set of things nearly all of these agencies agreed on. And that floor is where the loudest claims of the whole pandemic go to die.

Host: Start with the question that drove the most fevered speculation: was the virus built? Engineered in a lab, on purpose, by human hands? On that, the agencies were close to unanimous. The 2023 report.

Document voice: Almost all IC agencies assess that SARS-CoV-2 was not genetically engineered.

Host: Almost all, including agencies on the lab side of the split. And the 2021 summary had said the same thing, in its own confidence vocabulary, with one honest caveat the document is careful to keep.

Document voice: Most agencies also assess with low confidence that SARS-CoV-2 probably was not genetically engineered; however, two agencies believe there was not sufficient evidence to make an assessment either way.

Host: Most agencies: probably not engineered. Two agencies: not enough evidence to say either way. We keep that caveat, because dropping it would overstate the agreement. But the direction is unmistakable, across both documents: the engineering claim found almost no support inside the agencies.

Host: And then the most extreme claim of all, that COVID-19 was a weapon, built to harm. On that one, there was no split at all. None. The 2023 report.

Document voice: All IC agencies assess that SARS-CoV-2 was not developed as a biological weapon.

Host: All of them. And the 2021 summary had put it in the IC's flattest, plainest register, the kind of sentence the IC almost never writes without a hedge.

Document voice: We judge the virus was not developed as a biological weapon.

Host: Not a bioweapon. Unanimous. In both documents. This is the part of the record fairness requires saying loudly, because it cuts hard against a story a great many people were certain of. For a long stretch of the pandemic, the idea that COVID had been engineered, or built as a weapon and released, was not a fringe whisper. It was on cable. It was in fundraising emails. And the body of the government with the most reason and the most access to find a weapon, if a weapon were there, looked and said: it is not there. That finding deserves at least as much airtime as the claim it rebuts.

Host: And hold one distinction the IC held with great care, because the public conversation collapsed it constantly. A "laboratory-associated incident," the lab pole we drew earlier, means an accident. Something going wrong around ordinary research: experimentation, animal handling, sampling. That is a completely different claim from a designed pathogen, a virus engineered, or built as a weapon. The two agencies that landed on the lab side were pointing at the accident, not at a laboratory cooking up a weapon. The documents never let those collapse together, and neither will we. Saying "not a weapon, not engineered" is not leaning toward natural origin. It is reporting the one thing nearly everyone, on both sides of the split, actually shared.

Host: So here is the question underneath all of it, and it is the deepest one. Why did serious analysts, people with access most of us will never have, keep marking their judgments low confidence, on both sides, for years? You'd think, with all that machinery, somebody would eventually find the one document, the one intercept, the one sample that ends the argument. They didn't. And the documents are unusually frank about why.

Host: The answer is unglamorous. The evidence to settle it was never there. The IC said, in 2021, exactly what it would take to break the deadlock.

Document voice: unless new information allows them to determine the specific pathway for initial natural contact with an animal or to determine that a laboratory in Wuhan was handling SARS-CoV-2 or a close progenitor virus before COVID-19 emerged.

Host: Two specific things would have moved it. Find the pathway from animal to human, or place this virus, or a close relative, inside a Wuhan laboratory before COVID emerged. Either one, and the case could move. Neither one arrived. And the one actor who could have helped deliver them did the opposite.

Document voice: Beijing, however, continues to hinder the global investigation, resist sharing information and blame other countries.

Host: Hold that line precisely for what it is. A government obstructing an investigation is a fact about why the record stayed thin, and only that. It does not point at "lab." It does not point at "natural." It points at a hole in the middle of the file where the decisive evidence should have been, and was not.

Host: And into that hole, both poles reached for what they could weigh. The lab-leaning side gave weight to, in the document's words, "the inherently risky nature of work on coronaviruses" at the Wuhan institute: a city that happened to host a major laboratory studying exactly this family of viruses. The natural-leaning side gave weight to China's officials' "lack of foreknowledge," and to "the numerous vectors for natural exposure," the many ordinary ways, across markets and farms and supply chains, a virus crosses from an animal to a person. Notice that neither of those is a fact in dispute. The risky research was real. The natural vectors were real. They both sat in the same file. The split is what happens when you ask which one to weigh more heavily, and the evidence won't break the tie for you.

Host: So let me set the ledger down plainly. Once.

Host: Here is what the record establishes. That the United States Intelligence Community was formally and openly divided on where the pandemic came from. That by 2023, the Department of Energy and the FBI assessed a laboratory-associated incident as most likely; the National Intelligence Council and four other agencies assessed natural exposure; and the CIA could not decide. That the split rested on how analysts weighed a thin, contested record, not on the facts being in dispute. And that the agencies were near-unanimous the virus was not genetically engineered, and unanimous that it was not a bioweapon.

Host: And here is what the record does not establish. It does not establish that any one of these agencies is correct, "low confidence" and "moderate confidence" are the IC's own warnings against treating its conclusions as the final word. It does not establish that the split came from political pressure or bad faith; the documents attribute the disagreement to analytic weighing, and nothing more. It does not establish a lab origin. And it does not establish a natural one. This episode adopts neither, because the agencies whose job that was did not, and could not.

Host: This is a story about a government that looked at the same evidence and reached four different answers, and said so.

Host: It is tempting, watching a government argue with itself for three years, to read that as failure. It is closer to the opposite.

Host: There is an inversion at the heart of this, and once you see it you cannot unsee it. The louder the public certainty got, it was definitely a lab leak; it was definitely natural; it was obviously a bioweapon, the lower the actual confidence ran in the file underneath. On television, the question sounded settled. In the documents, it was marked low confidence and split four ways. The two registers ran in opposite directions: maximum certainty on the airwaves, minimum confidence on the page.

Host: The people with the most access to the classified record looked at it and, in large part, refused to claim more than it could bear. They split. They marked their confidence low. They told the public they had split, and why. Against all of that, the certainty that filled the airwaves was not a more rigorous reading of the evidence. It was a less rigorous one.

Host: The split is not the scandal. The pretense, on all sides, that the question had ever been settled is the thing the record does not support.

Host: All season, we've watched a private uncertainty become a public certainty. We've watched a precise word become a public denial. We've watched a single number govern a year of American life. And now we've watched a government look at one classified record and write down four different answers, natural, lab, undecided, and a floor they nearly all shared, and tell the public it could not agree.

Host: All season, we've kept saying the same five words: what the record does not establish. Next time, we explain exactly what we mean. How this project decides what counts as evidence. Why a confidence level is a warning, not a verdict. And what you should demand from anyone who tells you a thing like this is settled.

Host: Next time, on The Public Record: "How to Read the Record."

Host: For about a year, an enormous amount of American life was organized around a single number. Not a date. Not a percentage. A distance. Six feet.

Host: You can probably still see it. It was taped onto the floor of the grocery store, that little strip of tape, or the painted footprints, or the sticker telling you where to stand. It was printed on decals in the elevator, telling four people they were now a crowd. It emptied the middle seats of restaurants and turned the tables you could sit at into a checkerboard. And in classrooms, it pulled the desks apart and left half the chairs unused, the reason a child might attend school two days a week instead of five, because a room that held thirty children at three feet held only a fraction of that at six.

Host: For about a year, six feet organized the daily texture of American life. Not as a metaphor. As tape on the floor. As the empty chair you weren't allowed to sit in. You didn't debate it. You didn't look up the source. You moved your cart back, stepped onto the mark, and waited, because that was the rule, and the rule had the weight of science behind it. And the whole time, almost none of us asked the simplest possible question about the number we were all obeying.

Host: Where did the number come from? In January of twenty twenty-four, congressional investigators put exactly that question, under oath, to the man who had been the most visible face of American pandemic policy. To Anthony Fauci, long-serving director of the National Institute of Allergy and Infectious Diseases, the official Americans saw and heard, during the pandemic, more than any other. His answer was not a citation.

Document voice: You know, I don't recall. It sort of just appeared.

Host: Not a study. Not a date. Not a name. A shrug.

Host: Now, that sentence has been pulled hard in two directions since. One side heard a confession. The other heard a clip yanked out of context to embarrass a careful witness. Both sides were sure. And here's the strange part: they were both reading the very same paragraph. The same words, under oath, on the same page of the same transcript.

Host: So tonight we're going to do the slow version. We're going to follow one number, six feet, from the floor sticker, to the classroom, to the transcript where it was finally asked to account for itself. And then to the day, a year into all of this, when the same government cut it in half.

Host: From Fauci's Lies, this is The Public Record. I'm Brian. Episode four. Six Feet Apart.

Host: You don't need me to set this scene. You were in it. But it's worth slowing down on, because we got so used to six feet that we stopped noticing how total it was. For a stretch of months, that one measurement decided where you could stand, who you could sit near, how many people counted as too many. It was on the floor of the pharmacy and the tape across the church pews. It was the spacing of the line outside the hardware store, wrapping around the building in the cold.

Host: And here's what's easy to forget now that it's behind us: six feet didn't feel like a number while you were living inside it. It felt like the air had a rule attached. You learned to read a room by it, to glance at a checkout line and know it was too long, to flinch a little when a stranger drifted close. It became a reflex. And a reflex doesn't invite questions. You don't stop in the cereal aisle and wonder where the decal on the floor got its measurement. You just stand on it.

Host: That's part of why this story is worth telling slowly. The number was everywhere, and precisely because it was everywhere, almost nobody asked it to explain itself. It carried the authority of science, of the public-health establishment, of the experts on television every night, and that authority did the work a citation usually does. You trusted it the way you trust a speed-limit sign. Someone, somewhere, must have measured. And nowhere did the measuring matter more than in schools. Because in a classroom, six feet isn't a polite suggestion. It's geometry. And the geometry is unforgiving.

Host: Here's the math, and I want to take it slowly, because it's the hinge of the whole story. Spacing children six feet apart instead of three doesn't require twice the room. It requires roughly four times the floor area per child. Distance, in a grid, squares. You're spacing in two directions at once, front to back, and side to side. Double the gap in both, and you don't double the room you need. You quadruple it.

Host: So a room that could seat a full class at three feet held only about a quarter as many at six. A classroom built for thirty becomes a classroom for seven or eight. Which means the other twenty-some children are home. Which means hybrid schedules, two days in the building, three days on a screen. Which means, for a year, the difference between six feet and three feet was, very concretely, the difference between a school that could bring its students back and one that could not.

Host: Sit with how much rides on the choice between two small numbers. If the rule is three feet, the room works, the desks fit, the class comes back, the year looks something like a normal year. If the rule is six, the same room can't hold the same children, and the year becomes a patchwork of screens and half-days and parents trying to work while a child does long division at the kitchen table. Same room. Same kids. Same building. The only thing that changed was the number on the page. That's how much weight a single distance was carrying.

Host: And I want to be careful here, because this is an episode that's going to ask a hard question, and I don't want the question to arrive before the fairness does. This act is not yet a claim about anybody. I'm not telling you the number was wrong, or that distance is a hoax. I'm describing how completely one figure governed a year of American life, and why the exact figure, six versus three, turned out to decide so much. Which brings us back to the question from the top. The one the floor stickers never answered. Where did six feet come from?

Host: In January twenty twenty-four, Anthony Fauci sat for a two-day transcribed interview with the House Select Subcommittee on the Coronavirus Pandemic. Let me gloss that, because the format matters. A transcribed interview is a sworn, recorded session, conducted behind closed doors, where committee staff can slow down and press a witness line by line, with a court reporter taking down every word. It's not a televised hearing. No cameras, no audience, no soundbite to play to. It's slow, careful, and it's under oath.

Host: And that's why it matters here. What you're about to hear isn't a critic's characterization of Fauci. It isn't a clip an opponent cut to make him look bad. It's Anthony Fauci, in his own words, under oath, giving his own account of his own record. When a witness under oath is asked a direct question and answers it, that answer is about as close to bedrock as this kind of reporting gets.

Host: The investigators asked him when discussions of a six-foot threshold began, and how it came to be applied across businesses, schools, and public life. Here is the answer in full, because the meaning of this passage has been stretched both ways, and the only honest thing to do is let you hear all of it.

Document voice: You know, I don't recall. It sort of just appeared. I don't recall, like, a discussion of whether it should be 5 or 6 or whatever. It was just that 6-foot is —

Host: "I don't recall a discussion of whether it should be five or six or whatever." Stay with that. Not, we debated five versus six and chose six. Not, here's the meeting where we settled it. He could not recall a discussion of whether it should be five, or six, or anything else. And listen to how he describes the number arriving: "It sort of just appeared." That's not the language you use for a decision you made. It's the language you use for something you found already in the room.

Host: The questioner pressed on the obvious follow-up, the one you'd ask. And here I want to be precise, because we hold ourselves to the record: that question, as it reads in the source, is a paraphrase of the investigators' line, not a word-for-word quotation. In substance: did you see any studies that supported six feet? The answer is verbatim.

Document voice: I was not aware of studies that — in fact, that would be a very difficult study to do.

Host: Not aware of studies supporting that specific distance. And, to be fair to him, he adds something true: that it would be a very hard study to even run. He's right about that. You can't easily randomize a population, stand some people three feet apart and others six, and cleanly measure who got infected. Hold onto that point, because it cuts both ways, it's part of his defense, and it's also part of the problem. And then, asked once more to account for the figure, he reached for a word.

Document voice: I think it would fall under the category of empiric. Just an empiric decision that wasn't based on data or even data that could be accomplished. But I'm thinking hard as I'm talking to you… I don't recall, like, a discussion of, 'Now it's going to be' — it sort of just appeared, that 6 feet is going to be the distance.

Host: Empiric. That's his word, not mine, and we're going to keep it, because it's load-bearing. The plain-language version: an empiric decision is a judgment call, acting on experience and instinct, ahead of the completed data, because you have to act now and the studies aren't in yet. It is not a slur. It is not, by itself, an admission of anything wrong. In a fast-moving emergency, with a novel virus and people dying, an empiric call is often the only move on the board. Doctors make them every day. So when he reaches for that word, he is not confessing to recklessness. He's naming a recognized mode of deciding under pressure.

Host: But hear what he attached to it. He said the figure was "an empiric decision that wasn't based on data," and then went further than "we didn't have the data yet." He said it wasn't based on data "or even data that could be accomplished." A judgment, by his own account, that no realistic experiment was ever going to confirm.

Host: So let's be precise about what those answers establish, and only what they establish, in his own words, under oath. Three things. One: he did not recall a deliberation over the distance. There was no meeting he could point to where someone weighed five versus six and chose. Two: he was not aware of studies supporting six feet specifically. And three: he characterized the choice himself as "empiric," a judgment call that, in his words, "wasn't based on data."

Host: That's the record of the exchange. The most prominent defender of the six-foot rule, under oath, asked the most basic question you could ask about it, where did it come from, answering, three ways across one exchange, that he doesn't know, that he's not aware of supporting studies, and that it was a judgment call not based on data. Now, what does it mean? And, just as important, what does it not mean? Because this is exactly where the story usually goes off the rails, in both directions at once.

Host: When that exchange surfaced, it traveled fast, and it traveled in one particular shape. It got summarized, over and over, in headlines, in posts, in segments, as a confession. That Anthony Fauci had, in so many words, made up the six-foot rule. That he'd admitted he invented it.

Host: I want to be very clear about what I'm doing right now. That framing is not this show's finding. It's how the clip got captioned out in the world. And the investigation this episode is built on exists, in part, to correct it. So let me correct it, plainly, even though the correction helps the person this show is investigating. The transcript will not bear that reading.

Host: He did not say he made up the rule. He did not say he invented six feet, or that it was arbitrary, or that he pulled it from nowhere. Read it again, he said almost the opposite. He said he could not recall who decided it, or how. "It sort of just appeared" is a statement of not knowing the provenance. It is not a claim of authorship. "I made it up" is a confession of being the source. "I don't know where it came from" is a man pointing at a gap in the record where the decision should be. Those are not the same sentence, and that difference is the entire space this episode operates in.

Host: And on the record, Fauci wasn't even the author of the written rule. The six-foot figure that governed classrooms was written by the C D C, in its own guidance, in its own name. If it's easier, just think the federal public-health agency, the Centers for Disease Control and Prevention. We'll get to that document in a minute. The point for now: nobody in this story, on this record, said "I invented six feet." So this show won't say it either.

Host: And here's why the distinction matters, because it's tempting to wave off as a technicality. It isn't. The damaging answers he actually gave are damaging enough on their own terms. A number organized a year of American life, and the official most associated with it can't tell you, under oath, where it came from or whether any study backed it. You don't need to add anything to that. And inflating it into a confession that was never made is its own kind of dishonesty, you've done a smaller version of the very thing you're accusing him of, substituting the story you want for the record you have. It also lets the real problem off the hook: once the "made up" version gets debunked, people assume the whole thing was nothing. It wasn't.

Host: And the second misreading runs the other way: that this proves distancing was frivolous, a sham. It doesn't establish that either. Respiratory viruses spread at closer range; putting space between people is a long-standing, reasonable precaution. This show takes no position on the optimal distance, three feet, six feet, anything between. Nothing here treats early decisions under uncertainty as bad faith.

Host: So what is the argument? If it's not "he made it up," and it's not "distancing is fake," what's left? Something narrower, and, on its own terms, still serious. The investigation puts it this way:

Document voice: …about candor: a specific number, presented to the public for a year with the authority of science, that its most prominent defender would later testify was not, to his knowledge, based on data.

Host: That's the whole episode, right there. Not a lie. Not an invention. A gap between how a number was presented, with the full authority of science, and what its most prominent defender would later testify, under oath, about its basis. Both the prosecution and the defense are reading the same paragraph. We reproduced it so you could read it too. Now, why did that gap cost so much? Because of where the number landed. Schools.

Host: The six-foot rule was not an abstraction. Above all, it was a school policy. And it wasn't a vibe, and it wasn't a rumor, it was operational guidance. Operational just means it told institutions what to actually do, day to day. For K-through-twelve schools, that guidance put the number in plain text. Here is the language, as written.

Document voice: social distancing of at least 6 feet

Host: At least six feet. Five words. And notice the phrase isn't "about six feet" or "ideally six feet." It's "at least." A floor, not a target. In a room with fixed walls, "at least six feet" is the most demanding version of the rule you can write.

Host: And that requirement collided with the geometry we did a minute ago. Six feet rather than three needs roughly four times the floor area per child. The room that seated a full class at three feet held only a quarter as many at six. For a year, that arithmetic was the policy, the line between bringing your students back full-time and sending them home half the week.

Host: And here's the thing about guidance from an agency like the C D C: it doesn't stay on a web page. It flows downhill. Federal reopening materials picked it up. State and local health departments wrote it into their own rules. District administrators built reopening plans around it. Principals counted desks against it. By the time it reached a teacher measuring out a classroom with a tape measure, "at least six feet" wasn't a recommendation anymore. It was the constraint the whole building bent around, five words in a guidance document, translated into empty desks in ten thousand classrooms.

Host: So when the investigators asked, under oath, where six feet came from, that wasn't a trivia question. They were asking about the foundation a year of childhood was balanced on top of.

Host: And I'll hold the fairness here, because it belongs here too. I'm not telling you the C D C was reckless to write it, or that six feet was chosen in bad faith. I'm telling you the precise figure, applied to the unforgiving geometry of a classroom, decided whether kids went back. That's not an accusation. That's just the stake. Keep that stake in mind. Because a year later, the same agency changed the number.

Host: On March nineteenth, twenty twenty-one, a full year into the pandemic, the C D C cut the number in half. For schools. The setting where, as we just walked through, it did the most damage.

Host: And they didn't slip it out the back door. They put it in a press release, in the agency's own name, and they leaned on a single idea: new evidence. The release was titled to emphasize "New Evidence on Physical Distance in Classrooms." The agency said it was updating its K-through-twelve guidance "to reflect the latest science on physical distance between students in classrooms." And the new recommendation, with universal masking, read like this.

Document voice: students should maintain a distance of at least 3 feet in classroom settings

Host: Three feet. Half the distance. And the reason the agency gave —

Document voice: to reflect the latest science

Host: Now hold that release up against the testimony. Side by side. The six-foot figure, by Fauci's own sworn account, "wasn't based on data." A year later, the C D C reduced it to three feet, for schools, and the reason it gives is the latest science, new evidence. Which raises the question the whole episode has been circling: if the evidence is what justified three feet in twenty twenty-one, what was holding up six feet in twenty twenty?

Host: Picture the timeline as a straight line. At one end, twenty twenty: the six-foot rule, written into school guidance, reshaping classrooms. Then a year passes, a whole school year, with its hybrid schedules and empty desks and kids on screens. And only at the far end, March twenty twenty-one, does the C D C stand up and say, in effect, we've looked at the latest science, and for classrooms, three feet is enough. The evidence the agency cited to halve the rule arrived after the rule had already governed a year of American childhood. It didn't lead the rule in. It came along behind it.

Host: Which is the line this story turns on. The investigation says it cleanly, and I'll attribute it to the investigation, because it's a synthesis, not a raw fact: The number came first. The data came second.

Host: And I want to keep that sober, because it's easy to over-read. This is a sequence on the dated record, the rule, then a year, then the revision citing new evidence. It is not, in this show's voice, a claim about anyone's motive. I don't know why the evidence came late, and I'm not going to pretend I do. Maybe the early evidence genuinely pointed at six and later evidence revised it. Maybe six was always more cushion than the science required. The documents don't tell us which, and so neither will I. What they do tell us is the order, and that order is not interpretation. It's just the record.

Host: This is the part of every episode where we draw the line hard, between what the record actually establishes and what it doesn't. We do it every time, on purpose, because the whole value of working from documents is lost the moment you start saying more than the documents say. On this episode, that line carries the season's single most important correction, so I'm not going to rush it. Here is what the record establishes. Three things, each on a sworn statement or an official document.

Host: One. By the sworn account of the official most associated with it, the six-foot standard was an "empiric" decision, a judgment call he could not tie to specific studies, and that, in his words, "wasn't based on data."

Host: Two. That figure was written into C D C school guidance as "at least 6 feet," and that requirement drove whether and how schools could reopen.

Host: Three. In March twenty twenty-one, the C D C cut the school distance to three feet, citing new evidence, a full year after the original rule took hold.

Host: And here is what the record does not establish. I'm going to say each one out loud, because the absence is as important as the presence.

Host: The record does not establish, and this show does not assert, that Fauci "made up" the rule, or invented it personally. He testified he did not recall its origin. "It sort of just appeared" is a man saying he doesn't know where a number came from. That clip caption, the confession everyone forwarded, is not in the transcript. We checked. It isn't there. We won't run it as fact, because it isn't one.

Host: The record does not establish that physical distancing is ineffective. Let me be emphatic, because the "made up" crowd sometimes uses this exchange to argue distance never mattered at all, and that's not here either. Keeping space between people is a long-standing, reasonable precaution against close-range respiratory spread. This show takes no position on the single correct distance, not six, not three.

Host: And the record does not establish that anyone acted in bad faith. This is maybe the most important concession, so I'll be plain. Emergency decisions ahead of completed data are normal, often necessary. "Empiric" is not an insult and not a crime. In the first chaotic months of twenty twenty, officials made countless calls without finished studies, because waiting for certainty would have meant doing nothing while people died. The question was never whether officials acted before the studies were in. Of course they did. They had to. The question is narrower, and it's the only one the record actually supports: was the public told how thin the basis was?

Host: What's left, when you set aside everything the record won't support, is narrow, and it's the thing the record genuinely does show. A judgment made without data, presented to the public with the authority of science, and left uncorrected for a year. That's not a verdict. That's a candor problem. And it's the one we can stand behind, because it's the one the documents bear.

Host: There is nothing wrong with a public-health official making an empiric call in a crisis. There is something worth examining in presenting that call as though it were settled science, in letting the public believe, for a year, that there was firm data underneath a number its most prominent defender would later testify there wasn't. The failure, if there's a failure, isn't in the six feet. It's in the confidence with which the six feet was sold, and in how long it took for anyone to say out loud how thin the basis was.

Host: A government may have to act before it has proof. That's not the failure. Sometimes that's the job.

Host: What a government should not do is take a judgment made without data and hand it to the public as though it were a finding drawn from data, wrap a guess in the authority of science, put it on the floor in tape, build it into the walls of a classroom, and then leave the correction for a year later, after the cost has already been paid. And in this case the cost was paid in something specific: empty desks, half-attended years, kids who learned at a kitchen table because a room that fit thirty at three feet fit seven at six.

Host: The six-foot rule is a small thing and an enormous one. Small, because it is a single number, two syllables, a measurement you could pace off in a hallway. Enormous, because that number, by the testimony of the man who defended it, "sort of just appeared," and then shaped how millions of American children spent a year of their lives.

Host: The distance was real. The evidence, when it finally came, said three feet. The candor, in between, is what went missing.

Host: Across this season, we've been chasing one thing, the gap between how certain officials sounded in public and how uncertain the record shows they were underneath. We've watched a private uncertainty become a public certainty. We've watched a precise word become a public denial. And tonight, we watched a single number govern a year of American life, taped onto the floors, built into the classrooms, carried with the full authority of science, and then, a year later, quietly cut in half by the same agency that wrote it.

Host: There's one more split to follow. And this one isn't out in a grocery store or a classroom. It's inside the government itself, where different agencies read the very same origins question and reached different answers, at different levels of confidence. Not "the government concluded X." Different parts of the government, reading the same record, landing in different places, and saying so with very different degrees of certainty. What did the declassified record actually say? And how sure was anyone, really?

Host: Next time, on The Public Record: The Split.

Host: Here is a thing that can happen with words. Two people can stand in the same room, on the same afternoon, and say the very same three words to each other, and mean two completely different things by them. Neither one is lying. Neither one stops to say so. And everybody listening goes home certain they heard the same sentence.

Host: That is not a riddle. It is a transcript. Picture a hearing room in the United States Senate. The date is May eleventh, twenty twenty-one. The country is fourteen months into a pandemic, masks on the people in the gallery, careful distance between the seats, and the question pressing on everyone outside that room is simple, and frightened, and direct. Did we, did the United States government, pay to make a virus more dangerous?

Host: At the witness table is a man the whole country has been watching for more than a year: Doctor Anthony Fauci. For decades he has run a federal institute that funds research on dangerous viruses. He has testified in rooms like this one more times than almost anyone alive. He knows exactly how a single word can be parsed. Across from him, a senator and physician, Rand Paul of Kentucky, leans into his microphone, and you can feel the temperature in the room start to climb. And over the next ninety seconds the two of them are going to use the same three words, gain of function, to mean two completely different things, and neither one is going to say so.

Host: Paul makes his charge in plain English. He says the lab in Wuhan had been working to make a coronavirus better at infecting people, that this was gain-of-function research, and that American money helped pay for it. And Fauci answers. Not carefully. Not with a hedge. Flatly. Listen to the words as they sit in the official record of that hearing.

Document voice: it is not gain-of-function. And, if you look at the grant and you look at the progress reports, it is not gain-of-function.

Host: If you were watching that on television, you heard a denial. A plain, clean, English-language denial that the government had paid to make a virus more dangerous. He said it twice. It is not. That is what most people heard. And that is a reasonable thing to have heard.

Host: Now jump forward three years. Same man. No cameras. No clock running. Behind closed doors, under oath, congressional investigators sit Anthony Fauci down and ask him to slow all the way down and define his terms. And he explains that the denial you just heard had never meant what the country thought it meant. It had been true, but true in a very particular, very narrow, very technical way. A way the public was simply never handed.

Host: That gap, between the sentence people heard and the sentence he was actually saying, is the whole story. It is a story about a word, and about how a true answer can leave a room knowing less than it did before. So that is what we are going to do. We are going to take one phrase, follow it as it splits into three separate meanings, and show you exactly how that happens.

Host: From the newsroom of Fauci's Lies. This is The Public Record. I'm Brian. Episode three. The Definition Problem.

Host: Before we go anywhere, let me put a few people and a few words in your hand, because this episode lives or dies on keeping them straight. If you are new here, you lose nothing. We will build it from the ground.

Host: First, the phrase itself. Gain of function. In plain English, a gain-of-function experiment is one that gives something a new ability, or makes an ability it already has stronger. You take an organism, say, a virus, and you change it so it can now do something it couldn't do before, or do it better. That is the everyday meaning. Hold onto it. The whole trouble in this story is that the everyday meaning is not the only meaning, and the people in that hearing room were not all using the same one.

Host: Second, the people. Anthony Fauci ran a federal agency called the National Institute of Allergy and Infectious Diseases. The short name is NIAID. If it is easier, just think Fauci's institute. It funded the infectious-disease research at the center of this fight, and it sits inside the larger National Institutes of Health, the NIH. The senator pressing him that day is Rand Paul. For our purposes, Paul matters for one reason: the hearing produced an official transcript, and that transcript caught Fauci's answer, in public, on the record.

Host: And third, the money, because the NIH funded it can sound abstract, and it isn't. The path of the dollars went like this. Fauci's institute, NIAID, awarded a grant to a New York nonprofit called EcoHealth Alliance. Say it EE-koh-health. EcoHealth, in turn, sent part of that money overseas, to a research partner, the Wuhan Institute of Virology, to study bat coronaviruses. Federal institute, to New York nonprofit, to overseas lab. You don't have to hold the details. The point is just that public money funded research, and some of that research happened at the Wuhan lab. That money trail is real, and it is not in dispute. What is in dispute, the entire fight, is what to call the experiments at the end of it.

Host: And one more thing, before your ears get hot, because this is exactly where coverage of this story tends to go off the rails. This episode is about one narrow, almost boring-sounding thing: the funded experiments, and what you can and cannot truthfully call them. It is not about where COVID came from. That sounds smaller than the headlines. It is smaller than the headlines. But it turns out to be the thing the whole public fight was actually about, and almost nobody named it.

Host: Okay. With those in hand, go back into the room.

Host: Senator Paul makes his charge, and I want to attribute it carefully, because these are his words, his framing, not a finding of ours. Paul says the Wuhan Institute of Virology had been, in his words, experimenting to enhance the coronavirus' ability to infect humans. He calls that gain-of-function research. In another moment he sharpens it and says the work was meant to create super viruses. And he says the NIH paid for it.

Host: Notice which meaning Paul is using. He is using the big one, the everyday one, the one you are holding. Take a virus, make it better at infecting people, that is a new ability, that is a gain of function, plain as day. That is the broad meaning, swung like a hammer. And on its own terms, it is a perfectly reasonable way to use the words.

Host: And then Fauci answers. Same three words.

Document voice: it is not gain-of-function. And, if you look at the grant and you look at the progress reports, it is not gain-of-function.

Host: Here is the thing I need you to see, and it is the hinge of the whole hour. Paul swung the broad meaning. Fauci answered with a different meaning, a narrow, technical one he never names out loud. Two definitions passed each other in the air of that room, in the same sentence, in the same breath, on live television. And nobody watching could possibly have known it. There is no asterisk you can hear. There is no footnote in a man's voice. You just hear it is not gain-of-function, and you file it away as: the answer is no.

Host: And think about who was actually listening that day. Not a room of virologists. People at home, on the couch, more than a year into a pandemic that had killed hundreds of thousands of Americans, scared, looking for someone to tell them the truth in a language they could understand. The question underneath Paul's question, did we help build the thing that did this to us, is about as high-stakes as a public question gets. And the answer that comes back is three flat words. It is not gain-of-function. To a frightened person at home, that does not sound like a regulatory finding. It sounds like a man saying: no, we didn't do that. Rest easy.

Host: So which meaning was Fauci using? He didn't say. For three years, he didn't say. And then, under oath, he did.

Host: January eighth, twenty twenty-four. A different kind of room entirely, and the format matters, so let me tell you what it is. This is what is called a transcribed interview, run by the House Select Subcommittee on the Coronavirus Pandemic. It is not a TV hearing. There is no five-minute clock, no cameras, no cable-news segment to get to. It is investigators and lawyers in a room, on the record, under oath, with a court reporter taking down every word, and they can slow all the way down and make a witness define his terms, word by word, until there is nothing left to define. It is a congressional investigation, not a court. Its job is to gather records, not to hand down verdicts. And it is the one setting in all of Washington built to do the thing a televised hearing can never do: stop someone, mid-sentence, and say, define your terms.

Host: And that is exactly what they did. With all the time in the world, the staff asked Anthony Fauci to do the one thing the cable exchanges never had time for. They asked him to say what gain of function actually means. And he agreed, this is the key, that the phrase had been used in at least three distinct ways. I am going to walk you through all three, slowly, because once you hold these three apart in your head, this whole confusing fight suddenly becomes simple. And I should tell you up front, to be fair: that three-part breakdown wasn't a trap sprung by critics. The framing was put to Fauci by the minority's own counsel, the lawyers working for the side more sympathetic to him, and Fauci confirmed it. So this is his own taxonomy, on the record, agreed to in a room where he had every chance to push back.

Host: Definition one. The broad one. Call it the layman's one, the one you have been holding since the top of the show. In the record it is described as research that, and here I will let the record speak. Daniel.

Document voice: modifies a biological agent so that it confers new or enhanced activity to that agent

Host: New or enhanced activity. That is it. Change a thing so it can do something new, or do it stronger. By Fauci's own account, some scientists use the term that broadly, for any such change. And, this is important, he said this broad usage had, in his words, no formal, regulatory significance. Meaning: it is how people talk. It is the dictionary, not the rulebook. And under this first definition, an enormous amount of laboratory work counts as gain of function, maybe most of it. Hold that thought, because it comes back.

Host: Definition two. The twenty fourteen federal moratorium. Now we are moving from how people talk to what the government actually did. Back in twenty fourteen, the U.S. government got nervous about a specific slice of risky research and hit pause on the funding. A moratorium, a freeze. And it spelled out exactly what it was freezing.

Document voice: reasonably anticipated to confer attributes to influenza, MERS, or SARS viruses such that the virus would have enhanced pathogenicity and/or transmissibility in mammals via the respiratory route.

Host: Let me translate that out of bureaucrat, and flag a couple of words as I go. Pathogenicity just means the ability to cause disease, how sick something makes you. Transmissibility is how easily it spreads. And via the respiratory route means through the air, the way you'd catch a cold. So definition two is the government saying: we will not fund work expected to make flu, or MERS, or SARS viruses better at making you sick, or better at spreading through the air. That is narrower than definition one. It names specific viruses, it has a purpose. One quick footnote so you are not thrown: MERS is a different coronavirus disease. It is not COVID. It is in there only because the rule listed it.

Host: That twenty fourteen rule was a blunt instrument, broad strokes. And it stayed in force until twenty seventeen, when the government replaced it. Which brings us to the definition that actually matters.

Host: Definition three. The twenty seventeen framework. Its formal name sounds like a parking permit: P3CO. Say it as P, three, C, O. You don't need to memorize the letters. Think of it as a regulatory gate, a specific government checkpoint that a risky experiment has to pass through for extra, higher-level review. And the gate has a very narrow opening. Most research walks right past it, no stop required. It only swings shut for something the policy calls a potential pandemic pathogen, and here is the record's own framing of what that is.

Document voice: likely to be highly transmissible and spread widely... and likely [to] cause a high degree of morbidity and mortality.

Host: Read that slowly, because it is doing a lot of work. A potential pandemic pathogen isn't just a virus you have made a little nastier. It is a virus already capable of seeding a pandemic, already in the danger zone, that an experiment is then expected to make even more transmissible, or even more deadly. That is a high bar. A genuinely high bar. And the gate only slams down when both halves are true: a pandemic-grade pathogen, and work expected to make it worse. That is the function of the operative definition. It isn't a description of what is dangerous in the abstract. It is a switch that decides whether a particular experiment needs extra scrutiny before the money flows.

Host: So now hold the three of them up side by side, because this is the payoff, and I want it clean.

Host: Definition one, the everyday one, sweeps in almost any change that adds an ability. Wide net. No legal teeth.

Host: Definition two, the twenty fourteen freeze, a blunt government funding pause, aimed at making flu, MERS, or SARS viruses more dangerous to spread.

Host: Definition three, the twenty seventeen gate, narrow, legalistic, triggered only by experiments on viruses already capable of causing a pandemic.

Host: These are not the same thing. They are not even close to the same thing. And here is the sentence I most want you to carry out of this act, because it is the whole engine of the confusion: a great deal of research is gain of function under definition one, the wide one, and not gain of function under definition three, the narrow one. Same experiment. Two true, opposite-sounding answers. It just depends which net you are holding.

Host: And if you want a homely way to keep this straight, try this one. Think about the word organic. At a farmer's market, organic means roughly grown without the nasty stuff, a loose, everyday sense. But on a federal food label, organic is a legal term with a precise list of rules behind it, and a farm can fail that legal test while still, in plain language, growing food the careful way. Now picture someone asking a farmer, on live television, is your food organic, and the farmer says no, meaning the strict legal label, while the whole audience hears no meaning the everyday sense. Nobody is lying. But the answer lands wrong. That is the exact shape of what happened with the word gain of function. One word. A farmer's-market meaning and a federal-label meaning. And the audience was holding the wrong one.

Host: So when somebody asks Anthony Fauci, are you funding gain-of-function research, which net is he holding when he answers?

Host: He told us. Plainly, under oath, in that twenty twenty-four interview, and he gave the narrow one, definition three, a name. He called it the operative definition. His word. The one that operates. The one with legal force. He testified that the P3CO framework came out of years of deliberation and gave them, in his words, a very precise definition. And then came the move the public, back in twenty twenty-one, never got to hear, the moment he explained which definition lives in his head when somebody asks him the question. Daniel, the pivot.

Document voice: when I said we do not fund gain-of-function — gain-of-function according to the strict definition, which I refer to as the operative definition... when someone asks me, as a scientist, are you doing gain-of-function, is that gain-of-function, I always apply it to the operative definition of 'gain-of-function.'

Host: I always apply it to the operative definition. Always. That is the missing clause. That is the asterisk you couldn't hear in the hearing room. So when Rand Paul asked the broad question, did you fund work that made a virus more dangerous, and Fauci answered it is not gain-of-function, he was answering against the narrow regulatory gate. Definition three. Every single time. He just didn't say so.

Host: Now, I need to slow all the way down here, because this is the fork in the road where it would be easy, and cheap, to score a point. And here is where this show holds its line: the narrow definition was not a trick.

Host: Fauci is a scientist and a regulator. When a term has a precise legal meaning, and gain of function, in the federal rulebook, did have one, applying that precise meaning is precision, not evasion. The narrow definition was real. It was the binding federal standard. It was the actual rule that decided whether a given experiment needed extra government review. A scientist who hears are you doing gain-of-function and reaches for the regulatory meaning is using a word the way his field uses it. That move was legitimate.

Host: So let me locate the problem exactly, because precision is the whole point of this episode. The difficulty is not that Fauci used the narrow definition. He is allowed to use the narrow definition. The difficulty is this: when he used it in public, he did not say he was using it.

Host: That is the whole thing. That one sentence. He gave a yes-or-no answer, in plain English, to a room full of people using the plain-English meaning, and the answer secretly ran on a definition only he was holding. He answered a plain public question with a private dictionary, and he didn't hand the public the dictionary.

Host: And here is why that gap is so easy to miss, and so hard to defend at the same time. In a grant review, in a room full of people who all know the operative definition, Fauci's answer isn't just acceptable, it is the correct answer. The precise one. But move that exact same sentence to a podium, in front of an audience that has never heard the words potential pandemic pathogen in their lives, and the precision becomes its own kind of fog. The more technically careful the answer, the less the public can actually understand it. That is the strange trap of this whole story. The accuracy and the confusion are the same sentence.

Host: There is one more turn before we weigh all this, and it is the reason this episode is called The Definition Problem and not just The Definition. Because the operative definition didn't fall out of the sky fixed and final. It moved. It got redefined, by the government, on the record, in plain view.

Host: Go back to definition two for a second, that blunt twenty fourteen moratorium, the wide funding pause on making flu, MERS, or SARS viruses more dangerous to spread. In twenty seventeen, the government retired that blunt rule and replaced it with the narrow gate, definition three. And here is Fauci himself, in his own words, describing what that swap did.

Document voice: an even more specific set of definitions: potential pandemic pathogen, which is a multipart definition, and all sorts of carve-outs.

Host: A multipart definition, and all sorts of carve-outs. Hear what that means. The rule got narrower. It went from a broad class of dangerous viral enhancements down to a tightly bounded category, only the pandemic-capable stuff, with exceptions written in. And that is a perfectly legitimate thing for a government to do. Narrowing a rule so it captures only the genuinely catastrophic cases, instead of freezing every routine lab procedure, is a defensible regulatory choice. You can argue it was the right call.

Host: But it had a side effect, and the side effect is the quiet heart of this whole thing. As the official definition got narrower, the set of experiments an official could truthfully call not gain of function got bigger. The regulatory category shrank. The honest-deniability zone expanded. And meanwhile, out in the world, at kitchen tables, on television, the everyday meaning of the phrase didn't budge an inch. The public was still holding definition one, the big net. The government had quietly switched to a small one. The rulebook moved. The dictionary stayed put. And a person watching at home never sees that shift. They have never read the P3CO framework, they have one meaning of the phrase, and when the official says it is not, they don't hear a narrow regulatory finding. They hear no, we didn't do the scary thing. They are standing on ground that is moving, and nobody hands them the map.

Host: Which brings us to the question you have probably been holding this whole time. Fine, Brian, so was the Wuhan work gain of function, or wasn't it? Just tell me.

Host: And the only honest answer I can give you, the only one the documents support, is a question back. Under which definition. That genuinely is the answer. Let me show you both, because the record gives us both.

Host: Take the broad one first. Definition one. Did the funded work give a virus a new or enhanced ability? For that, we don't need anybody's opinion. We have the grant's own progress report, the paperwork the researchers themselves sent back up the chain to the agency that paid them. As this series documented in our last episode, the funded researchers built chimeric coronaviruses, engineered hybrids, viruses stitched together from pieces of others, and tested them in humanized lab mice. And one of those built viruses made the mice noticeably sicker. The report records it, in its own words.

Document voice: the pathogenicity of SHC014 is higher.

Host: Higher pathogenicity. Remember the translation, pathogenicity is the ability to cause disease. The engineered virus made those animals sicker. And under the broad, layman's definition, conferring an enhanced ability like that is a gain of function, almost by construction. That is not a leap. That is what the words mean. The researchers did something to the virus. The virus gained a function. The report says so in the grantee's own words. So under definition one: yes. You don't even have to argue. You just read.

Host: And I have to put a hard fence around that yes, because it would be so easy to run past it. Saying that experiment was a gain of function under the broad definition is not the same as saying it caused the pandemic. It is not. And it is absolutely not the same as saying it produced the virus that caused COVID. The record does not say that, and neither do I. The grant paperwork establishes one specific, narrow thing, that a modified virus was more pathogenic in the animals tested. That is the broad-definition answer, and only that. Hold that fence.

Host: Now the operative one. Definition three. The narrow gate. And here the answer flips, and it doesn't flip because we are being generous. It flips because the NIH itself says so, in its own name. In October of twenty twenty-one, the NIH's principal deputy director, a man named Lawrence Tabak, that is spelled T, A, B, A, K, wrote a letter to Congress, to Representative James Comer, addressing exactly this work. And I want to flag why this letter matters: this is not a critic's summary, and it is not a defender's spin. This is the agency speaking for itself, on paper. Here is what it said about whether the work tripped the regulatory gate.

Document voice: was not subject to departmental review under the HHS P3CO Framework

Host: Not subject to that review, because, the NIH's position is, these viruses were not potential pandemic pathogens. They weren't in the danger zone the narrow definition is built around. They didn't trip the operative wire. And on the bigger fear, the question of whether this work could have become the pandemic virus, the same letter was just as direct.

Document voice: could not have become SARS-CoV-2

Host: Could not have become the virus that caused COVID. That is the NIH, in its own name, on the record. So under definition three: no. That letter is a real document, a real position, and it answers a real question, and we are representing it here, in full.

Host: So now look at what you are holding. Under the broad definition: yes, the report says the modified virus was more pathogenic. Under the operative definition: no, the NIH says these weren't pandemic pathogens, the review didn't apply, they couldn't have become COVID. And here is the part that drives people crazy, and it is the truest thing in this episode. Both of those answers can be right at the same time. They are not in a fistfight. They are answers to two different questions. The broad answer is answering, did you add an ability? The narrow answer is answering, did you cross the regulatory line for pandemic-grade danger? Different questions. Different nets. Both answered honestly. And nobody at that Senate hearing told the audience which question was being answered. That, exactly that, is the definition problem, sitting right out in the open.

Host: This is the part of every one of our episodes where I stop building the case and read you both columns of the ledger, what the documents actually establish, and, just as important, what they do not. On this story more than almost any other, that line matters, so I am going to read both columns out loud, slowly, and plainly.

Host: Here is what the record establishes.

Host: One. By Fauci's own sworn testimony, the phrase gain of function carried at least three distinct meanings, and he told investigators that when he is asked, as a scientist, he always, his word, answers with the narrowest one. The regulatory one. The operative one.

Host: Two. In public, in twenty twenty-one, he denied funding gain-of-function research without ever indicating that his denial rested on that one specific, narrow definition. The room heard plain English. The answer ran on a term of art. The clause was true. The clause was also silent.

Host: Three. That operative definition had itself been narrowed in twenty seventeen, which means the ground under the word not had shifted. There was simply more that could truthfully be called not gain of function than there had been a few years before, even as the public's everyday meaning of the phrase stayed exactly where it was.

Host: That is the first column. Now the second one. And I want to say each of these out loud, in my own voice, because they are not footnotes, they are the guardrails of this whole episode.

Host: The record does not establish that Anthony Fauci lied, and this show does not assert it. Applying a precise regulatory definition is legitimate. The name on this masthead is not the verdict of this episode.

Host: The record does not establish that the narrow definition was cooked up to dodge this question. It wasn't. The P3CO framework was finalized in twenty seventeen. This controversy didn't boil over until years later. The definition predates the fight. It was the binding federal standard before any of this was a headline, before that Senate hearing, before most people had ever heard the phrase gain of function. Nobody invented it in a back room to give Anthony Fauci an escape hatch on live TV. You can't invent, after the fact, a rule that was already on the books.

Host: And the record does not establish that the broad critics are simply right, that this work cleared the bar for pandemic risk. It doesn't say that. The NIH's contrary position, not subject to P3CO review, could not have become the virus that caused COVID, is on the record, in its own letter, in its own name, and we have represented it here, in full, not knocked it down. If you walked away from this episode thinking we had proven the critics right, we would have failed just as badly as if we had called it a lie.

Host: And I want to be especially careful with one more thing, because it is the easiest place in this whole story to overreach. Nothing we have laid out tells you where COVID came from. Not one document in this episode answers that question. The grant's progress report tells you what one funded experiment did to some mice. The Tabak letter tells you these particular viruses couldn't have become the pandemic virus. The transcripts tell you which definition Fauci was using. That is it. The origin of the pandemic is a separate question, with a separate body of evidence, and anyone who tells you this definition fight settles it, in either direction, is reaching past what these papers actually say.

Host: And here is the fairness underneath all of it, in one line: both of the central people in this story have a real point, and you can hold both at once. Paul's point is real, under the meaning ordinary people use, an experiment that makes a virus better at causing disease is a gain of function, and the paperwork shows that happened. Fauci's point is also real, there is a precise federal definition with an actual gate attached, and by that definition, the answer is genuinely no. Neither of them is hallucinating. They are standing on opposite ends of the same word. The failure isn't that one of them was lying. It is that the one with the microphone, in the moment the public most needed the word defined, left it undefined.

Host: So if it isn't a lie, what is it? The failure the documents actually show is narrower than a lie, and sharper than a misunderstanding, and it has a precise name. It is about candor. A public yes-or-no answer that secretly turned on a definition the audience was never given.

Host: And here is why that matters. Our newsroom put it in a single image, and I will borrow it, because it is exactly right. A term of art is a fine thing in a regulatory filing. Lawyers and scientists use precise definitions all day long, and they should. But a term of art is a dangerous thing at a podium. When an official answers a plain, frightened public question, did we pay to make this virus more dangerous, with a word he is quietly defining more narrowly than everyone listening, the public isn't exactly lied to. It is something subtler. They are handed a true sentence they have no way to evaluate. They are quietly disqualified from the conversation about their own safety.

Host: And the remedy, this is the part that stays with me, the remedy was never a verdict on virology. It wasn't a lab report. It wasn't some grand confession. It was one clause. Four words. By the regulatory definition. If Fauci had said that out loud in twenty twenty-one, it is not gain-of-function, by the regulatory definition, he would have turned a flat denial into an honest, contestable claim. Something a citizen could actually push back on. Left unsaid, that same definition stopped being a tool for precision. It became a place to hide.

Host: So that is the answer to the question Episode two left you holding. When officials fought over whether the Wuhan work was gain-of-function, were they arguing about the facts, or the definition? Mostly the definition. The facts of the experiment were sitting right there, in the grant's own paperwork. What people couldn't agree on, what they were never even given the tools to agree on, was which meaning of the word to hold while they read it. They meant at least three different things. And in the one room where it mattered most, in front of the most people, nobody said which.

Host: We have now watched the same gap twice. A private uncertainty turned into a public certainty. A precise, narrow word turned into a flat public denial. Next time, we follow that exact same gap to a place every single one of us actually lived it, not in a hearing room, but in a grocery-store aisle, in a closed school, in a strip of tape on the floor telling you where to stand. The rules that governed how far apart we had to stand, which schools could open, and what the evidence behind those rules actually was, underneath the confidence.

Host: Next time, on The Public Record: Six Feet Apart.

Host: The Public Record is a production of the Fauci's Lies newsroom. Every quotation you heard read aloud in this episode was quoted verbatim from the public record, sworn testimony, the hearing transcript, and the NIH's own letter to Congress. The sources are footnoted in our companion investigation, Gain-of-Function, Defined and Redefined, at faucislies dot com. I'm Brian. Thanks for listening.

Host: Start with two numbers. Seventy-one percent. Twenty-five percent. In a laboratory in Wuhan, one group of mice was infected with a bat coronavirus called WIV1. Two weeks later, five of the seven mice were still alive. Seventy-one percent. A second group was infected with a virus that did not exist in nature. Scientists had taken the WIV1 backbone and inserted the spike protein from a different bat coronavirus. That spike was called SHC014. Two weeks later, only two of eight mice survived. Twenty-five percent. Both numbers come from the same document. Not a critic's summary. Not a cable-news segment. The grant's own progress report. That matters. A progress report is not an op-ed. It is paperwork from the grantee back to the funding agency, describing what the project did and what the researchers found. So for this episode, the strongest evidence is not someone else's description of the grant. It is the grant record itself.

Document voice: Five out of seven mice infected with WIV1 remained alive, while only two of eight mice infected with rWIV1-SHC014 S survived.

Host: This is The Public Record. Episode two. What the Grant Actually Funded.

Host: The first episode followed a paper. This episode follows money. The grant was numbered R01AI110964. Its title was Understanding the Risk of Bat Coronavirus Emergence. The purpose, on paper, was defensive. Researchers wanted to understand which bat coronaviruses might jump into people, so the world could prepare. That is a legitimate scientific goal. The question is not whether spillover research can matter. It can. The question is what this grant actually paid for, and how clearly the public was told about it. The money started at the NIH. NIH means the National Institutes of Health. It is the main federal biomedical research agency. More specifically, it came through NIAID, the infectious disease institute Anthony Fauci directed. NIAID sits inside NIH. Fauci directed NIAID for decades, which is why he is part of the funding story. From there it went to EcoHealth Alliance, a New York nonprofit. EcoHealth was the grantee. In plain language: NIH funded EcoHealth to do the project. And from EcoHealth, part of it went as a subaward to the Wuhan Institute of Virology. The Wuhan Institute of Virology was the overseas research partner that performed part of the work described in the progress report. NIH. NIAID. EcoHealth. Wuhan. That is the path. The grant had three broad aims. Find new bat viruses. Model which ones might infect people. Then test those predictions. Aim three is where the story turns.

Host: The progress report names Aim three plainly. Testing predictions of coronavirus inter-species transmission. In normal language: test whether viruses found in bats might be able to cross into other species, including humans. The phrase inter-species transmission is important, but it is not mysterious. It means a virus moving from one kind of host to another. From bats to another animal. Or, in the risk question everyone cared about, from animals into people. The technical tools were reverse genetics, cell-culture experiments, receptor tests, and humanized mice. Reverse genetics is a lab method for building or modifying a virus from genetic instructions. Cell culture means testing in cells grown in a lab dish. Receptor tests ask whether a virus can use a particular doorway into a cell. Humanized mice are mice engineered to carry a human receptor. In this case, the receptor was human ACE2. That matters because ACE2 is one of the doors coronaviruses can use to enter cells. So what did the researchers test? The progress report says they infected those mice with four SARS-related coronaviruses. One was a full-length recombinant WIV1 virus. The other three were chimeras. That word sounds abstract. Here, it means the scientists used the backbone of one virus and the spike proteins of others. Think of it as testing a swap. The backbone is the main viral structure. The spike is the part that helps the virus attach to cells. The experiment asked what happened when researchers changed that spike. The report says it directly.

Document voice: Three chimeric viruses with the backbone of WIV1 and S proteins of SHC014, WIV16, and Rs4231.

Host: If the phrasing is hard to follow, the grant's plain-language summary says the same thing more simply.

Document voice: We used the genetic codes of some of the other viruses we found in bats and inserted the spike protein genes of those viruses into the cultured viruses.

Host: Then the summary explains why that mattered.

Document voice: We also showed that some of these viruses cause SARS-like illness in mice that are adapted to have similar cell surface receptors to people.

Host: That is the center of the episode. The work was not just looking at bat viruses in a database. It was not just collecting samples. It was building chimeric viruses and testing them in mice engineered with a human receptor. Again, the stated purpose was to understand risk. But the method was not vague. The grant's own report describes it. And then it describes the result.

Host: The experiment ran for two weeks. The researchers infected mice with the different viruses and watched survival. All four caused lethal infection in some mice. But one stood out. The SHC014 chimera. The report says that after fourteen days, five out of seven mice infected with WIV1 were still alive. But only two out of eight mice infected with the SHC014 chimera survived. The report also says that this chimera was detected in the animals' brains at every time point, showed increasing viral levels after infection, and produced more significant lung lesions. Then the report draws its conclusion.

Document voice: The pathogenicity of SHC014 is higher than other tested bat SARS-related coronaviruses in transgenic mice that express human ACE2.

Host: Pathogenicity means the ability to cause disease. So translated carefully, the grant report says this lab-made chimera caused more disease in those humanized mice than the other bat coronaviruses tested. That does not answer where COVID came from. It does answer a narrower question. What kind of work did the grant fund? It funded the creation and testing of chimeric bat coronaviruses. One of them made humanized mice markedly sicker. And the SARS-related mouse experiment was not the only engineered-virus work described in the report. There was a parallel line involving MERS. MERS is a different coronavirus disease. It is not COVID. It matters here only because the same report describes another engineered-virus test in the same broader grant work. The report says researchers built a full-length infectious clone of the MERS virus and replaced one of its binding regions with binding regions from bat coronaviruses.

Document voice: These chimeric MERS coronaviruses were able to infect human cells from different tissues, including lung, liver, intestine, and kidney.

Host: That line is not there to say MERS caused COVID. It did not. It is there because it shows the pattern of the grant. Find bat viruses. Swap genetic parts. Test what the new construct can do in human cells or humanized animals. That is more specific than the public shorthand usually allowed. And for more than a year, that fact was not central to the public story. It became public in a very different way. Through a letter to Congress.

Host: On October twentieth, twenty twenty-one, Lawrence Tabak, a senior official at the NIH, wrote to Representative James Comer. Tabak was NIH's principal deputy director. That made the letter different from a critic's allegation. It was NIH speaking in its own name. Comer was a member of Congress receiving an official response in an oversight fight. The point is not Comer's politics. The point is that Congress had asked questions, and NIH answered on paper. The letter was only two pages. But it changed the record. Tabak wrote that the final progress report had been submitted to NIH in August twenty twenty-one, after what he called compliance enforcement efforts. Then he described the experiment.

Document voice: The limited experiment was testing if spike proteins from naturally occurring bat coronaviruses circulating in China were capable of binding to the human ACE2 receptor in a mouse model.

Host: And then came the sentence that made headlines.

Document voice: Laboratory mice infected with the SHC014 WIV1 bat coronavirus became sicker than those infected with the WIV1 bat coronavirus.

Host: That was the NIH describing the same basic result the grant report had recorded. Mice became sicker. Tabak also wrote that the result was unexpected. The researchers, he said, had not set out to make the mice sicker. But the grant terms mattered. They included a tripwire. If an experiment produced a one-log increase in growth, the grantee was supposed to report it immediately. That reporting was supposed to trigger a secondary review. Tabak wrote that EcoHealth did not report the finding right away.

Document voice: A requirement that the grantee report immediately a one log increase in growth.

Host: The Select Subcommittee later made EcoHealth's reporting failures a central finding. That Subcommittee was the House panel that investigated the pandemic response. Its reports matter because they gathered documents and testimony. But its stronger conclusions are congressional characterizations. That is the committee's characterization, and it should be kept in that box. But the NIH letter itself is enough for the basic fact. There was a reporting requirement. NIH said EcoHealth failed to report the finding right away. Now comes the phrase that turned this story into a years-long public fight. Gain of function.

Host: In May twenty twenty-one, Senator Rand Paul pressed Fauci at a Senate hearing about the Wuhan funding. This was the Senate health committee. The useful thing for this episode is not the cable-news argument that followed. It is the official hearing transcript. That transcript gives us Fauci's answer before the Tabak letter. Fauci's answer was categorical.

Document voice: The NIH and NIAID categorically has not funded gain-of-function research to be conducted in the Wuhan Institute of Virology.

Host: Five months later, the NIH letter described an experiment in Wuhan in which a lab-made bat coronavirus left humanized mice sicker. So was Fauci wrong? The answer depends on what someone means by gain of function. And that is the problem. In plain English, many people hear gain of function and think: an experiment that gives an organism a new or enhanced ability. By that everyday meaning, a lab-made virus making humanized mice sicker sounds like it belongs in the conversation. But the NIH was using a narrower regulatory frame. Tabak wrote that the research was not subject to review under the HHS P3CO Framework, because those bat coronaviruses had not been shown to infect humans. P3CO was the federal review framework for certain research involving potential pandemic pathogens. That phrase is clunky, so here is the listener version. It was a gate. If a proposed experiment fit inside that gate, it received extra departmental review. NIH's position was that this work did not fit inside that gate.

Document voice: Not subject to departmental review under the HHS P3CO Framework.

Host: That is not a minor footnote. It is the hinge. One side was arguing from the experimental result. The other side was answering under a regulatory definition. The same phrase carried two meanings. And the public was rarely given both at the same time. This show is not going to settle the label by shouting it louder. The record is stronger than the label. It shows the experiment. It shows the result. It shows the oversight tripwire. It shows the denial. And it shows why the denial depends on a definition most listeners never heard.

Host: Now the boundary. This record does not establish that this grant produced SARS-CoV-2. The NIH letter says the opposite. It says the viruses studied under the EcoHealth grant are not, and could not have become, SARS-CoV-2. It says even the closest known related viruses were far too divergent to have been the progenitor.

Document voice: Are not and could not have become SARS-CoV-2.

Host: That sentence matters. If someone tells you this grant proves the pandemic virus was created by this work, the documents do not support that. The record also does not establish that the gain-of-function label is settled. It is contested. The regulatory definition and the plain-English description point in different directions. The record does not establish that anyone intended harm. The project was framed as defensive spillover research. Surveillance and preparedness are real public-health goals. But the record does establish something narrower. And sturdier. It establishes that U.S. government funding flowed through EcoHealth to the Wuhan Institute of Virology. It establishes that the grant's own report described making chimeric bat coronaviruses. It establishes that one of those chimeras made humanized mice sicker. It establishes that NIH later acknowledged the result and the missed reporting requirement. And it establishes that public officials had answered the controversy using a phrase whose meaning was anything but clear.

Host: Episode one was about confidence. Private uncertainty became public certainty. Episode two is about scope. The public story was often small. A modest grant. Bat surveillance. No gain of function. But the grant paperwork tells a more specific story. Backbones. Spikes. Humanized mice. A survival gap. A reporting tripwire. And a definition fight hiding inside one phrase. The point is not that every loud claim was true. Many were not. The point is that the official simplification was not the whole record. And once you see that, the next question becomes unavoidable. What did they mean when they said gain of function? Next time on The Public Record: The Definition Problem.

Host: It is 10:32 on a Friday night. January thirty-first, twenty twenty. The new coronavirus has barely entered the public vocabulary. Its genetic code is online. Scientists around the world are staring at it, trying to understand what they are looking at. And one of those scientists sends an email to Anthony Fauci. Fauci, at that moment, was the longtime director of NIAID: the National Institute of Allergy and Infectious Diseases. That is the infectious-disease institute inside NIH. So if an American scientist had an urgent warning about a new infectious threat, Fauci was one of the people you would expect to hear about it. The scientist is Kristian Andersen, a virologist at Scripps. That matters because Andersen was not a pundit outside the field. He was one of the scientists looking closely at the new virus's genetic code in the first days of twenty twenty. He is not writing for the public. He is not making an accusation on television. He is writing privately, in the first days of a crisis, about something in the virus that worries him.

Document voice: Parts of the genome look inconsistent with expectations from evolutionary theory.

Host: That sentence is careful. It is also opaque. Four years later, under oath, Andersen was asked what he meant. His answer was much plainer.

Document voice: I think this could be engineered.

Host: Could be engineered. Not was engineered. But serious enough to reach the man who ran America's infectious disease institute. Three days after that warning, a draft paper was in Fauci's inbox. Seven weeks later, that paper would help shape what the public was told.

Host: From Fauci's Lies, this is The Public Record. Episode one. Three Days in February.

Host: The story begins with a document. Not a rumor. Not a theory. An email. On January thirty-first, twenty twenty, Kristian Andersen wrote to Anthony Fauci. Andersen had been looking at the new virus with other scientists. They were examining its genetic code, and especially the spike protein, the part that helps the virus enter human cells. Something about what they saw did not look right to them. For a general audience, the technical details can get dense very quickly. There is a phrase called the receptor-binding domain. There is another phrase called the furin cleavage site. Both became part of the origin debate. But for this episode, the plain-English point is enough. Scientists were looking at features of the virus. One helped it attach to human cells. Another was a small feature on the spike protein. And Andersen told Fauci those features raised a concern. Later, in twenty twenty-four, Fauci gave his own account of that night. He said Jeremy Farrar, then head of the Wellcome Trust, called him and told him he had just spoken with Andersen. The Wellcome Trust is a major biomedical research charity based in Britain. Farrar was not a random intermediary. In Fauci's later account, Farrar was the person who connected the alarmed scientist to Fauci and helped gather the early origins discussion. Farrar told Fauci he should talk to Andersen directly. Fauci said Andersen called him. And then Fauci described what Andersen said.

Document voice: There was something about it that looked like it could have been engineered.

Host: Fauci's testimony matters because it does two things at once. First, it confirms the warning. Second, it gives Fauci's defense. In his telling, a scientist raised an alarming possibility. Fauci did not say to hide it. He did not say to ignore it. He said it needed to be looked at more carefully. That is the beginning of the timeline. On January thirty-first, twenty twenty, Fauci was told the new virus looked like it could have been engineered. That fact does not prove the virus was engineered. It shows that the possibility reached him privately, at the very beginning. And in Fauci's account, the response came the next day.

Host: On February first, twenty twenty, there was a conference call. Fauci was on it. Francis Collins was on it. Collins was the director of NIH. That made him Fauci's boss in the federal research system. So when you hear Fauci and Collins together in this story, think of the top infectious disease official and the head of the larger biomedical agency above him. Jeremy Farrar was involved. Several scientists joined. The House Select Subcommittee later described the call as a place for Andersen to present his findings and discuss a path forward. Pause on that phrase: House Select Subcommittee. This was a congressional oversight body looking back at the pandemic response. Its reports and exhibits matter because they released documents and questioned witnesses. They are not court verdicts. And when the committee describes motive or intent, that is the committee's interpretation, not a fact this show adopts in its own voice. That description comes from the committee's released exhibits. So it belongs in the right box. The call happened. The broad purpose is described in the released record. But the committee also uses strong language in other places. Words about what prompted the paper, or what the scientists were trying to do. When those words belong to the committee, this series will label them that way. We do not have to rely on the committee for every part of the roster. In his twenty twenty-four testimony, Fauci was asked whether specific people were on the call. He confirmed several names. Ron Fouchier. Marion Koopmans. Peter Daszak.

Host: Fouchier and Koopmans were Dutch scientists with expertise in viruses and outbreaks. Daszak mattered because his organization, EcoHealth Alliance, had helped channel NIH money to coronavirus research at the Wuhan Institute of Virology. That does not mean Daszak said anything specific on the call. We do not have that transcript. It means his presence matters because the origin question and the funding question would later collide. But there is a limit to what this part of the record can tell us. We do not have a transcript of the February first call. We know it happened. We know some of the people Fauci later said were on it. We know a paper started moving soon after. But we do not know every sentence spoken on that call. So the call is a fixed point, but the conversation itself is still partly hidden. The documents pick up the story again in the days that followed. January thirty-first: the warning. February first: the call. And then, February fourth: the draft.

Host: The draft arrived fast. By the early hours of February fourth, at two oh one in the morning, a paper was in Fauci's and Collins's inboxes. The title was The Proximal Origin of SARS-CoV-2. That means the new coronavirus, the virus that causes COVID. From here on, we will mostly call it Proximal Origin. Proximal origin is a scientific phrase. It means the immediate origin of something: where the virus most directly came from before it entered people. In public debate, the title became shorthand for a much larger argument. The paper would become one of the most cited scientific arguments against a lab origin. But at this point, it was a draft. And the timeline was compressed. Three days from private warning to draft paper. Three days is the compression. And it matters because the question is not only what the scientists concluded. The question is how confident they were, and when. The public would later hear a strong conclusion. But the private emails in February show something more complicated.

Host: The center of the episode comes on February eighth. Four days after the draft paper reached Fauci and Collins, Andersen wrote another email. If you remember only one document from this episode, remember this one.

Document voice: Our main work over the past couple of weeks has been focused on trying to disprove any type of lab theory.

Document voice: The scientific evidence isn't conclusive enough.

Host: Those two pieces belong together. The work was focused on trying to disprove a lab theory. And the evidence was not conclusive enough. That sentence cuts in two directions. The work was pointed one way. The evidence was still short of confidence. That tension is the story. This is why the word confidence matters so much. Science often moves from uncertainty to confidence. That can be normal. That can be responsible. But when officials later use scientific confidence to shape a public debate, the earlier uncertainty becomes part of the public record too. Especially when the subject is not a minor detail. In early twenty twenty, the origin question was explosive. A lab origin could imply an accident. It could implicate research practices. It could implicate funding decisions. It could change how the public understood the pandemic. That made the handling of uncertainty important. A few days after the February eighth email, the same tension appeared again. The authors were communicating with a journal editor. And in that exchange, they were not calling a lab origin a conspiracy theory. They were saying the opposite.

Document voice: A serious scientific theory.

Host: That is how they described the possibility. A lab origin, they said, had to be considered as a serious scientific theory. They also said they wished they could rule it out, but given the data, they could not. The authors wanted to move away from a lab theory. That is visible in the private record. But they also told an editor that the evidence did not let them dismiss it. So by mid-February, the private record still contained serious uncertainty. Then, five weeks later, the public paper landed.

Host: Proximal Origin was published in Nature Medicine on March seventeenth, twenty twenty. Nature Medicine is a major scientific journal. The paper's authors were serious scientists. The paper was peer reviewed. For a public official, that matters. A published paper in a major journal can be cited as science, not just as one person's opinion. But journal publication does not erase the earlier drafts, emails, and caveats. It gives the paper authority. It does not make the private timeline disappear. And the paper made a real scientific argument. This point matters because there are two bad ways to tell this story. One bad version treats the paper as if it were just politics dressed up as science. The other bad version treats the paper as if the private uncertainty does not matter. The record is messier than either version. The paper looked at features of the virus. One feature helped the virus attach to human cells. The authors argued that this feature did not look like the most obvious design someone would choose in a lab. Another feature was a small part of the spike protein. It had drawn attention early. The authors argued that similar features can arise naturally. That is the technical argument, simplified. The authors were saying: this virus does not look like a clean engineering job. Nature can produce strange results. And the features that worried people can be explained without design. The paper also included a caveat. It said the exact origin could not be proved or disproved at that moment. That caveat matters.

Host: But on the specific question of a lab scenario, the paper used much stronger language.

Document voice: No laboratory-based scenario is plausible.

Host: Set the timeline side by side. February eighth, private email: the evidence was not conclusive enough. Around February twelfth, private correspondence: lab origin was a serious scientific theory. March seventeenth, public paper: no lab scenario was plausible. That shift is the heart of the episode. The shift is real. It is dated. It is documented. And it matters because of what happened next.

Host: On April seventeenth, twenty twenty, Fauci stood at the White House coronavirus briefing. A reporter asked about the possibility that the virus had come from a Chinese lab. Fauci answered by pointing to a recent study.

Document voice: A group of highly qualified evolutionary virologists looked at the sequences.

Host: He said the mutations were consistent with a jump from an animal to a human. He said the study could be made available. The study was Proximal Origin. That is why this moment matters for the series. The paper was no longer just a manuscript moving among scientists. It had become a public reference point for the country's most visible infectious-disease official. Here is the sequence. Fauci had received the January thirty-first warning. He had joined the February first call. The draft had reached his inbox on February fourth. And on April seventeenth, he pointed reporters to the published paper, from the most important public health podium in the country. The sequence is the point. Private warning. Conference call. Fast draft. Private uncertainty. Public paper. White House podium.

Host: Four years after the February emails, Fauci sat for two days of transcribed interviews before the House Select Subcommittee on the Coronavirus Pandemic. The transcripts run hundreds of pages. That format matters. This was not a television clip or a press conference. It was a formal congressional interview, with lawyers asking questions, witnesses answering under oath, and the transcript becoming part of the public record. It also matters because Fauci was able to give his own explanation. On the basic timeline, his testimony confirmed much of the record. He confirmed the January thirty-first call. He confirmed the engineering concern. He confirmed the February first conference call. He confirmed several participants. And he described Jeremy Farrar as the person who got people together. Fauci's framing was consistent. In his account, this was responsible science. A concern was raised. More experts were brought in. The evidence was followed. That was Fauci's account under oath. The documents add the other half. The public got a cleaner story than the private documents contained. The public heard confidence. The documents show uncertainty. That is why this is not just a fight about one paper. It is a fight about public trust. In a crisis, officials often simplify. They have to. The public cannot live inside every scientific caveat. But simplification has a cost when it erases uncertainty that later turns out to matter. Especially when the erased uncertainty concerns the origin of a global pandemic.

Host: So let us put the record in plain view. The record establishes that on January thirty-first, twenty twenty, a leading virologist told Fauci privately that the virus looked like it could have been engineered. That is supported by Andersen's email and by Fauci's later testimony. The record establishes that a conference call followed on February first. It establishes that Fauci and Collins were involved. It establishes that Fauci later confirmed several people as participants. It does not establish every word spoken on that call. The record establishes that by February fourth, at two oh one in the morning, a draft of Proximal Origin was in Fauci's and Collins's inboxes. That draft moved toward a natural-origin conclusion. The record establishes that on February eighth, Andersen privately wrote that the work had focused on trying to disprove a lab theory, while the evidence was not conclusive enough for high confidence. The record establishes that days later, the authors told a journal editor that a lab origin had to be treated as a serious scientific theory. The record establishes that on March seventeenth, the published paper said no laboratory scenario was plausible. And the record establishes that on April seventeenth, Fauci cited that paper from the White House podium. Now, just as important, here is what the record does not establish. It does not establish that SARS-CoV-2 was engineered. It does not establish that the paper was knowingly false. It does not establish that the authors failed to change their minds honestly.

Host: It does not establish that Fauci intended to mislead the public. That may sound unsatisfying. But it is the discipline the story requires. The claim is narrower than the loudest versions of the argument. It is also sturdier. And it gives us a rule for the rest of the series. When the documents support a claim, we will say it plainly. When the documents stop short, we will stop with them. When a committee, a scientist, or an official is making an interpretation, we will say who is making it. That rule is not decoration. It is the whole project. Because the public record around COVID is full of half-quotes, summaries, redactions, defenses, and accusations. The only way through it is to keep separating the dated record from the meaning people put on it. There was a documented distance between private uncertainty and public confidence. That distance was not abstract. It involved specific emails, specific dates, specific people, and a paper that shaped the public conversation. And that is enough to matter.

Host: The early pandemic was full of uncertainty. People were scared. Hospitals were preparing. Officials were trying to speak clearly while science changed by the day. That is the context Fauci and the scientists were operating in. But context does not erase accountability. If officials ask the public to trust them, the public is allowed to ask how uncertainty was handled. Not only whether the final answer was right. Also whether the public was told what kind of answer it was. Was it settled? Was it provisional? Was it a judgment call? Was there a live debate behind the scenes? In February twenty twenty, behind the scenes, there was still a live debate. The people writing the paper were still saying the evidence was not conclusive enough. They were still saying a lab origin had to be considered seriously. And then the public paper arrived with a much cleaner conclusion. Again, that conclusion may have been sincere. It may have been wrong. It may have been partly right. This episode does not settle the virology. It follows the confidence. And confidence is not a small thing. Confidence is what lets an official stand at a podium and say, in effect, this is what the science shows. Confidence is what lets newsrooms decide which theories are respectable. Confidence is what lets institutions decide which questions belong inside the bounds of normal debate. So when private confidence and public confidence do not match, the gap deserves scrutiny. That is what this series is about. Not certainty where there is no certainty. Not accusation without evidence.

Host: The public record. The paper is one thread. The money is another. Because while scientists were arguing over origin, another set of documents was sitting in grant reports and agency letters. Who funded what. Who reported what. And what NIH later acknowledged. In October twenty twenty-one, an NIH letter would say that an experiment it funded in Wuhan made laboratory mice sicker than an unmodified virus would have. That does not answer where COVID came from. But it opens the next question. What exactly had the United States paid for? Next time on The Public Record: The grant.