It’s not about getting “the scoop”, it’s about getting it right | Origin of COVID-19: My emails with former NYTimes reporter Donald McNeil
In a recent book, former New York Times reporter Donald McNeil falsely accuses several subject matter experts, myself included, of having misled him in early 2020 on the possibility of a SARS-CoV-2 “lab leak”.
A “lab leak” origin of the COVID-19 pandemic has always been possible and in early 2020, given a paucity of data, was a reasonable and scientifically justifiable (even, necessary) hypothesis. However, given current evidence, there is not a single, consistent, hypothesis supporting the “lab leak”, and McNeil makes the mistake of mixing up different ideas, timelines, and the nature of evidence-based science.
I will dive into more details below, however, let me make it clear that I mentioned several times in my emails to McNeil that a “lab leak” was a possible, even plausible (at the time), hypothesis for the origin of the pandemic, including stating on February 7, 2020:
“The question about potential lab escape is not necessarily a crackpot theory though and should always be taken seriously and investigated scientifically.”
“As I mention in my previous email, many less subtle scenarios (e.g., culture) would be expected to look much the same as spillover, so it’s going to be very. hard/impossible to distinguish between those sorts of scenarios.”
And on February 14, 2020:
As to your question — it’s impossible to rule out everything here. There’s nothing to suggest a bio weapon (somebody would have used SARS or MERS — not a novel bat virus). As for lab escape, as I mentioned in my previous email, that scenario would likely look like natural selection, so it’s impossible to assess at this stage. What I can say is that the data is fully consistent with natural selection with bats being the ultimate reservoir. Again though, a lab escape might look much the same.
Ultimately, our “Proximal Origin” paper (published on March 17, 2020; preprinted on February 17, 2020), as well as later studies, ended up providing evidence for a natural origin of SARS-CoV-2, while never excluding the possibility of a “lab leak”. Our earlier conclusions have therefore only been further supported by evidence over time.
You can download my emails with Mr. McNeil here.
COVID-19 “origin” hypotheses
Before diving into the emails, some background on the origin hypotheses for the COVID-19 pandemic and my changing views on them.
A few points to make clear:
- Our knowledge in early 2020 was not the same as our knowledge today — back then, we knew little about the virus, the disease, cases, and the ensuing pandemic.
- While I have investigated the emergence, evolution, and spread of many viruses throughout my career, in early 2020 my knowledge of coronaviruses was limited. Several of my co-authors had more extensive expertise on these viruses, however, and my own knowledge of coronaviruses rapidly increased.
- We were actively investigating a scientific question, with our views continuously being updated by additional data, analyses, and a general increase in our own understanding of coronavirus evolution. This meant that our views on likely origin scenarios changed over time.
One should not confuse “the possibility of a lab leak” with “the probability of a lab leak”, with the latter being an informed version of the former. While a lab escape remains a possible scenario for how SARS-CoV-2 originated, given current knowledge, I do not find it probable. A view that is shared by a majority of experts.
The “lab leak” is not a consistent hypothesis, but, rather, a collection of unsupported conjectures that are all unable to reconcile the consilience of evidence we have from the early outbreak in Wuhan and the emergence of SARS-CoV-2. Further, while this is not true for all “lab leak” conjectures, most of them are mutually exclusive and conspiratorial in nature.
From a hypothesis generation standpoint, considering a potential “engineering” origin of SARS-CoV-2 is not the same as posing conjectures about “bioweapons” or a virus “engineered with intent/purposefully manipulated” (i.e., the de novo creation of SARS-CoV-2 with the goal of creating a virus with a specified phenotype — scientists simply do not have the ability to do this). In a recent congressional testimony, I laid out these hypotheses in more detail and summarized here in a slide from a talk:
The science of investigating COVID-19 “origin” hypotheses
In the early days of the COVID-19 pandemic, our understanding of the emergence of SARS-CoV-2 was based on limited data and knowledge. For example, in a February 14, 2020 email to McNeil I noted, “…it’s impossible to rule out everything here. … lab escape […] would likely look like natural selection, so it’s impossible to assess at this stage”. These statements were based on available data and my understanding at the time, and reflect the scientific method’s inherent uncertainty in the face of emerging evidence and analyses.
In my emails, I also emphasized that our conclusions were always tethered to the latest available evidence. Our skepticism regarding a lab escape was not a dismissal, but a reflection of a scientific view supported by the data and analyses presented in our scientific publications. For example, in Proximal Origin we stated (March 17, 2020):
Although the evidence shows that SARS-CoV-2 is not a purposefully manipulated virus, it is currently impossible to prove or disprove the other theories of its origin described here.
More scientific data could swing the balance of evidence to favor one hypothesis over another.
Similarly, in our “Critical Review”, we stated (September 16, 2021):
… there is substantial body of scientific evidence supporting a zoonotic origin. Although the possibility of a laboratory accident cannot be entirely dismissed, and may be near impossible to falsify, this conduit for emergence is highly unlikely relative to the numerous and repeated human-animal contacts that occur routinely in the wildlife trade.
It is crucial to understand that in science, opinions are not static but evolve with new evidence and an increasing understanding of a scientific topic. My email exchanges with McNeil in February 2020 are examples of that.
For example, my comments on pangolin CoV sequences supposedly being “99% identical to nCoV” were based on preliminary reports, which later turned out not to be true. All of this is clear in my early statements to McNeil and when I stated “it’s likely that they [pangolins] might have served as an intermediate host”, I also caveated this by stating “These numbers have been discussed on Twitter and (I think) in various news outlets — however, without the data it’s difficult to conclude anything from them as it’s difficult to know what they mean”. These are all examples of science in action — a process of building knowledge piece by piece and, during a rapidly evolving crisis, under conditions of significant uncertainty.
Our initial hypothesis was that SARS-CoV-2 could have been an engineered virus. That, however, was short-lived and my view on this — as has already been reported and is clear from our published papers, as well as released emails and Slack messages — changed in just a short period of time.
By the time McNeil contacted me (February 6, 2020), we had largely discarded the idea of “engineering” (and barely considered e.g., “bioweapon”, “intentional engineering”, and “RaTG13 descendent“ worthy of further scientific inquiry, as they can all be quickly dismissed). However, at the time, I still believed that SARS-CoV-2 could have been the result of culturing an intact virus isolated from e.g., bats. All of this is clear from my responses to McNeil’s questions (see, below) and can also be seen from an early summary that we wrote in the evening of February 1, 2020:
Early correspondence
McNeil first contacted me about SARS-CoV-2 on February 6, 2020. He also talked to other scientists at the time, including Dr. Richard Ebright from Rutgers University. Here is a copy of his first two emails (note, the time stamps keep switching between Eastern and Pacific time throughout):
At the time McNeil contacted me, myself, Eddie Holmes, Andrew Rambaut, and Bob Garry (with later participation from Ian Lipkin), had just initiated our scientific investigations into plausible scenarios for the origin of SARS-CoV-2 (among many other things related to the pandemic — most of my work was, and continues to be, focused on response and broader infectious disease research).
As has been widely reported, our initial hypothesis was that SARS-CoV-2 could have been the result of a laboratory escape from the Wuhan Institute of Virology. As later described in our Proximal Origin publication, however, that initial hypothesis did not hold up to scientific scrutiny and we instead concluded that SARS-CoV-2 was very likely the result of a mundane zoonosis associated with the (illegal) wild-life trade in China.
I replied back to McNeil:
“unless the virus was a really obvious recombinant virus, a virus from culture vs an intermediate host would probably be indistinguishable.”
“While the RaTG13 bat sequence is interesting, it still too divergent from nCoV to have anything to do with the current epidemic — the genetic distance is simply too great.”
“From a genomics perspective, the theories Richard Ebright lay out would likely look the same — there would be no way to distinguish between them”
Note, in retrospect, several of these (and other early 2020) comments are premature, reflecting my knowledge at the time. For example, we can distinguish “culture” vs “intermediate host” and there are many other data points to consider outside just SARS-CoV-2 genomics. As above, these are examples of science in action.
McNeil followed up:
To which I responded:
“The question about potential lab escape is not necessarily a crackpot theory though and should always be taken seriously and investigated scientifically.”
Reemphasizing my previous point related to a potential lab escape:
“As I mention in my previous email, many less subtle scenarios (e.g., culture) would be expected to look much the same as spillover, so it’s going to be very. hard/impossible to distinguish between those sorts of scenarios.”
For context, below are the emails from Dr. Ebright to McNeil. I agree with much, if not all, of this.
Follow up correspondence
McNeil followed up on our earlier correspondence on February 14 (left):
In my reply (right, above), I continued to emphasize my earlier points about a potential lab escape (I, if not necessarily all my co-authors, still believed “culture” was a plausible hypothesis at the time).
These responses are fully consistent with the conclusions in our preprint of Proximal Origin just a few days later. Following additional analyses, knowledge, data, and peer review, those early conclusions were further updated in our published paper.
Below, our statements in the preprint (left; February 17, 2020) and published version (right; March 17, 2020) of Proximal Origin:
Correspondence following preprint
After we had released the preprint of Proximal Origin, McNeil sent me additional emails, now talking about “spooks” and “national security” and wanting personal information to “add[s] color” to his story.
“Could we talk about that? Not just about the science, but what went on in your head and in your life while this was unfolding?”
Anything that adds color to the story would be useful.
This email made me uncomfortable. First, I did not have any insights into “spooks” and “national security”, and second, I strive to stay focused on the science. I also made this latter point in my private correspondence on Slack just a few days prior:
While some may disagree with this approach, I feel that it is an essential component of science communication to avoid unnecessary speculation to help ensure accuracy. This is reflected in my response to McNeil, where I point him to our preprint, state our main scientific conclusions, and make it clear that we can loop back on this topic once a final paper has been peer reviewed and published. I shared this response with my colleagues in private communications and McNeil uses this message, out of context and with no reference to our prior communication, to falsely accuse us of misleading him.
As would have been clear to my colleagues to whom I was corresponding, my comments about:
… deflect the fact that I’m dismissing him
Can’t ignore him and can’t just give him the scientific story — that would only lead to follow up question.
… and I should add — I really fucking wished my life wasn’t this exciting…
were clearly in reference to the “personal” aspects of his request, and not the scientific questions, which I had already answered. He also quote mines a comment from our private communications on February 6, 2020 out of context where I state:
“And for Don — I gotta say, he pretty much nailed it. Let’s not tell him”
As is clear from my the emails, I did in fact “tell him”. I explained to him the various origin scenarios — whether “lab leak” or zoonosis — and explained to him my initial interpretations of the data available at the time. None of this is misleading — it is focusing on the science when communicating with journalists on topics that are actively being investigated, with insights still developing.
McNeil followed up with another email even more focused on personal aspects of my life (those first two sentences are on-point, by the way):
Correspondence following publication
In my February 19, 2020 reply to McNeil I stated that “We are hoping that all of this will be finalized within the next couple of weeks, so happy to loop back with you once all of that is complete”. Once the final paper was published, I followed up:
In this email, I specifically included an invitation for McNeil to get some more background on the paper and the process, given his earlier interest outside “just about the science”:
There’s of course more to the story than just the paper and I would be happy to go into some of this if there is still interest.
With a peer reviewed paper published—laying out the scientific questions and our conclusions — I felt that I had an appropriate reference point to discuss some of the background on how the paper came together.
As is clear from his reply, however, he did not show any interest in following up:
Correspondence about McNeil’s book
After the release of our private communication by the Select Subcommittee on the Coronavirus Pandemic, McNeil contacted me and my colleagues, making several false accusations, including that we misled him:
Among other things, McNeil stated:
… come up with answers that, while technically accurate, were very misleading in that you clearly thought — at least for a while — that there was a good chance that there had been lab engineering — a furin cleavage site insertion and/or passaging in human cells or humanized mice. You jointly decided to omit any mention of that […]
This is obviously false. As explained above, by the time McNeil contacted me, we had already dismissed several of these “lab leak” ideas, with “culture” being the hypothesis we were still seriously considering. I also explained to him why we did not consider other potential “lab leak” scenarios to be probable. This was mentioned several times in my emails and there was no “omi[ssion]” of any hypothesis that we were still seriously considering. So my answers were not “technically accurate”, they were just “accurate”.
I explained this to him in my reply, also mentioning:
my responses to you, including the one that prompted you to reach out again, reiterated what we published.
accuracy in reporting is important, and speculating on still-changing early hypotheses that later ended up being wrong, would have been reckless.
I see this type of speculation as being reckless and I did not want to talk about my personal “life” to a NYTimes reporter during a very stressful part of my life. My response to you therefore simply reiterated the science and our conclusions at the time (which still stand). This has nothing to do with “misleading” anybody — it’s making sure the science is clear.
speculation based on hypotheses we had already dismissed at the time you contacted me, is reckless.
In his reply—having just written a chapter in which he falsely accused subject matter experts of misleading him — McNeil made it clear that he had not read the original emails from more than two years prior, before rewriting his chapter:
This is shocking behavior from any reporter and the opposite of what one would (should) expect from a journalist interested in truthful reporting. More emails followed, with a new layer of misrepresentation:
McNeil lamented that we had not answered his early questions:
Is there any possibility that: it could be from the Wuhan lab? And, if it was — — would there be any way to tell? … Is there anything in the sequences posted so far that suggests the virus has been manipulated by human hands in any way? (Sequences from another virus inserted, deletions that seem unlikely to occur in nature, anything like that?)
As described mentioned above, on February 6, 2020 I had told him:
unless the virus was a really obvious recombinant virus, a virus from culture vs an intermediate host would probably be indistinguishable.
And on February 7, 2020:
“The question about potential lab escape is not necessarily a crackpot theory though and should always be taken seriously and investigated scientifically.”
“As I mention in my previous email, many less subtle scenarios (e.g., culture) would be expected to look much the same as spillover, so it’s going to be very. hard/impossible to distinguish between those sorts of scenarios.”
And on February 14, 2020:
As to your question — it’s impossible to rule out everything here. There’s nothing to suggest a bio weapon (somebody would have used SARS or MERS — not a novel bat virus). As for lab escape, as I mentioned in my previous email, that scenario would likely look like natural selection, so it’s impossible to assess at this stage. What I can say is that the data is fully consistent with natural selection with bats being the ultimate reservoir. Again though, a lab escape might look much the same.
Those are exact answers to McNeil’s questions. Which, after thorough and collaborative scientific investigations, were followed up with several peer reviewed publications.
McNeil also mentioned what he would have considered “a fully honest answer”. As explained above, and re-emphasized in my many emails to him, the fact that I did not give him the ‘perfect’ answer that he, given 20/20 hindsight, would have wanted at the time does not make my responses misleading — it just makes them, well, different.
I explained that in my follow-up, including stating:
20/20 hindsight is an amazing thing and as I have already stated, throughout our conversations I have always given you thoughtful and scientifically justified answers. Not just that, but I have gone to great lengths to ensure that my answers to you have been accurate and, I hope, helpful, and spent many hours answering your questions in emails and on the phone, including during the most stressful and busy times of my entire career.
.. there was very clear acknowledgement of a potential lab leak, reflecting my understanding at the time.
.. there is a very big difference between bouncing ideas around with colleagues as studies are ongoing versus openly engaging in fact- and evidence-free speculation in the NY Times, before actually having completed the required studies (which were ongoing at the time).
Further, McNeil complained that we did not mention“We’ve alerted the American and British governments about our worries [and] have discussed it with government officials..” As I stated in my reply to McNeil, none of this was ‘secret’ and was reported at the same time McNeil contacted me, including my involvement. So I did not “omit” this information in my communication with McNeil — it was already public knowledge at the time and a quick Google News search would have revealed as much.
It’s not about getting “the scoop”, it’s about getting it right
The final two emails in the thread serve as a stark contrast — McNeil wanted “the scoop”, while I was (and continue to be) focused on “getting it right”.
The ‘failure’ of McNeil for getting “the scoop” in early 2020 on what remains unsupported conjecture, fueled by increasingly more outlandish conspiratorial beliefs, had little to do with the answers he received from me and other scientists.
As I said in my reply, “It is fair to disagree on what you think I should or should not have said to you at the time, however, it is a long way from there to accusing me, and my colleagues, of being dishonest, which is exactly what you are doing in your draft chapter.”
In conclusion, I will simply reiterate what I also said to McNeil:
This isn’t about getting “the scoop” — it’s about getting it right. If you balance my early answers with the large body of evidence we have on the origin of the pandemic, you will see that they have held up remarkably well.
As for the larger question of how the COVID-19 pandemic began, all the evidence points in one direction — we are to blame.
Not because of research activities, which had correctly predicted the risk posed by SARSr-CoVs, but because of our unregulated and dangerous trade in wild animals, which we bring into large population centers — an immense lab performing millions of ‘gain of function’ experiments on a daily basis under BSL-0 conditions.
Despite the many lessons of the pandemic, we have done little to counter this, leaving us especially prone to future pandemics caused by novel influenza viruses, coronaviruses, and the many other viruses out there.
Given population expansion, deforestation, climate change, and other human activities, such emergence events will very likely occur with increasing frequency.
We are not ready. Scapegoating the very people actively working to make us better prepared will not help.
References
Our peer reviewed “origin” research
Andersen KG, Rambaut A, Lipkin WA, Holmes EC, Garry RF
The Proximal Origin of SARS-CoV-2
Nature Medicine 26(4):450–452 (2020)
Preprint
Holmes EC, Goldstein SA, Rasmussen AL, Robertson DL, Crits-Christoph A, Wertheim JO, Anthony SJ, Barclay WS, Boni MF, Doherty PC, Farrar J, Geoghegan JL, Jiang X, Leibowitz JL, Neil SJD, Skern T, Weiss SR, Worobey M, Andersen KG, Garry RF, Rambaut A
The Origins of SARS-CoV-2: A Critical Review
Cell 184(19):4848–4856 (2021)
Preprint
Pekar JE, Magee A, Parker E, Moshiri N, Izhikevich K, Havens JL, Gangavarapu K, Malpica Serrano LM, Crits-Christoph A, Matteson NL, Zeller M, Levy JI, Wang JC, Hughes S, Lee J, Park H, Park MS, Ching KZY, Lin RTP, Mat Isa MN, Noor YM, Vasylyeva TI, Garry RF, Holmes EC, Rambaut A, Suchard MA, Andersen KG, Worobey M, Wertheim JO
The Molecular Epidemiology of Multiple Zoonotic Origins of SARS-CoV-2
Science 377(6609):960–966 (2022)
Preprint
Worobey M, Levy JI, Serrano LM, Crits-Christoph A, Pekar JE, Goldstein SA, Rasmussen AL, Kraemer MUG, Newman C, Koopmans MPG, Suchard MA, Wertheim JO, Lemey P, Robertson DL, Garry RF, Holmes EC, Rambaut A, Andersen KG
The Huanan Seafood Wholesale Market in Wuhan Was the Early Epicenter of the COVID-19 Pandemic
Science 377(6609):951–959 (2022)
Preprint
Selected podcasts with more background on our work
Interview with Worobey, Andersen & Holmes: The Lab Leak
Decoding the Gurus, March 2023
Eddie Holmes: Zoonotic Diseases, China, Fauci, and the Lab Leak Hypothesis
Bad Boy of Science, December 2022
The Origins of COVID-19 (with Kristian Andersen)
In the Bubble w. Andy Slavitt, September 2022
Is The COVID Lab Leak Hypothesis Dead? With Kristian Andersen, Mike Worobey & Philipp Markolin
Bad Boy of Science, July 2022
