If a defendant refuses to disclose their whereabouts during the time of a crime, we have every reason to suspect that the defendant was at the scene of the crime until proven otherwise. In law, this inference is called “adverse inference”. The silence or absence of requested evidence from a concerned party suggests the evidence would be unfavorable to the concerned party.
Most of us learned adverse inference on the playground and parents regularly use adverse inference. If your kids are acting shady, holding something behind their back, say they’re up to “nothing” yet refuse to show you the hand behind their back, then you’re wise to assume they are absolutely up to something, that something is probably something they shouldn’t be doing, and you’ll probably learn what it is by looking at the hand behind their backs. In the GIF below, adverse inference is the inference we’d make if Bart Simpson refused to show his hands.
In the study of COVID origins, we have a lot of evidence. We have evidence of a grant in 2018 proposing to insert a furin cleavage site in a bat SARS-CoV infectious clone in Wuhan. One month after DEFUSE was rejected, the Chinese Academy of Sciences funded DEFUSE-like research in the Wuhan Institute of Virology. In the fall of 2019, 3 coronavirus researchers in Wuhan were so sick they reported to the hospital for symptoms consistent with COVID-19 (and other influenza-like illnesses). In January 2020, the world saw that SARS-CoV-2 emerged in Wuhan in 2019 looking like a bat SARS-CoV infectious clone with a furin cleavage site, highly unusual among wild coronaviruses and exactly what one would expect of a research product from the proposed work.
The Chinese government initially claimed the virus was zoonotic and originated in the Huanan seafood market (HSM). However, the Chinese CDC sampled the HSM and found zero animals out of 457 testing positive, and animal and wildlife vendors’ stalls were less likely to have SARS-CoV-2 than vegetable stalls and sewage. The outbreak emerged as a singular event in Wuhan, and not a more geographically widespread outbreak consistent with animal trade & food distribution networks. The Chinese government’s claims of a zoonotic origin were unsubstantiated, at best.
There is more evidence, but it starts to have diminishing returns in comparison to the potential information contained in evidence we still don’t have. The most important evidence we don’t have relates to coronaviruses in research labs in Wuhan, and the research activities of those labs. This evidence is not hidden in an animal in a far-off jungle or cave; it is in the possession of the Chinese government, yet the Chinese government refuses to share it. The refusal to share evidence necessitates we use adverse inference to learn from the evidence not shared.
Based on the actions of the Chinese government, the evidence not shared seems to be extremely important. Rather than share the evidence with the world, the Chinese government has instead proposed a shifting series of progressively more ridiculous narratives from zoonosis to India to frozen imported goods to Maryland. As the Chinese government desperately tries to deflect attention from the research & research products in its military biodefense labs, the CCP is caught lying about its military surveillance balloons, boorishly sending aircraft carriers to Taiwan, and more, all while refusing to share essential evidence on the origins of COVID-19 that could fit in an email. The world’s largest instance of “that could’ve been an email” would be comical if only it didn’t concern millions of deaths around the entire world.
With the evidence in the possession of the Chinese government, if SARS-CoV-2 did not emerge in a lab in China, then the Chinese government would know. Sharing this evidence would exonerate their labs and perhaps make their shifting narratives marginally less ridiculous. Yet, instead of sharing this evidence in their possession that we are asked to believe would exonerate their labs & government, the Chinese government creates a series of progressively more ridiculous alternative narratives. The narrative shifts in China are not random - they appear to be trying to blame their adversaries, effectively admitting that they know the origin of SARS-CoV-2 is a matter of great geopolitical consequence. The CCP seeks to blame adversaries while withholding evidence that could blame the CCP. The transparently ridiculous nature of these narrative shifts was laid bare when Chinese Academy of Science teams claimed COVID arose in India during the Chinese government’s 2020-2021 skirmishes with India. Now, while sending military balloons to Montana and naval ships to Taiwan, the Chinese government claims the virus originated in the US.
The world’s effort to get an honest account from China is starting to feel like parents interrogating a kid about missing cookies from the cookie jar. First, the kid says “what cookies?”. Then, the kid blames the (raccoon) dog, then the sibling they’re fighting (India), then the parents (the WHO and every other country in the world), all while refusing to let the parents see the chocolate stains and crumbs on their hands. Sadly, we’re not missing cookies, so this is not a silly game. We’re missing 18 million lives.
Beyond all the CCP’s smoke and mirrors, my eyes are fixed on the data not being shared by the Chinese government. Researchers in labs keep very detailed notes of their research and databases of their research products, and those notes and databases could help us rule out (or confirm) a Wuhan lab origin of SARS-CoV-2. The Chinese government without a doubt ran an audit of the labs. Yet, the audit, the databases, and the lab notebooks are all being withheld.
For completeness, it’s worth considering other evidence not shared. Researchers tied to zoonotic theories of SARS-CoV-2 origins claim that “China is hiding animals” in their own attempt to use adverse inference in support of a zoonotic origin. However, this claim falls apart when we imagine what would happen if the Chinese government sampled (or hid) animals under the two theories. If there were a zoonotic origin, and the Chinese government uncovered it by finding a progenitor like they did with SARS-CoV-1, the Chinese researchers would get global fame and the Chinese government would be in a position of credibility to connect with the world and better manage zoonotic viruses. Raccoon dogs are native to and legal in China, much like armadillos in America, and so claims that China’s trying to hide “illegal animal trade” don’t seem to hold up. If the Chinese government “hid” animals, then some Malaysian or Western scientists would probably find the progenitor of SARS-CoV-2 and get all the credit. Under a zoonotic origin, it’s not in the Chinese government’s interest to “hide animals” - they would have every incentive to sample animals & find reservoirs.
If, on the other hand, there were a lab origin, especially following bioengineering as appears likely, then sampling animals only makes the lab origin more obvious, incriminating Chinese labs and undermining the Chinese government’s efforts to deflect accountability. The more animals we sample without finding a progenitor, the more anomalous the lack of infected animals becomes, the stronger the evidence in favor of a lab origin. On the other hand, if the Chinese government knew of a lab origin, they’d know animals would test negative, and so refusing to test animals would keep our current level of uncertainty that allows them to sow doubt and disinformation about the origins of SARS-CoV-2.
Sampling animals doesn’t just provide evidence on the epidemiology of SARS coronaviruses in animals, it also provides evidence on the evolution of SARS coronaviruses. Under a lab origin, the more natural coronaviruses we find, the closer we’ll get to inferring the state of SARS-CoV-2’s pre-pandemic ancestors, the less evolutionary time there is for the series of unusual events in the SARS-CoV-2 genome, and the more obvious it would become that SARS-CoV-2 is not like natural coronaviruses, that it differs from its pre-pandemic ancestor in precisely the ways bioengineers modified the progenitor. Sampling wildlife viruses under a lab origin would focus our evolutionary microscope on this clade, revealing with greater focus how SARS-CoV-2 was engineered, so refusing to sample animals would prevent the discovery of closer relatives to SARS-CoV-2 and thereby prevent us from obtaining higher certainty about the bioengineering steps behind a lab origin.
Under a lab origin, sampling more animals would reveal the lack of SARS-CoV progenitors in animals and amplify the lab-like genomic anomalies of SARS-CoV-2. Consequently, if China is hiding animals and we try to ask “what are they hiding?”, then the refusal of the Chinese government to sample animals (if this truly exists) seems more consistent with a lab origin coverup than a zoonotic origin. A lab origin coverup is also more consistent with the databases, lab notebooks, and lab audits also being withheld.
Looking at the evidence not shared, there is a clear asymmetry in what that evidence may reveal, and what the Chinese government might be hiding. In all cases, under a zoonotic origin it’s in the best interest of the Chinese government to share this evidence. Under a lab origin, it’s in their best interest to not share it. The Chinese government knows what this critical information is, and so their refusal to share it is revealing. If the Chinese government’s audit of labs in Wuhan turned up no evidence of a lab origin they would believe a reservoir is still out there in China; for both national security and scientific glory, they would exonerate their labs and sample all the animals. If, on the other hand, the Chinese government’s lab audit revealed a lab origin, it has the clear incentive to withhold databases and lab notebooks, create a series of narratives blaming other governments to so confusion like a whoever farted blaming it on everyone next to them, throw up smoke & mirrors, not sample animals, and obstruct investigations.
We identify stars by the light they emit, and astronomers discovered the existence of black holes by the absence of light. When there is a concerned party suspected of a crime, their silence or refusal to share information that could shine light and exonerate them is highly suspicious, a black hole where something of massive importance appears to be. Where we have light and evidence on SARS-CoV-2 origins, we see strong evidence in favor of a lab origin. Where there is no light, the shadow is cast by the Chinese government withholding data and obstructing investigations There is a major asymmetry in this evidence not shared. The evidence not shared could disprove a lab origin and exonerate Chinese labs… or it could reveal a clear lab origin of SARS-CoV-2.
Only the Chinese government knows this evidence, yet they refuse to share it, allowing us to infer the unshared evidence points to a lab origin, not the exoneration of Chinese labs. The behavior of the Chinese government, from blaming India and the United States to obstructing international audits of its own labs, further suggests SARS-CoV-2 emerged in a Wuhan lab and the Chinese government knows it.
Hi Alex, nice post. To what extent would you apply this principle in general, especially with respect to so-called "conspiracy theories"? After all, countervailing narratives will almost always be harder to find and much weaker sourced than mainstream narratives due to funding, publishing, reporting, reputational and other incentives. How would you re-weight available evidence in light of this, if at all?
Interesting reading. As always, you lay out a well-argued and persuasive case. But I think you overstate your case. Let me tell you why, in a way intended less as criticism and more as a suggestion.
First, drawing adverse inferences is a tricky business. As Cicero is said to have said (but didn't): “Though silence is not necessarily an admission, it is not a denial, either.” As a logical matter, lack of evidence is not evidence. Silence is ambiguous. It doesn't prove anything one way or the other.
Drawing inferences from any piece of evidence is necessarily subjective, and that's doubly so for drawing adverse inferences from silence. That is an inductive process, not a deductive one. It is suggestive, not probative, and thus subject to abuse.
That potential for abuse was clear enough even back in 1789 that a prohibition was written into the Constitution against drawing adverse inferences from a criminal defendant's silence. To this day a judge will instruct a jury that it cannot draw any such adverse inference.
And that's a good rule, proven time and again. Demanding that a party produce evidence to prove their innocence, and then making an adverse inference against them if they don't, doesn't lead to truth and justice. Just the opposite. Truth and justice demands that a defendant be presumed innocent unless proven guilty.
That's why demanding that China produce documents to exonerate itself doesn't help the search for truth, but hurts. China might have valid reasons to want to keep its documents secret. Even if China did produce exonerating documents, those making the demands could just call them fake (which might well be true). Drawing adverse inferences is the argument of conspiracy theorists, not people looking for the truth.
Second, those who argue that a lab leak is the likely cause of the pandemic (and I am one) really do only have inductive evidence to rely on. There's little if any deductive evidence (what you call direct evidence) to support the lab leak theory. And that's a problem. Inductive evidence is only "consistent with" evidence, and to build a case you need "proof of" evidence.
That's an important difference, one similar to that in the scientific method. When using the scientific method you are free to use inductive evidence (statistics, probabilities, intuition, guesses, models, observations) to build your hypothesis, but you cannot use inductive evidence to test your hypothesis. You need deductive evidence to do that.
Third, does overstating your case really matter? I think so. Overstating your case (a) is counterproductive, (b) is unfair, and (c) destroys trust. If we demand that people provide evidence or respond to our claims, their response is rarely to do what we demand. Instead, they clam up (and hire lawyers). If we accuse people of causing the death of 18 million people but don't provide proof, we are not being fair. If we claim that we have proved our case when we can offer only inductive evidence, we destroy trust (especially if we make that claim as scientists or lawyers).
Much better to understate our case rather than overstate it. Instead of saying, "I'm 99% sure that the virus leaked from a lab", we might say, "I have seen a lot of evidence that is consistent with a lab leak, but it's hard to say I've seen any real proof". That approach will (I think) be more productive, more fair, and more likely to build trust.