The New York Academy of Sciences is hosting a panel discussion tonight about H5N1 influenza, “dual-use” research, and scientific censorship. Of course this stems from the ongoing debate about the alleged development of mammal-adapted H5N1 strains (see my earlier summaries here and here).
So far, I’ve found the public discussions on this issue disappointing, not only as a virologist and journalist, but also as a former debate coach. Both sides are advancing arguments, but there’s been a distinct lack of clash. In order to have a debate, each side has to listen to what the other is saying, and then respond directly, point-by-point, to the claims. Simply restating your own claims doesn’t cut it. If the arguments don’t clash, we can’t find out which ones are sound. Let me give some examples.
Members of the National Science Advisory Board for Biosecurity (NSABB) have made numerous public statements and published several essays about their recommendation to censor the new research. In all of those statements, they repeat the claim that H5N1 flu is a concern because it is highly lethal. In fact, they almost always quote a specific number: 59% of people infected with this virus die.
In blog posts and academic journal articles, virologists have repeatedly pointed out that this statistic is quite likely several orders of magnitude too high. The response of NSABB members has been simply to repeat the erroneous statistic. This is a crucial question that lies at the center of the debate, but there’s no clash.
Meanwhile, several commenters (including Howard Markel in an Op-Ed in today’s New York Times) have pointed out that redacting key data from the new papers, as the NSABB advocates, is a pointless gesture. As Markel explains:
In this case, censorship is too little, too late. The data generated by one of the research teams was already presented at a conference in Malta in September, where copies of the paper were distributed. But even if the data weren’t already available, the key details could likely be inferred from other information that is already available. I recently spoke with several prominent influenza scientists, all of whom agreed that, based on the knowledge that certain mutations can make H5N1 highly transmissible in ferrets, they could consult previously published literature and probably figure out what those mutations are.
In addition, none of this research was conducted in classified facilities. Hundreds of people without security clearances have already seen the data, which have also been emailed across multiple minimally-secured servers in at least four countries. If evil-doers wanted this information, it would be trivial for them to figure out which door to kick down to get it. The only people who actually can’t get access to the data are the scientists who might actually be able to use it beneficially.
The NSABB response to this criticism has been deafening silence. Again, it’s a crucial issue, but there’s no clash.
In the NSABB’s official policy statement explaining the rationale for their recommendations, we are simply told that they found the risks outweigh the potential benefits. For the past month, scientists have pressed the group, collectively and individually, to explain exactly what evidence they considered in reaching that conclusion. The policy statement, though, provides no more detail than their earlier editorials. So even when it comes to the central thesis of the argument, we have no clash.
At this point, I strongly suspect that the scientific backlash against the censorship recommendation took the NSABB by surprise. The group appears to have been looking for a test case on which to launch a discussion, and when the new H5N1 work came up they decided this would be it. Board members talked amongst themselves, made their recommendation on (probably inadequate) evidence, and expected the scientific community to go along peacefully. The scientists had other ideas. Having drawn a line in the sand, though, the NSABB now finds itself unable to retreat from it without looking foolish.
So here’s what both sides of the debate need to accomplish tonight. The virologists need to state their counter-arguments once more, but in the process they need to insist on direct responses from the NSABB. Don’t let them simply restate incorrect figures for the fatality rate, or just stipulate that redacting the data will prevent a terrible harm. Be nice, but press firmly. Don’t take “because we said so” for an answer. However, be sure you listen to the responses – it’s likely that NSABB members have some criticisms of the data you’re citing, and you’ll need to answer those as directly as you expect them to answer you.
The NSABB should begin by explaining their decision process, transparently and plainly, and acknowledging that it may have been flawed. Hey, we’re all people, and we all make mistakes sometimes. Next, address each and every one of the virologists’ counter-arguments directly. No dodging or pretending not to hear. Be prepared to concede any arguments for which you don’t have solid, evidence-backed answers. That will actually boost your credibility. Being wrong isn’t a character flaw. Being wrong while insisting you’re right is. On issues where you do have evidence, though, go ahead and press on. You had reasons for making the recommendation you made. Explain them. Finally, back off the fear button. Telling scientists that they have to agree with you or Congress will implement “Draconian restrictions on research,” or asserting (without solid evidence) that terrorists are on the verge of developing high-yield bioweapons, is fear-mongering. Stop it.