I think you’re confusing several things here. I guess it would be great if no one lived near the San Andreas fault, but that ain’t in the cards. In the interim, earthquake research has contributed a huge amount to understanding very practical things about building design and soil behavior that get translated into practical benefits through better materials, building codes, zoning ordinances, and emergency planning. This research comes from scientists who are simultaneously advancing knowledge but doing so in the context of a practical problem. There are innumerable problems, from managing fisheries, to preparing for natural disasters, to improving water quality, to monitoring and preventing infectious diseases, to improving agricultural productivity, that demand and benefit from good science—often of a very fundamental nature, and usually carried out in the context of use. Of course scientific knowledge is not “the limiting factor” in dealing with complex social problems, but that doesn’t mean that new insights can’t improve our ability to deal with such problems. This is even more so because, given the huge limits on our ability (political and intellectual) to implement policies that deal successfully with complex social problems, sometimes the best we can do is improve incrementally on the back of better knowledge and technology.
So this leads to the next problem with your position, K. You’re lumping all science together as if it’s one thing, or should be one thing—scientists just doing whatever they want. But the failure of the “traditionalist” model that you defend is precisely that the model doesn’t acknowledge the context within which scientists work. Who pays for it is only part of the issue. As Daniel Engber’s mouse model article shows, it’s also about why and how problems are chosen, which is influenced not only by who pays, but by trends and fads, by peer pressure and disciplinary blinders, by today’s theories and long-standing institutional culture, by whose shoulders a scientist happens to be standing on, and so on. So, the mouse model problem is actually the opposite of what you suggest—it’s not that scientists are driven “down a path of short-term optimization, as in the case of mice and biomedical research [in an] attempt to link scientific advances with societal needs.” It’s that mouse models have become the way thousands of scientists spending tens of millions of dollars do what they are incentivized to do—publish, and get grants, and advance knowledge. This has nothing at all to do with whether the science is moving fast enough, or is closer to or further from potential utility. In fact, most animal model research is totally isolated from potential utility. On the other hand, pretty much everything we know about the links between science and the advance of practical problem-solving show the importance of institutional settings where scientific problems are pursued in the broad context of application, and where real-world practice provides feedback into the research process. In biomedical research, it is the strong connection between clinical practice, and the pursuit of scientific knowledge, that has led to much of the advance of medicine.
As for DOD, whatever DOD’s failures may be in delivering national security or spending money efficiently, in fact U.S. soldier casualties, both in absolute terms and, more importantly, in relation to enemy casualties, have immensely declined in combat over the last decades, and significantly because of technological advances from pilotless drones to better body armor to improved medical protocols for wounded soldiers. But the point I was really making had to do with the DOD’s central role in driving innovation in almost every major technological system that we’ve now come to take for granted, from computers and telecommunications to GPS and satellite imagery to aviation and advanced materials. In playing this role, DOD consciously created new fields of scientific inquiry, such as materials science and computer science, and funded enormous amounts of fundamental research at universities to create the knowledge base—and the scientists—that DOD needed to pursue its own agenda.
And I totally agree with you on the implausibility of Francis Collins’ “translational science” proposal—but perhaps for the opposite reason. You want scientists to be more liberated from any thought of application, and you seem concerned that Collins’ proposal will simply choke off scientific creativity. Yet you and Collins are really on the same side here. He believes, with you, that the benefits of science come when society reaches in and “translates” the findings made by scientists pursuing their creativity wherever it may lead. But the lessons of real-world, everyday science are quite clear: scientific creativity and real-world problem-solving are both at their best when they can feed off of each other.
Where we both agree is that all the silliness about unprecedented rates of scientific discovery and innovation seems like some combination of self-congratulatory hype and fatalistic voodoo. Apparently, the number of scientific journals has been growing exponentially since the 17th century, a 300-year precedent. But what’s the rate of production of knowledge and innovation that can make crucial contributions to our well-being and future prospects? That would be the important question, and no one has figured out how to ask it in a meaningful way, let alone provide a meaningful answer.