Mackenzie's Certainty Trough, Nuclear Missiles and "Science Abuse"

The MRG once again outclassed their rival UPERG in their second meeting of the semester. On deck was Donald Mackenzie’s “Nuclear Missile Testing and the Social Construction of Accuracy.”

The article represents an interesting case study demonstrating how uncertainty calculations for missiles became a chess piece in a political game where decidedly value-laden policy positions are justified in terms of technical arguments. For example, 1964 presidential candidate Barry Goldwater was a long-time skeptic of missiles, which have never been rigorously tested, and made arguments against their supposed accuracy an issue in his campaign. Mackenzie offers a variety of examples of how political factors drove efforts to establish the accuracy of the missiles.

Many know of Mackenzie’s article for his “certainty trough.” Mackenzie’s point is that people who are intimately connected with knowledge production often are more unsure about their knowledge claims than those who indirectly rely on that knowledge. Further, in the case of missiles, those who were alienated by ICBMs (ie supporters of increased spending on bomber craft as a primary US deterrent like Goldwater) felt that the uncertainty was much greater.

The continuum of the certainty trough could be more coherent. We are not measuring certainty in relation to one single independent variable (like familiarity or intimacy with knowledge generation processes). Though a certain type of proximity is surely still in the mix, advocacy positions for or against missiles based on uncertainty are due to political configurations (e.g. what branch of the military you work in, whether or not you’re running for president, your views on how to “win” the cold war, etc.). So, the observation that uncertainty has to do as much with politics as it does with scientific results is still valid, but the simple graph is a bit misleading.

I personally felt very unsympathetic to some of the later critics of missile accuracy, which made me question some of my sympathies with critics of general circulation models (GCMs). As I see it, there can be a variety of skeptical positions on climate models.

A) There are the extremists who view climate models to be senseless garbage, perhaps constructed at the whim of conspiracy theorists.
B) There are more moderate skeptics of a variety of flavors. Some think that the level of predictive accuracy for long-term predictions is impossible, others think that we can't properly understand the complex atmospheric interactions that could cause warming, etc.
C) Some, including myself, are skeptical that science can directly determine what policies should be done. (I might also be skeptical in some b-type ways; I don't believe in a sort of Laplace's demon, but all but the most extreme modelers don't either.)

With nuclear missiles, there seem to have been two types of skeptics who believed:
D) The missiles wouldn't effectively serve as "counter-force" to knock out enemy silos.
E) The missiles’ accuracy was overstated or difficult to clarify.

There are significantly different contexts: with the models on nuclear missiles, there were explicit political goals. Policymakers wanted missiles that would be able to knock out enemy nukes in their bunkers. Given Mackenzie’s history, it seems as though post-1960 there was a reasonable level of confidence that the missiles would be able to fulfill their task. Because there were specific goals for military leadership, the science could ‘found’ the policy. E-critics may have been right in certain aspects, but D wasn't effectively challenged (I could be convinced otherwise on this thesis, but I do think that eventually D became effectively true in an epistemic sense). Many of the critics of the missiles were intelligent political actors, but their actions were exclusively political and "an abuse of science."

With climate models, there is no explicit goal for their use. Even if the climate models could make perfect (or perhaps consensus) predictions about climate change, there is no clear sense of what should be done in consequence. Thus C-type criticism of GCMs is likely justified. However, I also think that climate models are dealing with much more complex systems than ICBMs (although ICBMs are themselves highly complex). Thus I don't feel inclined to say that type B skepticism is necessarily an abuse of science.

Complaining about abuse on science is often pointless and sometimes stupid (as is perhaps the case with the Republican War on Science thesis). Many science and technology studies folk buy onto the idea that science can't found policy.

But perhaps the more refined idea is that that mature science can found policy in politically uncontroversial areas. So, if missile calculations were based on mature science and were based on an issue that eventually became uncontroversial, then the later critics become ‘science abusers.’ GCMs are not based on as mature of a science and are controversial politically, so it's hard to understand what abuse could really mean.