A correspondent of mine, Prajwal Kulkarni, is a physicist who is a fellow of the American Association for the Advancement of Science (AAAS) at the US Environmental Protection Agency. He is working on a submission to the 2011 annual meeting of the AAAS and asked me for some comments from an HPS perspective, which I originally posted at The Galilean Library. Here is a substantial excerpt from his proposal:
This symposium discusses how contrarians have exploited the public voice of science for their goals. Panelists will examine how public discourse gives all scientists the authority to speak for science, inevitably conferring credibility to skeptics. This false unity of science makes it difficult to argue that only climate scientists should be trusted on ACC [anthropogenic climate change]. Rather than portraying a single image of science, emphasizing its diversity might better communicate ACC. Panelists will also discuss how the collective message from scientists and their institutions may unintentionally foster denialism in two ways. First, an image of science as providing absolute answers and positive proof makes it difficult to describe uncertainty and doubt. Second, an implicit endorsement of naive falsification as the model of scientific advancement implies that theories are discredited with a single incorrect fact.
To help situate all this, the following quote is from a paper by Galison called Ten Problems in History and Philosophy of Science, in which he made a link between climate skepticism and doubt:
The debate over the dangers of smoking tobacco continued for a long time, but industry took pride—in secret, of course—in boosting that uncertainty: “Doubt is our product” became a guiding slogan of the tobacco lobby. “Teach the controversy”, militated the Creationists and Intelligent Designers, demanding that their version of the beginning of things be taught alongside Darwin. The “global warming controversy” became a major political talking point for those bent on blocking international accords. But this appropriation of doubt, this use of doubt as a weapon, raises difficult questions for HPS: What is controversy? What is scientific doubt in a world where it can no longer be treated purely as an offshoot of this or that scientist’s work? What role does HPS have in handling such matters, when one of the standard means of research – examining controversy – would itself reiterate and reinforce one side in a political confrontation with major consequences?
The implication here is that the inevitability of doubt in science allows critics of theories to drive in an epistemic wedge, so to speak. Methodologically, and other things being equal, this should be a good thing: regardless of how highly-confirmed a theory is, there are always grounds for the development of alternatives; if this were not so, theories would never change or be overturned. The question is: when does this incompleteness, coupled with a focus on anomalies for a given theory and an insistence that it may be wrong, become abuse – using our uncertainty to delay or undermine theories rather than developing alternatives?
We can argue from the history of science that theories considered hopeless in the past proved to be worthwhile much later, so it may be a bad idea to limit science to only those theories with something to recommend them now. Similarly, we can argue from both the history of science and the philosophy of science that there are good reasons to proliferate research programmes and not define science restrictively via a particular (simplistic) methodology [of which more in a subsequent blog entry]. However, in doing so we leave very little room to make the kind of claims we typically want to make with regard to research programmes, such as that x is a hopeless notion with no support whatsoever or that y is being supported or opposed by contrarians because they do not like the consequences, and so on.
The notion that doubt can be (mis)appropriated seems problematic because it turns partly on the tension between expertise and the ability to discuss science. If we want a model of science that is philosophically sound then doubt and its abuse (assuming, for the moment, that it makes sense to call it abuse) is inevitable; if we prefer instead to limit the harm (same caveat) that doubt and malicious skepticism can do, we need a more simplistic model of science. Perhaps we typically have both: when it suits us, science is the complex tale we tell ourselves, which accords with the results of philosophical critiques and comes somewhere close to matching the empirical reality of doing science; when it does not, we rhetorically treat matters as far more straightforward and use terms like consensus and denier to try and combat people we believe have no genuine interest in science. This latter reaction typically involves arguing a very simplistic philosophy of science and resorting to method-talk, requiring that science be defined via the mythical scientific method; however, the holes in this approach make it easy for critics to shoot them down.
The (mis)appropriation of doubt comes in here: if science is supposed to be characterised in a particular way and critics of a well-confirmed theory are able to show this model to be flawed, do they therefore lend credibility to alternatives? Logically speaking, they plainly do not; similarly, engaging with these same critics does not grant their ideas greater validity, no more than counting the number of scientists forming a consensus on any side can imply anything about the truth (or verisimilitude) of a theory. This is basic logic, but of course it works wonderfully in rhetorical terms. The trouble is that by resorting to this approach and trying to win public debates by referring to political phenomena like a consensus, it leaves opponents with an obvious course of action in response. This is why we see skeptics seizing upon errors in the IPCC report; after all, if we are going to reduce the validity of scientific theories to talk of a consensus then we can hardly complain if critics avoid discussing science and restrict themselves to our simplistic model of science instead.
For my part, I have argued in the past that this is a mistake for scientists because it pushes what should be empirical debates (at least in part) to philosophical areas. Framing the behaviour of critics as “denialism” is also problematic: if we are going to make any claim to doing science then we have to acknowledge the existence of gaps and that evidence can persuade but never remove doubt or make conclusions incontrovertible. Since everyone accepts that science is always incomplete and that skepticism is a vital part of it, how can we object to this skepticism in action? I propose an answer below.
Connecting these comments to Praj’s proposal, it is probably a good idea to emphasise that diversity is preferable to advocating unity. This is especially pertinent to ACC since it is straightforwardly the case that there is no unity, notwithstanding the consensus, due to the criticisms and research of scientists like Lindzen. Moreover, when errors inevitably come along, are reported and are buttressed by comments from plainly well-qualified skeptics, they can destroy the consensus in the public imagination since the unity claimed is shown to be false, even if it turns out that the errors can be corrected and the critics are wrong. If we were talking about science in terms that advocated diversity and recognised limitations of science then these errors and the existence of well-qualified skeptics like Lindzen would be expected, rather than anomalies. Involving certainty and unity in our rhetoric about science make it difficult or impossible to assimilate the doubt, uncertainty and errors that are an inevitable part of scientific practice.
The more simplistic our philosophy of science, the quicker it collapses on inspection, making our certainty and claims of unity look like political manoeuvring, which is probably a fair description of how many laymen now view subjects like ACC where the economic consequences are easier to understand than the justifications for policies to address ACC. A case like Intelligent Design, wherein it is argued that religious motivations are involved, is still more difficult in these terms and deploring scientific illiteracy does not really address any of this. Here we see that Praj’s point about naïve falsificationism is crucial: if we advocate falsification as defining scientific method, as many scientists and non-scientists alike still do, a single counter-example falsifies a theory. Falsificationism does not work, for a variety of reasons (see here for a discussion of the subject), but there is still the unintended consequence: if we talk in these terms about falsification, a single error or a single qualified expert arguing against a theory can suggest that it has been defeated. Attempting to redress the perception of damage after the fact may work to an extent but perhaps it is the model of science – falsificationism – that causes the problem, rather than whether the skeptics’ arguments actually falsified the theory or not.
What, then, is the difference between someone who seizes on doubts to develop a new theory and someone who is merely a contrarian or else actively opposes a theory because of its perceived consequences? For my part, I think there is another way to look at science, which may help avoid these pitfalls (although I don’t think science can ever be divorced from politics; indeed, the possibility that pure science is impossible – if not always then particularly under the institutionalised approach employed today – might be something else worth considering.) This approach involves accepting that values play a vital part in science and I discussed it previously when looking at the “distinction between ‘respectable’ people and cranks”.
This perspective reveals the difference between someone like Lindzen and ACC “denialists” (if we accept that term): Lindzen has his doubts about ACC and is actively working with them, even if he may ultimately be wrong, whereas a “denialist” offers only negative arguments, or else seizes upon mistakes and emphasises them instead of developing a counter-theory. Importantly, we have to also apply this distinction to the other “side” and identify those who are developing ACC and those who are pushing it rhetorically beyond where the science supports their claims, which is an important dialogue that Judith Curry is involved in, amongst others. I would say that all talk of a consensus is an example of this rhetorical pushing; another example is demanding that only suitably-qualified scientists can speak credibly about ACC, which immediately runs into the “Lindzen problem” and leaves us counting numbers, which is – or should be – obviously a ridiculous way to conduct scientific debate. This is not to impugn anyone’s motives: clearly both the issues of environmental catastrophe and economic development are important to a lot of people. The values-based approach is clearer still when applied to debate about Intelligent Design, in which the participants are also inspired by deeply-held convictions and principles. It does not answer the question of to whom and how to assign research funding, but I think this is a far more general difficulty due to science outgrowing its self-image and origins.
The related issue of expertise is an interesting one, which I hope to pursue further at some stage. If everyone has to possess specialist knowledge in order to be permitted to comment on a scientific theory, and if science has become so deeply specialised that even scientists in the same discipline can barely speak to one another, then perhaps no one can really comment on anything? In particular, interdisciplinary projects like the IPCC could never work. (It would also suggest that scientists are not allowed to comment on the philosophy of science either…) This may even be a new example of the incommensurability problem, in which we have to translate expertise (instead of scientific terms) in order to say anything about a larger theory. I will come back to this but the question remains: is this the kind of science – and political discussion of science – we really want?