by Judith Curry
The main intellectual fault in all these cases is failing to be responsive to genuine empirical concerns, because doing so would make one’s political point weaker or undermine a cherished ideological perspective. – Heather Douglas
I have spoken often and publicly about my concerns about the integrity of climate research. When I have used the words ‘integrity of research’, I have been referring generally to the adherence of the Mertonian norms of science and a general sense of ‘trustworthiness’.
The role of values in scientific research, and whether research is value laden or should be value free, is a subject of extensive debate. A perspective on all this that makes sense to me is that provided by philosopher Heather Douglas.
As of late, the term “scientific integrity” has been used as an overly broad slogan encompassing everything good in research ethics. In this paper, I provide a more precise and narrow account, where scientific integrity consists of proper reasoning processes and handling of evidence essential to doing science. Scientific integrity here consists of a respect for the underlying empirical basis of science, and it is this scientists are often most concerned to protect against transgressions, whether those transgressions arise from external pressures (e.g., politicization) or internal violations (e.g., fabrication of data to further one’s scientific career).
If this value of science is to be protected, evidence must be able to challenge currently held views. This requirement creates certain demands for the structure of how other values (whether ethical, social, political, or cognitive) can play a role in science.
Depending on where one is in the scientific process, values have different legitimate roles they can play, with legitimacy determined by the need to protect the value of science. Consider the following two roles values can play in our reasoning: direct and indirect. In the direct role, values are a reason in themselves for our decisions. They evaluate our options and tell us which we should choose. An indirect role for all kinds of values (political, social, ethical, cognitive) is needed and acceptable throughout the scientific process. Science is thus a value-saturated process.
This view of values in science can now provide us with a clear definition of scientific integrity. First, as described here, scientific integrity is a quality of individual scientists, their reasoning, and particular pieces of scientific work. Thus, a person, a paper, a report can all be said to have scientific integrity. The crucial requirement for scientific integrity is the maintenance of the proper roles for values in science. Most centrally, an indirect role only for values in science is demanded for the internal reasoning of science. When deciding how to characterize evidence, how to analyze data, and how to interpret results, values should never play a direct role, but an indirect role only. This keeps values from being reasons in themselves for choices when interpreting data and results. In addition, values should not direct methodological choices to pre-determined outcomes, nor should they direct dissemination choices to cherry-pick results. This restriction on the role of values, to the indirect role only at these crucial locations in the scientific process, is necessary to protect the value of science itself, given the reason we do science is to gain reliable empirical knowledge. We do science to discover things about the world, not to win arguments. Protecting scientific integrity as so defined thus protects the value of science.
What does this view of scientific integrity mean for our understanding of the politicization of science? Clearly, political forces could cause a scientist, either voluntarily or through coercion, to violate the proper roles for values in science and thus violate scientific integrity. Examples of this include scientists pressured to (or for their own political purposes deciding to) fabricate evidence, cherry-pick evidence, distort results, or stick to a claim even when known criticisms which fatally undermine the claim remain unaddressed. The main intellectual fault in all these cases is failing to be responsive to genuine empirical concerns, because doing so would make one’s political point weaker or undermine a cherished ideological perspective. It is to utilize a direct role for values and have that determine one’s results. It is to use the prima facie reliability and authority of science, which rests on its robust critical practices and evidential bases, and to throw away a concern for the source of science’s reliability in favor of the mere veneer of authority. It is to turn science into a sham. No wonder scientists get so upset when violations of scientific integrity occur.
For example, a failure to respond to criticisms raised repeatedly and pointedly is a clear indication of a problem. If a scientist, or a political leader using science, insists on making a point based on evidence even when clear criticisms undermining their use of that evidence have been raised, and they fail to respond to those criticisms, one is warranted in suspecting that the cherry-picked evidence is but a smokescreen for a deeply held value commitment serving an improper direct role, and that ultimately, the evidence is irrelevant.
Violations can also be detected in overt or covert interference with the activities of scientists. Political actors may not like the results produced by scientists, but their response should not be to declare them by fiat to be otherwise. Instead, politicians can legitimately question whether the evidence is sufficient to support certain policies, whether other policy options might be preferable, or whether value commitments should demand contrary courses of action.
In addition, one needs to assess whether a sufficiently diverse range of scientists (to ensure adequate criticisms of each other’s work are being raised) are working on a range of projects that do not just serve a narrow set of interests. If power and money draw the efforts of scientists into a narrow range of projects, society will not be well served. Even if the science being done is performed with perfect integrity, the results may be distorted and politicized simply because they are the only results available. This is a much harder problem to track and assess, and has not been the main area of concern with the politicization of science. But I suspect it will become a key area of debate in the coming decades.
JC comments: Points that I find to be particularly insightful and relevant to climate science include:
• If this value of science is to be protected, evidence must be able to challenge currently held views. Premature declarations of ‘consensus’ and attempts to marginalize those that disagree have become institutionalized in climate science, with strong statements of advocacy being made by professional societies (e.g. AGU, APS).
• . . . failing to be responsive to genuine empirical concerns, because doing so would make one’s political point weaker or undermine a cherished ideological perspective. JC: Climate science is rife with such examples, the most notorious example being the ‘hockey stick’. Another example is Lindzen’s iris hypothesis (which is the topic of a forthcoming post).
• If a scientist, or a political leader using science, insists on making a point based on evidence even when clear criticisms undermining their use of that evidence have been raised, and they fail to respond to those criticisms, one is warranted in suspecting that the cherry-picked evidence is but a smokescreen for a deeply held value commitment serving an improper direct role, and that ultimately, the evidence is irrelevant. JC: Well this pretty much sums up the approach being used by President Obama and his advisors with regard to climate change.
• One needs to assess whether a sufficiently diverse range of scientists (to ensure adequate criticisms of each other’s work are being raised) are working on a range of projects that do not just serve a narrow set of interests. JC: This is an issue of key importance for climate science, which was raised recently by the post Is federal funding biasing climate research?
Of direct relevance to the concerns raised by Hayward, Joe Duarte writes about Ideologically-fueled research, pursuant to a comment on his recently published research Political diversity will improve social science. Duarte focuses on an example from the social sciences, but these ideas easily generalize to climate research. Excerpts:
If you believe your ideology is true, but look out upon the world and see that large numbers of people don’t embrace it, it can be frustrating. You have a list of issues you think must be urgently addressed by society, yet society is not addressing them, perhaps doesn’t even see them as problems to begin with. This can create a lot of dissonance – why don’t people see what we see or think as we think? One way to resolve that dissonance is to assume that there must be something wrong those people, that there must be “causes” behind their positions other than simple disagreement, much less any wisdom on their part. So the next step is to inventory the uncharitable reasons why people don’t embrace your ideology, the ideology you just know is true and noble.
Environmentalism is a rather new political ideology, and possibly a religion or a substitute for traditional religion, and it’s alarming that social psychologists are promoting it and trying to convert people to it. Embracing new, abstract, and somewhat ambiguous values like “nature” and “the environment” is just assumed to be equivalent to rationality or something. Environmentalist values are contested by scholars all over the place (though not so vigorously within academia), but the field seems unaware of this, and unaware of their status as values, as ideological tenets, as opposed to descriptive beliefs about the world.
What’s more, we often see researchers declare outright that their motivation is to advance their ideology, to spark political action, and so forth. I think it’s impossible to argue that the field is not biased when researchers declare themselves to be political activists and that their research is an outlet for said activism.
This researcher has already decided that holding a particular position that she disfavors has a certain class of “causes”, including behavioral and neural bases. She has pre-emptively shrunk reality, the reality that she will allow herself to see. Rather, she is extremely likely to find what she is looking for.
Science requires us to be more sober than this. We can’t go in having decided already what kinds of causes must be in force.
It seems to be in the nature of ideology to convert ideological tenets and value judgments into descriptive facts/concepts in the mind of the ideologue. It’s a good protective immune system for an ideology to have, to pre-emptively marginalize and de-legitimize dissent as corrupt or ignorant and thus deter one’s members from closely examining alternative schools. In any case, a valid social science needs to immunize itself from this sort of ideological embedding.
The ideology that I am concerned about is what I have termed UNFCCC/IPCC ideology. In the way that I have defined it, there is nothing wrong per se with an ideology; the problem is with ideologues – absence of doubt, intolerance of debate, appeal to authority, desire to convince others of the ideological ‘truth’, and willingness to punish those that don’t concur.
If the community of scientific researchers was sufficiently diverse to accommodate a range of ideological perspectives, ideology wouldn’t have much impact on the overall scientific oeuvre. However, when a single ideology is adopted by the professional societies and enforced by the political party in power, then we have a serious problem.
As an individual scientist, navigating all this in a highly politicized environment can be a real land mine. But the problems – with only a few exceptions – aren’t with individual climate scientists, but with the institutionalization by professional societies of a particular ideology, the general liberal bias at universities, and arguable biases in federal funding of climate research.
It is very good to see philosophers and social scientists tackling these issues; it would be even better to see non-partisans from these fields analyze the situation in climate science.