by Judith Curry
Among the best indirect indicators available to nonexperts is the overwhelming numbers of scientists testifying to anthropogenic climate change. Yet the evidential significance of such clear numbers turns substantially on our nonexpert assessment of these scientists’ trustworthiness. Absent trust, even without active distrust, the numbers’ evidential weight drops considerably. – Ben Almassi
One of the most insightful articles I’ve come across on the topic of consensus is a paper by Ben Almassi entitled Climate Change, Epistemic Trust, and Expert Trustworthiness [link to abstract; unfortunately behind paywall]. Excerpts:
The evidence most of us have for our beliefs on global climate change, the extent of human contribution to it, and appropriate anticipatory and mitigating actions turns crucially on epistemic trust. We extend trust or distrust (or refrain from extending either) to many varied others: scientists performing original research, intergovernmental agencies and those reviewing research, think tanks offering critique and advocating skepticism, journalists transmitting and interpreting claims, even social systems of modern science such as peer reviewed publication and grant allocation.
This broad epistemic dependency need not thereby make our knowledge of climate change somehow intellectually or ethically suspect; nor does it make our relationship to climate science so different than other domains of knowledge toward which many of us (including those with expertise in other domains) stand as nonexperts relying on others. Yet attending to trust (and lack thereof) in contemporary public understanding of climate change may help illuminate moral issues overshadowed by the rightly urgent matters of international and intergenerational justice in responsibility for and response to climate change.
Epistemic dependence and scientific consensus
Even if I have a working appreciation for the greenhouse effect, were I to pursue epistemic autonomy on climatological phenomena, this working grasp combined with my scant firsthand observations of weather patterns and temperatures would fall well short of justifying particular beliefs on global climate patterns, let alone the extent of our human contribution. Indeed, even the notion that my basic understanding of greenhouse effects is sufficiently similar to contemporary atmospheric processes, as to be an edifying rather than misleading comparison, itself depends crucially on expert validation of the analogy. Whether my personal apprehension of hotter summers and unusual weather phenomena ought to be seen as confirmation of anthropogenic climate change, or instead as irrelevantly anecdotal, is itself something I myself cannot determine, something for which I must rely on expert corroboration. While I may be informed that the last years of the twentieth century are among the hottest in centuries, and I may then on reflection take this to be strong evidence of anthropogenic global climate change, nevertheless that inference is epistemically dependent on others. Just as I cannot collect annual global temperatures myself, without relying on experts I also cannot justifiably regard these temperatures as decisive evidence for anthropogenic global climate change over other candidate hypotheses. I may not even know which (if any) competing hypotheses are also capable of accommodating this temperature data, let alone judge for myself which hypothesis does so best.
It should be no surprise, then, if public understanding of and beliefs about global climate are parasitic on public understanding of and beliefs about climate science as an area of expert knowledge. Following Hardwig, we may even recognize this epistemic dependency as rational and responsible. What’s striking about public understanding of climate change is not the dependency, but rather the nonexpert assessment of climate scientific expert opinion.
The public presence of dissenting voices presents a substantial social-epistemic problem for nonexpert appreciation of the state of contemporary climate scientific knowledge. While experts may confidently and justifiably dismiss a few dissenters by appeal to their own first-hand understanding of climatologically significant empirical evidence, we nonexperts cannot easily make such assessments ourselves from our nonexpert perspectives.
Nonexpert assessment of conflicting claims
Consider the situation of a third-party nonexpert who is not part of the relevant scientific community yet trying to assess whose allegedly expert testimony to trust, and to what degree. Should she clearly put greater epistemic trust in the testimony of loyalist rather than dissident experts? How do speakers‘ credibility within a scientific community translate into trustworthiness (or lack thereof) for epistemically dependent outsiders?
Dialectical Superiority. Goldman observes that non-experts often are provided only with experts‘ conflicting conclusions, sometimes only with partial sketches of experts‘ argument for these conclusions. Furthermore, even when arguments are provided in detail, much of their content remains opaque to non-expert listeners. What sort of dialectical superiority does he have in mind as an indirect indicator of expert trustworthiness? It is in part a measure of smoothness and quickness in responding to rival experts‘ claims and objections, and here precautions about demeanor must be noted.
We should also be cautious about non-expert ability to weigh the genuine social-evidential relevance of apparent failures to rebut or respond to competing criticisms. In some cases lack of rebuttal is due to an inability to successfully defend one‘s case against critique, and in some cases lack of critique is due to an inability to offer cogent critique. In some cases duplicitous speakers may sometimes say that their criticisms have not been rebutted though they have; in other cases one side may sincerely regard their criticism as remaining unrebutted, and assert as much, while the other side takes themselves to have offered a rebuttal. Alternatively, further criticism or defense against criticism may have been made in another venue or environment of which non-expert listeners may be unaware, their lack of awareness owing precisely to their lack of expertise. Further still, what appears on non-expert analysis to be an unaddressed criticism may be something the other side has simply ignored, perhaps for good reason, as sufficiently irrelevant or repetitive as not to warrant response. On reflection, then, apparent dialectical superiority may result from a variety of factors, including but not limited to superior access to truth.
Expert Agreement. Ward Jones offers a pro-loyalist stance according to which non-expert third parties should side with the dominant position in expert disputes. He identifies loyalists as community members who endorse the dominant position in a scientific community and dissidents as other community members who dispute the dominant position. Is the position endorsed by a supermajority, majority, or plurality of the community? Do all members get equal voice in the measure; is the dominant position a product of power and influence in the community (who publishes most, is cited most, has the most students, edits the journals, controls grant funding, etc.)? Without a clear idea of how much different experts‘ assessments counts in favor of one side or other, non-experts can only guess whether a dominant position is also one with better testimonial evidential support.
As a paradigm case of an evidentially meaningless concurrence of opinions Goldman cites a guru with many slavish followers who believe and say whatever he believes and says. If the listener is justified in believing that these followers never break from blind allegiance to their guru, Goldman reasons, then it does not matter how many such followers there are; the followers‘ testimonies give the listener no more reason to believe the testified claim than was provided by the guru‘s testimony.
More broadly, two members of a scientific community may agree about a hypothesis because their endorsement or rejection of that hypothesis may follow from a set of shared theoretical, interpretive and methodological commitments. We may reasonably ask why these scientists agree on their assessments of evidence or their theoretical, interpretive, or methodological commitments, of course. Perhaps Y is just uncritically following X‘s lead on these matters, or vice versa, or perhaps both experts are mindlessly following a third expert Z or have uncritically accepted the typical standards of their epistemic community. Or it may be that they have come to their evidential assessments and commitments through an amalgam of independence reflection, warranted trusted in peers, deference to tradition, career considerations, social circumstances, luck, and other factors.
Trust and the Numbers. Yet the social-evidential significance of a supermajority of expert testimony in favor of anthropogenic climate change varies for different members of the nonexpert public, depending in significant part on our differing assessments of the scientists in and out of agreement. Here public trust and distrust of scientific expert opinion matters a great deal, I suggest, in our varying interpretations of the evidential significance of expert agreement. Those of us who trust will thus have reason to put greater evidential significance in the numbers of agreeing climate scientists because we trust that these scientists have come to a common commitment not through deceit, conspiracy, or lazy deference, but rather through considered assessment of the empirical evidence and their fellow scientists‘ rational credibility. Those who actively distrust, by contrast, will have reason to put far less evidential significance in an occasion of lopsided expert opinion in favor of anthropogenic climate change; for them, unsavory and epistemically irrelevant explanations of concurrence such as deceit, conspiracy, or groupthink are far more plausible.
Among the best indirect indicators available to nonexperts is the overwhelming numbers of scientists testifying to anthropogenic climate change. Yet (perhaps more surprisingly) the evidential significance of such clear numbers turns substantially on our nonexpert assessment of these scientists‘ trustworthiness. Absent trust, even without active distrust, the numbers‘ evidential weight drops considerably.
While the author comes across as supporting the consensus, the paper presents some insightful perspective on the ‘consensus enforcement’ by the establishment and why a substantial portion of the public is not buying the expert consensus on climate change. It boils down to a lack of trust, and concerns about deceit, conspiracy and groupthink.
Where do these concerns come from? Climategate and explicit advocacy by scientists are two obvious sources. Disagreement portrayed in the media and distrust of the government’s politicization of the issue are others.
How can trust be rebuilt? This brings to mind my post Climategate essay On the Credibility of Climate Research: Towards Rebuilding Trust – transparency and uncertainty were two major themes. Also:
The openness and democratization of knowledge enabled by the internet can be a tremendous tool for building public understanding of climate science and also trust in climate research.
The internet makes establishment consensus enforcement very difficult. Since I wrote that essay in 2009, I now understand the climate problem to be a wicked mess, and the idea of a consensus among experts to be misguided. As discussed on a previous post, consensus messaging doesn’t seem to work with the public in any even.
On this five year anniversary of Climate Etc., has, or how has, Climate Etc. helped to rebuild trust about climate science?