by Judith Curry
We shed new light on the epistemic struggle between establishing consensus and acknowledging plurality, by explicating different ways of consensus-making in science and society and examining the impact hereof on their field of intersection. – Laszlo Kosolosky and Jeroen Van Bouwel
I first met Laszlo Kosolosky at the Workshop on the Role of Climate Models. Laszlo is a Ph.D. student in Belgium, the title of his thesis is ‘Science Versus Society, Democracy Versus Expertise, Consensus Versus Plurality: A Social-Epistemological Study of Central Notions in Scientific Practice, i.e. Expertise, Consensus, Peer Review, Epistemic Integrity, and Values.’ He has already published a number of papers. His research is of obvious relevance to climate science, and I have flagged a number of them to include in future posts.
Here is a paper on consensus, which provides some important and useful insights into the consensus building process .
Explicating ways of consensus-making in science and society: distinguishing the academic, the interface and the meta-consensus.
Laszlo Kosolosky and Jeroen Van Bouwel
Abstract. In this paper, we shed new light on the epistemic struggle between establishing consensus and acknowledging plurality, by explicating different ways of consensus-making in science and society and examining the impact hereof on their field of intersection, i.e. consensus conferences (in particular those organized by the National Institute of Health). We draw a distinction between, what we call, academic and interface consensus, to capture the wide appeal to consensus in existing literature. We investigate such accounts as to put forth a new understanding of consensus-making, focusing on the meta-consensus. We further defend how (NIH) consensus conferences enable epistemic work, through demands of epistemic adequacy and contestability, contrary to the claim that consensus conferences miss a window for epistemic opportunity. Paying attention to this dynamics surrounding consensus, moreover allows us to illustrate how the public understanding of science and the public use of the ideal of consensus could be well modified.
A link to the paper (a chapter in a book) can be found [here].
Excerpts from the paper:
In our society, there are these moments in which establishing a scientific consensus seems imperative to solve urgent problems, for instance, as concerns climate change; achieving consensus on the causes and extent of global warming would facilitate policymaking and, moreover, send a convincing signal that doing nothing will have dire consequences. On the other hand, philosophers studying plurality and heterodoxy in science have raised questions concerning the ideal of the scientific consensus and the pernicious effects the consecration of scientific consensus might have.
If we aim to elaborate on the tension between consensus and plurality mentioned above, we first need to establish what aspects of scientific consensus we would like to address here. We draw a distinction between a ‘technical, academic consensus’ and an ‘interface consensus’. The former points at a consensus being established among scientists or experts in a certain field related to a certain topic. The latter relates to a consensus being established at the border between science and society, typically including a wider range of actors apart from scientists (i.e. laypeople, interactional experts, government representatives, etc.).
The relationship at play within academic consensus is one between experts. In the academic world, every scientist/academic is regarded to be an (equal) peer and everyone serves as an authority within his or her field. These people are generally regarded to be on the cutting edge of research and are expected to be among the first to notice changes occurring within their field of expertise. The relationship at play within the interface consensus is one between expert and layman, grasping the interface between science and society. This type entails a relation between expert and layman grounded on authority, trust, and mutual respect, where the actors are not regarded to be on equal (epistemic) footing. The difference in interaction is important to bear in mind when we want to have a look at what’s at stake in each of them.
Besides distinguishing the academic from the interface consensus, we would also like to introduce the notion of the meta-consensus. Instead of focusing on the simple level, that is, as the result of alternative theories/models tested against one another eventually – thought to be – leading to some consensus outcome, we could learn a lot by shifting to the analysis of the meta-consensus that stipulates the procedure to be followed.
Kristina Rolin (2009) stipulates the epistemic role outsiders to particular scientific communities can play. She argues that an epistemically responsible scientist has a duty to respond to outside criticism in certain circumstances insofar as it includes an appropriate challenge to her views. A meta-consensus taking contestability into account differs from both aggregated judgment – in supporting dynamical, diachronic interaction – and rational deliberation – avoiding groupthink via contestation. Obviously, contestability comes in degrees and is present to a greater or lesser extent in the existing formats for consensus-making.
In his earlier work on experts (2006), Beatty already captures part of the tension between consensus making on the one hand and the intrinsic value of plurality on the other hand. First, what he calls simplification, entails that scientists instead of simply telling us what they know, they might tell us simply. In this manner, a lot of crucial information gets lost along the way. Second, what he identifies as the intentional withholding of information, means that scientists often agree amongst one another as to retain information from the public or silence discussions, which results in a distorted view of consensus amongst the public and nourishes the ill-conceived expectations they might have. Both pitfalls find their reasoning on either paternalistic or protective grounds. The former meaning that experts state that it might actually be in the public’s advantage if they speak with one voice rather than with many, whereas the latter hints at the fact that experts, in this manner, could in fact guarantee that their status remains intact and can prohibit others from gaining the authority and trust to do their work.
The debate about scientific consensus then moves to consensus on epistemic procedures, i.e. finding a form of meta-consensus. Thus, the tension between scientific plurality and consensus is not tackled on the simple level, but on the meta-level. This is analogous to how democratic societies deal with value pluralism; the focus is not on getting rid of value pluralism, but on establishing a framework – a meta-consensus – within which pluralism can be dealt with satisfactorily. The meta-consensus can be one that prescribes rational deliberation (in line with models of deliberative democracy), or aggregation (stipulating a procedure for adding up the available views), or agonistic pluralism (developing a procedure or constellation – conflictual consensus – that wants to optimize the epistemic fecundity via agonism), etc.
Establishing scientific consensus is highly valued by philosophers, scientists and the public. The emergence of a scientific consensus replacing competing accounts is often interpreted as a proof of scientific progress and a marker of truth; ideally all scientific inquiry and debate would result in a consensus. Finding scientific consensus is then understood as a proof for the validity of a theory and – indirectly – of the public policy based on the consensus theory.
The back side of the coin is that the lack of scientific consensus often is used to undermine or criticize science and the public policy based on it (e.g., former US President Bush on climate change). When scientists agree, their results are taken more seriously than when they disagree, even though such an agreement or consensus might hinder scientific progress because of critical, heterodox theories not being taken seriously (e.g., the theory of continental drift was accepted by geologists only after 50 years of rejection, and the theory of helicobacter pylori as the cause of stomach ulcers, was at first widely rejected by the medical community).
These observations might question scientific consensus as an ideal or as the goal of inquiry and marker of truth; enforcing ‘consensus’ might be dangerous or not desirable, hence the importance of scrutinizing carefully what is actually going on in establishing scientific consensus. Communicating this variety of formats to the public, helps qualifying the actual span of scientific consensus-making and the oracle like features it might sometimes have. Our reasoning is in line with Inmaculada de Melo-Martin and Kristen Intemann’s recent paper, where they show that ‘focusing on dissent as a problematic activity sends the message to policy-makers and the public that any dissent undermines scientific knowledge.’ Encouraging and providing mechanisms of dissent can also be important to reassuring the public that the consensus view has undergone rigorous scrutiny. Events such as “climate-gate” reinforce the public perception that climate scientists are resistant to criticism and have a “bunker mentality”.
Through our analysis we have argued in favor of the following claims:
a) Consensus-making as discussed by philosophers of science, should be aware of the difference in aiming for an academic consensus or an interface consensus in science and society. A broader understanding of the different structures and functions of consensus-making helps us to see more nuances.
b) The difficulties of achieving consensus understood as an unanimous outcome – a seldom-attained ideal stipulated by a range of criteria – should make us shift to a procedural approach in which the emphasis is not so much on establishing the consensus, but dealing with plurality in a consensual way, i.e., framed within a meta-consensus that agrees on how to disagree.
This paper makes a very important point about the distinction between an academic and interface consensus. The climate change consensus is really an interface consensus, negotiated by the IPCC.
The primary example used in the paper is NIH (public health) consensus conferences. While this application shares some commonality with the climate change consensus building, the big difference lies in the type of problem being addressed. The NIH problems are relatively tame, in that they are constrained and there is no conflict of values (there are uncertainties). On the other hand, climate change is a wicked mess, with complexity and changing dimensions, solutions that are suboptimal with unintended undesirable consequences, and major conflicts of values.
For a wicked mess such as climate change, I’ve argued in my paper No consensus on consensus:
The climate community has worked for more than 20 years to establish a scientific consensus on anthropogenic climate change. The IPCC consensus building process arguably played a useful role in the early synthesis of the scientific knowledge and in building political will to act. We have presented perspectives from multiple disciplines that support the inference that the scientific consensus seeking process used by the IPCC has had the unintended consequence of introducing biases into the both the science and related decision making processes. The IPCC scientific consensus has become convoluted with consensus decision making through a ‘speaking consensus to power’ approach. The growing implications of the messy wickedness of the climate change problem are becoming increasingly apparent, highlighting the inadequacies of the ‘consensus to power’ approach for decision making on the complex issues associated with climate change. Further, research from the field of science and technology studies are finding that manufacturing a consensus in the context of the IPCC has acted to hyper-politicize the scientific and policy debates, to the detriment of both. Arguments are increasingly being made to abandon the scientific consensus seeking approach in favor of open debate of the arguments themselves and discussion of a broad range of policy options that stimulate local and regional solutions to the multifaceted and interrelated issues of climate change, land use, resource management, cost effective clean energy solutions, and developing technologies to expand energy access efficiently.
Note, I will be presenting my No Consensus on Consensus arguments at a philosophy of science Conference (Laszlo is on the organizing committee) entitled ‘The epistemic value of dissent in climate science’, scheduled for next fall.