Charles Rotter
I read this so you don’t have to.
A new paper published late May in Environmental Communication presents what is described as a “two-dimensional framework” for analyzing how climate scientists experience the public sphere. The lead author is Victor Avramov of the Athena Institute at VU University Amsterdam. The co-authors are at the University of Amsterdam and the Rathenau Institute in The Hague. The work was funded by the Dutch Research Council through the Dutch Science Agenda, file number NWA.1397.21223.014, paid for by Dutch taxpayers.
ABSTRACT
Climate scientists are increasingly drawn into a polarized public sphere, challenging relations between science and society. In this study, we interviewed thirty-five climate scientists – diverse in discipline and seniority – working in the Netherlands about their perceptions of, and experiences with public engagement. Based on our empirical material, we construct an analytical framework with a politization and participation axis on which we position their statements. Demarcating their public activities along these dimensions, climate scientists highlight concerns for scientific credibility, political efficacy, normative responsibility and individual capacity. While there is a clear opposition between those compelled to advocate for stringent climate policies or tackle misinformation and those who believe their main role is to provide solid knowledge and leave the normative choices to activists or politicians, only few scientists collaborate with stakeholders. Letting different stakeholders speak and participate in knowledge productions, we argue, may provide a solution to the science vs politics stranglehold.
The sample size is thirty-five. The sample consists of thirty-five climate change researchers from the Netherlands. The statistical method is to interview them for fifty to ninety minutes, transcribe the interviews, code the transcripts inductively, and then plot the resulting codes on a two-dimensional chart. The axes are labeled “Politization” (sic, throughout) and “Participation.” The axes have no units. The chart contains no data points and no scale, only four softly colored blobs of varying size labeled Demarcation, Seeking Impact, Activism, and Co-Creation. The bigger blobs represent more interviewees having said roughly similar things. The placement of the blobs within their quadrants is determined by the authors having decided that is where the blobs go.
This is what social science research looks like in 2026.
What the thirty-five interviewees have to say
The interview quotes, which take up most of the paper, deserve to be read at length. The plotting is less informative than the quotes themselves.
Here is interviewee P23 explaining why scientific neutrality may be overrated:
Besides being a scientist, I am also a citizen, and I am scared shitless about what is happening. Why should I not be allowed to speak out about it and why would it hurt my credibility, if it is actually science, which helps me understand what is happening?
The paper does not consider whether being scared shitless is a sound foundation for professional conduct. The paper plots P23 in the upper-left quadrant, which is labeled Activism. This is presented as a legitimate role for a scientist to occupy.
Here is interviewee P28, an Earth sciences professor, comparing the effectiveness of formal scientific advice to the European Union with the effectiveness of street protests:
I’m going to Brussels soon to talk to the EU. I could be talking in a void … but sometimes I doubt the effectiveness of that role compared to what we reach when we just take to the streets.
The paper does not raise the question that suggests itself, which is whether a scientist who has concluded that street protests are more effective than scientific consultation is still a scientist in any meaningful sense. P28 appears in Activism. P28 supplies several other quotes the authors find illuminating.
Here is P28 on the burdens of being a moral exemplar among the unenlightened:
I do feel estranged from some colleagues that I was good friends with because they never wanted to apply the climate crisis to themselves, to their own political lives.
The paper treats this as evidence of institutional friction with activist scholars. The friction is real. The other reading, which the paper does not consider, is that the friends in question may have grown tired of being lectured.
Here is P15, in the same general region of the plot, on the indignity of being labeled what they evidently are. The interviewee, the paper reports, admitted to censoring
themselves out of the fear of being called an activist.
The paper presents this as a problem. The chilling effect of accurate labeling is preventing the activist from doing more of the activism. The accurate labeling is the problem. The activism is the data.
The chart
The methodological centerpiece is Figure 1, a two-dimensional framework on which the four clusters of interview statements are plotted by their relative degrees of politization and participation. The axes have no units. The blobs have no coordinates. The placement of the blobs in their quadrants was determined by the authors in collaboration. The figure caption notes that higher color density represents higher frequency,
which gives the chart its only quantitative dimension: how often the interviewees said things that fell into each cluster, as determined by the authors’ own coding decisions.
The chart has axes. The axes have labels. The chart has clusters. The clusters have positions. The chart has a color-coded legend.
This makes the chart scientific in the sense that it is shaped like science.
What the chart tells the reader, on inspection, is that out of thirty-five interviews, the largest cluster of statements fell in the upper-left quadrant labeled Activism. The next-largest fell in the middle, labeled Seeking Impact. A smaller cluster appears on the right, labeled Co-Creation. A thin sliver appears on the lower-left, labeled Demarcation. Demarcation is the position of scientists who believe their job is to do science and to leave politics to politicians and activism to activists.
The paper treats this thin sliver as the limiting baseline to be overcome. The activist blob, in the upper left, is presented as a more mature and developed position. The co-creation blob on the right is the position the paper ultimately recommends.
Where this is going
The discussion concludes that the way out of the “science-versus-politics stranglehold” is participation. Specifically, citizen assemblies in which scientists and members of the public deliberate together about climate research. The closing paragraph helpfully discloses:
In a current research project, we are exploring the potential of the democratic innovation of citizen assemblies in the domain of science communication.
The paper that diagnoses the stranglehold and prescribes citizen assemblies as the cure is written by people who are funded to study citizen assemblies. The diagnosis arrives matched to the prescription as if by appointment. The Dutch taxpayer is paying for both.
The discussion further commends intellectual humility
and being an honest broker of knowledge
and observes that scientists running citizen assemblies must be trusted as neutral.
The thirty-five interviewees, plotted across the framework, contain almost nobody in the demarcation quadrant who could plausibly be described as trusted to be neutral. The neutral scientists, in the paper’s own data, are the small marginal cluster. The activists and the participation enthusiasts are the bulk. The paper’s recommended fix for the trust problem is that those activists and participation enthusiasts should run the citizen assemblies in which they are trusted to be neutral.
A note on the broader context
There is a serious point underneath the methodological theater. The thirty-five interviewees represent a profession that has spent the last fifteen years building public credibility on the strength of being scientists. The credibility has been pledged to scenarios that have now been formally retired (see our extensive recent coverage of the death of SSP5-8.5 and SSP3-7.0), to attribution claims downstream of those scenarios, and to a political program that has consistently overstated the certainty of the science. Public trust has shifted accordingly. The paper’s thirty-five interviewees are part of the system that produced that shift and have noticed.
The Avramov paper is what happens when the people inside that system look around, notice that things have gone sideways, and conclude that the problem must be the boundaries between science and politics. Not the substance of the science. Not the choice of scenarios. Not the alignment of published conclusions with the funding priorities of state research councils. Not the fifteen-year pattern of overstating findings to support pre-existing policy preferences.
The problem, in this paper’s framing, is that the scientists have not been doing enough activism.
The solution is more activism, conducted under a different name, with the participation of citizens brought along through deliberation. On the paper’s own evidence, the activist scientists want to be trusted while also being political. They want to be neutral while also being activists. They want to be heard while also lecturing. They want to be considered scientists while also taking to the streets. They want to be paid by the state while also speaking truth to it. They are caught, the title says, in the fray. They are also scared shitless. The paper plots all of this carefully.
The thirty-five interviewees of the Netherlands are now on record. The Dutch Research Council, having paid for the insight, will presumably commission more.
That is the grant cycle doing its job, which is what the grant cycle has always been for.
