CB 500

1 November 2022
STS, platform capitalism and the conundrum of expertise
What a difference a few decades can make!
by Philip Mirowski
Image Credit // The Circus Bazaar Company
Philip Mirowski
Philip Mirowski is Koch Professor of History and Philosophy of Science, Univ. of Notre Dame, and author of many books, including Science-Mart (2011) and with Eddie Nik Khah, The Knowledge we have Lost in Information (2017).

They might be loath to admit it, but many Science and Technology Studies [STS] scholars today long for the halcyon days of the so-called “Science Wars” of the 1990s. Back then, those scholars were relatively united in their defenses against a loose coalition of natural scientists and their fellow-travelers in the humanities, who sought to delegitimize STS work by suggesting it demeaned and degraded the cultural acceptance of and acquiescence in scientific expertise. STS scholars tended to argue that the Science Warriors misunderstood much of the STS methodology, which did indeed undermine older Mertonian stories of the pristine separation of science from society and its demands, but that this was done with the intention of supporting and clarifying the work of science in action, to cite one of its champions, Bruno Latour.1 While the STS formation may have temporarily risked a modicum of its academic credibility, there was never any sense that the stakes were much higher than, for example, a figure like Latour getting turned down for a plush sinecure at the Institute for Advanced Study at Princeton, or someone else’s article being ridiculed by Alan Sokal.

What a difference a few decades can make! These days, almost everyone anguishes over what they perceive to be pervasive attacks on scientific expertise, be it anti-vaccine resistance to global warming denial to the sperm counts of urban males, and much else. The entire population appears split over whether one should “trust science” or not, although there is little consensus over whether that is even a cogent way to phrase the nature of the antagonism. Not only is the trust divide regarded as political, in the conventional sense of partisanship, but its implications for the widespread breakdown of democratic structures are often mooted as having the utmost significance. The political stakes in the role and nature of expertise have clearly risen not only for the general populace since the 1990s, but also for the participants in STS research. Consequently, a rift appears to have opened within STS, between those who pride themselves in regularly challenging the shape and content of scientific expertise, and those who have begun to insist that science in the large must be defended from the general culture of disdain and disrespect. STS journals are full of disputations over the extent to which science studies may or may not be responsible for this perilous juncture. While the dominant tendency has been to plead that the crisis should not be laid at their doorstep, others have adopted a much harsher position: “Contemporary science and technology studies (STS) erodes the cultural importance of scientific expertise, and unwittingly supports the rise of populism.”

While I have sought to map the current STS terrain concerning the politics of science elsewhere, here I want to concentrate upon a major weakness in the camp of those inclined to defend science from its detractors. Some of the representative STS figures I have in mind are Harry Collins,3  Naomi Oreskes,4  and Darrin Durant. While these scholars offer admirable arguments why scientific consensus should be granted various privileges in the modern polity, I want to suggest that they have overlooked one rather large vulnerability in their case. While none of these scholars can be accused of ignoring the historical character of science, when they seek to justify the indispensable role of expertise, they do tend to describe the self-correcting character of science in relatively abstract terms. Collins tends to harken back to his theory of “Core Sets” of specialists vetting claims, while Oreskes appeals to general social structures that vigorously vet claims, reach agreement, and then move on.5  The flaw in their arguments resides in the fact that those regulatory social structures can change dramatically over time, and that the dependability of those vetting procedures may themselves be diminished as a consequence. Indeed, changes in the social structures of science may go some distance towards helping to explain the modern distrust of expertise. This observation will dovetail with another theme often broached when bemoaning the lack of trust in science: the significance of the rise of social media and their grounding in the social structures of post-1990 platform capitalism.

Collins and Oreskes are certainly aware that science has increasingly been subject to commercialization pressures since the 1980s, yet neither has sufficiently appreciated the extent to which the reconstruction of science under market imperatives has been a rolling, cumulative process. Once the first phase of extending intellectual property to cover scientific discoveries had passed, the next phase was to begin to apply market logics more intently to the process of research.6  The very notion that you and your immediate peers might devote your entire life to one discrete research program in some small byway of a single discipline was rendered obsolete, if only because a dearth of profitability would dictate that investment in such an area cease forthwith, and be shifted to some more promising arena. More relevant, the research process was broken up into component modules of research activity, under an injunction to open them to a range of labor at each stage, subject to external scrutiny, and better, to themselves be monetized to the maximum extent. This re-engineering of scientific research was promoted under the rubrics of “open science,” open access, banishment of legacy journals and limited in-group peer review, freely accessible databases, “democratization of citizen science”, and much more.7  Many of these innovations are based upon the models of prior social media platforms, and upon the general frameworks of “platform capitalism”.8 

What Collins, Oreskes and others have overlooked is that the older social structures of the discipline-centric, journal-identified and university-based science that provide their templates for “scientific consensus” are giving way to a newer model of flexible gig labor organized through for-profit internet platforms. People under the age of 40 are fully aware of the growing importance of such platforms as LinkedIn, Research Gate, Mendeley, Walacea, Zenodo, PeerJ, PubPeer, and even Twitter to their research profiles and practices. Some universities have incorporated internal profile platforms into their translational medicine units.9  Moreover, “virtual labs” such as Science Exchange and Emerald Cloud allow for arm’s-length conduct of what was previously in-person labor at the wet-lab bench. Evaluation of the “truth production” functions of the researcher is happening more and more in real time, mediated by surrogate end-points, and registered though numerous automated measures dubbed “altmetrics.” Even patent counts have been demoted as happening too late in the research process. Scientific research has been accelerated and subjected to any number of market-like evaluation points, with consensus itself fragmented and replaced by so many accounting devices, depending heavily upon the internet, the better to repackage the results as fungible information to be sold to the highest bidder. 

One very salient symptom of the New World of science evaluation has been reported by Mario Biagioli and Alexandra Lippman.10  They provide a number of illustrations of a novel sort of misconduct in modern science: namely, not direct falsification of the content of a scientific paper, but rather the manipulation/falsification of the various metrics of research products that have come to stand in for the validation and vetting of research validity. They range from spoof peer reviews, to fake author bylines, to paper authorship for sale, to ghost management of medical publications, and far more baroque options. Biagioli himself opines that all manner of cheating is rife in some precincts of modern science, and of course is rendered possible by the digitization of such data; yet he seems not to realize that the decoupling of research activity from researcher identity is the hallmark of re-engineered modern science. Since “The Market” is expected to provide evaluation of the worth and significance of each bit of research, the linking of any finding to a particular community member and their affiliated peer group bears no special significance, and indeed, can itself be rendered subject to market trade. Perhaps this explains why Biagioli struggles so hard to identify an operant definition of nouveau “scientific misconduct”: manipulation of the algorithms that stand in for the contemporary evaluation of truth seems to resemble crafty entrepreneurship as much as it may be redolent of sleazy misrepresentation. To be offended by such behavior runs the risk of missing the underlying shift in forms of scientific veridiction.

The extension of market logics through platform structures throughout the research process (particularly noticeable in the biomedical sciences) has fundamentally altered what it means for a disciplinary community to criticize and vet scientific research. Bluntly, critique by peer review is slowly becoming trumped by market evaluation. Nick Srnicek describes how platform capitalism seems to spin fortunes out of data: “[Data] have come to serve a number of key capitalist functions: they educate and give competitive advantage to algorithms; they enable the coordination and outsourcing of workers; they allow for the optimization and flexibility of productive processes; they make possible the transformation of low-margin goods into high-margin services; and data analysis is itself generative of data”.11 Indeed, most of the altmetrics numbers are generated by for-profit entities, which bear no obligation to submit their own algorithmic appraisal procedures to any scientific community. (After all, these procedures are protected intellectual property.) As if by magic, science itself has been transformed into the proverbial self-licking ice cream cone. As a consequence, the process of assessment has been progressively outsourced to various funders and investors.

What I think that Collins, Oreskes, et al. miss is the extent to which the unholy alliance of “open science,” platform capitalism, and the culture of social media have had deep and profound consequences for the politics of science. Patently, the injunction “Trust the scientists” no longer possesses the probative force it may once have enjoyed during what might be dubbed the Cold War era. Mistrust of vaccines is not entirely illogical when it seems to be Pfizer, rather than the Center for Disease Control or the FDA, who pronounces that booster shots will soon be required in order to combat COVID-19. A modicum of discomfort with the Green New Deal might be vindicated by demonstration that the initial development of the Gaia hypothesis was funded by oil companies.12 Peer review is openly being eroded in a world where ghost managers can buy the illustrious authors they choose to grace the masthead of a paper, which has been compiled and composed by a CRO and a nameless cadre of medical writers. A fragmented phalanx of gig workers cannot possibly underwrite the validity of some research output. The epistemic authority of the disciplinary peer group is being increasingly diluted by devices like altmetrics and open comment platforms, so no wonder the man in the street feels justified in tweeting the results of his own “research” and Google searches to be as equally valid as any pronouncements by “experts.” The pervasiveness of commercial valuation itself has direct influence on political attitudes, both by those who suspect scientists of unsavory pecuniary motivations, and conversely, by those who subscribe to the neoliberal proposition that markets always know more than any self-identified savants. The hermeneutics of suspicion cuts both ways. Ahistorical assertions by some STS scholars that “science has always been commercial” cannot begin to comprehend the profound ways that modern monetization of the research process has altered the general epistemic stakes when it comes to the political legitimacy of experts.

Thus there is little plausible prospect that a populace drenched in social media and increasingly employed by platform capitalism will return to attitudes towards scientific experts prevalent before the Science Wars of the 1990s. The more pressing issue for STS is, rather, what social formations will replace them?


  1. For a nice summary of the Science Wars, see Gordon Katic. 2019. “Science Studies and its Mea Culpas,” Cited: https://www.citedpodcast.com/article-sciencewars/
  2. Collins, Harry; Evans, Robert; Durant, Darrin & Weinel, Martin. 2020. Experts and the Will of the People. Cham: Palgrave Macmillan.
  3. Collins, Harry & Evans, Robert. 2017. Why Democracies Need Science. Malden: Polity
  4. Oreskes, Naomi. 2019. Why Trust Science? Princeton: Princeton University Press.
  5. Ibid, p.141.
  6. Mirowski, Philip. 2011. ScienceMart. Cambridge: Harvard University Press.
  7. Mirowski, Philip. 2018. “The Future(s) of Open Science,” Social Studies of Science, (48):171-203.
  8. Srnicek, Nick. 2016. Platform Capitalism. Malden: Polity.
  9. Robinson, Mark. 2019. The Market in Mind. Cambridge: MIT Press, chapter 4.


STS, platform capitalism and the conundrum of expertise

They might be loath to admit it, but many Science and Technology Studies [STS] scholars today long for the halcyon days of the so-called “Science Wars” of the 1990s. Back then, those scholars were relatively united in their defenses agai...

by Philip Mirowski

STS, platform capitalism and the conundrum of expertise

They might be loath to admit it, but many Science and Technology Studies [STS] scholars today lon...

by Philip Mirowski