Media coverage of scientific retractions risks feeding a narrative that academic science is broken – a narrative which plays into the hands of those who want to cut science funding and ignore scientific advice.

So say Joseph Hilgard and Kathleen Hall Jamieson in a book chapter called Science as “Broken” Versus Science as “Self-Correcting”: How Retractions and Peer-Review Problems Are Exploited to Attack Science

Hilgard and Jamieson discuss two retraction scandals that readers of this blog will be familiar with: the 2014 STAP retractions from Nature and the 2015 Michael LaCour paper in Science. Both of these incidents involved high-profile work that was retracted when it became clear that misconduct – data fabrication or manipulation – had occured. However, while both cases involved misconduct by a single “rogue” scientist, media coverage of the two cases was rather different.

In the STAP cells case, the responsibility for the case was generally ascribed to Haruko Obokata, the first author on the retracted papers and the one found to have manipulated experiments. The self-correcting nature of the scientific process was generally praised:

Reports of the Obokata retraction featured episodic framing that blamed Obokata for her conduct and, to a lesser extent, thematic framing that blamed Nature and the peer review process for being deceived. A few reports briefly praised science for self-correction or highlighted ongoing efforts to safeguard the integrity of science. Some accounts noted that Obokata’s papers previously had been rejected by Cell, Science, and Nature, suggesting that prepublication peer review and editorial process had performed, at least to some extent, appropriately…The “science is broken” frame was not a prevalent one. Instead, most articles focused on the individual researcher.

But the case of Michael LaCour, in contrast, was widely described as symptomatic of a problem in science itself. This discussion took on a political dimension, probably because LaCour’s retracted paper reported on a succesful program to raise support for gay marriage. In the eyes of many, the fact that LaCour’s fraudulent paper was published is evidence of the liberal bias of peer-reviewed science:

The LaCour retraction was characterized by a causal frame that located blame not solely in the actions of an individual but also in the ideological disposition and biases of an entire
scientific field… The Weekly Standard’s Andrew Ferguson (2015) argued that the liberal bias of peer reviewers led to uncritical scrutiny of a study whose conclusions reinforced liberal assumptions

The Wall Street Journal rather excitedly called the LaCour affair “the most spectacular scientific fraud in a generation”, suggesting that LaCour’s paper was published because it

Flattered the ideological sensibilities of liberals … LaCour’s purported findings let them claim that science had proved them right… Similar bias contaminates inquiries across the social sciences, which often seem to exist so liberals can claim that “studies show” some political assertion to be empirical.

Hilgard and Jamieson say that media coverage of retractions can be, and is, used to suggest the idea that “science is broken”. However, the same retractions could also be seen as evidence that science is a self-critical, constantly self-correcting enterprise. Whether the “broken” or “self-correcting” narrative dominates depends on how the issues are framed, and Hilgard and Jamieson suggest ways that scientists and science communicators should emphasize the positives.

In my view, this is an interesting chapter, but perhaps the most interesting point is this almost throw-away comment towards the end of the piece:

In many ways, the strengths of science – its self-criticism, transparency, and self-correction – lend themselves to exploitation in a partisan public sphere.

From a public relations standpoint, it would be better if there were no retractions in science at all. It would look better if all scientists agreed with one another about everything, and never criticized published work (at least not in public). But that wouldn’t really be science. It would be a cult. We shouldn’t be concerned about retractions – quite the reverse. We should be concerned about people and groups that never admit their mistakes.

I have myself previously written that “science is broken” and that it needs “fixing”, but I was using these terms in a specific sense: by “science” I meant the particular system of publication and evaluation of research that we currently have. I do think this system is broken, but this system is identical with ‘science’ only in a narrow sense.

In a broader sense, science is the whole community of people who are researching how the world works. In this, broader, sense, science is self-correcting and everything I’ve written about “fixing science” is meant as a contribution to the grand process of self-correction.