December 25, 2012

On Bullshit (in science)

More and more is being written about the reproducibility of biomedical research, including prominent recent editorials in the Guardian, New Yorker, and Atlantic.

Perhaps the most famous argument was advanced by Ioannidis in his essay "Why Most Published Research Findings Are False." Ioannidis' provocation is certainly worthwhile in drawing attention to the subject, but the largely phenomenologic nature of his argument does not help to explain the cause of the problem, other than to suggest the existence of flawed human behaviors.

It appears that scientific misconduct accounts for the majority of retracted scientific publications. Indeed, Retraction Watch and Science Fraud post almost daily notices of published papers that are admitted to have been fabricated.

To what extent is scientific fraud intentional, .e.g. a purposeful act of deception, as opposed to an incidental outcome of otherwise purely selfish human behavior? The latter is a common human vice, such as empty talk that is intended to impress. Colloquially, this is called "bullshit," but the vulgar nature of this term should not distract from its serious consequences in academic scholarship.

Philosophers have grappled with the nature of bullshit, and Frankfurt's essay "On Bullshit" is essential reading on this subject. In particular, the definition of bullshit is "not that it is false but that it is phony." Since bullshit is created "only on account of [its] supposed indispensability to a [selfish] goal other than deception itself," fraud is an inaccurate characterization of the problem that plagues biomedical research.

Instead, scientific bullshit is a more apt description. And as a result, changes to the current practice of science that limit the importance of impact factors, personal mythology, and self-promotion, and not vexatives about truth, integrity, or ethics codes, are likely to be more effective. Otherwise, I am afraid they may be just a bunch of hot air.