Abstract

One of the Commentaries in this issue is something of a departure for the JSE. The paper by Jaime A. Teixeira da Silva concerns retractions in scientific publications, a topic that has been receiving increasing attention in recent years, apparently coinciding with an increasing number of retractions over the same period. Because the SSE and JSE focus not only on specific (usually controversial or neglected) domains of scientific investigation but also on broader issues concerning the practice of science itself, I figured that the cluster of issues surrounding retractions might be of both theoretical and practical interest to JSE readers. The sheer number of retractions is enough to give one pause. A recent survey by R. Grant Steen of the PubMed database from 2000 to 2010 identified 788 retracted papers (Steen 2011). For 46 of those papers, Steen was unable to find formal retraction notices. So his survey dealt with the remaining 742 papers for which he could obtain such notices. The reasons for retraction were broadly identified as fraud and error. The former included data fabrication and data falsification, and the latter included (among other things) plagiarism,1 scientific mistake, and ethical issues (violations of accepted publication practices—for example, IRB [Institutional Review Board] violations). Steen found that the reason for retraction was more often error than fraud—73.5% as compared to 26.5%.