Association for Public Policy Analysis & Management

Sir Isaac Newton once said, “If I have seen further it is by standing on the shoulders of Giants”. Appreciating predecessors are often symbolized by citing predecessors’ works. Because of their perceived instrumental and symbolic roles, citations became a common evaluation method in science policy. Yet scholars have criticized the uses of citation counts for a number of reasons. Some scholars argue that citation fails to capture the “fair” distribution of recognition. Others view citations as a rhetorical device in the construction of facts rather than a form of idea recognition. In this paper, I argue that under some circumstance, citations are more likely to be functioning as myth and ceremony adopted by scientists. The evidence is illustrated from analyzing bizarre citation behaviors of some scientists, citation of retracted articles. Surprisingly, around 35% of citations from retracted articles are made after the articles have been retracted. This paper shows that such “sloppy” citation behaviors can be explained by both the ignorance (scientists’ cognitive distance from retracted research area) and field competitive pressure perceived by scientists.

This paper exploits the population of retracted articles and their citing articles from the Web of Science and PubMed database. The analysis is based on 104,404 articles (excluding self-citations), which cited 5,717 retracted articles. I argue that field ignorance and strong field competition are circumstances where citations are more likely to be functioned as myth and ceremony. Both text and discipline vector distances between the retracted and citing articles were positively correlated post-retraction citation. That is, scientists who are unfamiliar with the field of retracted research were more likely to cite them after retraction. Meanwhile, scientists who are in a highly competitive environment, which is measured by topic intensity, were more likely to cite retracted articles.

Using citations information form retracted articles, this paper discussed different circumstances where citation practices are more likely to be a reflection of institutionalized practice rather than functioning of peer recognition. Not only this paper sheds lights on growing literature on bureaucratization of science, but the finding is also a cautionary tale against uses of direct citation counts as a performance measure by policymakers.