Nested self-citation: the citation of a paper’s least divisible unit

Abstract

Try to imagine that a figure, a table or an explanatory box in your main manuscript gets cited, in addition to citations to the main paper. Some scientists would no doubt be ecstatic at this unrealistic opportunity of gathering additional citations. This paper highlights a case in which a text box (Unger and Couzin in Science 312(5770):40---41, 2006. doi:10.1126/science.312.5770.40) within a larger paper (Couzin and Unger in Science 312(5770):38---43, 2006. doi:10.1126/science.312.5770.38), as well as the paper itself, are both cited, 33 and 8 times, respectively, according to Clarivate Analytics' (formerly Thomson Reuters) Web of Science. Both papers were published in AAAS' Science. This paper explores details of these citations and shows how four papers between 2007 and 2015 have cited both papers, including the text box. The argument is put forward that citation of least divisible units of a paper, in this case, a text box, are unfair citation practices, and since they refer to the citation of a part of the same paper, the term "nested self-citation" has been coined. Given the attention given in recent times to citation manipulation, citation rings and inappropriate citations, the risks of nested self-citations, including the skewing of citation counts, and of not correcting potentially misleading information, need to be explored.