“I often quote myself. It adds spice to the conversation.”
Introduction
The opprobrium of self-plagiarism makes one a scientific pariah. This paper provides a critical evaluation of the discourse of “self-plagiarism” in scientific publications. We first show that “self-plagiarism” is a misnomer and should be replaced with “unacceptable duplication.” We distinguish acceptable from unacceptable duplication in scientific communications, providing a typology of both. We then review issues of copyright infringement and propose a preventive ethics approach to unacceptable duplication.
“Self-plagiarism” is a misnomer
Scholarly work in scientific peer-reviewed journals is expected not to have been published previously. This requirement for originality is clearly violated by plagiarism: the representation of the original ideas or words of others as one’s own. “Self-plagiarism” is used to characterize alleged violation of the requirement that scholarly work in the peer-reviewed literature be original: the reuse of one’s own text. This gives the false impression that the ideas and words are original.
The precise description of this violation of intellectual integrity is “unacceptable duplication.” The use of “self-plagiarism” communicates 2 ideas: the work is not original and the work is stolen from someone else.
Duplication is not original. But it is also not theft. As Lindey has stated, “A man can’t steal his own umbrella or his own automobile. He can (though he may not) steal his own brain child. If he takes a published work of his, alters it… and puts it forward under a new title, he wrongs his first publisher, cheats the second, and swindles his readers.”
“Self-plagiarism” is a misnomer
Scholarly work in scientific peer-reviewed journals is expected not to have been published previously. This requirement for originality is clearly violated by plagiarism: the representation of the original ideas or words of others as one’s own. “Self-plagiarism” is used to characterize alleged violation of the requirement that scholarly work in the peer-reviewed literature be original: the reuse of one’s own text. This gives the false impression that the ideas and words are original.
The precise description of this violation of intellectual integrity is “unacceptable duplication.” The use of “self-plagiarism” communicates 2 ideas: the work is not original and the work is stolen from someone else.
Duplication is not original. But it is also not theft. As Lindey has stated, “A man can’t steal his own umbrella or his own automobile. He can (though he may not) steal his own brain child. If he takes a published work of his, alters it… and puts it forward under a new title, he wrongs his first publisher, cheats the second, and swindles his readers.”
Acceptable duplication
Throughout history, the creative archives are littered with duplication. Johann Sebastian Bach frequently raided his own work to integrate into new pieces, the familiarity delighting audiences. The point is that there is no expectation of originality in works of art from the same creator. This is very different from scientific publication in peer-reviewed journals.
In our judgment, there are 4 types of acceptable duplication in scientific literature.
Textbook chapters and non-peer-reviewed articles
Readers of medical textbooks and non-peer-reviewed articles expect a review of the current state of knowledge and practice and expect that the chapter will not include original scientific work. When any previous work is referenced, the result is an acceptable type of duplication.
“Background” and “Methods” sections of papers
Scientific authors often repeat verbatim the description of their methods or the background in subsequent publications, since these sections often do not change. Citation to the original source should always be made.
Acknowledgement of permission for reuse
When one wants to use one’s own previous words, including tables and figures, one should conform to publishers’ requirements for doing so. Translations of one’s work into another language for historic or landmark papers to reach the widest audience possible are acceptable, provided acknowledgement is made. Previously published work may also be duplicated, with permission, in an authored book.
Abstracts turned into papers
An abstract is not expected to present all of one’s pertinent data and discussion. Prior presentation can be acknowledged.
Unacceptable duplication
The pressure to “publish or perish” can and does lead to authors to resort to reuse some or all of text that they have published elsewhere in an attempt to produce another publication. Without acknowledgement, this is unacceptable duplication. There are 2 types.
Concealed redundant publication
Up to 10-20% of all biomedical publications may have some duplication within them. An Australian study in 2007 by Bretag and Carapiet found that “over 60% of authors in the sample [had] re-used text from a previous publication without appropriate citation.” This repetition violates the requirement and expectation of originality and wastes precious journal space and reviewer’s time. Counting the same data more than once leads to an inappropriate weighting of results that in turn distorts the available evidence, which could negatively affect best practice and health policy.
Data fragmenting and data augmentation
Data fragmenting is sometimes referred to as “salami publication” or “salami slicing” and is a variant of the smallest-publishable-unit/least-publishable-unit strategy whereby authors seek “the minimum amount of information that can generate a publication in a peer-reviewed journal.”
Should study outcomes always be published together as the Office of Research Integrity insists? Some would argue not, as shorter articles with narrow outcomes may be preferred by both readers and journal editors, the latter who can gain more from 2 published articles over 1 longer one. When this strategy is used in absence of citation to one’s relevant previous publications, a subtle type of unacceptable duplication occurs. As a consequence, quality of publications is then lost to quantity, perhaps from motives of self-interest such as improving the author’s reputation or chances for promotion. If study outcomes are reported separately, the author must carefully consider in each instance if the division is necessary.
Data augmentation involves the collection of additional data that are added to the original study and republished without citation. This creates the false impression that the second data set is completely original.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
