A while back, on the same day as a court ruling on the Steven G. Salaita case, the UIUC chancellor, Phyllis Wise resigned. She was embarrassed by being caught in a self-plagiarism case. Most graduate students are well aware of plagiarism, copying someone else’s work without citing it, but self-plagiarism can be confusing. How do you copy from yourself?
In Dr. Wise’s case, she republished a paper that had 80% of the content duplicated verbatim from an early work – without ever citing that earlier work. Additionally, she had done the work with a co-author, whom she dropped for the re-publication.
The crux of the issue with self-plagiarism is the concern that material being presented as novel has previously been published. The reason it’s viewed as inappropriate is that someone is trying to multiply their credit for the same work. On the other hand, many academics do work that evolves, and it is natural to present the work at each stage as it is expanded on.
It can sometimes be difficult to understand what exactly qualifies as “novel”.
For example, say I’ve done a political survey and published the results. Can I republish the work in its entirety elsewhere? If not, can I republish the work elsewhere with a new commentary/interpretation of the data? If not, can I republish the work elsewhere if I gather more data, merge it with the old data and reinterpret the results? If not, can I republish the work if I add a second study and rewrite the paper in its entirety with a focus on contrasting the two studies?
Each of the above would be acceptable or unacceptable in different contexts within different research domains. In my research area (Computer Science), the view is you can republish a workshop publication at a conference, or a conference publication in a journal (the lengths of these increase, so you would be adding material). Sometimes a journal will solicit well-received workshop or conference publications, and republish them verbatim or ask that they be expanded upon. It is made clear that these are reprinted from the conference proceedings. The rule of thumb I’ve been told (for my research domain) when reusing material is that it isn’t self-plagiarism if there’s at least 40% significant new content – for example, just rewriting a section wouldn’t count as “new content”.
Being so frightened of self-plagiarism that every piece of information and data in a paper is completely novel strikes me as a mistake. There are times when new work builds on old work, and presenting both clearly makes sense. There are also times when it’s worth exposing separate research communities to work they might not find if it was only published in the other area. Part of the contribution, in this case, is placing the work in the context of another research domain.
Because it’s a more recent issue, opinions are all over the map regarding what is acceptable or unacceptable. My advice would be to talk to your advisor, other professors in your department, other graduate students, and the editors of any venues you’re applying at. Once you’ve heard everyone’s viewpoint, decide for yourself what level of reuse you think is appropriate. Make it very clear when you’re using material that has been published before, citing the previous publication and perhaps drawing the editors’ attention to it.
How much have you reused from past publications? Have you ever had any problems doing so? What is your reaction to Dr. Wise’s situation?