Standards for Self Plagiarism

Question

Caveats and Context

I have been out of academia and the publishing game for nearly 2 decades so this might just be a question about changing mores, difference between disciplines, or just general mis-remembering on my part.

My question is driven by this question where someone is accused of self-plagiarism for reusing text from a grant proposal.

Comments on that question are making me feel like the definition of self-plagiarism has significantly shifted and things like below are not OK by today's standards.

Type 1: Self-Plagiarizing Proposals

This is almost a category error in my mind, of course text is recycled from proposals all the time. I know that I have seen or even done:

• Revising and resubmitting a proposal to a different funding agency when it was rejected elsewhere with minimal changes.
• Reusing entire sections describing methods or facilities used in the research.
• Reusing sections defining the problem space, notations, framing what we are working on.
• Doing the verb tense dance to convert a proposal into a report, "we will investigate" -> "we investigated".

Type 2: Publishing Incremental Versions

It was not uncommon in my area to see the "same" paper move through the levels of rigor in venues as the work is refined. For instance you might see a paper start off as an unrefereed seminar paper, then move to a refereed conference paper and then finally be expanded to a longer journal article. I thought this was normal and nobody tried to hide that they did this, there is always some text or comment describing how Paper X was an extended version of Paper Y that had been presented at BLAHConf in whatever year.

I even remember one of the paper indexing sites that would annotate the PDFs for you so you could see what text in particular was different between versions.

I don't think that this was the same thing as playing the MPU (Minimal Publishable Unit) game, but maybe it was.

Type 3: Reboiling the Boilerplate

This is one I was and am (in industry) flagrantly guilty of--all the time--every day.

Maybe this is just the result of working in an empirical data driven type field, but a significant fraction of presenting a result is recording the methods and techniques applied to the problem. I don't feel bad going back to my previous text and pulling the paragraph describing my Turbo-Encabulator and then tweaking it to mention that it had been malleated by 1.5 nofer trunnions.

One important thing in this case is I am not actually referring to the previous text, that text and the new text are actually describing two similar configurations in "my lab". It feels ludicrous that this would also be considered to be self-plagiarism in any useful meaning of the word.

Honest Question -- Not Yelling At Clouds

This question is not meant to sound ranty, I am safely removed from academia, I have no dog in this fight.

My question remains, are those cases of self-plagiarism by today's standards? Has the standard shifted over the years? Must all text be generated anew for every publication?

Answer 1

Proposals are tightly-controlled by rules. The only thing that really matters is legally handling the money. Different funding agencies often explicitly allow you to submit the same proposal. They will likely reject it anyway. If you get funding from both and you proposed the same thing, you will only be allowed to accept one of them (or else go to jail).

Also many journals have a cutoff for incremental improvements, for example of 30 percent new material that is required. Given that number of submissions to journals has increased dramatically, this kind of requirement probably has as well, with many journals insisting on entirely new material. In ye olden days you could also get away with only a few references. Now your reference section needs to be some kind of tour de force of reference thoroughness too.

There is finally the issue of copyright and the (rather problematic) fact that journals want to sell our work as a product. Which we gave them for free. Plus helped them screen the products we would most want to buy for them. So they tend to push the motivation for demanding the text be as novel as possible, and now they have automatic tools to help enforce this.

You'll note none of the above required the concept of "self-plagiarism", which I tend to think is an unnecessary concept we already have covered (when it actually makes sense) by other rules of ethics.