77

I am planning to write several papers exploring various aspects of the same scientific question. Each of these papers must have an introduction which motivates it and explains the relationship between the problem and what others have studied in the past.

It would be fantastic if I could simply copy-and-paste the same introduction, or, at least, 90% of it. This seems to me to be ethically unproblematic. After all, I need to say the exact same things every time, and I certainly don't mind the self-plagiarism. Am I hurting the reader in any way? I suppose I might be, if the reader desired an introduction which consists of original material, but that is an odd desire, isn't it? Its the research, the stuff that follows the intro, that is original.

Sadly, I have gotten wind that the majority of the research community apparently does not agree with the sentiments expressed in the previous paragraph. This leads me here to ask a series of related questions:

  1. To what extent is self-plagiarism in non-technical bits considered acceptable? I often see authors recycle paragraphs but I have never seen anyone cut-and-paste the entire section outright.

    I'd be particularly interested in learning whether norms on this vary across different scientific communities.

  2. How often do scholars find themselves trying to same the same thing in different words to avoid self-plagiarism?

  3. Supposing I insert a sentence to the effect of: "The introductory section 1.2 is taken verbatim from the author's earlier paper [1]." How likely are journal editors and reviewers to complain about this?

By the way, I am fairly certain they would be very likely to complain about a sentence to the effect of "We refer the reader to [1] for motivation to study this problem and a discussion of its relation to prior work."

  • 5
    Why not put all various aspects of the same question together into one paper? Can they fit? – scaaahu Aug 19 '12 at 4:42
  • 2
    They sort-of fit, in that they are asking different questions about the same model. However, it would end up being a very, very long paper that, as a result, would likely go unread. More importantly, I am probably going to be thinking about these questions for years, so I'll have to face up to the dilemma in this question one way or another. – morgan Aug 19 '12 at 4:50
  • 2
    I've always wondered about this regarding my posters which all look the same. – bobthejoe Aug 20 '12 at 9:00
  • 1
    I tend to agree with the comment about citing earlier uses. I'm curious what people think of the influence of "double-blind reviewing" where if you use a lot of earlier text and cite it you are also un-blinding yourself to the reviewers. – Fred Douglis Dec 4 '12 at 23:31
  • 5
    If we weren't in the entirely artificial game of "peer-reviewed [sic] publication" as the only operationally viable sense of "publication" (e.g., literal publication=making-public on the internet doesn't count? Crazy sense of "publication"...), then we'd not find ourselves discussion "self-plagiarization", which, as a deleted answer observes, is semantic nonsense. While "semantic nonsense" is not a fatal judgement, it is certainly a sign that something's been perverted... as I strongly believe it has if/when there is any point in discussing "self-plagiarism" ... [cont'd] – paul garrett Nov 27 '14 at 1:09
42

Firstly, quoting any work, including your own, without explicit attribution of the quote, is going to earn you a very bad reputation for plagiarism sooner or later. Just say no.

Secondly, and to state the obvious, the introduction is there to introduce the rest of the paper. As each paper will be unique, so should the introduction be: the point is to lead the user into the paper, rather than to give a grounding in the whole subject area.

Thirdly, why not wrap up all of the things you want to say in all of the papers' introductions, into a single review paper? It would contain all of the literature review, motivations, unsolved problems, areas for future research, that you want to reference in your other papers. And then, those other papers can just briefly & concisely properly reference and quote the relevant bits of that review paper in their introduction. It won't be as brief as "for review and motivation, see my paper DOI:ABC.DEF.GHI", but you should be able to distil the review paper down to a paragraph or two for each subsequent paper, and those words will be unique for each subsequent paper, drawing only on the elements you need for that paper.

  • 2
    I agree with this sentiment. While I don't think there's anything wrong with self-plagiarism per se, I have come across a set of papers written in this way, and as a reader it was pretty annoying to find the same text in each introduction. It meant I had to skip through stuff I'd already read to figure out what new stuff was being presented in the current paper. Having a unique and to-the-point introduction for each paper makes it easier for the reader to differentiate them. It should be done not for ethical reasons but because it's better writing. – Nathaniel Aug 24 '12 at 17:53
  • 44
    I have to disagree with your first paragraph. Plagiarism means taking others ideas and/or words and passing them off as one's own. It follows that it is not possible to plagiarize oneself. Multiply publishing one's own ideas, words or work can be problematic, but I think it is confusing and unhelpful to call it plagiarism. Repeating yourself too much is inadvisable, but the repetition of one's own words and ideas is not inherently unethical or problematic. – Pete L. Clark Dec 8 '12 at 18:33
  • 3
    But here the trouble is the specific broken contract. If I publish a paper and then later write a chapter in a book based on that paper, I'll be okay unless I've signed some insane contract with the paper's publisher. If I get suitable permission I can even include the paper in its entirety in a text. However, suppose I got permission from some publisher to include someone else's paper in my text and found that it was okay with them if I removed the author's name and replaced it with my own. I'm still committing the grave sin of plagiarism: no one party can excuse me of that. – Pete L. Clark Dec 8 '12 at 18:45
  • 2
    I agree with "As each paper will be unique, so should the introduction be". However, parts of the intro might be similar. I do find it annoying to rephrase perfectly good sentences simply because it appears in an earlier paper of mine. (Maybe I'm being overly scrupulous about this.) – mhwombat Sep 20 '14 at 18:37
  • 10
    I completely disagree with the philosophy of encouraging somebody to shove everything into a single paper. Each paper should tell a story. Maybe the OP has three different stories. – Peter Shor Sep 20 '14 at 21:31
26

Typically, you will not write all of these papers simultaneously in parallel, but one after another. And each time you write a new paper related to the topic, you will have a learned a lot more, and you will have a much better idea of how to explain the setting, what are the right definitions, etc.

Therefore you have a very good reason to re-write your introduction for each paper, and tailor it for the specific question studied in this particular paper. And this way there is no risk of anyone accusing you of self-plagiarism.

Incidentally, this way you can even go as far as experiment with different ways of explaining the basic setting of your work, and see if you get different feedback... After all, your introduction is the most important part of your work from the marketing perspective, and it is not easy to get it right.

  • 4
    That is true. But it may be that even as your understanding of the technical parts evolves, your reasons for studying it do not. Moreover, you still need to cover your bases by saying "A similar question was studying 30 years ago in the paper [1]..." in the same way. So even while there is good reasons to rewrite a portion of your intro, copy-and-pasting a substantial portion of it would still be tempting (provided it were acceptable). – morgan Aug 19 '12 at 19:44
  • $+1$ for this. It is easy to say "I already know how to motivate it, and I motivated it perfectly before, let me just reuse that". But it is very unlikely that the way it was motivated before was actually perfect. There must be many ways to improve the motivation, in light of more recent work in the field, and in light of the new approach of the new paper. – 6005 Jun 26 '17 at 23:15
  • A paper which has the exact same introduction as previous papers will cause people to wonder: "then why is this paper any different than the other ones? Why should I read them and not this one?" It is a red flag which suggests that your paper's introduction and motivation is not unique enough to justify the research that you did. – 6005 Jun 26 '17 at 23:16
22

There is a very dangerous (for your career) issue here. If your papers are so similar that the same introduction could be repeated almost verbatim in each, it may be that the content of the individual papers don't represent a truly significant advance. As an obvious example, if you are just changing one parameter in your code and running it again, with no new analysis or insight, you will run into trouble and may even get banned from submitting to offended journals.

I realize this is not a direct answer to the question, but I think it may be useful to people interested in this question.

  • 1
    I agree that this may be the case, but it isn't necessarily going to be the case. – morgan Aug 20 '12 at 18:50
16

As a referee, I would not have problems with either way of referencing your previous paper. I would be more pissed off if I checked one of your earlier papers and found that a section was ripped off that without notice.

  • 2
    Even if the section had no technical content? – morgan Aug 19 '12 at 19:40
  • 11
    Yes, even if the section had no technical content. (But then I'd already be miffed that you included a section with no technical content; even a discussion of motivation and previous results needs some technical detail.) – JeffE Aug 20 '12 at 2:03
  • 1
    @JeffE , you took my meaning too literally here; every section in a technical paper will have some technical content, of course. My meaning was that the section would not have any theorems, proofs, or technical arguments - just summaries of what other people have done as well as references to where, in practice, the problem under the question has come up. Apologies for stating my meaning imprecisely. – morgan Aug 20 '12 at 18:54
13

You should definitely cite everything you copy verbatim, no matter what. (I.e., you should say something like "the following background information is repeated nearly verbatim from Section 2 of BLAH", so the situation is unambiguous. It's not enough to cite the source without making the copying clear.)

To see why, note that there are four possible types of people:

  1. People who think it is OK to copy your own text without citation, and think it is OK with citation.

  2. People who think it is not OK without citation, but is OK with citation.

  3. People who think it is not OK in either case.

  4. People who think it is OK without citation and in fact you must not cite it.

I don't believe people of type 4 exist. This means you should always cite any text you copy (even if you wrote it in the first place). There's no way you can lose: people of type 2 require the citation, people of type 1 don't mind, and people of type 3 are a moot point since they don't want you to copy it in the first place.

In practice, I think a large majority of academics are type 2, but it really doesn't matter.

10

To what extent is self-plagiarism in non-technical bits considered acceptable? I often see authors recycle paragraphs but I have never seen anyone cut-and-paste the entire section outright.

There is no strict rule about that, apart from the fact that "the overlap between two papers must not be substantial". I think it's mostly a matter of content than size. For instance, if you were in your paper to reuse a technique/theory/tool that has been previously defined, by you or someone else. Then having a short section reintroducing the tool (for the sake of the self-containment of the paper) that is a copy/paste from another paper is OK. You found a nice and concise way to present a technique/theory/tool, there is no real need to reinvent it.

However, the introduction is quite different, because that's the part that motivates your paper, and present the results detailed in the paper. There is of course some high-level paragraphs/sentences that you can reuse (The need of general-appraoch is paramount in the context of general-problem because of general-reason), but somehow, if 90% of your introduction is the same than for another paper, then that probably means that either the 10% is not enough as a contribution alone (which doesn't seem to be the case) or that your introduction does not focus enough on why those 10% are important. If the major part of your 90% is background and related work, then move them to explicit sections.

I'd be particularly interested in learning whether norms on this vary across different scientific communities.

I couldn't speak for other communities, but something that is usually frowned upon (but many people still do it) in CS is "incremental research", where your next paper is just a small improvement of a previous paper of yours, and where your strategy from the beginning is to maximize the number of publications you can get from one idea, instead of trying to publish directly the full idea. In this context, self-plagiarism, especially in the introduction, could be seen as a blatant proof of incremental research.

How often do scholars find themselves trying to same the same thing in different words to avoid self-plagiarism?

Personally not that often, as I will mostly reuse "technical" parts, and try to write from scratch non-technical ones for each paper.

Supposing I insert a sentence to the effect of: "The introductory section 1.2 is taken verbatim from the author's earlier paper [1]." How likely are journal editors and reviewers to complain about this?

For the above reasons, I would complain about it: if your introduction is taken verbatim from another paper, then maybe the contribution of the paper I'm reviewing is not that novel. For background/related work section, I wouldn't bother mentioning that it's a copy/paste from another paper, however, I would certainly like to be directed to "full" version of the sections. Also, and that might be something field-specific, some submissions are double-blind, i.e., the reviewers are not supposed to know the identify of the authors, and obvious self-references are forbidden.

2

Recently, we submitted a paper to an IEEE Journal, the student in charge, apparently copied large parts of the text from a previous article we had in an IEEE conference.

We got a response letter from the editor that said something like this:

Every IEEE paper is checked for instances of plagiarism (including self-plagiarism), and we found that you made extensive use of copy-paste from your previous article BLA to write the current article BLA. I'm afraid I can't accept the paper with so much previously Published text.

Just a word of caution for these practices in IEEE Journals and Conferences

  • In these days, checking copy-paste is easy. I am not surprised at all. – scaaahu Jan 25 '13 at 9:21
  • 1
    Indeed, just wanted to help OP before he entertained the idea – Leon palafox Jan 25 '13 at 9:28
  • 20
    "the student in charge" NO, that would be the authors. Signing on as an author means you take full responsibility for the work getting credit when it is deserved and taking blame when necessary. – StrongBad Jan 28 '13 at 11:22
  • 3
    Such "practices in IEEE Journals and Conferences" need no "word of caution", in my view, but rather some applause. – Did Sep 21 '14 at 17:33
1

iThenticate has a white paper called The Ethics of Self Plagiarism. Within it are reasons not so self-plagiarise, namely copyright issues and the fact that some definitions of plagiarism include copying one's own work (which is contrary to other definitions).

They also go on to give guidelines on how to avoid self-plagiarism, which are helpful:

• Guideline 10: Authors who submit a manuscript for publication containing data, reviews, conclusions, etc., that have already been disseminated in some significant manner (e.g., published as an article in another journal, presented at a conference, posted on the internet) must clearly indicate to the editors and readers the nature of the previous dissemination.

• Guideline 11: Authors of complex studies should heed the advice previously put forth by Angell & Relman (1989). If the results of a single complex study are best presented as a ‘cohesive’ single whole, they should not be partitioned into individual papers. Furthermore, if there is any doubt as to whether a paper submitted for publication represents fragmented data, authors should enclose other papers (published or unpublished) that might be part of the paper under consideration (Kassirer & Angell, 1995). Similarly, old data that has been merely augmented with additional data points and that is subsequently presented as a new study is an equally serious ethical breach.

• Guideline 12: Because some instances of plagiarism, self-plagiarism, and even some writing practices that might otherwise be acceptable (e.g., extensive paraphrasing or quoting of key elements of a book) can constitute copyright infringement, authors are strongly encouraged to become familiar with basic elements of copyright law.

• Guideline 13: While there are some situations where text recycling is an acceptable practice, it may not be so in other situations. Authors are urged to adhere to the spirit of ethical writing and avoid reusing their own previously published text, unless it is done in a manner consistent with standard scholarly conventions (e.g., by using of quotations and proper paraphrasing) (pg. 19-25).

-1

Imo self plagiarism is an absurd concept. Because the very name implies that you have stolen your own ideas. Which is impossible.

A lot of if not all of what gets labelled as self plagiarism should really be called academic dishonesty.

There is also a huge cultural divide here. In some countries/cultures reusing your own work is not only acceptable but is in fact encouraged. After all if I have already written an image manipulation library that utilises the gpu writing it again from scratch is absurd but updating it to support new gpu features, fixing bugs and releasing it under the gplv3 licese is of course very desirable.

If everything is properly cited (what constitutes a proper citation is again a cultural issue) I really don't see any issue with using this code for several courses if it is applicable to them.

Also if an academic institution allows students to take courses similar enough that copy&pasting things between them is viable such an institution is at least partially at fault too.

And then there is the whole semantic issue since self plagiarism is in many languages similar to theft from oneself or to self victimisation (how exactly am I a victim if I take some lsd (well eth-lad since it's better but whatever) and have a nice 10 hour long trip with some nice sensory and mental effects).

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.