Science

Is it still plagiarism if you copy yourself?

Rage is rampant in the chemical community over the issue of self-plagiarism.  It smacked me in the face today when I was editing a coworker's conference abstract due to four identical sentences. These sentences were on the same topic as our previous work even though the science is different. In my mind they seemed viable. However, with this issue blowing up, we knew the risk was not worth taking. 

Ronald Breslow of Columbia University, has been 'charged' with self-plagiarism. To begin, I want to fully elucidate this guy's rock start status:

-He used to be president of the American Chemical Society.

-He has also won their highest award and MANY others. 

-He is a full professor at an Ivy League university. 

Not convinced yet?

-He is also a member of the US National Academy of Sciences.

Basically, he is a big deal. So when someone who is a big deal is surrounded by controversy then it is a big deal. 

So what is the uproar about? He published his own text verbatim in several different journals. The research topic recently got some press that resulted from an American Chemical Society press release titled: "Could advanced dinosaurs rule other planets?". Anytime something like this is said, people start to notice. And notice they did… 

The first two papers were published in 2011 in the Israel Journal of Chemistry and Tetrahedron Letters. These journals each have different publishers and are dissimilar to where his most recent paper got published, the Journal of the American Chemical Society (JACS). Publishing papers here is generally perceived as prestigious. It is one of the most well read and cited journals in the chemical community. So when he used redundant text here, people began to catch on. Stuart Cantrill, chief editor of Nature Chemistry, did a side by side comparison of his previous works to the current JACS paper. He (@stuartcantrill) tweeted these pictures earlier this week: 

In case you were wondering, it is the highlighted text that is copied verbatim from his other works.

That is some serious self-plagarism. The American Chemical Society specifically makes a statement in its code of ethics about this issue: 

Authors should not engage in self-plagiarism (also known as duplicate publication) – unacceptably close replication of the author’s own previously published text or results without acknowledgement of the source. ACS applies a “reasonable person” standard when deciding whether a submission constitutes self-plagiarism/duplicate publication. If one or two identical sentences previously published by an author appear in a subsequent work by the same author, this is unlikely to be regarded as duplicate publication. Material quoted verbatim from the author’s previously published work must be placed in quotation marks. In contrast, it is unacceptable for an author to include significant verbatim or near-verbatim portions of his/her own work, or to depict his/her previously published results or methodology as new, without acknowledging the source.

It seems that Breslow is catapulting across this boundary. Since the publication of his JACS paper, it has been removed by the publisher. The editor, Peter Stang, is determining whether or not an ethics violation has occurred. Either way, some serious politickin' will occur to determine whether or not the paper is retracted. Allegedly, the science is different from the other two papers (it got removed before I could take a read myself), but it is obvious that significant portions of the paper is redundant. 

I think a major line here has been crossed, but it is not the first time I have noticed something like this in the scientific literature.  On more than one occassion I have seen some *achem* similar works.

However, this leads me to my next question. What is a "reasonable person" standard when deciding what constitutes self-plagarism? I am sure the language here is purposefully vague. 

They specifically mention that 1-2 sentences will likely be overlooked. However, as an author this creates some concerns. For example, methodology sections are often similar to authors' previous papers. Reserchers generally use the same techniques, so is it self-plagarism to copy your own methodology sections? Once the quality of the writing is optimized, I think it hurts the overall product if you have to rewrite it just so you don't plagarize yourself

Today I took an abstract from a recently published paper that we wrote and performed some minor rewrites on a similar abstract for a different system. Let me be clear, the science is different. To be honest, I am not sure if it hindered the overall quality of the writing, but was it necessary? 

The initial sentence:

Hydroxyurea is the only FDA approved treatment of sickle cell disease. It is believed the primary mechanism of action is associated with the pharmacological elevation of nitric oxide in the blood; however, the exact details of this are still unclear. 

The rewritten sentence:

Hydroxyurea, the only FDA approved treatment of sickle cell disease, is believed to increase pharmacological nitric oxide (NO) levels as its primary mechanism of action. Despite this fundamental understanding, specific details are still undetermined.

Sure these are minor changes. I only showed you 1-2 sentences and that might be overlooked, but I changed a total of 3-4 redundant sentences.  The remainder of the abstract was completely different. Again, the science is different and I believe that this is the important part.

Is there a line? Would I have crossed it by leaving 3-4 verbatim sentences? Should following ethical guidelines leave this much gray area? Does following ethical guidelines always leave this much gray area? There isn't a consensus about Breslow's self-plagarism yet, but the fall out will certainly be interesting. 

Jacqueline

Jacqueline, a true Floridian, wandered up to the tundra of Athens, Georgia to receive her PhD in computational quantum chemistry. Returning to her roots, she is currently working as a postdoctoral researcher in Tampa in the field of computational biochemistry investigating the wonders of penicillin-like drugs. When she is not slaving over the computer, her varied interests include international travel, Brazilian jiu jitsu, kickboxing, fancy food, (American) football, and Belgian quadrupels. She is also the founder of EligibleReceiver.com, a football blog with an exclusive female writing staff. Check out her sports ramblings there or follow her on Twitter @jhargis9.

Related Articles

39 Comments

  1. Speaking as a writer I don't think you self-plagiarized, and as a non scientist I could certainly see the logic of cutting and pasting the methodology sections if they are the same — I mean, if you have to do a polymerase chain reaction in two different sets of experiments do you have to rewrite the explanation each time? That seems silly.
     
    Also, nobody accuses the AP, Reuters or Bloomberg writers of self-plagiarism even though at a wire service you have the same story with minor tweaks appear five times in a day. You have to do it that way because you are given like 20 minutes to add the update and it's easy to end up writing a thousand words a day. Without the self-plagiarism of the wire service, the reporter would die of exhaustion.
     
    Now, when I write different stories I try not to plagiarize myself with the same turns of phrase but it is hard to not do that unless you stay alert. I've been practicing it, so I can do that, but somene who does not do this fo a living might not (just as I would be utterly unaware of proper lab technique).
     
    Breslow said in one instance he was not writing a science submission but a kind of roundup of the state of the art. If he isn't an author or that practiced a writer to begin with I could see this happening if he just had less time to do a proper job.
     
    That said, I don't think you should wholesale C&P your stuff if you told your editor they were getting a new story. But that's a journalist's view, not a scientist's. So ultimately I can't say for sure if this is a serious ethical violation — I feel liek there's a few facts I woud want to get first and a little beter understanding of what he was supposed to be doing in each case and what the expectations were. As I read this the ACS standard seems pretty reasonable, tho.
     

  2. Honestly I don't understand the problem.  how is it an ethics violation to repeat yourself?  just to give someone something slightly different to read?  are you promising the journal completely new wording each time?  you could say it's lazy but "unethical" I don't see.

    1. The problem is that scientific publications are considered to be new research. Publishing the same paper and pretend it came from different experiments, creating the illusion your data has been replicated is unethical.

      I wouldn’t see the problem in repeating yourself in abstracts, since abstracts usually start with background information about what you are studying. Previous information should not be changed unless it gets refuted.

  3. I'd be especially interested to hear what people think about 'self-plagiarising' methods sections. I'm pretty sure that's what we used to do back when I was a grad student (decades ago). Is that no longer the practice? As you said, once you've found the clearest way to explain a technique, you don't want to screw with it.

  4. This is something I also worry about. A lot of my writings share similar background and methods sections. There is a point at which you run out of truly new ways of saying somethings. And it's BS to say that if you change a couple words it's not plagiarism. That wouldn't fly if you were rewording someone else's work. That would still be plagiarism. There simply must be a different standard when you're working with your own intellectual property.

    1. That said, based on that highlighting, Breslow's copying looks pretty blantant and definitely over the line.

      1. I only see his copying as a moral issue if he presented the material as though it were new when it wasn't.  No evidence was given for this.  Beyond that, it's a matter of contention between publishers — and if they didn't bother to check whether the information had been released before, they shouldn't have assumed it wasn't.  Writers and researchers shouldn't be obligated to mark the exact date at which they first recorded every paragraph they put into an article or paper.
         
        The idea that you have to quote yourself, when using your own words, is almost ridiculous on its face.

  5. I think if you apply common sense there should not be much of a problem. Imho you can copy complete introductions to papers (I mean your own) without behaving unethically but you might run into copyright issues. With methods, unless the method itself is the new thing in the paper, you would usually write something like: “The data has been analyzed according to the previously described method [Citation]. Briefly…[verbatim copy].” That is how people usually do it. Omitting that you have previously published it is I think bad style. If it is unethical I don't know.

    1. I doubt there was a copyright claim involved here.  In order for that to be the case, ownership of the rights would have to have been transferred from the author to the publisher.  As far as I'm aware, that is not typical for academic papers.  If it is, though, it's a moral injustice and another reason* to disdain the rather conservative science publishing industry.
       
      * The first reason was that they construct paywalls around what was presumably supposed to be a knowledge building system, preventing the public from freely accessing and evaluating many papers.  No library would get away from charging those kinds of per-view, per-work fees.

      1. Actually, at least in the sciences, it is quite common to transfer your copyrights to the publishers. Only with the newer open access journals, or if you opt-in for an open access article when you publish you retain the copyrights yourself.

  6. I can ALMOST see this – the journal gets publication rights of material it accepts, but some of the same sentences it has published are showing up in other publications. Why can’t there be a simple disclaimer on later articles to the effect that “some statements have been previously published?” Or with so much of this apparently going on, perhaps EVERY article could state “Some material may have been previously published.”

    While not the same thing (music instead of prose), I’m reminded of John Fogerty being sued by his old record company from when he was in Credence Clearwater Revival, because his 1980’s hit song “The Old Man Down The Road” sounds a lot like the older CCR hit “Run Through The Jungle” (they’re both on Youtube for your listening pleasure), though there was no dispute that Fogerty wrote both songs. I just listened to them both to remind me, and they DO sound quite similar, being one main chord through most of both songs.

    John Fogerty won the case, but he’s now famous for not just his music, but for being sued for plagiarizing himself.

  7. I think the idea of self-plagiarism is stupid.  Really.  Of course, I appreciate the concern of publishers, but it's not *their* work.  
     
    Really.  Stupid.

    1. I think the problem here is not only that you might actually have signed over the rights to a particular work by publishing in a journal, but also that duplicate publishing is bascially "double dipping." In academia, advancement is based on publishing– publishing *a lot*. Publishing the same thing more than once can be considered unethical primarily for that reason. You are supposed to be publsihing new work.
       
      I don't think that really applies to very similar methods sections, though.

      1. You're right about the causes as to why academia sees this as a problem.  They want PAPERS, PAPERS, PAPERS, all new, all the time.  Your entire advancement in some fields may be based on the number (though not necessarily quality) of the papers published.
         
        It's hardly unethical to multi-publish anything, unless the author in question explicitly deceived someone in order to do so.  No academic paper of any consequence is ever standalone; it is always based on previous work, sometimes that of the same author(s).  Trying to force people to re-write their old publications is silly.  Now, occasionally it may make sense to leave a citation or brief summary instead of actually repeating yourself — but in other cases a reference or summary simply wouldn't explain the necessary information in sufficient detail.

    2. Not to mention the tenured, "rockstar," full professor's copypasta is taking up the space in which younger scholars need to publish in order to get or keep a job.

      1. That concern assumes there is a max length limit on the size of the journal, or issues thereof.  This is increasingly irrelevant with fewer and fewer journals actually publishing print issues and instead transferring to completely digital means.
         
        Quality, first and foremost, should be the main concern from the publisher's perspective.  If there is a reduction in comprehension due to excessive redundancy, that's a real problem.  That wasn't the concern claimed here, though.

        1. Again a comment for kagerato that doesn’t show that much insight in the matter, which is strange considering the number of posts. Although your argument is reasonable for journals only appearing on-line and not in print, it is not so for journals that are published in both forms (as is the case for both Tetrahedron Letters and JACS). Further I think, and this is supported by the fact that if you publish in an printed journal you sometimes have to wait for months before your article gets published, that publishers do not want to have their issues increase (almost) unlimited in size. Labor costs would go up while not generating more revenue, but more importantly it reduces the “exclusivity” of the journal. So I fully agree with delictuscoeli that he takes up space when he is verbatim copying his previous work.

      2. It wouldn’t have mattered at all though. This wasn’t an actual paper but a prospective that he was invited to write.

  8. Self-plagirism really doesn't make sense to me.
    If you wrote something, then it's your right to rewrite or copy it into whatever it is you're writing.
     
     

  9. Even if the science in these papers is different, the copying of that much text is plagiarism for the following reasons:

    – Most scientific papers are written by multiple authors (since I do not have access to the paper in question, I do not know if this pertains to the original paper).  These sets of authors typically change from paper to paper, with a 'corresponding' author remaining constant .  If the set of authors from the original paper are not identical to the sets of authors on subseqent papers that used the same text, this is literally plagiarism, with the newer authors stealing the work of previous authors (even if one or more of the original authors are also authors on the plagiarised work).

    – It is common in scientific papers to cite previous work instead of repeating text from other papers.  For example, in a materials and methods section, instead of copying and pasting the description of a method that had been published before, it is typical to write 'as has been previously described (citation)'.  This is done for brevity, clarity, and to prevent plagiarism.  There is no need to repeat large sections of text in a new manuscript.

    – Abstracts are a one paragraph summary of the work.  3-4 very similar sentences in one paragraph is a significant amount of copying.  Also, most papers are around 10 pages, give or take a few, so at least four pages of copied/pasted text is significant.  A few sentences in the introduction that are paraphrased from a previous paper when describing the same thing is much different than several pages of copied text.

    1. Your first point is more about mistaken perceptions of the reader than a problem on the part of the writer.  No assumptions should be made in multi-author works about who did what, unless exact specifications are given.  When one doesn't know, instead of simply assigning credit arbitrarily, ask.
       
      Also, if subsequent papers share some but not all of the same authors as older ones, then it is fair to think that any shared passages between the two must be the work of the authors in common alone.

      1. First you note that one should not assume which sections of a paper are written by which authors, and one sentence later  you state that you should assume which authors wrote which section if the original work appears to have been plagiarized.
        However, multi-authored papers are often mostly written by one of the authors.  The other people are included as authors not because they wrote a significant amount, but because they contributed in other ways, such as running some of the experiments and being the person who won the grant money to fund the work. 
        'Authorship' in a scientific publication does not mean the same thing as 'authorship' in other contexts.   Because of this, the authorship of papers listing multiple people is generally considered shared among all the authors.  The work cannot be chopped down to assign authorship based on who wrote what paragraph because all the authors contributed intellectually to the paper.  If some of the original work needed to be mentioned in a new paper, one should quote or paraphrase and cite the original work.  If this wasn't done, it is plagiarism.

        1. "First you note that one should not assume which sections of a paper are written by which authors, and one sentence later  you state that you should assume which authors wrote which section if the original work appears to have been plagiarized."
           
          Two different things there.  One, don't assume you know something you don't.  Two, you may infer authorship to some degree if you see repeated words from a previous work.  This doesn't mean you should infer that, obviously, it's just attempting to apply logic.  I don't assume plagiarism by default in either case, and the concept of self-plagiarism pretty much ruins the entire point of the concern.
           
          As to the second part, yes.  Sometimes only one of the people listed actually wrote the paper, and the others did research or investigation to produce data/images/whatever else.  Unless this is recorded somewhere, though, you still have to ask to really know.  Again, you might assume several things based on contextual information you have about the authors, but if those assumptions turn out to be wrong it would be the reader's fault, not the writer.

  10. I think the problem here is in trying to pass off substantive writing as if it is something new. Methodology is more procedural writing, and if your method stays the same, I see no problem with recycling the language. But the findings that follow from the methodology are only novel once, and publishing them in multiple journals seems a bit sketch.
    I worked on a legal journal this past year, and played a role in editing a paper by an author who copiously cited his own work, often in large verbatim chunks. On the one hand, he did cite his own articles and books as sources, so there's kind of an intellectual honesty there, but on the other hand, I kept thinking to myself, why doesn't this guy find something new to write about? Instead he keeps publishing minor variations on the same topic over and over again. When part of your career advancement stems from the number of articles you publish, it seems disingenuous to keep regurgitating the same basic material instead of actually continuing to contribute more substance to the scholarship.

  11. I think it's also important to point out that there's a difference between plagiarism and copyright infringement, although the concepts can overlap. Plagiarism is about attribution, primarily. You can copy verbatim without plagiarizing as long as you use quotation marks and cite the source, and you can paraphrase without plagiarizing as long as you attribute. Attribution really has no bearing on copyright infringement, however. Copyright is about a collection of rights that an owner of a copyrightable work is entitled to. You can infringe a copyright by reproducing a work in whole or in part, verbatim, or with some variation. You can even be liable although you have attributed the source of the original. Attribution really plays no part. In fact, with some very small exceptions, attribution is not one of the rights than an owner of a copyright is afforded, so failing to attribute has no bearing on the copyright infringement analysis.

  12. Is there a failure of the referee process going on here?  Isn't originality one of the things referees are supposed to assess?  Aren't they also supposed to check if citations actually support the claim being made?
     
    For example, if you write "As demonstated by xxxx (2010), blah blah blah in the case of y.  We find that blah blah blah is also true in the case of z, using similar methods", do they actually check that xxxx did that, and that the methods you describe are similar to those used by xxxx?  Shouldn't they notice if xxxx's words (even if you are xxxx, so the words are your own), are substantially identical to yours in the new paper and call foul?  Or are they unlikely to do this if you are a rockstar in the field?
     
    If I repeat this post on purpose three of four times, would I be doing the same thing, or just annoying everyone?   Is it just as funny to ask everyone to imagine I had done that, and not waste the bandwidth?
     
    Finally, did everyone notice that every single sentence in this comment is a question?  Should I shut up and go away now?
     

    1. Academic publishers are supposed to verify several things before they commit an article permanently to their journal(s).  Mainly, that there are no egregiously or obviously wrong factual claims, that the work is at least partly original, that the work meets basic scientific process, that the authors named are correct, and probably several others.
       
      Doing that is work, though, and a lot of bad papers make it through.  Whatever your field, an interesting experiment to do is check how many papers are published which have zero or very few citations anywhere else.  There are multiple causes for that, including network and popularity effects, but one simple cause is that the work is not really especially original or insightful.  It happens a lot; good science is far from easy.

  13. This is beyond ridiculous. He re-worked an article to re-publish in a different journal. Unless the journal that published the original article owns exclusive rights to the article, there is NO simply no legal or breach of trust issue here.  

    1. As far as I can tell, Breslow is not in any legal trouble for his re-publishing.  It's merely some academics chastising him for bad process and presumably deceiving them (even if he actually didn't).
       
      Even actual plagiarism doesn't land you in legal hot water unless someone would have also have a valid copyright infringement claim against you, as bunsnip explained above.

  14. It is absolutely unethical in this specific case because it's explicitly against the ACS code of ethics. If the author does not agree with that code of ethics, he should not publish in that journal.   As someone has already mentioned, plagiarism is more about dishonesty than it is about legal or copyright issues. It is an ethics issue first and foremost. The thing that makes plagiarism in general unethical is not just the stealing, but also the attempt to pass off information as new and original. If the text is not original to the document, it must be attributed regardless of the source. Self-plagiarism is a case of fraudulent behavior, not of stealing.  
     
    It is considered self-plagiarism when a student turns in a paper to two different professors for credit for two different courses. Why is it any different when someone publishes something that is factored into merit review? It erroneously inflates publication records. And, really, how hard is it to re-phrase or re-write parts of your work that are redundant? Sounds like a bit of laziness.

  15. I've published a couple of times in academic journals, and I have always been required to indicate whether and what portions of the paper have been published previously. I can answer something like "Preliminary data and versions of Figures 1-3 were presented at the XYZ Conference." In the academy, it is absolutely considered unethical to rewrite the same results for another paper. (While this kind of thing happens more than we'd like, each individual journal paper is supposed to present something novel.) Submitting what is effectively the same paper is absolutely beyond the pale, and I guarantee that Breslow had to misrepresent himself in order for this to be published.
     
    I will admit that I find summarization of previous results to be a very difficult section to avoid self-plagiarization. I've described another person's technology in one journal paper, two conference papers, and a poster session, and it's tough to use new words to describe the same thing over and over. While my current research is novel, the work I'm citing from three years ago is the same. I can see where this would be even harder for methodology sections, and I strongly dislike the the convention of simply citing earlier work ("prepared the gels in the method presented previosuly [4]"). I'm currently trying to use a procedure from a group that cites themselves like this, but small alterations were made with each paper, leading to the newest results. Having to cobble together the entire process from multiple papers is quite annoying for people who want to replicate and build off of your work. 

  16. Good point, Will. Authors choose to sign off on publishers’ ethical standards in order to get their work published in their journal/s.

    From the iThenticate “Ethics of Self-plagiarism” paper:

    Publications manuals have a set standard regarding self-plagiarism. When an author publishes in a journal, the author often signs over rights to the publisher; thus, copyright infringement is possible if an author reuses portions of a previously published work.

    Copyright law “protects original works of authorship” (www.copyright.gov).

    The Chicago Manual of Style (2010) provides the author’s responsibilities in guaranteeing authorship: “In signing a contract with a publisher an author guarantees that the work is original, that the author owns it, that no part of it has been previously
    published, and that no other agreement to publish it or part of it is outstanding” (pg. 142).

  17. Self-plagiarism like this is because the plagiariser needs / wants more papers on his resume. Many papers makes it easier to get a new job.
    Some universities have solved this by instead of asking for resume listing all the applicants published papers they ask applicants to submit five of their best papers. Obviously these papers cannot be plagiarized (as that would not look good). In this way the point of self-plagiarism disappears.

  18. The ambit of self-plagiarism has been extended to ridiculous proportions. No one should misrepresent published data as unpublished, but expecting people not to have a few sentences in common is taking it too far. I’m in the health science community, and I’m tired of having to modify sentences in the introductions of my papers just so they sound ‘different enough.’

    To extend this logic, if I take some information from another publication and give them due credit, I should be allowed to reproduce their statement(s) verbatim. But even that is frowned upon. So people tend to modify the sentences from the papers they’re quoting–often erroneously–and corrupt the spread of information.

  19. In this case, the author seems to have copy-pasted a lot of stuff. This seems over the line. It’s one thing to have words and phrases in common between papers–most research builds on previous research. But when this guy says so much of what he’s said before, he should really just cite that paper and move on. By punishing him for this, though, we are just driving the problem underground. People will disguise their words more skillfully. There isn’t an easy solution to this.

  20. My old university had a policy that would absolutely nail this. It’s less about plagiarism, and more about academic dishonesty.

    If the point of a university is for a student to produce new (at least to the student, if not the course), and original research, then copying one’s own material as a new paper is dishonest.

    What this person did was not only academically dishonest (to say nothing of lazy), but it defeats the purpose of the very core of academia: to f-ing *learn*.

  21. But why stop there?! I mean, look at all those people who give speeches, and repeat themselves. Why, I bet some of them are charging each audience, while presenting some/all of the same material! Shout it out: “UNETHICAL!”

    I have a great idea. Let’s make it so you have to hold your hands up in the shape of parentheses and say your name and the year you are citing. Each time you say something you have said in another speech, be sure to cite yourself in this manner.

    If this works, after a while we can also add citing other people in the same way. Won’t it be fun?!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Back to top button