Last summer I wrote here about a plagiarism scandal involving more than 60 arXiv preprints, more than thirty of which were refereed and published in at least 18 different physics journals, some of them quite prestigious ones (see also the page at Eureka Journal Watch). At the time I wondered what action the journals involved in this scandal would take in response to it. Nearly six months later the answer to this question is now in: essentially none at all. As far as I can tell, almost uniformly the journals involved don’t seem to have a problem at all with being used to publish plagiarized material.
Unlike the journals, the arXiv has taken action. It has withdrawn the papers, replaced their abstracts with lists of where they plagiarized from, and put up a web-page explaining all of this. After the scandal became public, one journal, JHEP, did withdraw the one rather egregious example of plagiarism it had published. This was only done after JHEP originally refused to do anything about this when first contacted last March, arguing that since the plagiarized articles were cited in the paper it was all right, and besides, they would only consider doing something if the plagiarized authors filed a formal complaint. Copies of the correspondence about this (and much else) are at this web-site.
The nature of the plagiarism varied greatly among the papers withdrawn by the arXiv. Sometimes all that was involved was self-plagiarism (large parts of one paper were identical with others submitted by some of the same authors), but mostly what was being plagiarized was the contents of papers by others. Mustafa Salti, a graduate student at METU, had his name on 40 of the withdrawn papers, many of which have been published in well-known journals. I checked a few of the online published journal articles corresponding to the withdrawn papers and, besides the JHEP paper, I didn’t find any others where the online journal article gave any indication that the paper was known to be plagiarized.
A more complicated case is that of Ihsan Yilmaz, where the arXiv lists three of his eight arXiv preprints as withdrawn due to plagiarism and one as withdrawn due to “excessive overlap” with two other papers of which he was co-author. Very recently one of his Physical Review D papers, a paper that was not one of the ones on the arXiv, was retracted with the notation:
The author withdraws this article from publication because it copies text, totaling more than half of the article, from the articles listed below. The author apologizes to the authors of these papers and to the publishers whose copyright was violated.
After the scandal broke, Yilmaz had a letter published in Nature where he justified the sort of plagiarism found in his articles, claiming “using beautiful sentences from other studies on the same subject in our introductions is not unusual.” Evidently the editors of the journal General Relativity and Gravitation agreed with Yilmaz. They decided not to do anything about the papers they had published that were withdrawn from the arXiv, writing an editorial in which they defended the papers, while noting that “we do not regard such word for word copying of introductory and descriptive material by others as acceptable.”
I heard about the GRG editorial via an e-mail from a group of the faculty at METU, who write that:
The note is clearly quite unacceptable and insufficient in the fight against plagiarism. We cannot help but ask whether the Editors seriously believe that those who cannot compose their own sentences are in fact capable of producing genuine research worthy of publishing in General Relativity and Gravitation.
and note the retraction of the Physical Review D article, which they regard as a much more appropriate response
Update: Someone helpfully sent me pdfs of the two GRG articles, marked up to identify the plagiarized passages. Looking at these, I find it hard to understand why any journal would not withdraw such papers if they made the mistake of publishing them.
Update: The journal Astrophysics and Space Science is retracting four of the plagiarized papers, by putting up errata on-line which appeared today and are dated January 11, 2008, saying:
After investigation and at the request of the President of the Middle East Technical University (METU), Ankara, Turkey, the Editors of Astrophysics and Space Science have decided to retract this paper due to extensive plagiarism of work by others.
The papers involved are gr-qc/0505079, gr-qc/0602012, gr-qc/0508018, gr-qc/0509022.
The reaction from some folks with secure jobs will be – no one is reading these papers anyway, so in the end the problem takes care of itself.
The problem comes when one is competing for a job with the young guy with 40 publications in a system where all that matters in job applications is the number of papers published weighted by impact factor. Or when the dean with whom he is plagiarizing is the dean of the school to which one is applying for a job. Then this is a serious matter.
The other observation is this: folks learn how to behave from their advisors and the other folks in their environment when they are grad students and postdocs. Too many basically decent folks learn this way to think that awful habits are professional and normal. This is why it is important that there be voices exposing what goes on.
The GRG editorial is not so bad. An author who copies word for word introductory material from his own article is in no sense a plagiarist (can one plagiarize oneself?). This person has done nothing unethical if the new article, in which the copied introductory material is used, reports substantially different results than did the original article. What’s so horrendous about copying one’s own definitions? If one quotes one’s own theorems, one might very well have good reason to quote them as one wrote them originally, particularly if one wrote the thing well the first time. There is a problem if the author is submitting to different journals articles which are essentially the same, in the sense that the results they report are essentially the same; this is unethical, though it is also not plagiarism – it’s also quite common, at least in a mollified form – we all know researchers who publish four papers on a given topic when one would have been enough – and we know they do this to inflate their paper counts (and their citation counts – sometimes they and their collaborators publish a series of interciting and repetetive papers on the same theme) – and many of them do this without seeing anything wrong with it – rather, they think this is what is their job – publishing papers. This is unethical because it means the originality of the papers submitted is misrepresented and if the paper is published space is taken up by something repetetive, when it could have been used to publish something noval.
If copying is genuinely limited to copying one’s own introductory material, I don’t see much wrong with it – the problem is that in the cases identified by the ArXiv administrators this was not what happened.
Nothing happened to 15 plagiarist with 70 plagiarised publication
announced by arXiv.
Education writers of some biggest nation-wide journals did not write anything about arXiv plagiarism scandal.
In spite of this, arXiv scandal became the most popular mass plagiarism scandal which found place in Turkish media history.
———————————
Osman Demircan announced himself as the investigator of 2007 COMU Plagiarism Ring (6 plagiarist, 20+ plagiarised publications) declared by arXiv.
First day he spoke to journals :
“These are not plagiarised.
These are libels by Turkish and foreign scientists who are against Turkey.”
———————————
Who is Osman Demircan :
He is vice president of COMU since more than 8 years who decides who can work in COMU and who must be fired only according to his and his rings’ partisanship criteria.
He is a member of physics dept. which the 6 plagiarists of COMU belong to.
He gave jobs in COMU to all these 6 plagiarist.
He is the head of this plagiarism ring, and much more.
———————————
In 2001 and 2002 he claimed me to YOK (Higher Education Council; or the Higher Ideologic Despotic Counsil) and other university and
wrote that I should be dismissed since I am guilty since I’m a PhD student in that other university.
He illegally made me dismissed me from PhD education just a few months before completion of it.
He is one of the main organizers of the underground relations who illegally prevented me to have my PhD diploma for 5 years.
Pingback: thanks for linking | neverendingbooks
functor,
The problem with the GRG editorial is that this is not just about authors copying some small amounts of text from the section of another of their articles. Sure, they’re nothing especially wrong with that, and I don’t think the arXiv would act to withdraw a paper because of that. In the case of one of the GRG papers, the copying was from other people’s papers, and in the judgment of the arXiv administrators, was extensive enough to make them withdraw the paper. The judgment of the GRG editors was very different, not to take any action at all, even not to add a note that extensive parts of the paper were plagiarized from others. Lacking the arXiv’s software, I’m not able to automatically identify which parts of the GRG paper are plagiarized, but I trust them to not have withdrawn a paper simply because of small amounts of use of other people’s language in an introduction.
One other thing to keep in mind is the author’s retraction note for his PRD paper, where he acknowledges that more than half of the text of the paper was plagiarized.
I wonder whether these journal editors are doing nothing because they are lazy, because they are embarassed that their journals accepted these papers, or because they don’t see anything wrong with plagiarism. I’m not happy with any of these reasons for inaction.
This is somewhat off-topic, and I don’t want to encourage excessive splitting of results over several papers, but I do want to give a small word of caution. There is a danger in combining all your results into one big paper. If you stick a result into your big paper which is only tangentially related, people may not realize that it is there, and may publish other papers duplicating your result.
I have come to conclusion that the right way to judge whether you should split your paper is that every paper should have one story to tell (it may be a long and complicated story, or a short and simple one), and anything that doesn’t fit into this story should be put into a different paper.
Someone helpfully sent me pdfs of the GRG articles, marked up to identify the plagiarized portions. See the update at the end of this posting for links to these.
I think the university involved in this scandal is totally rotten. The president of the university apparently tried to cover up the scandal and had it not leaked to the media, they would not do anything about it. Another issue I am concerned about is the responsibility of the supervisors of the Ph.D. students expelled for plagiarism. Where were these folks while the students were publishing plagiarised papers on their own? I think the guilt lies way higher in the rank in this case and the university tried to cover up the issue not to punish the supervisors and as far as I am aware, no action was taken against them in any way.
The next time anyone tries to tell me how important traditional journals are for quality control, and how conferences, the arXiv, etc. just don’t provide high enough standards, I will show them the following “cut & paste is OK with us!” letter from JHEP, which I think deserves a special place in academic history.
Dear Professor Akbulut,
We have looked to your letter regarding plagiarism.
First of all, in matters of plagiarism it is the injured party that must take action. We are not involved in administering justice.
The authors of the paper in question seem to cite the articles from which they copy sentences or paragraphs. This weakens the case.
It may be that the paper should not have been published and that thereferee erred. Unfortunately weak papers get sometimes published.This may happen when the work is in a rather minor subject or not in anactive and competitive field. If the results are oustanding or contradict well known physics this would not happen.
Therefore we fear that we cannot help you in this case.
Sincerely yours,
JHEP Journal
I used to teach upper-div physics at one of the universities in Oregon. After grading mid-terms using an `instructor’s-only’ solution manual provided by the publisher of the text, I noticed one student repeatedly gave answers which were verbatim copies of the ones in my manual.
I showed them to the chair, undeniable evidence of cheating by a physics senior, only to be told in so many words “Tut, tut, my boy: work around it “!!!
So these journals not being overly perturbed by plagiarism is ALL in the spirit of the age: Go with the flow…Don’t rock the boat….
I chose to rock the boat and failed the student.
As a journal editor I find the letter quoted by Scott Aaronson disturbing in several respects. Authorship and attribution are part of the reward and recognition system of science. Peer-reviewed journals play a prominent role in that system. While it may be the that as editors we are not “instruments of justice,” neither should we be the instruments of fraud and misrepresentation. Publication of plagiarized work does harm the journal by lowering its academic and ethical credibility; a journal and its editors does have standing as an injured party.
Rejection of papers that misrepresent authorship and even informing an authors institution of academic misconduct (after appropriate procedures) is part of our ethical responsibility. The inclusion of a citation is not exculpatory when a author represents substantial potions of another’s work as his own. That applies whether one plagiarizes others or himself (yes functor, self-plagiarism is recognized as a form of academic misconduct; it may also be a civil tort of copyright infringement). This past year I have rejected 4 papers for self-plagiarism.
I would be surprised if the large majority of editors of scientific journals did not have similar views. To be sure instances of misconduct do get b; hopefully with better plagiarism detection software on the horizon the number of such cases will diminish. Even so, good refereeing is our best line of defense.
Jimbo, wb,
The practice of “not rocking the boat” and sweeping embarrassing incidents under the table is part of the “don’t let’s wash our dirty linen in public” philosophy that permeates most professions. Physics, to be fair, is probably a lot less culpable in this area than, for example, the medical profession, where lawsuits often fail when professional colleagues close ranks to protect “their own”.
That journals would be reticent to act, I can understand, if not condone. But that JHEP would pen such a response as quoted above is astonishing. A massive gaffe. That journal better undertake some damage limitation, and quickly.
Peter,
You’re right that an article which plagiarized (necessarily work of other authors) should be rescinded, and that in that sense the GRG editorial is inadequate. Whether this is due to naivete or to pressure from the publisher or to some other motivation, I cannot say.
There are journals that behave properly; there are many more that do not. The best ones are the ones for which the profit motive is not important, often those published by university departments or professional societies. The problematics journals are, no surprise, mostly published by Elsevier, Springer, etc.
What’s going on with the commercial journals is simple. There is a cadre of professors who want to publish papers, and want to do so in journals with big impact factors; these journals provide that service, though at a handsome price (Which is, however, born by the researcher’s institution’s library’s budget rather than the researcher). Some of these unscrupulous individuals wind up as editors, and their editorial motivations are anything but publishing the best papers, or seeing to it that articles are refereed in a timely or fair fashion. There’s certainly no interest in bringing attention upon themselves of publishers of nonsense with low standards.
Look at the quality of the editing of the English in many of these publications. It is lousy. Any journal publishing in English but unable or unwilling to make the effort to guarantee that the articles which it publishes are written in correct English is likely cutting other corners too. This happens much less in certain journals than in others – it is one indicator of the seriousness of the editors.
Outright plagiarism and dishonest copying of ideas are huge problems. Repetitive publication of the same result is a problem too, but of a different nature; it is stupid, perhaps unethical, but not as fundamentally dishonest. Decent blokes can be deceived into thinking it a reasonable way to go. There are a lot of folks not having success with their research programs, but with a lot invested in trying to, with families to feed and mortgages or rent to pay, who lack the courage to suffer the consequences of their own limitations and engage in questionable but easily rationalized practices; this is humanity. Plagiarists and thieves are a different matter.
The upper tier of US universities has less of this nonsense, but in countries with undeveloped or developing research communities it is more common – there are not enough researchers of quality, and their voices are perhaps not even heard, so that even those within the community of serious (in the professional sense) researchers are able to tell the good from the bad. Also if one learns in an environment in which dishonest practices are the norm, then one learns that those practices are ok. This is why often famous researcher A develops a legion of students who behave like a mafia; they learned more than physics from A.
..Tansu, characterizing the whole turkish scientific community as modeling their scientific research on mafia-like structures and habits seems possibly, not that I had deeper knowledges of the context involved, a bit too stereotype. Even if it matches your personal experience, it may suffice to feed well-known stereotypes in the ‘western’ world, regarding the scientific culture of islamic countries. As it seems the above linked two pdf-files mostly display self-plagiarism (the blue marks?), and some copy&paste of more or less peripheral blah-blah, which could clearly be induced by insufficient mastery of the english language (one could wonder how many US-scientists spoke turkish, by the way). So while multiplying its own contributions by excessive self-plagiarism is unethical, at least from my point of view copying motivating verbal formulations in the exact sciences to compensate for insufficient english is not, one should possibly focus on the actual content of the papers in question, it is not impossible to plagiarize the essential content of whole books without even using one sentence of the plagiarized work explicitly. It is obvious that, again , the ‘little fishes’ get all the attention while the ‘big ones’ remain in shades.
DB,
I think JHEP already took as much damage control as they intend, by withdrawing the paper. The e-mail quoted was sent long before this became public, at which point they seem to have realized that their response to the e-mail was untenable.
I suspect that if one contacts any of the rest of the journals involved to ask why they don’t withdraw the papers, one will get a similar response. Unless you’re a member of the press or one of the plagiarized authors threatening legal action on copyright issues, in which case they might actually do something.
The GRG response seems more disturbing to me. Ultimately JHEP did the right thing, and changed their initial untenable position, which was in a private e-mail. The GRG editors considered these papers carefully, then despite the obvious serious problems with them, publicly defended them and refused to withdraw them.
a.k.,
“mostly display self-plagiarism”
I can’t believe we’re looking at the same files. In the second one, none of the plagiarism is self-plagiarism, and most of the first half of the paper is nothing but directly plagiarized material from other papers. In the first, there is some self-plagiarism, but also lots of plagiarism from other sources. The blue section (which is the main content of the paper) is described not as word for word plagiarism, but as following closely two other papers which are not cited, one of which is by the same authors, one of which isn’t.
..I wasn’t aware of the exact meaning of the ‘blue marks’, if the second paper on which this section relies, is not at least associated to the working environment of the authors, it is without any doubt severe (non-self) plagiarism. In the second paper, the first half seemed more or less what I characterized as blah-blah, as long as at least most of this plagiarized material is taken from cited papers, this could still be a ‘boundary case’. In any case, the affair also puts a light on isolated scientific contexts, the best way to avoid plagiarism is not by electronic tools but by global non-formal communication and personal networks, which could detect those attempts very early. In this sense, one should avoid putting the stigma of mafia-like structures and dishonesty on whole scientific communities and cultures, this does not help to avoid but to enforce exactly those structures.
Self plagiarism is persistent and wide spread. I know of a number of well known names in numerous fields who recycle essentially the same paper. The level of self plagiarism varies tremendously from mild variants of previous papers that have been mostly re-written to almost perfect copies. Stopping this is almost hopeless for several reasons. First, submission, acceptance, publication cycles are long & it is easy for multiple copies to be in the system unknown. Second, who is going to cross read all this stuff? Third where do we draw the line here? Fourth, is recycling your own writings & ideas a problem? Lastly in all its manifestations, this problem is huge for all journals.
I see even severe self-plagiarism as a judgmental call by the journal’s referees and editors. It depends on their goals, policies or standards. Low tolerance standards for self-plagiarism may be a negative reflection on the journal but not on the author’s morals. I do not see a big issue here.
Plagiarism of another author is another issue though. A clear line has been crossed.
It should be noted that legal exclusive rights such as copywrite trademark & patent all clearly state that no one has exclusive rights to either ideas or words. People just are not allowed to duplicate exactly how you use the ideas or words.
I see self-plagiarism as simply scientists “playing the career game”. To reduce its incidence job and promotion awarding bodies have to look beyond a paper count. I think this is easier said than done. The culture of “publish (a lot) or die” is everywhere and I can’t see any serious moves to tackle it.
As an experimentalist, I’m lucky that I’ve never had the pressure to self-plagiarise and produce X papers a year. On a large collaboration it is normal and accepted that an active physicist can only be expected to perform the data analysis for and publish around 1 paper per year. My other papers such as conference and summer school proceedings will undoubtedly look very similar to each other (though never word-for-word identical). However, it would surprise me if anyone could legitimately classify them as examples of unethical self-plagiarism – the remits of each invited paper were often the same.
There is one more dimension to this plagiarism scandal: withdrawing papers published online, or even altering online content smacks of rewriting history and compromising the historical record. (I mean JHEP. As is well known the print edition of JHEP ceased with 2000 and from 2001 and on JHEP is published online only.) Once you start on this slippery road with the best of intentions you can end up in hell.
Scott, isn’t your comment along the lines of “let’s not bother doing something if it cannot be done perfectly”? just imagine a world when you don’t have to pass any bar before your publication appears on your CV or in a popular science magazine (OK, bad example, but hopefully you get my point).
there was a high-profile scandal like this, in organic chemistry last year.
http://pubs.acs.org/isubscribe/journals/cen/85/i11/html/8511sci1.html
For some years professor Armando Cordova from university of Lund has been developing a reputation of a odious young man who acquires information about unpublished research from competing groups and repackages them as his own “contribution”. Eventually he got caught in the act when he tried to rip-off results from another group – and he clearly misunderstood the results that he was trying to pass as his own!
The investigation done by his university concluded that yes, Cordova is the unscrupulous bastard that everybody says he is – and the outcome was to give him a warning not to do it againg in the future – and provide him ethical counceling. The department head would also oversee Cordovas future publication activity. (Meanwhile Cordova tried to re-publish yet another ethically-challenged paper.)
I realized Cordova is at Stockholm Uni, not Lund. Sorry for the error.
To Scott Aaronson and anyone else who thinks that journals are ineffective for quality control I issue the following challenge:
(a) Write a deliberately trivial/uninteresting paper on an issue of central importance for some subcommunity of serious researchers (the specific choice of topic is yours) and see if you can get it accepted for publication in a major journal, e.g. one of the Physical Review journals. (Of course, if you actually manage to do this you should tell the editors afterwards that it was just an experiment to test publication standards and that they shouldn’t publish the paper.)
or
(b) Same as (a), except the issue addressed in the paper may be anything you like (not necessarily of interest to any serious researchers) and the target journal you have to get it accepted in is Physical Review Letters.
And no, it isn’t enough just to find some previously published paper that you consider to be an example of (a) or (b). If the quality control is really as bad as you claim it is then you should easily be able to manufacture such a paper yourself!
Imo the current plagiarism spectacles don’t disprove the value of journals; they simply show that when evaluating someones publication record one needs to consider not just the names of the journals and number of publications but also the topics and issues addressed in the papers (whether or not serious people care about them, how centrally important they are in the general scheme of things..). (Of course, for PRL publications the latter consideration is superfluous 😉 ) I’ve already elaborated on the reasons for this in tedious detail elsewhere so will try to resist doing it again here.
The Journal Astrophysics and Space Science has just announced that they are retracting four of the plagiarized papers, noting that this was done “at the request of the President” of the METU. See the update to the posting.
It’s interesting to see here that people from fields outside theoretical physics (e.g. Scott and Peter Shor) are appalled by this story, those from within (e.g. Moshe and amused) seem mainly interested in defending the current refereeing system.
I don’t think there’s any need to try to submit fraudulent papers to a range of theoretical physics journals to see what happens. The experiment has been done now by Salti et. al, and the results are known. The question is what should be done about the fact that the refereeing system in this field is seriously broken.
To keep amused happy, I’ll stipulate that this doesn’t apply to PRL, and advise our librarians that, when they cancel all subscriptions to theoretical physics journals on the grounds that they’re no longer worth the paper they’re not even printed on anymore, PRL should be spared.
I am a particle theorist who is certainly outraged by this plagiarism, and the reticence of some journals to withdraw the offending papers is both surprising and worrying. However, I can see a reason why theoretical physicists in particular might be less phased by this than researchers in other fields: because new research results in theoretical physics are not communicated via journals.
Do you think it would have been easy to verify that all the material was original, is there software available for this that does this well?
David,
The arXiv has some software for this, I don’t know if it’s publicly available. Once you know that there is plagiarism involved, it’s fairly easy to find the sources just by picking phrases from the article and using Google. One can argue about whether the referee should have realized that the paper was plagiarized, but once the plagiarism became known, it’s remarkable that the journals involved were unwilling to immediately retract the papers (and many have still not been retracted).
In talking to people on the subject of journals and peer review, the main difference I see is where people are from, and that includes myself. When I was a postdoc, benefiting from a constant stream of inside information, I too was tempted to think that most people know which papers are correct and worthwhile without having to rely on other people’s judgment. I now realize this is not true, and view peer review as service to the community at large. By and large I still think it is done well in my community.
So, I am not convinced that there is a wide-spread problem, certainly not in the hep-th community I am part of. In any event, I am confused about the logic of the argument of those outraged by what they see as lack of quality control: surely getting rid of journals and peer review altogether is only going to make thing worse…
Milkshake, that (the Cordova scandal at Stockholm U) is interesting. If you enjoy this kind of digging, see if you can find the (pretty obvious) link to JHEP and JCAP…
Moshe,
I don’t see anyone here arguing for the elimination of all peer review. Rather, there is a very real question about whether many journals actually are doing a legitimate job of it. If they aren’t willing or able to put the effort necessary into it, they shouldn’t be selling a product that claims to be something it isn’t. The attitude towards plagiarism demonstrated by the GRG editors and by whoever at JHEP responded to the initial plagiarism complaint should be disturbing for this reason, even though they don’t seem to bother you.
Several people have pointed out to me that the problem of journals publishing plagiarized or incompetent articles is one that people at elite institutions in the West are able to ignore because they can just ignore those articles, and the people making hiring decisions are able to recognize that they aren’t serious scientific work. For people at less elite institutions, where hiring decisions are often in the hands of authorities able to only count publications and journal impact factors, the problem caused by the behavior of these journals is very serious. In particular in this case I think it seems that what happened was that this sort of plagiarism worked well for one or more people in terms of advancing their career, encouraging others to join in and do the same thing. Those that weren’t willing to work this way had to compete for jobs with those who were. For such people, seeing the refusal of journal editors to do anything about this is profoundly discouraging.
Peter,
Scott has a couple of papers in Phys.Rev.A., looks like he might be a semi-physicist.. Same goes for Peter Shor, who seems to have quite a fondness for PRL..
My impression (and someone please correct me if I’m wrong) is that none of the papers in this scandal can be characterized as addressing issues of central importance to some subcommunity of serious researchers. So Salti and his fellow plagiarists haven’t already done the experiment I challenged Scott to perform. For this class of papers my impression is that refereeing standards are generally pretty good, and I expect the odds of Scott being able to complete the challenge are very small. When he fails I hope he will reconsider his view that journals are ineffective for quality control.
Besides that, I see this scandal as being the extreme edge of the well-known phenomenon that lots of rubbish is being published in physics journals. The minor ones are full of it, and some of it seeps through to the major ones. That is of course a serious problem for people working at institutions where they are evaluated simply by counting number of publications. The solution: stop doing it! That might be easier said than done though, since I imagine that at many such institutions the people with power got there through publishing rubbish in junk journals and won’t be keen to change a system that worked for them. But simply getting rid of journals isn’t going to solve the problem…
At any rate, thanks for sparing PRL 😉
Peter, it is a small step between “the journal system is broken” (a statement I disagree with) to “let’s not bother with journals, everyone reads the ArXiv anyhow”. But maybe this is not what Scott meant…in any event I think we both would like to see a stronger peer review system, any ideas on how to produce one? as you probably know JHEP decided to give a small financial reward to their referees, seems like a good start.
Moshe,
I’m afraid I see things in a different order than you. It seems to me that “let’s not bother with journals, everyone reads the arXiv anyhow” is already the de facto situation. Jacques Distler doesn’t seem to have submitted some of his papers in recent years to journals and from what I can tell he’s not the only one. Most particle theory papers are sent to JHEP to be published, and Mustafa Salti did a good job of showing what their standards are. The letter from JHEP responding to the fact that they had published a plagiarized paper is quite remarkable. If you are looking at a paper on the arXiv, you could check to see if it had been refereed and published in JHEP, but would that really tell you much? I suspect that people rarely actually do this, partly because of what “amused” describes as “lots of rubbish” being published in physics journals, including some in the major ones like JHEP.
Given that that’s where we are, I think the problem is that people won’t go to the next step and acknowledge that there even is a problem with the journals, partly since fewer people are paying any attention to them. The obvious answer to low refereeing standards is to raise them. That’s could be done, but it’s hard to get people to do if supposedly there is no problem.
While the problem of refereeing in this day and age is a complicated one, the question of papers involving significant amounts of plagiarism like the GRG ones isn’t. They have no place in scientific journals, and if editors don’t see this, it’s hard to believe they’re going to ever address even tougher problems.
Yeah, there is a small minority of people who think journal publication is redundant, and I strongly disagree with them. Not sure about Jacques, never discussed the issue with him, but my impression from his comment was that Scott is leaning that way, and that as usual he may have interesting reasons for his position. Anyhow, looks like it is too late for that, maybe next time…
Peter,
The point you make about what happens outside the elite institutions needs to be reiterated (it may happen inside too – I don’t know, I’m not there). In a country such as Spain the use of paper counts and impact factor is institutionalized to the extent that often by law the ranking of candidates for positions or contracts is based on a score in which an impact factor weighted paper count plays a prominent part. This encourages self-replicating publications and has a corrosive effect on notions of what is a good researcher. The goal of every Spanish jefe is to be one of the ISI most-cited researchers, and this goal can be accomplished most easily by the following scheme: jefe and coauthors A and B and former students X, Y, Z, etc… write repetitive, silly papers of short length which are published in mediocre Springer and Elsevier journals with unjustifiably large impact factors and which cite other articles by the same authors (preferably a disjoint subset of said authors, because this avoids the self-citation problem); since no one else but the others in the circle is terribly interested in the properties of the twelve times iterated tangent bundle, it is almost guaranteed that the referee will be a sympathetic collaborator. After some years, researcher A has 100 papers, each with 3 or 4 citations, and he can present himself as a researcher of quality to bureaucrats and persons working in other areas, so he is easily funded and has no trouble ascending to the position of chair of the institute he founds, and in which he employs B, X, Y, Z (and maybe some students of a friend of Z in France).
Usually A does not exactly copy his own papers, but when A is operating in some small rural place rather than Madrid or Paris, he might go a bit further. Secure in the knowledge that absolutely no one is reading his papers, and that absolutely no one will read his papers (he assigns his students to read the papers of his thesis advisor in France, a very famous man, so that they will know how good he is, being the student of a famous man), he begins to copy them wholesale and send them for publication in obscure journals like the Former French Colony Low Energy Physics Journal (fictitious example which somehow has on the editorial board his friend B), and he rationalizes this behavior with the good feelings that come from having lent the prestige of his highly cited first world (well almost) name to a journal published in a developing country.
I suppose most of the readers of this blog can continue the story.
What creates the dynamics whereby this sort of behavior is rewarded and even encouraged is the false profesionalization of scientific inquiry and the weird conflation of frequency of publication with quality.
In all cases where I’ve seen “broken systems” it was because people were responding rationally to incentive systems, but the incentive systems were perverse. (Also, most people have the attitude, I didn’t make this system, I just live with it.)
Every scientist ought to occasionally step back and review what it is that he/she really wants to accomplish and whether the “system” is helping them get there.
One of the plagiarized papers (according to arXiv) is gr-qc/0011027. It was published in PRD and according to Spires it has 1 citation by other authors, 2 in total.
Its clone, gr-qc/0505079, published in Astrophys.Space Sci., has 37 citations, at least 15 of which are by non-Turkish authors.
It’s interesting how the clone was so much more successful attracting citations than the original.
Peter, David,
Isn’t there some online service that many universities subscribe to, which checks for plagiarism in students’ term papers against stuff that is online and/or from “essay mills” (like schoolsucks.com and others)?
Maybe arxiv has a program similar to this. I would imagine the algorithms would be very similar, in systematically comparing long strings of words + characters against one another.
Article about arxiv plagiarism detector program.
http://ptonline.aip.org/journals/doc/PHTOAD-ft/vol_60/iss_3/30_1.shtml
I’m surprised by arxiv’s stance on plagarism in papers. It may take 10^500 string theory papers to investigate the landscape fully, and most will achieve similar results (failures). In this case, plagarism (repetition) of the text in each paper would seem appropriate, whereas making up different wording for each paper’s introductory paragraphs and conclusions is irrelevant because nobody is interested in reading 10^500 variations of the same basic message. The difference between each paper will be the numerical moduli values guessed for the particular Calabi-Yau being ruled out. Why bother making up different wording for each paper? Plagarism is sensible for such situations.
Preprint on the algorithm used for arxiv’s plagiarism detector.
http://arxiv.org/abs/cs.DB/0702012
“Plagiarism Detection in arXiv”
by Daria Sorokina, Johannes Gehrke, Simeon Warner, Paul Ginsparg
Reynold.
It may be that once a paper starts picking up citations (in this case by the Turkish physicists) it is somehow seen as having an impact. Other authors may well then be more willing to cite it as work which has been taken seriously by the community.
I found this comment at the Ars Mathematica blog, made by prof. Irfan Acikgoz.
http://www.arsmathematica.net/archives/2007/08/22/plagiarism-at-arxiv/#comment-55199
does anybody have informations about this letter was or wasn’t sent to arXiv and if the arXiv admins gave any answer?
thanks
a scientist,
I assume that letter was sent, but don’t know what if any response they got. There’s nothing in it likely to have changed the minds of the people at the arXiv who decided to withdraw the paper. Three of the signatories of the letter are the authors of the JHEP paper that I provided a link to marked up to show the plagiarism. That paper is just about completely plagiarized, and their is no way to deny this.
Looking at the annotated pdfs, I’m struck by how they’ve cut together a mishmash of many sources. To someone more knowledgeable in the subject area, do these papers read as coherent arguments or do they seemingly flit from topic to topic in the same way that they cut from one plagiarized paper to another? If the latter, how did they ever get accepted?
Reynold wrote that
“It’s interesting how the clone was so much more successful attracting citations than the original.”
Dear Reynold,
another interesting search would be to look who are cited in these plagiarized papers and also who cites these papers.
It may be possible that this mafia-like organization is more deep then what the community thinks.
I am studying the subject of plagiarism as a science and communication sociologist.
The turkish plagiarsim scandal raised some complex questions: no doubt about plagiarsim, if not detected, can raise an “author” visibility and help his carrer. But it’s true that if it is detected (and we see how it’s easy to do it today), the sanctions are very strong.
So, why a researcher should run this mortal risk? this is the question I ask myself when I try to consider all the variables which affect the case. The first answer that I give to this question is that the search for visibility is a weak reason, if compared to the eventual destiny of the plagiarist. Then, I try to think for a moment how hard to treat are the concepts of originality and relevancy in a scientific communication context.
Acting out of the theoretical physics context it’s a bit hard for me to analyze some specific aspects of your scientific communication.
That’s why I’d like to ask you (Peter) and all the physicists who will reply, one question, that you may find silly:
I don’t know exactly how many paper a year are published in theoretical phys, but from my data I can suppose that the number is expressed in terms of thousands.
My question, therefore, is:
is it reasonably and scientifically possible to produce thousands of papers every year in your field, which content being ORIGINAL and also RELEVANT for the progress of knowledge or are relevancy and originality two complex concepts, hardly translated in communication practices?
This is a very important point for me to understand how much the “publish or perish” pressure does affect the quantity and the quality of scientific literature and how much originality and relevancy are true features of this literature
Thanks in advance