38

In my field of research, it is common to include a supervisor as a co-author of a student's PhD research related papers. A supervisor mainly contributes by helping to improve the writing of a paper, such as emphasising its contribution or clarifying the content, etc. Which means that a supervisor does not contribute anything directly related to the research carried out in a paper, for instance, a supervisor may not even understand a method used in a paper.

Based on all the answers of this question, it is clear this custom is not applied to all fields, and I am not trying to argue if it is correct to do so or not.

I want to know how a supervisor explains to a new PhD student that he/she should include a supervisor's name in his/her papers. What happens if a student refuses to follow a custom?

PS: when I started my PhD, I already knew about this custom, so my supervisor did not need to explain it to me. But how about those that don't know or know differently?

  • 10
    I strongly suggest that you rephrase your question to ask how to convince somebody to have a supervisor help writing the paper. This is something entirely different and much more ethical than just adding a supervisor to the lists of authors. – Wrzlprmft Sep 23 '15 at 19:55
  • 72
    This is a very loaded question. If it is so difficult to explain this concept to your student, maybe it's because the custom is wrong? If I may propose a (very) over the top analogy, it is a bit like asking back in the era of slavery what would be the proper way to explain to your newly arrived slave that slavery is the custom of the land and they are supposed to be your property and follow your orders; and "what if the slave refuses to follow the custom?" The answer is: don't. If explaining the custom causes such discomfort, that to me is a strong sign that you shouldn't practice it. – Dan Romik Sep 23 '15 at 21:45
  • 10
    @DanRomik While I agree that there are cases when co-authorship of supervisor is not earned, in my opinion the psychology of such prep-talk is a very poor judgement about the situation. There are people who are very confident about their importance and power, others are less comfortable to discuss them in great details and commanding manner - and these can poorly correlate with their contribution. Also, it is a totally plausible situation that a student has no or wrong idea about what are the requirements for authorship, yet his/her ego prevents a light, graceful discussion on the topic. – Greg Sep 23 '15 at 23:49
  • 31
    a supervisor may not even understand a method used in a paper I don't see how this person could claim authorship then. And I'm in a field where supervisors co-author student's papers because it's impossible for students to do the research on their own. – Cape Code Sep 24 '15 at 8:33
  • 19
    @CapeCode If you assume that everyone listed as authors must be able to understand all methods used in the paper, you probably haven't done a collaborative natural sciences paper. – March Ho Sep 25 '15 at 2:40

12 Answers 12

67

If the student truly is writing a paper independently of the advisor, then the advisor shouldn't be a coauthor and it would be inappropriate for the advisor to demand this.

In the more usual situation in which the advisor provides the the dissertation topic and works directly with the student in developing the dissertation (holding regular meetings, reviewing drafts of the thesis, meeting with the dissertation committee) and then papers based on the dissertation are published, the participation of the advisor would normally merit coauthorship. In these situations the authorship of papers should be settled before the research work begins.

I make it a point to explain this to students before starting work with them on their theses and dissertations. I simply wouldn't agree to supervise a student who would not agree to sharing authorship in this way.

  • "I simply wouldn't agree to supervise a student who would not agree to sharing authorship in this way." Given that you and Paul Garrett are both mathematicians, I find the difference between your policy and his to be rather striking. (It's not a criticism, and my own policy definitely lies in the open interval between them. For instance, the paper that comprised my last student's PhD thesis is jointly authored with me. It is one of my strongest papers and I certainly contributed in an essential way, so ducking out of the authorship would go both against my ethics and my self-interest.) – Pete L. Clark Sep 25 '15 at 1:21
  • 5
    @PeteL.Clark a glance at our profiles and publications will make the difference clear. I'm an applied computational mathematician who does interdisciplinary work with hydrologists and geophysicists. I work at an institution where the conventions of the physical sciences completely dominate the institutional culture and my performance evaluation for research is based on how successful I am compared with my colleagues in the physical sciences and engineering. Paul Garrett is on the other end of the spectrum in pure mathematics. – Brian Borchers Sep 25 '15 at 2:03
  • I'd add that I think Paul's answer is good for "mathematics", if you define mathematics so as to exclude applied and computational mathematics, statistics, computer science, etc. as "not mathematics." I'd disagree with the answer applying to the broader category of "the mathematical sciences." – Brian Borchers Sep 25 '15 at 2:23
  • 8
    In case it wasn't obvious from my answer- when I coauthor a paper with a graduate student I've always done enough work to justify coauthorship under the ICMJE rules. I wouldn't agree to be a coauthor if I hadn't made that level of contribution. However, I'm having a hard time understanding how any responsible advisor could be that uninvolved in a graduate advisee's thesis or dissertation that their contribution wouldn't merit coauthorship. – Brian Borchers Sep 25 '15 at 4:25
  • 4
    @Brian: I’m a topologist. I have no joint papers with my adviser, who in fact probably never actually internalized the definition of the property that I was studying for my dissertation – and who was a perfectly wonderful adviser for me. When her students worked on something on which she happened also to be working, there was likely to be a joint paper or two; when they worked on something else, there wasn’t. No big deal either way. – Brian M. Scott Sep 27 '15 at 18:12
36

To an outsider, this sounds very much like gift authorship. It is likely that a student could come back with examples where what you are suggesting is considered academic misconduct. The only way I think you could convince the student is if you can provide a written policy from a relevant journal or professional body.

  • 1
    I would be very interested in an example of either, as I have so far not seen any official statement from someone to the effect that they do not wish to follow the Vancouver protocol (whereas I have on this site seen that it is very common in some fields to have some very different customs of what merits coauthorship than what is stated there). – Tobias Kildetoft Sep 24 '15 at 7:22
  • 10
    @TobiasKildetoft: "any official statement from someone to the effect that they do not wish to follow the Vancouver protocol" - that might be connected to the circumstance (issue?) that those who follow a different set of somewhat agreed-upon conventions simply will not mention the Vancouver protocol. Throughout my whole time in academia, I have never heard about the Vancouver protocol, neither from conference organizers, nor from other researchers - the only way I am familiar with it is because "some people on the internet (read: users on Academia SE) mentioned it". – O. R. Mapper Sep 24 '15 at 8:12
  • 1
    @O.R.Mapper As I said, "to the effect", by which I mean proposing a convention that disagrees with the Vancouver protocol, not necessarily mentioning that protocol directly. – Tobias Kildetoft Sep 24 '15 at 8:36
  • 1
    @TobiasKildetoft: This question mentions such a journal. Also, I recently learnt about a journal having comparable guidelines (they request a professor as a corresponding author). – Wrzlprmft Sep 24 '15 at 8:39
  • 3
    This. If the supervisor should be an author on the paper, it should be self-evident based on standard authorship guidelines. – Fomite Sep 24 '15 at 20:04
33

Instead of insisting to be named "just because you are the supervisor", I suggest that as the supervisor, it is your role to be sufficiently involved in the research of your students that you qualify for authorship "the conventional way" (for your contribution to the paper).

In other words - do the job of supervisor properly, and there will be no argument over your co-authorship. And if you really do the job properly, your student will eventually be able to write papers on which you will not be listed as author - and you should be proud when that moment comes instead of mourning a lost citation.

18

This is pretty common in computer science. You may be underestimating your contribution as an advisor. I assume it goes further than just grammatical editing of the paper. In my experience, my advisor's contributions were:

  • Setting the overall direction of my research
  • Providing (access to) equipment used in the research
  • Finding relevant related work based on his experience
  • Having regular meetings during the research to discuss progress, what to do next, how to tackle problems and so on.
  • Discussing how best to present the results and helping decide where to publish
  • Providing basic assistance with actually writing the paper including writing some introductory text, reviewing, proofreading, etc.

Based on my interaction with other PhD students, this was the common pattern, and it was very rare for the advisor not to be listed as the last author on the paper.

  • 5
    Good point, but I am curious how would your answer address the "a supervisor may not even understand a method used in a paper" at the end of the first paragraph of the question......Also IMO, one has to be carefull as a supervisor might also overestimate his contribution, and as a supervisor I would rather err by underestimating... – Nick S Sep 24 '15 at 17:34
  • 2
    @NickS: well, if the paper uses multiple methods at different points then I think it's potentially quite a tricky point whether every co-author needs to fully understand every step of every section in order to claim co-authorship. Obviously it'd be better if they did, before putting their name to it, but I wonder in practice how many solid collaborative papers would actually have zero authors if this was strictly followed ;-) – Steve Jessop Sep 24 '15 at 18:51
  • @SteveJessop I agree, with the comment in general. But this answer, if I understand it right, is saying that in many situations even if the student is doing the research, it is actually the supervisor which is subtly guiding the student on the right path, without the student knowing it. While this is actually often the case, I highly doubt it can be done in a very subtle way, so subtle that the student doesn't realizes he is guided, without the supervisor really understanding what happens there. – Nick S Sep 24 '15 at 18:59
  • 2
    @Eric. Despite your points, this is not the way to go about this. Credentialed academics will ALREADY know that a paper is under the auspices of a given department, university, and (therefore) the present head. How far would you argue this? Did the President not also influence the direction of the department? Why not just include the President and all past Presidents of the United States? – TheDoctor Sep 25 '15 at 2:34
  • 1
    @TheDoctor Like it or not, this is the way it works in computer science. Advisors are the lowest level managers of academia and as such they do contribute but in ways that the student does not always recognize or appreciate. – Eric Sep 25 '15 at 11:19
17

I want to know how a supervisor explains to a new PhD student that he/she should include a supervisor's name in his/her papers. What happens if a student refuses to follow a custom?

You seem to be trying to push them into putting you as a coauthor, just because [of a custom], which may make the student more likely to think you are trying to take advantage of them - which is how it would be viewed in other fields.

If this is as customary as you say, it might give them a bad image to peers or other people in the field when looking for help or advice.

If you feel that you deserve to be put as a co-author, other than because of the custom, why is that? That is what the student needs to hear if they are going to do it.

  • Possibility 1: You've had a great deal of input on the paper, enough to warrant co-authorship and the student doesn't acknowledge this. Point out the things in the paper that you've had a hand in changing or adding. Point out that if you were to ask them to remove anything added by you, they would not have a paper. Try referring them to another peer about it. If they still decide to ignore you, they clearly don't need any out-of-the-way input from you in the future and you should better indicate what you expect from the student when you first agree to be their supervisor.

  • Possibility 2: You've had very little, if any, impact on the paper. The student doesn't acknowledge the custom. If you would expect that it would be best, for the student, to follow the custom then explain why that is. It could be that future potential coauthors might not trust them or that it might be a start to a poor reputation. Looking from the student's, and even some outsiders, perspective it seems like you are taking advantage of them. They might not trust anything you say, especially as empty as "your reputation may suffer". You might have to refer them to ask about it to another person on the faculty. With more people agreeing, it becomes a little more trustworthy. If they still decide to ignore you, then at least you tried - they'll just have to live with the consequences.

  • +1 I think it is the only answer here that actually tries to answer the question. – silvado Sep 25 '15 at 11:19
9

I took the liberty to peruse your SE profile and history. You strike me as an honest researcher who is interested in an academic career, but is nonetheless frequently troubled by doubts and anxiety about your publication record, leading you to post here many questions about ethically questionable practices (in a few cases involving a hypothetical "researcher in your lab"). In addition to the current question, to which I've already posted an answer of sorts in the comments section, some examples I spotted at a glance are:

Several of your other questions (e.g., this one and this one) do not seem especially problematic from an ethical point of view but still point to signs of your wish to bolster your publication record and academic success through strategic means that are tangential to just working hard and being talented.

I must say this pattern is beginning to seem alarming. I will try to be as non-judgmental as I possibly can (among other reasons because I feel your presence on Academia SE and the discussions you raise are actually very valuable) when I say that I think your preoccupation with publications, and particularly your tendency to imagine "easy" and often unethical or borderline ethical solutions to the problem of acquiring a good publication record, is excessive, unproductive, and harmful to you and to your future. The simple truth is that no amount of scheming to add your name to yet another publication will earn you the respect and credibility that you crave as a genuine researcher who works hard and comes up with new and innovative ideas. Only actual hard work can get you that. All the energy and thought that you are devoting to publications, which are just the superficial outward representation of the work scientists do, is to me a sign that you lack passion about the science and the work itself. Without that passion, I am sorry to say you will never make it very far in academia. And without honesty and integrity you will never make it very far anywhere.

  • It seems that you are assuming bad faith in the OP based on an assumption that he intends to do the ethically questionable things he asked questions about. I wouldn't go so far. – March Ho Sep 25 '15 at 2:44
  • 2
    @MarchHo, I'm assuming what the evidence strongly suggests, which is that the OP frequently contemplates unethical behavior, and has an unhealthy preoccupation with the number of his/her publications. I know nothing of OP's intentions, but what I know is already quite enough to be concerned about. If you see someone posting many questions on an online forum about, say, cannibalism, do you not get concerned and try to dissuade them, even if the questions are all "hypothetical"? – Dan Romik Sep 25 '15 at 2:59
  • 2
    @DanRomik: I would upvote you if I were as certain as you seem to be in your characterization of the OP. But - I'm not. In fact I suspect most upvoters have not actually read through everything s/he has published. So... I don't know. Also, your conclusions are too far-fetched; unfortunately, quite a few people make it in Academia without honesty and integrity (some might argue that higher-ups in academic administration are typically people lacking in honesty and integrity...) Still, thank you for writing this. – einpoklum - reinstate Monica Sep 28 '15 at 8:20
9

As I have noted on other occasions, the style/convention in mathematics, at least in my part of that world, is that advisors rarely, if ever, have, or insist upon, or accept, co-authorship of any sort involving their students' theses and related publications. In recent years I have seen a few examples, but this already surprised me.

That is, in my experience, in the best situations, and barring a handful of truly exceptional students, of course the advisor had a very good idea of how things should go, knew the key background, knew the pitfalls, and most likely could have written the thing up in an afternoon if they had nothing better to do, but they did have better things to do.

It did take me some time to fully understand the extent to which this is true in my field... since I, too, had bought into the mythology that beginners can become experts within a few months or a year or two, thus "bearding" their advisors. Well, I don't think that really happens in the way novices, especially, seem to want to believe. That is, an experienced expert can "catch on" soooo quickly to new facts that they operationally nearly-instantly assimilate that information, and integrate it with previous. And, unlike the novice, the expert can often see implications far in the distance.

So, again, in the part of the math biz within my view, of course the advisor has made a significant contribution... and, of course, there's no point in making the student "acknowledge" this by giving up some publication credit. Indeed, it seems that if a thesis gets published with advisor as co-author in math, it suggests that the student really didn't do much work at all, and is being dinged for that.

I am fairly consistently amazed at the seemingly-popular notion that some PhDs are done "independently" of advisors. Sure, some advisors may be technically incompetent... but how does a novice do something worthwhile that experienced experts would have wanted to do, but tried and failed? Does one have to be "The Chosen One" to do a PhD? Or are some fields so shallow that complete inexperience, or two months' experience, really makes an expert? Or is that merely a popular myth here, as well as elsewhere?

And, back to the original question, and as in other answers: don't try to coerce people to do things you cannot happily defend. Further, do not pretend that beginners are or should be on the same playing field as seasoned experts, nor that the experts should always "be sure" to grab credit for every little thing their apprentices manage to do under their guidance.

  • 1
    So just to be clear, you're arguing that even though as an advisor you've contributed substantially enough to a student's paper to merit coauthorship under the Vancouver protocol (ICMJE guidelines), you don't wish to be listed as a coauthor because in your field it's the cultural norm for advisors to let advisees take full credit for the work? – Brian Borchers Sep 25 '15 at 14:03
  • 3
    @BrianBorchers, within some approximation, "yes". For that matter, it goes both ways: perhaps I'd not want to be entangled with a write-up that "doesn't suit me" for whatever stylistic reason. Perhaps a novice positively needs to write out details that I don't need to write out. Perhaps that could even be the point of the paper... Extreme asymmetry in many regards. – paul garrett Sep 25 '15 at 15:38
  • 3
    My PhD was done independently of my adviser in every mathematical sense. Her contributions were (1) frequently coming back from conferences with questions that people were asking and handing them out in case anyone took an interest in one – I did, and the work grew into a dissertation – and (2) taking care of the administrative duties of an adviser. Oh, and cheerleading a bit. She was not interested in that particular topic, and I was doing fine on my own, so it all worked out perfectly. I’d not trade her for anyone else. (And I proofread a monograph of hers, rather than the other way round.) – Brian M. Scott Sep 27 '15 at 18:21
  • 1
    @paul: Not harsh at all; rather the opposite, really, since her students were encouraged to work on what interested them. In Mike Starbird's case it was something on which she was also working, and they got a couple of joint papers out it; in mine, at the same time, it wasn't. And it was fine for both of us. – Brian M. Scott Sep 27 '15 at 19:26
  • 1
    @BrianBorchers: According to my reading of the ICMJE guidelines, contributions of the sort Paul describes are enough to merit the opportunity for coauthorship. In order to get coauthorship one should also be actively involved in writing the paper, for instance. It is a standard cliche in my circle that advisor contributions stop short of this point. – Pete L. Clark Sep 28 '15 at 16:27
6

My take is that convention is not enough. Explaining to your students the authorship standards for your field is essentially, but once that's done, whether or not a supervisor is an author should flow naturally from those standards. And as a supervisor, you should definitely be more involved in a paper than just helping improve the writing - that makes you a proofreader, not a PhD supervisor. Either you are underestimating your contribution, or something else is wrong.

As a single point of data, in my field, "Which means that a supervisor does not contribute anything directly related to a research carried in a paper, for instance, a supervisor may not even understand a method used in a paper" and still claiming to be an author would involve lying on several journal submission forms.

Of the work that came out of my PhD, my advisor was only present on 50% of the papers, by his own request.

  • "convention is not enough. Explaining (...) the authorship standards for your field" - this seems to assume the "authorship conventions" and "authorship standards" of a field are two different, independent things. – O. R. Mapper Sep 25 '15 at 7:35
  • @O.R.Mapper Standards and conventions are very often two different things. For example, while a supervisor may be included out of tradition, I'd give good odds that there's few journals whose formal, written authorship criteria says "Oh yeah, and your supervisor too, that's an auto-include, even if if they've got no clue what you actually did." – Fomite Sep 25 '15 at 10:32
  • Then I'd call that two standards possibly clashing with each other. – O. R. Mapper Sep 25 '15 at 10:39
  • Except that the very definition of convention is "a rule, method, or practice established by usage" rather than a formal standard. For example, you can say "You know, your advisor should really be on your paper..." all you want, but when push comes to shove, to publish in many of the journals I publish in, your advisor has to sign a written declaration that they took part in the analysis in question. – Fomite Sep 25 '15 at 10:48
  • Conferences I usually publish on do not prominently propose any such written standards to my knowledge - or if they do, I'd probably take it as a way for them to justify their high participation fees, i.e. a possibly suboptimal suggestion by a commercial entity. From talks to other researchers in my CS-subfield, I have figured that the attempt to formalize such a standard, or follow one to the letter, would rather be perceived as "getting overly obsessed with details" and "overdoing things", thereby distracting oneself from actually conducting any research or being productive in any way. ... – O. R. Mapper Sep 25 '15 at 11:07
5

Point to the Authorship Guidelines for the journal to which the paper is going to be submitted. If your contribution warrants authorship according to those guidelines then point this out to the student. If the guidelines do not indicate that you contribution warrants co-authoriship then you should not ask the student for co-authoriship.

1
Which means that a supervisor does not contribute anything directly related to a research carried in a paper, for instance, a supervisor may not even understand a method used in a paper.

When the result is incomprehensible to someone experienced in the field, either the paper is bad (and it is the duty of a supervisor to make the author amend it) or the supervisor does not even invest the amount of work expected from the typical recipient of the paper, let alone a coauthor.

I have little problem with supervisors tending to be in a relation of coauthorship. But if they are neither actually being coauthors nor supervisors, this is stretching it.

1

This is the wrong way to go about this. If there is no direct contribution, then you don't include the supervisor as author. The proper thing to do is establish a convention for proper mention -- much like you do when referencing the work of another. The business sector has established conventions for just these sort of things which can be used as a guideline (like when a secretary is dictated a letter vs. writes her own letter under a boss`s name).

For example, the author line could say "Clara Smith; advisor: Dr. Herbert Norig" or something to that effect.

In any case, you don't want to dilute the notion of authorship just so department heads can get free citation points.

[Edit: Strangely my answer is voted down even though there is a strong vote in support of the general ideas expressed here (where it was the first A to the Q) in the form of other answers and the form of a highly-voted comment under the question itself, suggesting agreement. A question perhaps for meta...]

  • 5
    This is simply not true in some fields. As also mentioned in the linked question, in neuroscience, say, you always mention the lab's PI as Last Author (papers commonly have 3-5 authors). The PI likely contributed not only by providing funding, but in exactly the way described in the question: discussing, and helping to edit and get the paper into a publishable shape. But in any case they will be Last Author. – gnometorule Sep 23 '15 at 20:04
  • 4
    @Strongbad: As I read the question, the advisor reads the paper, puts it in perspective ("its contribution (to the field)"), and edits it ensuring or increasing the likelihood of publication. In my eyes, that's quite a contribution. – gnometorule Sep 23 '15 at 20:26
  • 6
    "Clara Smith; advisor: Dr. Herbert Norig" - depending on the field, this clashes with styleguide requirements for the paper, and it isn't possible to input such information in a fixed-structure form for meta-information in the submission system, either. – O. R. Mapper Sep 23 '15 at 20:38
  • 3
    @TheDoctor: "As a doctor, you don't have to wait for someone else to give you approval" - I have experienced this very differently. Both doctoral candidates and postdocs are supposed to independently choose and manage their research, but at the same time, both are employees of a department paid from the department's funds and thus are not supposed to "wander off" into arbitrary directions without a basic approval of their respective superior. (The superior, of course, is also bound to some extent as each department is run by the university with certain (albeit vaguely-defined) goals in mind). – O. R. Mapper Sep 23 '15 at 21:19
  • 13
    (that seems to be the convention in applied CS) — So many [citation needed]s. I have never head of such a convention, and neither has my department head. – JeffE Sep 23 '15 at 21:39
0

I'd say it's pretty simple - but not quite what you seem to be expecting:

  1. People who contribute significantly to a paper must be listed as authors.
  2. Explain point (1.) to your student.
  3. If it's a paper in which he's doing the write-up, leave it up to him to decide who gets put on the author list.

What happens if he doesn't name you? Then, well, he's either a person of questionable character, or he's obtuse, or maybe you haven't actually contributed enough.

protected by ff524 Sep 27 '15 at 5:33

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

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