A scientist's good name.

Ms. PhD at YoungFemaleScientist recently blogged about the joys of dealing with coauthors on a collaborative project. After doing her experiment for the project, she waited patiently for a look at the manuscript that resulted from the collaboration. And when nothing arrived (after a more than decent interval) she asked how it was going. The response was essentially: We're sending the manuscript out next week. Oh, you want to look at it before it goes out? We'll get it to you the day before it goes out to the journal.

Uh, no.

And, since Ms. PhD hasn't taken leave of her senses (or forgotten her ethics training), that was pretty much her response. It is a shabby thing to do to a coauthor to cut her out of the manuscript preparation process, especially when she is willing to play an active role in making sure the manuscript (both in the details of her contribution, and overall) is in good shape before it goes out. And, there is something not quite right about a coauthor who wants to keep a collaborator out of this part of the loop. Maybe there are sensible reasons motivating it, but if none are offered, it's hard to be confident that that's what's going on.

Ms. PhD's assessment of the situation, especially from the point of view of a more junior collaborator, is so on the money that I need to quote it:

My gut instinct is that if it's really bad, I wouldn't even want my data in there, but I'm not even sure I can ask them to take my data out of it, or if I should just ask for an acknowledgment, or what. I'm only going to have one day- maybe less- to decide. And I don't imagine they would put in any edits I might want to suggest, since they're not going to have time.

I want to be especially careful because someone we know (distantly) is having a really hard time getting her grants, after having to retract some collaborative papers she published with a coauthor who later got caught for fabricating results. To me, that's really a nightmare scenario, especially in this day and age where you can't possibly know for sure, since you're almost always collaborating across (sometimes way across) disciplines.

At some point you really have to wonder if it's fair to hold the whole list of authors responsible for the sins of one greedy person, whom you're depending on to be the resident expert in their field.

And when the review process fails to detect it, you really have to wonder if the other authors should be expected to know. Since everything thinks peer review is so great and all that.

Collaboration is certainly a mixed blessing for the junior scientist. You can get some visibility by being involved in a high-impact paper with a big(ger)-name scientist, but you may not be taken seriously in the process of putting together the paper that's going to make the splash. As your collaborators see it, you're there to contribute one piece of the empirical scaffolding; after that, you're supposed to get out of the way. Leave the manuscript writing and editing to those of us with experience, missy!

But, as Ms. PhD points out, if your name is on the paper, you're taking responsibility for the goodness of what's in it. If you're ready to profit from it professionally, you're also putting yourself on the hook if there's something wrong with the paper. The "something wrong" could range from mistaken conclusions arising from sloppy data management to fabrication, and even if it was a collaborator's screw up, everyone on the paper gets to share in the shame — even the well established scientist. (I'm thinking of a city in Maryland ...) Putting your name on the paper is putting your reputation behind it. So there ain't no way in hell a collaborator should be keeping you out of the writing and revising loop. Indeed, it seems like coauthors should insist that everyone on the team read a full draft and contribute comments on it. Not only would this ensure that each author's contribution is represented accurately, but it would also be a good way to check whether the various pieces of the research are understandable to scientists who don't necessarily work with those precise experimental systems or methods. Think of it as the in-house peer review before the peer review.

Why wouldn't someone want to get a collaborator into the writing-and-proofing loop? I suppose researchers who are horrible procrastinators, or who get really blocked writing new stuff, or who get overly obsessive about comma placement, might slow down the process. Someone on the team ought to be able to hammer out the rough draft in a timely fashion, but it seems like all the writing and editing contributions could be done according to the group's deadlines. (People set deadlines for getting experiments run, so why not for drafting and editing?) Possibly a principal investigator might be hesitant to get feedback from a collaborator with who he had a long-standing feud about whether or not dangling participles are evil. Or, for that matter, from a collaborator who disagrees about the significance of a particular experimental result reported in the manuscript. But here again, peer review starts at home: if all the authors in the collaboration can't share their reasoned views and together come to some kind of agreement, why should they expect that their manuscript will be able to survive peer review by scientists not involved in the project?

A collaboration is a commitment that you try to make work for the benefit of all involved. But, if someone wants your data without your full involvement, it's time to ask whether this person is thinking about what's good for you, or for the collaboration as a whole. It is not wrong to insist that you be involved. In cases where someone is trying to prevent your full involvement, it is not wrong to explain that this is contrary to the spirit of collaboration, nor is it wrong to assert that you and your data cannot in good conscience remain in an arrangement where your proper participation is thwarted.

People who really need the data can either acknowledge your rights and responsibilities as the scientist who generated the data, or they can get their own damn data. When you're a junior scientist, pretty much all you have is your ability to generate good data and your good name. Anyone who would ask you to risk your good name by staying out of the process of generating the manuscript to which that name will be affixed forever, in front of everyone in the scientific community, is someone who isn't taking your best interests seriously.


Technorati tags: authorship, collaboration, peer review