A few weeks ago, a colleague’s postdoc sent out a draft of a paper for comments. I looked at the paper and thought, “Wait, haven’t we published everything we had on this topic years ago?” To put things in context, I had an undergrad honors student working on the theory part of this project. Since the last publication, the undergrad had graduated, attended graduate school elsewhere, and is about to defend his PhD. So yes, it’s been a while.
Honestly, I almost completely forgot about this project; I certainly don’t recall any of the details. I would first have to go reread what we’d published before I can intelligently comment on this new draft.
Which really brings me to the topic of today’s post: When are you really done with a project? And how do you decide if an abandoned manuscript deserves resuscitation or if it’s best left to die?
I know people who more-or-less work on the same problem, or within the same narrow field, their entire careers. This approach doesn’t generally bode well for funding prospects in many fields in the US, but I admit that some who’ve pursued this strategy have been funded and productive on account of being the highest authority in the niche.
Most people do switch topics, or, more precisely, they slowly drift away from an old focus and toward a new one. As they do so, they might maintain some activity in the original area. I personally like to make more of a clean break after I’ve said what I had to say. At any point in time, I run several fairly disparate research thrusts; each will be active for maybe 2-3 grant cycles, which is usually enough for some nice results and papers and about as long as my interest can hold before I itch to do something else, after which I move on while downsizing and often completely shutting down the work on the old thrust. This does result in always feeling like an outsider in a new field — scary, but also invigorating.
I focus on getting papers out as fast as possible and don’t really have a history of sitting on manuscripts for no reason. However, I admit I currently have two papers that are semi-abandoned.
One I simply can’t bring myself to submit, because my gut tells me there is something wrong with it and I don’t want us to look foolish. The paper disputes the work of another group, which is led by an excellent scientist. Our argument is that the other group missed something fairly basic, and I honestly can’t believe that they did; instead, I fear it’s us who might be in the wrong and that the issue is far more subtle than it seems at first (my former student insists that it is that simple), but I just haven’t been able to devote to this problem the time and attention it deserves in order to convince myself one way or another.
The other paper that I have on the back burner was going to come out of a former student’s Master’s degree work; the student was supported during the study by their employer, a national lab. (The student was completely disinterested in getting a paper out of the work and only wanted a degree; I generally expect one paper at the level of a Master’s, but it’s not a formal degree requirement.) This paper would be a very small contribution, but would have a head and tail and a clear pitch, and I think it might review well in a suitable minor publication venue. The question is whether writing up this little nugget and the hassle of getting the approval to submit from the national lab are worth my time, when the contribution is incremental and well below the standard of novelty I like to set for my group.
Other than these two, everything else I have in the pipeline is quite fresh.
The colleague whose postdoc sent me that manuscript is not a procrastinator; he’s usually good about submitting papers, but does seem to have a number that are five-to-ten-years old yet haven’t seen the light of day. I was really surprised that we had anything left unpublished on that particular topic, and it’s interesting that this work is being resurrected right now.
Dear readers, do you have any unpublished manuscripts that have been in purgatory for far too long? If yes, why do you think they never got their chance at submission? How long has it been since the first draft? Do you think they will ever be submitted? When do you think a manuscript is officially past its expiration date (assuming no one’s scooped you)?
xykademiqz 
Oh, good data and good findings are never stale, even if they are starting to fossilize. Four years ago I published a paper based on data I collected more than 30 years ago: it was a cool little thing I worked on very early in my career before I’d even finished my PhD, a fun little side experiment my advisor and I worked on: we moved on to more important things (like my main dissertation research) but never published it… yet never totally forgot about it either. That data (and others) stored in folders on old paper printouts moved in boxes with me to four jobs in three cities. Nobody else had done anything else quite like that experiment, for some reason: an application of a particular technique to a particular situation.
Then about five years ago I read about Prestigious Conference coming up in Really Nice Location that this old stuff would fit perfectly well in, proceedings papers to be published in Special Volume Of Highly Respected Journal. I found and dusted off the data from 1986, put some new spins on it, presented it at the conference and published the paper. I think it even helped me get promoted to full professor.
So… assuming no one’s scooped you, it’s never past its expiration date.
I have a manuscript unfinished since dissertation work. It is modeling work scooped during writing, and since the other approach is similar enough down to formulation of the likelihood function, I am convinced we do not add further value or insight.
There are plenty of short student projects (rotations, MPhil theses, etc), that could be written up in principle. We have done so if the student is aiming for an academic path, and can take charge. Initially, I begrudgingly finished the manuscripts myself when the student was soon not motivated enough after moving on. Now, shaped by much swearing, I scope doing the science as a separate project from the writeup, potentially with different people leading. It is nice when students come and help us get coarse answers to interesting questions; they write a short report of their findings. Turning it into a paper requires much more digging, grit, dotting the is and crossing the ts, and much of the science still happens during writing.
We therefore have a mix of projects ending in short reports, and ones where the project itself is to write the paper based on existing findings, avoiding any tangents (those become next projects). These writing projects are relatively quick, and none have been abandoned yet!
Well, in these cases you are not the first author right? These papers likely matter a lot to the careers of the less senior first authors. It’s certainly your decision on how involved to be and even whether to stay as an author, but I’d say an advisor owes it to the former student to offer some help or at least not throw roadblocks up with papers. (In my field it would be unheard of to not involve your former advisor, at whose site you likely collected data etc.- though former advisor could demure and do little).
1) SO many
2) Lack of time (teaching and service) plus delays from collaborators which let my focus wander plus students leaving plus I’m really bad at the last 10% of olishong….
3) the oldest one is based on data and a part draft from 1996
4) some of them will, yes, in some form – actual data don’t go stale very fast in my field
5) when I can’t reframe it to relate to current directions in the field or something I currently care about…
I had a pretty proof in an appendix to my PhD thesis that I never wrote up for publication. That would be about 34 years ago now. I’m thinking of resurrecting the proof and getting it out there, though perhaps as a blog post rather than a formal publication.
I have recently published a genome-sequencing paper using eight-year-old data from sequencing platforms that are now obsolete (no longer available). The bioinformatics work I did for using the data from the platforms is now irrelevant, as it would be cheaper to use a newer sequencing platform than to hire someone to analyze old data.
I have lots of minor results from my protein-structure prediction days that will never see publication.
I had a thought about your first paper (where you may have found something wrong in a colleague’s published manuscript). In an ideal world, you could send the paper to your colleague, and ask for his reaction. If you were right, then he or she would submit a corrected version, acknowledging your contribution (or even have you as a co-author). If you are wrong, then I am sure you will be set right, and you can always blame your student! (Even better, have your student reach out in the first place).
I have something from National Lab that I would like to get out. It has not been published by anyone else, and the data is sound, though I have moved on and no longer work in that area. I’d guess the data is 10ish years old by now.