Is it reasonable for an author to withdraw an accepted paper because they have since done better work?

Question

I submitted a paper to a good journal, it was refereed, and after a year it got accepted for publication.

But now that a year has passed, I no longer find the results or methods of that paper to be that great (novel?). And much of my more recent work, which has been submitted to other good journals but has not yet been published, is of greater quality and improves on the earlier methods.

Should I withdraw the paper, or just go ahead and publish it? To me it seems that there are negative consequences either way:

• If the paper is published, then the readers may perceive the author to be shallow or inferior in some ways. Also, if the more recent work is published then people may think I'm over-using a particular method, or "salami slicing". Of course, the latter would not be accurate, as the most recent work was begun several months after the first paper was submitted.

• If the paper is withdrawn, then the referee and editor (whom are very respected) may hold a grudge, or simply have a lesser opinion. That could impact future submissions and my reputation.

---

Although, if I formed a habit of submitting and withdrawing because of better and better results... my work would eventually be so wonderful that nobody would know about it!

Answer 1

No, this is not reasonable. It's up to you to determine that your paper meets an acceptable standard before you submit it, and stand by that determination unless you find a critical error in the work. Maybe it's not as good as what you're working on now, or maybe you're just overthinking things out of perfectionism, but it doesn't matter. Not every paper is a breakthrough, and that's alright.

Practically speaking, one less-than-exciting paper isn't going to give you a reputation as a salami slicer, and even if it did, this is much less bad (think several orders of magnitude less bad) than having a reputation for wasting editors' and referees' time because you won't stand behind your own work.

I'm not sure about the legal fine points, but you might not even be able to retract the paper at this stage without the editor's agreement (and why would the editor agree if there's nothing wrong with the paper?).

Answer 2

The work obviously will be less novel (particularly to its author) a year later, but that does not seem like a good reason to withdraw it. If the paper contains good and useful work, withdrawing it at that point is just wasting everyone's time: the authors', the editor's and the reviewers'. Unless there are substantial errors in the paper, or the journal demands too taxing changes before publication, I'd say go ahead and have it published.

And much of your more recent work, which has been submitted to other good journals but has not yet been published, is of greater quality and improves on the earlier methods.

There is no guarantee that the more recent work will be published in good journals, so why risk the old work as well?

Also, if the more recent work is published then people may think the author is over-using a particular method, or "salami slicing". Of course, the latter would not be accurate, as the most recent work was begun several months after the first paper was submitted.

Usually the submission date is included in the published version of the paper, so it should be easy to disprove this.

Although, if I formed a habit of submitting and withdrawing because of better and better results... my work would eventually be so wonderful that nobody would know about it!

It would also be a good way to annoy editors (and possibly reviewers) making it harder to publish those eventual, wonderful results. Not to mention that never publishing is rather bad for future funding opportunities.

Answer 3

I doubt that reviewers hold grudges. Many won't even know whether a paper they reviewed was ever published. Editors may be a different story, if they expect the paper for an upcoming publication.

On reason your older work seems "shallow" is that you have advanced in the art since you wrote it. You would feel the same if it had been published long ago. Celebrate your growth. Poets and other writers often feel the same way about early work, though that isn't universal.

But if you don't think your newer work is an improvement over the old, then you haven't grown much.

My advice is to publish it. One important reason, in my view, is that the ideas in it may inspire some future student to advance along similar lines. The journal editor will be happier, as you note.

I doubt that most academics are judged by their earliest work, or by the work they have grown beyond, but rather by their overall contribution to the field and by their best work.

---

Of course, if you find errors in a paper that were missed by the reviewers it is a different story. The editor should be informed, at least.

Answer 4

It is completely normal for papers to describe work that's at least a year old by the time they're finally published. It is completely normal for the authors of papers to continue working on the topic and obtain new results. Of course your work today on this topic is better than the work you did a year ago.

No reasonable person is going to accuse you of salami slicing if you produce multiple papers on a topic over a period of years. Salami slicing is dividing a single piece of completed work into more papers than it needs to be. That's not what you did: you took a finished piece of work, published it and then extended it. If everybody waited until all possible consequences of their work had been fully determined before publishing then, for example, the entirety of physics would consist of a single (as-yet-unwritten) paper called "How the universe works. No, really, we figured it out." Meanwhile, you'll find it hard to get your next job if you never publish anything because there's always some way of extending the work.

It would be completely inappropriate to withdraw your paper, unless you've found that it's incorrect. The referees have spent considerable time reading it; they and the editor have decided that it is interesting and novel enough to publish. Good luck with your next paper; it sounds like you've already made a lot of progress on it.