mjhjmtu

Co-authors are increasingly required to report their individual contributions to a research paper. But can they report their (internal) disagreement?

Co-authors may disagree on parts of a final draft. Each may have own interpretation of (parts of) the results or view of their implications.

The different views can be of course expressed in the publication without attribution, e.g.:

Our results may mean x, but they may also mean y.

But some co-authors may find others' interpretations/views controversial, or they may wish to get "exclusive" credit for their own ideas.

In such cases, is it appropriate for co-authors to explicitly attribute certain interpretations/views in a paper to their owners? Should they do that?

Co-author A thinks results mean x, whereas co-author B thinks they mean y but not x.

It is rather unusual, but it has occurred before. In the paper

Piccione, Michele, and Ariel Rubinstein. "Equilibrium in the Jungle." The Economic Journal 117.522 (2007): 883-896.

each of the authors has their own conclusions, marked "4.1. Concluding Comments by MP" and "4.2. Concluding Comments by AR." It should be noted though that the writing in economics tends to be less structured (no such thing as a "method section") and this is a somewhat unconventional paper to begin with.

This might not perfectly fit your bill, but the paper Trialogue on the number of fundamental constants
by Michael J. Duff, Lev B. Okun, and Gabriele Veneziano
basically consists of three co-authors disagreeing.

Abstract:

This paper consists of three separate articles on the number of fundamental dimensionful constants in physics. We started our debate in summer 1992 on the terrace of the famous CERN cafeteria. In the summer of 2001 we returned to the subject to find that our views still diverged and decided to explain our current positions. LBO develops the traditional approach with three constants, GV argues in favor of at most two (within superstring theory), while MJD advocates zero.

In theoretical cryptography, we have a major paper by Canetti, Goldreich, and Halevi, showing that in general cryptographic systems which are proven secure in the so-called random oracle model are not necessarily secure in the real world. At the end, they each offer different opinions about what this means regarding the usefulness of ROM proofs. (Basically, Goldreich says they offer no evidence whatsoever of real-world security, while the other two say they're not ideal but better than nothing.)

Presumably, both possible conclusions have arguments backing them up. (If they don't, then neither "Co-author A thinks results mean x" nor "co-author B thinks they mean y but not x" belongs in a published paper.)

All co-authors should agree that these arguments are valid, even if they disagree about which ultimately carries the day. If they disagree that an argument makes sense, this is not a problem that can be fixed by having only one co-author sign their name to part of a paper. If you see a flaw in your co-author's work, you should point out the flaw, and work with them until it is fixed!

I will give an example from mathematics, which I feel most comfortable talking about. Perhaps there is heuristical evidence for Conjecture X: it can be shown to hold for almost all randomly chosen objects. Co-author A is convinced by this; co-author B points out that such and such obstacles could appear in very unlikely cases. Maybe each author can point to examples of other problems where the equivalent conjecture turned out to be true or false.

Both co-authors should agree that the argument about random objects holds water. Both co-authors should agree that a certain kind of obstacle could invalidate the conjecture. Both co-authors should agree that the examples of other problems are referenced correctly. So there is no reason why both co-authors can't stand behind both sides of the argument, even if they disagree about which conclusion is likelier.

It might make sense to mention that the co-authors disagree about the conjecture, because this makes it clear to the reader that the issue is very far from settled. Which author thinks what shouldn't be persuasive to the reader, but it might be worth mentioning anyway; maybe the reader will write to one of the authors to debate the conjecture.

It would be inappropriate for a co-author to put their name down on one side of the argument in an effort "to get exclusive credit for their own ideas". This is no different from writing "although this paper is by authors A, B, and C, author B didn't help at all with proving Theorem 3."

Yes, but I guess that this would usually set the tone of the paper, i.e. the paper would focus on a debate between the co-authors. For instance:

https://www.nature.com/news/does-evolutionary-theory-need-a-rethink-1.16080

But that is something different.

I don't think mentioning author names in the manuscript is a solution. All authors should take responsibility for everything that is written in the manuscript. So I would say present both interpretations as possibilities without committing to either.

An alternative solution (and one which I try to use) is to avoid views/interpretations that your co-authors don't agree with, and instead write your interpretations in a follow-up opinion paper. This looks much better than the authors disagreeing with each other, I believe, especially if this is a basic research paper.

If I may: See also my book with Martin Osborne: A course in Game Theory, MIT Press 1994. Could be downloaded freely from my home page arielrubinstein.tau.ac.il
In the book we have many points when we express our disagreements in separate paragraphs signed by MJO and AR.