In “Adding Authors After Submission: How Difficult Should It Be?”, I looked at the issue of authors being added after a paper had been accepted. There are legitimate reasons for changing authorship, but there are also obvious integrity risks, particularly when authors are added very late in the process.
That discussion raises a broader question. Why do we only see the final author list?
A journal will usually retain a detailed internal record of what happened to a manuscript. It knows when the paper was submitted, when revisions were received, what editorial decisions were made and which versions of the manuscript were uploaded.
A published paper may tell us that it was received on one date, revised on another and accepted later. We will not see the original submitted manuscript, the version after major revision or the successive changes that took place during peer review. We only see the final published paper.
So, if we are arguing that authorship changes should be visible, we are asking for something different. The question is whether authorship is important enough to justify that difference.
Authorship is not just another manuscript change
A paper may change substantially between submission and publication. Arguments can be refined, figures replaced, analyses added and conclusions altered.
Those changes matter, but we don’t need to see them all. It would be nice to, but we have to have some faith in the reviewers and editors.
But I would argue that authorship changes matter in a different way. Adding or removing an author changes who receives credit for the work. It may change who accepts responsibility for the research. It can affect careers, promotion, reputation and, in some systems, financial reward.
That makes authorship more than another component of the manuscript.
The author list is a statement about ownership, contribution and accountability. An authorship trail would also provide a valuable record of who was added or removed, when the change occurred and why.
For that reason, I think there is a case for treating changes to authorship differently from changes to the text.
Authorship changes are not necessarily suspicious
It is important not to begin from the assumption that any change in authorship signals misconduct.
There are many legitimate reasons why an author list might change. A substantial revision may require new experimental work, additional statistical analysis or expertise that was not part of the original manuscript. A contributor may later satisfy the criteria for authorship. Someone included initially may decide that their contribution was insufficient to justify authorship.
There may also be genuine mistakes that need correcting.
The key issue is not whether authorship changes. It is whether those changes are properly recorded, explained and agreed.
A journal could maintain a record showing the authors at initial submission, after each revision, at acceptance and at publication. Every change could be accompanied by a reason and confirmation that the affected authors agreed to it.
This would create what we might refer to as an authorship audit trail.
Journals should certainly keep the trail
Manuscript systems already retain enormous amounts of information about a submission. It should not be difficult to preserve each version of the author list and record when a name was added, removed or reordered.
Editors could also be notified automatically whenever that happened.
The purpose would not be to assume wrongdoing. It would simply ensure that a potentially important change did not pass unnoticed as the article passed through the various stages of submission, revision, acceptance and publication.
Some changes might deserve closer scrutiny than others. Adding one author during a major revision involving substantial new work may be entirely reasonable. Adding several authors immediately before acceptance might raise a few additional questions. Repeated changes might also be significant.
If a manuscript repeatedly gains and loses authors during the publication process, that pattern may be more informative than any one change viewed in isolation.
But should readers be able to see it?
This is where the argument becomes more challenging. We do not see every intermediate version of a manuscript, so why should we expect to see every version of the author list? My view is that authorship changes are so important that they justify greater transparency.
A published paper might contain a short, structured record such as:
A published paper might contain a short, structured record such as:
- Initial submission: A, B and C
- Major revision: A, B, C and D (Reason: D contributed substantially to additional statistical analysis requested during peer review. Approval: Changes were approved by A, B and C)
- Final acceptance: A, B, C and D (Approval: All authors approved the author list)
That information need not imply that anything improper occurred. In fact, it could do the opposite. If the change was legitimate, the public record would show that it was explained and properly managed.
Why authorship may deserve special treatment
If a paragraph changes during revision, readers generally do not need to know who changed it or exactly when. The peer-review process is expected to improve and reshape the manuscript.
But if an author is added or removed, the identity of the people claiming credit and responsibility for the research has changed. That is not a change resulting from peer review or editing. It is a change in who is attached to the scholarly record, made by the authors independently of the reviewers and editors.
In my mind, that justifies a higher level of transparency than we expect for ordinary manuscript revisions.
The real value may be in the data
A public authorship trail could also create something much more valuable than an explanation attached to a single article. If journals and publishers recorded authorship changes in a consistent, structured way, researchers could begin analysing them at scale.
How often do author lists change between submission and publication? At what stage are authors most commonly added or removed? Are late additions more common in some disciplines than others? Do particular journals show very different patterns from comparable journals? Do some publishers see substantially more authorship changes than others?
We could also ask whether certain types of changes occur repeatedly.
Researchers might identify journals where late additions are unusually common, fields where authorship changes occur frequently, or patterns associated with particular groups of submissions.
That would not prove misconduct. But it would create signals that could be examined. The important point is that the data would exist.
Transparency should not become accusation
There is an obvious risk. If authorship trails became public, readers might treat every change as suspicious. A paper showing that three authors were added during revision could attract unwarranted attention even when there was a perfectly reasonable explanation. That is why any public trail would need context.
Just a record of changes would be less useful than a record showing both what changed and why. Journals would also need clear policies explaining that authorship changes can be legitimate and that the existence of a change should not itself be interpreted as evidence of misconduct. The aim should be transparency, not suspicion.
A different kind of publication history
We should be careful not to pretend that scholarly publishing already exposes every stage of a manuscript’s development. It does not. We often see only the key dates and the final paper.
But authorship may be different. If authorship determines credit, responsibility and professional recognition, then perhaps changes to authorship deserve a level of visibility that ordinary textual revisions do not.
The final author list tells us who ended up on the paper. An authorship trail would tell us how that list came to be.
And if thousands of journals recorded those changes in a consistent way, the result could be more than a useful record for individual papers. It could become a new transparency dataset for scholarly publishing itself.
About the author
Graham Kendall is a university leader and computer scientist with a long-standing interest in research integrity, scholarly publishing and higher education governance. He has published more than 300 peer-reviewed papers and has served in senior editorial roles, including as Editor-in-Chief of an IEEE Transactions journal.
He is Vice-Chancellor of GlobalNxt University, Malaysia, an Emeritus Professor at the University of Nottingham, and holds honorary and distinguished academic appointments internationally.
My newsletters
I publish two weekly LinkedIn newsletters.
Publishing with Integrity looks at research integrity, scholarly publishing, peer review and the systems that support trustworthy research.
Governance, Leadership & Research explores higher education leadership, governance, research and the impact of artificial intelligence on universities and organisations.