Publishing a paper critiquing one's own work?

In our April "how can we help you?" thread, a reader asks:

I've discovered a weakness in an argument I published recently. It would be very easy to write a paper pinpointing this weakness and strengthening the original argument to avoid it. Is this advisable? If so, how far should I go to hide the fact that I am the author of both papers from editors/referees?

Great questions! There was a great case of this recently, in which Hanno Sauer (Utrecht University) published a rebuttal to a paper he published just a year earlier critiquing the history of philosophy. 

Although one doesn't see this sort of thing very often, I personally think it's a nice reflection of intellectual honesty. We're after the truth as philosophers, are we not? So, if you think something you've published is mistaken in some way, why not critique it? Further, if you can use the critique to strengthen the original argument, that seems to me like all the more reason to do it!

But these are just my thoughts. What are yours?

Comments

[Hüseyin Güngör]

I guess this is a great place to remind that one of our greats did this in the most hilarious fashion:

https://retractionwatch.com/2017/08/01/35-years-philosophy-journal-corrects-article-cat/

Here are the articles in question:

https://andrewmbailey.com/dkl/Veridical_Hallucination.pdf
https://andrewmbailey.com/dkl/Censored_Vision.pdf

[Assistant Professor]

This sounds like a nice way to refine your work. I would refer to your own work in the third person when drafting the new piece.

People have mixed feelings about the use of cover letters in journal submissions but this is a good use of a cover letter: say to the editors that you are critiquing your own previously published view, that you have prepared the paper in a way that maintains anonymity while allowing for proper citation to the previously published works, and that you are noting this in the cover letter to make clear that you (as the author of the previously published work being critiqued) are not an appropriate peer reviewer for this paper and b) to stave off critiques from peer reviewers that you got the original view "wrong" (as sometimes happens when peer reviewers don't know the author of the paper at hand is also the author of the original view they are saying the author got "wrong").

You can always ask to make minor revisions if the paper is accepted to change some of the grammar to be in the first person.

Good luck!

[PhilyOsopher]

I think that this looks cringy and smacks of a desperate 'marketing' attempt. When next you write about the topic, refine the view then and indicate in a footnote how it corrects an earlier weakness in an earlier paper of yours. Nothing more is needed to satisfy intellectual honesty.

[nonewopinion]

"It would be *very easy* to write a paper pinpointing this weakness and strengthening the original argument to avoid it."

If it is that easy, it may not warrant an independent paper. Maybe other people can also see the way to strengthen the argument when they engage with your view? But if it is a surprising, interesting and/or important move, then by all means make a paper out of it.

As for how much you should hide your identity, just follow the norm and discuss your paper in the third-person. Deanonymize it once it is accepted.

[PD]

I don't think there's anything wrong with this. I think we should, as far as possible, stop so strongly associating an argument in the literature with the person who gave the argument. There's now an argument in the literature that you think could be stronger. You want to write a paper explaining the weakness and how to strengthen the argument. That's a perfectly legitimate move to make, and I don't think that the fact that you are the author of the original argument changes that fact. For similar reasons, I think that we should stop thinking it either unfair or passé to criticize a view that author A has defended but has since changed. If the original view is still a live option and your reasons for objecting to the view are different from the original author, then there's nothing wrong with criticizing the original view. To think otherwise strikes me as a weird kind of ad hominem mistake. It's not about the author. It's about the argument.

[academic migrant]

This paper came to my mind.

Sauer, H. (2023). The ends of history: A reply to Sauer. Inquiry, 1-11.

It responds to

Sauer, H. (2022). The end of history. Inquiry, 1-25.

[Be Bold]

If all the new paper does is refine or reject the old argument, then that looks like padding your publication record. I also fear that, for many people, it suggests that you didn't put an appropriate amount of thought into your original paper. That's unfair and not a good attitude for incremental progress in philosophy, but something you have to consider.

A better approach would be to incorporate the new argument into a larger and more daring paper that builds on your original paper. For example, you could respond to objections that are common to both arguments, discuss the compatibility of your conclusion with related theses in that area of philosophy, and explore previously undiscussed implications of your conclusion/premises.

In general, I think if a paper looks easy to write, then it's probably too unambitious, and one should be bolder.

[Concerned about how it looks]

While I agree that I like the general idea (stop worrying about who wrote what and just engage with the ideas!), I am also worried that if this is too easy it might not be the best look. It might suggest that you didn’t take sufficient care in writing the original paper or that you are trying to squeeze too many publications out of the same idea, or, worst, that the original mistake was intentional to allow you to write the follow up.

This depends entirely, I think, on how serious the mistake is and whether the change to the argument is notable.

For example, if you are just replacing a quantified in a particular premise to avoid some objection about edge cases that don’t matter. Or replacing a strong principle with a weaker one that will do the job. Then I would be worried about the above perceptions.

On the other hand, if your original argument explicitly relied on appeal to physicalism but you know think the argument goes through for idealists and panpsychists too, then that sounds like the sort of thing that would be big enough to publish.

I think the case where you are explicitly rejecting your earlier conclusion, rather than merely amending the argument, doesn’t suffer from any of these worries.

To be safe I would hold yourself to a higher standard here then you would hold someone else if you were reviewing their response to your paper.

[citer]

A related concern to the ones raised above, if YOU are the only one critically engaging with your publications then that is not good. One can see - with the help of Google Scholar - how often the citations to an article are due to the author's own self-citation. High levels of self-citation is widely regarded as a problem.

[David Thorstad]

Luc Bovens published a reply to Bovens (https://academic.oup.com/analysis/article-abstract/70/3/473/157919).

If I recall, he took some good-natured joshing for this at his LSE retirement, but the general sentiment was that it was an admirably honest thing to do.