Is it ethical to post preprints?

Liam Kofi Bright, Remco Heesen, and I have argued for transitioning toward a crowdsourced approach to peer review. Yet, one common concern raised to our alternative approach is that it would effectively undermine anonymized review--the main mechanism in our current system for mitigating bias in favor and against authors on the basis of knowing their identity. I honestly appreciate this concern, though I still think the overall benefits of moving toward a hybrid peer-review model like that used in math and physics (which combines a crowdsourced approach with traditional journal review) are likely to outweigh the costs. 

But regardless, I recognize that not everyone agrees that peer-review should change, and this has gotten me thinking a bit about the ethics of particular actions surrounding these issues. For example, one thing that advocates for change might do is to post preprints on PhilArchive or PhilSci Archive, normalizing the practice. But, is it ethical to do this?  Last week, I asked whether journals or referees care whether authors post preprints--and I'm glad I did. For, as Jake Nebel chimed in, at least one philosophy journal, Ethics, does care about this. Their official editorial policy for Submission Eligibility reads:

We ask that authors not post or publish online manuscripts that they have submitted or will be submitting to Ethics, as doing so may compromise the anonymous review process or may disqualify the submission on grounds of prior publication.

So, I'm curious about a couple of things. First, irrespective of whether other journals have a policy like this, is it ethical for authors to post preprints--either on an online archive or on their person website--prior to journal submission (given the manner in which it might compromise anonymized peer review)? If it's not ethical, what alternative steps for promoting reform to peer review (for those of us who advocate for reform) are ethical? I'm really curious to hear what everyone thinks!

Comments

[Andy]

I don't see how there could possibly be a moral problem with posting drafts online. A few months ago I looked at all the papers I had under review and added up the total time they had been under review. It added up to over 8 years. As far as I'm concerned the only way to make many of my best papers accessible in a reasonable time to other researchers is to post drafts online. I may well be long gone from academia before some of them get published.

If the situation with peer review was different (i.e. with respect to review times and the wildly variable and inconsistent quality of refereeing) then maybe there would be a case for not posting drafts online. But right now it seems absurd not to.

[A thought]

I haven't thought deeply about this issue, but I'd be very happy to see some experimentation and evaluation of such practice. For example, one or few journals (say) Ergo can explicitly encourage all authors to post pre-print before submission like practiced in science, and still require editors and referees to respect anonymity as much as they can, and then see what happens. Such a policy might affect people's decisions to submit to such a journal, but still seems ethical. Without experimentation, I find it hard to wrap my mind around the possible complicated effects it might involve---though I can also imagine that such an experiment is not very feasible or has limited value.

[Matt Weiner]

Ethics states two concerns: that posting a preprint online might compromise the anonymous review process, and that it might disqualify the submission on grounds of prior publication.

The first doesn't hold up, in my opinion. It's true that, if you post a paper online, a potential reviewer might discover your identity before being asked to review it. This can also happen if you write a paper based on a talk you've given publicly. Or if you discuss your ideas with someone before you write a paper. Or if you've already published a paper where you argue for a related view. Or... you get the idea.

There's a difference of degree here, probably, but online preprints aren't going to reach so many people that it's impossible to find a reviewer who hasn't read them. The moment that happens is the moment we will have won the battle against journals.

The second concern, that posting an online preprint constitutes prepublication, is obviously not an ethical consideration against posting preprints. It's the journal threatening authors to try to preserve their business model. This may create a prudential reason not to post preprints, but certainly not an ethical one.

Caveat: As is probably obvious, I think the peer review system should change and am hostile to the journal system, so people who aren't that hostile might want to discount my views.

[Richard Y Chappell]

Sharing pre-prints advances the scholarly goals of sharing and improving research. How could anonymous review be more important than that? I guess if one believed that most potential reviewers were *so* biased that the higher chance of knowing the author's identity would *so* distort their verdicts that the journal would end up publishing worse papers overall. That is, even though each paper (on average) is better than it otherwise would have been, the journal ends up selecting worse papers to publish (e.g. prioritizing crummy papers by famous folks over stellar contributions from junior or otherwise lesser-known academics).

I guess that's logically possible. But it doesn't seem very likely to me, which makes me think that we should positively favour everyone sharing their pre-prints whenever possible.

[anon]

Sharing preprints, or posting them in many places, doesn't seem any more anonymity-compromising than doing things like presenting your paper at a conference. The idea isn't that our papers are supposed to be top secret, the idea is that the journal should hopefully be able to find someone who hasn't heard of the paper yet.

I guess there are some outside cases where you might post them in a place where simply too many potential referees could see the paper. For instance I don't think I would email all of Philos-L with a link to my paper saying, "Hey folks, please send me some comments before I send this out for anonymous review next week!".

[Epistemic]

Not sure if it’s ethical or not unless somebody deliberately does not follow a particular journal’s rules. But publishing into a pre-print archive by itself doesn’t seem like it would violate any ethical standards. In these cases, ethics doesn’t apply much. Instead, we should focus on the *epistemic* concerns of pre-prints. That would be more fruitful than the supposed ethical concerns in my view.

I do like pre-prints as well. I’m not entirely against some journals remaining orthodox in their peer-review process. It would be nice to have both to see which process produces better quality articles from a comparative approach.

However, I do think that pre-prints should moderate the comments to weed out useless comments that do not help the papers and readers at all.

[not nobody]

I have to agree with many of the others - this is the wrong question. It is imprudent to post a pre-print if you want to publish in journals that forbid it. Indeed, you are undermining your own goals. I do not post pre-prints because I cannot imagine getting constructive feedback from a stranger who just happens to find my paper. I am not nobody (really, who is?!), but I am not so famous are going to read whatever I post on the web. I want quality feedback - so I employ other strategies for getting feedback.

[NarsiMarsi]

Here's one for team immoral! Since graduate school, I (and likely you, too) have been repeatedly exposed to what we could call the "Suspicious view" of how publishing in philosophy works. Disclaimer: it's mostly a bitter, cynical caricature of some of the problems of our publishing world. Still, if this view is even vaguely in line with reality, a case can be made that posting online preprints is immoral, at least for affirmed scholars and scholars coming from prestigious institutions.
For the "Suspicious view", the publishing game is mostly a matter of politics. Journals, editors, and referees act as gatekeepers who pursue their own interests. The journal's interest is to pursue prestige, publishing big names from big universities to fatten their metrics. When a journal is mostly run by a specific department, it will also have a strong incentive to publish work from within the department, help their graduate students get a first paper out, and set the agenda in a way that is favourable to their interest (e.g. if the department is big in teleosemantics, they'll profit from publishing more on that). Editors and referees pursue similar goals. They reserve a favourable treatment to authors from their own cliques, i.e. the close circle who shares their views, invites them to conferences or knows them personally, and who is likely to be equally favourable to them.
Agents need not be malicious in this model – they may not even realise what they do. The point is just that many are (directly or indirectly) moved by these incentives. The result is that powerful institutions and academics thrive by gatekeeping access to a few "prestigious" journals, while younger researchers and scholars from smaller institutions struggle to get their work published.
In this world, anonymity is a precious countermeasure: it limits each agent's ability to pursue their own interests, and increases the incentives for selecting manuscripts based on quality (instead of own interest).
Arguably, posting preprints online can undo the goods brought in by anonymity. Now the author's identity is just a click away. For those to compare it to "presenting a paper at conferences", two observations. First, a numerical one. Preprints (especially from prestigious authors) get hundreds (sometimes thousands) of downloads, mostly from scholars who are genuinely interested in the topic (hence potential referees).The size of your audience when you give a talk isn’t even remotely that size. Second, posting preprints, but not giving talks, makes it POSSIBLE for anyone at any time to discover the author's identity, simply by googling a bit of the paper – which is enough to say that the paper is no longer really anonymous.
This is a situation where posting a preprint online (and bypassing anonymity) is extremely convenient for the agents who stand to profit from the incentives laid out above. Given that they stand to gain unfairly from sharing preprints, a good case can be made that FOR THEM posting the preprint is immoral. Not so for ERCs, scholars working in a small colleges, etc. For them, posting the preprint is simply a bad move – it is immoral only insofar as they are hurting themselves.
Of course, as mentioned in the beginning, the "Suspicious View" is a gross exaggeration. Whether "team immoral" has a decent point will ultimately depend on how far their exaggeration is removed from reality. Still, even if it turns out that the Suspicious View is merely a "loser's tale", invented out of resentment by those who can't get their papers published, I think that reflecting on these issues is important, and that’s why I’ve been playing devil’s advocate. If privilege, prestige, and power differentials play a role in increasing one's chance to publish, the moral status of posting preprints is actually more complicated than suggested in previous comments!

[not buying the (new) propaganda]

As NarsiMarsi himself points out, on the "Suspicious View", there is no anonymity: big names are published more often etc. So it is not clear how on this view "anonymity is a precious countermeasure". It would be if it existed, but it does not, so it is not.

[Assistant Professor]

I would advocate strongly for posting a (near final) draft of a paper once it has been accepted to a journal in order to maximize accessibility to the work - even if journals also have rules against this in order to drive traffic and money to their journals.

But without an obvious mechanism in philosophy to get feedback on pre-prints in the ways other fields have established (methods I take it Marcus and his colleagues point to in their paper) it is hard for me to see the ethically justification for posting pre-prints that outweighs the harms for undermining review, if as a field we continue to value anonymous review. I could even see reasons for authors to consider explicitly NOT naming their paper the same thing as the talks they have given on the topic to further enhance the likelihood of anonymous review. For example I am fairly certain at least one reviewer identified me as an author of a paper of mine they were assigned to review by finding a handout on a talk with a similar title I gave on my academia page (that was posted there to enhance accessibility during my talk). I could see that it was discovered by a google search the same day the paper was assigned to reviewers, per the journal. It can't be a coincidence. Of course I value making my work accessible, and hadn't foreseen this conflict, but I also often search for recent work on a topic of a paper I am reviewing to make sure I am not missing relevant recent work with which I might not be familiar. If I do this and someone posted their preprint I might easily discover it, just by trying to do due diligence of being aware of the landscape of recent pubs on the topic as part of my peer review.

If as a field we had good mechanisms for pre-print commentary and feedback I would feel like there are benefits that might outweigh the risks/harms. But as much as I value accessibility of work, it doesn't benefit me much as a scholar to read other people's pre-prints that I can't yet cite in my own work on a topic without it being published, it isn't clear to me that people posting pre-prints want unsolicited feedback, and it does mean I am quite likely to ID the person if I were peer reviewing their paper.

[F. Contesi]

Hi, Assistant Professor! Since the beginning of this year, there is in fact a way to give/get feedback on preprints posted on PhilArchive:

https://freelosophy.github.io/

We look forward to receiving comments on Freelosophy!

[Postdoc]

In response to the comment by "not buying the (new) propaganda":
Surely you are right that "sometimes mechanisms to ensure anonymity can be fooled". But this doesn't mean that these mechanisms are completely ineffective. So it's a big leap to conclude that anonymity cannot play its role of "countermeasure" for prestige bias, etc. On the contrary, I think the opposite point can be made: since anonymity is so precious, we should do our best to protect it, which includes not posting preprints online.