[abecedarius]

This seems new, and passed over in all the comments so far:

> None of them had ever heard of a paper in any field being disappeared after formal publication. Rejected prior to publication? Of course. Retracted? Yes, but only after an investigation, the results of which would then be made public by way of explanation. But simply disappeared? Never. If a formally refereed and published paper can later be erased from the scientific record and replaced by a completely different article, without any discussion with the author or any announcement in the journal, what will this mean for the future of electronic journals?

I'd had the impression online journals normally had some plan for archiving. So insiders were left out of the threat model?

[jessriedel]

> Retracted? Yes, but only after an investigation, the results of which would then be made public by way of explanation.

Just to clarify: when papers are retracted, they are not normally erased from the database and replaced with an explanation. See this Nature paper for an example; the entire paper is kept available online, just as before, except that the words "RETRACTED" is stamped on each PDF page

https://www.nature.com/articles/nature10167

Likewise, nothing published to the arXiv is ever removed. A retraction notice can be added as an update, but the complete version history is always available.

Articles are much more likely to be "un-published" (wiped from the archives) in journalism than scientific publications.

[thraway180306]

Sorry, but the ArXiv does have a moderation team and stuff is being removed. Alike, most online journals remove stuff by wholly replacing it with a retraction notice. Journals printed on paper issue retraction letters and leave it at that, but some also do restrict online access afterwards.

What no journal does is to replace erroneously published article with another entirely unrelated one in-place. This is a very poor behaviour on the part of the managing editor here, who is currently on leave and replaced.

Nature is unusual, for example they publish letters complaining about ArXiv's moderation, beside publishing a lot of editorial and opinion stuff.

[jessriedel]

> Alike, most online journals remove stuff by wholly replacing it with a retraction notice

Can you provide a copy, or even examples? I can pull up dozens of Science, Nature, and Physical Review retracted articles to support my point. If those huge families of journals behave as I say, on what basis are you claiming "most" journal don't?

> What no journal does is to replace erroneously published article with another entirely unrelated one in-place.

Did anyone ever suggest this was the case?

> the ArXiv does have a moderation team and stuff is being removed

I know several members of the moderation team. They of course reject some articles at the time of submission, but I'm not aware of any being retracted and un-published after being posted publically. Can you please link to such a retraction, or to evidence of this in the Internet Archive?

[jessriedel]

(Note: in the case on the arXiv, the process is called "withdrawal" rather than "retraction" since it's only ever done by authors themselves, and only on a voluntary basis. The arXiv does not have an editorial team who would make retractions.)

[mirimir]

Looks like there's no hard copy. It's totally online.[0]

I checked https://web.archive.org/ but the 2017/10/23 snapshot ends at p. 1539,[1] and the 2017/12/01 snapshot has the substitute paper.[2]

    Ajay Kumar, Niteesh Sahni, and Dinesh Singh 	 
    Invariance under finite Blaschke factors on BMOA 	1641

So unless someone grabbed it, it's toast. Except for the preprint, anyway.

0) http://nyjm.albany.edu/nyjm.html

1) https://web.archive.org/web/20171027011421/http://nyjm.alban...

2) https://web.archive.org/web/20171201163957/http://nyjm.alban...

[jaccarmac]

It's still available on the site, AND I've grabbed it. http://nyjm.albany.edu/j/2017/23-72v-orig.pdf