The n-Category Café

This is an important question. Perhaps Jim Stasheff, who I believe was on the arXiv mathematics advisory committee, will read this and respond. Or perhaps John could post the same query on sci.math.research, where it will quite likely be read by Greg Kuperberg, currently chair of the committee.

Of course, one of us could just ask by email, but there’s something to be said for getting a public answer to a question of such public interest.

For those who have a very clear understanding of what the arXiv could and could not legally do, and do not see the need for any policy statement, perhaps it would be helpful for me to reveal my own lack of understanding.

I’ve used the arXiv for eight years and put 20 papers on it, and I tend to be careful about reading help pages and other descriptive matter, so I suspect that among arXiv contributors, I’m at least average in my level of understanding of how it works. Among all mathematicians and physicists, I should therefore be above average. But as the following will demonstrate, there are many areas where I’m unclear. That’s why an explicit statement would be so valuable.

Jacques wrote:

Can the arXivs guarantee that a given eprint, currently on their server (and hence, currently open-access), will remain open-access?

Obviously not. The author always has the option of withdrawing the manuscript from the arXiv tomorrow (either because it’s wrong, or because he or his publisher no longer wishes it to be open-access).

This is not obvious to me. Actually, I thought it was impossible to withdraw a paper. I thought that all one could do was replace it with a new, empty version (like this), but that still leaves all the previous versions available for the world to see.

Jacques wrote:

Even though arXivs provide open-access to manuscripts on their server, today, could the Cornell Libraries start charging for access, tomorrow?

Obviously not. This would be in clear violation of the copyrights of thousands of authors, whose manuscripts are on the server.

Again this isn’t obvious to me (though it surprises me less than the previous point). I own the copyright to the book I wrote, but Cambridge University Press make money selling it: hence one party can make money out of another’s copyrighted work. Presumably this is OK because I signed a contract allowing them to do this.

So, if I’ve reasoned correctly, I understand. But to me it’s not ‘blindingly obvious’.

When I started using the arXiv it was at xxx.lanl.gov, and I don’t think I really knew who operated it. At some point I became dimly aware that it had become associated with Cornell. I didn’t know it was ‘owned’ by Cornell until I read it at the bottom of the arXiv front page a few minutes ago. I’m not too sure what ‘owned’ means.

I’m explaining these areas where I’m unclear not so that someone can put me right, but to illustrate how little a typical user might understand of the situation, and therefore how much need there is for an explicit statement.

If the ownership of the arXiv changed to another university, I wouldn’t mind. If it was bought by a commercial publisher and, as John puts it, ‘cluttered with pop-up ads for expensive journals and books’, I would mind.

I’d like to see a commitment that ownership of the arXiv will not be transferred to a profit-making company. I’d like to see a commitment that the arXiv will not start charging for access (even if there are laws preventing them from doing this anyway). I’d like to see a commitment that the arXiv won’t display advertising. (I wouldn’t want Ronald McDonald popping up next to my abstract.) I’d like to see it as a stated aim that all papers will always remain open-access except where the author has requested otherwise or in exceptional circumstances (e.g. defamatory material in a paper).

I think all these commitments and aims are reasonable, and I don’t think any of them are obvious. I’ll continue to do as I’ve always done and trust that the arXiv is run by the ‘good guys’. But I still think we should have that statement.

Jacques wrote:

Can the arXivs guarantee that a given eprint, currently on their server (and hence, currently open-access), will remain open-access?

Obviously not. The author always has the option of withdrawing the manuscript from the arXiv tomorrow (either because it’s wrong, or because he or his publisher no longer wishes it to be open-access).

But, he also notes that on the arXiv submission page one fills out a form saying:

I understand that submissions cannot be completely removed once accepted.

So, I don’t see why the answer to the above question is ‘obvious’.

I don’t care too much what the answer is, but it would be nice to have it be clear, and to have it readable somewhere before one actually submits a paper.

In the interests of transparency, I’d like to note another feature of the arXiv which should be addressed: from which sites are trackbacks accepted?

This issue came up last year, during a kerfuffle over trackback policy. I was a little surprised that a list isn’t readily available; one can find the most recent trackbacks, but I couldn’t find a page showing the sites which can send trackbacks. I asked Jacques Distler about this, since it’s the kind of information I’d like to put on EUREKA, and apparently the reason I couldn’t find that page is because it doesn’t exist. Unfortunate, but perhaps it will be remedied.

(I’d suppose this is a fairly simple SQL query to run, but I don’t know how Cornell has their servers set up.)