Home > Uncategorized > The arXiv should be supported by the NSF

The arXiv should be supported by the NSF

June 29, 2016

What the fuck is wrong with the NSF? Why isn’t it supporting the arXiv?

I have been offended and enraged recently to receive pleading emails from members of the hard-working Cornell University Library arXiv Team for money. As in, please give us $5.

This is a ridiculous state of affairs.

Right now arXiv, which hosts preprints from the fields of mathematics, computer science, physics, quantitative biology, quantitative finance, and statistics, plays an absolutely pivotal role in basic research in this country, especially given the expense and time-consuming journal publishing process.

It has an operating budget of less that $1 million per year, and is somehow left begging for personal donations, supplemented by small grants from the Simons Foundation.

If you look at the mission of the National Science Foundation, it’s first part is “to promote the progress of science.” Moreover, it has an annual budget of $7.5 billion. I cannot think of a better way for it to fulfill its mission than to support the maintenance and expansion of the arXiv.

Am I wrong about this? WTF??

Categories: Uncategorized
  1. mathematrucker
    June 29, 2016 at 8:49 am

    For-profit publishers?


  2. June 29, 2016 at 8:56 am

    Funding Controls. Centralized Funding Controls Centrally.


  3. June 29, 2016 at 9:01 am

    No, you are not at all wrong. However, I suspect NSF is worried about its current set of minders and, so, is acting very timid, not wanting to do anything in the way of being edgy.


  4. June 29, 2016 at 10:07 am

    I fully agree with you. I was shocked too to receive messages from the arXiv people begging for a few bucks

    Liked by 1 person

  5. tina
    June 29, 2016 at 10:10 am

    Physics, Cathy? You forgot physics preprints? (crying in a corner over there…)


    • June 29, 2016 at 10:14 am

      Haha, sorry! I copied and paste that from the arXiv’s website. I’ll add physics!


  6. June 29, 2016 at 10:12 am

    I certainly agree that the NSF should provide funding for arXiv. As for why they don’t, my guess is that the NSF doesn’t usually fund things that smack of maintenance, but only discovery/innovation. The NSF has certainly provided funding to the arXiv in the past, e.g. http://www.nsf.gov/awardsearch/showAward?AWD_ID=0926550, but this particular grant was to add new features, not just keep the lights on, which is the general pattern.


  7. msobel
    June 29, 2016 at 10:44 am

    You know, I started to donate because why not, but they want too much personal information including home phone and credit card info. Why should I trust Cornell with that? If they had a paypal, I’d donate.


    • June 29, 2016 at 12:04 pm

      totally agree… and what a state of the world we’ve reached when so many of us no longer feel comfortable giving out personal info to ANY organization across the Web (and it’s not paranoia if they really ARE out to get you 😉


    • voloch
      June 29, 2016 at 8:23 pm

      Meanwhile, Sci-Hub only takes donations in Bitcoin.


  8. Anshul Kogar
    June 29, 2016 at 12:07 pm

    I don’t think this is inherently a bad idea, but if they can survive without government funding, this may be a good thing. Government funding doesn’t come free, and one may not necessarily want that sort of intrusion on their mission of free and open science communication.


  9. Aaron Lercher
    June 29, 2016 at 12:11 pm

    One avenue by which NSF funding *might* hypothetically have been directed to arXiv is the public access requirement for publicly funded research: http://www.nsf.gov/news/special_reports/public_access/

    But from the documents linked from this webpage, there isn’t any reference to arXiv or pre-prints. The initial requirement was for peer reviewed publications to be deposited at a Department of Energy repository. But now there’s an NSF repository: http://par.nsf.gov/

    I was recently at the math and physics librarians annual meeting, where publishers talked about about making it easy to comply with federal open access mandates. Authors need to add funder data when submitting articles for publication. But you probably already know that.


  10. June 29, 2016 at 4:01 pm

    No you’re not wrong you’re absolutely right.


  11. anon
    June 29, 2016 at 6:56 pm

    I don’t think I agree. The arXiv is certainly worthy of NSF support, but why shouldn’t other entities be funding arXiv? I think a large part of arXiv’s budget is covered by academic institutions and I think these institutions should be supporting arXiv. Government funds should be used to support the progress of science and the arXiv certainly does this. But government funds shouldn’t be the only thing that supports scientific progress. Universities should also support it.


  12. June 29, 2016 at 8:20 pm

    I agree with you, and I would certainly recommend that the NSF fund it if they asked me, but I’d be much happier if the arXiv could build an endowment or acquire other long-term funding:

    1. As Nathan Dunfield points out, the NSF doesn’t like to fund straightforward continuing operations. This could push the arXiv to deploy new features to justify continued funding, some of which might be actively harmful. (For example, see Izabella Laba’s discussion of commenting at https://ilaba.wordpress.com/2016/04/10/arxiv-comments-and-quality-control/.) I would be unhappy to see the arXiv depending on an funding source that might push it to make changes for the sake of change, since there are a lot of obvious features that I would actively oppose.

    2. The NSF has a long history of abruptly declining to renew funding for large-scale projects people really value. This is not unreasonable, given that they explicitly do not claim to be a long-term funding source, but it can be incredibly disruptive. Fundraising is hard when potential donors know there’s already a big NSF grant, but suddenly you might have to ramp up your fundraising efforts dramatically if the NSF’s priorities shift. You end up with one very high stakes decision every five years or so.

    3. NSF funding introduces politics into the discussion. Are other countries paying their fair share, and how should this influence the NSF’s funding level (to avoid congressional complaints)? Will crackpots start complaining to congressional staff about how they are being shut out from government-funded services?

    These certainly aren’t arguments against accepting NSF funding, but I think they are a good reason not to count on it in the long run.


  13. RTG
    June 29, 2016 at 11:00 pm

    Do you have any indication that Cornell asked NSF for funding? I also agree with Henry Cohn about several reasons why it might be better not to seek government funding. But it’s also possible that NSF just hasn’t been asked to provide funding.


  14. Allan Greenleaf
    June 30, 2016 at 11:53 am

    The new issue of Nature has an article on this very subject.
    The fundraising does seem to be focused on providing resources for new features.
    Personally, I do think that the arXiv is a crucial piece of scientific infrastructure and the NSF should support it long term. I also don’t think that adding comments would be beneficial.



  15. June 30, 2016 at 10:38 pm

    NSF? Maybe. But why can’t a school with a $6 BILLION endowment support it?

    How is being asked to donate to arXiv different from the Berkeley ($4 B) solicitations or your Harvard ($38 B) solicitations?


  16. July 1, 2016 at 5:57 pm

    Maybe some of the professional societies should chip in? AMS, AWM, SIAM come to mind for math. I’d like to see my professional memberships doing something for the larger community (also my taxes, but let’s start small).


  17. mathematrucker
    July 2, 2016 at 10:41 am

    The maintenance argument sounds the strongest re lack of NSF funding but it may be a little smoke-screenish, for powerful academic publishing conglomerates surely wish arXiv nothing but an early death and would be doing whatever they can to bring one about.


  18. July 5, 2016 at 6:26 pm

    I sent them money. Took much less time than reading these comments.

    Of course, I am prejudiced. When last year I published on arxiv.org a review article, the review was picked up by the definitive review volume series.


    • mathematrucker
      July 16, 2016 at 11:15 am

      Perhaps it is time for arXiv to begin charging posters a modest fee for the privilege.


  1. No trackbacks yet.
Comments are closed.
%d bloggers like this: