Academic papers want to be free


There is an interesting discussion on Panos Ipeirotis’s blog about open-access publishing, and the ACM. He argues that the ACM should grant open access to its digital library because ACM’s stated goal is “Advancing Computing as a Science and a Profession,” and that this would be an effective way to do so. I’ve always thought that the ACM digital library fees were unnecessary. Like Panos, I don’t know what ACM’s expenses are, but I do know that conferences are profit centers, and that too many non-profitable years can lead to trouble for the sponsoring SIGs. Given that

  • conferences make money from attendees,
  • all typesetting costs are borne by authors these days,
  • conferences are starting to abandon print proceedings (or to charge extra for them)

what is the rationale for charging for subsequent access to these papers? I could see having a one-to-two month blackout period during which proceedings are accessible only to ACM members, but after that, free access should be granted to all. After all, ACM already grants authors the right to distribute their papers from the authors’ sites (as long as it is not for profit), and ACM serves the metadata to published papers for free. So with respect to conference papers, all people get for their membership is that they don’t have to run a web search for the title of the paper to find the PDF version. Hardly a monopoly worth protecting.

Journal articles, of course, are somewhat different because their costs are not offset directly by conferences. Yet many of these costs are a legacy of print publishing; dispensing with that can drive down the costs significantly. A comment on Panos’s blog post suggests that the publisher provides expensive services as part of the publication process, including typesetting, marketing, and hosting of digital content. But typesetting costs can be passed onto the authors just as they are with conferences, the ACM brand is sufficiently strong these days that no additional marketing short of including publications in the DL should be necessary, and the cost of serving content can be amortized over the cost of serving all the conference papers.

Traditional print journals have employed copy editors to produce “pretty” pages from authors’ manuscripts. But that was in the days when these same editors used less sophisticated tools than the modern word precessing system available to authors. High quality conferences publish papers produced by their authors, with only cursory checks on the formatting of the document. What checks are performed manually can either be handled automatically (Contest: write a program that checks submitted PDF or LaTex for conformance with ACM style sheet), or manually by crowd-sourcing this to reviewers, just as they already crowd-source the review of content. And since the conference format is already well understood by authors, simply relaxing the page limit should be sufficient. After all, readers care more about what the articles say than about gutters and margins and fancy footnote separators.

It is not clear to me what costs ACM needs to bear to market itself at conferences. Most people in the field are well familiar with the brand, and with its publications, and the fact that many important conferences in the field are sponsored by ACM SIGs, and their proceedings disseminated through the ACM DL should be enough of a brand. Why spend more money on this?

Finally, the software and hardware related costs of hosting journal articles can be borne by the system for publishing conference papers. Even if the number of journal articles is comparable to the number of conference papers, notes, and other publication types, the additional costs of disk space to serve them are negligible. As of this writing, a terabyte of storage costs $150/month or less on Amazon. While this does not include bandwidth costs for serving the data, this is still reasonably cheap, no matter how bloated the PDF files get! ACM should also be able to negotiate a reasonable deal with a dedicated service provider that beats the Amazon rates. Of course it could save itself even more money by relying an a service such as to host the content, and only manage the metadata on its own servers. Another interesting possibility is to have the NSF pay for the cost of hosting content in exchange for providing open access.

In short, it seems that ACM’s policies of charging for content stem from seeing itself as a traditional publisher rather than as a society predicated on the technological innovation. It’s time to change this!


  1. Twitter Comment

    Posted “Academic Papers want to be free” [link to post]

    Posted using Chat Catcher

  2. I see interesting parallels between NIH / PubMed and NSF / ACM. While I acknowledge I don’t know the administrative relationships of both entities, I would be happy to see NSF follow in NIH’s footsteps and require all publications resulting from NSF-funded projects to be openly accessible via ACM.

    As it is right now, everybody puts in free work, and the publisher sells off the product. Authors submit for free, reviewers review for free, typesetting is done by the authors themselves. Long after the traditional costs of paper printing and moving documents around by mail have been discounted by the Internet, continuing to charge for what is essentially the product of a culture seems counter-intuitive to the ACM’s mission of advancing computing.

    You’d think that an organization of computing professionals/academicians would be at the forefront of utilizing the Internet to serve its mission, not lagging behind other disciplines who’ve already started the walk down that path.

  3. I’d love to add the following to the comment posted above, but I don’t see an Edit button:

    Not only should access be free, but redistribution rights as well. We should, as a community, encourage academicians to license their work under a Creative Commons license that allows others to take their reporting, analysis, and even raw data, and do new things with them.

    No more requesting permission from publishers if all you want to do is to comment on a graph that appeared in a paper published decades earlier. In my case, after I requested permission, the publisher told me they couldn’t find whether they had the copyright to the work in question, so they wouldn’t offer me the permission. The graph I wanted appeared in a journal article published by said publisher.

  4. I was the lone dissenter among the Computational Linguistics editorial board when it came to production and copy-editing. I thought the standard of the ACL conferences was fine, but everyone else who expressed an opinion thought copy-editing and production assistance were necessary to create a “professional” product.

    I think this stems from thinking of a journal as an “archival publication” and therefore justifying an investment in quality. I wasn’t arguing from cost so much as time.

    The concern wasn’t just about converting Word docs or cleaning up non-native speakers’ grammar, but that everyone makes mistakes and copy-editing cuts down on them.

    My own experience has been mixed. I had a great copy editor for my first book at Cambridge, but MIT was just annoying when it came to my second book (insisting I remove the scientific “we” and replace it with “I”).

  5. In a perfect world, we could run the experiment of publishing the same material in both ways, and seeing if the pay-for-content model survives for more than a few months. My opinion is that professional copy editing is a “nice to have” rather than a must-have. I think that in most cases relying on volunteer effort to police standards will be adequate, just as it is for conferences.

    From the authors’ perspective, open access should increase citation, which is what academics really want. So what would you rather have: no citation, or a citation from a paper with some grammatical mistakes?

  6. I would argue that professional copy editing is not even a “nice to have” :-)


    Managing metadata and a centralized distribution point seems to be the only benefits that a publisher brings now. Is it worth the cost?

  7. I would add that through its SIGs ACM also manages the review process for conferences. But that is a profit center, not a cost. So I think we’re all in agreement that the value provided is not proportional to the costs.

    Not clear how best to change this situation, though.

  8. Twitter Comment

    Academic Papers Want to be Free #OA #Open #Access, by @HCIR_GeneG [link to post]

    Posted using Chat Catcher

Comments are closed.