Sunday, August 8, 2004

Joseph J. Esposito, The devil you don't know: The unexpected future of Open Access publishing, First Monday, August 2004. Abstract: With the advent of the Internet and online publishing, the notion has arisen that access to the world’s research publications could be made available to one and all for free, presumably by shifting the costs to other places in the value chain and disintermediating publishers, a circumstance called Open Access (OA) publishing. While there are many hopes embedded in this view (lower costs, wider access, etc.), it appears more likely that Open Access will come about not through a revolution in the world of legacy publishing, but through upstart media built with the innate characteristics of the Internet in mind. An unanticipated outcome of this situation will be that the overall cost of research publications will rise, though the costs will be borne by different players, primarily authors and their proxies. [Open Access News] Entertaining, but Exposito extrapolates widly from a remote starting point — blogs — rather than a much more appropriate one, over a decade of self-archiving in theoretical physics and computer science. The combination of self-archiving and search engines in these fields has made access to recent research much easier, and changed significantly how many of us conduct research. The complex new infrastructure that Esposito postulates as a foundation for new publishing businesses might be nice to have, but the one "must have" is up to date, comprehensive indexing. Google is doing the indexing pretty well, and the marginal cost of self-archiving is falling rapidly as most researchers already have Web sites based on cheap bandwidth and storage. Value-added services like arXiv, CiteSeer, or institution-supported refereed journals like JMLR have higher fixed costs, but still much lower than the costs imputed (without evidence) by Esposito to reviewing and editing. I agree with Esposito's skepticism about author-pays OA models, but I disagree that the alternative is some kind of fancy blog infrastructure rather than what most theoretical physicists and computer scientists have been doing with increasing success and at low cost for the last decade. His paper would be that much better if he had carefully analyzed the growth, costs, and successes of those efforts. 10:45:29 PM