Self-help getting to the New System
By Andrew AppelThe “New System” of scholarly publication need not rely on organized journals or repositories. It can consist simply of professors (and students, and anyone else) putting papers on their home pages, and readers finding these professors and papers by Google. In fact this is the way that most computer scientists communicate with each other: by Google. Another way to see the future is to see how high school students do literature searches: they bypass the library and go straight to the search engine.
The computer scientists are there first, but other academics will not be far behind. To investigate this claim, I did a bit of research. I chose a university with similarly ranked Computer Science and Political Science departments: UCLA (each department is ranked about 15th in its own discipline). Each department has a similar demographic profile, with mean and median year of PhD about 1985).
95% of the Computer Scientists maintain their own web pages, most with a full academic CV (or equivalent data) on their web page, and most with a substantial number of their own published and unpublished papers in PDF. 80% of the Computer Scientists put their papers in PDF on their own web sites, regardless of whether they’re also published in journals or conferences.
Only 40% of the Political Scientists maintain their own web pages; about 42% of the Political Scientists make their full academic CV (or equivalent data) available on the web. Only 21% of the Political Scientists put the PDF’s of their published papers on their own web pages; but 29% of them put PDFs of their unpublished papers.
Is it a generational thing? I hypothesized that younger scientists in both disciplines would be more hip to the web, and to some extent that is the case:
Computer Scientists did not wait for organized repositories or open-publication journals to come along before making their papers available; they’ve been doing it for years. They’ve been doing it regardless of contractual language in the copyright-assignment forms they sign when publishing. In consequence, the journals and conferences have had to adapt: many of them (especially the majors professional-society publications) explicitly permit authors to put papers on their own web sites. The other journals adapt by pretending not to notice.
So, for Computer Scientists, the future is here already. Quite a few Political Scientists have moved into the future, and perhaps more will follow.
October 19th, 2007 at 12:54 pm
Here I will put some methodological footnotes. “Maintain their own web pages” means that I could find some web page other than the standard blurb typed in by their department secretary. “Papers in PDF” includes, in some cases, Postscript or HTML. “Papers on their own web site” means that at least several of their recent papers are available from the same web server that serves their home page.
I counted only regular tenure-track faculty at all ranks, excluding adjuncts, lecturers, emeritus, and assistant professors appointed so recently that their department web site does not list an office or phone number. There are 41 such Computer Scientists listed at cs.ucla.edu, and 48 such Political Scientists listed at polisci.ucla.edu.
October 19th, 2007 at 8:14 pm
I applaud the digging that Andrew has done and love having the stats available that he’s provided. I just wonder what percentage of the papers that both groups have put up are in violation of the copyright transfer agreements that many of those faculty signed with publishers in order to get those papers published. Mind you, I’m _glad_ they’re putting the content up, but am at the same time concerned that it be done legally so as not to provoke publishers into overtly pushing back against such efforts.
October 19th, 2007 at 11:34 pm
I do not think most publishers will try to push back. The big publishers in CS still attract a lot of traffic to their websites by providing reasonably well structured paper repositories (see ACM’s Digital Library in comparison with Google Scholar). Authors putting their own papers online won’t hurt the publishers, which know (just as Google knows) that there is more value in the links that connect papers more than there is in a paper by itself.
October 25th, 2007 at 10:27 am
One reason more political scientists may not be posting preprints of their papers is that the submission guidelines of the journals published by the American Political Science Association can be interpreted as excluding from consideration any manuscripts thus made openly available, or technically “published” under copyright law.
As for the potential damage to publishers, consider this: if everyone posts preprints or, even worse, postprints. the major aggregator of the journal literature in the humanities and social scientists, Project Muse, may cease to exist. And if Muse goes down, so too do many journals. At Penn State most of our journals could not now survive without the income derived from Muse.