>sh> One excellent way to lower >sh> start-up costs and to promote OA at the same time would be not to have >sh> a journal archive at all! Make it a policy that once peer-review is >performed >sh> and the final draft is accepted by your journal, it is the author who >sh> must self-archive it in his own institutional OAI-compliant archive. The >sh> journal merely provides the tag certifying that the article as been >sh> refereed, has met the standards, and been accepted for publication. > > Is the idea that these articles are not copy-edited or properly > typeset? Or is the idea that the journal insists that the author get > the copy-editing and typesetting done themselves? If the former, then > some articles are going to be very hard to read, and certainly not > esthetically pleasing. The latter system, on the other hand, seems > inefficient, because many scientists (or, at least, many economists) are > not knowledgable about language and typesetting standards. Also, the > certification burden on the journal increases: it has to check not only > the scientific worth of the paper, but also the English and typesetting.
Word-processing is today converging on just a few platforms: LaTex, OpenOffice or Word covers most papers submitted today. XML editors are not far away either. A journal can decide how much editing and copy-editing it still wants to do in-house, over and above refereeing, if any, and can include that in its review/certification process; the rest can and should be off-loaded onto the author today. Typesetting is obsolete, for online-only material; it is today a matter of metadata tagging and formatting, and there too, as much or as little can be included in the journal's review/certification process as the journal chooses, the rest being off-loaded onto the author. There will even soon be web-based reference-checking and even reference-linking software too, along with spell-checking, etc. This can all be off-loaded onto the author, with the journal merely specifying and checking the format required for approval and acceptance (i.e., the journal's "look.") For example, Psycoloquy http://psycprints.ecs.soton.ac.uk/ provides prospective authors with an XML template to which submissions must conform. The principal metadata fields, after all, are a small finite number for an article: Authorname fields Address fields Email fields URL fields Date(s) fields Title fields Headings Paragraphs Qutations (+ links) Figures Tables References (Authors, Title, Journalname, Date, Volume, Issue, Page-span) There is not much more to a journal article than that. And it makes far more sense, and is immeasurably more cost-efficient, to off-load on the author the task of making each paper conform to the journal's template than to do it for the author. (Or the journal can offer the service as an option, for a fee.) Stevan Harnad
