Andrew Adams is a Professor of Information Ethics in the Graduate School of Business Administration and Deputy Director of the Centre for Business Information Ethics at Meiji University in Tokyo. This is the first of two Policy By the Numbers posts by Andrew on open access publishing.
Scholarly academic communications began with letters between early scientists. Groundbreaking work was sent to multiple people and as the early scientific and scholarly societies grew they started distributing their proceedings to their members, and making them available in libraries. As academia expanded and in particular as science and engineering grew to industrial scale enterprises in which the unknowing duplication of experiments was seen to be a waste of money (knowing duplication as an independent check on results remains a valid exercise though unfortunately one which is rarely given the importance it deserves) the need for wide dissemination of experimental results led to the emergence of scientific publishers of three types: scholarly/scientific societies with a specific disciplinary focus; university presses with a mission partly driven by dissemination of their own researchers' work and partly by general academic desire to improve communications; commercial publishers either founded specifically to meet an apparently unmet need in a specific area or general publishers using their expertise and economies of scale in typesetting, printing, binding and distribution to provide a valuable service to academia as a conduit for scientists to talk to each other.
The "gold standard" of academic publishing gradually emerged as peer review: consideration of the merits of a piece by other scholars/scientists with appropriate knowledge of the relevant field. Over the course of the twentieth century a large publishing industry emerged with a set of common practices followed by most publications and most publishers. (There are exception to these practices but there is no space to present them here.) Academics (and a small proportion of researchers in independent research institutions and commercial labs) constitute the main authors, reviewers, academic editors and academic editorial boards of the journals (and in Computer Science fully reviewed conference proceedings) in which most academic work is published. In the twentieth century, publishers provided typesetting, copyediting, printing and distribution services. They generally made a reasonable profit margin on these activities, whatever type of publisher they were. In return, authors (mostly academics) transferred the copyright in their work to the publishers, without charging any financial recompense. Likewise reviewers and editorial boards are generally unpaid. Editors may receive a small stipend and/or a small contribution to administrative support costs (though by no means always).
Since about 1990, the system has come under multiple pressures. Commercial publishers have merged or been taken over and a small number of large players, sometimes part of larger multimedia conglomerates and sometimes just large primarily academic presses (though usually also publishing at least textbooks and monographs as well, and often publishing other education/research-related materials). Scholarly societies and universities have come to expect profits from their publishing arms which will support their other activities. Commercial publishers have consistently increased their subscription prices well above both inflation and the funding available to most universities) while at the same time their costs should have been going down after an initial investment in digital reproduction technologies was covered. Meanwhile in a few fields such as Computer Science and Physics (most notably High Energy Physics) the new technology of the Internet was providing a parallel route for academics to disseminate their peer-reviewed articles, first using ftp sites, then the Web and finally through databases with web interfaces providing both machine- and human-accessible meta-data as well as the postscript, pdf, Word document, HTML etc format versions of the text.
In addition to this, a relatively small number of journals have been either founded with no subscription costs for electronic access, or moved to such a model. The funding to support these journals comes from a variety of sources including fees from authors who publish (occasionally fees from authors who submit for possible publication), academic society, university, research funder, government or charitable donation support and others.
While there remain some methodological difficulties and disputes, most of the studies carried out in the last ten years have shown that when articles are available without readers being restricted by payment (either institutional or individual subscription to the journal or by payment for the individual article) then that article is cited more often. Since the primary purpose of research writing is communication with other researchers and impact on the field, one crude measure of which is the number of times an article is cited, then it is clearly in the interests of research authors, their institutions and the funders of their research that their articles are made freely available. This is, in effect, a pre-requisite for a free market in scholarly/scientific ideas in which the best ideas rather than those easier to access, receive greatest attention.