Why, then, did the golden age pass away? Was it just because of all this slow and messy business of putting ink on to paper? I believe that the major reason why professionals came into the picture was because of the sheer quantity of scholarly material being published --that is, because of the growth of the scholarly community producing papers. A university library of a million volumes has to have a staff of professional librarians. And while a journal publishing 15 papers a year could be run on an "amateur" basis, one publishing 1500 papers a year cannot, regardless of the medium it is published in. The sheer administrative load of organizing the input, refereeing, copyediting, formatting, and distribution of that many documents (including the ones that get rejected, which generate work too) requires full-time staff. And since these people have to eat, they need a salary. Contrary to what some participants in discussions of electronic journals have alleged, it is this area of "first-copy cost" that is responsible for most of the cover price of a journal, not the paper, printing, binding and postage costs. Yes, a purely electronic journal is inherently somewhat cheaper than a paper one; but not a tiny fraction of the cost.
There is also the question of subsidy --an emotive word. I prefer to put it that the costs of running a high-quality scholarly communication system have to be covered from somewhere. Traditionally, one major route by which universities subsidized scholarly publication was by giving their libraries funds to buy journals. Controversy arose because commercial publishers, from the 1940s onwards and led by the unlamented Robert Maxwell, realized that there was scope for making lots of profit here. However, not-for-profit publishers --university presses and learned societies-- have a big presence in the scholarly publishing field and cannot be criticized for excessive profit-taking. The main cost is simply the pay of the people who do the work. Of course, these people can be (and in the case of the presently free electronic journals on the Internet, presumably are) subsidized in a different way, by the university that originates the journal paying for them. But for how long? And for how long will the network itself be entirely free of charge at the point of use to the academic community, anyway?
Another question --raised by Frank Quinn-- is how much of the work done by journal staff needs doing at all? Is copyediting necessary? The existing network journals are of necessity put out in straight ASCII text for the most part, while paper journals that are being experimentally offered in dual form (paper and electronic) acquire their page-image bitmaps by scanning the printed pages. The craft knowledge of typographers, graphic designers and even the despised copyeditors is not negligible. They all serve to turn a crude, possibly unreadable manuscript into a publishable paper. What an advance it was when Graphical User Interfaces like Windows replaced purely textual DOS screens --a great increase in user-friendliness. In the same way, a pleasingly designed and laid out printed page, written in correct and readable English, is more user-friendly than a typescript (however scientifically correct) in poor English. So even if no printed edition is published, I believe that the requirement for quality will mean that some copyediting and design work will need to be done by someone.
In case it is felt that I am a pure Luddite, let me finally say that I do believe that the networks have transformed informal academic communication beyond all recognition, and in particular have democratized the invisible college. Whereas in the past only those who actually received the personal letters or phone calls, or who could afford to attend the international conferences, were admitted to the invisible college, now anyone anywhere can join discussion lists or computer conferences or look at bulletin boards. This must be an improvement. And formal communication should certainly be quicker, and somewhat cheaper. The additional features available online, most notably the ability to append open peer commentary to papers, are very valuable too, and when the supernetworks come along we will be able to add multimedia features to "papers." But we should not kid ourselves that this will all happen at no cost and without specialist staff.
Fytton Rowland Research Fellow Department of Information & Library Studies Loughborough University of Technology J.F.Rowland@lut.ac.uk