Speaker: Washington Taylor (MIT)

Panel Discussion: Current Thinking Panel

There are a variety of comments that I'd like to make. The point I want to really focus on, however, is that, at least in the high energy community, we are rapidly shifting our work habits on-line, and I think there's a certain danger that if this happens too fast, the print journals will become functionally irrelevant before we have an adequate replacement for them on-line. I think this is something we really have to be careful about because it would not only be bad for the print journals, who would go out of business, but it would also be bad for the physicists, who would go through a period of some chaos while they tried to reconstitute a system. So I want to briefly review the situation we're presently in and where we want to go, and to give you a couple of visions of scenarios of how we might get there from here.

In the past, print journals filled many roles. Everyone has their favorite list of all the roles that print journals played and may or may not play currently. They played the role of providing people with the actual papers; that is, they were the primary information source. There is a slightly distinct role, which is the role of keeping people aware of what the current trends in the field are. There's the issue of validation which was touched on last night. I just want to make one comment in that context, relevant to this question of why people currently submit papers to journals. As a young, struggling physicist, I hear people comment that the number of published papers is not the most relevant thing in applying for jobs. But one of my main reasons for actually continuing to submit papers to refereed journals (aside from having them archived in a certain place) is that there's some fear that if I didn't do so, then when I applied for a job and there were zero refereed publications on my resume that might be looked at slightly askance by a search committee. So even though it may not be necessary to have 50 publications, I think having zero would be bad, at least in the current context.

Getting back to the roles of the print journal, there is the archival role. There's also the issue of maintaining standards which was mentioned last night. The fact that people are writing for a journal hopefully means that they will try to aspire to a certain standard which that journal expects.

So, we are currently in a situation in high energy physics where the e-print archives give us unfiltered access to pretty much everything anybody writes. For most of us doing work in that community, the e-print archives, along with some of the on-line search mechanisms such as the spires data base, are really the main, in fact the only, methods that we use for looking at papers that were written in the last couple of years. It's very rare that I will actually go to the library to find a paper that I know exists on-line, because it's so much easier to just click a few things and print the paper out, or to simply view it on my screen. Sometimes I even find myself printing out a paper and hoping that there were no substantial changes made to that paper before it was published.

At this point there are a variety of people across the network who are starting to develop secondary level structures on top of the network which Paul is running. Everything from private lists of papers to on-line journals run by a small group of people are beginning to appear. There is a group at BU developing something called the Interjournal which will ideally be a completely on-line, volunteer based journal. The on-line resources have completely taken over roles one and two, and because of that the focus of physicists is somewhat decoupled from what is in the print journals.

Where do we want to end up? Well, I think we all more or less agree that what we'd like to have is one single, large, integrated hypertext data base which fills all five of these roles which journals once filled. We don't want to have to go back and forth between totally disconnected parts of the network. There are various things which might play a role in this integrated hypertext network, but let me just mention three types of components. There would be something analogous to the current e-print archives; that is, the primary information data base which could be distributed or local. There may be a set of informally developed, not-for-profit resources that have been developed on the internet by private groups or individuals, and finally there could be more formal (for profit or not for profit) electronic journals run by groups like the APS which would have some revenue stream and which would actually pay editors, and perhaps even commission people to write good review articles.

I think it's possible for all three of these types of structures to play a role in the final system. But I think right now the direction we're headed in may lead to a slightly more chaotic situation. At least in high energy theory, usage patterns are rapidly weakening the print journals. People are providing usable secondary level literature on-line fairly rapidly. If we continue moving in this direction, it seems like the cancellation of journal subscriptions could cause the existing print journals to go out of business well before we have a system on-line which really fills all of the roles that we need.

We do not, at the current time, have something on-line which includes refereeing and/or commissions for review articles, such as has been suggested by Mike Peskin. And I think it's important that some structure like that exist before the existing journals are completely defunct. Of course, even if the most pessimistic scenario is realized, and we do have a brief period of anarchy, then eventually out of the rubble will arise some fairly structured journals which everyone in the community will have respect for, and which will play the role of the traditional print journals. But I think having this period of discontinuity would be unfortunate not only for the journals, but also for physicists. So I think it is really in the best interests of the physicists to work with the journals to try to develop a scenario in which all of these different things play a role.

In the more optimistic scenario, where the print journals are able to move on-line very rapidly, and bring their services up to date, they may be able to maintain some significance in the current context. But many things have to change. The journals must come on-line in a flexible electronic format, so that whatever method or protocol for viewing documents becomes popular they can rapidly change software. The current acceptance process takes far too long for anyone to be able to get useful and currently relevant information out of the print journals. They have to shorten the process to something more on the order of, I would think, three to six. Most importantly, I think that the journals really have to redefine their mission. One issue here is the question of whether they should consider themselves to be in the business of providing the information that is in the papers, or information about what is in the papers. Finally, I feel rather strongly that the on-line journals of the future, even the for-profit journals, should not expect copyright to be transferred to themselves when they accept a paper.

In short, I think it's feasible that journals can move on-line in a short time frame. However, they will have to develop sufficient value-added that it will worthwhile for people in the community to read them and use them, even if they simply contain pointers at an external data base of papers and additional information about interesting papers such as commissioned review articles.

(Question from the audience) And people will pay for them?

(Washington Taylor) Yes. I do think it's feasible that the journals can profitably move on-line in a short time frame, but I think that they have to very rapidly decide to make a decisive and dramatic change in their structure and that the physicists really need to help a great deal in this transformation process or it will not happen smoothly. Let me just make a few more points about why I think that.

There are many people in the community who I think would say: "Why do we need professional journals? Why won't it suffice to just have e-print archives and a variety of volunteer-based efforts providing secondary level information?" There are a couple of things I'd like to comment on there. First, in all my experience of large organizations, on the net in particular, there are very few instances in which a stable large structure has been developed on a purely volunteer basis, except in those situations where there is one individual whose somewhat superhuman efforts hold the whole thing together. (There are several examples of that; Paul's e-print archives are one, The free software foundation run by Richard Stallman is another.) I think it is unlikely that it will be possible to rapidly develop a completely volunteer-driven, stable, refereed data base on-line. I think it's really necessary to have some people on there with a revenue stream, even if it is a not-for-profit organization. Second, I think it is critical to have continuity in the sense of validity of the journals, which can only be achieved if the current print journals rapidly make their way on-line.

That is all I wished to say. Thank you.