We Are Publishing Too Many Conference Proceedings

As an editor, I find that book buyers dislike conference volumes more than any other category of book, and yet the partnership of scientists and publishers persists in producing them. How are we to explain the fact that scientists can be both disdainful and enthusiastic about these ephemeral publications? What is it that drives the community to support proceedings volumes at all? Can we afford to do without them? Last August, I attended the Twentieth General Assembly of the International Astronomical Union (IAU) in Baltimore, which lasted 11 days. Nearly 2,000 astronomers registered and participated in half a dozen parallel sessions. In addition, there were 10 or so satellite symposia and discussion meetings before and after the General Assembly. I estimate that the publications of all these meetings will require about 8,000 pages and the resources of at least four publishers. The total cost to any library that takes everything will not leave much change from $1,000. Worldwide we are talking about $500,000 of library funds to put the printed versions of this one scientific assembly on the shelves. There is nothing special about astronomers in all of this; exactly the same sorts of meetings and their associated publications are found in chemistry, computer science, and molecular biology. What concerns me is that too much material is being produced indiscriminately, and much of it is not subjected to any kind of reviewing process. Conferences themselves obviously play an essential role in the exchange of new ideas in science. And funding agencies and meeting organizers - along with at least some of those who read papers - have an interest in having the proceedings published. The NATO Scientific Affairs Division, for example, has for many years arranged the publication of summer schools and advanced study institutes. The Royal Society puts its discussion meetings on record in Philosophical Transactions. Publication of IAU symposia is mandatory for the organizers. In all of these programs, there is much that is good. The danger lies in the example that they set. What concerns me is a trickle-down effect whereby even quite small conferences are thought to merit the full apparatus of international publishing. Increasingly I encounter meeting organizers who plead to have the papers from "their" gatherings published. This is particularly the case if an annual cycle of meetings is envisaged or a new institute wants to put itself on the map. Organizers apparently believe that funding will be easier in the future if a concrete result, in the form of a whacking great book, can be sent to the sponsors. I suspect that personal motives and ambition also play a role sometimes, with researchers being attracted by the prospect of their name on the book spine. Thus does the pressure to publish conferences continue. Surely, we all know of cases too where younger researchers are only paid their expenses to attend a meeting if they read a paper. This requirement puts a serious and insidious pressure on organizers, who inevitably find themselves faced with some uninvited papers of marginal value. The scientific poster, for example, is an excellent way of presenting data for discussion at a meeting, but it seldom translates into a sensible printed paper. Unfortunately, it sometimes happens that the poster papers themselves get printed because younger researchers might otherwise be inhibited from getting their expenses. Duplication is also a problem. Active scientists in regular demand for keynote lectures tell me that the same paper often may appear in several proceedings - and that they resent the chore of producing the written version several times over for different publishers. Sometimes I am simply informed that the written version will be identical to that in an already published book; I usually find myself in complete sympathy with the reasons for this. I want to suggest that a solution to these problems is for fewer conferences to be published in book form. Those that are produced should be of shorter length, and therefore available to all at lower cost. Careful editing and selection of papers should be the rule, with the aim of avoiding duplication. The publicity function of a conference can be accomplished by the publication of highlights in the news section of journals such as Science and Nature. Alternatively, a report can be submitted to a relevant newsletter; Episodes, for example, publishes many useful summaries of earth science meetings. Perhaps those responsible for funds could be persuaded to accept reports of this sort as an adequate account for the public record, at least in the case of smaller meetings. When conferences are published, there needs to be more rigorous control of length. It is very difficult to sell worthwhile numbers of conference volumes longer than 500 pages, and an upper limit of 300 pages is frequently desirable. This message is being heard: The IAU has now established a limit of 500 pages for its symposia, which routinely had been achieving from 600 to 750 pages. Publishers are getting tougher too, and I no longer submit tenders for conference volumes without specifying a fairly severe upper limit. Unfortunately volume editors may respond to these length restrictions by allocating less space all round. As a result, there may be just as many papers, but they will scarcely make any sense because they are so short. I have just received for review, for example, a book on radio astronomy, based on a symposium. It has 155 two-page papers, some of them reporting observations five years old. There are 13 pages of introduction containing photographs of those who attended and their names, and 25 pages of discussion. Informative review articles, of the type that a graduate student would find helpful, occupy less than 250 pages of the book. The list price of the library edition is $126. I have to admit that in the past I have supported the publication of similar books. But I have now stopped doing so because they do not sell well enough, presumably because scientists do not find them useful. I believe the best type of conference book is one that restricts its scope to invited papers, preferably reviews. For those articles to be really useful, they must not be too short; they should be regarded almost as chapters in a conventional book. Perhaps the aim should be to produce a survey of a research field that new graduate students would find affordable and accessible. Certainly in those few cases where I have persuaded conference editors to take that position, sales have been twice the normal expectation. Would those who now insist on presenting posters and contributed papers accept this line? If they value professional standards, they ought to find this stance reasonable. I believe that the correct place to announce new results is in the primary journals. Too often, untested or unrefereed "preliminary" results presented at a conference acquire an unwarranted patina of respectability when they are slipped into a conference volume. I suggest that those who organize conventions must address the problem of how much genuinely merits permanent publication. No shame should be attached to nonpublication. Who knows, conference participants may be less cautious and therefore more stimulating in their presentations. I also suggest that those who provide grants and subsidies should reexamine their policies if publication is an invariable requirement for the release of funds. For my part, I am resolved to be more selective. There are plenty of good conferences I shall offer to publish, but I shall place more stringent conditions on what is acceptable. Science will be better served by fewer of these books, produced to higher editorial standards. Simon Mitton is editorial director of science, technical, and medical publishing at Cambridge University Press.

We Are Publishing Too Many Conference Proceedings

Interested in reading more?

Become a Member of