College and Research Libraries Research Notes Co·nference Proceedings in Physics H. H. Barschall and W. Haeberli A growing fraction of science libraries' acquisitions, botlz monographs and journals, are proceedings of conferences. This research note examines the impact of these publications on an academic physics library. A particular concern is the publication of proceedings in journals. PROLIFERATION OF CONFERENCES The rapid growth of the scientific lit- erature and the increasing cost of journal subscriptions have resulted in a finan- cial crisis for most science libraries. Li- braries have had to cancel journal · subscriptions and may not have enough funds to maintain their collection of monographs. The proliferation of con- ference proceedings aggravates this cri- sis. In this note, conferences is used to include all types of meetings, such as congresses, workshops, symposia, semi- nars, and summer schools. Proceedings re- fers to the publication of the full text or a condensed version of papers presented at such meetings. The regular meetings held by the national professional societies in many countries are not of concern here. They do not generate proceedings al- though brief abstracts of the papers that were presented may be published in the transactions or newsletters of the scien- tific society. The discussion will be limited to the field of physics. There are, however, many areas on the border with other sciences, such as astronomy, chemistry, or engineering, in which the decision as to whether to include a conference had to be arbitrary. The listing of conference proceedings in Volume 25 of the Science, Engineer- ing, Medicine, Technology (SEMT) part of InterDok contains about 3,200 pub- lished proceedings. 1 Of these, about 220 are in the area of physics. · The March 1991 issue of Europlzysics News has an editorial entitled "Explo- sions." The editorial comments on the pro- liferation of conferences. "Explosions of one sort are thankfully over ... [this refers to the end of the Gulf War]. Our explosions are nearer home. The European Physical Society database now contains 570 meet- ings ... of interest to European physicists in 1991 ... a dramatic increase."2 This means eleven meetings are held per week. Since there are about a dozen subfields of physics, such as atomic physics, nuclear physics, particle physics, this amounts to one conference per week in each subfield. In the November 1989 Newsletter of the DhTision of Nuclear Physics of the American Physical Society (APS), Herman Feshbach, H. H. Barschall is John Bascom Professor Emeritus of Physics at the University of Wisconsin, Madison, Wisconsin 53706. W. Haeberli is Steen bock Professor of Physical Sciences at the University of Wisconsin, Madison, Wisconsin 53706. 563 564 College & Research Libraries who was then the chair of the Commis- sion on Nuclear Physics of the Inter- national Union of Pure and Applied Physics, writes that the commission "has been seriously concerned with the pro- liferation of meetings and workshops. Be- tweenAprill, 1989,and theendoftheyear there are at least 50 meetings, more than one per week scheduled." As another ex- ample, the June 1991 Newsletter of the Divi- sion of Particles and Fields of APS lists forty-three meetings during the period from the end of July 1991 to the end of 1991, approximately two meetings per week in particle physics. NATURE OF PROCEEDINGS One publisher produced one quarter of the physics conference proceedings listed in InterDok. Other publishers pro- duced 10 percent or less of the total. About 10 percent of the sample were proceedings of NATO workshops. The cost of the individual proceedings varied from a couple of free volumes to a maximum of $359. The highest cost per page was 50 cents, but for the typical proceedings the cost was around 15 cents per page, i.e., a typical volume cost $60 and contained 400 pages. Explosions of one sort are thankfully over ... [this refers to the end of the Gulf War]. Our explosions are nearer home. The European Physical Society database now contains 570 meetings ... of interest to European physicists in 1991 .•. a dramatic increase. In this sample of physics conference proceedings, 19 of the proceedings were listed as published in journals, either in regular volumes or in special supple- ments. These tended to be the most ex- pensive publications and included the proceedings that cost $359 with a cost of 38 cents per page. None of the proceed- ings in this sample reached the cost of $1.30 per page, which was reported in the Newsletter for Serials Pricing for the proceedings of a conference held in Po- land in 1989.3 November 1992 In most cases the date of publication of the proceedings was the year after the conference was held, although a number were published in the same year. In one instance, a five-year interval occurred between the date of the meeting and the publication date. In a fast-moving field the proceedings of conferences lose use- fulness rapidly with time, since an in- creasing fraction of the papers will have been published also in journals. Most proceedings are reproduced di- rectly from typed contributions by par- ticipants although some proceedings. published in journals are composed to the same standard as the rest of the jour- nal. Proceedings usually contain both in- vited and contributed papers. Some of the invited papers in proceedings consti- tute important scientific contributions in that they summarize results obtained by many scientists, and they are often pre- sented by the most knowledgeable workers in the field. Contributed papers are usually brief, two to four pages. These are often pre- liminary announcements of results that are published later in journals, although there are many cases in which the pre- liminary short version in the proceed- ings in fact appears later than the complete paper. Papers are usually not refereed in the usual sense, i.e., they are not sent for review to referees not as- sociated with the conference. All partic- ipants in a meeting are normally allowed to present at least one paper. Although there is often an official statement that all contributions are refereed, the organizer of a conference will rarely omit a con- tribution, unless it clearly does not fit the topic of the meeting. Consequently, the quality of the contributed papers ap- pearing in conference proceedings is often inferior to that of regular journals, even if the proceedings are published in the same journal. Most proceedings do not have a subject index. Many do not even have an author index, which makes their use inconvenient. Nonetheless many conference pro- ceedings are essential resources for the research physicists. Many conference proceedings give an invaluable over- view of the current activities in a given research area and provide a convenient source of references to published work and work in progress. Conference pro- ceedings are consulted by most users of the physics library. PROBLEMS FOR SCIENCE LIBRARIES The problem for the library is to select from the hundreds of proceedings pub- lished every year the ones needed by the users of the library. Meetings vary widely in the breadth of subject matter covered. In some broad subfields of physics, international con- ferences are held at regular, typically two- or three-year, intervals. Some sym- posia cover quite narrow areas of re- search. Proceedings of such highly specialized symposia would be of inter- est only to libraries at institutions in which there is an active research pro- gram in that specialty. When there is an active research program, the library may be able to get a copy of the proceedings from a participant. In a fast-moving field the proceed- ings of conferences lose usefulness rapidly with time, since an increasing fraction of the papers will have been published also in journals. Decisions on acquisition of proceed- ings are more difficult than those per- taining to single-author books, since the proceedings may contain a small number of important papers by distinguished authors among many unimportant con- tributions. Usually even physicists work- ing in the general area of the subject of a conference cannot evaluate the proceed- ings from the published announcement of their availability. Libraries receive proceedings either by subscription to a series of proceedings issued by a given publisher, or by purchase of individual volumes, or as part of a jour- nal subscription. A small number of pro- ceedings are distributed free by government or international agencies, such as the Inter- Conference Proceedings in Physics 565 national Atomic Energy Agency. In a typical physics library, proceedings now constitute a large fraction of all mono- graphs (in the authors' physics library the fraction is about one-third) and a substantial fraction of some journals. PROCEEDINGS PUBLISHED IN JOURNALS When the organizers of a conference seek a publisher, they prefer one who will make copies of the proceedings available to the participants at a low cost. The registration fee usually includes a free copy of the proceedings, but a high registration fee may reduce attendance. The publisher can agree to a low cost for participants if the cost of copies for sale can be increased, for example, by publication in a journal to which many libraries sub- scribe. This option makes the offer partic- ularly attractive to the organizers and participants, since it guarantees a wide distribution of the proceedings and pro- vides to the participants a journal refer- ence, which makes the contribution more easily accessible. Some physics journals, for example, the Physical Review and Physics Letters, publish no proceedings. Others, for ex- ample, the Journal of Physics and the Jour- nal of Applied Physics, rarely publish proceedings. Some journals, such as the Journal de Physique and Physica Scripta, have separate supplements for proceed- ings. Some journals, such as the Journal of Vacuum Science and Technology, consist largely of conference proceedings. Most journals that publish proceedings do not offer the option of not subscribing to the proceedings, even if they are published as separate supplements. Exceptions are Journal de Physique and, since 1988, Nu- clear Physics B. The authors examined a sample of nine physics journals that published several proceedings during 1989. Five are published under the auspices of scientific societies, and four by commercial publish- ers. As far as the authors know, it appears that subscribers to these journals do not have a choice whether to receive the pro- ceedings as part of the subscription. The subscription prices of the issues published 566 College & Research Libraries in 1989 in the journal were used. When the subscription price was given in a foreign currency, the January 1989 ex- change rate was used to convert to U.S. dollars. When several subscription prices were given, the one applicable to the U.S. was chosen. Some subscribers may have been able to subscribe at a lower price be- cause some publishers offer discounts to subscribers of several of their journals or to certain classes of subscribers. There is only a small overlap between the proceedings listed in Volume 25 of In- terDok and those published in 1989 in the journals examined by the authors, pre- sumably because of the delay between publication of proceedings and their list- ing in InterDok. Libraries may be forced to cancel subscriptions to expensive journals that contain conference proceedings, although they might be able to maintain the subscription, if they had the option of getting a subscription without conference proceedings at a lower cost. Among the journals examined, the fraction of pages that contained proceed- ings varied from 21% to 80%. When the annual subscription price was multi- plied by the fraction of pages containing proceedings, the cost assigned to pro- ceedings varied from $80 to $2,400, with an average of $670. Actually the cost per page of publishing proceedings is lower than that of the rest of the journal be- cause there are usually no editorial costs and often no composition costs for the publisher. Nevertheless for journals for which the subscription price is based on November 1992 the number of volumes published during a year the number of pages in volumes containing only proceedings is usually about the same as that for . volumes containing only directly sub- mitted articles. The cost per page varied by a factor of 21. The least expensive was the Journal of Vacuum Science and Technology at about 9 cents a page. The average cost per page of the journals published by commercial publishers was about four times that for journals published by society publish- ers, a result consistent with that found in a survey of chemistry journals.• In some cases the subscriber receives conference proceedings published in journals at very low cost, in other cases the sub- scription price is a heavy financial bur- den for a library. Libraries may be forced to cancel subscriptions to expensive journals that contain conference pro- ceedings, although they might be able to maintain the subscription, if they had the option of getting a subscription without conference proceedings at lower cost. Because of the budgetary problems they face, many libraries need to restrict their purchases of proceedings to those of high interest to their user community. When the proceedings are published in a journal, libraries are forced to buy pro- ceedings they would otherwise not have purchased. Subscribers should let the publishers know when this is a problem. Journals which publish conference pro- ceedings should publish them in supple- ments and make the purchase of these supplements optional. Librarians should call this problem to the attention of scien- tists who use their library, especially scientists who are involved in the or- ganization of conferences. REFERENCES 1. InterDok, Directory of Published Proceedings 25 (Harrison, N.Y.: InterDok, Sept. 1989/June 1990). · 2. P. G. Boswell, "Explosions," Europhysics News 22:46 (Petit-Laney 2, Switzerland: European Physical Society, 1991). 3. Marcia Tuttle, ed., Newsletter on Serials Pricing Issue, NS 2 Oune 3, 1991). 4. J. 0. Christensen, "Cost of Chemistry Journals to One Scientific Library," Serials Review vol.18, no.3:19 (1992).