Lost Information: Unpublished Conference Papers
F.LIEBESNY
It is the practice of many learned and professional societies or institutions to publish the titles (and often short abstracts) of papers which were presented at meetings, symposia, conferences, congresses or conventions. On those occasions a large number of papers, often invited, is read. If, however, the abstract given in the reports of these meetings arouses sufficient interest in a reader to make him want to peruse a fuller account of such a paper, his chances are roughly even of being able to do so. The exact figure of 48.5% of unpublished papers has been arrived at by investigating 383 contributions to four conferences of American societies.
The papers presented at the following conventions were analysed: the 1948 and 1949 Annual Meetings of the Optical Society of America (called Meetings A and B hereafter), the 1949 National Convention of the Institute of Radio Engineers (Meeting C), and the meeting of the American Physical Society at Oak Ridge, Tennessee, on March 16–18, 1950 (Meeting D).
For this investigation the dates of the conventions were deliberately chosen as rather early so as to allow sufficient time for the papers to be published in full. In each case, the indexes to the journals in which the programmes appeared were consulted, and also the relevant abstract journals: Chemical Abstracts and Science Abstracts A for Meetings A and B; Science Abstracts B and Wireless Engineer Abstracts for Meeting C, and Science Abstracts A and B and Chemical Abstracts for Meeting D. Papers similar to the conference material or by fewer or more authors have been regarded as “identical” papers and counted as such.
It can be seen from Table 1 that about half of the papers read could not be traced as having been published in full. Of the published papers, about one third appeared in periodicals other than that in which the abstracts, of the papers appeared. It therefore seems necessary to scan more than one periodical. Furthermore, the time-factor is of great importance both to abstractors and
FELIX LIEBESNY The British Aluminium Company Ltd., London.
TABLE 1 Analysis of the Papers
|
|
Presented |
Published by Dec., 1953 |
% |
Published in other periodicals |
% |
Publisheda from date of programme publication |
|||
1–12 months |
13–24 months |
25–36 months |
More than 37 |
|||||||
A |
33rd Annual Meeting of Opt. Soc. Amer. (Oct., 1948) |
76 |
34 |
45 |
7 |
20 |
22 (65) |
4 (12) |
6 (17) |
2 (6) |
B |
34th Annual Meeting of Opt. Soc. Amer. (Oct., 1949) |
62 |
36 |
58 |
6 |
17 |
27 (64) |
4 (11) |
3 (9) |
2 (6) |
C |
IRE National Convention (Mar., 1949) |
172 |
88 |
51 |
33 |
27 |
42 (48) |
41 (47) |
3 (3) |
2 (2) |
D |
Amer. Phys. Soc. (Mar., 1950) |
73 |
39 |
59 |
17 |
44 |
23 (59) |
14 (36) |
2 (5) |
– |
|
Total |
383 |
197 |
51.5 |
63 |
32 |
114 (58) |
63 (32) |
14 (7) |
6 (3) |
a Figures in parentheses denote percentages. |
readers. 90% of the papers eventually published were printed within two years of the appearance of the programme. Two of the 383 papers investigated were delayed by as much as four and a half years before they were printed in full. Sixteen of the published papers (8%) appeared in journals which are published by the firms employing the author(s) of the original paper, e.g., RCA Review, Bell System Technical Journal, General Electric Review.
To discourage any undue optimism in the potential reader, the list of summaries of the papers read at the Convention of the Institute of Radio Engineers is preceded by the following note: “No papers are available in preprint or reprint form, nor is there any assurance that any of them will be published in the Proceedings of the I.R.E., although it is hoped that many of them will appear in these pages in subsequent issues”(1).
A similar warning prefaces the list of papers presented at the 1950 Winter Meeting of the American Institute of Electrical Engineers: “These papers are not scheduled for publication in AIEE Transactions or AIEE Proceedings, nor are they available from the Institute”(2).
Specially invited papers do not fare any better, as shown in Table 2.
TABLE 2 Analysis of invited papers
|
Invited Papers |
||
Presented |
Published1 |
|
|
A |
12 |
5 |
(42) |
B |
10 |
6 |
(60) |
C |
11 |
5 |
(45) |
Total |
33 |
16 |
(48.5) |
1 Figures in parentheses denote percentages. |
Thus nearly half of the information presented at such meetings appears to be lost, unless some preprints are available, or unless the author is approached direct for copies of the manuscript.
This deplorable state of affairs is accentuated by the policies of the major abstracting journals; while Chemical Abstracts do present long lists of titles and authors, Science Abstracts do not publish abstracts “of papers presented orally at meetings, unless precis of substantial length are published in the normal way”(3).
The above analysis deals with American papers only, but there does not seem to be any great difference with affairs on the other side of the Atlantic. When papers are read at meetings of a British society, no abstracts are published in periodicals, but a small percentage of those papers (about 10%) is published in full in the proceedings of that society. Thus no appetites have been whetted, and therefore there are no disappointments. The Institute of Physics, however, publishes Conference Reports in its two journals. These reports are summarized proceedings of conferences organized by groups of the Institute; the reports are by one or two observers, but the actual authors of the papers are not indexed in the name indexes of the respective journals. The proceedings of some of the conferences are, however, published in full as Supplements to either journal (Journal of Scientific Instruments and British Journal of Applied Physics).
The importance of such “lost information” can perhaps best be illustrated by some examples:
At a meeting of the American Physical Society in 1926, J.C.Stearns read a paper about the effect of moisture on the viscosity of air (4). The six-line abstract published with those of the other papers presented at the meeting is the only source available for his findings of a decrease of viscosity with water vapour; this result is in direct contrast to the widely reported increase with moisture found by other scientists. For over thirty years it has been impossible to verify the accuracy of this statement. Since the author of this paper is no longer alive and the only known copy of the full paper is a typescript version in the Library of the University of Chicago, it seems unlikely that a serious enquirer will ever be able to make sure that there was not a two-letter misprint in the abstract, converting increase to decrease. Yet for thirty-two years Stearns’ controversial findings have been referred to and commented upon by writers who have apparently never seen the original paper—they all refer to the abstract!
Three scientists (one of them a Nobel Prize winner), M.McMillan, J.M. Peterson, and R.S.White, worked on the measurement of the ratio of negative π: positive π in the production of mesons; on theoretical grounds this ratio
was expected to be unity, but had not been measured before. They published a result in the form of a pre-meeting abstract (5), which gave the ratio as approximately 10:1. At the meeting itself, the ratio had shrunk to 7.5:1 and was reported as such (6). By the time the complete investigation was written up, this ratio was down to 1.7:1 (6), i.e., reasonably close to the theoretical value. This example shows the potential unreliability of a hurriedly prepared pre-publication abstract, which may perhaps explain the above-mentioned controversial findings by Stearns, and also the danger of relying too much on such abstracts which yet may be the only information available.
The third example relates to the 132nd meeting of the American Chemical Society held in New York on September 8–13, 1957. This Society undertakes the publication of a booklet entitled: “Abstracts of Papers.” The 1957 meeting sported 1509 papers by 2700 authors and co-authors who alone would make a good-sized meeting. These papers were condensed into 680 pages in the above-mentioned booklet. The title of one paper, on “Russian Patents” aroused considerable interest; the abstract was barely five lines long and gave only general information about the subject. A direct approach to the author was made, which did not produce any reply. Yet this problem of obtaining full copies of U.S.S.R. patents which are now being abstracted in Chemical Abstracts still exists, and guidance from the author of that paper was not forthcoming, so that complicated and expensive arrangements had to be made to procure copies.
In the absence of preprints, which in any case were not available for general distribution prior to publication, any hope of obtaining further information was clouded by the prefatory statement that “presentation does not guarantee publication, but the majority of these meeting papers will doubtless be published. Many will appear in the publications of the Society.”
This statistical survey shows that the above majority is only just a mathematical majority by 1.5%, and that one third are likely to appear elsewhere than in the publications of the Society.
The state of affairs revealed by this analysis is rather alarming, when positive indication of the existence of information has been given, only to find that all the information consists of a title or a brief summary.
The question arises: what happens to the other 48.5% of the papers? Many are withdrawn completely or revised before publication. But the still considerable remainder leads one to the assumption that either the authors have died an untimely death (such as Mr. Stearns), or have suffered a severe attack of amnesia or laziness which prevent them from putting their knowledge into print.
UNESCO is also concerned with that problem, particularly the long delay
which often intervenes between submission of papers for congresses and the printing. We live in an era of conferences and congresses, and yet allow so much of their results to be lost. Can we afford such a situation?
REFERENCES
1. Proc. Inst. Radio Engrs., 37 (2), 160 (1949).
2. Elec. Eng. (N.Y.), 69 (3), 250 (1950).
3. Sci. Abstr. A58, viii (1956).
4. Phys. Rev., 27, 116 (1926).
5. Science, 109, 438–439 (1949).
6. Ibid., 110, 579–583 (1949).