College and Research Libraries HENRY VOOS The Information Explosion; or, Redundancy Reduces the Charge! The information explosion has been blamed for the backlogs in li- · braries and information centers and for the inability of these centers and libraries to provide relevant information quickly. It has been used to justify the use of automation. An examination of the literature since the early fifties has shown that the library world has been aware of the problem, but has done very little about it. Some preliminary ex- periments reveal extensive dual publication between technical re- ports and journal articles; dual indexing and announcement have also contributed to what is more properly called a paper explosion. Methods of decreasing dual publication are suggested. p U BLISHING' S WALL STREET gold rush of a few years back was set off by an attempt to capitalize on the "information explosion" and a belief that expenditures on educa- tional materials would continue to climb ever onward and upward. Now with cut- backs in Health, Education and Welfare funds signaling that the education boom is leveling off, and the information explosion sputtering on a slow fuse, it appears that the money men are tucking in their tails a bit. 1 The literature, or the information ex- plosion (these are not identical), has been blamed for the current problems facing librarians and information scien- tists in handling and utilizing the infor- mation that is needed by scientists and engineers. Burton Adkinson and Charles Stearns quote a Systems Development Corporation report which states: Dr. Voos is Associate Professor at Rut- gers University, Graduate School of Li- brary Service. This paper was partially funded by the Research Council of Rut- gers University. Modern information technology has made it possible to place much of the accumu- lated knowledge of the human race with- in reach of a man's fingertips, so to speak. But the capacity of executives, scientists, and scholars to absorb information has not increased. Therefore, as the amount of available information grows, there is a par- allel need for a more precise capability to retrieve specific data in any area of inter- est.2 The academic world is now confronted with having not too little but too much informa- tion available. The student's problem used to be that he could not find enough ma- terial on his topic, whether it was the cri£is in the Congo or butterflies in Brazil. N dw -with the aid of up-to-date bibliographical tools, abstract services, and indexes-he is surfeited with sources. Librarians and archivists are also feeling the effects of the publication explosion. They are overwhelmed with book catalogs and reprint edition catalogs, to say nothing of the ever growing size of Publisher's W eekly. 3 Adkinson and Stearns further elaborate / 7 8 I College & Research Libraries • January 1971 TABLE 1 CoNTENT ANALYSIS: No. oF MENTIONs Journal 1964 1965 American Documentation 10.00 15.38 Special Libraries 7.14 24.32 College & Res. Lib. 9.43 6.89 Lib. Res. & Tech. Serv. 8.16 13.79 Average no. of times used 8.34 15.09 that the problem in library utilization of computers "arises out of a long history of financial starvation of library manage- ment. It has been intensified in recent years by the exponentially increasing amount of information that has to be obtained and filed, and the increasing library manpower and space that this entails."4 This problem has been used in numerous articles as a launching de- vice or excuse for whatever the author was writing about. To determine wheth- er the use of "information explosion" or words to that effect occurred with regu- larity in the library literature and that of information science, a content analy- sis was performed on four American li- brary journals for the years 1964 through 1969: American Documentation, Special Libraries, Library Resources & Tech- nical Servioes, and College & Research Libraries. (See Table 1.) Excluded from the analysis were articles which were bibliograp~ies, reports of committees, de- scriptions of libraries, and state-of-the- art on such items as reproduction meth- ods. Although it is difficult to assess the significance of this mention (it is signifi- cant at the .05 level using the Friedman two-way analysis of variance by ranks), it is obvious that 1nany professionals in the field are using the "information ex- plosion" in their articles for a variety of reasons. Librarians today are concerned about something that has been called a paper ex- plosion. This is not the best description of a phenomenon that has been going on for thirty or forty years-perhaps "creeping 1966 1967 1968 1969 Average 15.78 3.84 16.36 20.00 13.56 13.04 5.55 22.22 20.69 15.49 4.44 4.17 7.01 9.61 6.93 6.82 4.55 8.57 6.06 7.99 10.02 4.53 13.54 14.09 neoplasm" would be better. Since it has been going on for so long, it must have been noticed, and something must have been done about it. In the mid-thirties, some people were so alarmed about the paper explosion that they proposed a two- year moratorium on all research and de- velopment so they could "catch up."5 There are many references concern- ing the extent of the explosion, some of . which are in the bibliography at the end of this article. However, not everyone has agreed that the explosion is as large as has been stated. During the past 10 or 15 years innumer- able papers were written which empha- sized in an introductory sentence the im- portant and far-reaching consequences of the so-called information explosion. The use of this phrase seems to imply that the tremendous increase in the number of pub- lications automatically produces a corre- sponding growth in "information" or knowl- edge. But is this true?6 It is possible that the so-called information explosion is not as serious as has been thought. 7 . . . despite frequent references in the press and elsewhere concerning an expo- nential growth in technical literature, there does not seem to be any evidence to this effect.s Nlost of the library and information science literature tends to concentrate on handling the flood rather than dam- ming (damning) it. One factor that pro- vides evidence that the explosion is real- ly not a flood of new information is the extent to which much of what is written is redundant. This paper will try to ana- lyze the redundancy. The problems seemingly created by the explosion might thus be more deeply analyzed and more fruitfully resolved by studying the input rather than bemoaning the output. The factor of redundancy can be looked at from at least two sides: the first is the multiple indexing of the same material in different indexes and in dif- ferent ways; and the second is the re- publication of the same or similar ma- terial in a variety of media. The more time a literature searcher spends culling through indexes for ma- terial that might be relevant, but which has already been scanned by him with- out his being aware of it, the more time is wasted, and the more it costs to search the literature. J. J. Ebersole in looking at this problem states_: As more indexing and abstracting services are created, as the volume of knowledge increases, and as it becomes increasingly interdisciplinary in character there will be an increasing problem of overlap among these services. The major manifestations of this overlap consist of two or more organi- zations indexing and abstracting the same document. He then goes on to analyze some of the journals being indexed and abstracted and states that "an analysis of 17,000 journals covered by 11 of the 18 profes- sion-oriented services in 1961 showed a 50 percent overlap in journal coverage among these 11 services alone." He con- tinues: Although these costs are dramatic, they are probably exceeded both absolutely and relatively by the duplication existing in the coverage of technical reports. This dupli- cation cosJ extends not only through the community of indexing and abstracting services, but thJough the vast complex of libraries and information centers operating in hundreds of companies and government agencies. 9 This view is also echoed by Alfred A. Beltran: The Information Explosion I 9 The open literature (articles and technical papers) listed in TAB is also listed in standard commercial indexes available in all technical information centers. Inclusion of this material in TAB seems to serve no useful purpose while adding to the dupli- cation already prevalent in commercial in- dexes. This merely increases the time re- quired by a literature scientist to complete a search as he will encounter the same item several times. 10 .. A prime attack was made on this by Martyn when he showed that, in certain fields, only an additional percentage point of information is gained by scan- ning more · than the prime abstracting journal. A reinforcement of Beltran's point is that the Technical Abstract Bul- letin (TAB) and U.S. Government Re- search and Development Reports (US- GRDR) both index and cite items by corporate authors, while most commer- cial indexes use either author or title, but not corporate authors. , Although Beltran claims that the open literature was only announced in TAB beginning with Bulletin no. 65-7 ( 1 April 1965) and although he found an average of 6.47 percent of "open lit- erature unavailable" reports, the author presents his findings based on a random sample from the TAB and USGRDR for 1961 and 1963-68.11 ( See Table 2. ) TABLE 2 REPRINTS ANNOUNCED AS REPORTS Year No. in Sample No. of Reprints % 1961 1222 152 12 1963 1118 79 7 1964 2852 512 18 1965 630 54 9 1967 71 21 30 1968 53 15 28 It is obvious from the ranges that a more definitive study with larger sample size is required, but the extent of this re- dundant indexing is certainly evident. Many of the announced research re- ports are later published in some form 10 I College & Research Libraries • January 1971 as articles. This adds the second di- mension of redundancy to the explosion. Reprinting, or dual publication, has at times been advocated in order to reach multiple audiences. F. W. Hunt explores the reasons for dual publication: publish or perish, immediacy vs permanence, more depth in indexing of the report vs the journal, assertion vs authenticity, un- availability of space in journals, and the philosophy that a report is not a publi- cation.12 To determine whether the redundan- cy factor could be verified in another way, a random sample ( 1 percent) of citations was taken from Chemical Ab- stracts for the years 1955 and 1960. Ex- cluded from the sample were patents and non-American articles. The Defense Documentation Center was then request- ed to run a search on the authors of the articles selected from Chemical Ab- stracts. The search showed that in 1955 ' 11 percent of the articles were avail- able in report form, and in 1960, 27 per- cent of the articles were available in re- port form. It is also interesting to note that in 1955 about 11 percent were theses or extracts from dissertations, and in 1960 only 6 percent fell within this cate- gory. Another form of redundancy is there- publication in book form of articles pre- viously available as either journal ar- ticles or technical reports. Examples of such republication are the many volumes of ''Readings'' in various fields. A cur- sory look at the 1970 Paperbound Books in Print ( v.14, no. 10) shows that about .3 percent of the 80,000 titles announced is in this category, and about .2 percent of the 245,000 titles announced as hard- bound in the 1967 Books in Print fall within this category. A further example of redundant pub- lishing as a contributory factor to the literature explosion is the republication of the writings of a single author. In the field of library and information science, books by Taube and Shera serve as ex- amples.13 The justification for republish- ing, as mentioned before, is the exten- sion of the audience. This justification has been used, as well as some of the others previously mentioned, for pub- lishing abbreviated versions of disserta- tions in journals. Redundancy as a factor has been brought to the attention of the informa- tion and publication fields before. The National Science Foundation reported in 1952 that in a sample of 84 reports in various fields, 56 percent had already appeared in whole or in part, 13 percent were in the process of publication, and 2 percent would be published. In addi- tion, 6 percent were easily obtainable~ and 21 percent were unsuitable for pub- lication. Of the 47 reports which had been published in whole or in part, 71 publi- cations have resulted.14 Note that a fac- tor of 1.5 seems to exist in terms of multiple publication. In 1956, as a project of the National Science Foundation, the Library of Con- gress did a survey using a sample from the Technical Information Pilot of 1952~ to determine the extent of republica- tion or dual publication of technical re- ports. This report showed that 48 per- cent of the information in the technical reports containing publishable informa- tion had been published in the open lit- erature in its entirety (this is on(y' 21 percent of all the reports published ) ~ another 14 percent of the reports had published more than 75 percent of their information in the open literature, an additional 8 percent had published more than half in open literature, and 3 per- cent had published more than 25 per- cent in the open literature. This left only 23 percent of unclassified, unlimited in- formation that had not also reached the open literature. The reasons for the non- publication of this 23 percent ranged from the fact that the material in the re- port had become obsolete, through rea- sons such as <Sp::; .... ~ op::; ~op::; ~ o...,. ~~ ~~u~ .... u~ c-10 0 Department of Agriculture 48 15 18 15 Commerce: National Bureau of Standards 80 80 Coast and Geodetic Survey 80 70 10 Patent Office 50 25 25 Weather Bureau 60 30 25 5 Public Roads 90 60 25 5 Department of Defense: Office of the Secretary of Defense 15 4 7 4 Department of the Army 40 33 5 2 Department of the Air Force 30 10 10 10 Department of the Navy N.A. N.A. N.A. N.A. Department of Health, Education, and Welfare: Food and Drug Administration 100 75 20 5 Office of Education 11 10 1 St. Elizabeth's Hospital 75 75 National Center for Health Statistics 10 5 5 National Institutes of Health 98 40 52 6 Bureau of Medical Services 75 75 Bureau of State Services (Environmental Health) 86 40 40 6 Division of Accident Prevention 90 80 10 Communicable Disease Center 75 75 Division of Dental Public Health and Resources 80 50 20 10 Division of Hospital and Medical Facilities 95 40 40 15 Division of Nursing 90 90 Social Security Administration 95 75 20 Division of Vocational Rehabilitation 55 46 6 3 Department of Interior 5Q-90 1G-75 25-90 5Q-90 Department of Labor Department of State: Agency for International 8Q-90 8Q-90 Development 80 80 Department of the Treasury: Coast Guard 100 100 Post Office Department (1) (1) Arms Control and Disarmament Agency (2) Federal Aviation Agency 5 5 Federal Communications Commission 100 25 25 50 Federal Housing Administration 90 50 25 15 National Aeronautics and Space Administration 25 6 12 7 National Science Foundation 95 65 20 10 Office of Emergency Planning 10 10 Smithsonian Institution 100 10 80 10 Tennessee Valley Authority 50 28 12 10 Veterans' Administration 45 41 4 1 Under 10. 2 Less than 5. N.A.-Not available. 12 I College & Research Libraries • January 1971 placed great emphasis on journal publi- cation, and that many of the reports did not contain anything publishable.15 In 1955, Alan T. Waterman stated that in a sample of 100 reports, 25 in each of four fields, the authors were asked whether the information in the re- ports had been or would be published in the open literature. 56 percent stated that they had also been published in whole or in part in the open literature, 16 percent were in the process of pub- lication, and 6 percent had been an- nounced in journals.16 The Elliot report provides some data on time lag and extent of publication of the technical report and the same item published in a periodical.17 Table 3 is copied from the report. A variety of recommendations have been made to reduce redundancy in the literature. 1. To reduce the time that a literature searcher must spend scanning the in- dexes and abstracts, the Clearinghouse for Federal Scientific and Technical In- formation should cease to announce re- prints, even if they are the products of government and government-contracted research and development. Since the journal articles are generally abstracted in the standard abstracting and indexing journals, much time and money could be saved in announcing, acquiring, cata- loging, and searching if they were not announced in -USGRDR. 2. Government contract requirements should be modified so that fewer reports need be issued. 3. The system of refereeing should be tightened. 4. Preprints should be a standard dis- tribution mechanism for many items that later find their way into print (which should not). Moore's survey revealed that 40 percent of the members distrib- uted preprints to as many as 200 col- leagues on an average of nine months to one year before publication.18 5. Make strenuous efforts to eliminate the "publish or perish" syndrome in uni- versities. The very bulk of scientific publications is itself delusive. It is of very unequal value; a large proportion of it, possibly as much as three-quarters, does not deserve to be published at all, and is only published for economic considerations which have noth- ing to do with the real interests of science. The position of every scientific worker has been made to depend far too much on the bulk rather than the quality of his scien- tific publications. Publication is often pre- mature and dictated by the need of es- tablishing priorities ... ,19 There is sufficient evidence, even with limited sampling, to show that redun- dancy contributes heavily to the litera- ture explosion. It also provides addition- al evidence that the explosion is not one of information, but rather one of paper. Additional research, including cost ben- efit studies on nonpublication might help to solve part of the problem. Rather than finding ways of coping with the explosion, more effort should be devoted to investigating its real causes and its ex- tent in other than gross figures. REFERENCES 1. Richard Lingeman, . "American Note- book," New York Times Book Review (Mar. 1, 1970), p.24. 2. B. A. Adkinson and C. M. Steams, "Libraries and Machines-a Review," American Documentation 18: 121 (1967) 0 ' 3. K. M. Glazier, "Paper Pollution," Wil- son Library Bulletin 44 :'856 ( 1970) . 4. Adkinson and Stearns, "Libraries and Machines," p.123. 5. M. E. Freeman, "The Science Infor- mation Exchange As a Source of In- formation," Special Libraries 59:86 ( 1968) 0 ' 6. Gertrude London, "The Publication Inflation," American Documentation 19:137 (1968). 7. L. H. Mantell, "On Laws of Special Abilities and the Production of Scien- tific Literature," American Documen- tatum 17: 8 ( 1966) . 8. Ibid., p.13-14. 9. J. L. Ebersole, "An Operating Model of a National Information System," American Documentation 17:33 ( 1966). 10. A. A. Belb·an, "Evaluation of DDC TAB from Various Points of View," in Regional Workshop on the Report Lit- erature (Albuquerque, N .M., 1965), Proceedings . . . Western Periodicals, 1965. p.84. 11. Ibid., p.83. 12. East Coast Navy Interlaboratory Com- mittee on Editing and Publishing, Some Views on Dual Publication, by F. H. Hunt. Oct. 1, 1968. (Monograph no. 5, AD676523), p.2-3. 13. M. Taube, Studies in Coordinate In- dexing, v.1-6 (Washington, D.C.: Documentation Inc., 1953-1965). J. Shera, Documentation & the Organi- zation of Knowledge (Archon, 1966). --,Libraries & the Organization of Knowledge (Archon, 1965). 14. National Science Foundation, Prelim- inary Study of Unclassified Govern- ment Technical Reports (July 31, 1952), 6p. 15. Library of Congress, Final Report on Study of Publications Stemming from Defense-Related Technical Reports by D. E. Grey and S. Rosen borg (Dec. 1956), 8p. 16. Alan T. Waterman, "The Future of Report Literature," in B. M. Fry and J. J. Kortendick, The Production and Use of Technical Reports (Washing- ton, D.C.: Catholic University, 1955), p.6-7. 17. U.S. Congress, House, 88th Congress, 2d Session, Select Committee on Gov- ernment Research, Documentation and Dissemination of Research and Devel- opment Results, Study no. IV (Wash- ington, D.C.: Govt. Print. Off., 1964), p.37. 18. C. A. Moore, "Preprints: An Old In- formation Device with New Outlooks," ] ournal of Chemical Documentation 5:126-8 ( 1965). 19. J. D. Bernal, The Social Function of The Information Explosion I 13 Science (London: Routledge, 1939), p.118. BIBLIOGRAPHY The bibliography which follows is not exhaustive, but rather, is indicative of the quantity of literature pertaining to the "in- formation explosion" through the years. It has appeared in a normal distribution fre- quency pattern, with most articles being published between 1965 and 1969. (This paper is my contribution to that same pa- per explosion.) Adkinson, W. A. and Stearns, C. M. ~~Li­ braries and Machines-A Review." Ameri- can Documentation 18:121-4 (1967). Bar-Hillel, Yehoshua. "Is Information Re- trieval Approaching a Crisis?," in his Lan- guage and Information, p.365-72. Reading, Mass.: Addison-Wesley, 1964. (Also in Amer. Doc. 14:95-8 (1963).) Beltran, A. A. "Evaluation of DDC TAB from Various Points of View," in Regional Workshop on the Report Literature, Al- buquerque, N.M., Proceedings, p.78-85. Western Periodicals, 1965. Bernal, J. D. The Social Function of Sci- ence. London: Routledge, 1939. --. "The Supply of Information to the Scientist: Some Problems of the Present Day." journal of Documentation 13:195- 208 ( 1957). --. "Transmission of Scientific Informa- tion: A User's Analysis," in International Conference on Scientific Information, Wash- ington, D.C., Proceedings, v.1, p.77-95. Washington, D.C.: National Academy of Sciences, Nat'l Research Council, 1959. Bourne, C. P. Methods of Information Han- dling. N.Y., Wiley, 1963. p.1-10. --."The World's Technical Journal Lit- erature: An Estimate of Volume, Origin, Language, Field, Indexing and Abstract- ing." American Documentation 13:159-68 (1962). Brown, L. "Opening Session Address," in International Conference on Scientific In- formation, Wash., D.C. Proceedings, v.1, p.3-8. Washington, D.C.: National Acade- my of Sciences, Nat'l Res. Council, 1959. Conference on Problems of Centralized 141 College & Research Libraries • January 1971 Documentation. Proceedings. . .. June 1949. (ATI 73852) p.37. East Coast Navy Interlaboratory Committee on Editing and Publishing. Some Views on Dual Publication, by F. H. Hunt. 1 Oct. 1968. (Monograph no. 5; AD676523) 5p. Ebersole, J. L. "An Operating Model of a National Information System." American Documentation 17:33-40 ( 1966) . Federal Council for Science and Technol- ogy. The Role of the Technical Report in Scientific and Technological Communica- tion. Dec. 1968. (PB180944) 108p. Freeman, M. E. "The Science Information Exchange as a Source of Information.'' Special Libraries 59: 86 ( 1968) . Glazier, K. M. "Paper Pollution." Wilson Library Bulletin 44:856-7 ( 1970). Goudsmit, S. A. "Is the Literature Worth Retrieving?" :Physics Today 19:52-5 (1966). Gray, D. E. and S. Rosenborg. "Do Tech- nical Reports Become Published Papers?" Physics Today 10(6):18-21 (June 1957). Also published as: Library of Congress. Final Report on Study of Publication Stem- ming from Defense-Related Technical Re- ports, Dec. 1956. Heumann, Karl. "Report Utilization Through Intelligence Sections," in B. M. Fry, and J. J. Kortendick, eds., The Pro- duction and Use of Technical Reports, p.111-15. Washington, D.C.: Catholic Uni- versity, 1955. "The Information Deluge." The Johns Hopkins Magazine 17 ( 4): 1-33 (Fall 1967). Johns Hopkins University. Reports of Stud- ies of the Publication Fate of Material Presented at National Meetings, June 1969. (TN 10, TN 12. PB185 469). Kent, A. "Resolution of the Literature Cri- sis in the Decade 1961-1970." Research Management 5:49-58 (1962). Lingeman, Richard. "American Notebook." New York Times Book Review (Mar. 1, 1970), p.24. London, Gertrude. "The Publication In- flation." American Documentation 19:137- 41 (1968). Mantell, L. H. "On Laws of Special Abil- ities and the Production of Scientific Lit- erature." American Documentation 17:8- 16 (1966). Mellanby, K. A. "A Damp Squib." New Scientist 33:626-72 ( 1967). Miller, E. E. "The Genesis and Character- istics of Report Literature." American Doc- umentation 3:91-4 ( 1952). Moore, C. A. "Preprints: An Old Informa- tion Device with New Outlooks." Journal of Chemical DocUmentation 5: 126-8 ( 1965). "The National Information Problem." SDC Magazine 9:1-15 ( 1966). National Science Foundation. Preliminary Study of Unclassified Government Tech- nical Reports, 31 July 1952. 6p. ·overhage, C. F. J. ccScience Libraries: Prospects and Problems." Science 155: 802-6 (1967). President's Science Advisory Committee. Science, Government, and Information. Washington, D.C.: Govt. Print. Off., 1963. p.19-20. Price, D. J. deS. "A Calculus of Science." International Science and Technology, no. 15, p.37-42 (March 1963). --. "The Ethics of Scientific Publica- tion." Science 144:655-7 (1964). --. Little Science, Big Science. New York: Columbia, 1963. Scrivenor, Thomas. c'The Growth of Sci- entific Literature," in ASLIB 38th Annual Conference, 1964. Looking Forward in Documentation. London, 1964. p.3-9. U.S., Congress. House. Select Committee on Government Operations, 88th, 2d sess. Documentation and Dissemination of Re- search and Development Results. Washing- ton, D.C.: Govt. Print. Off., 1964. (Study no. 4) 48p. Urquhart, D. J. "Rising Tide of Paper." Ad- vancement of Science 21:279-85 (1964). Wall, Eugene. "A Rationale for Attaching Information Problems." American Docu- mentation 18:97-103 (1967). Waterman, A. T. "The Future of Report Literature," in B. M. Fry and J. J. Korten- dick, eds., The Production and Use of Technical Reports, p.3-8. Washington, D. C.: Catholic University, 1955. Wilson, J. H. "As We May Have Thought," in American Documentation Institute, Pro- ceedings, v.3 p.117-122 Adrianne, 1966.