From theWorld Brain to the Informatorium

Eugene Garfield

3501MarketStreet,Philadelphia,PA19104,USA
Tel.:+12152432205; Fax:+1 215 3871266;
E-mail:garfield@codex.cis.upenn.edu

From June, 1962, until February, 1967, I published a half dozen short editorials in Current Contents®under the rubric "The Informatorium". Realizing the term had some negative connotations, the column was renamed "The Information Scientist". That continued until March, 1969, when we adopted the rubric "Current Comments", since the scope of topics covered were broader than information science.

"The Informatorium" symbolized the quest for the ideal library [1]. I had used this term in a Columbia University term paper in 1953 in which I speculated on the library of 2045-100 years from the end of World War II. I suggested that the already rapid increase in scientific information might accelerate the achievement of library nirvana. To qualify as the ideal library, The Informatorium would have to satisfy the "information requirements of a population that will be highly intellectual and scientifically trained". By the millenium there would be "a new Renaissance during which the population of the entire world will be thirsting for knowledge".

Whether the entire population of the world is thirsting for scientific knowledge is debatable, but the advent of World Wide Web has revealed an amazing public appetite for information of all kinds, both textual and visual in the world of combinatorial chemistry. Two months after discussing the ideal library, I asked "Who are the Information Scientists?" [2] and in that short editorial I forecast that every laboratory scientist would essentially become an information scientist. That prediction seems to becoming close to reality. Almost every scientist today is almost totally dependent upon computer-aided information resources. This is reflected in the growth of biomedical and chemical informatics. The distinction between wet and dry laboratories is growing increasingly thinner in the world of combinatorial chemistry.

As most of you know, H.G. Wells' World Brain [3] has become a metaphor for a futuristic view of information science and technology. Others prefer to credit Vannevar Bush's Memex [4]. However, I have always given H.G. Wells the priority and even commissioned a major and unusual work of art in 1981 by Gabriel Liebermann with technical assistance from Vernon Porter at Texas Instruments. Their holographic etching entitled "The World Brain" resides in the lobby at ISI [5] in Philadelphia. Wells was also on my mind when I wrote "Towards the World Brain" [6], which includes my testimony before a Congressional Subcommittee on Education and Labor of the US House of Representatives of the 88th (1963--64) Congress. Here is how I described the "Information Crisis" to a lay audience:

"We all take the telephone for granted. When we have to wait more than a few seconds for a dial tone, we grow impatient and frustrated. When we call information - seconds seem like hours. We also take for granted the telephone directory - that innocuous book which methodically lists names and numbers in alphabetic order. Imagine the chaos in the telephone company information centers if one day every other page in everyone's phone books were missing. Imagine your frustration if most telephone numbers were "unlisted" - if a special, prolonged, and elaborate effort was necessary each time you made a call."

"Contemplate the chaos in your city if there were hundreds of different phone books - some ar ranged by people's national origin, others by occupations, by district or by name - yet none of them complete. Each time you needed a phone number you would have to know whether your friend was Irish, or a janitor, or whether be lived in the north side of town. Suppose that in each city the sys tem was different - each used a different terminology or system of spelling - a janitor might be a superintendent or a maintenance engineer."

"Suppose each of these phone books, large and small, is only half complete and at least a year old when it arrives. Suppose that phone books were not free but cost so much that only libraries could purchase them. Imagine your frustration if you had to go to the library each time you wanted to make a phone call,'

"Now what has all this to do with the so-called information crisis? The situation I have just hypoth ecated is a fairly accurate description of scientific communication today. There are some obvious exaggerations. On the other hand, there are even more chaotic aspects difficult to convey by simply analogy. We all use the yellow pages, the classified directory, and frequently find it difficult to locate a number because of peculiarities in our language. Gas stations are listed under service stations and sell gasoline; gas companies may be listed under power companies and sell gas. In science, terminol ogy is constantly changing - faster than the lexicographers or dictionary publishers can cope with. Every scientific dictionary is obsolete long before it is published."

"In science communication we not only call local numbers - we are constantly trying to place long- distance transoceanic calls because science is international."

"Our telephone operators, the Information scientists and librarians, must be able to handle dozens of languages including Japanese, Russian, and other exotic tongues."

"However, this is only the beginning of the difficulties. After painfully identifying the telephone number of the scientific document he needs, the scientist can't simply dial the number. He must first identify the telephone exchange that handles this number. He may be lucky and find that it is a local exchange."

"Quite frequently he will find that he must call a Washington exchange or some other remote city. But scientists are stubbornly persevering, and having learned the proper exchange, put through the call only to find that the line is busy. In fact, the average waiting time is a few weeks - and by then - if that hasn't discouraged him - he may find that he called the wrong exchange, the number is out of order, or disconnected, temporarily or permanently. It is not surprising that by the time his call does get through he has sometimes forgotten why he called in the first place."

"The working scientist places hundreds and thousands of such calls each year. He would call more often if he did not anticipate, consciously or intuitively, delay and frustration. The net result is that he gives up and only makes a call when he is absolutely desperate. If he can afford the luxury he will turn the job over to someone else - an assistant or a librarian."

"Today's system of scientific communication is absolutely chaotic. That we are able to operate with it at all is a tribute to human perseverance. Science communication is still in the pony express era."
 

THE INFORMATORIUM  June 19,1962 The Ideal Library- The Informatorium   This is the first of a series of short editorial reports on various aspects of science information, its retrieval and dissemination, that will be written by me, IS[ staff members, and others. After llve years of self-restraint, I feel I can no longer resist the oppor tunity to preach the gospel of scien tific information. Since we haven't missed a publication deadline in five years, I now feel we can justify de- voting some time and energy to such projects as this series of articles. The tide, TheInformatorium, istaken from a paper I wrote several years ago as the first Grolier Society Fellow at Columbia University. In it I theorized on what I thought "the library of the future" would be like in 2045 -- 100 years from the end of World War 11. With the rapid increase in science information in the last several years, TheInforrnatoriummight come into existence sooner than expected. I think the title is espe cially apt for this series of articles, since most topics discussed here will be aimed at making TheInformatoriurna reality. As I see it, the ideal library must be able to supply information instan taneously. As an ideal library facility, Thelnformatoriumwill have to satisfy a wide variety of information require- menus for a population that will be highly intellectual and scientifically trained. In my paper, I predicted rather optimistically and hopefully, that by the turn of the millennium there would be "'a new Renaissance during which the population of the entire world will be thirsting for knowledge." I think that age is already beginning in spite of, if not because of, threats of atomic destruction. Basically,The lnformatoriumwill cover various subjects in the field of scientific information, make announce- ments, and discuss new experimental indexes, projects and services for   CurrentContentsreaders. For ex ample, we have been engaged in ex- periments in compiling new author indexes. Through personal correspond- mace I was rather surprised to find that some readers do not feel that an author index to each issue ofCurrentContentsis as important as a cumulated index appearing every few months. Many readers have suggested a subject index. We have recently started using a new paper that is whiter and lighter in weight, yet more opaque. It is also more expensive, but we feel it is an important investment. We have also changed the format of the address directory from a two column to a sin- gle column arrangement. It looks better and is easier m prepare. We would certainly appreciate any prac- ileal suggestions from our readers on new formats for this directory. In future issues I plan to discuss various experiments ISI is conducting, or has conducted in the past, includ ing Miniprint indexes and abstracts. the COPYWRITER.@ citation indexes, international cooperation and copy- right, selective dissemination of in- formation,  microform  storage and retrieval  of documents,  contents page formats, chemico-linguistics and scientific nomenclature in general. the future of scientific journals, the World Brain, Memex, and the future Of the information scientist.       Fig. 1. First CC essay "The Ideal Library", dated June 19, 1962.

While we have a long way to go, the information crisis in worldwide scientific communication, if it still exists, has changed in character. The crisis today is more accurately one of information overload. It is no longer the question of keeping tabs on what has been published. And there is the transitional tension

Fig. 3. Photograph of Dr Garfield
and J.D. Bernal.

between printed and electronic journals and e-print depositories, exacerbated of late by the proposal of Harold Varmus, Director of the US National Institute of Health [7,8].

The future prospects in information technology, particularly traffic capacity, itself will be covered more than adequately by Tony Cawkell. However, as an example of the rapid change in information technology, let me describe briefly the rapid transformation of The Scientist,a news journal I have published for the past 12 years. It was in fact the first continuously published print journal to appear on-line, having begun in the old gopher days. It has been transmogrified into a full 10-year archive in html, which I recently described in terms of the worldwide reprint culture [9].

Having mentioned H.G. Wells and Vanevar Bush, it is essential to recall one of my early mentors, J.D. Bernal. He was not my professor per se, but as I have stated many times [10], he was an inspiration from my high school days. I took great pleasure in meeting him at the International Conference on Scientific Information in Washington, DC, in 1958, just one year after the Dorking conference [11]. The e-print facility at Los Alamos essentially is the fulfillment of his plan for a centralized repository of scientific communications.

And I am proud to have been responsible for the annual Bernal Award of the Society for the Social Studies of Science.

The subject of scientific communication, per se, is not likely to be a major theme in a conference about informatics research. Nevertheless, we are here to honor Jack Meadows. Surely I can claim him as a co-worker in the area of scientific communication. His book onThe ScientificJournalis one part of the evidence [12]. While I have always taken great but ambivalent pride in being described as an information scientist, my preferred claim to fame is in the field of scientific communication. So I feel a special affinity for Jack and his contributions to scientific communication.

When Helen Henderson asked me to give this talk I hadn't remembered that I visited Cranfield in July, 1971, and in fact gave a paper at the Third Cranfield Conference on Mechanised Information [13]. It is not necessary for me to go over the ground covered in that paper on "The Relationship between Large Multidisciplinary Systems and Specialized Information Services". I guess my friends in the ab stracting and indexing world did not appreciate my assertion that they could essentially close shop and not be missed (now that SCI® was available). Much of the redundancy of the various services in spe cialized abstracting has largely disappeared with the adoption of author abstracts being fairly universal. Undoubtedly the specialized indexing provided by CAS, Medline, and others is still greatly appreciated by users, but I suspect that the availability of full-text searching on the web will affect that as well in the not too distant future. When we introduced automatic Permuterm indexing of tides it was considered a non-event. Medline and others have since realized how powerful his kind of natural language indexing can be.

1997- H.M. Collins
1996- David Bloor
1995-Bernard Barber
1994- Mary Douglas
1993- David Edge
1992- Bruno Latour
1991- Melvin Kranzberg
1990- Thomas Hughes
1989- Gerald Holton
1988- Dorothy Nelkin
1987- Christopher Freeman
1986- Michael Mulkay
1985- Joseph Ben-David
1984- Joseph Needham
1983- Thomas S. Kuhn
1982- Robert K. Merton
1981- Derek de Solla Price

As a final retrospective view of the original Cranfield Project, let me say how disappointed I was not to have been present at the 1997 meeting which celebrated the 1957 Dorking Conference. It was there that I first met Cyril Cleverdon. When Helen Henderson asked me to participate in a Cranfield meeting, I said that I could not say much about the Cranfield experiment, except to point out that it had not been able to deal with citation indexing. The notion of relevance, particularly with respect to citation-based retrieval, is ambiguous to say the least. Like beauty, relevance, particularly in the context of citation indexing analysis, is surely in the eye of the beholder. This is often called the Humpty Dumpty Syndrome. To paraphrase Lewis Carroll, "When I use a citation, it means what I want it to mean".

Henry Small wrote a very useful paper on the citation as a symbol [14]. One of the beautiful ironies of citation retrieval is that the least expected retrieval result can turn out to be the most interesting. That makes it all the more relevant. Someone who is simply publishing another routine citation analysis will cite the expected group of authors, including myself. But the author that I am least likely to encounter may provide the most interesting connection to my published work.

I have not said much about the future of information science and technology research. When I use the Verity search engine [http://the-scientist.verity.library.upenn.edu] to do full-text searches on over 1000 essays I have published, I marvel at how far we have come. And I marvel at the achievement of autonomous citation indexing achieved by Lawrence et al. [15], at NEC. But as elegant as is this system for displaying the context of citations, it is quite different from automatic selection of citations [16]. But as a further indicator of future research, I reflect on the past and think of John O'Connor doing full-text searching manually, trying to understand how to give meaning to the search for papers on toxicity. How do we retrieve them when in fact that word rarely occurs in relevant papers. For similar reasons, we do not yet have a completely successor system of mechanical translation. So the fundamental challenges of artificial intelligence are essentially the same as when we started four decades ago.

References

[1]  back to text  E. Garfield, The ideal library - The Informatorium,Current Contents(19 June 1962). Reprinted inEssays of an Informa-tionScientist,Vol. 1, ISI Press, Philadelphia, 1977, p. 1. http://garfield.library.upenn.edu/essays/Vlp001y1962-73.pdf.

[2] back to text  E. Garfield, Who are the Information Scientists? Current Contents (7 August 1962). Reprinted inEssays of an InformationScientist,Vol. 1, ISI Press, Philadelphia, 1977, p. 2. http://garfield.library.upenn.edu/essays/Vlp002y1962-73.pdf.

[3] back to text  H.G. Wells,World Brain,Doubleday, Doran, & Co., Garden City, NY, 1938, 194 pp.

[4] back to text  V. Bush, As we may think,AtlanticMonthly176 (1945), 101-108.

[5] back to text G. Liebermann, ISI's 'World Brain' by Gabriel Liebermann: The World's first holographic engraving,CurrentContents15 (28 December 1981), 5-11. Reprinted inEssays ofanInformation Scientist,Vol. 5, ISI Press. Philadelphia, 1983, pp. 348-354. http://garfield.library.upenn.edu/essays/v5p348y1981-82.pdf

[6] back to text E. Garfield, Towards the World Brain,CurrentContents(6 October 1964). Reprinted inEssays of anInformationScientist,Vol. 1, ISI Press, Philadelphia, 1977, pp. 8-9. http://garfield.library.upenn.edu/essays/Vlp008y1962-73.pdf

[7] back to text D. Butter, The writing is on the wall for science journals in print,Nature397(6716) (21 January 1999), 195-200.

[8] back to text  D. Butter and M. Wadman, Mixed response to NIH's Web journal plan,Nature399(6731) (6 May 1999), 8-9.

[9] back to text  E. Garfield, The evolution of the reprint culture: from photostats to home pages on the World Wide Web" A tutorial on how to create your electronic archive,The Scientist13(4) (15 February 1999), 14.http://www.the-scientist.com/yr1999/feb/comm_990215.html.

[10] back to text  E. Garfield, J.D. Bemal - the sage of Cambridge. 4S award memorializes his contributions to the social studies of science,Current Contents19 (10 May 1982), 5-17. Reprinted inEssays of an InformationScientist,Vol. 5, ISI Press, Philadelphia, 1983 pp. 511-523. http://garfield.library.upenn.edu/essays/v5p51 1y1981-82.pdf

[ 11] back to text  J.D. Bemal, The transmission of scientific information: a user analysis, in:Proceedings ofthe InternationalConferenceonScientificInformation(Washington, DC, 16-21 November 1958), Vol. l, National Research Council, National Academy of Sciences, Washington, 1959, pp. 77-95.

[12] back to text  A.J. Meadows ed.,TheScientificJournal(ASLIB, London), 1977, 300 pp.

[13] back to text  E. Garfield, The relationship between large multi-disciplinary systems and specialised information services. Presented at the Third Cranfield Conference on Mechanised Information Storage and Retrieval Systems, Cranfield, UK, July 20-23, 1971.

[14] back to text  H.G. Small, Cited documents as concept symbols,SocialStudies ofScience8(3) (1978), 327-340.

[15] back to text  S. Lawrence, C.L. Giles and K. Bollacker, Digital libraries and autonomous citation indexing,Computer32(6) (June 1999), 67-71. http://www.neci.nec.com/lawrence.

[16] back to text  E. Garfield, Can criticism and documentation of research papers be automated?,CurrentContents9 (4 March 1970). Reprinted inEssaysof anInformation Scientist,Vol. 1, ISI Press, Philadelphia, 1977, pp. 84-90; and includes reprint of E. Garfield. Can citation indexing be automated? From: Statistical association methods for mechanized documenta tion (Symposium proceedings, Washington, 1964), M.E. Stevens et ai., eds, National Bureau of Standards, Washington, Miscellaneous publication 269, 1965, pp. 188-192. http://garfield.library.upenn.edu/essays/Vlp083y1962-73.pdf.