Eugene Garfield, Documentation Consultant


Robert Hayne, Head, Document Analysis Unit, Smith, Kline & French Laboratories, Philadelphia, PA

Presented at the "Symposium on Storage and Retrieval of Scientific Information" of the annual meeting of the American Association for the Advancement of Science, December 28, 1955 in Atlanta, Georgia


To the layman, and to the average scientist, "documentation" still denotes the usually unpleasant business of providing bibliographical sources. High school teachers still persist (and rightfully) in telling students of their inadequate or poor documentation of term papers. Certainly this remains an important part of present-day documentation. The conscientious author who has had the benefit of effective documentation will provide an exhaustive bibliography. Conscious of his momentary role as a documentalist, he will spare no energy in reporting his sources completely and accurately. Poor documentation will limit the value of an author's work.

It is indeed strange that after so many years any discipline should still be as ill-defined as documentation. Recently the American Documentation Institute conducted a contest for a suitable definition of "documentation." It is remarkable enough that the contest should have been necessary, but even more remarkable - despite the gratifying participation of many younger and elder statesmen of the science � that none of the definitions submitted was considered entirely satisfactory.

We face a three-part problem. First, documentation has something to do with library science, but no one seemingly knows, or is able, or wants to define the relationship exactly. Is it "modern" librarianship, or a "modern" complement to librarianship necessitated by requirements of modern research and made possible by modern technology? This first part leads to the second. Documentation like librarianship has always been the victim of its own poor public relations work. Very few of those who stand to gain most from documentation know what it has done or what it can do. Until these people appreciate the potential of documentation, financial support for its proper development will obviously be lacking.

Third, there is the problem of the word itself. Documentation is a poor name. The disagreement among documentalists and the confusion among laymen reflect its inadequacy. Semantic problems have been of great concern to many documentalists, but few have tackled this important semantic problem. We certainly haven't appreciated what Korzybski has taught us about fixed associations. A successful name must either denote immediately what is intended to be expressed, or its familiarity must be intense enough to communicate meaning. "Documentation" does neither. It still denotes at this late date merely the supply or use of documents in support of a thesis. It is of historical interest to note that one cannot easily find the word "documentation" used prior to 1900 when the International Institute of Bibliography had its start. Almost thirty years later it was named the International Institute for Documentation. During this thirty-year period the word had gradually been applied in its larger sense, by a small group of documentalists. The founders of the Institute were really borrowing a well accepted term, documentation in its restricted bibliographical sense, and applying it to a new science, whose own acceptance and recognition is rather limited. As far as I have been able to discover,the word documentation makes little impression on the uninitiated hearer. Indeed the word often agitates him.

We need a new word, an interesting and arresting word. Accuracy of nomenclature need not necessarily be a consideration. New words at least have the advantage of iron-clad definition. One of my colleagues uses the word biblio-cybernetics. For the specialized area of documentation concerned with the use of computers and other devices, I use the word documentation. Instead of documentation, Mr. Hayne suggests "documetics" - the complex of activities which result in documentation. Alternative suggestions will be welcomed. Once the word is coined, we must accelerate our public relations program so as to give the new appellation the widest circulation.

By stubbornly retaining the word documentation we therefore define documentation as one identifies Shakespeare as the author of the plays and poems attributed to Shakespeare. Documentation is really anything and everything involved in the creation and use of documents. That includes a lot � writing and publishing the paper, analyzing it, indexing it, storing it, copying it, retrieving it, and using and evaluating the data in it. Our new word should, by definition, cover all these facets.

In a highly competitive world, the quality of documentation available to the industrial manager, the scholar, the scientist can make the difference between success and failure. Science depends, as much as business management, upon the rapid retrieval of complete data if expected results are to be achieved, and achieved in time. Thus documentation is the forerunner of intelligence � a word that, though it has received wide acceptance, also has a disadvantage. It often connotes cloaks and daggers. In actual practice more than ninety percent of governmental intelligence work consists of the painstaking collection, organization and analysis of masses of old and new information, most of which is originally available to anyone. This is dramatically borne out by the recent comments of Dr. Samuel Goudsmit. Let me read the following clipping from the Science News Letter:

No spies need apply for Jobs in scientific intelligence, if the advice of Dr. Samuel A. Goudsmit, given to the American Philosophical Society meeting in Philadelphia, is followed.

Dr. Goudsmit, now chairman of the physics department at Brookhaven National Laboratory, investigated why the Germans did not achieve the atomic bomb during World War II.

"There are no Mata Haris with doctors' degrees in physics or chemistry," Dr. Goudsmit said. "Scientific knowledge cannot be transmitted via laymen.

"The purpose of scientific intelligence is to estimate whether a country's scientists and laboratories will make a significant contribution in case of a war emergency. Such an evaluation requires an insight not merely in science but also in economic, political and historical questions. It can be achieved only through the cooperation of many competent specialists. The sources of information are usually openly available but visible only to those who are aware of the intelligence problems.

"The evaluation of technical and scientific progress is rather straight-forward. More difficult, and at least equally important, is to reach an opinion about the mutual influence between scientists and governments.

"Under Hitler and Peron and from time to time in the U.S.S.R., pseudo-scientists were the principal advisers to the authorities, which of course had detrimental effects upon progress."

Our own government, in creation of the Central Intelligence Agency, has taken a step whose counterpart must be found for science. The scope of the CIA is the totality of information pertinent to the safety and the progress of the nation. The time has come for an analogous scientific body, one whose scope is the totality of information pertinent to science and research. The instrument of such a projected body must be documentation. On the national level, science intelligence activities are non-existent and what is being done by various special groups is completely uncoordinated. Most of you ladies and gentlemen probably remember that the scientist was once supposed to work in an ivory tower.

We may have ivory towers, but they are all connected by cables under the ground or by invisible channels in air. Scientific progress�the accumulated effect of your individual contributions�depends upon a free flow of information, of thousands of minute facts, of millions of seemingly unrelated observations made and reported by scientists in diverse specialties. Freud credited much of his success in the study of cocaine to Billings's Index Catalog which cited the work of botanists, anthropologists, zoologists, pharmacologists, doctors of medicine, and others. Most of us are not so fortunate. With today's voluminous publication, a John Shaw Billings is impossible.

Hindsight allows us to say that it is unfortunate that modern documentation has developed in the era of scientific specialization. Documentation's major achievements have been made in the service of the specialties. They have too often remained specialty accomplishments and have not been applied with advantage to the other specialties or to science as a whole.

The time is past when the products of intense specialty studies are utilized immediately only by the specialty which produced them, if there ever was such a time. They are the concern of science as a whole. Furthermore, the problem of specialty coordination which has plagued science as a whole is indeed beginning to plague the specialties themselves�further specialization within specialties is developing�the steroid chemist for example�while what we may call inter-specialism�biophysics, psychiatric biochemistry�is also increasing. Both the sub-specialties and the inter-specialties require a unified coordination of scientific intelligence, which does not now exist.

In an article, "The Literature of Forestry: An Appraisal" there is reported, "The literature of forestry is scattered in many non-forestry publications, from a microscopist's journal to a monograph in applied chemistry. The journals regularly scanned have grown from 847 in 1947 to 1310 in 1955. Publications come from nearly 100 countries and appear in more than 30 languages." The facts in this quotation are significant for us for three reasons.

l. This pattern is repeated in dozens of scientific specialties.

2. The 1310 journals are being scanned certainly by other and probably by one hundred scientific groups. None can at present take advantage of any work done by the others to prevent costly, and, as we shall see, a potentially dangerous waste of scientific effort and talent.

3. Interest in forestry�like geology and geography�runs at comparatively low throttle in peacetime. It is when an emergency comes�military or economic-that the throttle is opened all the way out and demands are made of the specialty� and its documentation�that because of accumulated disinterest and years of complacency it cannot meet.

Who is to provide the order of priority for documentation of the technical and scientific specialties in anticipation of the requirements of some sort of national emergency?

Our country is fortunate enough to have been blessed with abundant natural resources and amazing productive energy. To use these effectively, however, requires proper timing. Surely no aggressor will fall to learn one lesson both world wars have plainly taught�it is courting defeat to dissipate one's resources unless the program necessitates like action by the United States. In the past, we have had time�strategically either the gift of our oceans or tactically the creation of our friends. We shall probably not have time again, if the world is so unfortunate as to experience a third world war. Part of the preparation which we must make in anticipating this lack of time is a coordination of scientific documentation, so that when an emergency arises, needed information and techniques will be immediately at hand, and we shall not have to dissipate manpower for documentation projects.

It is not, strangely enough, that our government does not realize the importance of documentation activities. Through UNESCO we support the establishment of national documentation centers in countries like India and Mexico, but not at home, where we have strategically perhaps more reason to do so.

In the Soviet Union a different situation prevails. Documentation, like the other sciences, is receiving support of the State apparatus. The story of Soviet scientific and engineering manpower, has already been reported by the NSF. In a few years we shall also be hard pressed to duplicate the documentation facilities available to the Russian scientist. I do not look to such centralized governmental control as an answer. We must, however, fully appreciate the dangers of being left behind.

A great deal of documentation literature today reminds me of the early Greek philosophers, who without practical techniques, theorized, sometimes brilliantly, upon the nature of matter. We have all read many, I think too many, papers on the philosophy and logic of various possible documentation systems. Our technology has already given us practical tools and we are making too little use of them. Not long ago Vannevar Bush, perhaps as a joke, proposed Memex, something like H. G. Wells's World Brain in the form of an electronic desk. The merits of his proposal have been lost in the plethora of idealistic and again theoretical comment which it provoked. When a man like Dr. Bush suggests a Memex, too many people feel that there is nothing left to be said. Certainly Bush's Rapid Selector could not be called a practical realization of Memex.

Some time ago testimony was presented before a congressional committee considering an appropriation for the Rapid Selector. I wrote the proponents and asked if they had prepared any information on what this machine would contain, and how much. The answer was not forthcoming. Furthermore, when I brought to the attention of certain government agencies an invention that was an actual realization of Memex, the attitude was taken that the Rapid Selector had solved that problem. Hence, one must differentiate between the mechanical problem posed by Memex and the intellectual problem it poses. In a coordination of documentation on a national level lies a practical realization of the intellectual problem of Memex. In the words of Otto Neurath:

"Comprehensiveness arises thus as a scientific need and is no longer a desire for vision only. The evolving of such logical connections and the integration of science is a new aim of science."

Because Memex and the Rapid Selector were not based on realistic needs, a disservice has been performed, by sidetracking us from objectives, which were clearly recognized by Neurath and others over twenty years ago.

There are difficulties in our path no doubt, but we have been intimidated, I think, by some old bugaboos. Again Neurath has something pertinent to say: "An encyclopedic integration of scientific statements, with all the discrepancies and difficulties which appear, is the maximum integration which we can achieve." The bugaboos are the familiar ones�the impossibility of achieving a universal language for the comprehensive, unified scientific index, a coordination of specialty terminologies, a solution of the problem of synonymy. These bugaboos will lose most of their substance in the reality of practice.

The history of Esperanto is a good example. One could fill a library with the analyses that have been made of the requirements for an international language. Discussion of Esperanto's linguistic imperfections is equally exhaustive. Yet Esperanto is enjoying an ever increasing success � not the widespread fadism it once enjoyed, but the success of use in scientific periodicals, and of use as a means of communication between serious men. In the November 19 issue of The Pharmaceutical Journal Ernest Paul stoutly defends Esperanto as the vehicle for his contacts with workers in Scandinavia, Japan, China, and countries behind the Iron Curtain. Mr. Paul does not defend Esperanto's linguistic perfection-that would be a loosing battle. He says rather, "We have yet to see whether Basic English can get off to a better start than Esperanto". That is the point. Esperanto got started, and for all its linguistic square wheels is still lumbering along. Similarly we must get started in coordinating science documentation.

"Novial" is a much better, more flexible, and esthetically less offensive language than Esperanto, but few people have ever heard of it. Jesperson gave it the benefit of years of linguistic theory, but very few have used it.

Interlingua, which shares most of Novial's good points, but which is not greatly superior, is receiving increased acceptance because the promotion by its inventors and supporters is active in increasing its use, as for example in "The Science Newsletter." We must get started.

We must get started on a master list, a universal subject heading list, one that embraces the totality of science, however imperfectly. It matters little whether it is in English, Interlingua or some other acceptable language. Translations would quickly appear. In actually getting started, we will be able to take advantage of two deep-rooted human propensities. First, the readiness to accept what is at hand, what is being used�the readiness to adopt, in other words, the language of this proposed master list as the international language of science. Second, the eagerness to correct and improve. Men who will never expend an iota of energy starting such a list will expend considerable energy and talent in removing the imperfections of one that does exist. Synonymy is really not worth discussing�the problem will be solved by arbitrary decision, once work on the master list is begun. The success of the WHO's "International Statistical Classification of Diseases, Injuries and Causes of Death" is a good illustration of this point, as is the "Chemical Abstracts" system of nomenclature.

In the past efforts to realize a unified or coordinated documentation of science have been generally discouraged because of the tremendous clerical jobs involved. Certainly, as Mr. Alexander will show in his discussion of documentation tools, there is no cause to continue in this discouragement. The tools are available�remarkable tools that are particularly capable of handling clerical jobs of large proportions. Some time ago I recommended the compilation of a unified index to science essentially involving the large clerical job of piecing together all existing indexes of journals and abstracting services to produce an index more complete and thorough than any of the fragments from which it was constructed. At least the duplications of indexing that do occur could be put to advantage.

The problems of financial support and manpower are considerable and at this point may loom very large indeed, but they are by no means insoluble. The existence of Chemical Abstracts, Excerpta Medica, the Current List of Medical Literature, Biological Abstracts, Nuclear Science Abstracts, and many others prove that money and manpower are not insuperable obstacles. The real problems are inertia and intellect.

In the past, librarians, if they had the resources or the will, could have attacked the intellectual problem. We should not forget Billings, the librarian and scientist, and his Index Catalog. But the librarians did not act. Scientists were not willing to wait. In a fragmentary way, specialty by specialty they attacked the documentation vacuums -- engineering, chemistry, biology, psychology, physics, etc. Application of old techniques such as the 3x5-card file from the beginning would have been perfectly acceptable, if the basic intellectual problem had been realized. Many current documentation services are still based on the 3x5 card, the Engineering Index, Battelle, Wistar, etc. Very simply, librarians did not make the transition from cataloging books on a national basis, as is still done so admirably by the Library of Congress, to the indexing of periodical articles and other reports so vital to the scientist.

Someone once said that the great virtue of classical German scholarship was that it always asked the right questions. Its great failing was that it usually got the wrong answers. Now documentation is in a similar position. It has been getting a lot of right answers, but it has been asking the wrong questions. I propose that we start asking the right questions�that we stop asking how we can do more quickly what has already been done well by traditional methods, and ask instead how we can do what has not been done at all. No one need prove that punched cards, microcards, minicards, etc. will be of considerable aid in any research program to which they are applicable, but how much time and paper is still spent on repeated proofs of exactly that. Descriptions of such applications usually are qualified by the statement: "But of course the efficacy of the system will depend in great part upon the code developed and the manner of its use". The intellectual problem again. Machine codes are a translation of some intellectual framework, whether it is a conventional dictionary of terms, semantic factors, uniterms, or descriptors.

We have returned to a master list, to a national center of documentation, the central science intelligence center. The purpose of each of these will be the coordination of the segmentalized scientific disciplines by issuance of comprehensive indexes, encyclopedias, handbooks, reviews, and other instruments that will erase the artificial boundaries of specialization.

For those of you who came here primarily to hear us report on "Documentation: Present Status and Trends," I would like to make some short summarizing remarks. The general theme of this paper has not been what you expected only because the original title was assigned in advance-but we were given, and took, carte blanche.

We have said a great deal about the status of documentation�as a term that lacks general acceptance; as a science that lacks unity within its own ranks, in which there is too much squabbling over minor points; as a matter of strategic importance, that is terribly neglected, as a career, for which proper educational facilities are all but completely lacking.

There are some very positive signs, however. Documentation activities, albeit fragmentary and mostly uncoordinated, are increasing. This indicates that there is a growing acceptance amongst scientists that great benefits can be derived from good documentation activities, that the literature is not only something one uses for general inspiration, but also as a tool to obtain information just as one performs experiments. This is reflected in the increased demand for scientists whose training or inclination is away from the lab per se. As part of a team they can take advantage of the literature through effective documentation techniques. We now find the counterpart of the literature chemist in most other specialties. Equally important is a growing acceptance by administrators that documentation activities, whatever they are called, are as equally important to research as lab work.

The principle trends to be noted are increased interest in the use of mechanization, the departure from traditional techniques of classification, the increased use of micro-reproductions, increased use of translation services in contrast to individual mastery of foreign languages, and experimentation with new bibliographical formats. There is no shortage of new ideas, but because our documentation efforts have been fragmentary they have not gotten too far.

Mechanization has been most successfully applied in indexing to the printing of indexes and index searching but little has been accomplished in actual mechanical extraction of the basic ingredients of articles. This problem merges intimately with that of mechanical translation. With specific regard to punched-card devices, suffice it to say they are well proven adjuncts to a scientific research program. Of course, one of the principle ingredients is coordinate indexing which provides the multi-dimensional approach needed in science. Up to this point mechanization activities have been able to command large financial resources and practical opportunities because of the attraction of the machine and its seeming potential for the elimination of manpower. However, the time quickly arrives when mechanization alone loses its attraction. Management then has to face the real intellectual problem of documentation again before an effective science intelligence can result. For if automation does anything, it creates a need for more qualified personnel to handle the great volume of information which the machines can handle and make accessible.

Rapid strides forward have been made in methods of reproducing documents, but there is still nothing that approaches the ideal from the scholar's point of view. He needs a small, portable or desk size instrument that will provide him instantaneous copies of text-as he may require at the moment.

The dissemination of scientific information is receiving a great deal of attention, but here again I think there is a failure to recognize that inadequate dissemination can be traced to inadequate tools of dissemination rather than the greatly emphasized problems of journal format, volume of publication, etc. There are three stages in the dissemination process, these follow logically, one from the other.

The contents type publication is back with us to stay. The tables of contents of journals and books if widely disseminated provide scientists with an opportunity to scan quickly a large portion of the publications that may contain useful information. These are the cream of the crop, perhaps one to two hundred journals for each scientist. Concurrent with this type of publication, are the journals that the scientist will scan in toto�a dozen or so which he will usually use for news and other information other than original articles as e.g. advertising. Then we have the abstracting services, which of necessity take more time to publish and which give more comprehensive coverage, article by article, to specialty interests. Finally there are the indexes for very precise coverage. I maintain that none of these is a substitute for any of the others. Each is essential. Together they provide a well rounded coverage of scientists' general needs. Certain gaps are filled by attendance at meetings. Other dissemination gaps still remain. For example, it is almost impossible for the individual scientist to know all the people who may be interested in his work, who are influenced by his writings and what implications his work may have outside his own limited area of research. For this reason, among others, I proposed a Citation Index to Science1 that would, among other things, enable individual scientists to know where or for what purpose their works were being quoted. I won't take the time to indicate the other serious bibliographical gaps that exist besides those mentioned earlier. Most could be filled as easily by the proposed National Science Intelligence Center as could the compilation of the Citation Index.

Attempts have been made to fill these gaps. The AAAS helps in this respect. But the attempt now must be a concerted one. If there is anything that the documentalist requires of the scientist it is that the scientist recognize that documentation is an area of research equally as rewarding, as productive, and as worthy of support as any others. The methodology which has elevated science to its present status must now be applied to documentation. Without this our national scientific effort will progress further but not as productively as may our competitors'. The AAAS is the logical group to encourage and support the establishment of an institute for research in the documentation of science. The Research Institute if established first could undoubtedly formulate the plans and scope of the National Science Intelligence and Documentation Center.

1. back to text Garfield, E. "Citation Indexes for Science: A New Dimension in Documentation
    through Association of Ideas" Science, 122(3159), p.108-11, July 1955.