I S K O
Encyclopedia of Knowledge Organization
Éric de Grolier
by Michèle HudonTable of contents:
1. Introduction
2. Major works
2.1 Théorie et pratique des classifications documentaires (1956)
2.2 A Study of General Categories Applicable to Classification and Coding in Documentation (1962)
3. Recurring topics
3.1 System of sciences and classification
3.2 The science of classification
3.3 Classification systems: 3.3.1 Universal decimal systems; 3.3.2 Other universal systems; 3.3.3 Special classification systems
3.4 Symbolization and notation
3.5 Technology and the future of classification
3.6 Classification research
4. Conclusion
Endnotes
Selected works by de Grolier
Secondary sources and references
ColophonAbstract:
Éric de Grolier (1911-1998) has published hundreds of papers on various topics connected to his lifelong interest for the fields of linguistics, terminology, political sciences, reading and libraries, communication and information science. This article focuses on de Grolier’s passion for the history, content and structure of bibliographic classification systems. Six recurring topics have been identified in papers published between 1956 and 1992: system of sciences, science of classification, classification systems, symbolization and notation, technology and the future of classification, and classification research. Each topic is presented to demonstrate that de Grolier’s standing within the narrow circle of francophone pioneers in documentation and information science is well deserved, and that his contribution to the advancement of bibliographic classification is indisputable.
1. Introduction
Éric de Grolier’s (1911-1998) career was remarkable by its length (~65 years) and by its scope. He was in turn, and sometimes even simultaneously, bookseller, bibliographer, special librarian, teacher, editor, translator, researcher and consultant. The man was known for his strong personality, his insatiable curiosity, his exceptional erudition, his seemingly limitless capacity to absorb knowledge, his noteworthy ability to synthesize and his acute critical mind. → Vickery (1991, 170) admired his readiness to argue with anyone about anything and his vast knowledge of modern culture; indeed, de Grolier described himself as an “interdisciplinary being” (Grolier and Fayet-Scribe 1996, 287).
De Grolier’s wide-ranging interests for linguistics and terminology, political science, economics, reading, communication and → information science (IS) were well served by his excellent knowledge of French and English, and by the quality of his writing style. In the course of his long career, he published hundreds of articles, proposals, comments, syntheses, etc., the complete list of which he could not establish himself, on topics as foreign to one another as propaganda and public opinion in the United States (1943), the history of the book (1954), the training of documentalists in France (1941) and the terminology of ethnic relations (1988b). Whoever is interested in these texts encounters a free thinker working equally assiduously on original historical studies as on directories and glossaries, while also publishing opinion pieces on controversial subjects. It is easy to picture him as hyperactive, simultaneously occupied with several projects, never staying for long in a particular professional position or institution; his thick resume, and the fact that his institutional affiliation tends to vary from one year to the next, attest to the fact.
Palermiti (2000) notes that de Grolier and his contemporaries J.C. Gardin (1925-2013) and R. Pagès (1919-2007) are little known in their own country. She presents de Grolier as a militant and a practitioner who took advantage of his eclectic training: a first diploma in publishing and bookselling obtained in 1929, formal studies in the fields of sociology, social and economic history, history of sciences and philosophy at the Sorbonne and at École pratique des hautes études, followed by a diploma in documentation from the Institut national des sciences et techniques de la documentation (INTD) in 1952. Palermiti points to numerous similarities between de Grolier’s career and those of Gardin and Pagès. All three were concerned with scientific communication in an era characterized by a transition from bibliography to documentation, the development of public reading in France, the estrangement of documentation and library science, the birth of IS and the mechanization of information processing. In this evolving context, all three reaffirmed the critical importance of document content analysis and representation. And all three, recognizing that existing → knowledge organization systems (KOS) were inadequate, tested the capacity of new languages to increase the efficiency of knowledge communication.
Young de Grolier is 16 years old and a bookstore clerk when he develops a passion for classification which will endure for the next 65 years. His first paper at a meeting of the International Federation of Documentation (FID), in 1931, marks the beginning of his work on the relation between the system of sciences and → knowledge organization (KO). From that date, de Grolier will be active in training programs and in the creation and management of associations, committees, etc. devoted to issues concerning → bibliographic classification (the Committee for a General Theory of Classification established by FID in 1951 [1], for example). A special issue of International Classification (now Knowledge Organization), published in 1991 to celebrate de Grolier’s 80th birthday and 60-year long career, is a testimony to the significance of his contribution to the evolution of classification theory.
This article focuses on de Grolier’s unwavering interest for the history, content and structure of bibliographic classification systems. His main ideas on → classification were presented in the 1950s, in the form of seven reports prepared at FID’s request; the content of these reports was summarized by Maniez in 1991. Till the end of his life, de Grolier continued to enrich these early observations and propositions, never failing to take into account the context within which classification activities were conducted.
The following presents an overview of de Grolier’s work on classification, emphasizes his most significant hypotheses and findings, and provides bibliographic suggestions for those interested in (re)discovering the contribution of this important figure of classification theory. Main sources are texts authored by de Grolier himself between 1956 and 1992 in the form of monographs, articles, reports and conference proceedings. Syntheses of scientific meetings prepared by de Grolier have also been used to illustrate his remarkable ability not only to summarize numerous communications, but also to identify recurring topics and the various ways to go forward (Cochrane 1991). These syntheses are significant because they were used by de Grolier to describe what he considered critical areas in classification research.
2. Major works
To understand Éric de Grolier’s contribution, it is important to present his two major essays, published respectively in 1956 and 1962. These essays integrate the enormous amount of knowledge already acquired and integrated by the author, then at the mid-point in his career.
2.1 Théorie et pratique des classifications documentaires (1956)
Vickery calls it the “Big book on classification” and describes it as the “most comprehensive study of classification that has yet appeared” (1991, 170). In his review, he suggests that the fact that this important treatise has not been broadly distributed could explain why its author was not more widely known in the English-speaking world.
De Grolier presents this work as a synthesis of notes and observations made over 25 years, warning that it should not be considered exhaustive because it had been physically impossible for him to consult the totality of relevant documentation. Conceived as early as 1938-1939 as a French equivalent to the Manual of classification for librarians and bibliographers published in 1926 by W.C.B. Sayers, the early text was reworked in 1952 as a thesis submitted to INTD, and finally published in 1956 after further revisions. De Grolier believed that this work provided an infrastructure for the development of more efficient classification systems (1956a, iii). He firmly held that it was necessary to examine classification from historical and theoretical perspectives, while acknowledging that practitioners and users were more concerned with pragmatic, application issues.
A look at the impressive table of contents (1956a, vii-xix) confirms that an enormous amount of research and analysis was accomplished by the author. Théorie et pratique… [Theory and practice of documentary classifications] is 400 pages long and includes 827 notes. In these notes, de Grolier refers to the numerous sources used to support the argument, includes quotes and explanations, and adds comments or criticisms relating to an element in the main text or in the note itself. A detailed index of names and subjects is provided. The fact that references are given within the notes rather than in a formal bibliography makes it difficult for the reader to establish a list of sources used; this is regrettable since such a list would constitute, on the one hand, an inventory of classification systems used over the past centuries and, on the other hand, a listing of the author’s own works, many of which were never published or are today impossible to find.
Théorie et pratique des classifications documentaires is divided into three sections. The first section (p. 1-96) introduces the general characteristics of classification systems and proposes an informal template for analysis. De Grolier applies this template in the second section (p. 97-259) to illustrate the parallel evolution of classification structures and the system of sciences. In the third section (p. 260-368), the author turns to the practice of classification and its future, submitting proposals for the coordination and standardization of classification systems, and stating his conviction that the construction of a single standardized universal scheme is not only desirable but also feasible. His recommendation that → users’ needs and behaviour should be taken into account is thought-provoking: indeed, nobody at the time was giving much attention to this issue.
Every topic which de Grolier will study and discuss later in his career is introduced in this first major work which includes, among other gems, an impressive historical survey of classification systems and a presentation of thinkers and systems rarely, if ever, referred to in textbooks.
2.2 A Study of General Categories Applicable to Classification and Coding in Documentation (1962)
The second of de Grolier’s major essays is a report prepared at the request of UNESCO with the objective of producing an inventory of general categories (logical, spatial, chronological and formal) common to classification systems used in special domains. Éric de Grolier found in this mandate the perfect opportunity to combine his passions for linguistics, classification languages and information transmission. The author explained how difficult it was to process such a large volume of data while keeping abreast of technological innovations applicable to information retrieval. Two full years were necessary to complete the analysis. The study was conducted in the context of a proliferation of domain-specific documentation centres and special classification systems, with the hope of clearing the way for the production of a universal standard in classification which could also regulate the process of subject analysis.
The Study focuses on general categories, or parallel divisions, found in classification systems since the 17th century, and on the devices controlling their application; in the Dewey Decimal Classification (DDC), for example, these correspond to the tables of common subdivisions and to number building instructions. It is supported by a solid bibliography which again only manifests itself, unfortunately, in the 450 notes grouped at the end of the report.
The first chapter (p. 19-63) presents a list of categories common to a number of universal classification systems, such as DDC, the Universal Decimal Classification (UDC), the Library of Congress Classification (LCC) and the → Colon Classification (CC). Consideration is also given to a few systems introduced after CC. The list of categories suggested by G. Cordonnier in 1957, for example, is provided in full: organizations and services, persons, living beings, body, material, actions, abstract concepts, document forms and time, a list de Grolier deems incomplete without, however, suggesting any addition.
Chapter 2 (p. 65-148) is the most extensive. In it, de Grolier describes general categories and relations found in 40 special classification systems. For convenience reasons, the author has grouped the systems according to country of origin rather than by discipline; France, England and the United States benefit from detailed analyses. For France only, 20 systems are presented, some already known to specialists (those of Z. Dobrowolski, R. Pagès, J.C. Gardin, Commissariat à l’énergie atomique), and others much less so (those of S. Tchakhotine, J. Samain, F. Isambert). To cover England, de Grolier chooses to review the works of some of the greatest names in classification theory (J.E.L. Farradane, D.J. Foskett, B.C. Vickery, B. Kyle), and to assess the importance of their contribution within the Classification Research Group (CRG). For the United States, the efforts of C.N. Mooers, J.W. Perry, H.P. Luhn, M. Taube and the U.S. Patent Office are described. The chapter includes brief introductions to lesser known classification systems originating from the Netherlands, Germany, Canada, the U.S.S.R, India and Japan.
Chapter 3 (p. 149-164) covers general categories found respectively in natural language, → artificial languages, and interlanguages. The author’s objective is to establish a parallel between natural language and controlled indexing languages. De Grolier provides multiple references to various schools of thought: the German School and the work of W. von Humbolt on language as structuring system, the Franco-Swiss School and the work of F. de Saussure on structuralism, the English School and the studies completed by the Cambridge Language Research Unit, the School of Prague and the Italian School. Several paragraphs are dedicated to proposals submitted in the United States by E. Sapir, B.L. Whorf, N. Chomsky and R. Jakobson respectively. The last pages offer a survey of the most popular research areas in linguistics at the time (classes of words, grammatical genre, case systems, derivation, and typology of languages), assessing their relevance for the development of a unified scientific and documentary language.
If de Grolier’s effort to complete such an analysis was no less than remarkable, its usefulness remained in fact limited. The study did not lead to a proposal for systematization or to a list of general categories, or even to a “positive conclusion” (Grolier 1962, 17) since it identified more problems than solutions. De Grolier’s personal wish that his study be at least reused as a basis for future studies was never fulfilled.
3. Recurring topics
The six topics that are at the heart of Éric de Grolier’s work have generally been discussed by the author in more than one paper. In the following sub-sections, the most relevant sources for each topic are referenced in parentheses following the section title.
3.1 System of sciences and classification (1956a; 1970; 1974)
Éric de Grolier believes that the history of classification is inextricably linked to that of the system of sciences, which is the foundation of knowledge and bibliographic classification systems. The author is critical of the strictly utilitarian perspective on classification exhibited by too many librarians. He is convinced that one must necessarily consider more than immediate needs when developing a classification system, if only to better reflect the way scholars, themselves frequent users of classification structures, organize the world. An important section in Théorie et pratique des classifications documentaires (1956a), and an article published on the occasion of the Ottawa Colloquium on the conceptual basis of the classification of knowledge held in 1971 (Grolier 1974), focus on the evolution of the system of sciences. Such evolution depends, on the one hand, on the evolution of the sciences themselves, which gradually replace arbitrary and incomplete theories by comprehensive ones based on hard facts and, on the other hand, on ideological modifications brought by social evolution (Grolier 1956a, 142).
De Grolier considers that the Alexandrian cosmogonies can be seen as the first system of sciences created in ancient times. The mathematical basis of science established around 550 BC by the Pythagoreans will influence all systems to follow, including the Quadrivium (arithmetic, geometry, astronomy, music) in the Middle Ages. Social and medical sciences appear in the same era, and de Grolier believes that there is not a single modern conception of the sciences that has not been seeded in these first proposals. Greece plays an essential role in the classification of sciences, distinguishing natural sciences and humanities as early as the third century BC. The Romans inherit this system and pass it on, barely moodified, to medieval philosophers. It is only during the Renaissance that classification structures built on different bases, the human faculties of memory, imagination and reason proposed by F. Bacon, for example, emerge.
De Grolier is familiar with the discourse on classification attributed to philosophers in the Age of Enlightenment (Siècle des lumières). In his opinion, the significant work accomplished in this area by Leibniz has been overlooked by history, while the most celebrated Diderot and d’Alembert have only made a thin contribution to the evolution of the system of sciences. It is during the 17th century that Descartes and Hobbes finally integrate scientific modernity to knowledge classification, without reference to medieval structures. De Grolier describes the period 1789-1848 and the noted appearance of social sciences in systems developed by Saint-Simon, Comte, Ampère and Hegel. From a classification point of view, the 19th century is effervescent, every philosopher intent on creating their own classification of the sciences.
The quest for the ideal system is certainly less apparent during the 20th century. This is no surprise for de Grolier who provides numerous examples of sciences / → disciplines / concepts that are difficult to classify and make it unfeasible to develop a structure capable of integrating all perspectives and filling every need (1974, 35). The author recognizes that many problems remain to be solved, mostly but not exclusively, in the humanities and in biology (1974, 57).
Not only does de Grolier describe the general characteristics of systems proposed by historical characters whose contribution he considers significant, but he also enumerates the main classes of those systems, adding → notation to reveal hierarchical structures, and commenting on their respective strengths and weaknesses with the support of several examples. This is a meticulous and amazing exercise, but the arguments are not always easy to follow for readers who do not have ready access to the original systems.
When he gets to the history of bibliographic classification, de Grolier specifies that he has considered structures used in encyclopaedias, in archives, in museums, in libraries and documentation centres, in World fairs and exhibitions. In this, he emulates Russian specialist E.I. Shamurin who has created, in de Grolier’s opinion (1969), the most complete history and inventory of bibliographic classification systems as of 1955. De Grolier likes that Shamurin situates classification structures within their social context, something that H.E. Bliss and Sayers were not able to do; like Shamurin, de Grolier believes that classification structures are concrete historical phenomena (1969, 653), and his detailed exposes on the → Bibliotecno-Bibliograficeskaja Klassifikacija (BBK), which describe the system as an accurate testimony on the state of a society at a specific point in history, reflect this personal conviction (1970, 110).
De Grolier recognizes that the development of bibliographic classification systems is lagging behind that of the system of sciences, a phenomenon he attributes to the mass of documents and inertia of institutions (1961). His history of bibliographic classification often takes the form of descriptions of library catalogues in which classification structures that are not any more accessible to us in complete form can be observed. We know little about classification systems applied in libraries in Egypt, Babylonia, Assyria, Greece or Rome, but we know that these libraries were equipped with classified catalogues. From the 9th to the 17th century, catalogues listed religious works before all others. During the Renaissance, printed library catalogues were organized on the model of academic faculties. The first theoretical texts on the classification of books were published in the 16th century, and in the 17th century G. Naudé and N. Clément opened a path to standardization. The Manuel des libraires et de l’amateur de livres, published by J.C. Brunet in 1810, introduced the Libraires de Paris [Booksellers of Paris] system; de Grolier believes that even Dewey and Cutter followed in the tradition installed by Brunet.
De Grolier regrets that some of the most interesting classification structures did not survive; one of them is the system developed by R. Merlin around 1840, conceived as a parallel to the system of sciences. Other examples are O. Hartwig’s Halle Schema, and the structure developed by N.A. Rubakin, which classified by domains and problems rather than according to the science which would study them.
De Grolier describes the classifications systems used over four centuries by the Bibliothèque nationale de France (BnF). His detailed analysis is however of little use today, unless one needs examples of the neglect of natural sciences and overemphasis on humanities that is common to most national libraries. In a note (1956a, 16, note 363), de Grolier suggests that BnF has reached the point where it becomes essential that a new classification system be adopted for printed materials; over the next decades, he will regularly bring back this recommendation, particularly in the context of the project of très grande bibliothèque planned in the 1980s and 1990s. In his opinion, this project was a dreamed opportunity for the national library to lead the way in the revision and improvement of subject access methods and systems in all types of French libraries (Grolier 1988a, 482; 1991b, 249).
In his paper at the Ottawa Colloquium of 1971, de Grolier has widened his perspective to include languages other than classification systems. His knowledge of linguistics allows him to appreciate controlled languages that remain close to natural language, but he does not agree that the latter offers a better performance in information retrieval (1974, 88). He believes that the increasing popularity of → thesauri is not justified; indeed, a few years later (1978, 69), he will state that thesauri are based on the outdated model of information systems based on manual indexing with a controlled vocabulary. He suggests that classification and other indexing languages must be seen as searching tools rather than used strictly for indexing. He also criticizes (1979a, 67) the appropriation by computer specialists of the concept of semantic fields, defined as loose networks, which in his opinion cannot be as efficient as the → logical subdivisions of classification structures.
In these historical essays, de Grolier reveals the extent of his erudition. He has perused a number of treatises and manuals published during different eras, and taken the pain to trace and analyze classification systems which did not leave their mark in history. De Grolier is thus able to compare systems separated by decades or even centuries, and to reconstruct the history of the systematization of science and knowledge. He notes, for example, the many similarities between a virtually unknown classification of arts and sciences developed by professor-engineer J.T. Tykociner and that of Cordonnier, a better known system that he judges viable and potentially efficient (1970, 102-103).
3.2 The science of classification (1956a; 1967; 1974; 1979a; 1982a; 1988a)
Éric de Grolier defines classification as a technique, a practice, an activity. This practice is the focus of taxonomy [2], an interdisciplinary field related to epistemology. Taxonomy has grown its own methods and techniques, but the author estimates that, in 1974, it is still in a developing phase and benefitting from the work of research groups such as the CRG. De Grolier is aware that the information sciences are not the only ones concerned with taxonomy. He has himself studied numerous disciplines in which classification is an object of study, citing philosophy, linguistics, semiology and semantics among others. De Grolier observes that classification problems are also at the heart of terminological research (1988a, 470), and he predicts that it will be impossible for the embryonic field of artificial intelligence to avoid issues related to the systematization of concepts.
De Grolier sees the origins of a theory of bibliographic classification in Naudé’s 1627 Advis pour dresser une bibliothèque. The theory is slow to evolve at first, with the fundamental principle of → bibliographic warrant, for example, defined in the 19th century only. E.C. Richardson’s treatise and Sayers’ canons at the beginning of the 20th century, Bliss’s work and of course → S.R. Ranganathan’s Prolegomena to Library Classification, issued in 1937, lead to more rapid developments. De Grolier believes that his own essay, Théorie et pratique de la classification documentaire (1956a), has an important role to play in the definition of basic concepts and principles, as does a treatise on notation published in 1964 by Dobrowolski. Though intrigued by the CRG proposals relating to a classification by object rather than by discipline, and by the theory of → integrative levels which he associates to Hegel’s theory of emergence, he sees there little more than a complex scaffolding (1965a, 186) while admitting that considerations and propositions of this type are essential to the progress of the science of classification.
De Grolier describes nine perspectives that can be used to study classification and classification systems (1974, 22-25); his explanations are reinforced by examples which allow his reader to appreciate the extent of his knowledge and his talent for multifaceted analysis. These perspectives are: technical, psychological (with Piaget’s writings cited), logical, linguistic (with numerous references to the works of Linné, Lavoisier, Engels, Benveniste, J. Lyons, Sapir, Whorf and Gardin), scientific, pedagogical, → mathematical (exemplified by a research of classification algorithms), sociological and ethnological (with references to Durkheim and Mauss).
De Grolier is familiar with the principal theoretical models; he mistrusts those that appear too smooth, too systematic and too mathematical. While appreciating the genius of Ranganathan, whom he calls a guru, a leader and a mentor (1979b, xxv), and the pertinence of his model, he does not consider said model either completely original or ideal (1965a, 102; 1974, 68). If the idea of using → facets as principles of organization applicable to all disciplines is an interesting one, he demonstrates the impossibility for the facets defined by Ranganathan, and even for the better circumscribed facets later proposed by Vickery, to adjust to domains such as linguistics, → fine arts, philosophy and religion. According to Maniez (1991, 76; 77), this explains de Grolier’s expressed preference for hierarchical structures.
Éric de Grolier approves of the application of statistical methods for the study of classification systems. He notes that Bliss and Pagès have both used statistics to ensure an adequate distribution of available symbols in the creation of their notation. Between 1952 and 1982, de Grolier (1982a, 19) applies statistical methods to contrast the contents of various systems, and to determine the importance given to subjects within the structure as compared to their importance in the system of sciences, in scientific literature, and in society more generally. This type of analysis helps him to distinguish static and dynamic classes, and it is useful to determine whether so-called universal classification systems do reflect cultural differences (Grolier 1956a, 56). Even if he admits (1982a; 1991a) that statistical approaches can be very subjective and scientifically limited, de Grolier uses his findings to imagine the ideal structure of a much needed standardized universal classification system. In a 1982 study, he uses statistical methods to analyze 15 classification systems, including one used around 100 BC, the 14th century Canterbury system, the Libraires de Paris system, the Merlin system used at BnF, the DDC and the BBK. The results illustrate a progression with the passing centuries of the importance given in classification systems to hard sciences and technology, and the corresponding reduction in space occupied by theology and law; indeed, between the beginning of the 19th and the middle of the 20th century, this proportion has been reversed.
De Grolier (1965a) examines problems related to the identification and representation of relations in classification and indexing languages. Because indexing languages are characterized by the number and the diversity of relations they can represent and differentiate, he studies proposals made by Ranganathan, Pagès, Farradane, Gardin and J. Perreault on the topic, contrasting them with the treatment of relations in natural languages. De Grolier does not favour heavy and complex relational networks; in fact, he sees them as one of the main causes of unsatisfactory recall, convinced that the more elaborate the network, the higher the risk of ambiguity, of overlap and of failure in information retrieval (1965a, 208; 1974, 85).
De Grolier wishes to see an expansion of theoretical and applied research on classification, but also an augmentation of the volume of scientific literature in French on the subject (2.1% only in 1988). At the planning stage of the “Très grande bibliothèque” project, the author hopes (1988a, 482) for the development of a new classification system which would integrate theoretical principles and procedures known at the time and would, finally, adapt to the capabilities and requirements of technologies then available. He deplores that a movement towards synthetic systems launched in the 1940s by Desaubliaux, Cordonnier and Dobrowolski has fizzled out and that the DDC and the LCC have become the systems of preference even in France.
According to de Grolier, several factors explain the under-development of the science of classification. Firstly, the fact that proposals for a general theory of classification are often a posteriori rationalizations of existing practices. Secondly, a propensity to construct a theory on the basis of a partial examination of relevant facts and a single point of view. Thirdly, the lack of research in the sociology of classification, even if it is agreed that bibliographic classification systems, even more than philosophical ones, must mirror their time as it is represented in the literature (1974, 78). The author does not share, however, the relativist position adopted by his contemporaries, Pagès for example, who believe that everything is a matter of perspective. Conceding that each society interprets the world using its own value system, de Grolier nevertheless believes that every object has its own reality, independent from all points of view from which it can be considered (1956a, 304); this makes it possible to classify systematically and universally. De Grolier is also convinced that the classification of knowledge must be based on a scientific perspective, and that systematization of elements of knowledge can be achieved by taking into account known connexions between facts verified through observation and experimentation, and more generally by application of scientific methods. De Grolier recognizes that it is difficult for a classification to retain its pertinence through time and suggests that the feasibility of periodically superseding a classification system should be examined.
3.3 Classification systems (1956a; 1970; 1974; 1976)
In his 1956 treatise, de Grolier divides his analysis of classification systems in three sections: universal decimal systems, other universal systems, special classification systems.
3.3.1 Universal decimal systems
De Grolier is no fan of the DDC, judged old and outdated, more empirical than its European counterparts. If the DDC is so popular, it is for social rather than scientific reasons. Indeed, with its apparent simplicity and its resolutely practical character, it has appealed to public libraries which were rapidly developing at the end of the 19th century in the United States and in Great Britain (Grolier 1974, 62). De Grolier also criticizes the young Melvyl for giving no credit to his predecessors Brunet and W.T. Harris, whose influence obviously permeates his classification system.
De Grolier prefers, and by a wide margin, Otlet’s UDC. He notes, albeit reluctantly, that the UDC owes much to the DDC, but considers the former infinitely superior in terms of expression and symbolization. Otlet had recommended, as early as 1933, a complete overhaul of his system; de Grolier makes the same recommendation 20 years later. Indeed, he would like to see the UDC adopted more widely in France, where the system is criticized for its arbitrary character, scientific errors, and radical distance from the classification of sciences. While de Grolier welcomes the integration in the UDC of a system of symbols allowing for the expression of relations between subjects, he recognizes the difficulty of exploiting its full potential. He also admits that the structure is not up to date, which leads him once again to explain the concept of periodization, as it is inefficient to organize books published in the second half of the 20th century with a classification structure dating back to the second half of the 17th century (Grolier 1956a, 187).
The author’s position on UDC remains, however, ambiguous. He criticizes its main components and mechanisms, but suggests that it is high time a new international standard for bibliographic classification be developed, whose constituents should be inspired by the UDC. Paradoxically, the classification structure which he has himself developed shares few formal characteristics with Otlet’s system. De Grolier is mistaken when he states, in 1955, that the DDC and the UDC can only continue to move away from one another and that UDC has a promising future, admitting however that special classification systems are more responsive to evolving needs and rapidly gaining ground. In 1974, he is clearly less optimistic when commenting on the bleak future of not only the UDC, but also of the DDC; time has shown that he misjudged the situation.
3.3.2 Other universal systems
De Grolier’s analyses of non-decimal encyclopedic classification structures, often national systems without any claim to globalization, are his most accurate and informative. He knows how to describe clearly both a little-known Marxist classification recently used in the USSR, and the situation prevailing at the time in the United States, where little theoretical research is conducted but where two systems, LCC and H.E. Bliss Bibliographic Classification compete with the widespread DDC. The author admires the remarkable work accomplished by the Library of Congress while identifying several problems in its classification system, including its evident disdain for the scientific classification of knowledge, its enumerative nature, and its limited practical application; this leads to his mostly accurate prediction that the system will not spread beyond North America. De Grolier believes (1974, 65) that the classification systems developed respectively by Bliss and by J.D. Brown in England (Subject Classification), both reflecting a neo-positivist trend, better replicate the system of sciences of their time; he considers them potentially more efficient than the DDC, the UDC and LCC, but without a future for lack of adequate financial and institutional support.
Ranganathan’s Colon Classification is not a favorite of de Grolier who insists that the limited number of relations between facets is too rigid, that the system’s creator has favored canonical and chronological over scientific divisions, that the symbolism and the notation are needlessly complex. The author compares the various attempts at establishing sets of facets to hairsplitting (1974, 83), but argues that the idea of linking concepts is a viable one and may lead to more satisfactory results in retrieval; this is of course pure speculation on his part since no systematic evaluation of faceted systems will have been conducted in his lifetime.
The situation of classification in France worries de Grolier. He describes in great detail the most original and interesting French classification systems, starting with the excellent system known as Libraires de Paris. He considers that it is L. Delisle, far from deserving of his reputation as greatest administrator the BnF has known, who brings the coup de grâce to this efficient classification system without even proposing an adequate replacement. De Grolier analyzes several French proposals: those of H. Clavier (published in 1936 and 1938), R. Desaubliaux (published in 1943) and Shoemaker (published around 1942); of the latter, Grolier says that it presents itself as a complete theory of classification, but criticizes its mathematical basis. The main quality of another scheme, that of Dobrowolski, lies in its notation; as is the case with the system developed by Pagès as part of his work on coded analysis, de Grolier agrees that it represents an interesting attempt to solve the problem of how to create a classificatory language capable of expressing relations in an appropriate manner (1956a, 248), and suggests that it should be taken into account in the construction of a standardized system.
To reuse the results of his own analyses, de Grolier engages in the perilous exercise consisting in developing his own universal classification system, named ALSYN in reference to its alpha-synthetic notation. De Grolier (1956a) accepts the assumption that there is an order in nature and a general consensus among classifiers: everyone starts to organize on the basis of the most general mathematical properties, considers physical properties before biological ones, and sees humans as the terminal link in the lineage of living beings in the world as we know it. Based on a report written by de Grolier for the FID, Maniez (1991) identifies the proposed classes in ALSYN and evaluates the space they each occupy in the structure: logic, mathematics and dialectics (10%), physics (18%), chemistry (7%), biology (6%), botany and zoology (5%), physical anthropology (10%), sociology, linguistics, history and geography (7%), technology, economics (10%), politics, law, morality, education, philosophy (8%), arts and entertainment (5%) and literature (7%). Although de Grolier dedicates many years to modifying and refining his first draft, the system will never develop sufficiently to allow for comparisons with other structures created at the same time, those of Dobrowolski, Pagès and Gardin for example, and its fair evaluation.
3.3.3. Special classification systems
Special classification systems have appeared during the Renaissance period to serve the needs of emergent field and discipline-based libraries. De Grolier knows of the existence of a very large number of special systems, but since he has not been able to see them and analyze their contents, his comments remain superficial. He believes that these utilitarian structures are much more satisfactory than the so-called universal ones, more up to date and less dependent on obsolete theories and ideologies. Their main weakness is their reliance on the editing work of a single person or at best a small team often working in a vacuum.
3.4 Symbolization and notation (1956a; 1956b; 1956c; 1957; 1962; 1974)
In the early 1950s, Éric de Grolier predicts that the decimal notation adopted by Dewey and retained by Otlet can only lead to a dead end. From that date, he becomes passionate about issues related to the set of symbols available to represent classes and specify relations, and about the notation itself. This is evidenced by the more than 50 pages allocated to these matters in his reports to the FID (Maniez 1991), by the fact that standardization of notation is dealt with before standardization of content in Théorie et pratique des classifications documentaires (1956a), and by the author’s professed admiration for Cordonnier and Dobrowolski who both proposed original solutions to the problem of symbolization (1974, 68-69). De Grolier’s growing interest for various branches of linguistics, and especially phonology, inspires many propositions which will evolve over the years. In his 1962 Study of general categories, he recognizes that his own earlier work is outdated and that the issue must be re-examined, considering both the progress made in the study of language, in phonological research, and on the symbolic aspect of sounds (1962, 165). De Grolier will never question the importance of notation. Indeed, as late as 1988, he persists in suggesting that the issue of symbolization is one important problem that must be solved if more efficient classification is to be achieved (1988a, 482).
The elements to be considered are clear: → expressiveness of notation, length of class numbers, → distribution of class numbers in proportion to the current and future importance of a field of knowledge, → pronounceability of class numbers, relational markers, and → mnemonics (1974, 87-88). De Grolier’s writings on symbolization are simultaneously his most theoretical and his most technical given the depth of his analysis of notational systems devised by others and his own proposals, particularly in the context of his work on ALSYN.
De Grolier is convinced of the superiority of the mixed notation, and maintains his preference for an alphanumeric set of symbols based on the 17 letters of the Roman alphabet which are universally recognizable by their sound (1957, 101-102). In his opinion, the phonological dimension of the notation is unavoidable; the phonological system is fundamental to all languages, and if classification is to be considered a language in its own right, it must possess its own phonological system (1956a, 273). One must therefore be interested in the phonetic transcription of symbols, in the order in which they must be used (although this requirement must be reconsidered in the context of mechanical selection (1957, 83-84)) and in their graphic transcription. Numbers 1-9 are used to identify the characteristics of each domain (forms, functions, types, spatial location, etc.) In ALSYN, de Grolier proposes, for example, a class re Linguistics, subdivided into re1 Methodology, re2 Physiological conditions, re3 Sign systems, re4 Grammatical system, and so on (1956a, 286-287). In addition to letters and numbers, four indispensable symbols mark integration (for example, the slash / in UDC), intersection (for example, the colon : in UDC and the apostrophe ‘ in Cordonnier), complex notion (for example, the dash - in Cordonnier) and form (to distinguish the segment of a class number that specifies a form of presentation from the segment representing class and subject, for example the 0 in DDC).
In recommending such a complex notational system, de Grolier necessarily isolates its semantic function, that of identifying general classes and specific subjects, from its pragmatic function, that of organizing collections of documents. Surprisingly, and although he does recognize the distinction between the semantic and pragmatic functions of classification (1956a, 272), de Grolier never discusses the issue at length, merely suggesting that it is probably impossible to use a single form of coding to all at once store documents, denote their contents, and ensure their subsequent retrieval (1957, 108).
With the aim of getting closer to the new forms of coding required for efficient machine operation, de Grolier estimates the respective benefits of using a totally random notation, a semi-randomly generated notation, or a pre-established systematic hierarchical notation. He observes that the choice of notation must depend on the type of machine used. He has the wisdom to predict that the situation can only evolve very quickly, without imagining, however, that concerns for this matter will not survive the transition to a new millennium. His essays on notation, to which de Grolier has devoted much effort, are now of historical interest only to a handful of specialists.
3.5 Technology and the future of classification (1956a; 1957; 1961; 1965a; 1965b; 1988a)
While enthusiastically supporting the use of information technologies, de Grolier is suspicious of fads and of miracle solutions materializing at regular intervals. He is also wary of the so-called limitations of automation imposed by computer specialists who seem reluctant, if not incapable, to design systems based on completely new models (1965a, 110, note 462). Disregarding the rumours concerning the grim future of paper-based documents and of traditional methods of description, indexing and classification, de Grolier is convinced that in no way do → automatic selection techniques suppress the necessity to use classification systems; he claims on the contrary that information systems can only benefit from the development and application of more detailed and more rational classification structures (1956a, 83-84, 1965b). De Grolier’s vision thus extends well beyond his own time and the first attempts at using computers to manage documents in IS.
But in 1961, de Grolier also admits that classification is at a turning point. His report to the Air Force Office of Scientific Research on Automation Information Retrieval and Machine Translation (1965a) focuses on the logical, linguistic and semantic bases available in such a context, describing theories elaborated by O. Jespersen, E. Cassirer, M.W. Dixon and N. Chomsky. De Grolier addresses briefly the issue of classification, but devotes more space to other languages used for subject representation and retrieval. Gardin’s SYNTOL is singled out. In de Grolier’s view, SYNTOL “represents the most ambitious and the most modern attempt made in France […] for constructing a language incorporating many features considered having a permanent value, for application with machine selection, of the older systems of classification” (1965a, 147). The system is remarkable because it acknowledges that the same concept can be classified differently by various information professionals. De Grolier questions, however, the long-term viability of SYNTOL’s network of very specific categories which overburden its structure.
The impressive report, despite its underdeveloped conclusion and lack of recommendations, demonstrates yet again its author’s understanding of linguistic theory. It also shows that de Grolier’s interests go beyond classification to encompass other types of indexing languages useful in subject searches, such as the thesaurus. The author even recognizes that, in the light of his essay, it can be concluded that UDC does not seem particularly well suited to an automated environment (1965a, 100), a surprising admission by one who has repeatedly expressed his fondness for Otlet’s classification system.
De Grolier never doubts that bibliographic classifications remain useful in information systems. He does not believe that word frequency counts or full text searching are efficient when it comes to distinguishing what is relevant from what is not (1961; 1965a, 2). But he also cautions that the future of classification can only be assured if standardization efforts are undertaken and lead to the development of a single universal classification system. Noticing the growth of interdisciplinarity and multidisciplinarity (1956a, 260-261), de Grolier perceives problems to come, and recommends that classification structures become more flexible.
De Grolier judges that no existing classification system, not even the UDC, could be used as a standard. Consideration must thus be given to the development of a new system. Regardless of the chosen solution, the author is convinced that it is not possible anymore to work in isolation; pooling of resources, cooperation and compromise must be the order of the day while exploring new territories in linguistics and semantics. De Grolier compares the work to be done to the one that has led to the evolution of the system of sciences, accomplished not by the speculation of this or that philosopher, even the greatest one, but by the coordinated efforts of many thinkers and researchers, each adding their own stone to the common edifice (1956a, 143). In 1957, he considers that the time has not yet come to build a new universal classification system. Cogitation and research must progress on issues related to symbolization and relations. Surprisingly, he suggests that special classification systems could serve as a basis, through a process of merging, for the development of a universal structure (1957, 99), a suggestion he will dismiss a few years later (Palermiti 2000).
In the meantime, Éric de Grolier insists that the management of existing systems must be centralized. He favours the creation of an international body and an up to date international descriptive bibliography; in this, and in affirming as late as 1956 that universal bibliographic control can still become reality, he follows closely in Otlet’s footsteps (1956a, 364).
3.6. Classification research (1957; 1965b; 1982b; 1988a; 1991b; 1992b)
Éric de Grolier judges traditional classification systems ill adapted to the new parameters of information retrieval. He suggests avenues for improvement, reminding his reader that IS share this object of study with semantics, anthropology and artificial intelligence among other fields.
De Grolier’s ability to synthesize clearly, in writing, the work accomplished during scientific meetings is remarkable. In these texts, the author begins by summarizing the progress that have been made, as described by various speakers, before clarifying with specific recommendations what needs to be done so that progress can continue. Cochrane (1991, 78) sees these summaries as coherent, succinct, lucid… if sometimes biased; they are a valuable source for IS historians who can use them to gauge progress over time, and to determine the significance of individual contributions.
De Grolier considers that the scientific study of classification begins at the end of the 2nd World War, when philosophers and classifiers start to question certain principles and assumptions, as well as the usefulness of existing classification systems (1957, 81). Such reservations are a step in the right direction, and the author urges his colleagues to pursue their evaluation and standardization work. In 1957, at the end of the 1st International Study Conference on Classification Research (ISCCR), held in Dorking, de Grolier confirms that the objectives of the meeting have been reached: national barriers have been crossed, and consensus identified. Indeed, he observes that there is agreement on the relevance of a multidimensional (or faceted) classification system, on the primacy of the criterion of flexibility, on the need to associate the study of classification, of languages used for automatic selection and of control of terminology, and finally on the priority to be given to the identification of general categories and essential relationships in a classification system. He then suggests research avenues, including an in-depth examination of certain traditional principles, such as collocation, hierarchy and specificity. De Grolier warns against going too fast and doing sloppy work, a barely disguised reference to the work of the CRG, but also, conversely, against excessive caution leading to inertia, aiming now at his French colleagues who are quick to criticize without offering workable solutions (1957, 83-84).
In 1964, at the end of the 2nd ISCCR, de Grolier is pleased to note that funds have been made available for IS and classification research, thanks to critical information needs related to national defense. He insists that the expansion of technologies and the influence of emerging fields, such as machine translation, are greatly beneficial to classification research. De Grolier estimates that the conditions are in place for taxonomy to be given the status of discipline. He describes precisely the relation between classification and language, and reminds his colleagues that several active fields of research deserve to be explored, including logic, semantics, semiotics, phonology and cybernetics. But numerous problematic issues remain: the feasibility of developing a common documentary language adaptable to all information processing environments, the distinction between what can be standardized and what should not be, the desirability of creating a model for the description of general classification and information retrieval systems (de Grolier believes that this model must combine logical and linguistic components) and the critical need for testing and evaluation techniques (1965b).
The 4th ISCCR, held in 1982, allows de Grolier to insist again on the necessity for taxonomy to acquire the status of science. To this end, it is essential that empirical research programs on user behavior, quantitative analysis of classification systems and relations between classes and between disciplines be developed. Practitioners, researchers and theorists must find ways to take advantage of information technologies without being dominated by them. They must integrate the concepts of universalism and interculturality, and combat dogmatism. De Grolier recommends that the following principles be accepted by all: the complementarity of natural and indexing languages, of universal and specialized classification systems, of alphabetical and systematic order and of paradigmatic and syntagmatic relations (1982b, 162-167).
The summary written by the Grolier at the end of the 1st International Conference of the International Society for Knowledge Organization (ISKO), held in 1991, testifies to the maturing thoughts of someone whose passion for classification has now been developing for decades. De Grolier replaces classification in the broader context of information retrieval systems, and maintains that classification systems must continue to be used for information representation and research. He pays attention to other types of documentary objects (archives, museum objects, etc.) that must also be classified and made accessible. The problems he raises are not restricted to the study of classification, but he can explain their impact on taxonomy. The first problems mentioned, as was already the case in 1982 (1982b), concern the use of technology, in the face of which de Grolier sees two potential reactions: the adaptation of traditional systems conceived in the 19th century so that they become exploitable in a mechanical context, or the short or medium term disposal of these instruments. In his opinion, it would be premature to abandon the old systems (DDC, UDC, LCC) immediately, but the need is there to plan their replacement by a flexible structure adapted to a very different environment; de Grolier acknowledges the difficulty of developing such a structure. The author reiterates the need to focus more on what is happening outside IS, and to seriously consider the possibility of reusing concepts and techniques used in the fields of conceptual analysis, semiology, structural linguistics and terminology. To finance theoretical and experimental research, he suggests turning to international organizations and applying for grants available in the humanities and social sciences (1991b, 248-250). To show the way, de Grolier identifies gaps, shortlisting topics not yet addressed or whose discussion by the Conference’s speakers appears incomplete or unsatisfactory: user behaviour, probabilistic methods, and communication (or lack thereof) between national “schools” interested in taxonomy (the Indian, British, Russian and French schools); concerning this last point, de Grolier expresses the hope that ISKO may provide a fertile ground for exchanges, in particular through its support of transnational projects (1991b, 250-251).
At the 5th ISCCR, also held in 1991, Éric de Grolier is the only registrant who was also a participant in the four previous meetings (held respectively in 1957, 1964, 1975 and 1982). The summary and recommendations offered on this occasion reflect his growing curiosity for other domains of the social sciences, especially terminology; he describes his personal involvement in these fields, particularly in the framework of the UNESCO INTERCOCTA project (1992a, 225). De Grolier observes that the obsolescence of traditional KO systems is accelerating, not only under the effect of technological innovations, but also as a result of the rapid evolution of all sciences in the second half of the 20th century. He believes that the re-examination of ideologies and theories, combined with the gradual disappearance of boundaries between disciplines, are among the causes of the diminishing relevance and effectiveness of classification systems (1992b, 419). In what will turn out to be his last published synthesis, de Grolier recommends that classification research focus on the use of information technologies, the construction of classification structures and language tools capable of improving communication between different information systems (switching languages), the search for devices able to prevent ethnocentrism, the integration of classification and terminology research, the study of logico-mathematical models and numerical taxonomy, and the analysis of systems of notation (1992b, 419).
De Grolier reaffirms the need to develop a new standardized universal system for the classification of documents, a recommendation he has been reiterating for decades. A paper by the author at the 1991 ISCCR is in fact devoted to this universal system, which would of necessity be combinatorial. De Grolier describes previous and current attempts at developing a new system, explains the possible causes of their inefficiency, and lists the conditions to be fulfilled and the dozen or so problems remaining to be solved before such system can see the light of day (1992a, 228-232). It is interesting to note that de Grolier now seems to doubt the existence of a set of categories and relationships applicable to all fields of knowledge (1992a, 231), 30 years after the publication of his study on this very subject (1962).
In the course of his career, de Grolier did not discuss research exclusively in summaries prepared at the end of scientific meetings. Indeed, he provides in most of his papers a list of research topics which have been of interest to him throughout his career, many of which are still relevant today: the relationship between the classification of macro-documents (monographs) and micro-documents (articles, reports, etc.), the adaptation of classification systems to different types of libraries and information centers, the relationship between universal and special classification systems, and the problems associated with reclassification made necessary by the evolution of the system of sciences (1988a, 483).
4. Conclusion
Éric de Grolier’s position in the narrow circle of francophone pioneers of documentation and IS is undoubtedly well-deserved, if only by the diversity of his interventions and the volume of his writings. His contribution to the theory and practice of bibliographic classification is undisputable, though it cannot compare, in terms of influence, to that of Otlet and Ranganathan. De Grolier was an observer and an analyst rather than a theoretician or a developer. In dense texts taking the form of widely encompassing literature reviews, he recognized original ideas, problems, controversies and contradictions in what was presented by others, before submitting his own conclusions and propositions which he never hesitated to correct or withdraw in the light of new information. Having suggested in 1957 (p. 99), for example, that it was possible to create a universal system by merging special classification structures, he admitted later that this was a bad idea (Palermiti 2000). Towards the end of his career, he also expressed doubts that there existed one single list of general categories and relations applicable to all fields of knowledge (1992a, 231), a position he had developed thirty years earlier in a major essay (1962) and supported since. One could hardly blame de Grolier for these changes of direction, since they bear witness to the thoroughness of his thinking and accurate perception of the context, but they do result in a lack of precision in his definitions of key concepts; in his work, it is somewhat difficult to clearly differentiate classification of objects, of concepts and of subjects, facets, classes and categories, and even classification and coding.
De Grolier did not view himself as a researcher but he believed in the necessity of experimentation in the search for scientific truth; his approach was described as cautiously experimental by Maniez (1991, 73). He relentlessly criticized existing classification systems, but always in an informed fashion, providing required justifications and numerous examples. In de Grolier’s opinion, one needed to fully understand existing systems before launching new projects, the development of a single classification system for all domains of knowledge, for example, a possibility in which he appears to have believed till the end of his life. It was also imperative for him to observe what was going on in other disciplines, in various sub-fields of linguistics, for example, in order to improve IS processes and tools.
Blanquet (2008) suggests that the analysis of forgotten classification systems, brought back from the distant and more recent pasts by de Grolier, may be his greatest contribution to the field of KO. De Grolier’s papers help us perceive the sheer volume of thinking on classification and the number of projects in subject access that were conducted between 1940 and 1970, as automated information systems were gaining ground. Paradoxically, a large segment of this important production is neglected nowadays, when the lack of theoretical foundations and models considered essential for the standardization and efficiency of our practices remain a major concern.
Éric de Grolier was a visionary, well ahead of many of his French contemporaries in his thinking. He was lucid when he described the changing function of the classification system which he eventually presented as a searching instrument rather than as a tool used strictly to organize documents. Concerned with the obligation to meet users’ needs and to take their behaviour into account in systems development, he went as far, in 1965 (1965b, 12), as to suggest that users’ requests could be used as one basis for classification; he never came back, unfortunately, on this intriguing idea.
But de Grolier also made mistakes and many of his original proposals did not withstand the passage of time, the evolution of technologies and the digital revolution. The importance he gave to symbols and notation, for example, remains difficult to explain. Pretexting the necessity to align classification with the coding system allowing for the mechanical processing of information (on punched cards at the time, let’s not forget!), he expended vast amounts of time and energy in the analysis of theoretical propositions that were clearly going against the needs of librarians and information users, suggesting that an arbitrary system of notation (that of UDC, for example) be replaced by another arbitrary system even less connected to natural language, as in his own ALSYN system. If de Grolier was among the first to recognize the practicality of using technology to process information, he gave surprisingly little attention to the wave of technological changes that was already forming at the end of the 1980s, a wave that would confirm the supremacy of content (the organization of concepts) over form (the notation) in the classification system. Neither did he perceive that problems associated with the “weight” of library catalogues, as described in the 1950s (1956a, 69-70), would be of little significance in catalogues maintained and queried online.
Éric de Grolier was craving communication and knowledge, and he would be comfortable in the hyper-connected, hyper-performing and constantly evolving information environment of the 21st century. Most likely surprised to see that DDC, for better or for worse, remains the standard in bibliographic classification, he would approve of the popularity of facets as structuring and searching tool on the Web. He would observe, as he had predicted, that information and not documents is now at the heart of librarians and information specialists’ preoccupations. And he would most certainly persist in claiming that the classification specialist will one day be “master of the world” (1957, 85), as he did six decades ago in Dorking.
Endnotes
1. The Committee was chaired by S.R. Ranganathan and co-chaired by de Grolier, who had stated the need for such a committee as early as 1932.
2. In papers written in French, de Grolier uses the term taxilogie rather than taxonomie (1974, 21; 1988a, 468).
Selected works by de Grolier
Grolier, Éric de. 1941. La formation professionnelle des documentalistes et des bibliothécaires: rapport pour l’Union française des organismes de documentation. Paris: UFOD.
Grolier, Éric de. 1943. Propagande et opinion publique aux États-Unis: étude de psychopolitique. Paris: Institut d’études politiques et sociales.
Grolier, Éric de. 1954. Histoire du livre. Paris: Presses universitaires de France.
Grolier, Éric de. 1956a. Théorie et pratique des classifications documentaires. Paris: Éditions documentaires industrielles et techniques.
Grolier, Éric de. 1956b. “Nouvelles recherches sur la symbolisation des classifications documentaires”. Revue de la documentation 23: 13-21.
Grolier, Éric de. 1956c. “Symbolisation normalisée d'une classification encyclopédique internationale”. Revue de la documentation 23: 128-139.
Grolier, Éric de. 1957. “Concluding Survey”. In Proceedings of the International Study Conference on Classification for Information Retrieval, Held at Beatrice Webb House, Dorking, England, 13th-17th May 1957. London: Aslib, 81-85.
Grolier, Éric de. 1961. “Points de vue rétrospectif et prospectif dans la classification”. In The Sayers Memorial volume: Essays in Librarianship in Memory of William Charles Berwick Sayers. London: CRG, 96-119.
Grolier, Éric de. 1962. A Study of General Categories Applicable to Classification and Coding in Documentation. Paris: UNESCO.
Grolier, Éric de. 1965a. On the Theoretical Basis of Information Retrieval Systems: Final Report, Washington, D.C.: Air Force Office of Scientific Research.
Grolier, Éric de. 1965b. “Current Trends in Theory and Practice of Classification”. In Classification Research: Proceedings of the Second International Study Conference, Held at Hotel Prinz Hamlet, Elsinore, Denmark, 14th to 18th September 1964. Copenhagen: Munksgaard, 9-14.
Grolier, Éric de. 1967. “Synoptic Critique: Proceedings of the International Symposium on Relational Factors in Classification”. Information Storage & Retrieval 3: 385-397.
Grolier, Éric de. 1969. “L'histoire des classifications: une interprétation soviétique. Compte rendu de Shamurin, 1964-1967”. Bulletin des bibliothèques de France 14: 652-657.
Grolier, Éric de. 1970. “Quelques travaux récents en matière de classification encyclopédique”. Bulletin des bibliothèques de France 15: 99-126.
Grolier, Éric de. 1974. “Le système des sciences et l’évolution du savoir”. In Les fondements de la classification des savoirs: Actes du Colloque d’Ottawa sur les fondements de la classification des savoirs, du 1er au 5 octobre 1971. Munchen: Verlag Dokumentation, 20-119.
Grolier, Éric de. 1976. “Classification 100 Years after Dewey”. UNESCO Bulletin for Libraries 30, no. 6: 320-329.
Grolier, Éric de. 1978. L’organisation des systèmes d’information des pouvoirs publics. Paris: UNESCO.
Grolier, Éric de. 1979a. “In Search of an Objective Basis for the Organization of Knowledge”. In Ordering Systems for Global Information Networks: Proceedings of the Third International Study Conference on Classification Research, Held at Bombay, India, during 6-11 January 1975. Bangalore: S.R. Ranganathan Endowment for Library Science, 64-73.
Grolier, Éric de. 1979b. “A Tribute to Dr Ranganathan”. In Ordering Systems for Global Information Networks: Proceedings of the Third International Study Conference on Classification Research, Held at Bombay, India, during 6-11 January 1975. Bangalore: S.R. Ranganathan Endowment for Library Science, xxv.
Grolier, Éric de. 1982a. “Classifications as Cultural Artefacts”. In Universal Classification: Subject Analysis and Ordering Systems: Proceedings of the 4th International Study Conference on Classification Research, Augsburg, 28 June-2 July 1982. Frankfurt: Indeks Ver1ag, 19-34.
Grolier, Éric de. 1982b. “Synthesis of the 4th FID/ICR Conference”. In Universal Classification: Subject Analysis and Ordering Systems: Proceedings of the 4th International Study Conference on Classification Research, Augsburg, 28 June-2 July 1982. Frankfurt: Indeks Ver1ag, 163-167.
Grolier, Éric de. 1988a. “Taxilogie et classification : un essai de mise au point et quelques notes de prospectives”. Bulletin des bibliothèques de France 33: 468-483.
Grolier, Éric de. 1988b. Glossaire des relations interethniques, version provisoire. Paris: UNESCO.
Grolier, Éric de. 1991a. “Some Notes on the Question of a So-called "Unified Classification"”. In Tools for Knowledge Organization and the Human Interface, Vol.2: Proceedings of the lst International ISKO Conference, Darmstadt, 14-17 August 1990. Advances in Knowledge Organization 2. Frankfurt: Indeks Verlag, 85-l08.
Grolier, Éric de. 1991b. “Conference Summary”. In Tools for Knowledge Organization and the Human Interface, Vol.2: Proceedings of the lst International ISKO Conference, Darmstadt, 14-17 August 1990. Advances in Knowledge Organization 2. Frankfurt: Indeks Verlag, 248-251.
Grolier, Éric de. 1992a. “Towards a Syndetic Information Retrieval System”. In Classification Research for Knowledge Representation and Organization: Proceedings of the 5th International Study Conference on Classification Research, Toronto, Canada, June 24-28, 1991. Amsterdam: FID, 223-234.
Grolier, Éric de. 1992b. “Summary and Conclusions to the Conference”. In Classification Research for Knowledge Representation and Organization: Proceedings of the 5th International Study Conference on Classification Research, Toronto, Canada, June 24-28, 1991. Amsterdam: FID, 419.
Grolier, Éric de, and Sylvie Fayet-Scribe. 1996. “La passion de l’organisation des connaissances: entretien avec Éric de Grolier”. Documentaliste – Sciences de l’information 33: 287-293.
Secondary sources and references
Blanquet, Marie-France. 2008. Éric de Grolier: une vie bien remplie. https://www.reseau-canope.fr/savoirscdi/...
Cochrane, Pauline Atherton. 1991. “Éric de Grolier: the Analytico-synthetic Summarizer”. International Classification 18: 78-86.
Hudon, Michèle. 2020 « Eric de Grolier et la classification » Cahiers du numérique 16, no. 1 : 73-103.
Maniez, Jacques. 1991. “A Decade of Research in Classification”. International Classification 18: 73-77.
Palermiti, Rosalba. 2000. Vers des précurseurs: Robert Pagès, Éric de Grolier, Jean-Claude Gardin, Communication au groupe de travail Théories et Pratiques scientifiques (TPS) de la SFSIC, le 3 mars 2000. http://www.iut2.upmf-grenoble.fr/RI3/Mise_jour_06/TPS_precurseurs.htm
Vickery, Brian C. 1991. “Éric de Grolier’s Big Book on Classification”. [Review of “Théorie et pratique des classifications documentaires »]. International Classification 18: 170.
Visited times.
Version 1.0 published 2020-03-11, last edited 2021-02-09
Article category: Biographical articles
This is a translated and revised version of Hudon (2020).
©2020 ISKO. All rights reserved.