ADVANCES IN INFORMATION SCIENCE
The French conception of information science: “Une exception française”?
The French conception of information science is often contrasted with the Anglophone one, which is perceived as different and rooted mainly in Shannon's mathematical theory of communication. While there is such a thing as a French conception of information science, this conception is not totally divorced from the Anglophone one. Unbeknownst to researchers from the two geographical and cultural regions, they share similar conceptions of the field and invoke similar theoretical foundations, in particular the socio-constructivist theory. There is also a convergence of viewpoints on the dual nature of information science, i.e., the fact that it is torn between two competing paradigms—objectivist and subjectivist. Technology is another area where a convergence of viewpoints is noticeable: Scholars from both geographic and cultural zones display the same suspicion toward the role of technology and of computer science. It would therefore be misleading to uphold the view that Anglophone information science is essentially objectivist and technicist while the French conception is essentially social and rooted in the humanities. This paper highlights converging analyses from authors based in both linguistic and geographical regions with the aim to foster a better understanding of the challenges that information science is facing worldwide and to help trace a path to how the global information science community can try to meet them.
Aims and Scope
The objective of this review article is to contrast the French conception of information science with the Anglophone one. It is not an exhaustive account of views held by French and Anglophone scholars on information science; rather, it focuses on areas of convergence. Indeed, there is a widespread belief among members of the academic community in France that the Anglophone conception of information science is very different from theirs, in that it is rooted mainly in Shannon's mathematical theory of communication (Fondin, 2005). I have sought to point out that while there is such a thing as a French conception of information science, this conception is not totally divorced from the Anglophone one. Perhaps, unbeknownst to researchers from the two geographical and linguistic regions, they share the same concerns on the lack of a coherent body of theories underlying research in the field, on the lack of visibility in the field, and on how the discipline should position itself with regard to other neighboring disciplines such as communication, semiotics, sociology, linguistics, and computer science.
The rejection of the domination of the physical or object paradigm is another area of convergence. The emphasis on the social aspects of sense making and on the systemic-constructivist approach to information–communication problems are two other points of agreement. There is also a convergence of viewpoint on the dual nature of information science, i.e., the fact that it is torn between two competing paradigms—one that is objectivist-systems-driven and another that is subjectivist-human-oriented (Bates, 2005; Buckland, 1999; Cronin, 2008; Fondin, 2001; Robertson, 2008; Saracevic, 1999). This can be seen in the fact that the name of the field oscillates between the singular form—“La science de l'information”—and the plural form—“Les sciences de l'information”. In English, the field is also either referred to as “information studies” or “information science.” Technology is another area where a convergence of viewpoint is noticeable: Scholars from both geographic and linguistic zones display the same suspicion toward the role of technology and of computer science (Davallon, 2004; Hjørland, 1998; Hjørland and Albrechtsen, 1995; Jeanneret & Ollivier, 2004).
After reviewing the origins of the discipline in France, I will describe how the cohabitation of information science with communication science has affected the evolution of information science in France and how it is perceived there. I will then look at the epistemological question raised by the use of the plural form to refer to the name of the field. Next, I examine the rapport of information and communication science in France with technology and computer science before offering some perspectives for the future. While France provides the background for this study, comparison with Anglophone scholarship will be made wherever applicable. This study is animated by the conviction that pointing out the convergence in views may create a framework for a better understanding of the challenges that information science is facing worldwide and so help trace a path to how the global information science community can try to meet them. Some of the issues discussed in this article were raised in the author's professorial thesis,1 written in French (Ibekwe-SanJuan, 2010).
Origins of Modern-Day Information Science in France
It was the foundational works of historic figures such as Paul Otlet and Henri La Fontaine in the late 19th century and Suzanne Briet in the first half of the 20th century that laid the foundations for what would later become information science in the 1970s. In the first half of the 20th century, French library and documentation associations had close ties with Otlet's International Institute of Bibliography (Fayet-Scribe, 2000). Other types of research carried out by 20th-century pioneers, however, were needed in order to make the link with modern times, i.e., with the exponential growth in production of printed documents from the 1960s onward, the start of mechanization and automation of documentary processes.
On its way to academic recognition in France, information science has traversed four periods, each with its own body of work but with little or no connection to each other:
- Period 11895–1950, corresponding to Paul Otlet and Suzanne Briet's foundational works on bibliography, classification systems, documentation, and documents.
- Period 21950–1974, corresponding to Jean-Claude Gardin's, Eric de Grolier's, and Robert Pagès's works, which saw a marked shift in focus from bibliographic-level analysis to content analysis and the automation of documentary processes.
- Period 31975–2000, corresponding to the official recognition of information and communication studies (ICS) as a single interdiscipline, these were the first two decades of existence of information science in which contact was lost with the pioneering works in bibliography and knowledge organization and scholarly focus shifted to applied work in artificial intelligence (AI) and information processing to the detriment of theoretical research.
- Period 42000 to the present. Owing to intellectual pressure from communication science scholars, as well as the retirement of the first generation of information science (IS) professors whose primary intellectual point of reference had been the “hard” sciences, IS is witnessing a repudiation of the physical system-oriented paradigm, of information retrieval (IR), and of technologically oriented research in general. The current mindset is a powerful swing back to the social sciences from whence ICS originated.
Pioneering Work on Knowledge Organization and Information Retrieval in the Post-WWII Era
Salaün (1993), then Palermiti and Polity (2002), have given interesting accounts of the pioneering work that prepared the ground for the transition from bibliography to documentation and then to information science in the post-WWII period, i.e., from Briet's time at the French National Library (1924–1954) to the mid-1990s. I summarize here the focal points of these studies.
The development of new information practices in the late 1950s that would eventually coalesce into what we now know as information science, on the one hand, and as computer science, on the other, led a group of researchers to shift the focus of study from the form or container (documents, books) to the contents of documents (indexing and retrieval). The study of the former was left to library management (“bibliothéconomie”) and to librarians. In this early period, research on information-related topics was carried out mainly by scholars from other fields. Robert Pagès (1919–2007) and Jean-Claude Gardin (1925–) were from the social sciences.2 Gérard Cordonnier (1907–1977) was a brilliant mathematician who came to be interested in problems of documentary information classification and retrieval. These researchers were faced with knowledge organization problems in their own disciplines. Eric de Grolier was one of the rare pioneers to come from documentation. All these researchers were born in the first quarter of the 20th century. Much of their research was supported by national institutions like the Centre National de Recherche Scientifique (CNRS), the Union Françaises des Organismes Documentaires (UFOD), the Ecole des Hautes Etudes en Sciences Sociales (EHESS), and by international bodies like the United Nations Educational, Scientific and Cultural Organization (UNESCO).
A pioneer in documentation and knowledge organization, Eric de Grolier and his wife Georgette de Grolier were advocates for public libraries and the promotion of reading. They were instrumental in fostering a closer relationship between library and documentation associations before and after World War II (WWII). De Grolier was also instrumental in setting up the first training courses for documentalists at the UFOD in 1939. He laid the foundations of a subject heading system which later evolved into the well-known RAMEAU (Répertoire d'Autorité-Matière Encyclopédique et Alphabétique Unifié) system. He sought ways to make specialized classification languages more compatible through the normalization of classification schemes. In today's language, we would call this a programmatic effort to achieve interoperability. Despite Suzanne Briet's and the de Groliers' efforts to foster a close relationship between librarians and documentalists, a breach between the two professional bodies was to occur in the post-WWII era that prevented any fruitful collaboration. Both bodies (librarians and documentalists) had enjoyed a fruitful collaboration in the period between 1895 and 1944 (Fayet-Scribe, 1997). Consequently, documentalists and researchers working on information-related problems tended to disregard classification schemes in favor of research on thesauri whose influence was attributed to North America. This, in turn, resulted in a lack of research on classification schemes from the 1960s onward. Ranganathan's facet classification was hardly implemented in France. While the arrival of computation and computers was seen elsewhere as giving a new lease to classification research—in the UK, for instance, the Classification Research Group (CRG) flourished—this was not perceived as an opportunity in France. Classification schemes were traditionally used for organizing books on library shelves. French documentalists and researchers did not perceive at the time that they could become a means for searching documents online.
Robert Pagès was a social psychologist with a major in philosophy. He became interested in documentation as early as the 1940s and this interest intensified when he entered the CNRS in 1951. He had observed the lack of knowledge organization systems of his time (rigidity of universal schemes, incompleteness and inadequacy with regard to specific fields like social psychology). Critical of the principle underlying universal classification languages such as Otlet's Universal Decimal Classification (UDC) and the Dewey Decimal Classification (DDC), whereby objects are only seen from one dimension, Pagès (1955) advocated an n-ary dimensional analysis of contents (documents). He suggested that “documentation” should be a larger specialty subsuming library management because the latter was about books whereas the former was about documents, a category that subsumed books. Pagès wanted to study the relation between documents, books, and experiences and called for the introduction of psychology into the study of documents. His idea was that documents are made of signs and symbols that are subject to interpretation. These symbols acquire meaning outside of their context of production. Thus, a document is an instrument for accumulating symbolic activity. He studied other types of symbols like mathematical language and advocated the grounding of documentary classification languages on a formal scientific basis. This led him to create an analytic representation code for documents called “coded analysis” (“analyse codée”) or CODOC, which went into operation in 1954 in his Center for Documentation at the Sorbonne. Pagès's CODOC system was also used on Gérard Cordonnier's SELECTO cards. The CODOC system was inspired by the functioning of natural language, symbolic logic, and algebra. The idea was to design an extensible grammar and a lexicon that would enable the creation of new and unexpected classes, thus giving an infinite possibility for subdivision while being easy to memorize (mnemotechnical faculty). However, the result was an artificial language for indexing and classification that was inhospitable to memorization. Pagès worked on the normalization of specialized classification languages. Clearly, for Pagès, the focus was on analyzing and retrieving content, not organizing books. His idea of the nature of a document was not far removed from Briet's own wide conception. Also, his idea of a multidimensional approach to document content analysis was visionary given the rudimentary information indexing and retrieval systems available in his time.
Jean-Claude Gardin, the third post-1950 pioneer, is a versatile scientist who majored in political economy, history of religions, linguistics, and archeology. He became interested in information-centric problems after he was recruited as a researcher at the CNRS in 1950. He was confronted with the problem of sorting and comparing archeological objects referred to in scientific texts. He took note of the very little work that existed on retrospective research and the absence of a “répertoire” (catalog) of previous works done in a field. Judging the analytic compilation of previous work to be an important component of scientific research, he sought ways to reduce their labor-intensive nature and to systematize the conceptual analysis of the contents of scientific communication. It was in this context that he designed SYNTOL (Syntagmatic Organization Language) in 1964, a sophisticated system for facet analysis, indexing, and IR (Gardin, 1964). He also did much research on discourse analysis, i.e., the structure of scientific discourse with a particular focus on archeology.
The CODOC and SYNTOL systems share some similar traits: Each was designed by a CNRS researcher from the social sciences interested in scientific information representation and retrieval. Both systems sought to provide formal languages for content representation. They signaled a shift from bibliographical analysis to content analysis. They aimed to provide better access to contents of scientific publication by enabling a multidimensional and combinatorial approach to IR. This paved the way for research on formalisms to automate content analysis. Gardin and Pagès's work also formed the basis for some of the research on automating information systems conducted in the early 1990s by French computer scientists and first-generation information science scholars who hailed from the sciences. Gardin was very critical of the emerging fields of natural language processing (NLP) and artificial intelligence (AI). He was especially skeptical of the claim made by scholars therein that there could be a universal semantic representation of discourse. He argued that such methods could only work in relation to microdomains. History has since proved him right. Pagès was equally visionary in his defense of a science of documentation that he called “documentology” and that was part of symbolic communications, grounded in the humanities and populated with “researchers-cum-documentalists.” This, in his view, would ensure that documentation would not be reduced to a set of techniques aimed at solving practical problems. The term “documentology” would be unsuccessfully taken up later by Jean Meyriat (1983), one of the founding fathers of information science in France, in his attempts to find an adequate name for the discipline. Unfortunately, Pagès's vision did not come to fruition. The majority of works carried out in the information science field following the official recognition of the field were largely rooted in the system-driven paradigms of IR and cognitive science. Very little space was accorded to psychology and to sociology.
While the works of Otlet, La Fontaine, Briet, Pagès, Gardin, and de Grolier (among others) laid the theoretical foundations for the emergence of a French information science, the official existence of the field in the French higher education system did not come about until 1974. Information science was not recognized as a distinct discipline but as an interdiscipline merged together with communication science. The concept of interdiscipline means here that information and communication were considered as common or shared objects, at the crossroads of several disciplines. More details of the circumstances leading to the emergence of this interdisciplinary field are given in the next section. It is important to observe that the nascent French information science was not grounded in the foundational works by these early pioneers in the post-WWII era. Indeed, none of the pioneering figures in documentation and content analysis were associated with the emergence of the field. However, the first group of professors in information science were inspired by Gardin's works, especially the SYNTOL system and his discourse analysis. Unfortunately, Gardin's work, after receiving some echoes in the 1990s, fell into obscurity. Palermiti3 offered three possible reasons for this:
- These authors (Gardin, Pagès, de Grolier, Cordonnier) published in a pre-paradigmatic era, i.e., at a time when the discipline did not officially exist and no explicit paradigm was established.
- Apart from Eric de Grolier, the others belonged to other disciplines. However, I note that this is not a unique feature of French information scientists nor is it specific to the discipline. In other parts of the world many prominent information scientists began their scholarly careers in other disciplines.
- What would later coalesce into information science was not yet a scientific object of research but was mainly a body of professional practice (documentation). Training courses at the time were essentially practical in their orientation. Little or no research was devoted to theoretical and historical issues in information science. Doctoral programs at the time were also mainly concerned with solving practical problems pertaining to indexing and retrieval.
As Palermiti (2000) observed, it was the development of computer science and the Internet in the early 1970s that ironically overshadowed these early works. The focus of research had shifted to automatic translation, the development of expert systems, and understanding natural language. The consequence for French information science was that traditional documentalist research on knowledge organization was forsaken in favor of research models coming from the computer and AI communities where the focus was on information processing.
Also, very little work has been done on the theoretical foundations of information science and on knowledge organization in France (Polity, 1999). Hence, there appears to have been a disconnect between the different bodies of work leading up to the emergence of the field.
The Birth of a Discipline: The “Péché Originel”4
Several accounts have been given of the intricate web of events that led to the creation of what is known today as information science in France (Boure, 2002; Escarpit, 1991; Le Coadic, 1994; Meyriat, 1993; Palermiti & Polity 2002; Tétu, 2002). The birth of a new interdiscipline called “information and communication sciences”5 (ICS) was the work of a committee on ICS that later became the French Society for Information and Communication Sciences (FSICS6). The creation of this discipline was the result of three types of pressures.
First, France needed to develop its own information infrastructures (servers, databases, scientific and technical document processing) and thereby gain independence from the United States. Ministerial policies were thus focused on only one type of information: scientific and technical information. Second, there was pressure to professionalize training in information technology. Third, some professors interested in issues of communication wanted a brand-new discipline where they could expect better career prospects (Palermiti & Polity, 2002). The ICS was then carved out of existing humanities disciplines in 1974 thanks to institutional lobbying by three prominent figures: Roland Barthes, Robert Escarpit, and Jean Meyriat.
Roland Barthes was a renowned writer, a semiotician, a literary critic, and a philosopher. Robert Escarpit was also a writer and journalist before coming to communication studies. He was one of the first French literary scholars to raise the question of the role of the reader and to consider the literary act (writing) as a communication act. He first used the word “communication” in relation to literary writing in 1958. He was also the recipient of the first Chair in Comparative Literature in the French higher education system (Tétu, 2002).
Jean Meyriat7 studied political science before coming first to documentation and then to information science. Other prominent figures who took part in the committee for ICS were Algirdas Julien Greimas, founder of the most important semiotic school in France, and Oswald Ducrot, who imported Austin's speech acts theory into France and was also a pioneer in linguistic pragmatics. However, most of these prominent figures (apart from Escarpit and Meyriat) did not officially leave their discipline of origin to join the emerging ICS discipline; rather, they worked on communication problems from within their own disciplines (sociology, semiotics, linguistics). Hence, the creation of the field of ICS was not the result of a consensus on its objects, theories, and paradigms but rather an opportunistic coming together of professors who were interested either in communication science or in documentation but from the perspective of their own fields. This would have lasting consequences on the theoretical grounding of the field.
Given the backgrounds of its founding fathers (Escarpit, Meyriat, and Barthes), the French interdiscipline of ICS can be said to be born of literary origins.
The fact that information science was not recognized as a distinct discipline but rather as an interdiscipline merged with communication science, and that none of its pioneering figures were associated with its emergence, also had significant repercussions for the evolution of the field. With the notable exception of Jean Meyriat, the research carried out in the early 1980s by the prominent literary professors who are considered the founding fathers of the ICS discipline had little or nothing to do with library management, classification schemes, documentation, and content analysis.
It is also worth mentioning that the CNRS, which had funded Gardin's and Pagès' works on designing prototype indexing and retrieval systems has, to this day, steadfastly refused to create a section on ICS within its own administrative structure. Given that the CNRS is the major French research institute, its recognition of ICS would have given information science its “lettres de noblesse”—an acknowledgement by the scientific community that it had indeed risen to the status of a scientific discipline. A consequence of the fact that France's major research institution has not recognized ICS as a scientific discipline is that information science has been perceived as an ancillary field devoted to the pragmatic task of providing services to other research communities through dedicated computing services. Such is the mission of the CNRS-owned Institut de l'Information Scientifique et Technique (INIST), which hosts the two major multidisciplinary bibliographic databases, PASCAL and FRANCIS. Because CNRS has not acknowledged the need for fundamental research in ICS, there has been little funding available for information science research and what research has been done has been largely practical in nature. As a result, France lost the historic advantage it had up until the mid-20th century in information science. In Ibekwe-SanJuan (2012), I analyze how the successive ministerial policies in the last quarter of the 20th century helped to shape the landscape of current information science in France.
Having reviewed the context of emergence of information science in France, I now come to the main issue of this paper, which is to analyze the French conception of information science and see if it is indeed “une exception française.” Information science theories, concepts, and paradigms cannot be discussed in the French context without reference to communication science, as the two are bound together in one interdiscipline. I will therefore begin by summarizing viewpoints on the nature and scope of ICS as a whole before focusing on the information science branch.
Information and Communication Sciences: An Unbalanced Union
The linking of the two concepts “information” and “communication” in the same discipline was based on a general sentiment shared in the mid-1970s that the “more concrete notion of information would make precise the vague notion of communication. This coupling also had the advantage of serving the interest of many distinct groups of specialties without adopting a clear stance on the epistemology of the field” (Palermiti & Polity, 2002). It is hardly surprising that the first debates within the emerging discipline were about its name. Indeed, since its birth ICS has been dogged by incessant debates on what constitutes its object (or purpose) of study and where its boundaries lie. The situation is further complicated because of an imbalance in the internal configuration of the field: Communication science is at least three or four times bigger than information science in terms of academic staff, students, and courses. With the notable exception of the National Higher School of Librarians (ENSB, École Nationale Supérieure des Bibliothèques), which was created in 1963, information schools have no separate existence in the French higher education system. As a result of government reforms aimed at bringing library schools closer to the university system, the ENSB was renamed the National Higher School of Library and Information Sciences in 1992 (ENSSIB, École Nationale Supérieure des Sciences de l'Information et des Bibliothèques). However, the overall orientation remains pragmatic and professional rather than conceptual and theoretical. The primary mission of the ENSSIB remains the training of librarians (“les conservateurs”), although some information science courses have been added in the school.
The Quest for the “Object” of a Discipline
ICS in France is undergoing the same definitional process concerning the nature and scope of information science with which that the Anglophone world has long been familiar (Hjørland, 1998; Shera & Cleveland, 1985; Vickery, 1997; Brookes, 1980; Buckland & Liu, 1995; Bates, 1999). It has been particularly difficult for ICS to distinguish itself as a separate scientific field from other neighboring fields such as sociology, psychology, anthropology, ethnology, semiotics, journalism, and even computer science.8 Rather than recall every single proposal that has been made to distinguish ICS as a separate field, I will try to summarize what seems to have emerged as “zones of consensus” from these debates.
In his pioneering book, General Theory of Information and Communication, Escarpit (1991) wrote that “[i]nformation is perceived as a product of an act called communication”; similarly, Jean Meyriat (1981b) saw information as the “cognitive content of an act of communication.” Indeed, many first-generation ICS scholars perceived information as the tangible part of the communication act, capable of obtaining recognition by ministerial bodies and by society at large. Thus, in the first years of its official existence, information science appears to have enjoyed a relatively good rapport with communication science scholars. These early definitions formulated by the founding fathers are being questioned today. The concept of information is a perennially elusive and widely debated concept. Communication is probably just as elusive and omnipresent. In their account of the emergence of ICS as a discipline, Palermiti and Polity (2002) recall that the focus in the early 1970s was on obtaining official recognition for the field in the higher education system, hence little attention was paid to questions of its underlying epistemology.
An official definition of the discipline which seems to have won some consensus was given by the National Committee of Evaluation (Comité national d'évaluation, CNE) in 1993, according to which ICS concerns itself with “the study of information or communication processes, that arise from organized or finalized actions, that may rely or not on technical tools and that partake of social and cultural mediations.”9
From 2000 onward, second-generation scholars have emerged who contend that ICS cannot be satisfactorily defined by the ontologically oriented question, “what is information and communication science?” In a special issue of the Hermès journal published by the CNRS, Jeanneret and Ollivier (2004) gathered some of the most significant contributions on the topic.10 They contend that scientific disciplines have two ways of coming into existence—they are either built around an “object” (ontological question) or around a “project” (constructivist epistemology). The specific way in which the object of a discipline is defined is itself a matter for debate. In the case of hard sciences, such as life sciences, astronomy, and physics, their object may be more or less clearly identified. For the social sciences and humanities, it is quite a different story, for the frontiers of “disciplinary objects” claimed by one or other discipline keep shifting:
Disciplines have more or less an object, meaning that this object can be more or less clearly defined but even for the well-established disciplines, it is not certain that this object does not slip away. If demography deals with variations in populations, it can be said to have a recognizable object, just as gastroenterology or astrophysics do (at least for the novice). The object of sociology is already more difficult to circumscribe, it can be everywhere. That of linguistics is not at all clear for the non-linguist since it is based on a category created by linguists who invented two categories, language and speech, and decided to study what the former meant to them. As for philosophy, its object is most in(de)finite. (Jeanneret & Ollivier, 2004, p. 1411)
In the same Hermès issue, Davallon (2004, p. 31) acknowledged that ICS “reuses, experiments, and adapts concepts and methods built for other objects in other scientific domains.” ICS scholars seem to agree about the futility of choosing objects, problems, and methods that would belong solely to ICS, in a bid to claim disciplinary status:
It was also necessary to refuse at once a certain number of complexes and representations that have for long dogged debates on the scientificity of information and communication sciences. Resist, as much as possible, the idea that objects, problems and methods could be the properties of specific disciplines.12 (Jeanneret & Ollivier, 2004, p. 17)
Interestingly, scholars outside of France have made similar observations that objects are not the properties of specific disciplines (Bawden, 2008; Buckland, 1991, 1999; Floridi, 2002, 2004). Capurro and Hjørland (2003) pointed out that information science is not the only field to study information and that other disciplines are concerned with this (astronomers, historians, photographers, journalists, etc.). In fact, the whole of human society is involved one way or the other with processing information.
Jean-Baptiste Perret (2004) in the same special issue of Hermès observed that difficulty in distinguishing research objects is common to all “constituted disciplines.” I am assuming that by “constituted,” he means disciplines that are not from the life sciences. He goes on to say:
The desire to delineate the scope of a discipline comes up against two classic impasses related to the criteria for recognition and validity of a science:
- On the socio-historical level, the circle of relativism: a discipline is what scientists in the field decide that it is. Its identity relies more on a consensus within the community of scientists than on conceptual agreements, and depends above all on the state of the power relations therein.
- On the theoretical level, the circle of knowledge: any judgment on the relevance or validity of an assertion relies itself on the implicit recognition of a certain paradigm, hence on another judgment that cannot be proved. Hence, there is not, and there cannot be, a scientific definition of scientificity nor any “theory of a good theory.”13 (Perret, 2004, p. 122)
Since its official recognition and up to the present day, ICS has oscillated from one definitional axis to another. Definitions that revolve mainly around one discipline tend to be self-serving, aimed at legitimating the place of their authors within the field. The center of gravity at a given time, in terms of the disciplinary orientation toward which ICS is leaning, depends heavily on the “rapport de force” (power relations) within the ruling body—the National Council of Universities (CNU) for the ICS—rather than on any scientific proof of the superiority or adequacy of one epistemological approach over another. The current center of gravity is very much in favor of theories, paradigms, and methods derived from the social sciences and humanities (except linguistics) and very much in disfavor of computer science and the sciences. Bates (1999) observed the same oscillations and a current swing back to the social sciences with regard to methods and theories within the Anglophone information science community.
Having acknowledged that the field cannot be defined by the ontological question naturally leads to the second alternative according to which ICS can only be defined by its project, i.e., by its purpose or agenda.
Systemics and Constructivist Epistemology as Theoretical Foundation for Information & Communication Science
The avowed impossibility of strictly defining the objects of the field and claiming ownership of these objects has given rise to a second consensus: what distinguishes ICS from other disciplines is the “communicational look” it bestows on objects, be they technical or not (Davallon, 2004, p. 30). This viewpoint is linked to the constructivist epistemology which holds that scientific objects do not exist independently of a subject (that is to say, the person beholding them). The constructivist epistemology adopted by the Palo Alto group has been fiercely championed in France by Jean-Louis Le Moigne (1995), a prominent scholar on systemics and constructivism. Constructivist epistemology is compatible with the systemic theory of communication considered by many communication scholars as the most adequate theoretical foundation for ICS (Mucchielli, 2000). As to how this communicational look or approach might be deployed as a methodology, Mucchielli (2000, p. 43) writes: “To adopt a communicational approach to a phenomenon is to analyze it as an element of a system contributing, in a circular movement, to the emergence of another phenomenon.”14
Hence, the communicational approach introduces a circular causality whereby communication is seen to take place in a system where interactions are circular (chain of retroactions), thus placing this conceptual approach also within the paradigm of complexity. According to this viewpoint, it follows that it is the manner in which objects are regarded that guarantees the uniqueness of the ICS discipline, not the objects themselves (in this case, information and communication), since these can be claimed to also be the objects of investigation of other disciplines.
I see a rapprochement between Mucchielli's viewpoint (2000) and the one defended by Hjørland and Albrechtsen (1995), and later by Hjørland (1998). Although Mucchielli was proposing a communication theory while Hjørland and Albrechtsen were proposing theoretical foundations for information science, both approaches advocate the anchoring of the discipline on a socio-constructivist rather than on the individualistic cognitivist theory. Both advocate nonlinear, holistic approaches to the study of information and communication phenomena that serve communities of practice and discourse rather than individuals.
Is Methodological Purity Desirable or Even Possible?
The theoretical debate naturally spills over to the methodological level. What kinds of methods are acceptable for the discipline of ICS? Several papers published during the 16th Annual Congress of the French Society for Information and Communication Science held in 2008 concluded that the diversity of objects of study, of epistemological traditions, and of theoretical approaches to information and communication show that no one methodological approach or theory can account for all the research that falls under the scope of the discipline. Indeed, to try to impose a single unified theory is not only illusive, it is considered a totalitarian maneuver (Bouillon, 2008). Not unlike information science in the Anglophone world, research in ICS also seems condemned to borrowing models, theories, and methods from other sciences. This is evident when one analyzes the methodologies deployed in doctoral dissertations defended in the ICS discipline. Buisson-Lopez (2008) noted, for instance, that “doctoral [theses] focused on machine-mediated communication borrow theories and methodologies from engineering sciences, while dealing with the study of messages, senders and receivers of such communications. Those concerned with interpersonal communication are in close and often explicit contact with anthropology and social psychology whereas dissertations focused on discourse analysis and on institutions flirt with education, history or law.”15
This situation is not unique to ICS nor to France. A similar observation was made by Marcia Bates (1999, p. 1049):
A final comment on methodology: regarding the great methodological shift sweeping through the social sciences, the shift to the qualitative, multiple-perspective, post-Modernist approaches—these new techniques simply add to and enrich the armamentarium of techniques available to the information scientist for studying the subject matter of our field. For reasons that have already been argued, this field requires multiple methodological approaches to conduct its research. In mid-20th century social science we have had a series of waves of methodological fashion—each wave declaring the prior approach to be hopelessly bankrupt and inadequate. It is to be hoped that it is finally recognized that all of these methodological approaches can be powerful and useful—especially in information science.
The French Vision of Information Science
Researchers who identify with the information science branch within the ICS discipline have naturally attempted to define the scientific object of this branch (Fondin, 2001, 2002; Le Coadic, 1994; Meyriat, 1981a,b; Salaün, 1993). Polity (1999) echoed the same sentiments as Jeanneret and Ollivier (2004) and argued that a field cannot be defined through lexical or ontological definitions but by its objectives, the problems it proposes to study, and the methods it employs to solve these problems.
The Dual Nature of Information Science: A Shared Viewpoint
Underlying the difficulties faced by scholars who attempt to define information science as a scientific field are the historical difficulties and debates in defining such notoriously ambiguous concepts as “information,” “knowledge,” “truth,” “concept,” “life,” “love,” and “happiness.” Without wishing to reopen this debate, I observe that a convergence exists between views held by French and Anglophone authors on the lack of operativeness of the concept itself. Fondin (2002) and Buckland (2012), for instance, have observed that one cannot usefully employ this term without specifying which meaning of information one is referring to. In the same vein, Fondin (2001) wrote:
[…] every word is thought of in the context of a theory or in reference to a more or less explicit model. It [the word] thus acquires specific properties. To talk of “communication,” and be understood, one needs to first indicate the communication theory in which one is situated. The same thing goes for “information.” A specialist cannot speak of this term without reference to an underlying theory. At the very least, s/he must always employ “information” with a qualifier or an explicative in order to be understood by other specialists.16
Furner (2004) has also argued that information is an inadequate concept for defining the discipline because of its inherent ambiguity. He suggested that it could be replaced by one of its surrogates, namely, “relevance,” but this latter term is also fraught with heavy theoretical and practical problems (see Buckland, 2012, for a short discussion of these problems; see also the literature on relevance measures in TREC-like evaluation campaigns). Analyzing the literature on the concept of information, Furner (2010) identified three broad categories: a semiotic family (those accounts that make distinctions between “information-as-thing” [concrete objects, signs] and “information-as-knowledge”; e.g., Buckland, 1991); a socio-cognitive family (accounts that lay emphasis on “information-as-process”; e.g., Belkin, 1990; Brookes, 1980); and an epistemic family (philosophically oriented accounts that focus on the properties that a resource must have in order for the information it emits to qualify as justified, true belief, and so constitute “conceptions of information-as-evidence”; such approaches derive fundamentally from Shannon's mathematical theory of information, augmented with philosophy of language or informational semantics (e.g., Dretske, 1981; Floridi, 2010).
Nevertheless, as Buckland (2012) has pointed out, information is a very fashionable concept and is not likely to disappear any time soon—not so long as the general public continue to think positively of the concepts of “information society,” “the information highway,” and “information technology,” all of which in the popular imagination are associated with a science of information.
More fundamentally, Fondin (2001, 2002) acknowledged, as some of his Anglophone colleagues had done previously (Saracevic, 1999), that information science is torn between two competing paradigms: (1) an objectivist paradigm, which he attributed to the Anglophone world (as if this world were homogeneous in its analysis and viewpoints) and (2) a subjectivist conception, which he attributed to the French viewpoint but which is not peculiar to it, as I will show. In a later article published online, Fondin (2005) tried to distinguish the French conception of information science from what he perceived as the North American one as follows:
Is information a real tangible object or a social object? Embodied in this interrogation is the whole question of ‘meaning’ and its attribution. Meaning, this coherent mental representation that every human being constructs or deduces from things observed in his environment, what he calls information if the meaning is shared: is it immanent (intrinsic) because laid out in the document by its author, or constructed because contextually built by the beholder, in this case the reader? In the former case, and this is explicitly or implicitly the thesis defended by advocates of information processing, those situated in the North American vision of information science, information is a discrete element. For them, all operations of extraction of elements by locating linguistic or other forms in texts are possible. And given that they are working on the original document, the results of the processing are all the more faithful to the text and to its author.17
From the above excerpt it would appear that the North American viewpoint of information is steeped in the information retrieval paradigm—that is to say, in a positivist approach in which documents are perceived as having an innate subject, inherent in the words, just waiting for the reader to pick them up.
Fondin attributes the second human-centered conception of information (the constructivist conception) to the French approach. In this conception, the notions of the “immanence” of information and the validity of automated processing of information are rejected because only humans—that is to say, readers in real-life situations—can construct meanings. Meanings are compulsorily linked to a context, that in which it is received. In this viewpoint, information—i.e., the content of a document—cannot be a fixed, definitive or eternal thing.18
Sylvie Leleu-Merviel (2010, p. 8) defends a similar constructivist viewpoint when she writes, “Patterns, and therefore information, are a construction of the interpreter or beholder.” She also argued that it is not data itself that constitutes information but the perceived relations between elements of data.19
However, I think that there ought to be limits to individual construction of meaning. Taken too far, it may lead to serious ethical, historical, and practical problems. For instance, how many different meanings can be construed from Hitler's Mein Kampf? Is it acceptable that an individual construction of meaning leads to asserting that this book is about tolerance when indeed the very opposite is the case? Fondin (2005) seemed aware of this pitfall when he acknowledged that although meaning is constructed by individuals, it does not authorize each person to deduce whatever s/he chooses; he acknowledged that there is such a thing as collective sense-making or shared interpretations based on social contexts and that words do matter. Recently, Furner (2010) reviewed different philosophical views on the aboutness of documents. He identified at one end of the pole the idealist view, which holds that there is no way in which the aboutness (i.e., the subject) of documents can be determined. At the other end, the realist view holds that there is a “regular procedure by which a work may be analyzed in order to discover its subject,” notably through “linguistic expressions that comprise subject statements.” One may well wonder whether each of these poles does not have something to contribute to an operational understanding of information.
This binary polarization of conceptions of information and information science is found in the texts of many French ICS scholars. Obviously, for them, only the social conception of information is acceptable. Paradoxically, their overwhelming rejection of Shannon's mathematical theory of communication in ICS became the norm after some authors had unsuccessfully tried to apply it to information and communication studies (for examples of some unsuccessful attempts, see Baltz, 2007a; Salaün, 1993).
The opposition between the “objectivist-physical” paradigm and the “subjectivist user-oriented” conceptions of information and communication is not a ‘Franco/French’ thing. Several authors from the Anglophone world have already analyzed the influence of these two opposing paradigms in research in information science (Bates, 2005; Buckland, 1999; Cronin, 2008; Hjørland, 1998; Robertson, 2008; Saracevic, 1999) especially with regard to its relationship with information retrieval (IR).
Saracevic (1999) noted that this historic opposition dates back to the origin of the discipline and is attributed to two historic figures: Jesse Shera and Gerard Salton. Anglophone authors who have analyzed the consequences of this opposition—a parallel existence, on one hand, of the IR community around Salton's work, and, on another, a more user-oriented approach embodied by the American Society for Information Science & Technology (ASIS&T)—have also called for a better merging of the research agenda of both approaches (Saracevic, 1999) as the only way forward for information science. Although this goal has not always been achieved, members of the Anglophone community in information science have adopted a pragmatic approach and reached a modus vivendi on how to accommodate social and user-centered approaches with a more technological one. They seem aware of the fact that information science cannot ignore technology altogether, nor can it fail to concern itself with the design of information systems as a means to ensure better access to information. Bates (1999, p. 1049) also contended that the methodological substrate of information science is of a socio-technical nature and observed the same duality in information science by stating that the two most important bodies of research methods from which the field draws are those of the social sciences and engineering sciences.
No Information Without Communication?
For ICS scholars, the act of communication is intentional—that is to say, triggered by humans for a given goal. In this perspective, the object (or rather the purpose) of information science is to study the “modalities and processes of this finalized communication” and this should be done “within a global approach, whether based on a device (tool) or on a social system” (Fondin, 2005). In an earlier article, Fondin (2001) stated that this goal was specific to IS20 and so could justify the disciplinary status of information science. Four years later, he seemed to retract this assertion by stating that the “the term ‘science’ is inadequate to qualify this sector with a very technical agenda. One maintains a confusion between science and technique or engineering” (Fondin, 2005).21 Seeking to further distinguish the French conception of information science from the Anglophone one, he writes:
No North American scholar can imagine having anything in common with communication scholars [in France]. Anglo-Saxon information science considers itself a separate science and claims this status. The only enormous problem is that some forty years after its birth, information science is still chasing after its recognition. Indeed, how can one envisage a study of information that excludes the phenomena of communication that accompany it? How can one study the content of a message without considering those that create it, those that transform it and those that use it? These activities, because they are highly complex ones with high stakes, are communication activities. Hence, in this light, information science cannot not belong to Information and Communication Sciences. It is this refusal to acknowledge communication that explains why the ”historical“ information science [i.e., the Anglo-Saxon one] remains locked in a technical conception that seems to have no future, on problems of the modalities of production, of dissemination and usage, while ignoring the human factors underlying these activities.22
As I have observed, many North American information scientists share this communicational conception of information science with their French counterparts. Marcia Bates (1999) stated that “the field's interest is in human-produced information, and therefore, how human beings relate to this information—how they seek it, use it, ignore it, retrieve it—is of central research importance also.” She also claimed that information science has a close affinity with communication science when she wrote, “In communications research, a cousin to our field, the emphasis is on the communication process and its effects on people; in information science we study that process in service of information transfer” (Bates, 1999, p. 1048).
Much earlier, Buckland (1991) in his book Information and Information Systems defended a human-oriented conception of information. He adopted a wide view of “information systems” from which mechanical and machine-based processes were excluded. He observed the inappropriateness of including data processing (information technology) in information studies and argued that an “exploration of information systems” must “include the social, economic and political contexts.” Information studies without this social dimension, he argued, would be incomplete (Buckland, 1991, p. 9). He specifically advocated that information science should “include communication both at interpersonal and mass levels.” Not unlike Fondin (2005), he argued that information systems were “supposed to inform people, but in practice, they deliver physical stuff such as books, papers and signals on glowing screens” and that “all information systems deal directly with and only with physical objects such as coded data or documents” (Buckland, 1991, p. 10).
Later, Buckland (1999) defended the viewpoint that information in itself is not important and that what is more important is its relation to knowledge and to communication. Another advocate of the social dimension of information, Saracevic (1999) proposed a definition of information science in which he emphasized the relation between information and knowledge and on the social function of information science in mediating between people and information resources.23 As one can see, these viewpoints are remarkably close to those of French ICS in their conception of how the field should be defined—namely, by its projects and approaches rather than by the objects of investigation. It would appear that these texts are not widely known in the French ICS community and that language is indeed a barrier. If this is the case, then we have a serious communication problem (no pun intended).
In reality, Fondin's criticism applies more to Shannon's linear theory of communication and to research carried out within the field of IR, which is mostly of computational inspiration, but not to library and information science as it is developed in Anglophone countries,24 where much emphasis has been placed on user studies. At any rate, the French view of IS is not a monolithic bloc upholding a social and humanistic approach against an object-oriented approach, attributed to Anglophones. There are at least three French professors who have defended Shannon's contribution to the development of ICS in France. Abraham Moles was a French ICS scholar who came to communication from engineering. He is considered the best French specialist of Shannon's work. He wrote a preface to Shannon's 1948 book in French with an introduction from Warren Weaver (Weaver & Shannon, 1975 25). He personally met Shannon and Norbert Wiener. Jacques Perriault (2004, 2007), also a professor of ICS, is a strong advocate of the role of technology in ICS. Claude Baltz (2007a, b), another interpreter of Shannon's work, has deplored how badly his mathematical theory of information has been misunderstood by communication scholars.
Convergence of Research Agenda on Both Sides of the Atlantic
In terms of research orientations, there also seems to be a remarkable convergence between some North American and French scholars. The three research questions outlined by Bates (1999, p. 1048) for information science align quite nicely with Jean-Paul Metzger's (2002) proposal of research poles around which information science in France should be structured:
- The physical question, “What are the features and laws of the recorded-information universe?,” in Bates (1999) can be aligned with the “formalization and computation” agenda for French information science in Metzger (2002).
- The social question, “How do people relate to, seek, and use information?,” in Bates (1999) would correspond to the information search behavior (user studies) pole identified by Metzger (2002).
- The design question, “How can access to recorded information be made most rapid and effective?,” in Bates (1999) can be partially aligned with two research poles in France named “information production processes” (circuits de production de l'information) and their “modes of diffusion” (modalités de diffusion de l'information).
Although Metzger's three poles do not yet form a consensus in France, they nevertheless reflect the main research orientations that information science has taken since its official birth in the mid-1970s. There is also an analogy between Bates's (1999) proposal to consider information science as a meta-discipline serving other disciplines by focusing on the form rather than on the content, and the ambition of the founding fathers of ICS, such as Norbert Wiener (cybernetics), Claude Levi-Strauss (structuralism), Roland Barthes, and Gregory Bateson, who wished to make it a kind of “super-science whose research problematic would irrigate almost all the disciplines that have traditionally been recognized and included in the classifications of the sciences” (Miège, 2005, p. 99).
The recurrent debate in France about where the boundaries of information science and ICS lie strongly echoes similar debates in the Anglophone world. With their usual pragmatism, our Anglophone colleagues have decided that not only can we not provide a sound scientific argument in support of any boundary but that this debate is a waste of time:
It seems a remarkable waste of time and effort to worry about setting up disciplinary boundaries, and debating who is in and who is out. (Bawden, 2008)
I am wholly in agreement with this statement.
What's in a Plural? “La” or “Les sciences de l'information”?
Behind this deceptively simple question lies a series of epistemological questions about the very object of information science. The exact name of this branch within the ICS discipline is a point that is rarely debated in the French academic community. As Palermiti and Polity (2002) have observed, the opportunistic coining of the field's name “Information and Communication Sciences” was done at the expense of clarification of the scope of the conjunction “and.” Is it a distributive “and” that signifies a conjunction of two separate disciplines, namely “information science” and “communication science”? Is it an additive “and” that correlates one or more information sciences with one or more communication sciences? Or is it a combination of both? In the absence of this clarification, the door was left open to all kinds of interpretations and manipulations. This denominative faux pas can be seen as the “péché originel” under which ICS and subsequently information science have been laboring for decades. Information science scholars in France have not yet agreed on whether we have one information science or several information sciences. The name of the discipline is frequently cast in the plural form as “Les sciences de l'information” (Salaün & Arsenault, 2009; Staii, 2004) but also is expressed, less frequently, in the singular as “La science de l'information” (Fondin, 2001, 2002, 2005). If the plural form is retained, what bodies of work differentiate one information science from another? What are the underlying paradigms of each information science? Two hypotheses can be put forward:
- The plural form is used implicitly to convey the idea that information science is made up of several disciplines or specialties from which it is historically derived, such as documentation, library studies, and archival studies; therefore, it constitutes a branch of the humanities or of the social sciences.
- For epistemological reasons, given that information science is an interdiscipline irrigated by different epistemological approaches and paradigms, the plural form subconsciously reflects the idea that each epistemological tradition constitutes a type of information science.
An example substantiating the first hypothesis would be the following excerpt taken from the Web site of the School of Information and Library Sciences in Montréal, Québec (École de Bibliothécaires et des Sciences de l'Information, EBSI):
Information sciences build on the solid roots of the traditional professions pertaining to the document—library studies and archives—and is deployed in more recent avenues of strategic information, knowledge management and the multiple developments brought by electronic technology.26
It is also in this sense of a discipline made up of other sister specialties that Jean-Michel Salaün (1993) speaks of “les sciences de l'information” in an article written on the origins of the discipline within the larger ICS discipline. This also appears to be the justification for the use of the plural on the Wikipedia page27 dedicated to the discipline where it is stated that the French equivalent of library and information science—“Les Sciences de l'information et des bibliothèques”—is “used to designate a body of knowledge and know-how useful to people in charge of managing libraries or an information-documentation service.” The article goes further to specify that the official name of the discipline is “Sciences de l'information et de la communication” (information and communication sciences) and that this “field is characterized by its object (information, its nature, its properties and its transfer) rather than by its methods.” Further on, the article states that different approaches to the study of this object are possible but mentions mainly the physical-object approach involved in information transmission, which, it says, is the chief goal of “Les sciences de l'information.” It concludes by saying that “Les sciences de l'information” cover other specialties like library management, bibliography, cataloging, and indexing as well as “bibliology.” This definition portrays information science as the technological branch of ICS, encompassing other sister but practical fields. It does not reflect the viewpoint of current French information science scholars who wish to distance the field from such instrumental approach.
Francophone scholars who refer to the discipline in the plural rarely go into theoretical justifications for their choice of form. A notable exception is the article by Fondin (2002), where he analyzed the implications of using the singular and the plural form. In his view, the use of the plural form to designate the discipline is explained by the fact that information science is torn between two opposing and irreconcilable conceptual approaches or paradigms, as we have already noted above. This is in agreement with our second hypothesis above.
Further, Fondin argued that the question of the plural form hardly interests anyone else except the French since the Anglophones have their “Information Science” and “scholars in journalism (media studies) belong naturally to communication science, while computer scientists have their Computer Science.” In the French context, the plural form would designate a pluridisciplinary stance, whereby scholars from different disciplinary backgrounds cohabit under different conceptions of information science without defining a common coherent object or agenda nor taking an epistemological stance. This conception, according to him, is that defended by advocates of information science as an “intensional science” in a broad sense. Fondin wondered what specificity there might be behind this vision. He went on to say that “insofar as debates focus on the object of information science, i.e., information, many can claim to belong to the field.” The common ground would be an agreement on a pluralistic approach, thus justifying the use of the plural form “Sciences de l'information” but at the expense of a clear identity and research agenda. This enables computer scientists who deal mainly with data, journalists, and information producers (i.e., media, journalists) to also claim to belong to information science because information is not defined precisely (but is this possible?), thus allowing anyone to claim to study this object with his or her own tools and aims (Fondin, 2002).
The singular form “La science de l'information,” on the other hand, refers to the interdisciplinary viewpoint, thus evoking the idea of crossing or sharing around the object of information:
If, on the other hand, information science, like the other sciences, is defined by its object [purpose or agenda], and by what it aims to explicate or comprehend through the object of study, by its knowledge and methods, then the interdisciplinary approach is wholly justified. In this context, every scholar can use elements borrowed from other sciences but reconstructed to suit their object (project) of study. This specific object of study is yet to be defined.28 (Fondin, 2005)
What is implied in the singular form is an acceptance of one out of all the possible conceptions of a chosen object, for instance, scientific and technical information (STI). Fondin argued that such a viewpoint sees information science as an autonomous science and he assimilated it with the Anglophone viewpoint defended in the late 1960s by Harold Borko (1968).
However, Fondin observed that despite this quest for autonomy, Anglophone scholars have often talked about the plurality of information science. Indeed, Bates (2007) edited an “encyclopedia of the information disciplines” where she spoke of “the information sciences” in much the same way as we speak of the “social sciences.” According to Bates (2007), “The information disciplines all deal with the collection, organization, retrieval, and presentation of information in various contexts and on various subject matters. That social purpose, of collecting, organizing, and disseminating information shapes all the activities of the information disciplines.” This is in line with her earlier view of the information science as a meta- or orthogonal discipline, which, together with communication science, journalism, and education, serves all the other more traditional and content disciplines. What drives information science is societal need (Bates, 2007; Buckland, 2012). Furner (2010) also employs the plural in talking about “the scope of information studies and/or the information sciences.” Bates's and Furner's usage is congruent with our first hypothesis above, i.e., that scholars across geographical and cultural zones who employ the plural consider “information sciences or information studies” to be made up of a set of sister disciplines or specialties, thus elevating information studies/sciences to the same level as “the social sciences” or the humanities, even if in their minds information studies are also part of the latter. A number of schools outside France are being called “schools of information studies” or iSchools.
Fondin himself only claims one “Information Science,” that “which aims to understand the specific communicational process of information search.” I have also chosen to express the name of the discipline in the singular, thus implicitly agreeing with the interdisciplinary stance of the field, which although borrowing theories and methods from other sciences, aims to build a coherent research agenda geared toward a better understanding of information phenomena and its processes, involving both humans and machines. Fondin's analysis captures quite nicely the status quo in the French literature on this question. Regardless of who is concerned by this debate, the question that it raises, as Fondin rightly pointed out, is that of “inter-” or “pluri-” disciplinarity.
The “Attraction/Repulsion” of Technology
Technology has always been a sensitive topic for many fields in the humanities. In France, in particular, because of the peculiar circumstances of the emergence of information science, it has become an even more sensitive issue. The relationship of ICS toward technology and technically oriented research has been characterized by a double movement of “attraction–repulsion.” Anglophone scholars have also issued warnings about the dangers of a too technological approach to information science problematics (Hjørland, 1998; Hjørland & Albrechtsen, 1995; Saracevic, 1999). In particular, Hjørland and Albrechtsen (1995) expressed the fear that the physical paradigm (systems-driven paradigm) symbolized by Shannon's information theory and by computer science, if not checked, would turn information science into an application terrain for specialists coming from other disciplines who work on information-theoretic problems but from the context of their original discipline. This would make information science unecessary as a scientific discipline (Hjørland & Albrechtsen, 1995, p. 410). This threat was further discussed by Hjørland (1998) with regard to the evolution of IR research.
The early 1980s and 1990s witnessed a fruitful collaboration between information science and the physical systems-driven paradigm, when the quest for scientific legitimacy led scholars in France to seek alliances with the “hard” or experimental sciences. Models and methods from natural language processing (NLP), artificial intelligence (AI), cognitive psychology, and statistics were imported to develop systems in natural language understanding, automatic indexing, information retrieval, automatic translation, expert systems, bibliometrics, and scientometrics. With the retirement of the first-generation scholars who championed these research programs and who came mostly from the sciences (for example, Jacques Rouault, Richard Bouché, and Henri Dou), such strains of research are being phased out. It has become now “de bon ton” (fashionable) to be very critical of any technological approach to ICS problematics and to frown on what is perceived as a technicist approach to the problematics of the discipline (Jeanneret, 2007; Perret, 2004).
What communication scholars chafe most against is all the media hype about the wonders of technology, the revolution attributed to the Internet, to Web 2.0 applications, and to the Semantic Web, where it is often not highlighted that it is people that turn technological devices into tools for machine-mediated communication and that without people these technologies will remain clever feats of engineering and nothing else. For instance, Jeanneret (2007) considers the term “information technologies” to be improper because it maintains a confusion between two conceptions of information, one used in the mathematical sense to refer to data processing and the other to human-mediated information that is embedded within social practices. As Buckland (1991) had done more than a decade previously, Jeanneret (2007) observed that what computer programs disseminate are material objects (that is to say, signs) and not information in the human sense of the word and that so-called “information technologies” should, strictly speaking, be termed “semiotic technologies or text technologies.”
Overall, the attitude that ICS and information science scholars have toward technology and computer science is ambivalent. The picture, as often, is not just white or black but includes various shades and nuances. Some scholars from communication science have acknowledged that the conception that communication science researchers themselves have of communication is fundamentally a technical one (Davallon, 2004); they see it as a tool, as a means to an end, thus reflecting the subconscious belief that “there can be no communication without a tool.” As Davallon observes, “Society sees information and communication sciences spontaneously as a theory of technical objects—i.e., as a technology. Hence, any research concerned with another dimension (conditions of production, context of reception, etc.) will appear to belong to other disciplines such as economics, sociology, etc.”29 In his book entitled Shannon revisité (Shannon Revisited), Baltz (2007a), an information science professor, used role playing to portray this subconscious technical conception of communication. Students were asked to pass on some information from one person to another. They invariably looked for a device (tool) with which to pass on the communication. This illustrated the fact that “communication is first and foremost a question of technique, and often, with any means available to us”30 (Baltz, 2007a, p. 19). Jeanneret and Ollivier (2004, p. 88) also spoke of the fascinating power of technology and warned ICS scholars against “camping in an attitude of pure criticism.” They recognized that what distinguishes “communicational objects is that they have at the same time a social, technical and semiotic dimension.”31 Perriault (2004) and Staii (2004) have also exhorted ICS scholars to become actively involved in discussions on how to design systems that will better serve society instead of taking a passive “let's-wait-and-critique” posture because technological solutions will be built anyway without input from the field.
Earlier on, Polity (1999) observed that, because of its history, information science has been astride three types of platforms: fundamental research, professional practices, and the information industry. It is in the interaction between these three areas of endeavor that the discipline can nourish its research agenda, fuel its fundamental research by observing information processes and usage in real-life situations, and design information systems that can meet societal needs. To invoke the tree metaphor, denying one component is like cutting off a main branch of the tree, which while withering and dying, may well kill the whole tree (Polity, 1999).
I also observe that in the different tributes written to Suzanne Briet, her visionary image of what a document is has been justly lauded as the precursor of information science (Buckland, 2006; Day, 2007; Maack, 2004; Martinet & Day, 2008), but what about her enthusiasm for technology? Briet was a staunch advocate of the role of technology in documentation and information science and of competitive intelligence even if she never used these two terms (Blanquet, 2007). She saw technology as an indispensable auxiliary that information professionals must master if they are to remain efficient in an ever-changing landscape. She was quite taken with the technological inventions of her time (e.g., microforms) and marveled at the “progress” these had enabled the field to accomplish in terms of storage:
One can transfer a whole book with its illustrations on microfilms, on microcards. A thick file (dossier), microfilmed, can be slipped into a coat pocket. A whole library is held withn a handbag.32 (Briet, 1951).
Briet would certainly have been very excited by the “progress” accomplished by today's technology. She would have enthusiastically endorsed the Internet, the Web, e-books, the Semantic Web, the Web of Data, and the participatory Web 2.0. She was emphatic in predicting that “the special librarian (documentalist) will be more and more dependent on tools, the technicity of which is increasing at lightning speed,” and that the “Homo documentator should prepare himself to command, all senses alert, the robots of tomorrow. Machines will have the value of servants.”33
Although the last sentence is too emphatic and undoubtedly raises skepticism and objections, the rest of Briet's predictions have come to pass. She also saw the importance of linguistic knowledge in documentation for building multilingual terminological resources in order to assure a better dissemination of documentation languages.
Currently, there seems to be a double standard with regard to technologically oriented research in ICS. On the one hand, the discourse directed toward government bodies and funding agencies is very inclusive: ICS is presented as being concerned with technical and pragmatic solutions of which the society is in need, as well as with social analyses of communcations processes. In such contexts, ICS scholars have little qualms in claiming association with IR, with information processing, or even with data mining.34 However, when it comes to peer recognition and promotion via the national ruling commission for the discipline, technologically oriented research is viewed unfavorably. There is now a tendency to consider such work as being outside the scope of ICS, and as stemming from a narrow technicist approach.
From the above discussions, it appears that there is a consensus crossing geographic, linguistic, and cultural boundaries on the dual nature of the field, namely, that it is both technical (i.e., the objectivist paradigm) and social (i.e., the subjectivist paradigm). It then follows that information science needs research on both technical and social aspects to accomplish its agenda. This dual nature of the field is a strength but also a weakness. As Bates (1999, p. 1049) observed: “This is one of the reasons we have failed to coalesce as a field around one standard methodological paradigm. For one thing, we need this methodological variety to solve these problems.”
While the omnipresence of technology in society should not lead to adopting a purely technicist approach to scientific problems nor to a techno-euphoria, neither should it lead to the wholesale denial of the need to employ technology to address information problems. Technology has played a major role in bringing to fruition many aspects of the information science research agenda (e.g., design of information systems, OPACs, digital libraries, search engines, knowledge databases, man–machine interface, user studies, and knowledge organization artifacts). Robertson (2008) went as far as saying that research in the IR field is the best claim that information science can make to “scientificity” while at the same time acknowledging that IR research lacked theory.
The double oscillatory movement of “attraction” and “repulsion” between technology and human-focused research, between anchoring information science in the “hard”35 sciences or in the humanities and social sciences is not unique to the French context. Marcia Bates (2005) explained how the current mood swing in the Anglophone world has also shifted toward the social, subjectivist, and hermeneutic conception of information and of information science, whereas 50 years before humanities fields sought methods from the “hard sciences” in order to be accepted as scientific fields:
In more recent years, there has been a reaction to this approach [extreme scientism and logical positivism], with a concomitant swing towards the use of what are essentially humanities methods in the social sciences. Now the fashion is to deride the very scientific techniques so recently valorized and to insist that only highly qualitative and subjectivist methods produce credible results. Hermeneutic interpretation, detailed participant observation and historical analysis, among others, are now the methods of choice. Nowadays, it is seldom remembered, however, that the logical positivist approach was itself a reaction to what were deemed ineffective subjectivist research philosophies that preceded it. […] In like fashion, attitudes towards information itself have swung between highly objectivist and subjectivist interpretations. (Bates, 2005)
Buckland (2009) equally observed that the British Institute of Information Scientists (IIS), under the aegis of Jason Farradane, also sought to identify its methods with those of the hard sciences (notably, physics) in the early years of its existence before finally embracing a social turn much later.
Hence, the ambivalence that French ICS scholars display toward the sciences, especially toward AI and computer science, appears to be a universal trait: We invoke technology and the “hard sciences” when they serve our purposes, we decry them when they do not.
Whither Information Science in France?
A consequence of the disconnect in the research carried out by contemporary French information scientists and that done by the pioneering figures of the 20th century is that current French information scientists are barely visible on the international scene. The best known French contributors to information science are historic figures like Suzanne Briet, Eric de Grolier, Robert Pagès, Jean-Claude Gardin, Gérard Cordonnier, and a few late 20th-century authors like Jacques Maniez, Sylvie Fayet-Scribe, or Yves Le Coadic, most of whom have now retired.
The cohabitation of information science with communication science in the same (inter-)discipline brings its own trials that render the quest for identity and visibility more difficult in the French context. Communication science departments far outnumber information science ones and therefore wield greater negotiating power in defining the (inter-)discipline's center of gravity. An alarmist view would be to say that the very existence of information science in France is in the balance. The current shift in power is in favor of a social and theoretical research agenda. One consequence, as already mentioned, is that research of any technological orientation is seen as falling outside the scope of ICS, even when the actual modeling and programming are done by computer scientists. Yet many observers acknowledge that humans today cannot carry out their tasks adequately without resorting to technological tools. It is also clear to me that if information science abandons the task of systems design entirely to computer scientists, then it gives up any possibility of influencing the design choices to better meet societal needs. Information science and documentation have traditionally been concerned with the design of knowledge organization systems, not only in a strictly technological sense but also in terms of their conceptual and methodological dimensions, as a means of enhancing access to information.
To end on a more optimistic note, one can conjecture that the future may bring a mutual understanding of the information science research agenda by communication science scholars and vice versa. In the French higher academic system, one branch cannot survive without the other. Therefore, in order to gain sufficient negotiating power as a discipline, ICS needs to be united. There is also hope that the younger generation of communication science scholars, who are, by virtue of their age, more accepting of the larger role technology is set to play even in scientific inquiry (e-science and cyber-infrastructure initiatives are springing up in most developed countries), will be in a better position to understand the agenda of information science. They are also in a better position to perceive more easily the potential interest of combining theoretical with applied research into different aspects of computer-mediated communication (social networks, virtual identity, transformation of professional work practices, influence of technology on scientific research, etc.). This could foster a mutual reinforcement of information and communication sciences in France and allow them to attain a more visible place on the international arena.
I thank the anonymous reviewers and Jonathan Furner for their pertinent suggestions, which helped to improve an earlier version of this paper. I also thank Thomas Dousa for help in polishing the English and for flushing out the remaining recalcitrant French formulations.
“Habilitation à diriger des recherches” is the highest university diploma, obtained after the PhD, usually after working several years as Associate Professor. For more details, see http://en.wikipedia.org/wiki/Habilitation
This short biography is based on the one written by Palermiti (2000), available online at http://www.iut2.upmf-grenoble.fr/RI3/Mise_jour_06/TPS_precurseurs.htm
Original unpublished text available at http://www.iut2.upmf-grenoble.fr/RI3/Mise_jour_06/TPS_precurseurs.htm
The original sin, in reference to the biblical story of Adam and Eve eating the forbidden fruit in the garden of Eden. The analogy here is that ICS was flawed from birth.
In French “Sciences de l'information et de la communication” (SIC).
In French, “Société française des sciences de l'information et de la communication” (SFSIC).
Jean Meyriat regrettably passed away on December 26, 2010. He was aged 89.
The CNRS has a department called “Sciences et Techniques de l'Information et Communication” (STIC) which is often confused with the ICS field, as both have the “IC” in common (information and communication). However, the CNRS has funded some interdisciplinary research in this STIC program involving information science scholars but the STIC department is largely led by computer scientists.
The French title is Information et communication: théorie générale.
In the original text: “l'étude des processus d'information ou de communication relevant d'actions organisées, finalisées, prenant ou non appui sur des techniques, et participant des médiations sociales et culturelle.”
Jeanneret, Y., & Ollivier, B. (Eds.). (2004). Hermès, special issue on “Les sciences de l'information et de la communication.” Hermès, vol. 38. Online at http://documents.irevues.inist.fr/handle/2042/9030
In the author's original text: “Les disciplines ont plus ou moins un objet. Entendons par là que celui-ci se définit plus ou moins aisément et que, même pour les plus assurés, il n'est jamais certain qu'il ne se dérobe pas. Dans la mesure où la démographie traite des variations de population, elle a un objet saisissable, comme la gastroentérologie ou l'astrophysique (au moins pour le non spécialiste). L'objet de la sociologie est déjà moins facile à circonscrire. Il peut être partout. Celui de la linguistique ne l'est pas du tout pour le non linguiste, puisqu'il repose sur une catégorie produite par les linguistes qui ont inventé deux catégories, langue et parole, et décidé de traiter de ce que recouvre pour eux la première. Quant à la philosophie, son objet est le plus in(dé)fini” (Jeanneret & Ollivier, 2004, p. 14).
In the original text: “Il fallait aussi refuser d'entrée un certain nombre de complexes et de représentations qui ont animé longtemps des débats sur la scientificité des Sciences de l'information et de la communication. Repousser, autant que possible, l'idée que les objets, les problèmes et les méthodes seraient propriétés des disciplines particulières.”
In the original text: “Le désir de délimiter en principe le champ d'une discipline se heurte en effet à deux apories classiques concernant les critères de reconnaissance et de validité d'une science. Au plan socio-historique, le cercle du relativisme: une discipline est ce que les chercheurs qui l'animent décident qu'elle est. Son identité repose donc plus sur l'accord entre la communauté des chercheurs que sur des attendus conceptuels, et dépend avant tout de l'état des rapports de force entre eux. Au plan théorique, le cercle de la connaissance: tout jugement sur la pertinence ou la validité d'un énoncé repose lui-même sur la reconnaissance implicite d'un certain paradigme donc sur un autre jugement lui-même indémontrable. Dès lors il n'y a pas et il ne peut pas y avoir de définition scientifique de la scientificité ni de théorie d'une bonne théorie” (Perret, 2004, p. 122).
In the original text: “Avoir une approche communicationnelle d'un phénomène c'est l'analyser comme élément d'un système contribuant, dans un mouvement circulaire, à l'émergence d'un autre phénomène” (Mucchielli, 2000, p. 43).
In the original text: “Ainsi, les thèses centrées sur la communication via des dispositifs techniques de médiation empruntent parfois aux sciences de l'ingénieur tout en abordant l'étude des messages et des sujets producteurs et usagers de cette communication. Les thèses préoccupées par la communication interpersonnelle sont en contact étroit et souvent clairement énoncé avec l'anthropologie et la psychologie sociale. Enfin, les travaux centrés sur les discours ou les institutions flirtent avec les sciences de l'éducation, l'histoire ou le droit” (Buisson-Lopez, 2008).
In the original text: “Selon cette approche, tout mot est pensé dans une théorie ou en référence à un modèle plus ou moins explicite. Il a dès lors des propriétés particulières. Parler ”communication,“ c'est, pour être compris, indiquer d'abord la théorie de la communication dans laquelle on se situe [2, p. 28–33]. Il en est de même pour ”information.“ Un spécialiste ne peut en parler sans faire référence à une théorie sous-jacente. À tout le moins, il doit systématiquement utiliser ”information“ avec un qualificatif ou un explicatif afin de se faire comprendre des autres spécialités. Selon cette approche, tout mot est pensé dans une théorie ou en référence à un modèle plus ou moins explicite. Il a dès lors des propriétés particulières. Parler ”communication,“ c'est, pour être compris, indiquer d'abord la théorie de la communication dans laquelle on se situe [2, pp. 28–33]. Il en est de même pour ”information.“ Un spécialiste ne peut en parler sans faire référence à une théorie sous-jacente. À tout le moins, il doit systématiquement utiliser ”information“ avec un qualificatif ou un explicatif afin de se faire comprendre des autres spécialistes.” Fondin, H. (2001). La Science de l'information: posture épistémologique et spécificité disciplinaire. Documentaliste, Sciences de l'information, vol. 38, no. 2, p. 120.
In the original text: “L'information est-elle un objet réel, objectivable, ou un objet social? À travers cette interrogation est posée toute la question du ”sens“ et de son attribution. Le sens, cette représentation mentale cohérente que tout homme construit ou dégage de quelque chose observée dans son environnement, ce qu'il appelle information si ce sens est partagé, est-il immanent car déposé dans le document par son auteur, ou construit car élaboré contextuellement par les acteurs, ici par celui qui lit? Dans le premier cas, et c'est explicitement ou implicitement, ce qui est défendu par les tenants du traitement automatique de l'information, ceux qui se situent dans la vision nord-américaine de la SI, l'information est un élément discret. Pour eux, toutes opérations d'extraction d'éléments par repérage de formes linguistiques ou autres sur les textes est dès lors possible. Et du fait qu'on travaille sur le document original, les résultats des opérations de traitement ne sont que plus fidèles et au texte et à son auteur” (Fondin, 2005). Retrieved from: http://w3.u-grenoble3.fr/les_enjeux/2005/Fondin/index.php
In the original text: “Dans le second cas, et c'est une contestation forte de toute idée de sens immanent, et donc de la validité de tout traitement entièrement automatique, et donc de la SI ”historique,“ seul l'homme, ici le lecteur, en situation vécue, crée du sens. Et le sens est obligatoirement lié à un contexte, celui de la réception. L'information, autrement dit le contenu d'un document, ne peut donc pas être quelque chose de figé, définitif, éternel. Certes cela n'autorise pas pour autant chaque lecteur à lire n'importe quoi, mais cela redonne toute sa place à celui-ci. À chaque lecteur de construire, à travers le sens qu'il attribue, un espace social de partage de sens. Dans cette optique, l'objet de la SI ne peut, ne doit pas être un objet physique. C'est un objet éminemment social, avec des acteurs, des enjeux, des contextes…” (Fondin, 2005).
In the original text: “Finalement, ce ne sont donc pas les données elles-mêmes qui supportent l'information, mais les liens aux interstices entre les données, sur lesquels viennent se constituer les schèmes structurants” (Leleu-Merviel, 2010, p. 12).
In the original text: “Ainsi, la spécificité de la science de l'information est d'étudier les modalités mêmes—le processus—de cette communication finalisée. Cette étude est inspirée par le souci d'une approche globale, que ce soit autour d'un dispositif ou d'un système social. Ce souci doit être celui de la SI car aucune autre discipline n'a globalement ce projet.”
In the original text: “le terme ”science“ est inadéquat pour qualifier ce secteur aux projets très techniques. On entretient la confusion entre science et technique ou ingénierie” (Fondin, 2005).
In the original text: “Aucun chercheur nord-américain n'imagine avoir quelque chose en commun avec des chercheurs en communication. La SI [anglosaxonne, historique selon lui] se veut une science à part entière, revendiquant ce statut. Le seul et énorme problème, c'est que, quelque quarante ans après sa naissance, la SI court encore après cette reconnaissance. En effet comment envisager une étude de l'information en excluant de prendre en compte les phénomènes de communication qui l'accompagnent? Comment étudier le contenu d'un message sans considérer ceux qui les créent, ceux qui les transforment, ceux qui les utilisent …? Ces activités, qui sont d'une grande complexité du fait des enjeux qu'elles traduisent, sont des activités communicationnelles. Dans cette logique, la SI ne peut pas ne pas appartenir aux SIC. C'est d'ailleurs ce refus qui fait que la SI ”historique“ s'enferme dans une vision technique, qui paraît sans avenir, autour des modalités de la production, de la diffusion et de l'utilisation, en occultant trop tous les facteurs humains sous-jacents à ces activités” (Fondin, 2005).
“More specifically, information science is a field of professional practice and scientific inquiry addressing the problem of effective communication of knowledge records—”literature“—among humans in the context of social, organizational, and individual need for and use of information” (Saracevic, 1999, p. 1055).
It should be noted that while IR is historically considered a subpart of the Anglophone information science, this is not the case in France, where IR research is carried out mainly by computer science scholars. Although a handful of information science professors in France had worked on automatic indexing using NLP in early 1980s and 1990s, this line of research has since been abandoned to computer science.
In his preface to the French translation of Weaver and Shannon, Théorie mathématique de la communication, Retz–CEPL, Paris, 1975.
In the original text: “Les sciences de l'information s'appuient sur les racines solides des professions traditionnelles du document: la bibliothéconomie et l'archivistique, et se déploient dans les avenues plus récentes de l'information stratégique, la gestion des connaissances et les multiples développements induits par les technologies numériques.” Retrieved from: http://www.ebsi.umontreal.ca/
In the original text: “En revanche, si la SI, science comme les autres, se définit par son objet et par ce qu'elle cherche à expliquer ou à comprendre dans l'objet qu'elle étudie, et par les savoirs et méthodes convoqués, l'approche interdisciplinaire se justifie pleinement en ce que chaque chercheur peut utiliser, en cohérence, des éléments pertinents empruntés à d'autres sciences en les reconstruisant par rapport à son objet. La spécificité doit donc d'abord être celle de l'objet d'étude, ce qui, en outre, permettrait d'afficher devant les autres une réelle identité. Reste à définir cet objet spécifique” (Fondin, 2005).
In the original text: “la place faite par la société aux sciences de l'information et de la communication est spontanément celle d'une théorie de l'objet technique—c'est-à-dire, au sens strict, d'une technologie. Toute recherche portant sur une autre dimension (conditions de production, contexte de réception, etc.) semble relever par contrecoup de domaines scientifiques étiquetés par une discipline autre, telle que l'économie, la sociologie, etc.” (Davallon, 2004, p. 32). Retrieved from: http://documents.irevues.inist.fr/handle/2042/9421.
In the original text: “…il reste maintenant à redécouvrir que la communication est avant tout une question technique et, souvent, avec les moyens du bord” (Baltz, 2007a, p. 19).
In the original text: “Mais le propre des objets communicationnels est qu'ils ont à la fois une dimension sociale, une dimension technique et une dimension sémiotique” (Jeanneret & Ollivier, 2004, p. 88).
“On transfère un ouvrage entier, avec ses illustrations sur des microfilms, sur des microfiches, sur des ”microcards.“ Un épais dossier se glisse, microfilmé, dans une poche de veston. Une bibliothèque entière est renfermée dans un sac à main” (Briet, 1951).
In the original text: “le documentaliste sera de plus en plus tributaire d'un outillage, dont la technicité augmente à une vitesse grand V. ”L'homo documentator“ doit se préparer à commander, toutes facultés en éveil, aux robots de demain. La machine vaudra ce que vaut le servant” (Briet, 1951).
In a letter dated 22 February 2010 adressed to a Council for the Development of Humanities and Social Sciences by the President of the French Association for Information and Communication Sciences (SFSIC), he requested the inclusion of ICS in the list of humanities disciplines and gave a broad vision of the areas of expertise of the field: “Elles ont une expertise traditionnelle sur la veille, le traitement, la sélection et la qualification de l'information dont l'extrême abondance nécessite le développement de recherches dans des domaines aussi pointus que le ”data mining,“ la veille stratégique, ou encore les technologies de surveillance et de protection de la vie privée… .”
I am using the term “hard sciences” here for convenience, to avoid a lengthy debate, and also to portray a common if not unfounded belief that some sciences are “hard” whereas others are “soft.” I believe that such boundaries cannot be established convincingly.