Résumé des interventions

Birger Hjorland: The nature of information science and its core concepts.

Information science developed out of scientific documentation and of librarianship and is today much concerned with, among other fields, bibliometrics, information architecture and Internet studies. How does the concept of information fit in? And what difference does it make whether “information” is understood one way or another?

Thomas Dousa: Documentary Languages and the Demarcation of Information Units in Textual Information.

A central premise of information studies is that information can be decomposed into smaller units, which can be demarcated on the grounds of syntax (i.e., signs and sign sequences), semantics (i.e., concepts and propositions), or both (i.e., terms and sentences): in other words, it is possible to speak of “information units” isolable through an act of “information analysis”. Within the field of indexing theory, this premise gives rise to an interesting and important question: if one is going to characterize the information content of textual information, how does one identify and demarcate “information units” within text? The purpose of this presentation is to examine this question with reference to the theories of information analysis articulated by two significant indexing methods developed over the course of the 20th century: method of “systematic indexing” of J. O. Kaiser (1911) and the “machine language”/”semantic code” developed by James Perry and Allen Kent (1956, 1958): in particular, it will explore how both epistemological considerations and the communicative constraints of the documentary languages shaped Kaiser’s and Perry/Kent’s views of information analysis.

Bruno Bachimont: Information et communication : phénomènes empiriques mais concepts mal définis.

L'information et la communication sont des notions abstraites qui ne sont définies le plus souvent que de manière vague malgré certains domaines comme les télécommunications et le traitement du signal. De ce fait, nombre des analyses menées à leur endroit sont peu productives car elles manquent de la rigueur nécessaire. Par ailleurs, l'information et la communication ne sont pas seulement des concepts qu'il faut théoriser, mais des réalités empiriques incontournables qu'il faut penser. Nous tenterons donc de caractériser l'information et la communication d'une part en tant que faits empiriques comme des problèmes et de les rapporter à des concepts mieux fondés d'autre part. Cette dernière démarche s'appuie sur une analyse de la technique et la matérialité de l'information, et sur la manière dont le sens est manifesté par la matière et son conditionnement technique.

Søren Brier: The Cybersemiotic approach to a transdisciplinary and evolutionary theory of meaningful information, cognition and communication.

As soon as we want to base our theory of meaningful communication with language on a semiotic evolutionary theory it becomes clear that we do not have a transdisciplinary theory that combines quantum mechanics, thermodynamics, informations science and evolutionary behavioural biology and brain physiology with conscious meaningful cognition and communication as it is defined phenomenologically. The view from outside does not fit with the view from inside. The central goal of a theory of knowing and communication must be to find a theoretical way of putting the knower into a known with an ontology that is able to keep the knower, as well as the knowing process, conscious viable? But science – including information science - cannot tell us what life is phenomenologically. What can the frames for such a possible transdisciplinary theory be, when cognitive science and phenomenological semiotics seems to carry two different incompatibles ontologies and dualism is no longer an acceptable solution?

Luciano Floridi: Perception and Testimony as Data Providers.

What is the relationship between information and knowledge? Recently (Floridi [forthcoming 2], Floridi [2011]), I argued that semantic information — understood as well-formed, meaningful and truthful data — upgrades to knowledge if and only if it is correctly (Floridi [forthcoming 1], Floridi [2011]) accounted for. The basic idea is rather simple. Each piece of semantic information (e.g., “the beer is in the fridge”), is constituted by a Boolean question and answer (“Is the beer in the fridge?” + “Yes”), which, as a standalone item, does not yet constitute knowledge, but poses further questions about itself. Such further questions require the right sort of information flow in order to be answered correctly, through an appropriate network of relations with some informational source. If all Mary can do, when asked why she holds that the beer is in the fridge, is to repeat that is the place where the beer is to be found, the fact that the beer is actually in the fridge only warrants the conclusion that Mary holds the information about the location of the beer, but nothing else. For all we know, Mary might have uttered “the beer is in the fridge” as the only English sentence she knows, or she might have dreamed or guessed correctly the location of the beer. Indeed, the beer that she reports to be in the fridge might have been removed by John, but then more beer might have been placed in the fridge by Peter, making Mary right, yet only accidentally. The result of such analysis is an informational definition of knowledge according to which a knowing subject S knows that p if and only if:
1. p qualifies as semantic information;
2. A accounts for p, that is, A(A, p);
3. S is informed that p; and
4. S is informed that A(A, p).
The articulation of this analysis in terms of a network theory of account and its defence, especially against a potential Gettierization, are explicit tasks I have dealt with in Floridi [forthcoming 2]. In this paper I would like to explore an important consequence of the informational definition of knowledge: if knowledge is accounted information, what happens when we apply this definition to the analysis of perceptual knowledge and knowledge by testimony?

Sylvie Leleu-Merviel:Horizon de pertinence dans le processus informationnel

Le processus informationnel est revisité au regard des travaux de Mioara Mugur-Schächter qui explicitent et organisent la genèse et la structure du contenu de l’entier volume du conceptualisé, en particulier de la connaissance scientifique. Partant de l’infra-conceptuel, qui se présente comme un magma de perçus indistincts, les processus d’élaboration de représentations transforment les perçus en entités signifiantes. Des procédures agréées de légitimation permettant une normalisation du sens conduisent à des significations, c’est-à-dire du sens négocié et partageable, détaché de son contexte, communicable, enseignable, apte à circuler, dès lors institué en objet social. En montrant comment s’entrelacent des concepts, des opérations, des données factuelles, une sémantique symbolique et une syntaxe qui assurent la communicabilité, le « canon descriptionnel » instauré met en relief les diverses composantes d’un « horizon de pertinence » dans lequel les “données” ne sont pas “données” mais “saisies” conformément à des intentions spécifiques (et différentes).
Mots-clés. — Horizon de pertinence, représentation, communication, données, faire sens, signification, interprétation

Ira Noveck: The interface between sentence meaning and speaker meaning.

Paul Grice underlined how sentences carry meanings that go only so far in helping capture the speaker's meaning. The domain of linguistic-pragmatics has since evolved to try to account for the way interlocutors comprehend sentences. In this talk, I will show how experiments are useful for describing how listeners go beyond sentence meanings in order to comprehend a speaker's meaning while covering Sperber and Wilson's Relevance Theory.