The processes involved in distributing scientific information, i.e. information produced by and for the sciences, assume Depending on the conceptual point of view, the distribution processes are patterned in different ways. As physics is currently the leading scientific field, physical science models are often used. Two of them are studied here and their limitations are shown. But biology and biological models have also attracted certain information scientists, and again we show the limitations to these.All of these processes are, however, of a social kind, which indicates that only sociological models, and particularly those of mathematical sociology are appropriate in studying this phenomena. Various models are discussed here and a new conceptual approach is proposed. Terminotogy ' It is difficult not to be surprised when leafing through the many writings on the distribution of scientific and technical information, when we hear it spoken of by some as transfer, by others as dillusion. while some speak of trunsrnrssron and, still others as commulllcatioll, drsseminatron or Circulatioll. Sometimes, even within the same document, several of these words are used without any really apparent rea~on.This leads one to worry about the grounds of discussion in information science. The language used is more an ordinary everyday one than scientific, and the worry arises from the manifold examples of terminological problems in this science. For example, the literature on users' needs has been blurred by the inaccurate use of many terms [1]. The concepts are in fact not scientific concepts but linguistic ones, which are equivocal by nature. They cover several phenomena and are open to looseness in meaning, to metaphor and to incongruous associations [2]. All of these terms do in fact offer a richness in meaning with a concomitant wealth of ambiguity, thereby limiting their usefulness in theoretical discussion. Everybody ' knows' what communication means. We all do it and all know how to talk about it, without precision and often even without clarity. Without going so far as proposing a total separation of the &dquo;grain of 'truly' scientific concepts from the chaff of assuredly ideological words&dquo;, as Levy-Leblond suggests [3], a clean break has to be made with common sense representations if we want to establish objective knowledge on such concepts, without this point of scientific honour condemning us to a reductionist objectivism [4].But the difficulties involved are not only terminological. In fact. they reveal the incredible mess of conceptual and epistemological difficulties that afflicts the current pre-scientific state of information science. An example of confusion at the conceptual level giving rise to this multitude of names is that of considering the information' concept of the theory of ' information' to be analogous to the concept of information in the human communication process. If the phenomena studied were the same in both cases, we should have an unequivocal meaning characteristic of any scientific co...