Résumé
This study is divided into three parts: the first is problematic, the second is analytical and the third is programmatic. The first part presents the theoretical framework of semiotic engineering, its problematic, its heritage and its articulation with the problematic of diagrammatic representations. A historical and genetic analysis of operative thought, leading to algorithmic representation, forces us to acknowledge that diagrammatic representation is irreducible to algorithmic representation. For this reason, the diagrammatic representation is the bearer of an additional intelligibility of a semiotic order that needs to be characterised. This is what we try to do in the second part of this work, by proposing analyses of diagrams in different fields of knowledge, whether theoretical or applied. In fact, our aim is to propose a methodology of artificial intelligence that is characterized by the forgetting of the logicist and formalist position, namely the symbolic, effective and yet non-computational character of diagrammatic representations. The semiotic engineering of diagrammatic systems, as we call the artificial intelligence conceived and realised in this framework, is dedicated to the identification of visual means of elaborating theoretical and practical knowledge with the help of diagrammatic representations. For the diagrammatic system constructed must produce a meaning in the practical domain that is recognised by those who read it and use it as a tool for solving problems that cannot be solved algorithmically. This is why we argue that diagrams are semiotic machines. By articulating a theory of the material inscription of knowledge (B. Bachimont’s theory of support) with the theory of individuation processes (which comes from G. Simondon) thanks to diagrammatic semiotics (which comes from C. S. Peirce), we aim to support two complementary theses: on the one hand, that diagrammatic inscription corresponds to the recording of operative knowledge (thesis 1); on the other hand, that operative knowledge corresponds to the analogical intuition and actualisation of the diagrammatic inscription (thesis 2). The second thesis, which is the reciprocal of the first, presupposes the rehabilitation of analogy as a method of invention. By reinscribing diagrams in a long history of technique, we show in a third and final part that they open a new chapter in the history of writing, appearing as intellectual technologies constituting new ways of thinking and of equipping thought. Likewise, by reinscribing analogical reasoning in a long history of science and technology, we argue for the rehabilitation of the method of analogical reasoning as a rational and valid approach, as long as it is exercised under certain conditions and within certain limits. Finally, by arguing for the thesis that the diagram is a semiotechnical device at the heart of the living, wehypothesise that it is a relevant tool for understanding the processes of morphogenesis within theframework of a philosophy of nature.
Source: http://www.theses.fr/2021COMP2672
.