Séminaire doctoral PhilSci - Marie Michon, Henri Salha

Présentations de Marie Michon et Henri Salha

Marie Michon, « Pour une défense de la peur comme attitude non propositionnelle »
L’approche cognitiviste en philosophie et en linguistique considère généralement les émotions comme des attitudes propositionnelles. Cette considération pose cependant des défauts théoriques importants, que ce soit dans la manière de caractériser ce qu’est une émotion ou avec la définition même d’une attitude propositionnelle. C’est sur cela que portera cette présentation, avec pour illustration le cas d’une émotion spécifique : la peur.

Parmi les différentes manières de caractériser une émotion, je traiterai des travaux qui caractérisent les émotions en tant que  jugements (cf. Solomon 2004; Neu, 2002; Nussbaum 2003 par exemple), ainsi que de ceux qui caractérisent les émotions en tant que désirs (cf. Marks 1982; Oakley 1992). Ces deux exemples serviront de support pour montrer que la peur ne peut ni être assimilée à un jugement ni à un désir. Ces deux instances étant des attitudes propositionnelles, on pourra voir en quoi la peur ne peut pas être systématiquement une attitude propositionnelle. La seconde partie de ma présentation portera sur les défauts généraux inhérents à l’idée de peur comme attitude propositionnelle d’un point de vue linguistique.

Henri Salha, “Programs as tools for knowledge”

La présentation sera donnée en anglais mais les questions pourront être posées en français ou en anglais (traduction possible dans les deux langues)

This communication aims to discuss in which sense computer programs can be used as tools for supporting or enhancing knowledge-related activities. This topic may evoke first computer simulations, which have been well researched by recent epistemology. But as one thinks a little bit further, other types of programs come to mind: for example, databases as they organize knowledge and enable search, counting, groupings, and correlations between large bodies of information; these in turn enable other more specialized programs, like tools for stylistic analysis in literature studies; and then, as other examples pop up, one starts to think that any program, as long as it manipulates data with a given knowledge value, could be considered as a tool for knowledge. A supply chain management software could be considered as holding critical knowledge for the company, such as inventory value, or pending customer orders, and so on.

The questions which are raised are therefore the following: 1) can we measure the “knowledge value” of a computer program? Databases, as long as they only store and retrieve information, seem to have a low knowledge value added; on the reverse, simulations, which mint their own data
apparently from scratch, may hold much more added value; 2) how can we ascertain the epistemic link between a program, which is a “blind” symbolic process, and the portion of reality which it is supposed to represent? On this question, databases rely safely on the epistemic guarantee that is brought by the information they store, whereas simulations, as it is well known, have much more trouble to validate their link to reality.

These two questions are important for the philosophy of computing, as they may help to bring new perspectives in the understanding of the way computer programs can relate to reality and their semantics in a broad sense. Further, they may change our understanding of what is knowledge itself, if it is a capability that computer programs may enhance or even hold. Cognitive sciences’ ambition to understand mental behaviors – and knowledge in particular – through the paradigm of computing may be significantly rephrased when we understand how actual programs themselves are also vehicles and holders of knowledge.

In this communication, we try to demonstrate that three models of knowledge may answer to these questions, corresponding to three broad categories of software applications: functional applications, reactive systems and games. We view this classification as systematic.