Séminaire « Interactions entre informatique, logique et langage : histoire et philosophie »

Responsables : Liesbeth De Mol (STL,CNRS / Université Lille 3), Alberto Naibo (IHPST, Université Paris 1 Panthéon-Sorbonne, ENS), Shahid Rahman (STL, Université Lille 3) et Mark van Atten (CNRS/SND, Université Paris-Sorbonne)


fleche rouge1ère séance : Mercredi 21 janvier 2015, 14h-17h, Université Lille 3, Bât. B, salle Danielle Corbin (B1.661)

Giuseppe Primiero (Middlesex University) : « Software Theory Change for resilient near-complete specifications »

Yonathan Ginzburg (Université de Paris VII) : « Quotation, Diagonalization, and Dialogue »

flechejaune_droite_70voir l’argumentaire du séminaire et le programme de l’année.

flechejaune_droite_70voir les présentations des deux intervenants sur le blog du séminaire



Giuseppe Primiero (Middlesex University) : « Software Theory Change for resilient near-complete specifications »

Software change is a critical step in the life-cycle of computational systems. The process of modifying or re-defining specification properties of a system is partly due to its increasing architectural complexity, partly it is required to improve software quality. A major aspect of such process is that it occurs as part of the late life-cycle of the system, dictated by architectural degeneration  and flexibility requirements. But a crucial challenge is the development of a model of software change for the early life-cycle of systems, namely at their design and implementation stages, to assess and anticipate the presence of errors or bugs, and in general to determine system’s reliability. Formal specification and formal verification are the two essential approaches to this latter task, where reliability is intended as compliance of any execution of a program’s instance with respect to its intended behaviour. Despite progress in accuracy in both areas, the phenomenon of software malfunctioning is still one of enormous relevance and impact. The assumption at the basis of the present work is that software change can be modelled more similarly to scientific theory change than done so far, in order to account for both perfective and corrective changes in the implementation, when the latter is diverging from the model represented by the specification. We explore an approach to software change as operations on a model of a scientific theory, defined according to the formal operations of AGM belief change. A further assumption of this approach is that software engineering can be regarded as building and managing of the (possibly large) deductive closure of a set of axioms. But we constrain the revision operation to finite bases, as we assume that requirements specification of any however large software should be accounted for in terms of finite representation of a knowledge base. Specific to our analysis is the explicit account of both perfective and corrective changes at early stage of software development; we deal in some details with the formalization of revision in terms of safe contraction and expansion; and, most importantly, we use the properties of such operations to define a resilience metrics for the suitability of a system to adaptive changes, resulting from the accommodation of new standards.

Yonathan Ginzburg (Université de Paris VII) : « Quotation, Diagonalization, and Dialogue »

Despite its ubiquity, there are few attempts to integrate quotation into formal grammars (though many treatments within philosophy.). Part of the reason for this,
I will suggest, lies in the fact   that quotation constitutes a diagonalisation mechanism for (or against) grammars. I will explore the consequences this has for the treatment of quotation within a formal grammar for natural language. I will sketch one such account tying it to ubiquitous mechanisms of dialogue processing, which, I will suggest, has the potential of resolving some of the significant issues.

Laisser un commentaire

Votre adresse de messagerie ne sera pas publiée. Les champs obligatoires sont indiqués avec *