Nous avons le plaisir de vous informer qu'Armelle Thomas a pris ses fonctions d'ITA à l'IHPST. [More about]
Coming soonThursday, 17th May 2012, 12:00-17:00
Réunion du Comité Editorial de la Collection de Philosophie de la Biologie de Springer
IHPST – UMR8590
13, rue du Four
75006 Paris
Tél. : +33 (0)1 43 54 60 36
IST : +33 (0)1 43 54 94 60
Fax : +33 (0)1 43 25 29 48
Send an e-mail
International Conference: The Collective Dimension of Science
At the Archives Poincaré, in Nancy (France), December 8-10th 2011.
Organized by Cyrille Imbert (CNRS, Archives Poincaré) and Anouk Barberousse (Lille 1, IHPST).
This is a 3-day conference, with keynote speakers and submitted papers.
Keynote speakers include John Greco (Saint Louis University), Philip Kitcher (Columbia University), Paul Thagard (University of Waterloo), John Woods (University of British Columbia), Jesus Zamora-Bonilla (UNED, Spain).
>>> More on this conference
Published on Monday, 05th December 2011
Persons in charge : Anouk Barberousse
Field : Philosophy of physics and complex systems
Funding : Agence nationale de la recherche
The relationship between physics and computation has changed to a great extent during the last decades. Computers are now indispensable instruments used on an everyday basis in physics. This major "computational turn" in the physicists’ activity is likely to have deep implications concerning the epistemological aspects of their work.
The traditional notion of scientific understanding is certainly doomed to change radically due to the computational turn: traditionally related to the activity of finding out the (sometimes numerical) solutions of differential equations, it is now dependent on the possibility of reproducing certain aspects of natural phenomena by means of a computer. Moreover, in certain cases, the computer program does not contain any surrogate of the differential equations traditionally contributing to the explanation of the studied phenomena.
Other major philosophical questions implied by this new practice are: Are we witnessing a major turn in our conception of physical laws? Have the criteria of a good explanation radically changed? In the same way, a new analysis of the traditional notions of observation and experiment is required. What does it mean to observe something in a computer simulation? As we cannot, in a computer simulation, proceed the relevant information by ourselves or check the result independently (this is called "epistemic opacity"), we are obliged, in some sense, to rely on computer simulation: we loose any possible epistemic "keeping track" of what is going on in a simulation and cannot but trust the physicists managing it. This implies a renewed analysis of scientific expertise.
This project aims at bringing to light and rigorously analyzing these philosophical changes.
Unexpected links have been discovered during the last decades between statistical physics and complexity theory. In this way, first and second order phase transitions have been identified in many problems in computer science. Moreover, some methods which come from statistical physics have been used to study these problems. Some examples are the use of the replica method or of phase diagrams so as to describe the complexity of problems. One may ask what explains and justifies the success of methods stemming from physical systems in formal problems. Last but not least, such a situation allows to reformulate the theoretical status of statistical physics.
During the last two decades, studies about parallel computation have developed and various partly parallel calculators have been built. A first modest goal is to describe the cases in which parallel computation is useful in practice for the study of physical systems. Another question is to study the efficiency of parallel computation in physics and to see if this efficiency can be partly correlated with some characteristics of the physical systems (e. g. the locality of interactions). In any case, it is worth analyzing what the development of parallel computation implies for the practice of modeling. More generally, it seems fruitful to study the epistemological changes associated to the development of a science where many agents and calculators work simultaneously.
For two decades, significant efforts have been made to build quantum computing machines. These machines, in the framework of some problem classes, would be more successful than their digital counterparts. For that, the proposed computation algorithms use in a clever way some typically quantum properties of matter, as the notion of state entanglement. Are we witnessing the emergence of a new conception of what is computation, which would go further than its predecessors, because it would rely on a different physical basis? How may this computation model compell us to modify our philosophical conceptions of computation in general?
Moreover, quantum mechanics presently allows several interpretations, each of which seems to be associated in a privileged way with a mathematical formalism. This relation will be investigated and characterized. Besides, when a scientist wants to compute a given concrete result, he finds it often more convenient to use a formalism which is not necessarily the one associated with his favorite interpretation. We will look for an understanding of this cognitive tension, which seems to totally dissociate interpretation and formalism (and thus, computation).
