Vous
êtes
ici : DyXi, page
française avec
qques sous-titres en anglais --- You are here: DyXi, English homepage with some French
subtitles
DyXi
Dynamiques Citadines Collectives :
HétérogénéitésSpatiales
et
Individuelles
Urban Collective Dynamics: Individual and
Spatial Heterogeneities
CAMS
UMR 8557 CNRS -
EHESS 54,
bd Raspail - 75270 Paris Cedex 06
Projet bénéficiant
d'une aide de l'ANR (Agence
Nationale
de
la Recherche) dans le cadre du programme "Systèmes complexes et
modélisation mathématique" (SYSCOMM),
contrat
ANR-08-SYSC-008.
Project
supported by the Program
SYSCOMM
of the French National
Research Agency, the ANR.
S. Grauwin, E. Bertin, R. Lemoy & P.
Jensen,
"Competition between collective and individual dynamics", PNAS 2009
L.
Gauvin, J. Vannimenus & J-P Nadal, “Phase diagram of a
Schelling segregation model”, European Physical Journal B, Volume 70:2 (2009) pp. 293-304 (DOI:
10.1140/epjb/e2009-00234-0)
Gordon, M. B.,
Iglesias, J. R., Semeshenko, V. &
Nadal, J. P., “Crime
and punishment: the economic burden of impunity”, European Physical
Journal B,
Volume 68, number 1 (2009) pp.133-144
D.
Phan, “Other dependent preferences and moral sentiments in cognitive
economics
models: first steps towards a more “generalized” rational choice
model”, in Raymond Boudon: A Life in
Sociology, Edited by M. Cherkaoui and P. Hamilton, The Bardwell
Press, forthcoming.
Conferences, workshops
L. Gauvin, oral communication at the 2009 European
Conference on Complex Systems (Warwick, 21-25 Sept. 2009).
D. Phan & R. Waldeck, "Social norms, emotions and
cooperation in groups". 2009
European
Conference on Complex Systems (Warwick, 21-25 Sept. 2009).
L. Gauvin, Poster, Congrès de la Société Française de Physique (6-10
juillet 09,
Palaiseau)
J-C.
François et L. Sanders, "Pratiques de scolarisation et différenciation
sociale
de l’espace scolaire : 4 modèles pour un seul homme",
communication au
séminaire de Géographie-cités, Marly-le-Roi, 18-19 mai 2009
M. B. Gordon and J.-P. Nadal, Invited talks at the
Workshop
ANR-NSF on Dynamics in the Human Sciences: Cognitive, Behavioral, &
Social
Complexity, 27 et 28 avril 2009 Reims.
Events organized by DyXi
March 30, 2009, EHESS, Paris :
DyXi first general meeting
Dans ce projet de recherche
fondamentale, nous nous proposons d’utiliser des techniques de
modélisation de systèmes complexes dans l’étude des interactions
sociales et spatiales en contexte urbain : il s’agit de repérer,
expliquer, anticiper des patterns (statiques ou dynamiques)
macroscopiques qui émergent lorsque le comportement de chaque individu
est non seulement dépendant de facteurs idosyncratiques, mais est aussi
influencé par le comportement des autres, ou par les croyances de
l’individu [sur le comportement ou sur les croyances des autres, ou sur
le pattern macroscopique lui-même (sur le comportement collectif)].
Dans ce cadre général, nous nous focaliserons plus particulièrement sur
trois aspects :
Les modèles seront construits à partir de l’analyse de données
empiriques (concernant principalement Paris, mais aussi les banlieues
parisiennes et Lyon), et feront appel à des techniques de la physique
statistique des systèmes désordonnés, à de la dynamique des
populations, et à l’approche multi-agents. L’analyse statistique
permettra d’extraire des faits stylisés et les variables pertinentes.
La modélisation aura pour but d’une part de dégager les mécanismes
responsables de ces faits stylisés, et d’identifier les propriétés
génériques ou au contraire spécifiques, et d’autre part, à chaque fois
que cela sera possible, de rendre compte quantitativement des faits
observés. La modélisation – et en particulier la simulation - permettra
aussi de tester l’effet de différentes politiques. Enfin les
interactions entre les travaux sur les trois domaines choisis serviront
de base à une réflexion générale sur la constitution de « normes »
sociales en contexte urbain - l’adoption de règles implicites communes,
de comportements collectifs (méta)stables au sein de la population.
Au-delà de la seule comparaison avec les données, la validation et la
critique des modèles ferra largement appel à l’expertise des chercheurs
en SHS impliqués dans le projet.
The DyXi Project:
This basic research project, conducted
by an interdisciplinary team, uses techniques of complex system
modelling in
the study of social and spatial interactions in an urban setting. The
aim is to
identify, explain and anticipate the (static or dynamic) macroscopic
patterns that
emerge when the behaviour of each individual is not only dependent on
idiosyncratic
factors, but also influenced by “the others”. The main objective of
this work
is to explain the social phenomena that shape the dynamics of urban
segregation.
Three major objects are studied:
(1) antisocial behaviour and criminality, (2) urban
segregation and housing, (3)
school segregation,
all three
taken as measurements of social and spatial
inequality.The
models
will be constructed
on the basis of the analysis of empirical data, exploiting techniques
drawn
from the statistical physics of disordered systems, population dynamics
and the
multi-agent approach.
For the theme “urban segregation and
housing”, the aim is to explain the influence
of
social segregation on the price formation
of urban housing, taking into account the heterogeneity of individuals
and the
dynamics of their interactions. We
incorporate into models inspired by statistical physics not only the
income gaps in populations but also the dynamics of housing prices. We
hypothesise
that the price of housing is linked not only to its intrinsic
characteristics,
but also to what could be called its neighbourhood characteristics,
which can
evolve. By explicitly taking these effects into account, we shall
explain the mechanisms
at work in the evolution of a district, bringing to light the
conditions under
which mixed equilibriums will emerge (social mix), or separating
equilibria
(distinct sites of “gentrification” and pauperisation).
As regards the theme of “school segregation”, the
originality of our proposition lies in the way we take into account the
context
of the pupils and the establishment, on several levels. Here, our aim
is to
account for the effects of feedback between pupil behaviour and the
dynamic of
the setting in which they go to school. The aim is to
modelise the choice of
school made by families, with the help of a multi-agent system, drawing
on the
data and expertise of sociologists and geographers. By means of
simulations, the
model will enable us to test the effects of the abolition of
pre-defined
catchment areas on social and school heterogeneity. The main objectives
are to
better understand the rationales at work in the choice of school, by
taking
into account the different heterogeneities, evaluating the social and
spatial
inequalities in access to education and their dynamics, and then to
forecast
the main trends in the social division of schooling in the Ile de
France region.
As far as the “antisocial behaviour and criminality”
theme is concerned,
we shall be seeking to modelise in a domain where little such work has
been
carried out. One of our first steps will be to draw up an
original
database. We shall then modelise crime in
economic terms with direct application to the criminality of young
people in an
urban setting. Lastly, we shall consider the theoretical aspects of
criminality,
like the possibility of an epidemiological approach with analysis of
the dynamics
and characterisation of fluctuations in crime rates.