What progress and what risks come with the new
approaches in the social sciences? Where do
they stand in the history of demography?
Particular attention is given to answering the
following methodological and epistemological
questions: is it possible to define causality
and the search for explanation in the social
sciences? What programs are supporting research
on these approaches? Do the positions defended
by behavioural genetics have credibility, and
if not, by what can they be replaced? What is
the contribution of the multilevel approach
compared with more conventional demographic
analysis? Do agent-based models permit to
understand macroscopic regularities? What role
is there for the Bayesian approach, whether
subjectivist or logicist, in the social
sciences and in demography, where the
objectivist approach is usually employed? Is
there cumulativity in the social sciences, and
in demography in particular, and what forms
does it take? Are all the methods used in
paleodemography valid and which methods
can replace those that are not? What are the
paradigms of demography and can they be
expressed as axioms? What are the relationships
between mathematics, biology, and the social
sciences? How to modelize human migration and
the spatial distribution of populations?
This reflection – vital for demography and for
all the social sciences – has involved
researchers from around the world (see the list
of the project’s participating authors in the
detailed description of objectives). The books
in the ‘Methodos’ series published by Springer,
edited by Robert Franck, philosopher at the
University of Louvain, and Daniel Courgeau,
provide the project and its findings with an
international platform. Nineteen volumes had
already been published and two are in the publication process.
De nouvelles approches ont été développées au cours de vingt dernières années en démographie, telles que l’analyse biographique, l’analyse multiniveau, l'analyse séquentielle, l'analyse des comportements d'agents (agent-based models) ou la génétique du comportement, certaines se révélant fort prometteuses, d’autres étant sujettes à plus de discussions. Devant la complexité des méthodes que ces approches mettent en œuvre, il nous parait nécessaire de mener une réflexion plus approfondie sur les questions méthodologiques et épistémologiques qu’elles posent. Ainsi, il est intéressant de voir les progrès qu’elles peuvent apporter à la connaissance des comportements humains: permettent-elles de passer d’états de moindre connaissance à des états de connaissance plus poussée ou, au contraire, sont-elles basées sur des hypothèses erronées, qui conduisent à des résultats non valides? Egalement, il est important de replacer ces méthodes dans l’histoire de la pensée démographique, qui s’est développée depuis le XVIIe siècle, et plus généralement dans l’histoire des sciences humaines, pour voir plus précisément comment elles se situent dans l’ensemble de ces courants.