Titre | Les relations causales : problèmes de définition et de vérification | |
---|---|---|
Auteur | Raymond Boudon | |
![]() |
Revue | Revue Française de Sociologie |
Numéro | 1967, 8-3 | |
Rubrique / Thématique | Mathématique et causalité en sociologie |
|
Page | 389-402 | |
Résumé anglais |
Raymond Boudon : Causal relations, definition and verification problems.
What do we mean when we say that two sociological variables are linked by a causal relation and can we verify a causal relation under non experimental conditions, situation characteristic of sociology ? The author takes up a controversy which had already been taken up by Herbert Simon in his work and by Isambert in his article (see above résumé). He shows that the notion of cause may be exactly defined from the notion of implication. But this definition remains theoretical and it is necessary to use operational definitions in its place. Concerning the verification of causal relations the author shows that the use of mathematical tools implies in all cases a certain number of postulats. Consequently one can never make an absolute demonstration of the existence of a causal relation. In the second part, the author examines also the problem of the relation between regression analysis and mathematical models of causal analysis. Source : Éditeur (via Persée) |
|
Article en ligne | http://www.persee.fr/web/revues/home/prescript/article/rfsoc_0035-2969_1967_num_8_3_3202 |