L'étude des processus de création de connaissances souligne la complexité des interactions individuelles au sein des organisations productives. Cette complexité est telle que les théories de l'entreprise basées sur les incitations, focalisées sur les problèmes de traitement de l'information, peuvent ignorer une part substantielle des facteurs de décision individuels intervenant dans le contexte de l'apprentissage organisationnel, et plus largement dans les processus de production collectifs. Nous utilisons dans cette thèse la méthode expérimentale pour étudier les déterminants de la coopération, afin d'affiner les hypothèses comportementales sur lesquelles sont basées les théories économiques de la production collective. Nous montrons que les deux visions de la coopération portées par les théories de l'entreprise - un comportement devant être extrait d'intérêts divergents et une propriété émergente découlant des interactions sociales entre agents – sont étayées par les résultats expérimentaux. Par conséquent, nous concluons que les deux approches devraient être retenues et éventuellement intégrées dans un cadre d'analyse plus large. / The study of knowledge creation processes has pointed to the complexity of individual interactions within productive organizations. This complexity appears to be such that incentive-based theories of the firm, which focus on information processing issues, may fail to grasp a substantial part of the individual decision-making involved in the context of organizational learning, and more broadly in collective production processes. In this thesis we use experimental methods to study the determinants of cooperation, in order to refine the behavioral assumptions on which economic theories of collective production are based. We show that the two visions of cooperation embodied in competing theories of the firm - a behavior to be elicited from diverging interests and an emergent property stemming from social interactions among agents - find support from the laboratory experiments. Accordingly, we conclude that both approaches should be upheld and possibly combined in a broader, integrative, analytical framework.

The study of knowledge creation processes has pointed to the complexity of individual interactions within productive organizations. This complexity appears to be such that incentive-based theories of the firm, which focus on information processing issues, may fail to grasp a substantial part of the individual decision-making involved in the context of organizational learning, and more broadly in collective production processes. In this thesis we use experimental methods to study the determinants of cooperation, in order to refine the behavioral assumptions on which economic theories of collective production are based. We show that the two visions of cooperation embodied in competing theories of the firm - a behavior to be elicited from diverging interests and an emergent property stemming from social interactions among agents - find support from the laboratory experiments. Accordingly, we conclude that both approaches should be upheld and possibly combined in a broader, integrative, analytical framework.