||This article discusses the extent to which a responsive evaluation (RE) approach contributed to learning by stakeholders in a case of high complexity. Fishery management in Grand-Popo, Benin is characterized
by ambiguity, that is contrasting views among fishery stakeholders about what should be done, why, how, where, and when to resolve fishery problems like the depletion of fish-stock and absence of income alternatives. It was also characterized by great gaps (mismatches) between interventionists’ plans and actions, despite generations of interventions and evaluations of their effectiveness. The RE approach aimed at facilitating interactions between interventionists and fishing people to stimulate learning and hence reduce the ambiguity and mismatches. In this article, we take distance and evaluate the results of this action research approach. We found that in the interaction some learning indeed occurred. The fishing people learned among others that intervention resources are limited and that they should organize themselves to lobby for and monitor interventions to solve their problems. Interventionists learned that they could share knowledge about their roles and limited resources with fishing people so that the latter could lobby for more resources. Fishing people however, did not learn to adopt more sustainable fishing practices. Also, interventionists did not learn to influence politicians and financial partners themselves for sufficient resources. Both categories of stakeholders developed ideas for how to
collaborate to improve fishery management. We conclude that although some single-loop, double-loop
and social learning occurred, the learning was limited and reflect on the related challenges for RE in natural resource management.