Peter Oksen's comments on our paper about farmer–herdsman relations in Burkina Faso raise some interesting issues, notably regarding problems of interpretation of oral and archival sources and regarding the broader relevance of insights gained from an in-depth case study. Before answering straightforwardly to his objections, it is useful to clarify the misunderstanding which appears to exist about the meaning we attribute to ‘symbiosis’ and ‘symbiotic relations’. In our article we restricted the use of these terms to the way in which past relations between farmers and herdsmen, depicted as undifferentiated groups, are often represented. In discussions about the change of these relations, the emphasis is on progressive deterioration, again without attention paid to the possible existence of intra-group differentiation or of heterogeneity of relations across group boundaries. In this regard, it is remarkable that from colonial documents the interests of farmers and herdsmen emerge as equally irreconcilable as they are often considered today, and that, just as at present, competition over scarce natural resources constituted a major factor in inter-group relations. If we therefore reject ‘symbiosis’ as a correct description of formerly existing inter-group relations, we do not intend to imply that complementary links – such as those we describe for present-day relations between Mossi and Fulbe – did not exist in the past. A major aspect of our argument is that inter-group relations, whether past or present, cannot be subsumed under simplifying labels such as ‘symbiosis’. Neither can changes in these relations be understood in terms of uni-directionally processes of deterioration. Hence, present-day ‘complementary’ links across the ethnic boundary – established by certain, but not all, Mossi and Fulbe actors – are but a manifestation of the continued presence of diversity of relations, not of ‘symbiosis’. They point to mutual interests between certain actors belonging to different ethnic groups, not between the groups as such.