The author seeks to verify the validity of the model that he has worked out
to account for endogamous agnatic practices among the Fulani by using a va
st set: of comparative data on Africa and Madagascar. In all the examples t
aken under consideration, this analysis shows that the marriage between pat
rilateral parallel cousins reflects not an "endogamous preference" but rath
er the avoidance of a marriage with uterines. The way this model can be ext
ended to other types of marital systems is investigated. In this case, it w
ould provide an alternative to anthropological analyses of marriage systems
based on alliance or descent group theories.