sion" which begins with the ritual, indicates the new complexity of a dimension of female identity. Girls acquire new responsibilities towards their families, while their wish to engage in courtship demands a loosening of family control. Girls are symbolically handed over to the male domain, but they also experience self-empowerment in the ritual. After the ritual, however, their freedom of action is often reduced, as well as their autonomy in relation to the male domain--this tighter control being one of the causes of early marriage. The ritual thus expresses discontinuities as well as continuities ( Crapanzano 1992) in the process of becoming a mujercita.
The ritual cannot be reduced, as the CEBs' discourse implies, to its functions of family status differentiation and "consumption". On the other hand, the ritual cannot be fully understood if we focus only on its symbolic level: on the ways it marks sexual boundaries, and helps to construct the female body as a vessel which needs to be defended from male philandering. The ritual is a performative act, an experience which may or may not be part of the process of creation of female self-identity. Indigenous exegeses of the ritual (and, indeed, of reasons for foregoing it) need to be understood within the context of particular sets of family relations--which are very heterogeneous--and also differing contexts of religious discourse and perceptions of class and status. For a wide range of economic and other considerations affect peoples' decisions about whether to celebrate the ceremony and, if so, on what scale.
The material in this article is drawn from my thesis ( Napolitano 1995), based on eighteen months of fieldwork between the summer of 1990 and the spring of 1992, funded by the University of London Scholarship Fund. I am in great debt to Richard Fardon, Peter Worsley, John Gledhill and Stephen Hugh-Jones for their insightful comments on early and later versions of this article.