In a field usually characterized by its extremes, the middle ground of Javanese religious belief and practice is often overlooked. Studies of normative Islam in Java (e.g. Nakamura 1983, Peacock 1978a and b) have generally concentrated on urban and modernist Muslims or the traditional Islamic schools as representatives of orthodoxy, to the neglect of Islam as it is practised in ordinary villages. The typical informants in these accounts are knowledgeable and highly articulate with clearly formulated ideological positions. They are also, usually, members of fairly discrete communities and their contacts with non-santri are often limited to the workplace or market. The separation of santri religion from that of non-santri implied by this selective focus is further emphasized by recourse to typologies which exclude intermediate cases. For example, in Geertz's (1960) presentation, one gains the impression that practitioners of the three religious variants inhabit separate worlds and each is consistent in his or her separate identity. However, as we know, much of rural Java is populated by heterogeneous communities, and many individuals in these communities are neither clearly santri nor abangan but something in between.

It is this middle ground, of santri living intermingled with abangan, that I wish to explore here: a zone of compromise, inconsistency, and ambivalence which cannot be captured by a categorical opposition of santri versus abangan. My interest lies less in the delineation of types and subtypes (as within a static field) than in tracing the fluctuating interrelations between people of differing orientation – the compromises, role switches, and half-measures that make up daily life in a mixed community.