Abstract:
Ethnographic accounts recorded from a number of its islands illustrate that marae or ceremonial structures in East Polynesia are richly polysemic, their significance spreading out in an embarrassment of directions (cf. Geertz 1980: 105). Yet, it is fair to say that archaeologists have hardly approached the great variety of meanings encoded, in marae. Most previous studies have generally constrained inquiring into the local characteristics of ceremonial structures in order to build general models of social evolution. Critical reconsideration of these insights provides a new perspective derived from contextual archaeology. The principal aim of this study, therefore, is to interpret a variety of local meanings encoded in marae in the Cook Islands in accordance with this new perspective. For this purpose, I document the considerable diversity of ceremonial structures in this island group, mainly on the basis of data which were collected during the 1985, the 1989 – 1991 and the 1995 - 1996 Keio University Expeditions to the Cook Island Group including my own 1995 research in Tongareva Atoll.
It would, however, be impossible to interpret all the meanings encoded in Cook Island ceremonial structures. Therefore, I focus on the concept of ‘cultural landscape.' It is a product which has been formed by interaction between a local culture and a local environment or by human imposition of local meanings onto a local environment (cf. Ucko 1994: xviii). Seen in this light, we can view construction of ceremonial structures as a physical action of such an event. The concept of cultural landscape, thus, involves us in taking up the issue of why marae provided with particular morphological characteristics exists a particular locale. Data related to this issue include the morphology of elements comprising a marae, the configuration of these elements comprising a marae, the configuration of these elements, its orientation, and the various features related to its location. They also include the spatial associations with other architectural features, and more general aspects of the cultural context. And comparison of these visible characteristics among marae — along with ethnographic accounts — provide us suitable contexts relevant to examining various set of meanings encoded in each marae.
The substantive research includes three case studies. The first of them demonstrates that marae are intimately related to certain aspects of the Tongarevan prehistoric society in the Northern Cook Islands which include economic, social and political as well as ideological and cosmological factors. It is also noteworthy that every marae does not necessarily share the same set of meanings. This reflects a situation in which marae turn out to be associated with various socio-political levels — ranging from households to the entire atoll — formed within a society that is segmentary in nature. Marae thus carry a variety of meanings peculiar to each social unit at various levels, as well as serving as territorial markers at a main socio-political level, that of the level of largest and most stable social units called huaanga.
The perspective developed in the analysis of Tongarevan case — ceremonial structures are richly polysemic — is also the analytical basis of next two case studies of religious structures of Rarotonga and Mangaia in the Southern Cook Islands. The comparison among the three islands, however, provides still another viewpoint. Marae in the islands exhibit homogeneous or heterogeneous pattern in morphology and location, but these patterns are not directly related to the degrees of socio-political complexity. This suggests that the realization of the local meanings which vary relative to islands must be examined in detail before linking too facilely ceremonial structures with political hegemony or social evolution in some overall framework.
The above perspective derived from three case studies leads to a reconsideration of the diversity of ceremonial structures in the Cook Island Group and throughout East Polynesia. First, the diversity can be observed most directly in the Cook Islands. This is quite sufficient to cause a revision of the prevailing image that East Polynesian ceremonial structures all share basic morphological elements. If we do not aim at constructing a general model of ceremonial structures, but instead scrutinize the diversity, it becomes necessary to refer to sets of local meanings varying not only within island and island but also within marae by marae. I believe this may prove to be an effective approach for a new range of marae studies. Moreover, it would also contribute to the development of a more recent perspective designated 'landscape archaeology' which aims at approaching various aspects of the built environment including physical, social, economic and ideological ones.