The North Mekeo classification of society is homologous with villagers’ classifications of the territorial world and the human body. From what can be reconstructed reliably on the basis of early post-contact information, however, North Mekeo society was organised at a number of different levels or scales: lineage, clan, moiety, village and tribe. I will do my best in this section of the paper to sketch the patterning of social organisation at all of these levels, which I hope to show is isomorphic with the bisected-duality-type patterning of bodily, land and territorial classifications discussed above. In this context, however, the village (paunga) is simultaneously a territorial and a social entity, and it is this relationship that I will be seeking to characterise in most detail.

The Mekeo-speaking peoples occupy the middle reaches of two adjacent river systems, the Angabunga and the Biaru. Those Mekeo who historically lived in village communities along the Angabunga have come to be known as Central Mekeo, or Mekeo proper, while those Mekeo living along the Biaru River generally are known as the North Mekeo. While all Central Mekeo speak a single dialect, North Mekeo speak two distinct dialects (Jones 1998). From before the imposition of colonial control to the present, the Central Mekeo have been organised politically into two ‘tribes’, the Biofa and the Ve’e (Seligmann 1910; Hau’ofa 1981; Stephen 1974; Bergendorff 1996). Similarly, the North Mekeo have comprised two tribal groupings, the Amoamo and the Kuipa. In pre-contact times, relations between members of different tribes tended to be hostile. Accordingly, tribal groups have been by rule endogamous.

While individual tribal groups were named in earlier times and continue to be so-named today, there does not appear to be in the language a generic term for ‘tribe’. Among the Amoamo (North Mekeo) who I have studied most closely, the tribal unit consists of four dispersed patrilineal clans (ikupu) organised ideally into exogamous patrilineal moieties (ngopu). ^[6] Each clan is typically composed of more than one localised branch, and ideally each localised branch of a clan should contain a full complement of politically and ritually specialised lineages, relating to the well-documented fourfold division of Mekeo chiefly labour between ‘peace chief’ (lopia), ‘war chief’ (iso), ‘war sorcerer’ (faika) and ‘peace sorcerer’ (ungaunga) (Figure 3).

It is these ‘residential clans’ that are usually the main landowning groups. To the extent, then, that the territories of the residential clans composing a tribe are contiguous, as is usually the case, the tribe can be considered a territorial as well as a social entity.

There are several important qualifications to this generalisation, however, which must be noted. First, the decades preceding the establishment of colonial peace were marked by considerable migration of family, lineage and clan groups across the region. As a result, many lineage and/or clan groupings nowadays lay claim to parcels of land that lie outside their own village and tribal territories. Thus members of the Akaifu clan of Nganga village consider themselves ‘owners’ of land now occupied by members of other clans belonging to other villages. And some clans lay ancestral claim to lands that otherwise are now part of Biofa or Ve’e tribal territories. The exigencies of history have, therefore, greatly complicated the correlation of tribal groups with tribal lands. Secondly, related to this, as is common everywhere in PNG, practically every claim to ownership of land is contested by other individuals or groups either within the same village or tribe or in different tribes. It is, I believe, inherent in contexts of landownership that there is no one-to-one correspondence of a single group of owners with respect to a single parcel of land owned. And thirdly, to speak of the relation of people to land in simple terms of ‘property’ and ‘ownership’ seriously distorts the character of how villagers conceptualise their ties to land and, thereby, how they see themselves and their social relations (see Mosko 2002). This aspect of the problem invokes the distinctive sense of Melanesian personhood, agency and sociality on which I will focus later in this chapter.

Typically, the members of the several clans composing a tribal grouping do not live in a single location but reside in a number of nucleated villages, usually situated adjacent to a river or other water source. Ideally, each village is composed of members of the localised branches of two clans, one from each exogamous moiety. Thus the village of Nganga consists of members of the Pitoli and Akaifu clans plus the male members’ wives and children, minus their married sisters (unless they married into the opposite clan in the same village). The nearest village, Ainapa, consists of another localised branch of Akaifu clan and the senior and junior clan branches of Ofuenge clan. It is the ceremonial clan clubhouses of these two clans that usually stand facing one another at opposite ends of the village abdomen (Hau’ofa 1981; Mosko 1985: 25), and it is the peace chief of the clan who nominally owns the clubhouse who is responsible for proceedings that take place there. ^[7] Adult male members of the clan meet in their clubhouse daily to discuss matters of common concern. And importantly, it is in their clan clubhouse that members entertain visitors from other clans, particularly those designated in special exchange relations, termed kofuapie (or ufuapie) and pisaua (see below).

Again, it is clan groups typically that are the collective owners of hereditary land—village house sites and surrounding bush lands used for gardening and hunting. It is usually the case also that villages have been established on the land owned by just one of the two (or more) clans resident there. This means that the peace chief of the landowning clan is the nominal owner of the village, but aside from his and his clan’s special relationship with respect to the land itself, the peace chief’s authority does not extend any further. That is, in all other contexts, the two peace chiefs of the two (or more) resident clans engage with one another as ritual ‘friends’ (pisaua) without any precedence of one before the other. There is, therefore, no such indigenous status as ‘village’ or ‘tribal chief’ (cf. Hau’ofa 1981; Stephen 1974; Bergendorff 1996). Peace chiefs, along with the other three types of hereditary officials I will describe below, are essentially clan functionaries. Aside from the local situation of the chief of one clan owning the land occupied or used by members of another clan, there is no chief who wields authority over members of other clans beyond in-marrying spouses when they attend mortuary feasts and the female members’ offspring (papie ngaunga, or the ‘women’s children’ of the clan).

Nonetheless, at the societal levels of village and tribe, which correspond roughly with territorial contiguities as noted above, there is an important coordination of functions that corresponds with the processes of inside/outside transfer in relation to classifications of territoriality and the human body. To illustrate this, I will describe the indigenous conceptualisation of clan groups as themselves analogous to human bodies of two gendered types, male and female, with certain kinds of interactions between them. For it is in these terms of interacting clan bodies that village and tribal entities or ‘chieftainships’ are constituted. Here, particularly, the relevance of the inside/outside classification of space/time and body domains for societal organisation becomes apparent.