Many of the new ILGs at Lake Kutubu are the result of large clans such as Wasemi Fo'omahu'u Orodobo or the large Damayu and Fiwaga clans splitting into constituent subclans and lineages. These new applications are defensible in terms of population growth alone, which was the most common precipitating cause of clan fission. In Foi, ira (‘tree’) is the term applied to three generations of male descendants of a single man. Practically, it takes the form of a group of full brothers whose land is normally contiguous — in other words, they live near each other as well as being closely related. The ILGs that were audited at Wasemi were all ‘trees’ of the Orodobo and So'onedobo clans — the two biggest clans at Wasemi. I have previously identified the ira as the property managing and work-related cooperation group within the Foe social system (Weiner 1986, 1988a), and so there is nothing non-traditional about this kind of division — it merely gives formal ILG recognition to a unit that is already explicit and visible in the Foi social system.

The ILG, though based on principles by which clans are defined as landowning entities, is not the same as the clan per se. The Foi themselves are clear about the different functions of ILGs and their traditional clans. For example, a single clan may consist of two or more subclans for the purposes of ILG recognition and land stewardship. But the entire clan still acts as a unit, for example, in the collection and receipt of bridewealth for its female members.

A singularly appropriate feature of the LGIA is that it defines the land belonging to a landowning group not in terms of a discrete unbroken border, but in terms of a list of specific sites over which its constituents exercise what for all intents and purposes are the prerogatives of ownership. In this context, ownership is defined as control of access to the site or ground in question. Among Foi the local clan exercises a sort of nominal communal dominion over its territorial resources, but effective ownership, that is control of access, is always exercised by specific individuals or, at most, a set of full male siblings and their father (Weiner 1986, 1988a).^[8]

The issue at stake is not just control of resources but also the mechanism and locus of decision making at the local level. As a result of there now being two social units, the ILG and the clan, clan-wide decisions concerning things such as bridewealth and ceremonial will be split off from decisions concerning resource management and distribution. It is an open question whether, in terms of traditional custom, the control of land was seen as distinct from all these other clan functions.

The local Foi group reached consensus not by convergence upon a common interest but by the temporary rhetorical abeyance of the fissive mechanism that really ‘founds’ group formation (Goldman 1983). In traditional times, the local clan rarely acted as a single unit, except ceremonially, as in bridewealth distributions for example. Land decisions were the affairs of those directly involved. Although disputes over land within the clan were commonly adjudicated, they did arise. Disputes over distribution of bridewealth were also common.