In March 2006, the Secretary for Lands and Physical Planning was reported to have said that ‘incorporated land groups were not owners of the land but were people legally recognised as a group’ (Post-Courier, 10 March 2006). This provoked one member of the public to write a letter complaining that

It seems that the Secretary was making a distinction between the process of land group incorporation and the process of registering titles to customary land, or between the legal mechanisms for establishing land boundaries and land group boundaries, in order to show that ‘some of the land legislation needed to change’ if these two things were to be combined. His respondent, on the other hand, was articulating an ‘ideology of landownership’ which both reflects and distorts the social relations of resource compensation (Filer 1997, 2006a). This ideology asserts that ‘clans’ and ‘land groups’ are essentially the same thing, and their incorporation only serves to confirm the fact that these ‘customary landowning groups’ are the basic building blocks of Melanesian society. But when the author of this letter says that the State ‘subjects’ these groups, he exposes an interesting fault line in this ideology, for it is not clear whether he means to say that the State itself has been responsible for the creation and empowerment of these collective subjects, or whether he means to say that they are traditional subjects which have been subordinated to the power of the State and now need to reassert their relative autonomy.

One of the ‘Ground Rules’ which the Petroleum Policy Action Team adopted at the behest of the jolly Chevroid who acted as its facilitator was for its members to ‘have fun’. And one of the main sources of amusement was the recurrent banter between myself and Tony Power on the question of whether land groups, or groups of landowners, should or should not be described as ‘clans’ in any statement of national policy. At one of our many meetings, I laid two trump cards on the table, which were photocopies of two pages in a special issue of the journal Anthropological Forum. One was taken from an article by Dan Jorgensen, and contained a passage in which he described the type of evidence which his Telefol informants had prepared for his consumption, once they knew that he was coming to evaluate their claims to ownership of land in the vicinity of a proposed copper mine close to the Sepik River.

The other page was taken from an article by Philip Guddemi, and included a rather similar observation about the ‘invention of clanship’ in another community whose members were reflecting on the same prospect.

Guddemi went on to interpret the word ‘rear’ as one which corresponds to the Sawiyanoo vernacular term for ‘following behind’, rather than the English word ‘rare’, which is perhaps what Mr Hopkis had in mind. But Guddemi’s allusion to the existence of a ‘pan-Sepik’ ideology of patrilineal inheritance (if not patriarchal authority) made no mention of the role which my fellow teamster had played in the development of this ideology through his efforts to secure the registration of customary land in East Sepik Province. So I scribbled ‘Influence of Tony Power!’ in the margin of the photocopy before I passed it over to my target, and by this means was able to banish the word ‘clan’ from his vocabulary for the next couple of meetings.

But no amount of ethnographic evidence seems likely to affect the way in which this word is used by Papua New Guineans whose thoughts lie well beyond the proven influence of Tony Power. While the Action Team was actively debating the question of landowner organisation, one member of the national intelligentsia published a newspaper article whose title boldly declared that ‘The Clan Sits at the Heart of PNG Politics’ (Anere 1998). Oddly enough, the evidence deployed in support of this statement comprised the votes cast in the Namatanai Open Electorate in the national election of 1992 — an electorate in which there were only two candidates, and both accumulated substantial numbers of votes in each of the 12 census divisions into which it was divided. Nowhere could the footprint of the ‘clan’ be seen on this particular trail of numbers. At a seminar held shortly afterwards, I drew this fact to the attention of the author, and reiterated my own argument that ‘clans’ have only come to be seen as the homogeneous building blocks of national society because they ‘have actually become groups of landowners claiming compensation from development of their resources’ (Filer 1997: 168). But the other members of the national intelligentsia who were present at this meeting seemed to be unanimous in their endorsement of the view that ‘clans’ in all parts of the country are fundamentally alike, have always been the sole collective owners of all natural and cultural resources, and naturally vote as one in national elections. No doubt many national members of the Action Team thought likewise, which might help to explain why Tony Power soon reverted to the argument that ‘clans’ are not only synonymous with ‘land groups’, but are also ‘very independent of each other, even within one tribal (ethnic, language) group’ (Power 2000[1]: 7).

Anthropologists may find it somewhat ironic that they should now be cast in the role of opposing a ‘nationalist’ ideology that seems to recapitulate the outmoded version of their own discipline which formerly informed the colonial vision of ‘native society’. That vision was based on the practical need to establish representations of the native population that would place its component parts in the spatial hierarchy of colonial administration and subject them to the most effective forms of social control (Brown 1963; Strathern 1966; Wolfers 1975). That is why they adopted the ‘African model’ of segmentary social organisation in which the principle of unilineal descent gave rise to a set of ‘tribes’ divided into ‘clans’ and subdivided into ‘lineages’ whose members were literally lined up in front of the kiap when he came to conduct a census. Melanesian ethnography began to recoil from this African model in the early 1960s (Barnes 1962; Wagner 1974; Lawrence 1984), and most practitioners have since abandoned the concept of ‘social structure’ as a figment of the modernist imagination. But Australian government officials were not content with the application of one African model to the business of aligning customary group boundaries with those of a modern state. After 1960, they also tried to apply the ‘Kenyan model’ to the registration and privatisation of ‘communally owned land’ in order to realise Hasluck’s vision of an indigenous peasant society (NGRU 1971; Quinn 1981). Indigenous politicians objected to the legal instruments of this ‘agricultural revolution’ (Ward 1972) because they would ‘grant a small number of named members of the [customary] group the power to deal with the land as if they were the absolute owners’ (Ward 1981: 250). In other words, they would have used the ‘clan agent system’ as the basis for a wholesale transformation of landed property relations.

Like the Constitutional Planning Committee, the Commission of Inquiry into Land Matters was attempting to nationalise the body of colonial law by turning ‘native custom’ into ‘customary law’ and making this the ‘underlying law’ that would inform the new laws of the Independent State. If the colonial regime had divided the native population into ‘clans’ and ‘tribes’ for the purpose of ruling over them, the new national body of land law was careful to avoid the use of either of these words for reasons that most anthropologists would readily endorse. If a customary ‘land group’ were to be granted legal status through the process of ‘incorporation’, local custom would still need to determine the name, shape and size of the group in question. So how is it that the ‘clan’ has since been reinstated as the heart and soul of the independent nation?

One answer would be that there is no other word in the English language that could be adopted as shorthand for a ‘customary social group’, and the use of this word does not necessarily mean that the speaker subscribes to any particular model of social organisation or would deny the diversity or flexibility of the Melanesian social groups to which the label is applied. If Tok Pisin speakers now use the word klan, instead of the word lain, to talk about these customary groups, this might simply show that they have moved beyond the restricted vocabulary of the colonial social order. The creolisation of Neo-Melanesian political discourse might encourage English-speaking social scientists to think that they are hearing the construction of an ideology when a new noun, like klan or kastom, appears to turn a set of social relations into an object. However, it is not single words but whole conversations which have this ideological effect, and then only because of the way they reveal and conceal what people are talking about. And in the present case, the point at issue is not the way that people vote in national elections, but the way that people get involved in the social relations of resource compensation (Filer 1997; Holzknecht 1997; Ernst 1999; Hirsch 2001).

If ‘clans’ have become subjects or objects in an ideology of landownership, it might still be argued that this is an ideology articulated by the dominant class or party in these social relations of production. In that case, the ‘Neo-Melanesian’ mindscape might turn out to be a neo-colonial vision articulated by members of an indigenous political and bureaucratic elite who represent the constitution of their own society when dealing with bodies like the World Bank or a multinational oil company. Their enthusiasm for land group incorporation might then be seen as a betrayal of the values articulated by the CILM, either because it merely replicates the colonial project of subordinating customary social groups to modern political and corporate structures, or worse still, because it has become a backdoor route to the registration and privatisation of customary land.

There is no doubt that some members of the national elite, including the Secretary for Lands, would like to strengthen or streamline the connection between land group incorporation and the registration of customary land. Furthermore, as we have seen, this connection has already been established in the oil palm sector. But this does not mean that land group incorporation and the ideology of landownership are both aspects of a neo-colonial conspiracy to subordinate Landowners to the State or the Developers.

First, we must recall that the CILM and its legal advisers were never opposed to the registration of customary land, but were only concerned to ensure that this was not done in a way that would lead to the dispossession of customary owners and the creation of a landless proletariat (Fingleton 1981, this volume; Ward 1981). Second, land group incorporation finds absolutely no favour with those expatriate commentators who have recently begun to reiterate the demand for individual, transferable land titles as an essential condition of PNG’s economic development (Lea 2002a, 2000b; Gosarevski et al. 2004a, 2004b). Third, the white advocates of land group incorporation, who might be said to represent the ‘progressive’ strand of agrarian populism, no longer have their hands on the central levers of public or corporate policy in PNG. Fourth, there are many national managers of ‘community affairs’, especially those working in the forestry and petroleum sectors, who have many practical reasons to wish that they did not have to practice the art of land group maintenance (Koyama 2004). And finally, the ‘reactionary’ strand of agrarian populism, which opposes land group incorporation on ideological grounds, is in fact articulated by members of the national intelligentsia (Lakau 1997), while the demand for incorporation now seems to be coming from people who represent ‘the Landowners’ in the social relations of resource compensation.

These are all good reasons to deny that the ideology of landownership is merely recapitulating the ‘native narratives’ of the colonial regime. Colonial administrators were content to treat ‘tribes’ and ‘clans’ as the superficial form of the relationship between the State and ‘native society’. They did not seek to lift the ‘corporate veil’ by modernising, legalising or harmonising the internal constitution of these ‘customary groups’ (Taylor and Whimp 1997: 74), but only tried to establish the social and political responsibilities of the individuals who were appointed or elected as their leaders. The ‘clan agent system’ implied the existence of ‘clans’ without granting them any kind of legal status or demanding that they act in any particular way (ibid: 102). While anthropologists were busily discovering the diversity of ‘local custom’, colonial administrators lost interest in this subject as their project changed from ‘pacification’ to ‘development’. Although they were responsible for initiating several large-scale resource development projects in the final decade of their rule, they barely began to think of ways to ‘customise’ the social relations of resource compensation.

If nowadays we find a cult of incorporation amongst the Landowners themselves, this suggests that the ideology of landownership is not the property of any one party to the social relations of resource compensation, but is a form of the relationship itself. This new African model asserts the novel, newly rediscovered power of ‘custom’ and ‘customary groups’ against the powers of the State and the Developers, but because it retains a segmentary structure, it represents a customary shadow of the State that could either be a source of compliance or resistance in the process of resource development. Landowners anticipate their relationship with Developers before the Developers arrive on their land, both in their imagination and in their social practice. They generally want ‘development’ to happen, and they prepare themselves for it. If the Developers do arrive, they also come with expectations about the Landowners, which are derived from their understanding of national and local custom, national and provincial laws and policies, their own corporate policies and practices, and those of other companies with which they are associated. That is the broad ‘mental landscape’ within which actual relationships are forged, and that is the reason why the relationship of resource compensation reaches beyond the local context of specific ‘development projects’.

The State and the Developers might think they are incorporating land groups as part of a corporate strategy to control the wayward behaviour of clan agents, landowner company directors, or other leaders of the ‘resource development cult’ (Filer 1998a: 284), but these same land groups may then turn out to be weapons of political intrigue in the hands of these same leaders (Lea 2002a; Gosarevski et al. 2004b). If ‘the incorporated land group represents an attempt to deal with the reality of capitalist relations of production through what the legislators have conceived to be the traditional form of property’ (Lea 2002a: 83), it may also be one of the main reasons why ‘[t]he state is like a landlord which cannot get the tenants out of its house’ (Taylor and Whimp 1997: 127). At one moment, the process of incorporation is represented as a way of subordinating custom to capitalist relations of production; at the next moment, it becomes the vehicle by which those relations are customised, tribalised, or even demonised.

These are all partial truths or double movements that bespeak the presence of an ideology. What this ideology reveals is a desire to standardise the social organisation of customary landowners as a condition or consequence of their dealing with the social relations of resource compensation. What it conceals is the fact that this desire is nowhere near to being realised.

Much of the debate about land group incorporation seems to assume that land groups are indeed all much alike, just as ‘clans’ are all much alike, and this view seems to be shared by the most radical proponents and opponents of any kind of land reform. What I have tried to demonstrate in this chapter is that people have been incorporating land groups for different purposes and with different outcomes in different branches of extractive industry. But that is only half the story. For if the ideology of landownership also conceals a real variety of ‘local customs in relation to land’, we should also expect to find that Landowners in different parts of the country are not all equally willing or able to sustain the incorporation of ‘land groups’ for any particular purpose, whatever the policies or strategies adopted by the State or the Developers. And that is the point at which the art of land group maintenance invites the need for social mapping studies to debunk the ideology.