INTRODUCTION

In his influential essay 'Are there social groups in the New Guinea highlands?' Roy Wagner wrote:

When the white men first came to Karamui they felt a strong obligation
to discover groups. They were administrators, faced with the task of
building an inter-face between the native's 'institutions' and their
own, and intent on resolving a confusing array of names and
settlements into groups that could serve as the final (local)
constituents in a political chain of command (1974:115).

Early anthropological and linguistic studies in New Guinea were, similarly, concerned to demarcate and label groups that, on the basis of some selected characteristics, appeared coherent. 'Groups', Wagner wrote, 'were a function of our understanding of what the people were doing rather than of what they themselves made of things' {ibid. 97, original emphasis). Regions, languages, tribes, moieties, phratries, clans, lineages, and so forth were invoked in attempts to order the extraordinary diversity of human populations in New Guinea (e.g., Hays 1993; McElhanon 1971; Weiner 1988). Names that people used to connote relationships, and employed strategically in contexts that were forever shifting, were taken to be denotative, to express fixed structures when local concerns were, so often, with fluid process.

In the aftermath of Wagner's essay, an emphasis on the ways in which social relations were variously generated or dissolved, and strategically employed, came to characterize much Melanesian ethnography. (1) But as notions of process and flux were fore-grounded in scholarly discourse, Papua New Guineans themselves, both local people and an emerging bureaucracy, had recourse to the academic language of earlier times and put in place, or appeared to put in place, 'groups' that resembled the corporate entities of Western societies. Particularly in contexts of large-scale resource extraction projects, social collectives were increasingly represented as fixed in both composition and place--as unambiguously 'clanlike'--and, hence, to local people, visible to those who would distribute benefits and, to the bureaucrats charged with this task, legally manageable (e.g., Ernst 1999; Filer 2007; Jorgensen 1997; Wesch 2008).

There has been a sense, therefore, that as anthropologists and other scholars have deemphasized fixed structures in favour of fluid process, Papua New Guineans themselves have moved in the opposite direction. The scholars have opted for uncertainty. The nationals have opted for certainty. But, as Leaver and Martin (2016) rightly point out, such a characterization overstates the case. In Papua New Guinea (PNG), the names associated with places and with groups of people have always had both denotative and connotative aspects; they have been both 'markers and makers' of identity (Martin 2009). Nor, indeed, do supposedly modern corporate groups qualify as unambiguously bounded entities. They too are in flux, their members responding to changing contexts by strategic adjustments of the status quo.

This paper explores some ways in which people from a small, lightly populated area in the northeast of the Western Province of PNG have responded to the incursions of colonial officers, missionaries, companies intent on resource extraction, and national government by altering ways in which they employed names to variously connote or denote particular assemblages of people. Their most recent efforts--since the mid-2000s--have been prompted by the presence in their area of exploration and drilling camps associated with what became the multi-billion dollar PNG Liquefied Natural Gas (PNG LNG) project (Dwyer and Minnegal 2014; Ernst 2008; Goldman 2009). Two case studies provide our primary illustrative material. They concern people who have experienced outside contact, but little 'development', through the past five decades. They reveal ways in which people, with an eye to a desired future, have increasingly adopted terminological conventions with respect to language, group and place identification that differ greatly from previous practice. These new conventions were initially imposed by colonial officers and missionary-linguists who recorded a structural order that did not in fact exist. The conventions were subsequently implicit in the recording practices of the agents of petroleum companies and government departments tasked with ascertaining the 'owners' of 'bounded' areas of land that had not previously been either 'bounded' or 'owned' in the ways expected by those who required this information. The two cases also reveal, however, that though what people are doing appears to be directed at satisfying the requirements and expectations of outside authorities those people are simultaneously aware of both the potential repercussions of their decisions for local relationships and the consequences those decisions may have for challenging or safeguarding identity.

BACKGROUND

In this article, we are concerned with a relatively small area of PNG that is located between the Nomad (Giluwe) and Upper Burnett (Na) Rivers in the Western Province (Fig. 1). To the east this area is bordered by the Muller Range; to the west it extends somewhat beyond the Strickland and Murray (Ogua) Rivers. From south to north the landscape changes from lowland (with swamps to the west either side of the Strickland River), through foothills, to mountainous terrain north of the Carrington (Osio) River. Our two ethnographic cases concern areas north of the Cecilia (Baiya) River. (2,3,4)

In the 1960s, when a government station was first established at Nomad, less than 1500 people lived in this area. They were hunter-horticulturalists who were widely dispersed as small, longhouse-based communities (Dwyer and Minnegal 1992). They often moved the location of settlements, and the assemblages of people who came together at these settlements frequently changed in response to marriages, initiations, disputes, sorcery accusations, cannibalistic raiding, and a fragile demography (Minnegal and Dwyer 2000a).

Through the 1960s and early 1970s the reports of government officers patrolling out from Nomad include many names of purported languages, 'groups' and places (e.g., Barclay 1972; Cawthorn 1970; Johnson 1968; Paterson 1969a, 1969b). Agala, Bogaia, Bugoti, Daba, Eobi, Febi, Konai, Kubor, Orabia, Samo, Siali, Supei, Tsinally, Uwo, and Wato feature as possible languages north of Nomad while Alibu, Bedamuni (Biami), Bibua, Gebusi, and Honibo are named as language groups to the immediate south and east of Nomad. (5) North of the Cecilia River, names attributed to 'groups' include, among others, Augose (Augusi), Bilatie, Deima (Dema), Foisoso (Foisubi), Gumitie, Headubi, Hwotie (Wuotie), Kesomo, Kofebi, Nomu, Yawuasoso, Ulatie, Watia, Woson (Wosabi), and Wuo (Uwo). In most cases, places where people from these named 'groups' were encountered match places still associated with those names after 2000. In a few cases, government officers recorded 'group' names as place names; for example, the 'group' name Headubi became attached to a temporary village and features on the Nomad topographic map (PNG 1:100,000 Topographic Survey, Sheet 7385 (edition 1) Series T601, Printed 1979). Furthermore, place names themselves sometimes 'moved' when most of the people from a longhouse community at, for example, Soabi relocated to a new location and government officers named the latter Soabi 2 (Fig. 1).

Government and mission influence led to amalgamation of formerly dispersed long-house communities as larger and more permanent settlements, though these changes emerged only slowly north of the Cecilia River (Minnegal and Dwyer 2000a; Shaw 1996; Suda 1990). In parallel, language identification became more precise, with Samo, Kubo, Konai and Febi--the last sometimes named as Agala (6)--recognized by outsiders as closely related though distinct and, to the north, the unrelated language Bogaia grouped with Duna (e.g., Shaw 1986; Stewart and Strathern 2010). Similarly, many of the 'groups' reported and named in patrol reports came to be understood by colonial authorities as resembling patri-lines, lineages, or clans though, as might be expected, for the people themselves these 'groups' were fluid and contingent, lacking the definitional coherence implied by the outsiders.

Kubo spoke of these named groups as oobi, a term that translates, literally, as 'man mound'--an assemblage of people brought together, as a megapode (djago) rakes together the leaves on the forest floor to make a mound (djago bi) in which to incubate its eggs. Perceptions or constructions of similarity based on birth, fosterage, residence, and sustained use of particular areas of land may all influence oobi identification by self and others. Indeed, identification with more than one oobi was not uncommon, based in nurturing as much as birth, and strategic realignments to pursue personal or political agendas were widely accepted. While some groups were identified as 'brothers' based on real or Active territorial proximity, and marriage between members of these groups proscribed (Minnegal and Dwyer 1999:64), there was no prescribed or preferred marriage arrangement based on pre-existing affinal relationship or group affiliation. (7)

It is not uncommon in this region that different lineages within the same named oobi are associated with non-contiguous areas of land, sometimes in the territories of different language groups. For example, there are five groups of Headubi people, linked through origin stories, which are associated with non-contiguous areas as much as 25 km apart on Konai, Febi, and Kubo land. (8) Such groups tend to be distinguished by reference to focal sites on their respective lands. Crucially, however, this spatial separation may result in the eventual dissolution of perceived siblingship. (9) In this region, it seems, 'proximity rather than linearity invokes the proscription on marriage' (ibid.).

In the years following establishment of the government station at Nomad there was intermittent exploration for both mineral and petroleum resources. This activity intensified in the mid-1980s when wells were drilled in mountains north of the recently established Suabi mission station on the western fall of the Muller Range. The area became known as Juha and, by 2010, with five potentially productive wells drilled, was included as the most remote gas field of the PNG LNG project (Ernst 2008). A desire to access monetary benefits that were expected to eventually flow from these wells had major consequences for ways in which people in the area under consideration sought to be known both by outsiders and by each other (Minnegal and Dwyer 2017).

By 2014, in the area north of Cecilia (Baiya) River, Suabi village had an airstrip, a community health centre, and elementary and community level schools--though the last was only intermittently staffed. Contact with mission and medical workers beyond Suabi was provided by radio and, since 2011, there had been intermittent mobile phone reception. Komagato village, west of the Strickland River, supported a one-teacher elementary school. Facilities of these kinds were not available at other communities between the Cecilia and Burnett Rivers.

It is against this background that the shifting use of names and identities deployed by people in our two case studies must be read. The first case directs attention to ways in which a group of people asserting rights to widely dispersed areas of land present themselves to outsiders but do so in ways that establish the grounds for deflecting potential challenges from neighbours. The second case presents a finer-scale history of decisions made by one individual as he sought, on a national stage, to secure the rights of those to whom he felt most responsible and, simultaneously, on a local stage, asserted his right to make these decisions with respect to the area of land in question. In the text and Endnotes that follow, local people are referred to by pseudonyms.

THE PEOPLE OF GAMLIHAI

In the years 1986 to 1999, we learned of four groups of people who were named as Kesomo and we personally knew people from three of these groups. Abai hafi Kesomo and Tebesutie Kesomo held land north of the Osio River; Toio ho Kesomo held land south of Osio River and Ia hafi Kesomo, with no living members, had held land in the area where the stream Ia joined the Strickland River (Fig. 2). (1) These groups were spoken of as though they were distinct oobi. (11) The apparent distinction between these groups was reinforced by the existence of cross-cutting marriages--endogamy is regarded unfavourably by these people--but was weakened by the exceptionally strong and long-lasting bond established between men from different oobi who regarded their mothers as sisters, even in cases where those women were born into different Kesomo groups. (12)

In 2007, Tom Ernst (2008) undertook a full-scale social mapping and landowner identification (SMLI) study of Petroleum Retention Licence area 2 (PRL2) on behalf of ExxonMobil, the operator of what was to become the PNG LNG project. The focus of his report was with groups of people he considered to be members of the language group identified as Febi. In Table 9 of the report he listed these groups, as clans and subclans (2008:61-2). (13) A clan named Gosomo was reported to be made up of three subclans (Abaya-aufitie, Nauaufitie, and Debesutie) with the primary place of residence for Gosomo people recorded as Tobi 2. Gosomo, Abaya-aufitie, and Debesutie correspond to the names we render as, respectively, Kesomo, Abai hafi, and Tebesutie. Tobi 2 has been adopted by local people as an 'official' name for a village that was established in the early to mid-2000s at the junction of the stream Molo with Dogomo River. (14) The name Nauaufitie is more problematic though its referent is probably the junction of the river Na (or Nali) with the river Dio; that is, the junction of the Upper Burnett with the Lower Burnett River.

Kesomo people had little input into Ernst's report. But stimulated by knowledge of the work he was doing, and by their desired engagement with the PNG LNG project, some of them prepared a 13 page booklet titled The Origin of Kesomo (Taprin 2007-08). In a reported origin myth, the founding ancestor Kesomo emerges from a sago palm at a place known as Gamlihai, marries a woman from there and has four sons who establish the 'sub-clans' Tebesutie Kesomo, Abai hafi Kesomo, Dio hafi Kesomo, and Toio ho Kesomo. (There is no reference to Ia hafi Kesomo.) It is said that the originating sago palm is still present in the swamp at Gamlihai, that it never changes shape and has never flowered (Mora 2015:40-1). The group Dio hafi Kesemo presumably corresponds to Nauaufitie Kesomo of Ernst's tabulation, with the geographic emphasis now on the junction of Dio with the Strickland River rather than Na with Dio (Fig. 2).

The Origin of Kesomo tabulates purported descendants of Kesomo's four sons and records the names of many mountains, rivers and creeks, waterfalls, lakes, swamps, caves, and sacred places associated with each subclan. (15) A detailed map marks the location of many of these places for Tebesutie Kesomo, Abai hafi Kesomo, and Dio hafi Kesomo (Fig. 3). Several of those places, such as Gamlihai and Biguhai, have associations with multiple groups, the former as the origin place for all Kesomo, the latter of importance to people of Botie, Gumitie, and Ulatie clans in addition to the subclans Tebesutie Kesomo, Abai hafi Kesomo, and Toio ho Kesomo. Furthermore, in agreement with an annotation on the map that 'these subclans have only one land, no border' the subclan areas encompassing the mapped sites imply considerable overlap.

In 2009, and through the years that followed, possibilities engendered by the PNG LNG project were often focal in the imaginings of Febi and Kubo people. In interactions with government agencies and petroleum companies, people attempted to ensure that they, or the group of people with which they were aligned, were recognized and listed on official documents as legitimate beneficiaries of future wealth (Minnegal et al. 2015). Simultaneously, however, people came to understand that an unambiguous association with a specified area of land had the potential to elicit benefits if that area was found to harbour resources of interest to mining, petroleum, or timber companies. This latter understanding led to emergence of a concern with 'borders' and 'ownership' that had not existed in the pre-colonial era and, indeed, was merely insipient by the year 2000 (Minnegal and Dwyer 1999).

In May 2009, hundreds of stakeholders from all areas impacted by the PNG LNG project met at Kokopo, East New Britain, to ratify overarching arrangements for the distribution of benefits. In December of that year, representatives of people with claims to Juha met at Moro, Southern Highlands Province, to ratify arrangements for the future distribution of benefits derived from the Juha wells. In November 2013, the PNG Department of Petroleum and Energy (DPE) organized a 'clan vetting' meeting at Siabi, near the Juha well heads, with the intention of finalizing a list of clans and subclans, together with their named representatives, that qualified as legitimate beneficiaries of royalties to be paid when those wells eventually came into production. (16) Some Kesomo people had input into at least the second and third of these meetings. At the Moro meeting Kesomo was listed as one of 12 Febi clans, none of which was noted to be further subdivided. However, the five men who attended as Kesomo representatives were associated with, respectively, Dio hafi Kesomo, Tebesutie Kesomo (2), Abai hafi Kesomo, and Toio ho Kesomo. (17)

At the 'clan vetting' meeting of November 2013, Kesomo--previously depicted as a single assemblage of four equivalent groups, all descended from one of four brothers--was now presented as two 'major clans' named as Kesomo and Tebesutie. Each of these, moreover, was declared to be further divided. The former comprised nine subclans, four of which were Dio hafi, Abai hafi, Toio hobe, and Ia hafi. The remaining five subclans in Kesomo, and the seven listed for Tebesutie, had been devised for purposes of the vetting process and were named for mountains and other landscape features. (18) This was the first occasion at which Ia hafi Kesomo received formal recognition on a government document. And while previous depictions had identified Tebesutie, Abai hafi, Dio hafi, and Toio ho as constituting equivalent social entities, each descended from one of the four sons of Kesemo, now Tebesutie was elevated to equivalence with Kesomo as a separate major clan.

In 2015, Kesomo people learned that they themselves might be major players in the developing PNG natural gas industry. The petroleum company Repsol planned to drill two wells about 4 km west of Tobi 2, close to the junction of the rivers Dogomo and Abai. In mid-2015, and in accordance with the PNG Oil & Gas Act 1998/2001, Repsol contracted a preliminary social mapping study of the impacted area (Mora 2015). (19) Tobi 2 was the base from which information was gathered. On this occasion Kesomo was again treated by local informants--20 of 28 identified as Kesomo--as a single clan with five subclans, named as Tebesutie Kesomo, Abai hafi Kesomo, Dio hafi Kesomo, Toio ho Kesomo, and Ia hafi Kesomo.

What emerges from the above history is that, in the years from 1986 to 2015, the representations of Kesomo by others or by themselves have grouped or divided people in a variety of ways. The more recent representations, since the mid-2000s, have sought to satisfy the presumed expectations of government agencies by creating numerous subclans--probably more subclans than there are extant Kesomo families--and, in particular cases, to assert rights to land where there is little or no evidence of recent use by Kesomo people. On the latter count, the ways in which Dio hafi Kesomo and Ia hafi Kesomo have featured are significant.

The name Dio hafi Kesomo did not come to our attention until 2007-08, when it featured in The Origin of Kesomo and, as subclan Nauaufitie, in Ernst's (2008) social mapping report. In the 2014 DPE list of major clans and subclans, and in Mora's 2015 Repsol report, the name of only one man is recorded as associated with Dio hafi Kesomo. (20) The impression left by available information is that the label Dio hafi Kesomo is, increasingly, given prominence to ensure that land in the vicinity of that junction is explicitly associated with Kesomo people. It seems that the lower reaches of the Dio watershed have not been used since 1993, when most residents at the Bogaia-dominated community at Koge died in a landslide. It may even be that there are no surviving members of the Kesomo lineages that once resided in this area. And they certainly did once reside there for, ironically, the first recorded mention of Kesomo (as Kesomobi) people is from the lower reaches of Dio River where a 1970 patrol led by Cawthorn (1970) camped for several days and received an airdrop of supplies from Nomad. Two years later, Barclay (1972) mentions a 'KESOMO house and garden about 1/2 mile up from the confluence of the Strickland and Burnett Rivers, on the South bank up about 1000 feet'. Barclay's party also spent several days in the area where they received supplies from Nomad by helicopter. (21)

The case of Ia hafi Kesomo is even more clear cut. For more than 30 years there have been no living Ia hafi Kesomo and the land they once held has been regularly used by people from other oobi. It has not been used by people from other branches of Kesomo and, judging from The Origins of Kesomo, neither the land nor the name had much salience for those other branches in the years to 2008; no son of Kesomo is identified with Ia hafi. At the 2013 clan vetting exercise, however, and subsequently, some people have named Ia hafi as a subclan of Kesomo, in effect an opening bid for eventually asserting a claim to that land. They were, in this way, exploiting official mechanisms of asserting rights to land as a means of bolstering their position in an emerging dispute with people of other oobi who actually used the land, though they were not born of it. Those Kesomo who now claimed the land named a recently deceased woman from Komagato on the west bank of the Strickland River as Ia hafi Kesomo--though, in fact, that status belonged to the woman's mother and not to the woman herself. In the context of official documentation, where those named as representatives of groups are expected usually to be male, a Kesomo man, orphaned as a child, was listed as the representative for Ia hafi Kesomo. No other member of this group was named, and assertions that some members of this group lived at Komagato were without foundation.

Finally, it is noteworthy that from 1986 to 2016 it was only in the context of Ernst's social mapping study, where the specific focus was with Febi, that people who talked of Kesomo made any reference to affiliation with a named language group. The separation of Febi and Kubo groups depicted on Figs. 1, 2 and 4 emerges as a Colonial artefact though, with Osio River specified as the border, it is now routinely acknowledged by people who are asked.

'WE ARE FIRE CLAN'

In April 2014, we asked Martin to sketch the outline of his clan land on a topographic map that we carried. We did not name the clan. Martin was familiar with this topographic map. Very carefully, he outlined an area that was bordered to the north by Osio River and was cut through the middle by a stream named Hio (Fig. 4). We asked him to name the area he had enclosed. It was the land of Osumitie, he replied. At our request, he now divided the area up. From east to west he marked out areas he labelled Osumitie, Sowasoso, Bogua, and Tiamososo. (Hereafter, we refer to the larger collective as 'greater Osumitie' and the subsection of the same name as 'lesser Osumitie'.) The names were not those of the areas depicted. They were, rather, the names of sets of people associated with those areas, though for two of these sets--lesser Osumitie and Sowasoso--there were no living representatives. Indeed, Martin himself had never personally known lesser Osumitie or Sowasoso people. His knowledge of them had come from his now-deceased father. The four sets of people, he told us, share a 'special place' that is located on Bogua land, a mountain named Biyo where fire originated and was first found by people.

We now asked Martin to write on the maps the names of those peoples who bordered greater Osumitie. In the southwest corner he wrote Dobiti and, moving clockwise, Sisu, Kesomo, Gumitie, Yawuasoso, Koli, Nomo, Iodibi, and Udubi. Sisu was a branch of Tiamososo that, unlike their brothers, did not share the 'special place' on Bogua land, Kesomo was represented south of Osio River by a branch at the headwaters of Toio Stream and, at our request, Martin noted the location of Wuo northeast of Yawuasoso at the headwaters of Masi River and of Woson south of Osio River towards its junction with the Strickland River.

Martin told us that when 'social mapping' was done for the Juha gas fields, the 'western' border of the mapped area followed Hio to its junction with Osio River and then, for some distance, followed the latter river. The outcome of these external mapping decisions, he said, was that Sowasoso, Bogua, and Tiamososo were not included as potential beneficiaries of Juha.

Juha SMLI was undertaken in 2007 (Ernst 2008). At that time the focal area comprised 10 graticular blocks licensed as PRL2 (Fig. 4) though Ernst sought to accommodate all groups of people he understood to be Febi rather than merely those who had specified association with land in PRL2. He listed Osumitie as a Febi clan though he was uncertain where it was located (2008:562). Graticular block 1709, within PRL2, included much of the area Martin mapped for us as lesser Osumitie. Martin, it seems, had confounded the tentative and imprecise drawing of boundaries associated with social mapping, always a relational exercise, with the rigorous designation of boundaries required in taking out petroleum licences (cf. Wesch 2008). And, moreover, he assumed that where borders were mapped they would, logically, follow landscape features such as watercourses rather than 'lines' based on the abstractions of latitude and longitude.

At the 'clan vetting' meeting held at Siabi in November 2013, DPE officers planned to identify groups of people who qualified as 'landowners' of what had now become Petroleum Development Licence area 9 (PDL9), an area comprising six graticular blocks that had been excised from PRL2 (Fig. 4). In this exercise, DPE officers worked from a provisional list of potentially eligible groups of people that, in the case of those declared to be Febi, included all clans identified by Ernst (2008:561-62) irrespective of any geographic association with either PDL9 or the more extensive PRL2. On this basis, Osumitie was listed despite the fact that none of its land fell within PDL9. DPE officers, however, had heard that Osumitie was without living representatives and announced during the meeting that the name would be removed from the list of eligible clans. (22)

Martin was present, however, and responded. 'We are here', he called; 'we are fire clan.' He asserted that Osumitie was the name of a major clan that included six subclans: Osumitie, Sowasoso, Bogua, Tiamososo, Dobiti, and Sisu. He asserted, also, that the sub-clans he named were united in sharing a story about the origin of fire. The list produced by DPE adopted Martin's recommendations with the exception that, while Osumitie was named as a major clan, the same name did not appear as one of the listed subclans within that collective.

In Martin's understanding he had now done all that was possible, with respect to the PNG LNG project, to secure the future position of those to whom he was most closely related and to whom he felt his greatest responsibilities. To us, however, he insisted that the public position he had adopted was not ideal.

We first met Martin as a boy, about 8 years old, in 1987. He was initiated in 1991, married in 1999 and, by 2014, was father to four children. His mother had died when he was young. His father remarried, had three daughters by his second wife and died in 1999. In these years, Martin identified as, and was said to be, a member of an oobi named Bogua. His primary place of residence was Suabi, though he had spent lengthy periods at Nomad (vocational education), Balimo (training as an elementary school teacher) and Debepare (Bible school). As an adult, influenced by his father, he considers that lesser Osumitie, Sowasoso, and Tiamososo 'came from Bogua' and that it is the last named that is central with respect to the origin of fire. Indeed, with no surviving members of lesser Osumitie and Sowasoso it became the responsibility of Bogua to use and care for the land that had been associated with these groups. In short, to Martin, the lands of lesser Osumitie and Sowasoso should be seen as falling within the ambit of Bogua, and under his care, rather than Bogua being seen as subordinate to a greater collective named Osumitie. Indeed, to both reinforce his connection and claim to lesser Osumitie, Martin had given his eldest son, born in May 1999, the custom name Woiku in recognition of a long-deceased man whose 'land is Hio Hoi' (Hio River on the land of lesser Osumitie). (23) Martin's long-term plan was to change the name of the collective to Bogua but, in the meantime, he argued, it was best to use the name Osumitie because that is the name recorded by Government and, since only lesser Osumitie land falls within the nominated Juha area, it is necessary to identify with that 'name' to secure a share of benefits.

At the Siabi 'clan vetting' meeting Martin had named six groups as subclans of Osumitie and asserted that these were united as 'fire clan'. Later, he was insistent that in truth neither Sisu nor Dobiti qualified for this status. They did not share with the other groups the special place from which fire originated. He asserted that these two groups were invited to join the major clan Osumitie because otherwise they would not receive royalties from Juha. In fact, close patrilineal ties between Tiamososo and Sisu informed one of those decisions and long-term access to fish and sago resources accorded by Dobiti to mountain-dwelling Bogua informed the other. These invitations, however, did not mean that Sisu and Dobiti now shared the 'special place' on Bogua land. They were eligible to participate in registered Incorporated Land Groups established through reference to the major clan Osumitie but they were not, Martin stated firmly, 'owners' of that land.

In the years to 1999, we had assumed that Martin was Kubo though, in fact, we had never asked. We did ask in 2012. He equivocated, suggesting that he was in some way 'mixed'. In 2014, however, when shown Ernst's list of Febi clans he insisted that Ernst was wrong with respect to Osumitie: 'I am Kubo', he declared. By now, however, Martin was confident that Osumitie had been listed by DPE. He no longer felt that aligning with or as Febi might pay further dividends.

DISCUSSION

The case studies summarized above reveal ways in which the naming of social groups by some Kubo and Febi people shifted as they attempted to comply with their understanding of the expectations of outsiders while, simultaneously, attending to potential challenges by their neighbours. In a context where substantial financial benefits were at stake then, whether attending to external or internal concerns, people were motivated by considerations of their own future well-being or the future well-being of those with whom they were most closely aligned and to whom they had responsibilities.

In the years to 2000, people seldom mentioned the names of groups that had no living members. It was in the context of eliciting details of genealogy that we were most likely to learn of their previous existence. Named oobi, however, often had few members, sometimes a single lineage. Extinction was not uncommon and when this happened the land associated with that oobi was, quite rapidly, merged with and subsequently identified with the contiguous land of a 'brother' oobi. (25) Our failure, through a period of 13 years, to record the name Dio hafi Kesomo, and the limited reference to Ia hafi Kesomo, reflected the fact that, without living members, these groups lacked salience. It was what living people were currently doing on the land--where they were living, where they were gardening, where they were hunting or fishing, and who they were interacting with--that was of interest and, ultimately, came to establish understandings of proper connections between particular living people and particular places. Disputes about land were not a common emphasis of the local social geography of the people who lived in this area of Papua New Guinea.

To Kubo and Febi people, the advent of the PNG LNG project seemed to promise future wealth to those judged to be legitimate beneficiaries because they held land, or were connected to those who held land, in the area under licence to the petroleum company. For Kesomo people, none of whom held land within PDL9, their primary objective at the 2013 clan vetting meeting was to maximize representation on DPE lists and to do so in ways that were in accord with DPE requirements. They were fortunate in that DPE officers had no personal knowledge of either past or present connections between people and land, and were reliant on local advisors who did not challenge the legitimacy of any supposedly Febi clan. To achieve their ends, Kesomo people provided the names of clans and subclans that created a semblance of well-ordered structure and, simultaneously, accommodated each extant lineage (or family) by naming one person as the representative of each subclan. In these ways they sought to render both a quality of Kesomo-ness and the entity Kesomo, visible to the state (Jorgensen 2007:58).

While the decision by Kesomo people to present as two major clans rather than one, at the clan vetting meeting convened by DPE, may have been strategically wise in a context where other local groups were also ramifying (Minnegal et al. 2015), the decision to elevate only one of four previously equivalent named subgroups, Tebesutie, to major clan status will undoubtedly have been influenced by the manoeuvrings of particular individuals. But that strategizing itself was framed by opportunities generated by the PNG LNG project; as a major clan, Tebesutie would command a significantly greater share of future benefit payments than the descendants of Kesomo's other three sons, who remain subsumed within the clan Kesomo.

Similarly, the decision to include the names Ia hafi Kesomo and Dio hafi Kesomo among the names provided to DPE at that meeting had the potential, by increasing the number of recognized subgroups, to increase the share of future benefits that Kesomo people might receive from PNG LNG. But again, these names also altered relationships at a local level. In effect, Kesomo people took advantage of formal government procedures by registering the names, and thus re-making--as Kesomo--groups that, at one time, had been associated with the areas denoted by those names. In this way, they were establishing the terms of reference for assertions of Kesomo rights to land should future exploration reveal resources on those (unmapped) areas. They were exploiting procedures put in place by the state to pre-empt possible challenges from people in different clans who might invoke different histories of association with the land in question (based on long-term use in the case of Ia hafi land and on kinship links through Bogaia ancestry in the case of Dio hafi land).

When, 2 years after the 2013 clan vetting meeting, social mapping was conducted on what was unambiguously Kesomo land the local imperatives changed. People now presented a modified version of what had been recorded at the DPE meeting. Tebesutie was again presented as merely equivalent to the groups descended from other Kesomo sons. And for the first time, in a formal record, equal status (as subclans) was accorded to five named groups within Kesomo. Thus again, as at the DPE meeting, people put in place what amounted to a statement of claim over land that Kesomo people had not themselves used for a generation or more.

In their recent engagement with petroleum companies and the state Kesomo people are, in the first instance, using names to imply the existence of concrete, long-established groups and to simultaneously imply long-standing associations between those groups and fixed areas of land. They are using those names to 'mark' identity in the particular context of resource extraction and the financial benefits that could follow. At the same time, however, they are exploiting opportunities provided by the state to assert their rights, as an unclear amalgam of groups with uncertain membership, over areas of land that are themselves illdefined. They are using those names to 'make' identity in the particular context of foreseeable disputation. Their intended audiences differ. They 'mark' identity with the state in mind. They 'make' identity with neighbours who are not Kesomo in mind. They take advantage of the ambiguity that is entailed in the denotative and connotative possibilities afforded by names (Martin 2009; Wagner 1974).

The Osumitie case summarized above has points of similarity with the Kesomo case and significant points of difference. Martin's immediate concern, in the ways in which he represented groups as names to officers of the state, was with securing the financial well-being of his immediate kin and of others to whom he had responsibilities. Only those who were listed by DPE would, at some time in the future, receive royalties. With an eye to the state he was 'marking' identity with the assertion that a set of names, each of which in fact conjured an imaginary group into being, denoted a pre-existing and manageable entity. With an eye to local people with potentially competing aspirations, he was 'making' identity with the implied assertion that, as the guardian of the myth of the origin of fire, he alone held the knowledge and the right to declare proper associations between sets of people and the land at issue.

There are, however, three ways in which the Osumitie case differs strongly from the Kesomo case. Firstly, the land of greater Osumitie is geographically more confined and, at least in Martin's understanding, less ambiguously bounded than is Kesomo land. Secondly, it is only the land of lesser Osumitie that is, in part, included within PRL2 and, as the only living adult male with close ties to the place where fire originated, Martin considered that he could act as sole spokesperson for the land associated with greater Osumitie. In the Kesomo case, there were multiple stakeholders from multiple lineages and public representations entailed greater collaborative input. Thirdly, Martin rationalized his public statements concerning connections between people and land by specific reference to the myth of the origin of fire, at a site located on the land of one particular oobi, his own. Kesomo people made no analogous public statements; Gamlihai was acknowledged as the origin place for all Kesemo but, though located on land that is now primarily associated with Tebesutie, there was no assertion that Tebesutie thus had some precedence over other Kesomo. Martin's personal sense of identity, in contrast, was based in an understanding that the land of his natal oobi and, hence, the original people of that land, were primordial with respect to greater Osumitie. Within this frame Martin considered that the public position he had adopted was an unfortunate necessity. He planned that eventually it would be revised such that the name Bogua was restored to its rightful status.

On a national scale, Kubo and Febi people are minor stakeholders in the PNG LNG project. At most, there are 1500 people living relatively close to a site where five marginal wells have been drilled but, at best, are not scheduled for production until 2020. Their concerns, and their strategies for dealing with those concerns, however, are like those of other stakeholders. As Golub (2007:46) wrote of Papua New Guinea as a whole, 'understanding how... customary relation to land is structured requires an approach not just to local institutions, but to local cultural logics which are themselves irrevocably dynamic.' (26) Kubo and Febi people seek recognition by both the state and relevant petroleum companies as legitimate and deserving beneficiaries, they seek to maintain good relations with chosen neighbours, and they seek to pre-empt challenges to their position by potential competitors. In attempting to achieve these ends they present versions of the connections between names, groups and land that are intended to convey different messages to different audiences. These messages may vary through time as different concerns come to the fore. The Kesomo and Osumitie cases show quite clearly that, for these people at least, there can never be answers that will be valid across time and space to questions about the association of particular people with particular areas of land.

ACKNOWLEDGEMENTS

We thank the many Kubo and Febi people who have cared for us and taught us in the years since 1986. Special thanks to Henick Taprin for permission to reproduce the map from The Origin of Kesomo. Thanks to Willie Samobia and Ana'is Gerard for assistance and advice, the University of Melbourne for granting periods of leave, and the Australian Research Council for award of a Discovery Grant (DP120102162).

NOTES

(1.) The Daribi people of Karimui would not now be classed as highlanders by those who adhere to the value of this regional categorization of New Guinea groups. When Wagner wrote, there was much ambiguity about both the altitudinal extent of the 'New Guinea highlands' and the inclusion of particular language groups within a 'highlands' category that was socially coherent (Hays 1993). Wagner drew upon his knowledge of Daribi to make a point that he judged to be applicable to much of New Guinea.

(2.) In the years 1986 to 2014, we spent 30 months living with people on Kubo land and visited communities on the lands of Konai. Febi, Samo. and Bedamuni people. Our research has been independent of social mapping exercises or other contractual engagement with the PNG LNG project.

(3.) By 2007, surviving descendants of the Etoro-like language group that Barclay (1971, 1972) named as 'Siali' had re-aligned within Febi and Huli groups, and were perhaps best known as members of a group named Mora (Denham el al. 2009:4.23).

(4.) PDL9 is made up of six graticular blocks. In PNG, petroleum retention licences (PRL) and petroleum development licences (PDL) are granted over areas of land that comprise a set of graticular blocks. These blocks are predetermined as areas delimited by 5 minutes of latitude and 5 minutes of longitude and, as shown on the map, each has a unique numeric identifier.

(5.) The names of purported languages, groups of people, individual people, places, and landscape features are spelled in many different ways in both early and recent documents produced by anthropological, linguistic. mission, government, resource company, local, and other sources. In what follows we have standardized spelling in cases where ambiguity is unlikely but have biased spelling in favour of Kubo versions.

(6.) The primary referent of the name Agala is uncertain. Shaw treated it as an alternative language name for Febi (Shaw 1996:73; Evensen and Shaw 2015). Ernst (2008:10) suggested that it 'is probably a term used by people around Febi to mean "others'". Anais Gerard (personal communication 2012) was told by eastern Febi-speakers that it referred to the people who lived in the wedge of country, south of the Blucher Range, between the Murray and Strickland Rivers. In fact, from south to north people living in the wedge are speakers of Febi, Konai, and Bogaia languages. Arsjo (2016:13) considered it to be one of two terms used by Duna and Huli-speakers to refer to Febi and Konai, but commented that Reggie Howard, a Christian Brethren Church missionary who lived at Tobi in the years 1984-1990, used Agala with reference only to people living in the Murray-Strickland wedge. In 1995, at Omeibi (Ogwatibi), two in-married women were identified to us as Agala. These identifications were in response to the question 'e oobi e hun ko' [what is the name of her 'clan']. At Omeibi and Tobi we were told that the primary residential location of Agala people was the mixed Konai-Bogaia community of Tinahai, close to the Strickland River, about 20 km north of Omeibi.

(7.) Until at least the late 1990s marriages among Kubo were, ideally, exogamous and based in immediate exchange of sisters. Primary extra-familial bonds were those established between brothers-in-law. Spouses acquired rights of access to each other's land with the outcomes that no one was 'disadvantaged by the marriage with respect to either the wherewithal of subsistence or their standing within an essentially egalitarian community, a community in which the identity and performance of each individual in his or her own right is more highly valued than the identity and coherence of superordinate groups such as clans' (Minnegal and Dwyer 2006:125; see Note 6, pp. 131-32, for examples of ways in which Kubo may accommodate to the frequent complication that demographic exigencies disrupt the ideal of sibling exchange). Ernst (2008:62-4, 67) writes similarly of Febi, and comments that 'the system of marriage produces dense but small local kin networks for support, with close and trustworthy... relations between brothers in law and first cross cousins'.

(8.) Some present-day asserted connections of this sort are based on discovering similar mythological stories in contexts of travel across greater physical distances and encounters with people who speak quite unrelated languages. It is assumed that analogous stories--of ancestors having had a role, for example, in the discovery of fire--establish an historical connection the details of which are no longer remembered.

(9.) There is an historical dimension to the way in which different groups that share a name regard each other, with respect to siblingship and marriage prospects. In 1987, a young man and a young woman from differently named subgroups of Headubi wished to marry. Senior Headubi men argued that the marriage was legitimate because the two groups of Headubi were associated with non-contiguous areas of land and had been separated for a long, though indefinite, time. The facts that the two groups had regularly associated socially, that men had spoken of each other as brother, and that the man and woman had co-resided for a long period were not considered to jeopardize their marriage. In contrast to Ernst (2008:61), however, we do not think that marriages between subclans were a 'regular' occurrence.

(10.) Kubo speakers use the word hafi in reference to the junction of two watercourses and use the words hobe and ho in reference to the headwaters of a stream or river, Febi speakers appear to use the word aufi in place of hafi and, hence, Abai hafi may be rendered as Abai aufi or as Abai aufitie where the suffix tie connotes 'sleeping together' or 'sleeping in one place' (Ernst 2008: Appendix 4). Arsjo (2016:68) notes that for the closely related, and neighbouring, Konai language 'the verb tie with the meaning of "live" is an existential state verb. When it is conjugated as an experiential state verb it means "fall asleep/sleep'".

(11.) In those years people did not use the word 'clan' or its Tok Pisin equivalent, 'klen'.

(12.) Among Kubo and Febi, men who have married sisters maintain a long-term, affectionate joking relationship and address each other with the reciprocal term kiuwi. Though the men may be affiliated with different oobi, their children call each other 'brother' or 'sister'.

(13.) It is only in recent years, probably since 2000, that people have adopted both the terminology of 'clan' and 'subclan' and the sense that these terms connote a hierarchical relationship. In earlier years assemblages of people might be referred to as oobi but without a connotation that named oobi were necessarily equivalent. That is, on one occasion the overarching group name Headubi (or Kesomo) might be spoken of as an oobi; on another occasion named subgroups within Headubi (or Kesomo) might be spoken of as oobi. Ernst recorded the hierarchical status of named groups as these were reported to him. He was thoroughly aware that the potential benefits from large-scale resource extraction projects could influence the ways in which people represented group structure (Ernst 1999).

(14.) Many of the people who established Tobi 2 had, in the years preceding a 1997 drought, lived at Tagohai, north of Dogomo River and a few kilometres east of the Strickland River (Minnegal and Dwyer 2000b). The village that is officially named as Tobi 2 is also recognized locally as Tagohai 2 and Molo Village.

(15.) The lists of purported descendants of Kesomo's four sons include the names of people from multiple clans and three or four language groups, who are thus designated as potential beneficiaries of the PNG LNG project. They bear little resemblance to either actual or generally accepted genealogical connections as reported to us in the course of 28 years research.

(16.) 'Clan vetting' was a process devised by officers of PNG government departments after the Kokopo meeting in an attempt to resolve perceived difficulties with Social Mapping and Landowner Identification studies that were required under the Oil & Gas Act 1998 and were conducted as consultancies under contract to relevant petroleum companies (Koim and Howes 2016). Those studies did not name the individuals who might eventually be judged to be legitimate beneficiaries of royalties and other benefits, and this proved frustrating to both bureaucrats and local people. But clan vetting, too, did not identify individual beneficiaries. This, people understood, would occur with the registration of Incorporated Land Groups, each accompanied by a list of members. In 2005, Febi people lodged 11 applications for registration as ILGs (Ernst 2008: Appendix 3); local understandings were that financial benefits from the production of gas on their land would be paid only to registered ILGs. It is not known whether these original applications were successful, but they were no longer salient in 2011. In early 2014, Febi and Kubo people at all communities were engrossed in compiling lists of people and maps of land to accompany planned ILG applications (Minnegal et al. 2015). They were doing so in response to the clan vetting process of November 2013 and an understanding, never fulfilled, that government employees would assist them with ILG registration. In fact, as Koim and Howes (2016) make clear some people consider the clan vetting process to have been a failure, the judiciary has started to challenge that process, and since 2015 a new process termed Alternative Dispute Resolution has been implemented though with little result to date. The focus on drawing boundaries around people and land, in the identification of beneficiary groups and the process of registering ILGs, has undoubtedly informed changes in the ways that group names are used. But the specific changes in usage we describe cannot be reduced to an artefact of government-devised processes of mapping and registration.

(17.) The man associated with Abai hah Kesomo was one of the few survivors of a 1993 landslip that buried about 16 people at Koge, a small village in the lower drainage of Na River towards its junction with Dio. Most of these people were said to be members of the Bogaia clan Augose. Weka's father was an Augose man, his mother was of a Febi clan. In the years that followed the tragedy, Weka's affiliations changed such that by 2009 he had adopted the name of an Abai hafi man as father's name, and hence was eligible to represent that group at Moro, but by 2013 had readopted the name of his own father and aligned with a section of Hcadubi that had been upgraded to the status of 'major clan'. In effect, therefore, Weka had at different times in his life identified with, and been identified as, a member of three different 'clans' in two different language groups.

(18.) DPE officers required that people list the names of major clans and of subclans within each of these. People knew that the named clans of their more populous eastern Huli neighbours were very finely divided. They assumed that they were required to provide analogous detail and, to oblige, both upgraded the status of pre-existing subclans to clans and suggested names for subclans that had no previous standing as such (Minnegal et al. 2015).

(19.) The Repsol drill sites failed to yield and were abandoned late in 2016 (OIL SEARCH LTD 2017). Before Repsol departed, however, a 23-ft fibreglass dinghy, an outboard motor and life vests were delivered by helicopter as gifts to the community at Tobi 2 (Anon 2016).

(20.) In The Origin of Kesomo, this man was listed within Tebesutie Kesomo. He died in May 2017.

(21.) An origin story recounted by Febi people living at Tobi 1 refers to Gesemo, carved from bamboo by the first man, being sent to live 'near the big river' (i.e., Strickland) and told that 'this is your place' (Howard 1991:8-9).

(22.) We arrived at Suabi 2 weeks after the Clan Vetting meeting ended. Several different people, on a number of occasions, told us of events and negotiations that had occurred there.

(23.) An understanding, by others, that Bogua oversaw adjoining lands that were without people was evidenced in 1995 when a teen-aged girl told us that her land and Martin's land flowed one into the other, without a clear border. The girl was of the group named lobidi whose land is contiguous with that of lesser Osumitie but not with that of Bogua (Fig. 4).

(24.) After Osumitie people had obtained fire an old woman took its source and hid it on another mountain. Several different kinds of animals attempted to steal the fire but she chased them away. Eventually a black cockatoo was successful and distributed the fire to Dobiti people. The cockatoo was burned and now carries red marks on either side of its face.

(25.) By 2014, we had recorded the names of 35 Kubo oobi (clans) associated with land north of the Baiya River. Five were without living members; another four were without living males. In two cases, the senior male in the only extant lineage had been, as a child, adopted into a different oobi and now he and his children usually identified as members of the adopting oobi. At least 10 of the remaining 24 comprised a single lineage.

(26.) See Filer (2007), Jorgensen (2007), Weiner (2013), Zimmer-Tamakoshi (1997) for other contributions in this area.

REFERENCES

ANON. 2016. Boat flown into rural community. Post Courier April 15, 2016.

ARSJO, B. 2016. Konai Reference Grammar. Available at http://www.sil.org/pacific/png/abstract.asp?id=928474565677. Accessed 17 October 2017.

BARCLAY, R.I. 1971. Nomad Patrol Report 16 of 1970/71: Western District. Territory of Papua and New Guinea.

--. 1972. Nomad Patrol Report 3 of 1972/73: Western District. Territory of Papua and New Guinea.

CAWTHORN, W.A. 1970. Nomad patrol report 13/69-70: Western District. Territory of Papua and New Guinea.

DENHAM, T., A. BEDINGF1ELD, and U. GILAD. 2009. 4.2.1 Juha to Hides. In L. Goldman (ed), Papua New Guinea Liquefied Natural Gas Project. Social Impact Assessment 2008. Report to ExxonMobil Corporation, Papua New Guinea Liquefied Natural Gas Project. Abbotsford, Victoria: Coffey Natural Systems Pty Ltd., pp. 4.15-4.104.

DWYER, P.D. and M. MINNEGAL. 1992. Ecology and community dynamics of Kubo people, in the tropical lowlands of Papua New Guinea. Human Ecology 20: 21-55.

--. 2014. Where all the rivers flow west: Maps, abstraction and change in the Papua New Guinea lowlands. The Australian Journal of Anthropology 25: 37-53.

ERNST, T. 1999. Land, stories and resources: Discourse and entification in Onabasulu modernity. American Anthropologist 101: 88-97.

--. 2008. Full-Scale Social Mapping and Landowner Identification Study of PRL02. Report to ExxonMobil Corporation, Papua New Guinea Liquefied Natural Gas Project.

EVENSEN, S. and D. SHAW. 2015. Kamula Wordlist 1979. [Data set]. Summer Institute of Linguistics. Available at http://www.sil.org/pacific/png/abstract.asp?id=928474563952. Accessed 29 December 2015.

FILER, C. 2007. Local custom and the art of land group boundary maintenance in Papua New Guinea. In J. Weiner and K. Glaskin (eds), Customary Land Tenure and Registration in Australia and Papua New Guinea: Anthropological Perspectives. Canberra, Australia: ANU Press, pp. 135-173.

GOLDMAN, L. (ed.) 2009. Papua New Guinea Liquefied Natural Gas Project: Social Impact Assessment 2008. Report to ExxonMobil Corporation.

GOLUB, A. 2007. Ironies of organization: Landowners, land registration, and Papua New Guinea's mining and petroleum industry. Human Organization 66: 38-48.

HAYS, T.E. 1993. 'The New Guinea Highlands': Region, culture area, or fuzzy set? Current Anthropology 34: 141-164.

HOWARD, R. 1991. Lukluk long olpela Testamen. Wewak, Papua New Guinea: Christian Books Melanesia Inc.

JOHNSON, A.E. 1968. Nomad Patrol Report 2/68-69. Territory of Papua and New Guinea.

JORGENSEN, D. 1997. Who or what is a landowner? Mythology and marking the ground in a Papua New Guinea mining project. Anthropological Forum 7: 599-627.

--. 2007. Clan-finding, clan-making and the politics of identity in a Papua New Guinea mining project. In J. F. Weiner and K. Glaskin (eds), Customary Land Tenure and Registration in Australia and Papua New Guinea: Anthropological Perspectives. Canberra, Australia: The Australian National University E Press, pp. 57-72.

KOIM, S. and S. HOWES 2016. PNG LNG Landowner Royalties--Why so Long? Devpolicy blogpost, 16 December. Available at devpolicy.org/png-lng-landowner-royalties-long-20161216/. Accessed 10 January 2016.

LEAVER, A. and K. MARTIN. 2016. Creating and dissolving social groups from New Guinea to New York: On the overheating of bounded corporate entities in contemporary global capitalism. History and Anthropology 27: 585-601.

MARTIN, K. 2009. Names as markers and makers of contested identity: On social groups in the New Guinea Islands. Oceania 79: 162-176.

McELHANON, K.A. 1971. Classifying New Guinea languages. Anthropos 66(1/2): 120-144.

MINNEGAL, M. and P.D. DWYER. 1999. Re-reading relationships: Changing constructions of identity among Kubo of Papua New Guinea. Ethnology 38: 59-80.

--. 2000a. A sense of community: Sedentary nomads of the interior lowlands of Papua New Guinea. People and Culture in Oceania 16: 43-65.

--. 2000b. Responses to a drought in the tropical lowlands of Papua New Guinea: A comparison of Bedamuni and Kubo-Konai. Human Ecology 28: 493-526.

--. 2006. Fertility and social reproduction in the Strickland-Bosavi region. In S.J. Ulijaszek (ed). Population, Reproduction and Fertility in Melanesia. Oxford, England: Berghahn Books, pp. 110-135.

--. 2017. Navigating the Future: An Ethnography of Change in Papua New Guinea. Canberra, Australia: ANU Press.

MINNEGAL, M., S. LEFORT, and P.D. DWYER. 2015. Reshaping the social: A comparison of Fasu and Kubo-Febi approaches to incorporating land groups. The Asia Pacific Journal of Anthropology 16: 496-513.

MORA, S. 2015. Strickland 1--Preliminary social mapping report--PPL 269. Report to Repsol.

OIL SEARCH LTD. (2017) Exploration and Appraisal Drilling Update--December 2016. www.oilsearch.com (see asx releases). Accessed 5 January 2017.

PATERSON, W.R. 1969a. Nomad Patrol Report 14/68-69. Territory of Papua and New Guinea.

--. 1969b. Nomad Patrol Report 19/68-69. Territory of Papua and New Guinea.

SHAW, R.D. 1986. The Bosavi language family. Pacific Linguistics, Series A 70: 45-76.

--. 1996. From Longhouse to Village: Samo Social Change. New York: Harcourt Brace College Publishers.

STEWART, P.J. and A.J. STRATHERN. 2010. The appearing and disappearing world of the Bogaiya: A corner of Papua New Guinea cultural history. In K.A. McElhanon and G.P. Reesink (eds), A Mosaic of Languages and Cultures: Studies Celebrating the Career of Karl J. Franklin. Dallas, TX: SIL International, pp. 399-407.

SUDA, K. 1990. Leveling mechanisms in a recently relocated Kubor village, Papua New Guinea: A socio-behavioral analysis of sago-making. Man and Culture in Oceania 6: 99-112.

TAPRIN, H. 2007-08. The Origin of Kesomo. Unpublished manuscript.

WAGNER, R. 1974. Are there social groups in the New Guinea highlands? In M.J. Leaf and B.G. Campbell (eds),

Frontiers of Anthropology: An Introduction to Anthropological Thinking. New York: Van Nostrand, pp. 95-122.

WEINER, J.F. 1988. Mountain Papuans: Historical and Comparative Perspectives from New Guinea Fringe Highlands Societies. Ann Arbor, MI: University of Michigan Press.

--. 2013. The incorporated what group: Ethnographic, economic and ideological perspectives on customary land ownership in contemporary Papua New Guinea. Anthropological Forum 23: 94-106.

WESCH, M. 2008. Creating 'kantri' in central New Guinea: Relational ontology and the categorical logic of statecraft. M/C Journal 11(5) Available at http://journal.media-cuIture.org.au/index.php/mcjournal/article/view/67. Accessed 6 December 2015.

ZIMMER TAMAKOSHI, L. 1997. When land has a price: Ancestral gerrymandering and the resolution of land conflicts at Kurumbuka. Anthropological Forum 7: 649-666.

Peter D. Dwyer

The University of Melbourne

Monica Minnegal

The University of Melbourne

DOI:10.1002/ocea.5183