Halana: The Ways of the Land-Tree

The tree is already the image of the world, or the root the image of the world tree. (Deleuze and Guattari 1987: 5)

The primary function of the land-tree is to provide a map of relations between landholdings based on the history and colonisation of that land by ancestors. This is made more obvious in Figure 2, where the ‘tree’ is clearly the prime object of attention, and the more general cosmological elements are simply represented alongside the image of a globe, identified as such by the inclusion of lines of longitude and latitude. [13] As Jif Ruben explained it to me, the words that appear alongside the various branches and sub-branches of all of these land-tree images are the names of ancestral bwatun vanua, or the ‘foundation places’ of descent groups (see above). The collection of names is representative rather than exhaustive, and this mainly explains the significant differences between the two examples shown here. Some of the bwatun vanua appearing on these images are the names of contemporary hamlets, such as Asaola, Arevo, Atabulu and Amangao in Figure 1. Other names are said to represent the sites of hamlets long since abandoned.

Figure 2: Land-tree two

Figure 2: Land-tree two

The land-tree very clearly demonstrates the centrality of itinerant principles, and especially of series and divergence, in Sia Raga interpretations of social topography. The analysis of these can readily be linked to recent appraisals in Pacific anthropology of Deleuze and Guattari’s distinction between rhizomic and arborescent models. [14] In apparent contradiction to Deleuze and Guattari’s suggestion that Oceania, ‘in particular’, offers ‘something like a rhizomic model opposed in every respect to the Western model of the tree’ (1987: 18), their shape is clearly tree-like. Thus in Deleuze and Guattari’s terms, and like the figure Bwatmahanga (discussed above), they may be seen to incorporate a certain ‘tree logic’ of ‘tracing and reproduction’; something akin to an endless development of ‘the law of the One that becomes two, then the two that becomes four’ (Deleuze and Guattari 1987: 5).

Alan Rumsey (2001) has tested the application of Deleuze and Guattari’s models of rhizome and tree to a variety of indigenous forms of ‘topographic inscription’, such as they are described in anthropological accounts from Australia and Melanesia. His discussion juxtaposes what he sees as the essentially ‘rhizomic’ nature of Aboriginal dreaming cartographies with several more ambiguously ‘arborescent’ examples from Melanesia. (The Moorhead River and Iatmul regions of PNG, and the island of Tanna in Southern Vanuatu, provide his main examples from Melanesia.) In the terms of this discussion, Rumsey notes the appearance of striking similarities across Melanesian cultures

in the way the earth is interpreted as a surface of inscription, in the dense interconnectedness of named places along tracks established by the movements of protean creator figures, in the localisation of knowledge about these movements and the role that that knowledge plays in the reproduction and contestation of present-day people’s rights to land. (Rumsey 2001: 32)

Rumsey’s description of arboreal metaphor in the Moorhead River region of PNG, which draws mainly on the ethnographic work of Ayres (1983), is in many ways cogent to the form and movement of the images discussed here. For the Moorhead people and the Sia Raga, the tree seems to be the main figure by which people understand the spatiality of social relationships and the mapping of these onto the landscape. Indeed, this has also been noted as a common theme across the Austronesian language area (Fox and Sather 1996). Furthermore, in both cases the image of the tree as a fully formed object is not as important as an attendance to the nature and trajectories of its growth (Rumsey 2001:28). This final section is concerned with discussing the tree as just such an image, as a network of hala (ways, roads) by which the Sia Raga form an understanding of relationships to place, and also to each other, across space and time.

As is the case in Rumsey’s (2001) discussion of Moorhead interpretations, the Sia Raga identify two places of original separation and dispersal for the formation of the cosmos and the island itself, and the peopling of the land. The first of these is the home of Mugarimanga, whose name, somewhat like that of Bwatmahanga (above), suggests the opening of ‘sides’, here being the two valves of a cockle shell—mu (mother), gari (cockle shell), manga (to open). Mugarimahanga’s first-and second-born sons accidentally opened a hole in the ground from which issued a great flood of salt water, thus creating the sea that now divides Pentecost and Maewo, two islands that were previously joined. [15] Fleeing from the flood, Mugarimahanga’s 12 children dispersed and embodied the natural environment, becoming the rain, wind, waves, tides and stones. [16] Some time after the embodiment of ureure (the cosmos) by these primordial atatun vanua, two of Mugarimanga’s children met one day in the middle of the island of Pentecost at a place in the bush, somewhere above where the village of Bwatnapne is located today. [17] These were a woman called Mumate, who had become the ‘quiet’ western sea, and a man called Mauri, who had become the rough sea of the east. Without realising that they were in fact brother and sister, Mumate and Mauri met once again at Atabulu where they married and had children. [18] The first of these children, and therefore the ‘first man’ of Pentecost itself, was Bwatmahanga (the ‘foundation diverges’), primogenitor of the Sia Raga today.

The second location is the home and place of death of Bwatmahanga. This is the starting point of human dispersal, but also of return, and it is with this location that Jif Ruben’s land-tree image is primarily concerned. Anserehubwe is located at the extreme northern tip of Pentecost Island. Following Sia Raga directional orientation, in which to travel southwards is to go ‘up’ (hae), this point represents the very base of the island. It is thus located in the district called Ahivo (‘at down’). The Sia Raga identify Anserehubwe as the pre-eminent bwatun vanua (foundation place), the unique place of origin and departure for the human settlement of Pentecost. It is also a place of merging (hubwe), for it is through this place that people at death ‘return’ to abanoi (the Other world) and become fully merged as atatun vanua (people of the place).

From Anserehubwe, the dispersal of the Sia Raga population across the land is reckoned to have followed specific patterned trajectories. As Jif Ruben’s representations elegantly show, the image of the tree provides a binding metaphor by which this dispersal is traced, and eventually becomes a kind of cartography. Thus space and time are condensed within the same image. Attention to the manner in which the dispersal of people across the landscape is considered to have ‘taken place’—the growth of the tree—is made clear by the presence of circles and arrows in the second of the figures presented. As Jif Ruben explained it to me, circles represent bwatuna (foundations, roots, places of settlement) and arrows represent hala (ways, roads, trajectories).

As Jif Ruben further explained, the land-tree demonstrates that dispersal took place in a series of departures divided into two successive phases. Each phase is characterised by a distinct pattern of movement, the first being a progression of what Rumsey describes as ‘lineal segmentation from a unity’ (see Figure 3 [left], adapted from Rumsey 2001: 29), the second a mode of bifurcation from the points already established along that lineal segmentation (Figure 3 [right]). While obviously recognisable as a kind of ‘branching’, the latter kind of movement is clearly not identical to another mode described by Rumsey; rather than appearing as a somewhat rhizomic ‘scattering or fragmentation from a starting point or centre’ (Rumsey 2001: 29), what occurs here is serial bifurcation at specific nodal points (bwatuna) that are distributed evenly along a central ‘trunk’. This pattern may be further imagined as recurring on subsequent branchings, and may be realised ultimately as an infinitely reticulating fractal (images not presented here, see Taylor 2003). The image of the tree is the image of the island, and it is also the image of the leaf.

Figure 3: Lineal segmentation from a unity (left, after Rumsey 2001: 29), and serial bifurcation at specific nodal points (right)

Figure 3: Lineal segmentation from a unity (left, after Rumsey 2001: 29), and serial bifurcation at specific nodal points (right)

In explanation of these different itinerant modes, Jif Ruben told me that when the original departure from Anserehubwe occurred, the rules of marriage—the division of society into two exogenous moieties—had not yet been devised. Being devoid of the lateral dimension of ‘sides’ (tavalui), lineal movement was the only available option. While some people stayed at Anserehubwe a group of people moved upwards—or, in Western geographical idiom, southwards—to settle on a new point on the land. At a later stage, a division from this second group departed once again, continuing along the same trajectory, up the ‘trunk’ (bwatun gai) of the island. As this pattern of lineal segmentation continued, the children of those groups left behind formed their own new bwatun vanua (foundation places) at the places where they settled.

As Jif Ruben further explained to me, the main impetus for the succession of departures was fighting over land. Social tension was fuelled by the restrictive, linear nature of their pattern of movement and settlement. Nevertheless, throughout the course of their journeying, the people had been developing their knowledge. By the time they began to reach the ‘end of the line’, they came to a point where they were able to strategise new roads (hala) for the maintenance of peace (tamata). These roads, being the related rules of marriage and land tenure, were based on the designation of two moieties. These are known today as Bule and Tabi. With the division of the population into two tavalui (sides, halves, partners), successive generations began to move outwards rather than upwards, thus forming branches (raran gai) on either side of those previously established nodal settlements. With women moving across from one tavalui to the other in marriage, yet with land being retained within the matriline, the fair and equal distribution of land was ensured.