by Dr. Chris Ballard
If hameigini 1 are conceptual units, hameigini 11 refer to physical entities, both to the parish territories, and to the coresident communities or parishes living within their boundaries. The apical ancestors of hameigini 1 as descent units are essentially ancestors-in-place; the dindi malu narratives that underpin clans as descent structures also “ground” those clans a particular locale. Parish territories are ideally the products of these original or early ancestors who are held to have marked out the boundaries by performing certain actions, such as dropping the leaves of trees or shrubs that now grow at these locations, digging ditches or river channels, or “cutting” the land into territories in association with the ancestors of neighbouring parishes. Parish boundaries are predominantly water features such as lakes, rivers or drainage ditches. Topographic features such as ridgelines are generally employed as boundaries only in the headwaters of a parish. Figure B 13 maps the approximate boundaries for parishes in the Tari area. Parishes vary considerably in size, from 5 ha to 1800 ha (Wood 1984, Vol. II: Table A4.2), apparently in relation to variation in the local environment and in specific historical circumstances. Parish names usually correspond to the names of the resident agnatic clan though, in rare instances, the wholesale acquisition of parish land as a form of war indemnity has produced a situation where the parish name does not match that of the clan now held to be tene on the land. 1
Tagira nogonaga mabu agalinaga
Outside is for pigs, mabu gardens are for men [i.e. people].
Wandia walinaga agandia agalinaga. Ani lene.
Women’s houses are for women, men’s houses are for men. It is said thus.
Duliya, 17.8.91, 91/16A:400-403
The movement of women, in particular, was circumscribed by restrictions on entering men’s garden areas or houses, all ritual sites and bachelor groves. 2 Since the 1950s and with the active encouragement of the missions, the proscriptions on male and female cohabitation have relaxed considerably and houses in which married couples live and sleep together are common (see Figure C9). Conceptually, rights to parish land are derived from tene status, or through links to tene individuals as the “fathers of the land” (‘dindi aba’ ), reflecting the genealogical link between tene and the founding ancestor. The gender ideology underwriting Huli descent also accounts for the relationship expressed between the production in conception of bone (kuni) from sperm, and the primary status of agnatic tene in relation to ancestral land (dindi kuni) (Goldman 1983:76). Possession of an agnatic patronym is matched on the ground by material evidence (walia or muguni) in the form of ancestral bones and ritual stones, the ditches dug and the groves of hoop pine (guraya) trees planted by the ancestors, and the locations of ancestral houses (gebeanda) and of the graves where their spirits (dinini) are held to be present. 3 Notably, where land is acquired as an indemnity for the deaths of ancestors in warfare, it is through reference to the bones of these dead ancestors that the new tene assert their status on the land.
If the ideal form of a claim to parish land is through agnatic descent from a parish-founding ancestor, it is also fundamentally an expression of the belief that rights to land are created through use, a practice literally enshrined in the deference paid to the “cutting” actions of early ancestors through the performance of rituals at the gebeanda sites of their former residences. The same actions originally employed by ancestors in the initial demarcation of parish territory create new rights when performed by the living; planted trees and excavated ditches serve as walia evidence of rights developed through use, irrespective of one’s descent status. One of the primary arenas for disputes over land is precisely this contest over ownership between land-owning tene and land-using non-agnates (Goldman 1983:169-170). Again, the exceptional case of the assumption of tene status by the owners of land gained as a form of mortuary compensation illuminates the broader relationship between agnatic status and land rights: here, the unbroken passage of use of the land by the conventional number of five generations of “new” agnates is usually deemed sufficient for them to assert their new status as tene. Critically, where other Highlands societies assimilate non-agnates by converting them terminologically to agnates, often over the course of two to three generations (Cook 1970), the force of appeals to the doctrine of tene amongst Huli is such that it is the name of the ground that changes, and not the identity or the patronyms of the new parish agnates or “owners”. Thus, in another means of “parish conversion”, non-agnatic cognates occupying a discrete portion of a parish territory over a period greater than five generations may attempt to assert their status as tene on that land. 4 This accounts for the fact that some clans claim tene status in several parishes. Dobani clan, for example, are tene at three locations in the Komo, Tari and Haeapugua basins. 5 The keenness with which claims are made to the status of parish agnate is suggestive of the symbolic power of corporate tene claims to land, but even agnatic claims to parish land are confirmed only through the actual use of that land (Glasse 1968:40); this, in turn, reflects upon the role played by the recognition of rights in a community in that tene status, or links to tene in a parish require both demonstration and recognition. “Recruitment” to a parish thus rests upon appeals to an amalgam of principles of descent and personal histories of residence and cooperation within a community. The two major unit distinctions of tene and yamuwini that Huli employ to describe parish residential categories relate initially to the descent categories of tene and aba. Parish tene are those agnates of the resident agnatic clan who reside or maintain a residence in the parish. Note that not all of the tene of a clan (as a unit of descent) actually reside together in the same parish. The term yamuwini, meaning literally “placed (wini) by women (yamo or yamu)” (Goldman 1983:76), refers to those nonagnatic cognates who reside or maintain a residence in the parish by virtue of their links through a female tene. Both categories are thus defined partly in terms of descent and, as descent constructs, are not subject to change over time: yamuwini can never become tene, nor do they ever employ the patronymic prefix denoting the status of local tene. 6 Neither category corresponds perfectly to a descent unit because, in each case, the coresident community is composed of only some of the descent unit members. As Goldman (1988:90) stresses, the distinction between tene and yamuwini relates solely to residence in a particular parish; beyond this quite specific context, the label yamuwini has no meaning other than the general gender category connotation of female (B3.3). Tene, in Goldman’s (1983:77) useful distinction, are not so much wife-givers as landgivers.
Effectively, the distinction between tene and yamuwini residents in a parish is the realisation on the ground of aba links between a single tene patriline and a variety of named aba patrilines related at different segmentary levels to the tene patrilineyamuwini are those aba who reside with a tene clan. To put the same point in a different way, the tene residents of a parish are only some of the total number of
agnates of the clan which is tene in that parish, and the yamuwini are only some of the members of the lineages that stand in relationship as aba to the parish tene. It is important to stress that yamuwini residents do not form a single coherent unit. Rather, the category of yamuwini embraces a multitude of lineages of different segmentary sizes that are derived from different clans and attached to the single tene clan of the parish at a wide range of different segmentary levels. Tene and yamuwini thus behave in respect to one another as lineages related as aba. Prior to the post-contact cessation of public rituals, representatives of specific tene and yamuwini segments of the parish community performed complementary roles in parish-centred rituals such as tege and homa haguene. The concretising effect of cosponsored ritual and of cooperation in production, war and the financing of marriages and compensation produces a situation in which yamuwini, particularly where they have maintained residence in a parish for over five generations, are treated as “the brothers of tene” (tene hamene). The rights created over time through use of their land within a parish renders yamuwini almost as permanent a parish feature as tene, but it has to be stressed that yamuwini can, ideally, be evicted by tene (see also Frankel1986:48, Goldman 1988:90). This is, of course,
hotly denied by individuals committed to residence as yamuwini in a parish, and rarely voiced as an opinion even by uniresidential tene; like talk of group fission among the Wahgi (O’Hanlon 1989:31-32), discussion of even the possibility of eviction of yamuwini is not a topic that is lightly broached. Nevertheless, Huli history is replete with instances where particular yamuwini lineages have been ousted, usually by force and under some pretext, by the tene and other parish residents.
In a sense, the categories of tene and yamuwini are the only unit distinctions routinely made amongst residents of a parish, because they are the only category labels that identify corporate descent-related units within the parish community. A third “covert” category of non-cognatic residents in a parish is identified by ethnographers using the terms tara (“other”) or wali haga (“women +continue to stay”) (Goldman 1983:83). Individuals identified as tara or wali haga include friends, nonconsanguineal affinal relations and refugees from war in other parishes. Although I have heard both terms used in reference to non-cognatic residents, the minor debate over preferred use of either one or the other term (Goldman 1988:158 n.2) appears misplaced in the context of parish residence insofar as none of the people so identified would employ the label to describe themselves, nor were there formerly any contexts, ritual or otherwise, in which tara behaved as a distinct unit. The term wali haga is more commonly used as a generic label for all clans other than one’s own, as sources of marriageable females. Tara are attached to individual patrilines, usually to tene but also to yamuwini, through either of whom they derive their right to residence in the parish. While tene and yamuwini are enduring categories, in that the distinctions amongst them are reproduced over time and are held to be immutable, individuals or lineages identified as tara rarely maintain that distinction over time: as a source of marriageable individuals with respect to the parish tene, those tara that persist in residing in a single parish are usually swiftly incorporated as yamuwini through marriage. 7
The principles of cognatic kinship (but not cognatic descent) thus produce communities composed largely of cognatic kindreds (dame or aria). These consanginueal clusters of tene and yamuwini are described literally as people tied (baile) to one another (Goldman 1983:72), as distinct from those with whom one has no ties (nabaile), such as individual tara. The critical individuals in these linkages between broader communities, and particularly those whose aba relationships position them between adjacent communities, are literally “men in the middle” (dombeniali) or men “tied between” (bi bai). 8 These individuals play a critical role in mediating between different groups, embodying the third position (dombeniali) between two disputants (tene) in the normative structure of Huli disputes (Goldman 1988:94). The broadest possible appellations for a coresident community, incorporating both kin and non-kin, include the term hameigini, in the sense of hameigini 11, and the general term, “we here” (ina oali) as distinct from others elsewhere (e.g. uyuali: “those up there”).
Huli principles of descent and affinity play a crucial role in the structure of relationships between people and land (and this is what was being expressed in Huli ritual), the scope for “play” between the two furnishes individuals with a ‘negotiable face’ to social structure (Goldman 1983:71). The more pragmatic factors of propinquity and coresidence appear to play a major role in the quotidian constitution of social groups- groups that actually garden, finance exchanges and conduct warfare together, but whose composition seldom corresponds to the classificatory boundaries defined by descent and affinity. Before proceeding to an analysis of the composition of these “labour” or “project” groups (B4.5), some account of the ways in which people deploy themselves across the landscape is required.
The wide range of aba links available to any individual provides considerable scope for men, in particular, to maintain gardens and even residences in a number of different parishes. This scope for multilocal male residence is further extended through the rights of women to maintain gardens and residence in parishes in which their parents have maintained rights to land; married men thus gain access to (but not ownership of) the land owned or used by their wives. The advantages of a widely dispersed suite of gardens were even more evident prior to contact than they are now, diminishing the exposure of an individual’s holdings to natural hazards and the effects of war. During the period of Glasse’s fieldwork (1955-1960), 42% of the adult men of the parishes Glasse surveyed maintained residences in more than one parish. Allen (in press) has described the results of a 1979 survey of 44 parishes in which an average of 9.7% of all residents claimed multilocal residences; multilocality in individual parishes varied within a range from 1% to 43%.
The complexity of the composition of Huli hameigini II, as parish communities, reflects the cumulative consequences of a system of access through cognatic ties to land combined with a rigid definition of descent that have historically yielded a high incidence of multiresidence. Statistical variation in the ratios of tene, yamuwini and tara in different parishes is thus a function both of the scope for multiple residence and of historical circumstance, in terms of the activation of specific aba links and the differential growth of lineages within a parish. Table B6 provides data on parish composition from five different surveys at nine parish locations, showing considerable percentage variations for the categories of tene (19.9%-50%), yamuwini (37.2%-58%) and tara (6.1%-37.2%). The change over a period of two decades in the composition of Toanda parish, where the percentage of residents who are tene has risen from 20% to 49%, is particularly interesting in the light of Glasse’s method of determining descent rules from recruitment; obviously, it suggests that descent, as a conceptual resource, cannot be predicted from the statistical facts of parish composition. But, equally, if there are no prescriptive norms that dictate the necessity of multiresidence, how are we then to account for this distinctive practice?
The questions of the historical genesis and the role in Huli society of the practice of multiresidence have been addressed by several writers (Glasse 1968, Frankel1986, Wood 1984, Allen in press). Glasse’s (1968:83-84) original observations, that multiresidence reflected the paramount need for security and the advantage to be gained from dispersing the effects on crops of war and natural hazards, have not been challenged and are supported by Huli explanations for the practice. The forceful banning of warfare by the colonial administration and the introduction of a wider range of crops and imported foodstuffs are thus presumably responsible for the dramatic decline since the 1950’s in the practice of multiresidence: adult male residents of Toanda parish who maintained residence in another parish constituted some 70.4% of the total in 1959 (Glasse 1968:30), but only 19% in 1978 (Grant 1979) and 17.6% in 1979 (Allen in press). The roles of warfare and of the desirability of prime agricultural land in the historical distribution of the Huli population are returned to in B4.5; but the relationship between Huli social structure and the practice of multiresidence is raised here as a problem that deserves further analysis.
On current evidence, the formal structure of Huli society appears as the precipitate of the cross-cutting influences of agnatic descent and cognatic ego-focused kinship, producing a system of widely dispersed agnatic clans residing in cognatic communities within individual parishes. In his concern to identify Huli society as either cognatic or agnatic, Glasse was firmly in step with the theoretical interests dominant in anthropology of the time, but it was an orientation that led him to confuse the distinction between descent and kinship, or aba and yamuwini (Keesing 1970:761, Goldman 1983:87-88). 9 The importance of distinguishing between the conceptual categories of descent and the practical categories of residence is evident in comparing the Huli case with that of the neighbouring Enga.
Meggitt’s (1965) account of the principles of social structure amongst the Enga, undertaken during the same period as Glasse’s Huli work, made a case for agnatic descent, with ‘clan-parishes’ in which an average of 90% of the male residents were ‘putative agnates’. A brief comparison, in the light of this redefinition of Huli social structure, suggests that the significant differences between Enga and Huli society are located not so much in terms of descent as in forms of residence and in the types of category distinction these entail. Frankel (1986:51) has made the point that Enga terminological conversion to agnatic status of non-agnates after two generations of coresidence produces a considerably more “generous” definition of agnate than the more conservative Huli system of category labels. He suggests that the application of terminological rules similar to those employed by Enga would see the percentage of agnates in a Huli parish (presumably Hambuali) rise from 29% to 69% and concludes that, ‘[p]aradoxically, it is the ideological significance which the Huli attach to agnation which causes them to appear cognatic in comparson with other highland societies’ (1986:51).
Yet the terminological exactitude and genealogical depth characteristic of Huli kin reckoning is itself associated with (and might, in a functionalist account, be regarded as a consequence of) the practice of multiresidence. The differences between Enga and Huli social structure thus need to be located in regionally and locally specific historic circumstances, and the historical emergence of the practice of multiresidence emerges as a key problem within Huli history. The significance of multiresidence for an archaeological history is that it undermines the possibility of a straightforward correspondence between neatly circumscribed units of land and groups of people; the recognition of change in the relationship between people and land in history becomes more complex and requires that different, and possibly qualitative means of access to that history be sought.
(Reprinted with permission of Dr. Chris Ballard. The Death of A Great Land: Ritual, History and Subsistence Revolution in Southern Highlands of Papua New Guinea. A thesis submitted for the degree of Doctor of Philosophy of The Australian National University, Canberra, January 1995. pp. 69-76)
(Photo courtesy of Wigman Photography)
- Parish territories are not generally, as Wood suggests (1984 Vol.I:89), named after the founding ancestors of the agnatic clan; a common expression denoting parish territories takes the form “X andaga” (the “real place” or “true home” of X clan). To some extent, the spatial deployment of people maps the structure of Huli society onto the landscape. Parish territories are ideally divided into sub-territories corresponding to the higher named segmentary units of the tene clan. Thus the current ownership of individual gardens in Hiwa parish in the Haeapugua basin shows distinct internal boundaries between three descent-based groups corresponding to the three clan sections in Hiwa clan (Figure C19, Appendix C6: Gen.5); precedence is also inscribed on the land in this case, as the sections descended from the two older and thus stronger and more experienced siblings, Hari and Hiraya, are said to have been allocated the border strips, with the youngest (heyogone) section, Wamia, sheltered between them.
Gender distinctions are similarly employed in Huli spatiality, with a strict segregation of areas of male and female space. Formerly, men and women slept separately in men’s (agandia or balamanda) and women’s (wandia) houses within different garden areas. Formal pureromo adages enjoined the observation of social norms in relation to space:
Wandia agandiala There are women’s houses and men’s houses.
Wandia napaliabe, agandia paliabe
Don’t sleep in the womens’ house, sleep in the men’s house.
Tagira e hongoleni mbabuha hangabe gana wowa hangabe
Clear swidden gardens outside, dig ditches and make real gardens within [theditches]. (( The distinction made by Huli between swidden (e) and “real” gardens (mabu) is described in Section B4.4. [↩] - Areas used by both men and women were subject to particular practices: crossing bridges, for example, bachelors would use mhuambua (Erichtites lerianifolia) leaves to grip supports that might have been held by women. [↩]
- Goldman (1980:216) reports a speech by a Huli land mediator in which the ideal relationship between physical markers of ownership (walia), ancestral land (dindi kuni) and tene status is clearly set out: “The truth [about the ownership of land] should be said on the Casuarina trees, it will be seen on the drains, on the houses of both men and women, from the nut trees; these things are holding me. The ancestral land (dindi kuni) belongs to one only, you can’t pull him out and leave another there.” [↩]
- This is not to suggest that the past names of that ground are forgotten; the efforts of the refugee Bogorali clan to remember and reclaim the names of their former territory at Haeapugua are described in Part C. [↩]
- Dobani are also resident in Yaluba, where Goldman refers to them as “Tobani”. Here, they appear to have a form of tene status on a discrete territory, but as yamuwini to another precedent tene clan (Duguba or Dugube) on an adjacent territory (Goldman 1983:121). Possibly, Dobani land in Yaluba is undergoing the process of territory conversion described here. The historical circumstances of Dobani’s assumption of tene status in the Haeapugua parish are described in Section C2.4 and Appendix C2. [↩]
- In rare instances and for specific purposes such as disputes, individuals of particular renown, who are usually also tene hamene, will be “claimed” as tene and accorded the local patronymic prefix. Refugees are also on occasion “hidden” from their enemies by being referred to with the local patronym. [↩]
- This process is also suggested in an adage recorded by Goldman (1983:80) which observes that the clans and fathers (ie affines) of yamuwini are themselves tara. [↩]
- Haea-Yoge, 25.6.91, 91nA:427-451 [↩]
- Glasse never publicly responded or even referred to the critique of his analysis offered by Jackson, Goldman and Frankel, though there is some evidence in his later work (Glasse 1975:349,357, 1992:247) of an element of doubt about his unitary description of what have been described here as two quite [↩]