Table of Contents
While anthropologists have been alerted to the regimented and thwarted research which a strict adherence to regionally developed ‘gatekeeping concepts’ (Appadurai 1986) may effect, it is neither necessarily more productive nor less hazardous to export/import such concepts across regions. As a case in point, James Fox has remarked (1989:51–53, 1994) that there are inherent difficulties in applying Dumont’s concept of hierarchy (1980) to societies of eastern Indonesia. These do not have the encompassing religious coherence that Dumont has attributed to India; for this reason, hierarchy cannot be described as a single principle nor identified with a specific opposition, such as pure and impure. In Eastern Indonesia there are a variety of contending oppositions that are of considerable importance to the definition of hierarchy and it is not one opposition but the interplay among various oppositions that gives rank to elements of a whole in the relation to the whole (Fox 1989:51).
The validity of this general assessment of eastern Indonesian societies depends perhaps as much on the theoretical bent and analytical perseverance of the investigator as it does on the nature of the ‘contending oppositions’ themselves. But be that as it may, attempts to conflate hierarchy and stratification, to link the one to the other, or to insist that one generates the other (see several articles in the Transformations of hierarchy volume of History and Anthropology [Jolly and Mosko 1994]) — despite Dumont’s expressed formulations to the contrary (for example, 1980:65–66), is perhaps counter-productive. Anthropologists have long been accustomed to the possibility that not only indigenous terms but analytical ones as well can turn out on inspection to be polythetically constituted (Needham 1975). We have been warned that failure to recognize this may cause confusion and, correspondingly, pointless theoretical dispute. Will an insistence on pushing ‘stratification’ into ‘hierarchy’ lead to sharper models, or will it muddle them?
My own position on this issue is agnostic. But if my tentative skepticism is not misplaced, then we may be well advised to simply accept Dumont’s definition of the concept and the accompanying strictures — but with a proviso: to regard his definition as nomological — not essential. That way we can at least avoid a dialogue of the deaf. Thus I shall not use this particular term in the following, except when its meaning is unambiguously Dumontian, explicitly discussed, or when I quote or paraphrase other authors.
Since my argument, such as it is, is principally informed by data derived from field research among the Ngadha ethno-linguistic group, central Flores, I begin with a rough sketch of Ngadha social organization.[1]
Those who are familiar with (especially eastern) Indonesian societies will not be astonished to learn that Ngadha evince a pervasive concern with scalar, seriated classification. Houses are placed in ranked classes, plots and tracts of land are divided into ordered categories, and persons are ascribed to social strata. Asking what these classifications are basically about, I shall elaborate the programmatic answers I offer now as I go along.
With respect to Houses[2] and lands the classification is most abstractly to do with what one could label seniority, or authority; ‘an ordering on a gradient of stepped differences from low to high’ as Adams (1974:328) puts it with regard to Sumba intellectual order — the most apt word for the concept I wish to convey is perhaps ‘precedence’, as Lewis (1988) has used it. And the primary idiom employed, implicitly or explicitly, in this discourse of gradient order or degrees of precedence is that of ‘origin’. The closer the classified entity (House, land) is to the origin (conception, inception) — of which it is in a sense merely a later version, pale copy, or weak reflection — the grander, loftier, mightier, more important, and more valued it is held to be.
Central to understanding Ngadha social organization are first of all the categories sa’o (‘House’) and woé (‘House coalition network’), into which every Ngadha person is born. Every person is, accordingly, referred to as the ‘child’ (ana) of such and such House and woé. Whereas ‘House’ is an accurate translation of the Ngadha sa’o, the (provisional) gloss ‘House coalition network’ for woé is mine — other ethnographers (Bader 1953; Arndt 1954) have used instead ‘clan’ or other terms which misleadingly evoke ‘descent’ as the sole principle for recruitment.[3] Second is a principle of social stratification. Ngadha are divided into three strata: nobles or aristocrats, commoners, and (former) slaves, between which marriage is strictly regulated. With a view to the overall context of the discussion that follows, it should be pointed out that there is no record of any central institutions of political or religious power in Ngadhaland prior to the moderately successful installation of ‘rajas’ following the establishment of Dutch military control in the early twentieth century.
As seems to be the case in a great number of eastern Indonesian societies Ngadha Houses are simultaneously dwellings, corporate estates, ancestral abodes, ritual centres and repositories of heirloom sacra. They are also frequently partners in exchanges predicated on marriage. In short, Ngadha Houses are — if any such thing exists — prototypical Houses in the Lévi-Straussian (1982) definition of that term. As Fox formulated it in 1980, ‘house’ ‘is a fundamental cultural category used in eastern Indonesia to designate a particular kind of social unit’; its characteristics including an idea of localization or origin, a strictly stipulated physical structure, elder/younger hierarchical relations — yet not necessarily gender-specific, and a notion of a ‘flow of life’ (usually women) between the houses (Fox 1980:11–12). I must leave out the ethnographic details which show that in contrast to comparable units elsewhere in eastern Indonesia the Ngadha House is not ‘a primary descent group’ (Fox 1988:xii), but want to stress that traditional, strictly stipulated architecture is practised in Ngadhaland with remarkable vigour.
A visitor to a traditional Ngadha village (nua) will notice that at least two Houses look slightly different from the others. One will have a tiny model of a house, called ana ié, at the centre of the ridge of its roof, the other a human-like figurine, called ata, at the corresponding place. These Houses are the two initial ones of any one woé. As far as I know every Ngadha woé will have minimally these two Houses. The one with the little house on it is called sa’o (or saka) pu’u, the other sa’o (or saka) lobo.[4] These designations indicate the place allotted a male elder of the House in question when a new ngadhu or ‘two-pronged sacrificial pole’ is carried into the village. This is a rare event; a ngadhu may stand for a century or more and the political act of instituting a new woé is not a frequent one. The more senior of these two Houses — and positions — is pu’u. When the ngadhu — placed horizontally atop a huge bamboo scaffold — is carried from the place where it has been carved into the village, a war-clad, prominent male affiliate of the sa’o pu’u stands on the trunk/root part of it (which enters the village first) while an affiliate of the sa’o lobo stands atop its tip end.
Having pointed out that ‘at least’ two Houses look different from the others, I must emphasize that many villages include Houses of more than one woé. Where this is the case there may be a corresponding number of sa’o pu’u and sa’o lobo pairs — and indeed also of pairs or more accurately couples of ngadhu (a sacrificial pillar, the manifestly ‘male’ representation of the woé founding ancestor) and bhaga (a miniature house, the manifestly ‘female’ representation of the woé founding ancestress) at the centre of the village plaza. Thus a visitor to a Ngadha village will normally know at a glance the number of woé its residents are affiliated to — it suffices to count the number of Houses with a little house on the ridge purlin, or the number of Houses with a figurine, or the number of ngadhu, or the number of bhaga. It must be pointed out, however, that this ‘normal’ situation does not always prevail. Firstly, for certain (chiefly economic) reasons any one House, ngadhu, and/or bhaga in a village is possibly yet to be put up; secondly, a village may include ‘cadet’ Houses of woé centred elsewhere. I shall comment further upon the ngadhu/bhaga ‘emblematic couple’ shortly, but should point out here that they, together with certain House and woé heirlooms, belong to an unnamed Ngadha category of inalienable sacra. These are physical objects possessing to a degree certain attributes commonly ascribed to sentient beings; they are frequently conceptually gendered, they can feel abandoned, and experience hunger — especially for the blood of a sacrificial animal.
Decisions affecting the woé in its entirety are ultimately the responsibility of its ‘trunk House’ (sa’o pu’u), the rightful residents (office-holders, stewards) of which are said to have ‘full rights’ (ha benu). Marginally second in prominence is the ‘tip House’ (sa’o lobo). Accordingly, woé land is divided into trunk (House) and tip (House) land. While decisions on land use can be made with relative autonomy by any named House that possesses at least one symbolic ‘House digging stick’ (su’a sa’o), no woé land can be sold without the explicit approval of the trunk House.[5]
I shall discuss aspects of ‘marriage’ and related topics below, but should make clear at this point that the woé division into trunk and tip sides is never invoked with respect to past, current, or contemplated marriages; there are no moieties in Ngadha social organization.
Less prominent than the trunk and tip Houses are the wua gha’o and, subordinated to these again, kaka. These are all named cadet Houses and are collectively referred to as sa’o mézé, lésa mézé (great Houses), sa’o ngaza (named Houses) or lanu. The term wua gha’o exploits a fundamental Ngadha idea — one that will not surprise anyone familiar with ethnographies of eastern Indonesia — that human relationships are mostly asymmetrical: wua means ‘fruit’, ‘child’; gha’o, equivalent to the Bahasa Indonesia (BI) term géndong, means ‘carry on the small of the back or the hip, supported by the waist and one arm, often with the help of a cloth sling’ (Echols and Shadily 1990:184). The relationship between a sa’o mézé and its wua gha’o is thus metaphorically expressed as one between parent and child. The meaning of kaka is less specific, but among its many possible glosses are ‘member’, ‘belong to’, ‘support’, ‘protect’ and ‘assist’ (Arndt 1961:230). Ngadha men and women often explain to the investigator that just as large upright stones need smaller ones next to them in order not to fall over, men depend on women, and trunk Houses need the support of wua gha’o and kaka. Another conventional indigenous synecdoche for the structure of the House coalition network (woé) is the ngadhu itself. As a sacrificial pillar at which woé affiliates kill water buffaloes in ritual it is the emblem for the woé in its entirety, although this emblem is ineffective without the bhaga complementing it. The ngadhu’s trunk and tip stand for the trunk and tip Houses or, to put it in an equally valid way, the labels ‘trunk House’ and ‘tip House’ themselves refer directly back to the high-profile ritual when senior males of each House ride atop either end of the ngadhu, as just mentioned. The ngadhu’s branches — and the branches of its roots — represent the two sets of wua gha’o and kaka Houses.[6]
I have already mentioned that the ngadhu/bhaga couple are replicated in yet another set of miniatures: the ata/ana ié atop the two founding Houses of the woé. It is worth noting that the ‘male’ trunk House is adorned with the minuscule House model while the ‘female’ tip House features the figurine.
I should also reiterate that every House ‘belongs’ to one of the two sides: to the tip side or the trunk side. The entire woé is as it were split down, or along, the middle. Importantly, only when the one side is completely extinct can affiliates from the other move in to become permanent dwellers of those Houses. The main point is that every House (and person and plot of land) belongs to either the trunk or the tip side, and that each House relates to any other of its side in an order of precedence.
By the same token, the size of a traditional House — its physical proportions — must also correspond to its position in this scheme of precedence. Largest of any woé House is sa’o pu’u, the second largest is sa’o lobo. And Houses of each consecutive level (wua gha’o, kaka) must be, if only barely, smaller than the one preceding it. Moreover, socio-symbolic spatial categories, elements of construction, and a multitude of evocative carvings embossed on the House, are all salient to the many meanings of the House (cf. Smedal 2000).
With demographic growth further diversification takes place, viz. the building of subsidiary, unnamed Houses — invariably outside the traditional village compound (loka nua) — referred to as baru.[7] Finally, the inhabitants of most lanu and baru build also a field house (kéka) or two where they spend a great deal of time ‘guarding the maize’ as crops mature and tempt monkeys and wild pigs.
Notably, the number of named Houses in any one village is not fixed once and for all. But the expenses associated with the requisite ceremonies for turning an unnamed baru into a named lanu/sa’o, even when people aspire to do so, render such transformation all but impossible at present. An important reason why a great measure of organizational stability obtains in Ngadha villages, therefore, is that while politico-ritual relative autonomy is always attainable its cost is great.
Further establishment of Houses may occur when — for diverse reasons — people from elsewhere come to live on and work the woé territory. These may found their own woé or, over time, be incorporated into one already existing, with their own named Houses. Hence a woé need not consist only of people that are either consanguineally or affinally related. Conversely, just as Houses may be incorporated from the outside (a practice reminiscent of adoption) they may of their own accord break loose and establish their own woé, complete with ngadhu and bhaga, although, as just noted, this is economically prohibitive.
[1] Field research 1990–91 and 1993 totalling some 14 months was financed by the Norwegian Research Council for Science and the Humanities and by the Institute for Comparative Research in Human Culture, Oslo, and was conducted under the auspices of Lembaga Ilmu Pengetahuan Indonesia and Universitas Nusa Cendana, Kupang, Timor. The Department and Museum of Anthropology, University of Oslo, provided a six-month grant which enabled me to think through some of the material reported here. The first draft of this text was presented in Leiden in April 1996. I am grateful to Michael Vischer and the International Institute for Asian Studies for inviting and hosting me, to the Department of Social Sciences, University of Tromsø for financing my trip to Leiden, and to the Department of Social Anthropology, University of Bergen, for providing the opportunity to finalize the article.
[2] The term ‘House’ (with capital ‘H’) designates a social unit.
[3] Arndt offers a long list of meanings of woé of which several have potential relevance here: first of all ‘sib’, ‘family’, ‘clan’, and Rangstufe (‘level of social stratification’ or ‘place in hierarchy’); but also ‘friend’ (gender neutral), ‘troupe’, ‘gathering’, ‘people’; ‘to embrace’, ‘to bind’, ‘to draw in’, ‘to tie together’ (1954:204; 1961:57). Realizing that ‘House coalition network’ hardly evokes the semantic range — nor of course the poetic potential — of the Ngadha term, I still search for a better English gloss.
[4] Saka ‘to ride’; pu’u ‘trunk’, ‘stem’, ‘origin’; lobo ‘top’, ‘tip’. The correct and complete designations of the two ‘apex’ Houses is as follows: sa’o pu’u saka pu’u and sa’o pu’u saka lobo. Thus there are two sa’o pu’u for each woé. But houses are not usually referred to in this way; people always speak of the first as sa’o pu’u and of the second as saka lobo.
[5] I cannot deal here with the subtle symbolic significances of the su’a sa’o.
[6] Unlike the neighbouring Keo ‘clan’ péo (see Forth, this volume), Ngadha ngadhu/ bhaga couples are exclusive to each woé.
[7] A word not, I think, cognate with the BI word baru ‘new’; in (northeastern) Manggarai, immediately west of Ngadhaland, houses are referred to as mbaru (Nooteboom 1939:231). Erb remarks that until such a house has been ritually inaugurated it is called not mbaru but ‘garden house’ or simply ‘forest’ (1987:212).