2 Precedence and Hierarchy
2.1 Acciaioli on precedence and hierarchy
2.2 Sameness and difference
3 Kinship in Ende
3.1 Patrilineality
3.2 Affinity
4 Interactions of precedence and hierarchy
4.1 Matrilineality as disguised patrilineality
4.2 Matrilineality as disguised affinity
The main aim of this paper is to describe and analyse the functioning of matrilineality among the Ende people in the central part of Flores, eastern Indonesia. The first thing I want to emphasise, in this context, about Ende is that the Ende society is not what one would characterise as a matrilineal society. It is a patrilineal society and there is no ideological support for matrilineality.
What I want to show in this paper is how the matrilineality emerges from ideological (as it were) interactions between patrilineality and asymmetric alliance. I want to accomplish this aim by recourse to the idea of precedence (by James J. Fox) and hierarchy (by Louis Dumont).
So, before proceeding to ethnographic accounts of the Ende people, let me spend some time for elucidations for these two theoretical tools I am going to employ.
The idea of precedence (see (Fox 1989) and the following articles) as an analytic notion has proven to be remarkably fruitful in analysing various cases from the Austronesian societies in general (and the eastern Indonesian societies in particular). See for example, various articles collected in Michael Vischer (ed) (2009) . The success of the idea is due, partly, to its origin; it was conceived in close connection with the idea of ethnological study field (Fox 1980).
The idea is so successful that sometimes one tends to put all that there is in one's field into the category of ``precedence''; especially all the social ordering principles. I was about to do that in my field, when I felt very uncomfortable, because I was sure that there are substantial differences between these principles.
I suppose it is high time for the idea of precedence to jump out of the cradle of its study field and meet the world so that it becomes more focused.
As an example, I am going to pick up the concept of ``hierarchy'' devised by Louis Dumont (Dumont 1970) to contrast with that of precedence. It would have taken me a tremendously long time to capture and contrast the two concepts (precedence and hierarchy) theoretically as well as philologically, had it not been for Greg Acciaioli's excellent paper, ``Distinguishing Hierarchy and Precedence'' (Acciaioli 2009). I owe a great debt to this paper of his.
My starting point is his summary of hierarchy: that the basic assumption of hierarchy is that all the items have something in common. This ``something'' he calls ``unitary value'' (Acciaioli 2009, 57: 57). And the system is generated by ``the recursive application of a single differentiation in value'' (Acciaioli 2009, 57: 57).
I am in a complete agreement with him in this point. In the Indian cast system, the origin place of the concept of hierarchy, one caste is regarded as pure and another as impure; because the former has more purity than the latter. It is the unitary value, ``pureness'', that matters in this system. So, one could say that it is a matter of degree or quantity that counts in the system of hierarchy.
The second point to note about hierarchy is, he proceeds, that the rationale of the system is ``the successive . . . exclusion'' (Acciaioli 2009, 58: 58). Here I diverge from Acciaioli; I would like to stress that the rationale or the ``aim'' (Acciaioli 2009, 58: 58). of the system can be either ``inclusion'' or ``exclusion'', which point I will discuss fully in the main part of this article.
He then proceeds to characterise the precedence system; he says that it is ``diversity of . . . oppositions'' (or perhaps ``values'') (Acciaioli 2009, 58: 58) which marks precedence distinctively against hierarchy. I would like to say, in his stead, that it is the ``lack of values'' which comes into the comparison of items in the system. Two items are differentiated in the system of precedence, because they are different in nature (sharing nothing in common). In other words, it is matter of quality that counts in the system of precedence.
Acciaioli (along with Fox) contends that the aim of the system of precedence is ``inclusion'', and I would like to repeat the same point above: the system can function for inclusion as well as for exclusion.
What I am saying is that it is ``sameness'' (sharing the same property, more or less) that matters in hierarchy and ``difference'' (having nothing in common), in precedence. In the explanatory system of hierarchy, sameness is what is given in the system, and difference is what is to be explained by the given sameness. In that of precedence, on the other hand, difference is what is given and sameness is what is to be explained by the given difference.
The most definitive characteristics of precedence (as well as hierarchy, I would add) is, according to Fox (Fox 2009), that one term is accorded a value over the other. The two systems are, thus, systems of evaluation. Let me summarise how the evaluation process would work in each of the two systems.
In a hierarchical system, a is superior to b because a has more of a ``unitary value'' (whether it be pureness or ``closeness to the origin'' or what not) than b. In a precedence system, a is superior to b or a is different from b; and that is it. There is no explanation available behind the distinction.
I may be deviating, perhaps, from the original expositions of the two concepts; but I hope my re-rendition of them may lead to the clearer understanding of the ethnographical facts which the two intend to aim at. I will expound more detailedly my rendition of the two concepts in analysing empirical data.
The aim of this paper is, as I said at the beginning of the paper, to analyse kinship system among the Ende people, with the aid of concepts of precedence and hierarchy. The Ende people is an Austronesian speaking population, living in the central part of the island of Flores, eastern Indonesia.
The prominent principles of the kinship in Ende are: (1) patrilineality and (2) asymmetric alliance. From now on, in describing the Endenese kinship system, to avoid unnecessary confusions, I will stick to a male point of view, unless otherwise explicitly mentioned.
The principle of patrilineal descent (``patrilineality'' for short) is eminent, among others, in the sphere of inheritance. Communal land is owned by a patrilineal group, with an emphasis being put on the eldest line. Individual properties are inherited from father to son, with the same kind of emphasis on the eldest son.
By asymmetric alliance, I mean an ideology of (classificatory) mother's brother's daughter marriage. Even though people only occasionally contract such kind of marriages, the ideology serves to order an Endenese social world. The eminent result of this ordering is that affines are classified into two categories, that is, wife-giver and wife-taker, and that the relationship between wife-giver and wife-taker is supposed to remain the same over generations. This allegedly constant relationship between wife-giver and wife-taker, I will call, hereafter, ``affinity'' for short.
Ende is a patrilineal society, in the sense that patrilineally related members are regarded as of one and the same group. The ethics permeating the members of a patrilineal group is ``of one mind'' (se atE). The relation among agnates (the members of a patrilineal group) is, thus, ``sameness''; the ``rationale'' of the system is inclusion. Members of the group are the ``same people'' because they share descent. As Endenese say, members of a patrilineal group ``come down/descend (wa'u)'' from the same ancestor.
As the group is of one mind, it is an impossibility for members of the group to exchange, because it needs two discrete entities to exchange; the interchange of things between agnates is, thus, never called ``give and take'' (pati and simo), but bagi, ``to divide'' or ``to share''.
Inside the group, we can, however, discern hierarchy, generated by the very idea of ``descent'', based on seniority, that is, age and generation. The most eminent one is the opposition based on age, that is, between younger brother and elder brother (ari and ka'E). Elder brother is superior to his younger brother in various ways. Based on this opposition, a patrilineal group in Ende tends to be segmented into loosely hierarchical sub groups, with the descendants of the eldest being regarded the most privileged. The system works for exclusion in this aspect.
Even though there is a hierarchical division (functioning for exclusion), the most important thing about the patrilineal group is that members form (ideally) a harmonious whole, because they are ``of one mind'' (se atE). The relationship among agnates is almost always called ari ka'E; the two terms are not used separately (ari only or ka'E only) except when emphasising the hierarchical aspect. The first and foremost function of this principle (patrilineality as hierarchy) is, thus, inclusion.
Once outside one's own group (ari ka'E), one shares, by definition, no descent with those people belonging to other patrilineal groups; they are ``different'' people. Thus, in this explanatory system of patrilineality, the sameness is a given and the difference is what is to be explained by the given sameness: a is different from b because a belongs to a certain group (A), and b does not.
With most of non-agnates, an Endenese usually has another set of relationships: that is, wife-giver (ka'E embu) and wife-taker (weta anE). While the relationship between ari (younger brother) and ka'E (elder brother), or more precisely, between ari ka'E and ari ka'E (between agnates) is, first of all, egalitarian, that between ka'E embu (wife-giver) and weta anE (wife-taker) is not. A wife-giver is superior to a wife-taker in many ways. When a wife-giver, for example, demands something of his wife-taker, the latter is supposed to do everything to fulfil the wife-giver's demand. If not, it is said, the wife-giver will curse his wife-taker.
I used, according to an anthropological convention, the terms ``wife-giver'' and ``wife-taker'', yet the more appropriate rendition would be ``mother-giver'' and ``mother-taker''; when an Endenese is asked about who his wife-giver is, he would, in most cases, mention his mother's brother, a representative of his mother's natal group. The group (his mother's natal group) is called ka'E embu pu'u (ka'E embu of the ``trunk/origin''). The mother's brother is the trunk and the sister's son is the tip (rhombo) (see Fox [1971] on the botanic metaphors and Fox [1994] for the importance of the concept of ``origin'').
The relationship between wife-giver and wife-taker is, thus, that of precedence: they have nothing in common. They are prototypes of the ``other'' to each other in the Ende culture. The ethics permeating the relationships is ``being good to each other'' (papa pawE). The word papa (``to each other'') suggests, in contrast to se (``one'') of se atE, that there are two discrete entities. And the expression as a whole suggests that there is always a possibility of enmity emerging between the two. It sometimes happens that the enmity between wife-giver and wife-taker becomes so intense that they detach the relationship (papa ro'i). Until wife-taker visits wife-giver and begs for reconciliation (papa warhE) (with a set of proper ritual procedures), the wife-taker is considered ``like a witch (porho) or a slave (o'o)'', a being looking like a human being, but not fully human. This relationship is the arena of exchange, par excellence. The interchange of things between affines is called pati and simo, ``give and take''.
Affinity, as precedence, is, first of all, a principle of exclusion.
As Acciaioli emphasises, Fox's contention (Fox 1994) is that the aim of precedence is ``assimilating and incorporating groups'' (Acciaioli 2009, 59: 59). It is also true in Ende. The assimilating or inclusion process is most conspicuous visually in Ende at the time of bridewealth negotiation. The scene of the negotiation can be described most easily employing the Endenese terminology of ``waiting people'' (ata napa) and ``coming people'' (ata mai). On the day before the negotiation, in the bride's village, the bride's group is originally the sole waiting group, waiting for their wife-giving group to come. When they come, they come as coming people. They are received properly by their wife-taker, the bride's group. Then they are assimilated into the bride's group, and become ``waiting people'', waiting for another wife-giving group to come. At the end of the day, all the wife-giving groups (``coming people'') have been assimilated as one big ``waiting people'', waiting for the final ``coming people'', that is, the groom's group. The latter consists of the groom's group and its own wife-takers (assimilated into the groom's group).
Precedence as affinity thus works as inclusion principle, although the final opposition between the bride's group and the groom's group remain to the end of the ceremony. Yet, on the next occasion, say, when one of a female member of the groom's group is going to marry, the then bride's group will become one of the ``coming people'' to the then groom's group, and will be assimilated as a ``waiting people''.
Thus, even though precedence, as it is, is a principle of exclusion, it can work as a principle of inclusion.
So far, it seems that I have not diverged much from Fox's arguments (and Acciaioli's as well), if we confine our attention to the affinity principle (wife-giver/wife-taker relationship) as precedence in Ende. It looks as though the patrilineality principle as hierarchy is a totally separate principle, working in its own field by its own method, independent of affinity as precedence.
In the last section of this paper, I would like to show how these two principles interact with each other.
First, let us note rather a strange terminology for the wife-giver, ka'E embu; ka'E means, as was already mentioned, ``elder brother'' and embu means ``grand parent''. Here the superiority of precedence is couched in terms of superiority of hierarchy (age and generation).
As to the word ka'E (and ari for that matter), it should be noted that it not only designates ``(elder) brother'' but also ``father's brother's son'' as well as ``mother's sister's son''. The identification of father's brother's son (or father's father's brother's son's son etc) with one's own brother is well supported by the ideology of patrilineality: all of them belong to the same patrilineal group and are ``of one and the same mind''. It is not the case, though, with mother's sister's son: he is neither an agnate nor an affine. He is, theoretically as it were, a matrilineal relative, but there is no ideology of matrilineality in Ende.
If it were just a matter of kinship terminology, we would dismiss the identification as a mere historical coincidence; but the problem is that the Ende people do behave as if this category of people (one's mother sister's son and other matrilineal relatives) were agnates, real ari ka'E. Actually, on various social occasions, matrilineally related ari ka'E often come to the fore, as often as the real (patrilineally related) ari ka'E.
The underlying logic is this.
Let me, first, recapitulate the logic of patrilineality. Agnates (patrilineally related people) are regarded as one and the same because they share the descent: they descend/come down (wa'u) from the same ancestor.
In Ende, as I described, your mother's brother is the representative of the wife-giver, the most important ``other''. You are said, in this context, to ``come from'' (mai) your mother's brother, the ``origin'' (pu'u). When you die, a set of substantial payment called ``head'' (urhu) has to be given to your mother's brother. You come from your mother's brother and your head returns back to him.
Now about your mother's sister's son (a matrilineal ari ka'E): he also ``comes from'' (mai) his mother's brother, who is actually your mother's brother too, because his mother and your mother are sisters and, thus, have the same brother. So you and your mother's sister's son, the Endenese people say, ``come together'' (mai bou) from the same person.
In the patrilineal complex, you and your agnate (say, your father's brother's son) ``come down/descend from'' (wa'u) the same person (an ancestor), so you and your agnate are regarded as one and the same (ari ka'E). Inclusion (you and your ancestor) and inclusion (your father's brother's son and his ancestor) produce inclusion (you and your father's brother's son). In the matrilineal complex, you and your matrilineal relative (say, your mother's sister's son) ``come from'' (mai) the same person, so you two are regarded as one and the same (ari ka'E). Exclusion (you and your mother's brother) and exclusion (your mother's sister's son and his mother's brother) produce the inclusion.
Matrilineality emerges, one would conclude, in disguise of patrilineality, depending on the same kind of rhetoric (``descending together'' and ``coming together'') even though the underlying logic is distinctively different from each other: the patrilineal inclusion is generated by the accumulation of inclusion, while the matrilineal inclusion is generated by the accumulation of exclusion.
In other cases, it also happens that matrilineally related persons (mother's brother's son and mother's brother's son, for example) are regarded in the relationship of affinity, that is, wife-giver and wife-taker, especially when the mother's brother is missing. According to this convention, an elder sister and her (patrilineal) descendants are regarded as a kind of wife-giver to her younger sister and her descendants.
People explain this convention by saying that an elder sister is ``like'' a mother to her younger sister.
It is not rare, in the context of patrilineality, that people convert an age difference to a generational difference, in order to emphasise the hierarchical aspect of the relationship of brothers: ``An elder brother is like a father to his younger brother''. The hierarchical aspect of the relationship is, as I pointed out, a matter of degree, or quantity. In thus doing, people merely enhances the quantitative difference of seniority between elder and younger brothers.
In the context of matrilineality, the same procedure produces a qualitatively different situation. From a woman's point of view, once she marries out from her natal group, she becomes a member of her husband's group; her mother (along with her father) is, now, a member of her (and her husband's) wife-giver. In short, the relationship between mother and daughter is that between wife-giver and wife-taker; so, to say that an elder sister is like a mother to her younger sister is to say that they are like wife-giver and wife-taker to each other. In this way, your mother's elder sister's son (your matrilineal ari ka'E) is now regarded as your wife-giver.
This is the way how age intervenes the matrilineal complex and transfigure the matrilineal ari ka'E relation into affinity (that is, precedence). The conversion by age produces, in this case, a qualitative difference in the relationship concerned.
Let me summarise what I have done in this paper.
I, first, contrast the idea of precedence against that of hierarchy of Dumont, emphasising the unitary value in hierarchy against the lack of it in precedence. By so doing, I contest that, against Fox (as well as Acciaioli), precedence is a system of exclusion and that hierarchy is that of inclusion, in so far as they stand as they are.
In the second part, by analysing ethnographic data from Ende of Flores, I proceed to show that patrilineality as hierarchy (system of inclusion) can sometimes produce exclusion, and that affinity as precedence (system of exclusion) can sometimes produce inclusion.
My main aim is expounded in the last part, where I show that matrilineality, not directly supported by any of Endenese ideology, emerges sometimes as quasi-patrilineality, and sometimes as quasi-affinity, as a result of interactions between the two principles (patrilineality as hierarchy and affinity as precedence).
I just wish that my argument will somehow make a small contribution to the great legacy of Prof. Fox.
This is a rewritten version of the paper I read for the pre-symposium, held by Universitas Indonesia in July 2016. I am grateful for the comments I got on that occasion, especially comments from Greg (Acciaioli) as a moderator of my session and from Pak Jim (Prof. Fox) himself. Even though I am not sure I can answer all the criticisms and comments I received on that occasion, I am thankful to the two of them and other participants. And in my rewriting process, Greg has been a constant inspirer of the ideas and been a helpful advisor. I owe a great debt of gratitude to him. I am to blame, of course, for all the shortcomings and faults found in this paper.