Symbolic Anthropology in the Netherlands

Front matter

SYMBOLIC ANTHROPOLOGY IN THE NETHERLANDS

VERHANDELINGEN

VAN HET KONINKLIJK INSTITUUT VOOR TAAL-, LAND- EN VOLKENKUNDE

Introduction

STRUCTURAL AND SYMBOLIC ANTHROPOLOGY IN THE NETHERLANDS TODAY

I

The collection published here was brought together to form an issue of the Yearbook of Symbolic Anthropology. The Yearbook had embarked on what proved to be too difficult a project — that of bringing to world attention a number of the peripheral traditions of structural and symbolic anthropology, namely those outside the major world schools domiciled in France and the United States. These smaller anthropological centres appeared, to me to offer solutions whose importance was underestimated in the centres where the basic theories had been formulated. Apart from the Dutch collection, a Danish, Brazilian and Soviet collection had been planned.

The fact that the Dutch collection now appears in the Netherlands, after years of delay, shows that the position of what I called the smaller centres is even more precarious than I had supposed. It was generous of the publishers of this series to permit me nonetheless to make the points I had in mind in the context of their series.

The basic virtue of the smaller anthropological centres is that they have tended to counteract dangerous fissional tendencies within structural and symbolic anthropology. It is not my purpose here to discuss these fissional tendencies in general terms, but a good example of them, clearly evident in the contributions of the present volume, is the polarizing of structural and symbolic anthropology into two supposedly opposed camps whose leading figures would be Claude Lévi-Strauss and Victor Turner.

The inconvenience of such intra-clan feuds lies in the threat by powerful outside forces to destroy the whole tradition of social anthropology and thus finally dispose in one blow of the Turnerians and the Lévi-Straussians together. They should therefore be mindful of what unites them as well as what divides them. The value of the articles brought together here, quite apart from their particular ethnographic and semiotic virtues, is that they demonstrate this unity in their practice, even if not always in their theory. By and large, phenomenology and rationalism are here kept in equilibrium, as indeed they have been in Dutch thought over the centuries.

When I first conceived this volume, I had hoped that van Baal and Pouwer would have been represented here, as their work exemplifies very clearly the kind of equilibrium I had in mind. Aware of the compelling force of both structuralism and phenomenology, they have used both as tools to bring intelligibility to the symbolic systems of their chosen culture area. The younger authors in the present volume have not always been fully conscious of the contradiction between these two tendencies, nor of the methods by which it can be transcended, but if all contributions are taken together they point at the same message which, in any case, is made explicit in the contributions by de Josselin de Jong, Locher and Moyer.

The present contribution will demonstrate the point made above with reference to the collection of articles presented here. In a further section, it will take up the crucial question. posed in de Josselin de Jong’s article, namely within what limits anthropology can be and to what extent it should be deductive. For it seems that a balanced reply to that question would provide us with a viable working relationship between structuralism and phenomenology. It would also seem that the answer given by de Josselin de Jong is very much in the main stream of the Leiden tradition. Finally, we should like to examine the relevance of that answer for anthropology in general.

II

The present volume appears to have affinities with the work of both Lévi-Strauss and Victor Turner, as well as having been influenced by various forms of phenomenologist thinking. Some influence of Marx is discernible only in the work of Schefold. The contributions here take none of these schools for granted but abound in critical questions in regard to all of them. Moyer sums up well this eclecticism when he points out that the specific interest of “de Leidse richting” lies in the articulation of the lived-in and the conceived orders. Moreover, this articulation is, as Moyer correctly argues, proposed not in a universal model but only on the level of a particular culture area or “studieveld”.

Yet a studieveld, in this sense, is by no means to be equated to a single culture, nor exactly to a culture area. It has, for instance, been used in the context of making a comparison between Indonesian and Proto-Indochinese agricultural rituals. The relationship between the conceived and the lived-in order in a sociocultural unit of such range cannot therefore be established by a phenomenological inquiry, such as that by Victor Turner of the Ndembu. On the other hand, there are still important uniformities within the studieveld, such that structural relationships need never be justified by some blanket theory about universal mental structures, but rest — at the very least — on a broad history of communications within a given region. Moyer points out that Lévi-Strauss’ own work has a similar historical foundation, even though the studieveld in this case extends from South America to the Northwest Coast. Thus the concept of studieveld is broadly consistent with the structuralist method.

As such, it is apt to occupy an uncomfortable intermediate position between major schools outside the Netherlands. An excellent example of the discomfort of this position is furnished by Locher’s description of how he was caught in the crossfire between Boas and Preuss in the 1930’s. Locher, like his thesis director J.P.B. de Josselin de Jong, was opposed to Boas’ particularism, nor could he be identified with the theories of Preuss, even though the latter appears to have come to his rescue. The ancestry of Locher’s methodology, as he explains in detail, was Dutch and French, hence focussed on the holistic analysis of specific systems of varying range. In the end, this approach won adherents also in the United States so that Irving Goldman’s recent standard work accorded Locher the international recognition he had deserved much earlier, for having pioneered the systematic study of Kwakiutl mythology.

Another case of the same search for articulations between the conceptual and lived-in orders is given by Kamma’s article, which should be read together with his well-known book on Koreri movements. While the earlier book describes the strictly religious content of these movements in historical context, the article published here articulates them with two action systems, namely the repertoire of rituals and the political, ritual and religious levels of the relationship between Biak and the Tidore sultanate.

Some might argue with Kamma’s assessment of the political relationship, as all State formations, even the most oppressive, may well have grown from such seemingly innocuous beginnings, but it remains a fact that we have few descriptions as detailed as

Kamma’s of the religious and ritual transactions of this seminal phase. The feudal lord might never be able to become the oppressor he often turns out to be, if the early phases of his ascendancy did not give rise to the religious adulation, in fact to the love of participation in the State formation, so well evoked by Kamma.

During the phase of culture when ritual closure is effective, it would appear that millennialism offers a solution to the contradiction between forced tribute payment and autonomy. Koreri offers a symbolic sultanate in opposition to the physical one. It is only when we learn about the magico-religious transactions with the Sultanate that the Biak Koreri movements become entirely clear in their articulations with the lived-in orders. Kamma’s article lies thus entirely within the tradition Moyer attempted to interpret.

The problem of resolving contradictions between the Lévi- Strauss and Turner approaches is especially evident in the articles by Schefold, Visser, van Beek and Creyghton. Though these authors do not seek to resolve this contradiction on the theoretical level, they all seek their individual ways out of the dilemma without becoming camp followers.

Creyghton comes close to suggesting that the contradictions between Lévi-Strauss and Turner are reducible to the contradiction between “sign” and “symbol”. On this complex issue, I shall say here only that in two vital passages Lévi-Strauss proclaimed himself a follower of Peirce and that his method displays several Peircian characteristics (though not, to be sure, the vocabulary). When Creyghton writes six illuminating “equations” to specify the meanings of the opposition open/closed among the Khoumirians, she presents in fact six connotations, ‘‘interpretants” or “transformations”’. This series is in principle an infinite one, in that the act of tying a girdle too tightly or of putting a safety pin into the mouth can always be fitted in as part of the series. This kind of semiosis is practised by Lévi- Strauss as well as Turner.

Creyghton does, however, put her finger on the real contention between the two theorists as she evokes, in the opening sentences of her contribution, the phenomenological starting point of her own enquiry, for it is at that point that a subjectivist element enters. She evades the traps of subjectivism, at every step of her argument, by reducing the sense-data to fundamentally cognitive categories. Questions such as spirit attack, barrenness, loose morals, loose bones and especially the deflowering of virgins are never resolved in the end by reference to the inherent qualities of ritual symbols (e.g. the potency of blood), but by reference to the problems of excessive openness or excessive closure.

While this reduction is essentially structuralist, the problems Creyghton sets herself are close to those traditionally posed by phenomenologists. She thus effects a skilful confluence of the two approaches.

Visser, on the other hand, gives the impression of being rather critical of Turner. She is concerned to show that the famous triad red/white/black is not really a triad but two binary oppositions. A structuralist would now proceed to argue that colour on its own is not analysable but only colour in particular relations. Visser however does not choose this road, for in her conclusion she calls the white/black opposition “metonymic”? while the white/red opposition is called ‘“‘more metaphorical’. Like Christopher Crocker, who draws similar distinctions, she seems to occupy middle ground between Turner and Lévi-Strauss.

Van Beek, in contrast, claims that his contribution proves “testable hypotheses” based on Lévi-Strauss’ theory of totemism. He manages to articulate ethnozoological classification of taboo animals with social categories among the Kapsiki (Northern Cameroons). The social category of the blacksmiths is said to be “defined” by the type of taboo animals they are exclusively allowed to eat.

It turns out that the “test” is based on a selection, made by the investigator, of six properties one or more of which are supposed to be shared by 61 species eaten exclusively by blacksmiths. The six properties were not elicited as such from informants; they were supposedly “‘self-evident” from “‘symbolic analysis”. Some of these properties are highly specific, others (like “special function”, “oddness”’) lack precise boundaries. Nonetheless, van Beek raises a question of the greatest interest: why the blacksmiths, in spite of their supposedly low status, are provided with a massive supplement of protein foods. Is their low status an objective reality or is it an ideological fiction to cover over profound status ambiguity? Perhaps van Beek’s basic orientation to this question is as close to Turner or Douglas as it is to structuralism.

Schefold’s article is the most complex of this set, as it not only seeks to find a bridge between Turner’s and Lévi-Strauss’ symbol analysis but in addition appears to adopt a Marxist evolutionary model in which the opposition: use values/exchange values acts as the basic variable to explain differences between the Sakuddei of Mentawei and the Khmir of Morocco. The chief problem he poses is what is the objective efficacy of symbolic activities — how symbolic values in a culture become “‘activated’’. He seeks to answer it by a close analysis of some well-chosen events in each culture examined.

His opening section restates Lévi-Strauss’, Turner’s and Geertz’s methods of symbolic analysis, but without bringing out the opposition between them. He argues that none of these methods fully answers his question. This is basically because Schefold distinguishes at least two kinds of symbolic efficacy: ‘‘direct”’ (like the uma of the Sakuddei) and “indirect” (like milk among the Khmir or national flags). Where we find the latter kind, there is a “figure established at a higher level in the order which is symbolized in various rules concerning behaviour with” the symbol. Groups where symbols are ‘“‘directly” efficacious are those where the symbol (e.g. the uma) is seen as the equal partner of man, where ideology emphasizes the need for equilibrium between partners, where economic activity Is centred on the production of use (rather than exchange) values and where class contradictions are insignificant.

Whereas Schefold uses a basically Levi-Straussian terminology when discussing the Mentawei, his discussion of the Khmir case follows more closely Turner’s well-known approach to the topic of milk. Like David M. Schneider did on one occasion, Schefold appears to allocate different social theories to appropriate geographic regions where they fit best. In effect, he is rejecting all existing theories and looking for his own.

These summaries serve to exemplify in what sense Dutch anthropology is very eclectic today and refuses to identify itself with foreign schools. Other articles, however, can be more easily classified as having a place — albeit a very personal one — in particular schools. Platenkamp, for instance, offers what is essentially a classical structural analysis, blending Lévi-Straussian and Leachian methods. De Ruyter, on the other hand, appears to adopt a firmly functionalist perspective, following Nutini and (by implication) Nadel.

As for de Ruijter’s concern to render a difficult Levi-Straussian text intelligible, may we suggest that the distinction between mechanical and statistical models was part of the apparatus of the American configurationalist school and that it was first developed by Margaret Mead in proposing the methodology of her Arapesh studies. As Lévi-Strauss presented his theory first in the United States in 1952, when Margaret Mead’s work occupied a dominant position there, he may have hesitated to make overt reference to it, as he made many changes in Mead’s method. The audience will nonetheless have recalled Mead’s arguments why the statistical method was not useful to her Arapesh work and why every case, every exception had to be accounted for if a valid “‘pattern’’ for Arapesh culture was to be derived. Neither Margaret Mead nor Lévi-Strauss were ever concerned with the distinction between real and ideal norms. If Mead’s distinction between mechanical and statistical methods be accepted, then it seems reasonable to use the former for elementary forms of marriage exchange, while mechanical models are clearly useless on their own as soon as non-reversible time enters into the analysis.

HI

It has been rightly said (in Paris) that structuralism has many mansions and no doubt the direction indicated by the work of P.E. de Josselin de Jong is one of these mansions. Moyer’s article in this volume outlines well enough for our present purpose the general characteristics of that mansion. Far from eliminating history from his investigations, de Josselin de Jong chooses settings in which the naked eye (or rather, the non-structuralist eye) perceives only an intricate web of forces of history. De Josselin de Jong then proceeds to show the interesting results that follow by operating a structuralist reduction on this mass of data. The vast resources provided by the studieveld approach minimize the risks always involved in such reduction.

His article published in the present volume is directed against the facility with which functionalist theorists construct supposedly universal “‘social laws’’, in this case laws specifying the prerequisites of matrilineal societies. His point is that such “laws” lack all serious predictive power. He demonstrates it by the case of the Minangkabau system which none of the theorists appear to know very well. But if their theories were truly predictive, they ought to be able to predict its prerequisites. De Josselin de Jong shows that they were unable to do this.

Yet, he could hardly deny anthropology has predictive power in certain circumstances. His own analysis of Minangkabau society is, after all, based on a mathematical model which he uses deductively. Moreover, as I shall show, deductivism is by no means absent from the article in which he attacks “‘deductive anthropology”. The question before us is therefore not whether we should accept deduction but rather: within what limits we should accept deduction and in what conditions it is actually indispensable?

De Josselin de Jong's own practice shows clearly enough where he thinks it indispensable. The Minangkabau have an claborate system of adat which, as they pretend, describes how their society works or should work. But a good deal of this apparatus (whether on the level of real or so-called ideal norms) serves to obscure not only their practices but also the rules to which the practices actually conform. There are no lies in adat, but there are many omissions — rules which are not in the adat but which modify the operation of the adat. Not only can the existence of such rules be deduced from repeated hints in the published literature, but they can be deduced also on logical grounds when the systemic interrelations of the adat are examined.

When I studied the Minangkabau in the field, I found that de Josselin de Jong’s deductions were not exactly popular with so-called adat experts, because they would elevate to the status of “rules” certain practices which they considered “wrong”, which “should not exist” and of which they were very ready to deny the existence. The whole of de Josselin de Jong’s work is therefore a sort of deduction. It is a set of hypotheses that there are hidden characteristics in that society that anthropologists ought to be able to find if they systematically looked for them.

Among these hypotheses I shall quote three because they figure prominently in the paper in this volume. The first one is that the status of a matrilineage is affected by the status and genealogy of male ancestors, the so-called bako. This hypothesis was derived very skilfully (1951) from scattered hints, but no field work was ever done to test it until Ms Ok Kyung Pak, very recently, gathered stratification data of sufficient density to show the ubiquity of this tendency so often denied. The second hypothesis is that the Minangkabau have a system of circulating connubium between kaum based on male hypogamy!. This likewise has the status of an hypothesis derived from a model, as the only way of testing it adequately is to find the marriage rules of every kaum in a given community (these rules are at the level of the kaum), and again, such data were not available to de Josselin de Jong. Moreover, the existence of such rules are denied even more hotly by the adat experts, because the latter are ideologically committed to an egalitarian model of Minangkabau society which they hope to fully institute. Their strategy in that laudable objective is to deny the ranking system and express surprise that there are still some villagers backward enough to marry according to that system. Here the anthropologist has a serious problem

because he does not wish to fight egalitarianism, yet scientific

objectivity appears to require us to say de Josselin de Jong’s hypothesis is quite right.

The third hypothesis is one that appears for the first time, to my knowledge, in the present volume. It is that the Minangkabau system generates some centrifugal force causing male mobility on a vast scale and ranging, today, over the whole of Indonesia and beyond. This is the so-called merantau complex, which was little studied until Mochtar Naim wrote a book about it recently. Naim does convincingly establish a strong link between the merantau complex and tensions within the matrilineal kinship system, but he does not clarify merantau as a cultural construct (i.e. its ethnosemantics), nor does he show the effective cause of the phenomenon. De Josselin de Jong, hypothesising from his earlier theory, postulates merantau is effectively caused by the excessive responsibilities generated by the bilocal residence of males. To my knowledge I have been the first to test that hypothesis in the field. This is not the place for reporting my findings, but it is relevant to say briefly what such a test involves.

First of all, on the level of individual life histories, merantau is explained by a variety of economic, but also educational, political and religious causes. Moreover, the institution has gone through a number of profound historical modifications. I had in fact to do a microstudy in Minangkabau history. With regard to kinship structure, two questions proved relevant:

1. In view of the proved fact of hypogamy, to what extent does male achtevement provide the subtle differential that makes a woman proud to accept as husband a man whose genealogical status is not (and on a global view, cannot be) superior? Is merantau a road to such achievement?

2. Given the fact that the status of a male suffers if he appropriates much of the yield of his kaum’s land, and that his wife’s kaum will not give him anything except payment for specific physical services rendered, to what extent does the traditional bilocal system allow the male room for economic self-Improvement? Are his prospects better if he goes outside his own and his wife’s kaum? Is merantau in fact often motivated by a man’s search for the best way of furthering the prosperity of his mother’s house? l

It will be seen that the answers to these empirical questions are

necessary for the testing of de Josselin de Jong’s hypothesis. We

also need to know what are the responsibilities of the male under bilocal residence and how taxing they are, whether they are resented and whether they are the effective cause of merantau. Let me only say here that preliminary impressions of the results of this investigation support the general hypothesis.

If predictive anthropology is sometimes useful (as we have seen), what are the limits of its usefulness? It is clearly within the limits of permissible prediction to say: because male bilocality is an essentially unstable arrangement, the society must have some institutionalised escape valves. Having said this, one can then go out and look for them; this is a good way to make progress in anthropology.

It is not, however, within the limits of permissible prediction to say: all matrilineal societies in the world must be of one of a specified finite number of types. It is not permissible, precisely because someone will always be able to find cases not included in the original hypothesis, and because these cases do not constitute types. There is, of course, no Minangkabau “type of matrilineal society”; for at the point of specifying such a type, we have to account for all the changes in the Minangkabau system throughout history. There always has been a centrifugal force driving males out of the bilocal system, but at the same time there has been restructuring so as to retain the principle of bilocality. Kato showed recently that in the 19th century, two different subsystems developed in Minangkabau, each with very different social values, epitomized in the terms darek and rantau. Every man henceforth has two roads, the road of darek and of rantau, between which he can choose. Kato shows this development to be quite specific in its historical determination, hence essentially unpredictable.

Now there are essentially two ways of relating ‘predictive anthropology” to history. We may be able to predict at the cost of eliminating history from the equation, as Moyer appears to advocate. Secondly, we may reduce aspects of history to a model. Of the authors represented in this volume, both Kamma and Schefold have tried this. De Josselin de Jong’s merantau hypotheses seem to point in the same direction. Male bilocality is part of what one may call a timeless model of Minangkabau. But due to developments which started in the 19th century, bilocality today has two dimensions. The ancient dimension (where the two places are the sister’s and the wife’s kaum) has not disappeared, but in addition the male tends to divide the mental world into the provinces of darek and rantau, the first of which 1s based on the immemorial adat, whereas the second is, as it was explained to me, like climbing a stiff hill, which cannot be climbed if one carries one’s brother on one’s back. It is, then, the province of adventure and self-fulfilment, antithetical to the values of the first, but not incompatible with it, for many perantau bring the fruits of their merantau gains back to the mother’s house and marry well on the reputation they thus acquire.

Structural analysis thus reduces historical change to a logical transformation — from one form of bilocality to another, thereby making the change intelligible on a symbolic level. Yet history is not set aside by this kind of prediction, for we have treated it throughout as a phenomenal reality. Our model has accounted for the specificity of merantau, for many features that distinguish it from mobility in other cultures. This is, I think, the kind of predictability which is possible if structuralism is combined with the studieveld approach and with an unrelenting preoccupation with the mysteries of phenomenal reality. If there is a distinctive Dutch tradition in anthropology, or at least a distinctive movement identified as “de Leidse richting”, it must lie in the combination of these qualities. The present volume is but a fugitive moment in its development.

NOTE

1 I.e. the exchange among females of male marriage partners who are of higher status.

MYTH AND LAW!

The term “Dutch structuralism”’ is perhaps unfortunate. In English one finds the expressions “Dutch treat” (which isn’t really a treat at all), “Dutch courage” (which at best may be viewed as synthetic courage), and “Dutch uncle” (which is a negation of the normal English associations of “‘uncleness’’). Thus at an unconscious level the term “Dutch structuralism’’ may imply a structuralism that isn’t really structuralism, an artificial structuralism, or indeed the negation of structuralism, and more important, by implication, that “real” structuralism is French structuralism. Faced with such a problem it seems best to resort to the normal anthropological solution, i.e., the use of the native term. It is significant, in this regard, that the literal translation of “Dutch structuralism” into Dutch, 1.e. “Nederlands structuralisme”, is rarely used. Instead one finds the expression “Leidse richting”? used most often to refer to the study of structural phenomena in the Netherlands. The word “Leidse” is the adjectival form of the word “Leiden”. And if one wants to be thoroughly pedantic one might argue that the word “Leidse” refers as much to the city of Leiden as it does to the University of Leiden, for in its formative days the “Leidse richting” was more closely associated with the National Museum of Ethnology (which, largely as a result of the efforts of J.P.B. de Josselin de Jong, is still situated in Leiden) than with the University. Now, however, the term “Leidse richting” is intimately associated with the study of anthropology at the University. Though this is the customary usage, this is not to say that all Dutch structuralists are or have been associated with Leiden. Just to name one exception, H.G. Schulte Nordholt, an important Dutch structuralist, studied at Utrecht under Fischer.

Though these remarks on the association of the practice of structural analysis in the Netherlands with Leiden are perhaps interesting, it is the use of the term “richting” that is significant and revealing. The normal meaning of the word “richting” is

“direction” or “trend”. But more significant is the use of this word instead of the more usual, and indeed stronger terms “school” (school) or “traditie” (tradition). As the usage implies, the study of structural and cognitive phenomena at Leiden has always avoided orthodoxy and pursued its work on a broad front. Thus the appearance of Lévi-Strauss’ Les Structures Elémentatres de la Parenté was welcomed at Leiden as fascinating and stimulating even though there were certain rather basic divergences with the views set forth by Lévi-Strauss. Though these differences of opinion existed and still exist one could argue that Levi-Strauss’ work fits in with the broad spectrum represented by the “Leidse richting”. This is not to say, however, that it 1s impossible to distinguish Leiden structuralism from Paris structuralism.

One can denote several general characteristics of Leiden structuralists that distinguish them from their Parisian counterparts. First, until very recently the “Leidenaars” have been almost exclusively concerned with the analysis of Indonesian material. Furthermore, this regional interest rests on and is supported by an important analytical concept, 1.e., the “ethnologisch studieveld” (usually translated as “field of ethnological study”). In accordance with this concept most Leidenaars have accepted that in the first instance the analysis of structural phenomena should be confined to a group of closely related cultures.” In practical terms this has meant that the Leidenaars have tended not to seck out cultural universals as the Parisians have done. The Leiden view on the search for universals has tended to be that while such universals undoubtedly exist, it is more practical to begin with a closely defined regional field of study. Parenthetically, it is worth noting that the Leidenaars see in Levi-Strauss’ preference for the material from Central Brazil and the Northwest Coast the tacit use of a “studieveld” concept.

Another important feature of the practice of structural analysis in Leiden is also directly related to the focus on Indonesian societies. Because many Indonesian societies are highly literate, one of the most obvious sources of data on these societies has been textual material, not only that collected and transcribed by Europeans but also texts written and composed by the local people in their own language. Many Leiden anthropologists have done their own philological and translation work in preparing these texts for analysis. This is due not only to familiarity with the indigenous languages but also to the sheer bulk of the available material. When translations and critical editions do exist, the