Corps de l’article

Introduction

Leiden structuralism constitutes a tradition or orientation[1] that developed in the 1920s. It derives its inspiration first and foremost from Durkeim and Mauss; after the second world war from Lévi-Strauss; and in the 1980s and 1990s also from Dumont. In this paper, I explore the impact of Mauss on the Leiden structuralists with special attention to the notion of the field of anthropological study as it was applied to Indonesian and Inuit cultures and societies (cf. Saladin d’Anglure 2004).

Marcel Mauss spent the spring of 1898 in the Netherlands. Considering the importance of Mauss to the development of structuralism in Leiden, surprisingly little is known about this visit. Tarot (1999) does not discuss it, and even in Leiden the visit of Mauss only became a topic of interest in the late 1970s. Mauss’s biographer Marcel Fournier (1994: 123-130) devoted several pages to his visit to Holland. Mauss arrived in late December 1897 and departed for England in April 1898. Apparently, Mauss was not particularly impressed by the intellectual and academic life in the Netherlands. In a letter dated 1897 he contrasts the sparkling intellectual life in Paris with the rather dull academic traditions in the Netherlands, traditions that emphasise accuracy, detail, clarity and competence (ibid.: 127). These virtues were indeed considered the hallmarks of the Dutch orientalist tradition of that time. Mauss used his time in Leiden well. Even before he arrived in the Netherlands, he was well acquainted with many leading Dutch scholars of that period. He had studied the works of Steinmetz[2], Caland[3], De Groot[4], and Wilken[5], and in Leiden he mastered the Dutch language to the extent that he was able to read their publications in Dutch. Thus he refers to studies in Dutch by Caland, Van Ossenbruggen and other scholars. According to Fournier (1994), he met Kern[6] and Tiele[7], but worked mainly with Caland. His influence on Leiden anthropology was to become profound and lasting, but it took some time to take effect.

In 1972, P.E. de Josselin de Jong[8] devoted an excellent paper to the relation between Mauss and Dutch structuralism. In it, he shows how especially the study on primitive classification (Durkheim and Mauss 1903) and The Gift (Mauss 1925) stimulated research among the early Leiden anthropologists in the 1910s and 1920s. He notes that whenever the point of views of Durkeim and Mauss diverged, the Dutch scholars tended to follow Mauss explaining that like him they were mainly concerned with non-Western societies (de Josselin de Jong 1972: 80). Moreover, like Mauss, the Leiden scholars tried to combine an external perspective of culture and society with the internal perspective of the participants (ibid.: 82). De Josselin de Jong and Vermeulen (1989: 298) related that Mauss came to the Netherlands in 1898, to arrange an exchange between the Bijdragen of the KITLV (the Royal Institute of Linguistics and Anthropology) and L’Année sociologique[9]. They suggest that Mauss probably met De Groot who held the chair at Leiden and who was also a member of the board of the KITLV. They conclude that the visit by Mauss may have “served as van Ossenbruggen’s introduction to the members of the Année sociologique group or at least have stimulated him in keeping track of their work” (ibid.). However that may be, the initial impact of Mauss’s visit to Leiden on Dutch anthropology was small, and it took another 20 years before the theoretical perspective of the Année sociologique group began to take shape in van Ossenbruggen’s[10] study of the Javanese monca-pat system in 1917.

The origin of the structuralist tradition in Leiden

The development of Dutch anthropology is largely founded on its relationship with the Indonesian archipelago, formerly the Netherlands East Indies. Already in the 19th century the Dutch Government developed training programs to prepare civil servants for service in the colonies. In 1836, a chair in the geography and ethnography of the Malay archipelago was founded at the Royal Military Academy in Breda. In 1877, a chair in geography and ethnography of the Netherlands East Indies was founded in Leiden, and occupied by P.J. Veth (1814-1895). He published an extensive monograph on Indonesia in three volumes (1875-1884)[11]. His successor G.A. Wilken was born in Manado, Sulawesi, as the son of a Moluccan-Dutch woman and a missionary. He was a fine ethnographer and the first to place data from Indonesian societies in a comparative evolutionist framework, but unfortunately died young at the age of 44 in 1891. He was succeeded by the sinologist J.J.M. de Groot. De Groot’s impressive study in six volumes on the religious system of China (de Groot 1892-1910) had a deep impact and was extensively used by Durkheim and Mauss (1903) in their study of primitive classification. At that time, two other important institutions played a major part in the development of Dutch anthropology: the National Museum of Ethnology, founded in 1837, and the Royal Institute of Linguistics and Anthropology, founded in 1851. From 1877 until 1908, when Steinmetz was appointed in Amsterdam, Leiden was effectively the only centre of anthropology in the Netherlands.

When de Groot accepted the chair of sinology in 1904, he was succeeded by A.W. Nieuwenhuis who held the chair of geography and ethnology until 1934. Nieuwenhuis was an evolutionist with a strong interest in the relationship between society and the natural environment. He was not particularly interested in the theoretical issues raised by Durkheim and Mauss that were introduced in Dutch anthropology by van Ossenbruggen in his paper on monca-pat in 1917. Van Ossenbruggen was a very careful scholar and one of the founders of the studies of adat (customary) law in the Netherlands. Mauss knew the work of van Ossenbruggen well and reviewed three papers by him in L’Année sociologique (Mauss 1904, 1906a, 1906b). In his paper on monca-pat, van Ossenbruggen extensively refers to the work on primitive classification by Durkheim and Mauss. The monca-pat (lit. ‘four-five’) system takes shape in a centre with four surrounding cardinal points. Van Ossenbruggen derived his theoretical inspiration from Durkheim and Mauss, and related this principle of spatial organisation to an original social organisation in exogamous phratries and clans that once must have existed. He argued that principles of classification based on this model still operate, even though the social organisation they generated no longer exists:

Wherever a higher degree of cultural development has been achieved, the earlier tribal organization fades into the background, eventually to be lost. Since ideational systems must evolve to keep pace with [social] organization, few if any traces of earlier tribal structure can be detected in such a societies. Instead the basis of organization has become, exclusively, the perceived structure of the universe itself.

in de Josselin de Jong 1977: 43

Thus van Ossenbruggen focused on the temporal dimensions of systems of classification. This problem continued to dominate the theoretical perspective of the Leiden anthropologists. They were well aware that the dual organisations they postulated to explain the organisation of Indonesian cultures no longer existed, and therefore considered them as models developed by the participants themselves in contemporary culture. Van Ossenbruggen drew the attention of a curator of the National Museum of Ethnology, W.H. Rassers[12], to the work of the Année sociologique school. In his PhD dissertation on the romance of Pandji, Rassers (1922) refers to van Ossenbruggen as an inspiration: “If the core of the Pandji tales is a myth […] for which the division of society into two exogamous phratries served as the model, we must now proceed to consider the development of this myth in the light of what Mr. van Ossenbruggen taught us on the origin of the concept of montja-pat” (in de Josselin de Jong and Vermeulen 1989: 301).

Thus the notion of the model came into play at an early stage of the development of the Leiden structural tradition. Duyvendak, the first of 22 students to pursue his PhD under the supervision of J.P.B. de Josselin de Jong (Fox 1989: 503), argued in his analysis of the social and territorial organisation of Ceram that the two most important groups, the patasiwa and patalima, never constituted two exogamous halves of a tribe but that the relation was only mapped on such a model (Duyvendak 1926: 175). The division of the Kakean (a secret society on Ceram) therefore did not necessarily represent the patasiwa and patalima groupings. The main point was that the two contrasts were homologous. In the use of theoretical concepts such as “model” and “homologous relations,” the Leiden anthropologists in many respects anticipated on the discourse of structural anthropology in France and the Netherlands after the Second World War.

Rassers drew the attention of his younger colleague J.P.B. de Josselin de Jong[13] to the importance of the Année sociologique school. De Josselin de Jong was a curator at the National Museum of Ethnology until he obtained a full time position as professor of ethnology at Leiden University in 1935. His 1913 PhD dissertation compared animate and inanimate categories in Indo-European and Algonquin languages and cultures[14]. Most Leiden anthropologists had received considerable training as philologists and some of the leading scholars (such as Rassers and Held) based their PhDs on the study of texts, but de Josselin de Jong set an important trend in the Leiden tradition by combining the studies of language and culture. His position at Leiden University enabled him to act as supervisor of a series of PhDs that all bore the hallmark of the Leiden tradition of anthropology. An important part of this development took place between 1922 and 1935. Nieuwenhuis and de Josselin de Jong had a completely different orientation as the former’s background was from the natural sciences whereas the latter was oriented towards linguistics, archaeology and ethnology. Most promising scholars turned to de Josselin de Jong. The student debating club WDO,[15] founded in 1928, became the centre of an enthusiastic group of young anthropologists discussing new developments (see van Wengen et al. 1995).

By the time de Josselin de Jong took over the chair of ethnology in 1935, the Leiden tradition of anthropology was fully established. In that year, two important PhD dissertations were defended. Held’s (1935) dissertation on the Mahabharata focused on dualism as a fundamental pattern of the Mahabharata and van Wouden’s (1935) dissertation developed a comparative approach of Eastern Indonesia that in many respects prefigured Lévi-Strauss’ magnificent study of 1949.

The Leiden tradition in anthropology founded by Rassers and J.P.B. de Josselin de Jong never was a strong institutional group. Before the Second World War many students of de Josselin de Jong did not hold positions at a university. This situation changed after the Second World War when many of his pupils accepted key positions in Dutch anthropology[16]. These scholars went different ways and in no way represented a distinctive group, although they all acknowledged J.P.B. de Josselin de Jong as their teacher.

The field of ethnological study

In his 1935 inaugural lecture to his chair of ethnology, J.P.B. de Josselin de Jong developed the notion of Indonesia as a field of anthropological study. He opened his lecture with a statement of the problem: “All attempts at classifying mankind into smaller or clearly delimited groups according to race or culture have been utterly fruitless” (de Josselin de Jong 1977a: 166). Classifications and typologies of cultures had only resulted in a “growing aversion to this kind of characterisation and description of culture […] in modern ethnology for the past twenty years” (ibid.: 167). De Josselin de Jong also argued that ethnological research on “individual groups is conducted with a hitherto unknown accuracy and intensiveness, while at the same time there is a concern with wider areas of ethnological research, perhaps best designated for the moment as ‘fields of ethnological study’” (ibid.). He defined the field of ethnological study as “certain areas of the earth’s surface with a population whose culture appears to be sufficiently homogeneous and unique to form a separate object of ethnological study, and which at the same time apparently reveals sufficient local shades of differences to make internal comparative research worth while” (ibid.). As he referred to Australia as “one of our finest fields of study” (ibid.), the Leiden anthropologists assumed that Radcliffe-Brown’s study of Australia constituted the main inspiration for J.P.B. de Josselin de Jong. Thus P.E. de Josselin de Jong (1980: 318) stated: “Recollecting de Josselin de Jong’s lectures and seminars and the reading he required—admittedly some fifteen years later—I think it most likely that he had been particularly impressed by Radcliffe-Brown’s (1930-31) series of articles […].” But as Leiden anthropologist Jos Platenkamp (1991) quite correctly pointed out, the definition of the field of ethnological study was clearly derived from Mauss who was such an inspiration to the Leiden structuralists in the 1920s and 1930s.

Indeed, in his famous 1906 paper Seasonal variations of the Eskimo, Mauss (1979: 19) stated: “But we must not lose sight of the fact that the Eskimo occupy an extensive coastal area. There exist, not one, but many Eskimo societies whose culture is sufficiently homogeneous that they may be usefully compared, and sufficiently diverse that these comparisons may be fruitful.” Moreover, even the notion of a privileged field of study itself was used by Mauss when he stated, “The Eskimo offer such a privileged field of study because their morphology is not the same throughout the year” (ibid.: 19). While focusing on social morphology, Mauss was questioning the anthropo-geographical approach proposed by Ratzel with its main emphasis on geography. He proposed:

[…] to study here the social morphology of Eskimo societies. By this term, social morphology, we refer to the science whose investigations are intended not just to describe but also to elucidate the material substratum of societies. This includes the form that societies assume in their patterns of residence, the volume and density of their population, the way in which the population is distributed, as well as the entire range of objects that serve as a focus for collective life.

Mauss 1979: 19

Mauss argued that the settlement, not the tribe, was “the true territorial unit” in this field of study. In his view, the settlement is not just a territorial unit:

The settlement has more than just a name and a territory; it also possesses a linguistic unity as well as a moral and religious one. Although these two categories may appear initially to be unrelated, we have purposely linked them because the linguistic unity to which we wish to call attention has a religious basis and is related to ideas about the dead and their reincarnation.

Mauss 1979: 27

The settlement thus constitutes the basis of Eskimo morphology. In his article, Mauss tried to demonstrate how the life of the settlement is decided by a seasonal rhythm of intensive social life in winter, and a much more individualist life in summer when individual families disperse out on the land. In his conclusions, Mauss points out that such a pattern can be found among many cultures suggesting a relative autonomy of the seasonal rhythm of society with respect to the climatic and geographic conditions.

In his 1935 inaugural lecture, J.P.B. de Josselin de Jong also attempted to define the Malay Archipelago as a field of ethnological study in terms of its social features arguing that they form part of an integrated whole or system. “The deeper we delve into this system the clearer it shows itself to be the structural core of numerous ancient Indonesian cultures in many parts of the Archipelago” (de Josselin de Jong 1977a: 168). He described this structural core in terms of four distinctive features, namely asymmetric connubium, double descent, socio-cosmic dualism, and a special kind of resilience in reacting to cultural influences. This structural core implied a research program focusing on the relationship between the social and cosmic organisation of society, and the temporal framework of continuity and transformation in Indonesian societies. These topics continued to dominate academic research in the Leiden tradition of anthropology in the next decades.

The research issues of Eskimo society and of Indonesia were obviously quite different. Mauss referred to an area of linguistically and culturally closely related people. In contrast, the Malay Archipelago has great linguistic and cultural variety, and the unity of the field of ethnological study was largely determined by the Dutch colonial context. However, for de Josselin de Jong the core of his argument was not so much the definition of the area, which was already given by the colonial situation, but the development of a research program that allowed the researcher to deal with the problem of cultural variation.

Structural anthropology at Leiden after the Second World War

The Dutch scholars around de Josselin de Jong continued to closely follow developments in French sociology. After the Second World War, they were immediately aware of the significance of Lévi-Strauss on anthropology as his 1949 publication Les structures élémentaires de la parenté was very well received. J.P.B. de Josselin de Jong devoted an extensive study on Lévi-Strauss’s book in 1952. He qualified it as “one of the most important contributions to anthropological theory of the present century” (de Josselin de Jong 1977b: 315). He paid special attention to the problem of reciprocity since Mauss’s (1925) essay on the gift had been used extensively by the Leiden anthropologists in their studies of exchange in Indonesia. De Josselin de Jong argued that Lévi-Strauss underestimated the intrinsic value of the gift referring to the exchange of male and female goods in Indonesian marriage systems.

Lévi-Strauss appreciated the positive reception his work received in Leiden and returned the compliment with a contribution to the Dutch journal Bijdragen in 1956 arguing that dual organisations derived from more complex organisations and not vice versa. In the same issue of the Bijdrage, P.E. de Josselin de Jong published a paper on the participants’ view of their culture, exploring the apparent incompatibility between the views of the researcher and that of the participants. “It strikes a student (I shall not go into details about this) that quite a number of structural principles are not, or hardly, recognized by participants” (de Josselin de Jong 1977: 248). When the paper was published, the issue of views discrepancy between researchers and participants was still mainly an academic question not yet central to the anthropological debate (see also Oosten 2005).

Both J.P.B. and P.E. de Josselin de Jong were excellent anthropologists and inspired many young scholars. They were by no means prolific authors, but excellent teachers and like his uncle, P.E. de Josselin de Jong trained many young scholars who later acquired leading positions in various universities (e.g., Miyazaki, Platenkamp, Geirnaert, Moyer). The Leiden tradition of anthropology was identified with structuralism after the Second World War, when Lévi-Strauss developed structural anthropology as a research program. P.E. de Josselin de Jong was very much inspired by this approach. In his 1977 publication, he traced the origin of Dutch structuralism in the early 20th century. However, the style of Dutch structuralism was in many respects quite different from the French tradition. Indeed, anthropology at Leiden University was part of the Faculty of Arts until the early 1960s and close connections between the study of language, culture, and history remained a distinctive feature of the Leiden tradition[17]. Furthermore, the importance of fieldwork was always emphasised as P.E. de Josselin de Jong (1977: 233) stated:

Cultural anthropology as practised and taught by Professor J.P.B. de Josselin de Jong is characterized by the importance it gives to fieldwork and to structural research. These two aspects of anthropological work, which some hold to be incompatible, are actually complementary. A structural analysis of a society that is not based on data in the field is, of course unconceivable; and one tries to understand what has been observed by considering it as a manifestation of a structure which may itself be hidden.

de Josselin de Jong 1977: 233

By contrast French structural anthropology always tended to subordinate history to structure and give priority to theoretical and general discussion, making fieldwork a secondary aspect of anthropology. However, fieldwork opportunities for the Leiden Indonesianists disappeared after the Second World War, when the struggle for independence in Indonesia first led to Dutch military intervention. When Indonesia became independent in 1949, the Dutch Government refused to hand over Papua New Guinea to Indonesia and relations remained strained until the Dutch finally agreed to the Indonesian claim to Irian Jaya in 1962. Then political tensions and struggles in Indonesia itself prevented anthropological fieldwork. It took de Josselin de Jong until 1972-1973 before he was able to set up a fieldwork project, and even then, politics again intervened with his work in the field and he returned greatly disappointed. He continued to focus on Indonesia, but no longer attempted to conduct fieldwork. Instead, he supervised the fieldwork of his students resulting in many PhD dissertations on Indonesia.

De Josselin de Jong also further developed the theoretical concept of the field of ethnological study. He realised that by connecting cultures, a better framework of explanation can be provided: “The advantage of studying any culture as a member of a field of ethnological study is that the culture under scrutiny and its field of ethnological study congeners become mutually interpretative” (de Josselin de Jong 1980: 319). He argued that “our aim should be to use ancient as well as modern data to enlarge our model for the structural core of the whole field of ethnological study, and then investigate the extent to which the model applies to the culture under scrutiny” (ibid.: 322). He felt that “both LesMythologiques (1964, 1968, 1971) and La voie des masques (1975) are firmly based on a field of ethnological study” (ibid.: 325-326) as it was hardly conceivable “that such a mutually interpretative study of a large number of cultures could have been carried out in areas defined only in geographical terms.”

The field of anthropological study

In 1982, P.E. de Josselin de Jong organised a conference on the topic of Indonesia as a “field of anthropological study” (or FAS), as the field of ethnological study was now renamed. There, de Josselin de Jong (1985: 243) cautioned against an essentialist approach: “members of the FAS-group are not constrained frantically to hunt for core elements in the society they are studying” and “one is not searching for ‘hints and survivals’ or imperfect resemblances, but considers the various transformations of certain basic principles” (ibid.: 246). He thus conceived the field of anthropological study in terms of transformations, not in terms of core elements, as it provides a framework for local comparison:

Local comparisons within the broader FAS-comparisons help to reveal local manifestations of basic elements for comparison which one might otherwise have overlooked; it can explain them when they appear to be anomalous in the context of a single culture. Furthermore, such comparisons serve to elucidate absences of phenomena one expected to find.

de Josselin de Jong 1985: 261

A central issue in the debates was the role of language in the construction of a FAS. In an interview with James Fox, P.E. de Josselin de Jong stated that to the best of his knowledge the idea of the FAS “was inspired by the idea that we recognise families of languages that are genetically related” (Fox 1989: 502). Fox (1980: 333) himself had emphasised the importance of a shift from formal models to comparative category analysis. He suggested that by considering those categories as dynamic metaphores one might better understand the vitality of Eastern Indonesian societies. At the 1982 conference, Blust (1984) argued that a linguistic criterion could be applied whereas Platenkamp (1984, 1996 [1990]) showed that Eastern Indonesia could be fruitfully studied as a FAS despite its linguistic heterogeneity.

In the mean time, colleagues at Leiden University applied the concept of the field of anthropological study to other areas. Adam Kuper examined bridewealth and marriage in Southern Africa, and explained “The regional approach encourages the study of concomitant variation, structural transformation and historical change while imposing a sense of context and meaning of cultural practices” (Kuper 1982: 4). In my own research, I applied the concept of the field of anthropological study to Indo-European cultures (Oosten 1985) as well as in the comparative perspective of the study of Inuit shamanism (Oosten 1986). In both cases, the relationship between language and culture is close.

P.E. de Josselin de Jong retired in 1987 and was succeeded by R. Schefold in 1989. Schefold continued to work in the FAS tradition (see Schefold 1986, 1989, 1994, 2001), but in his work the emphasis in theoretical perspective shifted from structuralism to symbolic anthropology. Thus, he focused in his inaugural lecture on the comparative study of harmony and rivalry as conflicting principles in Indonesia (Schefold 1990).

The structural perspective of Leiden anthropology took a new turn in the collaboration between the French ERASME team (Équipe de Recherche en Anthropologie Sociale: Morphologie, Échange) and the Leiden structuralist research group CASA (Platenkamp 1991). The French research team, founded by Louis Dumont and directed by Daniel de Coppet applied the key notions of social morphology and exchange developed by Mauss in a holistic comparative perspective of society. Like the Leiden structuralists, the ERASME team emphasised the importance of fieldwork and careful ethnography. Furthermore, Cécile Barraud of the ERASME team participated in the supervision of three PhD dissertations on Indonesia in Leiden[18].

Inuit cultures as a field of anthropological study

The notion of the field of anthropological study inspired by Mauss’s famous study of the Inuit seasons, provided a theoretical framework for the structural study of Indonesian societies to the Dutch structuralists in the Leiden school. In the second half of that century it was also applied to other societies before it was applied to Inuit societies again. Since the mid 1980s, I have consistently worked from the perspective of the field of anthropological study in research projects on Inuit culture and history, combining it with the Leiden emphasis on the view of the participants. In a paper on exchange in Inuit winter feasts by Blaisel and Oosten (1997) it provided a comparative perspective for the study of the Sedna feast in south Baffin Island and the Tivajuut in the north Baffin area.

In a series of courses and workshops with Inuit elders developed in close collaboration with Frédéric Laugrand (Université Laval), this approach proved particularly fruitful and has continued to guide our joint research. In Laugrand and Oosten (2002a), we used the FAS approach to map the complex symbolism of the dog in Inuit societies in Nunavut. Applying the principle of P.E. de Josselin de Jong that in a FAS the data become mutually interpretative, we analysed the position of the dog as a structural mediator in Inuit culture. In Oosten and Laugrand (2006), a comparison of epic traditions in the Western Arctic and the various tales, customs and practices in the eastern Arctic revealed the basic pattern underlying the symbolism of the raven in those areas[19]. We also applied the FAS approach in Laugrand and Oosten (in press) and argued:

In this approach linguistically and culturally closely related areas are defined as a field of anthropological study and examined as cultural variants linked by transformations in time and space. Only by charting out the cultural diversity and richness of local traditions can we do justice to the complexity of a field of anthropological study. In our comparison we will therefore not assume that different groupings can be clearly distinguished from each other, but examine the variations and patterns, the diversity and richness of the cultural variation in the whole area under study.

Laugrand and Oosten in press

The application to Inuit culture of the concept of field of anthropological study that was developed by the Leiden structuralists implies a return to some of the fundamental questions raised by Mauss. Whether the settlements really were the constituent units of Inuit morphology, as Mauss thought, is hard to say. The winter settlements he was referring to have long disappeared in the Canadian Arctic and they have been replaced by permanent large communities that are quite complex in their composition and history. Ethnographic sources as well as Inuit oral traditions show that the winter settlements of the past were by no means closed units and that Inuit were travelling around and moving from one settlement to another all the time. Oral traditions of the Inuit as well as archival evidence show that Inuit were always on the move. Local traditions are always connected to other local traditions. Today, the situation is no different. The modern communities are quite heterogeneous in their composition and connections to other communities play an important part in social life.

One may study Inuit peoples from East Siberia to east Greenland as a field of anthropological study or focus on the Central Eskimos as Boas initially did. In 1888, he proposed to consider the Central Eskimo, the Inuit inhabiting Canada, as a separate group. Later Boas (1907: 258) refined his classification by distinguishing the following areas in Greenland and Northeast Canada:

  1. East Greenland

  2. West Greenland

  3. Smith Sound

  4. Ponds Bay, Fury and Hecla Strait, Western Hudson Bay, and Boothia Felix

  5. Southampton Island

  6. Labrador and Southern Baffin Land

In the early 1920s, the members of the Fifth Thule expedition refined the classification of Inuit in Northeast Canada by distinguishing various groupings such as the Netsilik Eskimos, the Iglulik Eskimos and the Caribou Eskimos. At that time, these groupings comprised approximately 500 to 600 people. This classification was a little unfortunate as the names Netsilik Eskimos and Iglulik Eskimos are derived from subgroups, the Netsilingmiut and Iglulingmiut, the inhabitants of respectively Iglulik and Netsilik. The Iglulik Eskimos comprised the Tununirmiut, the Iglulingmiut and the Aivilingmiut. By identifying the whole grouping as Iglulik, parts tend to become identified with wholes. The various subgroups consisted of small local groups, varying between 20 to 60 people in size. The members of each group used to collaborate in the winter settlements and small families dispersed over the land in the summer season. Each family had its own favourite hunting grounds, its own traditions and knowledge. Mathiassen (1928: 91) emphasises that the distinctions are more geographical than ethnological.

Inuit themselves identified groups on the level of the connection between a people and the place or area where they lived or originated: Iglulingmiut, people of Iglulik, or Sallirmiut, people of Salliq (Southampton Island). People frequently migrated to other areas and maintained contacts with neighbouring groups. Thus the Tununirmiut maintained relationships with Akunnirmiut on Baffin Land, while the Avilingmiut were influenced by the Akunnirmiut and the Qainirmiut. Mathiassen (1928) stresses antagonism between the Netsilingmiut and Iglulingmiut, but at the same time the members of the Fifth Thule expeditions found Netsilik immigrants living among the Iglulik Eskimos. Rasmussen’s descriptions of Iglulik shamanism combine especially Nattilik as well as Aivilik traditions.

The demarcations proposed by the Fifth Thule Expedition were basically accepted and can also be found in the Arctic volume of the Handbook of American Indians (Damas 1984). Here Copper Eskimo (Western Canadian Arctic), Netsilik, Iglulik, Caribou Eskimos, Baffin Land Eskimo, Inuit of Quebec, and Inuit of Labrador are distinguished as separate groups. However, on none of the levels discussed above do we find clear boundaries and distinctions. The search for such clear boundaries and distinct units of social life is a quest of anthropologists that makes no sense to Inuit themselves. They know perfectly well that social life is always in flux. People move over the land and change over time and even the social rhythm described by Mauss does not yield the clear distinctions he was looking for. Inuit always endlessly varied their annual cycles and modes of livelihood never conforming to the patterns described by Mauss. Instead of trying to identify clear boundaries, between areas or seasons, we have to accept shifting boundaries that change over time and depend on context. Differences are perceived and constructed all the time. Boundaries and identities change as new modes of livelihood or new frames of references are adopted.

Variation is the name of the game. Each area, each village, each family can be seen as a particular mode of Inuit culture. In different areas we see familiar patterns reorganised in a variety of ways. In Boas’ description of the Sedna feast in Cumberland Sound, the harpooning of Sedna plays an important part. In the descriptions of the Tivajuut in Iglulik provided by Rasmussen (1929: 241-143) and various Inuit elders, we find no reference to the harpooning of the sea woman (cf. Blaisel and Oosten 1997). In the Nattilik areas, the harpooning of the sea woman, Nuliajuk, plays an important part in shamanic traditions, but we find no references to the feast (e.g., Rasmussen 1931: 226). Among the Umingmakturmiut, we do not find the Sedna feast or the Tivajuut, but instead the ascent of the sea woman plays a central part in shamanic practices. Rasmussen (1932: 24-27) provides an extensive description of how the sea woman was lured to the house where a séance was held by the Umingmakturmiut shamans.

When the whalers arrived the Sedna feast in south Baffin Island began to decline. The influence of whalers meant that Inuit in some areas such as south Baffin Island adapted their modes of livelihood to the opportunities provided by the whalers, whereas elsewhere Inuit tended to continue their traditional modes of subsistence hunting. When whaling collapsed, the fur trade had a great impact on Inuit modes of livelihood. The foundation of trading posts as well as missionary posts also affected the movement and settlement patterns. The adoption of Christianity implied that Baffin Island became predominantly Anglican and the Kivalliq and Nattilik areas, mainly Roman Catholic. The two religions competed for a long time, but what lasting effects this competition will have on Inuit culture and society remains to be seen. Thus the composition of a FAS changes over time and distinctions that may be useful for one period, are irrelevant for another.

Conclusion

The influence of Mauss on Inuit studies has been tremendous as can be easily discerned in the work of two leading scholars in Inuit studies, Ann Fienup-Riordan (e.g., 1983) and Bernard Saladin d’Anglure (e.g., 2006). In their research they show how contrasts such as human/non-human and male/female organise social and cosmological categories.

An excellent example of a comparative approach is provided by Saladin d'Anglure (2006). He compares the Canadian myth of a young woman who becomes a shaman with a mythical variant from Alaska about a cross-dressed man who gives birth to a whale calf. Both the man and the woman cross the boundaries between the genders and in doing so become mediators between human beings and the great spirits evoking the structural ambiguities of the shaman, the mediator par excellence between human beings, animals and spirits. As the FAS approach does not look primarily for similarities and generalisation, but for homologies, variations and transformations, it corresponds to the valorisation of cultural differences by Inuit themselves. Indeed, they emphasise that each community, even each family, has its own traditions and they caution against generalisation.

The FAS provides a framework for the study of change and variation as it avoids an essentialist approach that reifies temporary distinctions and boundaries. It perceives a cultural area in terms of transformations and transitions in a spatial as well as a temporal perspective. In its valorisation of change and variation, it does justice to the perception of Inuit themselves and provides an adequate framework for the study of the rich variation of Inuit culture and society as a privileged field of study.