Since the 1940s the term structuralism has become generally used for a certain approach (not a school or dogma), particularly in linguistics, social anthropology and psychology. Although there is some difference between the way it is applied in these disciplines, and between American and European usage, it generally refers to types of research in which the object of investigation is studied as a system. Because a system is a ‘set of connected things or parts’ (Oxford English Dictionary), this entails concentration on the relations between the elements which constitute the system; in the words of Dumont (1970 [1966]):
We shall speak of structure exclusively…when the interdependence of the elements of a system is so great that they disappear without residue if an inventory is made of the relations between them: a system of relations, in short, not a system of elements.
Ferdinand de Saussure (1931 [1916]), who is generally regarded as the founder of structural linguistics, makes a basic distinction between the study of language as parole (speech, that is, language as produced by a speaking individual) and of language as langue: as a system. The system is essential, while parole is ‘contingent and more or less fortuitous’.
The structuralist emphasis on the relations between the elements in a system appears, for example, in Saussure’s discussion of the value (valeur) of words:
the value of words in a language which express similar ideas limit each other’s scope. The value of synonyms, like ‘to fear’, ‘to dread’, ‘to be afraid of is entirely determined by their mutual opposition. If ‘to dread’ did not exist, its meaning would be adopted by its neighbours.
The field most developed by later linguists is structural phonetics, usually termed ‘phonology’ in Europe and ‘phonemics’ in the USA. The aim of phonetics, as a parole discipline, is to give the most accurate description of speech sounds; phonology, on the langue level, is concerned with the question of which speech sounds function as phonemes, in other words, the smallest units which differentiate between the meanings of words. That is to say, it is not the phonologist’s concern whether there is a phonetic difference between the English -p- sounds in pin, prone, up, etc. It is his concern that the opposition between -p- and -b-, -d-, -f-, for example, makes for the distinction between words with different meanings such as pin and bin, pin and din, pin and fin, and so on, that is, that the -p- in English is a phoneme.
Roman Jakobson developed a means to specify what distinguishes each phoneme, that is, what are any phoneme’s ‘distinctive features’: ‘The inherent distinctive features which have so far been discovered in the languages of the world…amount to twelve basic oppositions, out of which each language makes its own selection’ (Jakobson and Halle 1971). These basic oppositions are, for example, vocalic versus non-vocalic, abrupt versus continuant, voiced versus voice-less, and so on. This approach is typically structural, as it defines phonemes as the elements in a system, by considering what distinguishes each element from the others, that is to say, by concentrating on the relations between the elements. The relations, in this case, are of the most elementary type: binary oppositions. It is not surprising that this linguistic method made an impression on the structural anthropologist Lévi-Strauss. Another aspect of the distinctive feature analysis affected Lévi-Strauss more than any other anthropologist, namely its universal applicability: it refers, as we saw, to ‘the languages of the world’.
Two other concepts which anthropology owes to structural linguistics are ‘syntagmatic’ and ‘paradigmatic’ relations. Elements in a language, for example, words in a sentence which, are arranged in a certain sequence, form a syntagmatic chain. A paradigmatic (or, in Saussure’s now obsolete terminology, ‘associative’) set comprises elements, for example, words, which are equivalent in one or more respects. For example, ‘unpardonable, intolerable, indefatigable …’ and ‘teach, teacher, taught…’ are two paradigmatic sets (de Saussure 1931). These concepts have been applied, again particularly by Lévi-Strauss, in the analysis of myths. The events narrated in any single myth form a syntagmatic chain, while the personages and events occurring in a myth, or a corpus of myths, can be studied as members of paradigmatic sets.
It is not unlikely that Saussure was familiar with the works of Emile Durkheim, which were very influential in his time. Durkheim can be considered as one of the founding fathers of modern sociology, but structural anthropology can also trace back its origins to him, his collaborators, united around the journal Année Sociologique, and his pupil and successor Marcel Mauss. The work of this school can be exemplified by Durkheim and Mauss’s joint publication ‘De quelques formes primitives de la classification’. The opening sentences of this long article are typical.
Contemporary psychology has shown how very complex apparently simple mental operations really are, but ‘this operation…has been only very rarely applied as yet to operations which are properly speaking logical’. The authors then demonstrate such logical operations in several non-western societies by describing systems of territorial classification:
In totemic societies it is a general rule that the tribe’s constituent groups, namely moieties, clans, subclans, arrange the territorial sectors which each of them occupies in accordance with their mutual social relationships and the resemblances and differences between their social functions.
Durkheim and Mauss (1963 [1903])
For example, when the entire Wotjoballuk tribe of New South Wales is (temporarily) united in one territory, one of the tribe’s moieties must always occupy the northern, the other moiety the southern area. The two clans with the sun as their totem occupy the eastern portion of the settlement, and so on.
It was particularly this work from the French school which served as an inspiration and an example to (amateur, and later professional) anthropologists in what was to become the other consistently active centre of structural anthropology, The Netherlands. Their earlier (roughly pre-1950) works, usually based on data from Indonesia, give clear evidence of their origin. They are concerned with orderliness or system as it appears in the deeds, words and works of the members of the investigated societies—this is common to all structural anthropology. But the earlier Dutch writers, like their Année Sociologique exemplars, concentrated on ‘ordering’ rather than on ‘order’; on structures of which the social participants are aware, and which they deliberately construct. In addition, the nineteenth-and early twentieth-century structuralists in both countries shared the idea that social structure serves as the model for all other classification systems. From the study of territorial classification, Dutch structural research fanned out, as it were, to the structural principles, particularly binary oppositions, in Javanese material culture, traditional theatre, and mythology. In the 1930s, however, kinship and marriage systems became the focus of interest. A major discovery was the frequent occurrence of ‘asymmetric connubium’, a system whereby marriages between individuals are so arranged that they conform to a regular connubial relationship between groups (clans or clan segments): one group always acts as ‘bride-giver’ to a second, while this ‘bride-receiving’ group gives women in marriage to the males of a third group (de Josselin de Jong 1977). This brings us to more recent times, and back to France.
Claude Lévi-Strauss is the foremost exponent of structural anthropology. His first major work (1969 [1949]) might be called a rediscovery of asymmetric connubium (which he terms échange généralisé), but he places it in a much wider and richer context: ethnographically, by using material from Siberia, China, India, South East Asia and Australia, and, above all, theoretically, by making the échange system shed light on the concepts of incest and exogamy, and on the opposition, which is to be fundamental in all his subsequent work, between ‘nature’ and ‘culture’.
‘In language there are only differences’ (de Saussure 1931 [1916]); Lévi-Strauss applied this typically structural viewpoint in his book on totemism (1962 [1962]). Totemism does not associate each clan with one animal species as its totemic ancestor, but consists of a classification of the animal world, based on the ‘distinctive features’ of each species, and a classification of clans on the same basis. When the two classification systems are associated with each other, the result is totemism, in which ‘it is the differences which are similar’ (1962). Lévi-Strauss’s study of a set of cultural phenomena as a system of variations has its climax in the four volumes of Mythohgiques (1970–9 [1964–71]).
Myths are the purest manifestation of La Pensée sauvage (1966 [1962]): thought which, in contrast to ‘domesticated thought’ does not aim at practical results, but tries to solve problems as an end in itself. The problems dealt with in the 813 myths of South and North American Indians analysed in Mythologiques are, principally: Why do we humans prepare our food, and animals not? Why can we take off our clothing and ornaments, and barter them with foreign groups, while the animals can not? How did this come about? In other words, the problem of culture versus nature.
In analysing myths, a basic precept for the investigator is to avoid ‘mythemology’, that is, the interpretation of each single mythical personage or event in isolation (the heroine ‘stands for’ fertility, travelling by boat ‘stands for’ long life, and so on). Here again, it is not the elements, but the relations between the elements that is essential (for example: the lizard, as a land animal, stands in opposition to the aquatic crocodile; hence they are also in opposite relationships to the human hero: the hero of Myth 1 chases lizards, the hero of Myth 124 is chased by a crocodile).
By the same token, a myth can never be understood in isolation, but should be studied in its relation to other myths. All the American Indian myths studied in the book are to be considered as variant versions of one another, linked together by ‘transformations’. That is to say, one does not compare myths when, and because, they are similar, but because of their differences: the myth corpus, like a language, is a system of differences—sometimes even of perfect oppositions.
Lévi-Strauss uses the same method in La Voie des masques (1975). A certain type of mask used by the Salish Indians of British Columbia is his starting-point for a study of comparable masks in the same region: comparable, not on the grounds of similarity, but of systematic ‘transformations’. Part II of the book discusses the Dzonokwa, a type of mask which is the opposite of the Salish mask in every respect: in form and colour, in its ritual function and performance, in the myths about its origin, and in the way it is obtained and inherited. By introducing the concept of transformation, Lévi-Strauss has added a new dimension to comparative studies in general, and revivified comparative anthropology. Also in contrast to the earlier French and Dutch structuralists, he does not confine his research to structures of which the cultural participants are aware. On the contrary, his frequent references to the ‘structure inconsciente de l’esprit humain’ indicate his particular interest in not only unconscious structures, but also in basic structuring principles which are not culture-specific, but (probably) of universal occurrence.
Outside anthropological circles there is a tendency to equate structural anthropology with Lévi-Strauss; this is a popular misconception. It is striking that, while Lévi-Strauss has a tendency (stronger in some works than in others) to study the products of the ‘unconscious structure of the human mind’ as closed systems, the aim of many of his French congeners is to link the conceptual structures more closely to social problems. Georges Balandier demonstrated that three binary oppositions (male-female, elder-younger, superior-inferior) are frequently the basis of conflicts. Roger Bastide applied the insights of structural anthropology to the problems in developing countries. Louis Dumont (1970 [1966]) demonstrated the fundamental difference between inequality and hierarchy: the latter is exemplified by India, for ‘the caste system is above all a system of ideas and values, a formal, comprehensible, rational system’. Maurice Godelier remedied structuralism’s neglect of economic factors, and thereby achieved a synthesis between structuralist and Marxist anthropology. Although Roland Barthes can also be mentioned in this context, as he studied the effect of ‘mythologies’ on modern western societies, his principal achievement was his comparison between myth-as-language and natural language, thus making good use of the old association between structural linguistics and anthropology. Georges Condominas also closed a circuit, by directing the research of his active Centre for South-East Asian Studies to the topic of l’espace social, that is, territorial classification.
Among British anthropologists there are several whose works are typically structural, although they might not call themselves structuralists. In view of the British tradition of social anthropology, it is not surprising that the tendency, just discussed in connection with the French group, also appears in British structuralist publications, be it in a different form.
Rodney Needham’s works on matrilateral cross-cousin alliance started from a position very close to Lévi-Strauss’s Les Structures élémentaires de la parenté, but diverged sharply with the introduction of the concepts of ‘prescriptive’ and ‘preferential’ alliance. E.R.Leach’s position moved in the opposite direction: sharply critical of the Structures (which he called a ‘splendid failure’), he came to be more and more in sympathy with Lévi-Strauss’s views in his later publications on myths and belief systems. While one of Mary Douglas’s (1966) best-known books could be called ‘Lévi-Straussian’, the social context (in the form of the pressure exerted on an individual by his ‘group’ and the society’s ‘grid’) plays a dominant role in a later work (Douglas 1970). Victor Turner’s principal works are concerned with ritual. Lévi-Strauss has only very seldom dealt with this subject, perhaps because rituals, by their very nature, have also to be studied as socially operative. Turner devoted much attention to this aspect, for example, in the case of ‘rituals of affliction’.
Of structuralists outside linguistics and anthropology, the most prominent are the psychologist Jean Piaget, the historian Fernand Braudel and other members of the Annales group. As a structuralist, Piaget is noted for his emphasis on self-regulation as a characteristic of structures. Braudel is very close to Lévi-Strauss when he recognizes structures as one of the three types of history, history de langue durée, of which the participants in the events are not conscious.
P.E.de Josselin de Jong
University of Leiden
References
Douglas, M. (1966) Purity and Danger, London.
——(1970) Natural Symbols, London.
Dumont, L. (1970 [1966]) Homo Hierarchies, London. (Original edn, Homo Hierarchicus, Paris.)
Durkheim, E. and Mauss, M. (1963 [1903]) Primitive Classification, London. (Original edn, ‘De quelques formes primitives de la classification’, Paris.)
Jakobson, R. and Halle, M. (1971) Fundamentals of Language, The Hague.
Josselin de Jong, P.E.de (ed.) (1977) Structural Anthropology in the Netherlands, The Hague.
Lévi-Strauss, C. (1969 [1949]) The Elementary Structures of Kinship, London. (Original edn, Les Structures élémentaires de la parenté, Paris.)
——(1962 [1962]) Totemism, London. (Original edn, Le Totémisme aujourd’hui, Paris.)
——(1966 [1962]) The Savage Mind, London. (Original edn, La Pensée sauvage, Paris.)
——(1970–9 [1964–71]) Introduction to a Science of Mythology, London. (Original edn, Mythologiques, 4 vols, Paris.)
——(1982 [1975]) The Way of the Masks, Seattle. (Original edn, La Vote des masques, Geneva.
Saussure, F.de (1931 [1916]) Course in General Linguistics, New York. (Original edn, Cours de linguistique générale, Paris.)
