Durkheim and Mauss on Civilisation

In our view, the most promising, least Eurocentric, conception of civilisation in classical sociology and anthropology was the one forged by Durkheim and Mauss in Reference Durkheim, Mauss and Schlanger1913 (Schlanger Reference Schlanger2006, text 3).Footnote ¹ Emile Durkheim did have a theory of social evolution, which was singular (from mechanical to organic solidarity), and you might therefore expect that he would have had a singular theory of the evolution of civilisation. But surprisingly he and his collaborator and nephew, Marcel Mauss, stressed the histories of civilisations in the plural and rejected connecting them to some hypothetical general evolution of humankind, as Auguste Comte had done. What has come to be shared, a civilisation, may have occurred through the spread of institutions, techniques, myths, or other practices and products from a single origin or, they point out, by exchanges among a number of contiguous societies.

Durkheim and Mauss acknowledge the predecessors of their theory in ethnologists, ethnographers, and cultural historians in North America and Germany and museums in France and Sweden. They had established, for Durkheim and Mauss, a less than clear but still distinctly social phenomenon whose spatial extent is bigger than political society upon which their own theories of social order had been based. Tools, styles, language families, institutions of organisation, such as chiefdoms, and types of kinship, spread spatially over time. These sets of institutions have no clear bounds, no single social organism, yet they are linked to each other in an integrated but not a functionally interdependent system. Such a system is properly called a civilisation, which can be constant across languages and political societies. The examples they list at one point are Christian civilisation, Mediterranean civilisation, and Northwest American civilisation. Because civilisations are social phenomena, like all social phenomena they are, to Durkheim and Mauss, moral milieus – they determine a certain cast of mind and of conduct, yet they travel and spread across social boundaries of all kinds over long courses of time.

In a later text, dated 1929 or 1930, Mauss, now writing on his own (Schlanger Reference Schlanger2006, text 7), defined civilisation as ‘those social phenomena which are common to several societies’ ‘more or less related to each other’ by lasting contact ‘through some permanent intermediaries, or through relationships from common descent’ (61). A civilisation is, then, ‘a family of societies’ (62). We can imagine what these permanent intermediaries are when we think of tributary or diplomatic, trading or marital relations. In the technical terms of Mauss’s and Durkheim’s sociology, a civilisation is the spread through such intermediaries of collective representations and practices, which are the social aspect of the materials of civilisation. Mauss says they are ‘arbitrary’, which means they are not universal but preferred modes of making and doing things. In other civilisations the same things are done in different ways, functions performed by different things.

In the actual order of analysis, to say these things belong together as a civilisation is, as he and Durkheim stress, to infer from archaeological, ethnological, and historical evidence a common set of practices and meanings, not one dominant characteristic, design, or thing, but the way they hang together, and to trace their evolution over time and space. Note that these inferences mark limits of civilisational spread. Beyond them are the further spreads of bartered or marketed goods that are accepted for their strangeness, or exoticism, rather than the symbolic meaning or the practice and conduct that goes with them within the civilisation from which and within which they are produced.

Within a civilisational spread there are other boundaries of more coherent social and cultural structures and their centres. These singularities enclose and differentiate themselves from others in similar ways, ways that in fact characterise a civilisation. In this sense, the civilisation, as a way of defining inside and outside, logically precedes and gives societies and cultures a mode of self-definition and internal coherence.

The variation among them increases with greater geographical distance until a civilisational border region is reached where even greater differences are to be found, namely differences between civilisations. But even there in these border zones, local societies and cultures will be creative mixtures of civilisations, related to both or more sides and their centres. And what comes from elsewhere through these border regions may well be absorbed into those centres. Civilisations are centred mixtures.

Mauss relied on there being cores and origins of civilisation (Schlanger Reference Schlanger2006: 67). But he included in ‘civilisation’ the societies of hunters and gatherers, such as the Australian aborigines, and he envisaged four regions of what he speculated might be the huge civilisational spread through the coasts and islands of the south Pacific (63). On land, such as that of Central and West Africa or the Amazon basin, when tracing non-hierarchical societies or a series of small states and the shallow hierarchies of each to claim that they are similar, differentiating themselves from each other in similar ways, the space across which the series runs can be very broad. It is also harder to detect a civilisational border region than when we are dealing with steeper hierarchies and their centres.

For Mauss’s conceptualisation of ‘civilisation’ the idea of a culture area (Kulturkreis), one of whose main ethnologists was Adolf Bastian, was a critical predecessor, criticised for its propensity to single out a cultural object or trait and survey its diffusion, whereas for Mauss you could only make sense of either when you saw how they were related to others in a complex of objects and traits. For Bastian, ‘culture’ and its local variation in a geographical region was the effect of an historical adaptation to the changing ecology of that region, whereas for Mauss there was the additional factor of a culturally autonomous (or arbitrary) process. We too will stress the interdependence between political economy, adaptation to changing ecology, and conquest or other kinds of involvement with other cultures on one hand and the processes of cultural adaptation and transformation as a relatively autonomous and formative history.

Bastian is well known to have been the teacher of Franz Boas, founder of an anthropology of singular cultures. Unlike Boas as well as Bastian, but like Mauss, sharing his stress on ways of making and doing across cultures, we have chosen to stay with the word ‘civilisation’ instead of the almost cognate ‘culture area’.

We also follow Mauss and Durkheim in their moral project – a way of knowing what they called a moral milieu is also a way of knowing how to reform a world lacking moral sense, a distinctly political project. Mauss’s concept of civilisation is no less part of this project than any of his other writings. Like the rest of his and Durkheim’s work it is both an analytic and a critical concept. We will shortly give our view of this critical potential.

While we share with Mauss the centrality of moral aspirations as formations of humanity, we do not stress the higher reaches of civilisations. We reject the idea that those who are at the bottom or at the margins of hierarchical civilisations are any less part of those civilisations and any less human than those who have the accomplishments that each civilisation ranks high. Indeed, it is among the heterodox, at the margins, and at the lower reaches of a civilisation where we often find within civilisations critical disputes and challenges to the claim of being civilised or human. Further, those who retreat from civilisational empires cannot be understood except by reference to what they seek to escape and indeed to some extent still aspire to reach.Footnote ²

We follow Arnason’s (Reference Arnason, Arnason and Hann2018) high regard for Durkheim’s and Mauss’s concept, but we seek to take it in certain directions not followed by Arnason. One is to stress spread, mix, and variation, while Arnason takes from Mauss his more expectable stress on systemic coherence, albeit a looser coherence than that developed by Talcott Parsons in his systemisation of Weber, Durkheim, and Pareto. A stress on systemic coherence can lead to the empirically false idea that civilisations do not borrow and become changed in borrowing from other civilisations. New civilisations emerge out of such fusions at and from their margins, where the hopes and aspirations raised by the criteria of civilisation are dashed or denied.

We are interested in differences between civilisations and in their comparison, as were Mauss and the early twentieth-century ethnologists to whom he referred. But in Mauss’s and our own emphasis on spread and mixture, we reject the idea of ‘clash’ of civilisations put forward by Samuel Huntingdon, late Professor of the Science of Government at Harvard University. In 1993 he published an article entitled ‘The Clash of Civilisations’, forerunner of his famous book defining civilisation as the highest level of cultural grouping, understanding ‘culture’ in the holistic terms of 1960s cultural anthropology. He argued that there had been a historical evolution from the wars of kings, to the wars of peoples, to the wars of ideologies, and now the wars of civilisations and of identification with them. Merely differing in the matter of scale from the classic idea of a culture as a self-defined whole derived from Franz Boas, Huntington used his definition to criticise the West for maintaining a monolithic idea of the history of civilisation in the face of what he described as the emergence of a number of distinct ‘civilisations’ (such as the Muslim and the Sinic). As the veil of the Cold War was stripped away, the reality of these differences was revealed. Huntington’s prediction of a future of endless conflict and difference was soon opposed in terms that we entirely endorse, stressing both the reality of distinctions and at the same time of exchange, cross-fertilisation, and sharing (Said Reference Said2001).

For us, the most interesting characteristic of Mauss’s conception is one that he might have considered to be a weakness and is certainly an abomination to world-history strategists such as Huntingdon who promote holistic notions of culture/civilisation. It is the loose integration of its elements, not a holistic integration. Even though it can be said of a civilisation that it is reproduced, just as social relations, or moral milieus and systems of meaning and material practices are reproduced, we need not feel compelled to put all these together into a single totality and its reproduction. Civilisation is like ‘culture’, but it emphasises the spread of culture. It is like ‘society’, but it is partial, forcing us to think and to infer how elements of a culture or society carry with them habits of relating to others, practices, and ways of making things, differentiating itself from other cultures or societies in a similar manner, namely the manner that is true of its civilisational integument. Further, the integument is transformed with different additions from elsewhere, from other civilisations. ‘Civilisation’ is a grand but not a totalising concept of social, moral, cultural, and material life. It forces us to analyse mixtures, not just the ways in which cultures distinguish themselves from contiguous other cultures but also the spreads of culture into each other and in combination with each other.

As Mauss writes in a third text, not considered by Arnason:

Mauss’s inspiration in thinking about civilisation is clearly ethnological. It resists creating the division between ethnology and sociology that was necessary for the paradigm break leading to the foundation not just of cultural anthropology but also of sociology and social anthropology in the 1920s. Instead it constantly emphasises that there exist phenomena that are not limited to a specific society or culture. They are phenomena common to a larger or smaller group of societies and cultures. These are phenomena – particularly material practices – that are what he called ‘fit to travel’. They overflow boundaries and do not themselves have fixed boundaries.

Societies exist in larger, shared sets of material practices and characteristics. It is not the Durkheimian principle of order that fascinates so much here; rather it is the chaos/order and outside/inside binaries that are deemed necessary for the cohesion of society. Mauss is making the startling point that far from civilisations being forms of society, civilisations are logically prior to expositions of these binaries, logically prior to and maybe also historically necessary forerunners of the societies that form themselves within civilisational spreads.

Transcendence, Immanence, and Writing

Mauss included religious civilisations in the sketches he outlined. So did another comparative sociology that seeks to derive a universal human civilisation from the comparison of religions. Max Weber distinguished ‘universal religions’ from all other religious institutions. They are ‘universal’, he claimed, first because they address a transcendental state of being, namely one or more spiritual beings above and beyond the experienced world. More vitally, they are universal because they profess truths valid for the whole of humanity. Every other kind of ‘religion’ or ritual is, for Weber, ‘magical therapy’ for healing, long life, and wealth by contact with immanent spirits and demons, ancestors, and functional gods.

On this foundation, Karl Jaspers, psychiatrist, existential psychologist, and philosopher of history and a younger colleague and admirer of Weber, in the course of developing a history of thought that attempts to go beyond Western philosophy (2009, original 1949), detected what he called an ‘axial age’ of transcendental thought and self-cultivation. He referred in this way to the extraordinary proliferation, independently of each other, of foundational thinkers between the years 800 and 200 bpe, followed by later founding preachers and teachers of what became state-sponsored religions (though this institutionalisation was in his view against their spirit). They include Socrates, Plato, Aristotle, and the historians and scientists of Greece, the prophets of the Bible, and later on Jesus and Muhammad in the Middle East, Confucius and Mencius, the author or compilers of the Daoist classics (Laozi, Liezi, Zhuangzi), the egalitarian Mozi and the legalist school in China, the authors and compilers of the Upanishads, Gautama (Buddha) and the founders of Jainism in India, and Zoroaster in Persia. They founded religions or schools of thought that were widely influential and extremely different from each other (so different that the age of the mentioned thinkers in China was known as the period of ‘a hundred schools of contending thought’). But they had one thing in common. This was their addressing universal humanity and the cultivation of an inner being to accord with an external, transcendent, beyond but also as a potentiality within human beings. For Jaspers, they constituted the birth of humanity, by which he meant a transcendental, self-reflecting humanity capable in the history of thought of replacing or becoming the Being that was God. Potentiality of human being was for him an adequate interpretation of God or gods. That potentiality is to be cultivated by the exercise of critical reason and independent judgement, as in prophecy, as well as by humane conduct, love, and respect for others. Various and quite different models of such conduct could be found in the Christian turning of God into human flesh, Daoist and Buddhist self-cultivation and perfection as harmony with the Way of the universe or a compassion for all things, and Confucian discipline through propriety and rites. Jaspers’s is a unitary concept of humanity and its potentiality, of a single possible civilisation with several variants. It permits of only one, cosmopolitan perhaps but single and potential civilisation of human being. What was urgent and attractive to Jaspers and to such close followers of his philosophy as Hannah Arendt and the sociologist of religion Robert Bellah is the constant renewal of the religions of transcendental and critical humanity in the face of a human reality full of suffering and increasing scales of violence.

We certainly do not reject this concept of a universal human possibility. But we note that it has to be both transcendental and immanent, a possibility within history but beyond current realities. It is a philosophical aspiration and a discipline of self-cultivation that can be part of any civilisation.

More strictly historically considered, the axial age is, quite possibly, a fact of the evolution of civilisations and religions in Eurasia, but not civilisations in the rest of the world, including Africa and the Americas. There are many possible speculations about the reasons why something similar occurred in such different parts of the Eurasian continent at around the same time. For instance, Morris (Reference Morris2010) has noted that Axial thinkers accompany the emergence of high-end sovereign states. ‘High-end’ means having a bureaucracy and a standing army and relying on a tax-generating and collecting system rather than just on alliances with lesser kingdoms and vassalage (feudal) ties to noble families that raise their own armies. Sovereigns of these high-end states rely on a human mediating a relation to divinity rather than claiming themselves to be gods or shamans to establish superiority to their peers. The civil advisers and ministers of these supreme but human mediators, bureaucratic and highly armed sovereigns, came from minor lineages, were not nobles or royal, but were of course highly accomplished in the arts of literacy. In sum, ‘universal’ or Axial religions and schools of thought were the product of scholars, separate from the sovereigns of states but needed by them, for instance in guiding princes in the arts of achieving long life, good or just and effective rule, or (in China) eventual immortality as a sage. But also and at the same time their advice was available to anyone who could read or who could hear the preaching and other ways by which their thought and methods for self-cultivation were transmitted. And they were themselves adepts at the self-cultivation that could attain the peaks of civilisation. Their advice was not confined to their princes.

Morris (Reference Morris2010: 298ff.) detects a second axial age in which the salvation religions emerged out of the prophets and philosophers in what he calls high-end states: Paulian Christianity in a now divided Roman empire, while in the break-up of the Han dynasty in China, Daoist healing cults and Mahayana and Pure Land salvation emerged. It may well be that this dissolution of high-end states and the violence of their dissolution and re-establishment reinforced the critical transcendence and ritual imaginaries of an ideal state, against which actual states and states of life and death were measured. But all this is only true of civilisations in Eurasia, not the other continents, an indeterminate number of other civilisations for comparison with the Eurasian civilisations and their transformations in the long centuries of their axial and post-axial transcendentalism.

This type of civilisation was, after Jaspers, theorised from a sociological perspective by Shmuel Eisenstadt. He used Durkheim, Mauss, and their colleague Van Gennep but in the main his work was a continuation of Max Weber’s sociology of comparative religion. For him, the key is the distance established during moments of experience of the transcendental from everyday practicality. But he notes, relying on Arnold Van Gennep, that such a state is already available in the middle phase of rites of passage, which the later anthropologist Victor Turner named the ‘liminal’ stage. Eisenstadt locates Weber’s ‘charisma’ in this experience of liminality, a liminality that is common to all ritual. For Eisenstadt, the transcendental is a state of mind as it is for Jaspers, but it is universal and it is ritually induced. It could also be induced by disciplines of solitary self-cultivation.

We can elaborate this insight. Charisma is a property of all ritual processes and an attribute associated with having the knowledge of how to conduct rites. Knowledge gained in the experience of subjection to rituals creates at the very least a distinction between those with and those without such knowledge. Rituals create and recreate a sense of a world divided between what can be sensed or lived, the world of the life cycle and mortality, and a beyond, the invisible. The invisible world of the dead and of the spirits has gateways, accessible to those with ritual expertise. The immanent that can become transcendent is an apprehension of a world, or of a cosmology and its genesis, a sense of the giving of life and the reconquest of the world of the living, in Maurice Bloch’s (Reference Bloch1992) reinterpretation of Van Gennep.

This is our preferred appropriation of Eisenstadt’s insight. But Eisenstadt and his followers, including critical followers such as Johan Arnason (Reference Arnason2010a), conceive of civilisational analysis not simply as the articulation and disclosure of a world and the practices of being in the world. For them civilisation is a separable articulation reliant on writing and human potential. That potentiality is not a fact until it is a self-professed and distinct level of articulation, institutionalised through writing and all that writing implies. In short, for them until transcendence replaces or transforms immanence, we cannot speak of ‘civilisation’. And this is civilisation in the singular, leading to the global civilisation in which we now live. So, despite all the modifications of the original thesis based on Jaspers, for instance Eisenstadt’s own finding that Japanese civilisation was not axial but nevertheless produced a modernity and Arnason’s own stress on the archaic civilisations (Reference Arnason2010b), the Eisenstadtians consider civilisation in the Weberian teleology of what might lead to ‘modernity’ as a single civilisation.

We reject this reasoning, which is a history of the preconditions of modernity. At the same time we acknowledge the irreversible effects of the invention of writing and of the capacity in religions of renunciation of the world to set a particular pattern of transcendence over the world of the living and therefore to question it. But patterns of reconquest of the world of the living are a property of rituals in all societies, and so they too are civilisational. They do not require writing. But the emergence of literate elites is important, for instance in the steepening of hierarchies of aspiration and exclusion.

Our concept of civilisation starts and ends in material practices, including those of oral transmission, hospitality, ritual practices, and forms of divination, which of course continue even while being affected by the invention of writing and the ‘axial’ emergence of the universal claims of what Weber and others define as world religions, renunciation, and transcendence. Transcendence and immanence coexist in all cultures, but the relation between them is irreversibly affected by the invention of writing. Writing is crucial for axial civilisations. But the main moment of the evolution of written religious civilisations, the moment that makes some of them ‘axial’, is when inscribed and transmitted transcendental moments become so distant from vernacular and practical reality that they can become sources for the criticism of practical reality. Then there is a possibility of seeing all the world of the living as mundane, or as another reality radically different from though related to the transcendent.

Eisenstadt differentiated axial civilisations according to their styles of bringing together the transcendental and the mundane, whether they are this-worldly or other-worldly. Confucian, Greek, and Roman are relatively this-worldly (an historical rectification of the world), Hindu and Buddhist (and Daoist, we could add) are relatively other-worldly (ascetic, monastic, aspirations to a relatively impersonal state, or transcendence of and in the body to leave or relieve suffering in the world). It seems to escape his attention that a number of quite different civilisational religions co-existed and were combined in China, both this-worldly and other-worldly, and that this is within the one polity and its economic sphere of influence. If what combines them in China is a single civilisational style, this fact should warn us against thinking that civilisation is commensurable with a single religion. On the other hand, there may be a spread of more than one civilisation in the same geography and population.

What remains important for us in the sociology of axial civilisations, in particular Arnason’s take on them, is the very long-term historical conception of trajectories once a civilisational pattern has been detected. Put another way, a slow rupture sets the pattern of various, often conflicting, articulations of being in the world. ‘Civilisation’ alerts us to a temporally long and spatially patterned spread, including its own transformation.

Writing and Archaic Civilisations

It has been a convention to confine the term civilisation to those cultural complexes that include writing. We want to stand back from this limitation and consider writing to be one among other inventions in material culture.

The anthropologist Jack Goody (Reference Goody2006: 101) notes that along with writing came other inventions of the Bronze and Iron ages, such as the plough, the wheel, and animal traction, which together allowed a single piece of land to produce a large enough surplus for what Eric Wolf (Reference Wolf1982: 79–88) had called ‘tribute’ and a kind of political economy, a ‘tributary mode of production’ – rejecting the Eurocentric distinction between feudal and Asiatic modes of production. This places writing among other techniques and technologies of statehood and economy. In other words they are key elements in the histories of political economy and the production, exploitation, and expropriation of surplus that we would distinguish from civilisation and its eventual hierarchies, shallow and steep. Plainly, each affects the other, new inequalities and classes are important ways in which a hierarchy is created or its maintenance transformed. In the other direction, a civilisation, a way of differentiating a polity from neighbouring polities and a way of self-cultivation, self-restraint, and aspiration, can absorb new – class – relations. Marshall Sahlins (Reference Sahlins2015) has been saying the same about the embeddedness of economies in cultures. But we, more than he does, allow for a greater autonomy of political-economic change, and its absorption into ongoing cultures.

Tribute supported the building and growth of cities. Centres of tributary accumulation were also centres of empires. So writing is an attribute of civilisations forming a tributary mode of production and a class system, and so a steepening of hierarchies. Modes of production with less settled agriculture, or without agriculture, had shallower hierarchies. Age grades, for instance, along with the rites of passage through them, are a shallow hierarchy. Admiration for the dreams and other ways of finding and successfully killing prey are a charismatic form of aspiration among hunters. Diviners and in particular shamans are often leaders, who are venerated in pastoral, nomadic societies.

Writing establishes a record. It is a technology for a shared record and its transmission, outside and beside human memory and oral transmission. In relation to oral transmission, it records what is spoken or sung. A written record is usually placed in a hierarchical position above continuing oral transmission.

Jack Goody (Reference Goody2000) in a summation of his previous publications, including his answers to critics, showed that in nearly all cases their objections were based on too crude a version of what he had been arguing. Reading through his answers to his critics, his argument goes like this: Writing is a technology, and like all other technologies, once invented, used, and spread, it has implications, leads to further inventions, and so it leads to irreversible change – you cannot easily wipe the technology out once its use has spread, but of course, as with any technology, it does not automatically spread. Not only does it have social implications, it also has social prerequisites before it can be widely used for more than one purpose. It has taken many thousands of years, in this instance, for literacy to spread beyond a small minority in those societies where writing existed. On the other hand, once writing has been invented, other inventions follow that have equally far-reaching and widespread implications: printing, and now computer data storage.

Goody argued that writing produces power over those who depend on oral transmission through (a) its technology of the intellect, which is a power of superior knowledge and superior capacity to store and sort out information; and (b) control of the means of written communication, which is a means of domination.

Let us elaborate some crucial extensions from Goody’s argument. First we note that not only does oral transmission continue, but modes of learning by rote are used for learning to read and write just as ritual practices and artisanal skills are learned by copying. Second we note that logographic scripts, like the Chinese – despite every character having a phonetic clue to its pronunciation, these phonetic clues never became a syllabic alphabet – can be read out and comprehended by speakers of many spoken languages. This indicates a universal fact of written language, namely that it always tends to diverge from the spoken, or vernacular, languages to which it is related. There is a built-in archaism in script, which at its extreme becomes the keeping of old texts in dead or archaic languages, and archaic languages, like Latin or classical Arabic, are then treated in many civilisations as sacred.

What is learned by oral transmission and by repetition of the written texts and other parts of ritual practice can and does turn into rebellious movements for the reinterpretation or vernacularisation of writing, or the inspiration of new sacred texts. Texts’ distance from speech brings about the constant questioning of the meaning of texts, of how they should be interpreted, of what they are as a record and as evidence, of how and by whom they were produced, and so forth. But all this occurs in a hierarchy of privilege, which has often been an exclusive privilege of the literate and their claims to superiority over the non-literate. They rely on a ranking of knowledge and literacy into high (calligraphy, knowledge of the classics, and so forth) and low (numeracy, story-writing and -telling or -acting, for instance).

We do not want to confine ‘civilisation’ to civilisations that include writing, however significant and irreversible are the transformations entailed by that invention. But from this consideration of writing we can conclude that every hierarchical civilisation is a hierarchy of orientation to one or a number of centres where it is imagined those who are most civilised live and perform the arts of civilisation, places that are always partly mythical and only secondarily located, such as the real but mythical city of Jerusalem, whether it be imagined in England or in Ethiopia. Cultural historians and archaeologists find such centres. They are also centres of assimilation to a point of reference, points of self-definition of a ‘we’ who aspire to being whatever they consider to be civilised. But we emphasise that in addition there will be similar but rival centres to the same named civilisation.

The Longue Durée and Political Economy

Sociologists of axial civilisations and anthropologists like Goody are not the only ones to emphasise long durational and gradual processes of change. Archaeologists habitually do so. And among historians, longue durée was coined to describe projects following in the footsteps of Fernand Braudel. In his ‘structural history’ (Braudel Reference Braudel1972) the past is conceived as an interacting set of temporal processes combining the short term (events), the medium term (economic cycles and demographic cycles), and the long term (ecological adaptation and its continuities). We would agree with Braudel’s focus on ‘civilisation’ as a spatial and temporal mapping of combinations of material practices, often quite mundane and everyday, that articulate ways of making and doing things that link culture and production in ways that are reproducible over long periods of time. We would concur with his choice of a sea (it could also be an ocean) of communicative transport as a civilisational region. Intellectual association with the techniques et cultures tradition in French anthropology equally focuses attention on how local taxonomies concerning materials are conceptualised and organised (Lemonnier Reference Lemonnier1993). What emerge are striking continuities in the distinct forms of civilisations, often persisting over thousands of years and across the transition from prehistoric communities to dynastic states.

On the other hand, we acknowledge that events of great turbulence or of extended duration can occur at any of the levels Braudel wrongly distinguished into short, medium, and long. Events at any of these levels can create irreversible change not only at their level but affecting the other levels, a criticism already made by Corfield (Reference Corfield2007: 208–10).

Further, as distinct from Braudel’s stress on ecological adaptation, we stress the production, through the technologies of food production and its sharing, through the technologies of the making of pots and implements including weapons, of cosmologies that are equally long-lasting. It is these that convey aspiration and the senses of being in a world surrounded by invisible forces that we consider to be central to ‘civilisation’. Going further, we do not consider ecological adaptation to be determinant so much as a condition. Different cultures and therefore civilisations can prove equally adaptive to the same or similar ecological conditions.

A long-term history of political economy and empire, or what Ian Morris (Reference Morris2010) indexes as ‘social development’ (the formation of urban centres and their size, energy capture per head of population, military capacity, and information capacity), is necessary. But it is not the same as a long-term history of the civilisations that absorb political economies. The two are interdependent, but the ways of making and doing things and what they convey, an encompassing world and aspiration, in hierarchies of the recognition of high status is not the same as techniques in relations of production, the formation of classes, systems of appropriation of surplus, and exploitation, and the politics of command over resources, including human resources, summed together as political economy. We will focus on the former and hint at the latter in this book.

Civilisation and Its Discontents

This is the title of a major metapsychological enquiry by Sigmund Freud (Reference Freud2002), first published in 1930. Kultur, translated as Civilisation but covering more than the French word, according to Freud is an extension of the human organism, a prosthetic in which we sublimate multisexuality and anal eroticism in aspirations to beauty, cleanliness, order, and perfection, to be like the gods who represent those ideals and to resist the demons that threaten them. And the order achieved is an extension of the compulsion to repeat, not to have to choose. The sociologists and anthropologists Marcel Mauss, Norbert Elias, and Pierre Bourdieu have called this compulsion, formed in different ways, habitus.

Freud’s treatise is on the need for and the effect of any civilisation on any human psyche. The need is to prevent harm of two kinds. One is the harm that social reality poses and that is the harm of brute force. The other is the harm and dangers of the non-human natural world. The effect is restraint, the channelling of instincts by character formation, by sublimation and by renunciation. It is also the tendency of the libido, the seeking of sexual gratification once its aims are inhibited, to expand, to comprehend other humans with whom one might interact and who could be harmful. Its discontents stem largely from the inhibition of sexual gratification and the inhibition of aggression, the other main drive to destruction of self and others, displaced onto neighbouring cultures, the external objects of cultural selves, the demonic threats to self in the narcissism of minor differences.

Civilisation is the intervention of the father in the Oedipal drama, the entry not only into language but also convention, manners, and rituals. What Durkheim had from the point of view of this third person called the internalisation of collective representations as images, Freud designated the superego. Vincent Crapanzano (Reference Crapanzano1992: chapter 3), an anthropologist who was also a Lacan-influenced psychoanalyst, has pointed out that the apparent fixity of social and cultural order in the third person is an alienation that allows for self-definition and the idiosyncrasies of desire. In other words, the rules of manners and convention are not fixed but constantly extended or curtailed and reinvented in their application and in the fantasies to which they give rise, including the constant anxiety to make self–other dialogical definition and characterisation consistent. The civilisational third person, initiator, giver, and controller of the word and of habitus, is itself an object of uncertainty, of a desire for recognition, the structure of which has constantly to be remade according to the biographies of each precariously consistent self. You could say this anxiety is a primary ground of discontent. On the one hand the third person hierarchy of recognition and aspiration is idealised and on the other hand it is varied in its application.

The Civilising Process and Modernity as Civilisation

Mauss considered modernity to be a civilisation that had spread across the planet, and he attempted to find in it a civilisational quality or a humanity that could counteract its violent and destructive, anti-civilisational tendencies. This quality hinged on a reminder of the collective representations behind individualised exchange. But we leave that aside in order to review other accounts of modernity as ‘civilisation’ that have become prevalent since Mauss’s writings.

Norbert Elias (Reference Elias and Jephcott1994, original 1939) described a trajectory in European governance, starting with the court of Louis XIV, in which a state assumes the monopoly of the legitimate use of force and governs by guaranteeing realms of public peace. Peace was accomplished, argued Elias, by the inculcation of what was described at the time as ‘civility’ and then as ‘civilisation’, which was a new kind of noun, a verb-noun, describing a process. Elias fixes our attention to civility and courtesy as court aristocratic manners that bourgeoisies aspired to and adapted to their own emergent national characteristics, which in the case of Germany included a celebration of a more poetic Kultur and all-round cultivation, Bildung, terms used to oppose the over-rational, less intuitive ‘French’ concept of civilisation.

All these terms, according to Elias, indicate self-restraint, repression of violent passions, and stopping oneself from intruding upon and violating the bodies of others. Force was still used legitimately to prevent the crimes of the uncivilised and to wage wars in the international system, including of course imperial wars against other European nation-empires and to bring civilisation to the uncivilised. Beyond the boundaries of self-restraint are regions of turbulence and barbarity, where the use of force underlies trade and diplomacy. The use of force protects spaces of peace, which can fragment into smaller spaces of peace under conditions of warlord power, as Georg Elwert later argued through his concept of markets of violence (Reference Elwert, Georg, Feuchtwang and Norbert1999). Peace and civilisation are never secure. The monopoly of state violence can break down and processes of decivilisation ensue, as they did with the rise of fascism. Decivilisation is a form of rule denying civility and aspirations to it for whole categories of population, demonising and dehumanising them in fantasies that have been transmitted by the same one civilising process (Elias Reference Elias and Keane1988).

Elias relied on Freud in his emphasis on the work of shame and embarrassment in repressing volatile desires and drives. His is a study of a particular process that occurred in late medieval Europe and continues today. It emphasises the social conditions for a process of increasing self-restraint that is historically specific. He did not go further and engage in comparison with other similar processes of self-restraint in other traditions, for instance in other religious civilisations. But Elias’s concept was not specific to the European conditions in which the word ‘civilisation’ emerged. Self-constraint and internal peace occur in all figurations. Self-constraint is not ‘modern’.Footnote ³ For instance, the Hadza, a hunting and gathering African society, have constantly to reinforce the necessity to share and not to hide and hoard (Woodburn Reference Woodburn1982). Indeed, Woodburn described all immediate-return hunters and foragers as ‘assertive’ egalitarians, with the emphasis on ‘assertive’.

For Michel Foucault (Reference Foucault and Sheridan1977), the work of discipline in European modernity is far more extensive than Elias’s government of the passions. Discipline and the various bodies of knowledge acquired and deployed through its implementation for maintaining social life are diffused and extended through the institutions of society, including those of the state, but also those of the family, of charities, of schools, and of all kinds of professional training, including military training. For Foucault the state is not separate, as it is for political scientists. It is a centralising function of disciplinary powers.

But, whatever their differences, both Foucault and Elias emphasise the result in capitalist modernity: the self-regulating individual. It could be argued that the self-regulating individual is a key ideal and ideological assumption of the civilisation of modernity. Certainly it is an assumption in law. It may have spread from Europe as part of the ideals of the Enlightenment in its cult of reason. But it is also argued by the anthropologist Sidney Mintz that the atomic individual may well have been produced by the cruelties of European enslavement outside Europe (Mintz Reference Mintz1996). From these sources, or from the English individualism that preceded its industrial revolution, we would have to count a particularly individualist self as the moral person within that ‘modern civilisation’ which was varied in its absorption into the civilisations of the rest of the world through large-scale industrial capitalism and its imperialist spread. But we note also that there is a danger of an assumption of methodological individualism in the very concept of ‘self’ constraint.

Indeed, a variant of modern civilisation is the cultivation of a collective self and the guarantee of work and welfare, which Kotkin (Reference Kotkin1995) calls Stalinist civilisation, describing the ‘break’ into a new ‘war communism’ that constructed from scratch the heavy industry and the new proletariat of the Soviet Union. A similarly described ‘socialist civilisation’ was similarly constructed to mould a new socialist human in China under the leadership of Mao.

Distinctions between inner and outer states of being, formation and subjectivity, of intention and extension, are not necessarily based on a prior inner ‘self’. They can also be based on the prior assumption of an already relational being and relations among a multiplicity of intentions. In our continuing use of such terms as ‘self-realisation’, ‘self-cultivation’, or ‘self-fashioning’ and ‘self-restraint’ in this book we are therefore referring to selves or inner states already defined by and in their relations of mutual being, from the intimacy of kinship onward through the species-specific human capacity of empathy with and of trust in a third person (Feuchtwang Reference Feuchtwang2013).

Elias used the Latin word habitus to describe the learned but unconscious moulding of emotions by senses of shame and embarrassment. Pierre Bourdieu’s use of the word is better known. But in both their uses, they emphasised the habitual dispositions learned in and acting upon a social context, including interpersonal conduct and therefore an embodied sense of propriety – which is a useful expansion of the meaning of ‘moral milieu’. We can now add that any concept of civilisation includes a history of long duration and lasting habitus, which may outlast the reproduction of a particular political economy, even including new and short-lived civilisations such as those of Stalinism. And we can learn from Elias one further extension of his theory of the civilisational process: civilisation, particularly that of modernity and its technologies of violence and vast organisations of force, but all civilisations, including those with less forces of coercion at their disposal, have as their counterparts and in their own states of decline not just concepts of otherness as ‘barbarian’, but act out their own realisations of barbarity. Civilisations bear their own seeds of decivilisation.

For instance, Franz Steiner’s universal history of civilisation, written in 1938, describes a process of civilisation as a way of organising the avoidance of danger and contagion in a history of increasing domination, not over nature, but over other people and peoples. The least civilised, least dominating, place danger and others on their borders. Danger and otherness are ambivalently strange, wonderful, and demonic for these civilisations. The most dominant is the civilisation imposed on and extending across the world from Europe in which the demonic and dangerous as well as its policing are internal (Steiner Reference Steiner, Adler and Fardon1999).

For us, this is a reminder that our concept of civilisation is not evaluative, but is about evaluation, although we will have to deal with a civilisational spread that has, since European imperialism and corporate globalisation, become as large as humanity and been absorbed into many civilisations, including its and their barbarities. Conquest is the barbaric and political-economic aspect of civilisation, as it was for the spread of the Roman empire, which post-conquest admitted into its civility the leaders of the speakers of the other languages and cults absorbed into ‘Rome’.

Historical Human Types: Hierarchical and Anti-Hierarchical Civilisations

There is in the meantime one aspect of most but not all civilisations in the world today with which we have to reckon: hierarchy. So let us now pay particular attention, through Louis Dumont’s treatment of what he calls an ‘ideology’, to the concept of hierarchy.

For Dumont, ‘ideology’ is a system of practices and images that encompasses political economy. The system is that of a hierarchy, the ideal type of which is the Brahmanic Indian one, which determines the relations between jajmani, castes and subcastes, and their functional interdependence by the simple structural opposition of purity to pollution, an opposition that is intrinsically hierarchical in that purity governs its opposite and thus governs all its parts, subcastes. It also encompasses, as a container of its contrary, kingly sovereignty and the status hierarchy of varna, which are also functional groups, similar to estates in feudal Europe. Once the purity–pollution structure has emerged, the holders of positions in the hierarchy of sovereignty and status evaluate themselves in terms of purity and pollution, for a key instance in the gifts they offer to their priests, members of the priestly caste defined as the only purifiers.

In many ways, Dumont’s ‘ideology’ is similar to what we conceive as ‘civilisation’, in its relation to political economy and as a historical product. But for Dumont, the territorial extent of the authority of the ideology is a secondary consideration because he has found it to be the type of hierarchy as such. For us territory is primary both for cosmological reasons and because we emphasise variation across space. In addition, for us there is not a pure type of hierarchy, but an indeterminate number of hierarchies, not all of them necessarily functioning as a structure of complementary opposition. With all these qualifications we accept Dumont’s principal criterion of encompassment and a set of values (which we call aspirations and ideals) by which practices, including everyday material practices, are validated and judged.

One further reservation. For Dumont, ideology and hierarchy are a totality that determines its parts and that either exists or does not exist, whereas for us civilisation is not a totality. It is a totalising process, told us by its cosmology. As much is restated by Rio and Smedal (Reference Rio and Smedal2008: 35–41) in demonstrating that Melanesian egalitarianism is encompassed by a striving of each to gain access to life-giving substances and is therefore hierarchical: an aspiration to become a person whose will is a greater version of the will of each other person. Civilisations and their cosmologies feature totalising objects such as a Big Man or an outrigger sailing canoe and its material as well its analogical relations to gardens and islands of forest in the Kula system (Damon Reference Damon, Sather and Kaartinen2008).

The exception to hierarchical civilisations must be the assertively egalitarian societies of hunter-gatherers that through collective teasing and exposure quash every manifestation of domination and accumulation, which James Woodburn called immediate-return hunters and foragers.Footnote ⁴ Such societies exist in a large swathe of the forests of the Congo basin, but also in Namibia and Botswana and outside Africa in parts of India and Southeast Asia. There is good genetic evidence that at least the Congo examples represent in their chief characteristics the hunter-gatherers of 30–40,000 years ago. Those characteristics include a gender division of labour and female as well as male coalitions in which the female coalition by song, rituals of spiritual prey hunting, and bawdy sexual banter match and prevent male prowess (in hunting and in gathering honey) from becoming dominant. There is no hierarchy in this kind of human society, unlike the hunter-gather societies of more extended scale and of delayed returns of reciprocity, as in the New Guinea Highlands. But there is civilisational restraint, ideology, and encompassment (Lewis Reference Lewis, Dor, Knight and Lewis2014).Footnote ⁵

The example Lewis gives is the Mbendjele BaYaka pigmies in Central Africa. They have initiation rituals in which the myth of the gender division of social organisation is enacted. Ancestors are recalled, but they are not extensions of age grades. Significantly, the BaYaka perform polyphonic music in what Lewis calls ‘spirit plays’, which act as a model of egalitarian division of labour and of being part of a ‘society of nature’ that involves listening to and mimicking the forest and its fauna, which also listen to each other. Masks representing this relation with the surrounding forest are an important part of this singing of ‘spirit plays’ and we would argue that the mask and the euphoric experience of the collective performance are encompassing objects.

Anti-hierarchical peoples also live in Amazonia, descendants of what were parts of or escapees from chiefdoms of cultivators on the alluvial soils along rivers. These hunter-gatherers have certainly evolved their own cosmologies. There may be several such spreads of similar cultures and their cosmologies. But there is a question whether they are encompassing, in the sense of implying a scale of inclusion that contains all lesser spirits or animations. For instance, the Urarina in Amazonia differentiate themselves as more human from their more violent and dominating neighbours (Walker 2013: 208). When they die, Urarina spirits join their ancestral thunder people, described as ‘beautiful’ and celestial, an encompassing reality but not a higher rank. Their cosmology is a world on the verge of apocalyptic collapse, prevented by shamanic virtue and psychotropic trance songs through which the shaman cares for the lives of individuals and of the world (Walker 2013: 179–84). Theirs is a civilisation of sharing, caring, and self-fashioning. It is also ideological, or an ideal, in that there may in fact be dominance by middle-aged men, but it is disavowed.Footnote ⁶ Other egalitarian societies honour ancestors and teach respect for parents in their family or descent group dwellings. This is not hierarchy, because there are no ranks. But it does indicate an asymmetry of roles, for instance between the caring spirit and the human or animal subject or the shaman mediator and the sick person or the world in danger of collapse. But the most powerful spirit of the Urarina, the spirit of Ayahuasca that enables shamans to make their spirit journeys, enabling hunting success and securing the precarious universe, is not an encompassing spirit. A common feature of this civilisation is that it relates to an outside of affinity or notional affinity that makes internal consanguinity possible, thus constituting an expansive inclusion from the outside, including an appeal to the care of outsiders with superior resources and power. But there are no borders of insides and no centres. So, a distinctive cosmology, a self-fashioning but not a centring civilisation, it still has a hierarchy of spirit-masters.

It is juxtaposed with the mestizo agrarian cultivators and pastoralists in the Amazon, just as the BaYaka have for even more centuries been juxtaposed with Bantu-speaking agrarian and hierarchical peoples that they consider to be barbarian and which in turn rank them in the lowest of their own categories. We would describe this juxtaposition, in which the BaYaka are placed in the bottom in the Central African economy, as an excluded rank of an agrarian civilisation, not as a mixture, since the BaYaka retain their own idea of humanity and civilisation. It is an equivalent to conquest without the conquered accepting the standards of the civilisation of the conquerors. In sum, egalitarian societies of hunter-foragers had hierarchical encompassments, a point forcefully demonstrated by Sahlins (Reference Sahlins, Sahlins and Graeber2017) but in order to make the quite different point that the hierarchy of encompassing spirits are proto-political.

Louis Dumont described another kind of egalitarianism altogether in his pair of ideologies: those of homo aegalis or homo minor and homo hierarchicus or homo major. But because this is a comparison of two, the latter is the type of hierarchy, determined by a single structural principle. The former is of modern society and of European society in particular, set off against hierarchy as such, which is of traditional society as such, although there are remnants of hierarchy in modernity. Against Dumont’s singling out equality and individualism and Rio and Smedal’s (Reference Rio and Smedal2008) endorsement of this and thus his conclusion that individualism is an ideology in denial of hierarchy, we rename and reconceive Euro-North American ‘equality’ as the ideology of a meritocratic hierarchy. Its ideology is that those who are at the top have achieved their supremacy by merit, based on equality of opportunity in which there is individual and family mobility up and down, according to ideals of merit in learning and its accomplishment or of risk-taking and its just rewards and of work and its just fruits. For Dumont this is not hierarchy, it is the authority of power and wealth, and the ideology of equality is not a structure that encompasses political economy and its classes. Yet that is precisely what the ideology of a meritocratic hierarchy does and how its ranked parts evaluate themselves. As in Dumont’s hierarchy, the ideology of meritocracy contains but does not determine the command that statuses, in Weber’s sense of relative life chances, secure over their lifestyles through the accumulation of wealth. Further it contains but is not identical with the institutions of authority in either the pronouncement of truth or the exertion of discipline and of force.

Ideologies are ideals, dominant ideals, and the reality of class relations is not a realisation of these ideals either in India for homo hierarchicus or in Europe (or North America, or anywhere else in the world) for homo aegalis. So, one problem shadowing Dumont’s account of these two ideologies is how they and other ideologies are affected by or in turn affect the processes of political economy. The relation between Brahmans and kings is particularly important in this respect (Fuller Reference Fuller1992), the pairing and separation of Brahmanic responsibility for the cosmos on one hand and on the other hand sovereign responsibility for rule. Their separation meant that there never was a single Brahmanic empire, just kingdoms linked by the Brahmanic hierarchies of caste and ritual. From one side or the other, Brahmanic ideals breeding critiques of the world or of political sovereignty adapting those ideals, breed new ideologies or transformations, such as the emergence of Buddhism from the civilisation of the Upanishads and of Dalit (out-caste) adoption of Buddhism in the twentieth century.

Dumont’s account offers no way of saying how the hierarchy might be subject to such transformation, and itself has been the result of structural transformation, such as the struggles between classes and groups that Norbert Elias showed to result in the national characters of civilisation in Europe. It is deficient in one further respect for Hindu civilisation itself, in that it leaves out of account except as a self-excluded counter-example the Hindu tradition of the renouncer, an individual who disowns the world, including the world of caste hierarchy. What makes this objection even more substantial is that there is another individualist Hindu tradition, that of bakhti, which affirms the world of human existence as recognised in its devotion to the supreme deity, a tradition that predates and accompanies the emergence of Dumont’s version of Brahmanic hierarchy. Bakhti comes from all castes and none, form congregations that can also merge with temple cults, or otherwise form networks of teachers, mendicants, and saintly exemplars. Their encompassing supreme deity overlaps with the encompassment of the Brahmanic hierarchy and shares characteristics with Sufi Islam. In sum, Hindu civilisation like the Chinese is a family resemblance of several hierarchies and shallow or non-hierarchical traditions (Fuchs Reference Fuchs, Arnason and Hann2018). Dumont has singled one, steep hierarchy from among them and made of it an ideal type, crucially ignoring an individualism within this complex that is world-critical, both immanent and transcendent, and long predates the Reformation individualism of homo aegalis.

He leaves us with a vexing comparative problem. Dumont has set up binary opposites: aegalis : hierarchicus, in which aegalis stands for modernity, now globally spread, and hierarchicus presents general characteristics, as well as Indian peculiarities of the generalities of traditional, pre-modern hierarchy. The word ‘civilisation’ as we use it, instead of Dumont’s ‘ideology’, replaces this binary with the possibility of defining several such homo as long-persisting but historical human types, of which meritocracy-aegalis and Indian-purity-and-pollution are just two. Indeed, Dumont himself refers to different civilisations, each having their own temporality (Reference Dumont1972: 242). Instead of proposing that one of them, the Brahmanic, is a pure type, as Dumont does, we think it would always be better to conduct a comparison of more than two in order to avoid dichotomies and to specify each instance as a historical human type.

Evaluative and Conceptual Challenges to Civilisational Dominance

The ideology, or the ideal of any hierarchy, includes, as in many instances the body of the ruler includes, all within or under it. This is only valid if those within or in the lower reaches of the hierarchy accept the principles of representation as versions of themselves and of ranking. Whether they do can and should be empirically doubted. But the principle that a hierarchy is ideologically justified as encompassing remains, even if the particular circumstances of a hierarchy may not be accepted. More pertinently, the particulars of those in encompassing ranks may be known to betray the ideals of the hierarchy. Therefore, alternative realisations of the ideology may be espoused by the lower ranks: the ideals serving a critical function.

Returning to there being many hierarchies, or historical human types, even within the complexities of civilisation in India or China, we are including all human cultures in a broader and more linked-up concept of civilisation. Civilisations are processes of self-fashioning constraint. Seen from within these practices, every civilisation conveys senses of a world with reference to its cosmology, which in the case of meritocracy is a temporality of modernisation and a world of nature known, partially controlled, or destroyed by its supreme product: human consciousness. At the same time, within this same civilisation, another post-Enlightenment conception of the human imagination and of its being part of nature was ‘invented’, as Wulf (Reference Wulf2015) points out in her portrait of Alexander Humboldt.

It has become a convention to decry as Eurocentric and as a danger the nature:culture divide of Enlightenment ontology. Forgotten in this convention is that it was decried as soon as it was celebrated as the conquest of nature by science and industry. Another version, just as much Enlightenment, of what had been Creation is that of nature, immanent with a force of its own creation both in science and in poetry, in philosophy and in theory, including ideas of organic integration of humanity with the rest of nature. Against the older science of classification, Alexander Humboldt used measurement to show how everything is interrelated. William Wordsworth and landscape painters, such as John Constable, in both their writing and their painting, as well as in notebooks and prose publications, illustrated intimations of transcendent vision in depictions of the particulars of selected subjects, of what Wordsworth called ‘spots of time’ in his long poem Prelude. Jonathon Wordsworth (Reference Wordsworth1982) shows how the poet worked through conceptions of nature as a capacity to be formed by it through imagination and thus to form others. Wordsworth and Coleridge no less than Goethe, Schelling, and Humboldt wrote of being formed by as well as forming the perception of nature that combined science, philosophy, poetry, and prose. The true vision of a harmonious universe was vouchsafed for smallholding farmers, a pedlar, or a shepherd, though the creative capture by imagination in poetry and paint would be needed to spread the truth of the unity, the organically linked universe that the lyrical scientist Alexander Humboldt conveyed in his already extraordinarily widely read publications. Sublime visions of natural harmony are human as well as being of nature, glimpsed amidst human fear, guilt, fancy, destructive vanity, and greed. This is a version of the natural as cultural and just as much the converse, the cultural as natural. It predates the finding by anthropologists of other cultures’ senses of trusteeship of the plants and animals on which they depend and with which they identify themselves. Maybe it anticipates such findings. In any case it too is a cosmology, shared scientifically with the cosmology of the domination of nature. But each provides a different sense of the worth, the morality of persons, be they scientists, farmers, or business people, and therefore each has its own version of what is natural and also what is unnatural or barbarous. Each is a heterarchy to the other.

There may be several evaluative hierarchies, or heterarchies, in any one civilisation and its spread, co-existing and complementing each other, let alone distinguishing themselves from other civilisations. India is another case in point, since during the period from which Dumont drew his ideal type of Brahmanic hierarchy, it co-existed with Muslim ideology and encompassment and each hierarchy encompassed its own idea of sovereignty and the two borrowed from each other. Another is the African spread of shallow hierarchical centres, each a heterarchy to the other in contiguity rather than within a single-centred civilisation.

Local senses of the world and particular totalising products and practices are not necessarily unified, but they bear a family resemblance to each other in a civilisational spread. In describing them, the anthropology of civilisations does not of course endorse them as a universal standard, even though they may in their own terms claim to be universal. The important point is that a civilisation is a transmission of self-fashioning and aspiration to ideals that might be realised over a time-frame that may include many generations or simply a life course. Most important is that these practices are embedded in everyday material practices.

In an earlier attempt at a comparative historical anthropology of cultures and civilisations in the 1950s and 1960s, organised and heavily influenced by Robert Redfield and his concept of Great Traditions, every (hierarchical) civilisation was seen from its centres downwards. We reject this model in favour of including everyday practices, and doing so without endorsing or prioritising textual traditions, high-status practices, or capital cities and tops of hierarchies in general, in order to see the work of transmission at all levels and how, or if, they work together. Indeed, the fact that the main centres accommodate themselves to less powerful centres within their regimes as well as on their frontiers shows the reverse of so-called Little and Great Traditions is possible. Contrary to the Great Tradition concept of civilisation, we do not reserve a term (Great) for a description of the top echelons, the courtliest, the textual producers. Rather we see civilisation to refer to aspirations towards moral ideals of cultivation and that includes quite ordinary habits of eating, preparing food, and, especially, of rituals of offering and of hospitality, even if as in the Brahmanic hierarchy those below have no expectation of reaching the higher levels of purity.

In every case, but especially once states have been formed, there is political selection from the transmission of civilisational ideas, products, and practices to add to the cultural disavowal of mixture and borrowing. This can be a denial of borrowing from conquered civilisations, of for instance their expertise in healing and hunting, which accompanies exclusion of the conquered from mobility through the politically selected aspirations of self-fashioning.

Civilisations have histories, partly because they can be self-critical, partly because they have to absorb the contingencies of political relations, wars, influences from other civilisations, and the political economies that they encompass. What is more, the concept of civilisation can be used critically, exposing the ideological usage that justifies continuation of privilege and denies the civilisational aspirations of others in the self-justified hierarchy.

Marshall Sahlins and Others on Spreads of Cultures and Civilisation

One reason why we like the concept of civilisation is that it raises the question of transformation through time. Another reason is that once put into the plural, instead of establishing a universal standard for humanity, as it did in its first European usages, it describes the same sort of thing as does ‘culture’ but as a spread, not as a unit.

Sahlins’s (Reference Sahlins2004 and Reference Sahlins, Otto and Bubandt2010) idea of spread is that it is structured by the making of distinctions between contiguous peoples or places that are in warring or raiding relations, in which each defines itself against the other. Each people is dependent on the immediately outside other for its self-definition, and this of course goes from one set of neighbouring peoples to the next. Each is a centre of representation and hierarchy defined against other centres of representation and of hierarchy. But since relations of marriage and treaty with gifts link each to the other, and by conquest the outside can become the other at the centre, the differentiations are internal as well as external. Over long periods of time and contiguity these differentiations become faultlines for intensification and escalation of local conflicts that break out into civil wars within each and of wars of alliances between neighbouring peoples.

There is no whole, just parts defined by structural opposition in regions that can in principle be extended ever outward by their contrasts and their relations to external conditions, in which mythic figures of potential domination which are out of human or internal control are realised in actual external political powers. Each centred culture is defined by that upon which its carriers and creators depend, an outside or an otherness and the compulsion to appeal to or to incorporate what is outside.

In this structural fashion, with the aid of the pervasive figure of the stranger king and of internalised strangers that are created by marriage, Sahlins can include various kinds of spread. They range through empires of hegemony but not direct rule, what Tambiah (Reference Tambiah1973) called ‘galactic systems’ radiating from civilisational centres, to relations of raiding, war, or conquest. Sahlins can show that each identification of a polity, small and large, is also what he calls a ‘cosmocracy’ defined by mytho-historic representations of its actual others. These mythic representations are enacted in rituals of command over life and the sources of fertility, of the giving of life, by an outsider who is also outside the control of ordinary practices and can on occasion deal death and disease instead of life.

Sahlins takes this approach, literally, to China as a galactic centre from maritime Southeast Asia via the Kachin in mainland Southeast Asia and the mediating kingdom of Nanchao on the direct tributary fringe of the Chinese empire. He relies heavily on the image of the stranger king and to some extent on Dumont’s concept of encompassment. The basic idea of alterity and of mediation with the beyond or outside is itself hierarchical and in this broad sense a hierarchy of encompassment of parts can be retained to refer to many kinds of sovereign or of ritual experts, not just those specified by Luc de Heusch, the Africanist originator of the concept of the stranger king or by historians and anthropologists of India.

James Scott (Reference Scott2009) approaches China by the same route as Sahlins. But in his case he uses the idea of China as a civilisation, and turns his back to its centre, describing its ever-receding peripheries as places of an escape from civilisation, a definition against the very centricity and hierarchy that is Chinese civilisation. In a gradual process, over many centuries, people of the mountains, deserts, and swamps have been formed as escapees from civilisation, he claims. But it must also be said, turning in the other direction, that the escapees have their own claims to the civilisation that they have escaped, as manifested in the stories of having had their writing stolen or having themselves lost the skill of writing by a careless act. Such stories transmitted orally are an ironic comment on the fixity and control exerted through the techniques of literacy, in tax registers, cadastral surveys, and censuses of empire. They are the eventual results of flight because of rebellion against over-high taxes and labour demands, including slavery and other forms of forced labour, or from famine and disease, or the seeking of refuge from correction of their ritual practices. Scott calls this process of seeking refuge from civilisation a ‘self-barbarisation’, against the hierarchy of a civilisational state. Ecologically it is a move to swidden from sedentary agriculture.

Scott’s is a one-sided argument from the side of the anarchic. It plays down the opposite pulse between margins and centre, namely the movement from the margins to the centre, from the anarchic to the hierarchic, not only the shallow hierarchies of, for instance, Highland Burma, but back to the steep hierarchy of kingdoms on the great fringes of the hierarchical centre – the tributary or independent kingdoms based on trade and their own agricultural sedentarisation, or the oasis states of Central Asia incubating new civilisations out of the flows of travellers and traders from larger political civilisations. Indeed, as Leach (Reference Leach1977: 240–9) argued, the Nanchao and other Shan states that spread into what is now Central Burma were centres of a civilisation that combined both Hindic and Sinic institutions.

Then there are the much more powerful pulses of the pastoral economies and control of trade routes forming federations and empires in Inner Asian states of aristocratic houses and their vassals (Sneath Reference Sneath2007: 195–8) that made a counterpoint with Chinese empires from the first emperor of China onwards, between which wives, counter-gifts to tribute, trade (for instance for horses), and war could be close to an equivalent of the competition and wars between Athens and Sparta (Sahlins Reference Sahlins2004: 82).

In any case, these pulsations and mixtures can be conceived in Sahlins’s way as part-cultures defined by their contiguous others, if only to invert and oppose their hierarchies. Equally, it should be possible to move in the other direction, from centres of a civilisation outward, along the same path of part-societies and part-cultures that define themselves as distinct and whole, across a number of civilisational centres and on outward.

Civilisation as Knowledge

The problem of recognising uniformities over large areas in an anthropology that also takes localised differences into account has been considered creatively within the framework of social knowledge. Fredrik Barth, for example, accounts for cultural variation by showing how many different groups draw from a similar vocabulary, from a pattern of material cultures and social practices, defining a tradition as an overall pattern in the distribution of knowledge and ideas (Barth Reference Barth1987: 78). He recognises that differences occur in the modes of transaction and handling of knowledge over time so that detecting commonalities between Bali and Melanesia shows how different modes of transmission have channelled their development in very different trajectories. In Bali, gurus increase cultural capital by disseminating knowledge widely and by objectifying it in complex temple and court systems to which all have access. By complementary contrast, in Melanesia, elders, who have been initiated into secret knowledge, hoard and control access to it as a means of retaining status.

A similar attempt to grasp a sense of higher unity shared by particular local traditions can be seen in Tambiah (Reference Tambiah1973) writing on Theravada Buddhism in Thailand, Burma, and Sri Lanka as a tradition that takes different forms or divergent trajectories in these three settings. In other words, Barth and Tambiah both argue that cultural difference occurs as a result of the transmission of knowledge and of distinctive modes of transmission of knowledge although this is only possible because they share cultural commonalities. The particular attraction of Barth’s approach lies in the recognition of how deep analogues in substantive ideas – in his case between Bali and Inner New Guinea – can be combined and shown by comparing the modes by which knowledge is transmitted, adopted, and transformed over time (Barth Reference Barth1990: 640). He suggests that the processes that result from these interactions ‘generate regional trends over time, but also discontinuous variation and incompatible syntheses in different parts of the same region’ (Barth Reference Barth1987: 80). ‘Culture areas are then not only the product of past history; in a very real sense they are being made now, by the efforts of different intellectuals elaborating different kinds of knowledge’ (Barth Reference Barth1990: 650). We could call these different modes of knowledge transmission different civilisations. But we would have to add that their interaction within a region may also indicate a regional style, which might also be counted as a civilisation, a local civilisation made from a mixture of civilisations.

Fred Damon (Reference Damon, Mosko and Damon2005) presents another way of thinking about spreads of knowledge, as a patterning effect, using chaos theory and the pattern of fractals, in which master symbols are media of learning and communication. His example is the outrigger sailing canoe and its construction out of particular species of wood, and the garden cut out of the swidden, which link the societies and cultures touched by the Kula ring. Beyond that he is now also moving back to China along the route that Austronesian peoples migrated across the South China Sea and into the southern Pacific, to enquire whether and how a land-based transformation of the same pattern might be found.

To us, all these different ways of seeing patterns over long periods of time and large extents of spread, within which societies and cultures are parts, not wholes, are inspired by the same sense of scale and duration that reintroduction of a concept of civilisation offers. What these emphases on the transmission of knowledge teach us is that civilisation is a mode of learning, not just of transmission but of absorbing influences and practices that transform even while the mode of learning continues.

Barbarity and the Transformation of Civilisations

Finally, we must include in this review the opposing characterisation of civilisation: barbarisation. It is a convention of most if not all civilisations to nurture a sense of a territorial outside of the spread of civilisation that is given the name connoting ways of life that are less than or un-civilised. On the other hand, this outside can also be considered a resource of great power, to be brought inside or exploited. But in addition to these evaluative schema, which are the materials for the study of any particular civilisation and its comparison to other civilisations, we are reminded by Norbert Elias that civilisations can de-civilise, become barbarous by their own previous standards, or disavow the barbarities they commit by denying the humanity of those violated by conquest.

It is possible to describe the European maritime imperial conquests in this way, barbarising the remnants of the civilisations they destroyed and disavowing their own barbarity, including the barbarity of enslavement, by various justificatory discourses of civilisational evolution that produced scientific racism. This is the singularisation of civilisation that we write against. We apply the evaluative standards of Enlightenment civilisation to show its dark barbaric side, a side intrinsic to the Enlightenment, and move towards a recasting of civilisation as plural and challengeable, from within as well as by other civilisational standards.

Does conquest and destruction spell the end of a civilisation? Unlike empires, civilisations do not have end dates in great events. Not even the burning of the library in Alexandria spells a neat end to the ancient civilisation in Egypt. In any case that civilisation had already undergone great transformations. By then it had become mixed with the various civilisations in the Alexandrian Greek and then the Roman empires. The conquest of the American empires may have spelled a more sudden fragmentation and admixture, with the burning of the Mayan and Aztec codices. But the question of continuity through this violent rupture is still open.

Our main contention remains that civilisations are material modes of learning and self-fashioning that are transformed by long processes of assimilation from each other. So-called ‘lost’ civilisations have not so much ceased as been transformed beyond recognition by internal processes and external borrowings or impositions.