The State and People Who Move Around
How the valleys make the hills in Southeast Asia
On 14 December 1998 James Scott, Eugene Meyer Professor of Political Science and Anthropology at Yale University and former President of the Association for Asian Studies, delivered the IIAS Annual Lecture in the Academy Building of the University of Utrecht. An edited version of his speech is printed below.
By James Scott
In preparing this talk, I have come to appreciate quite consciously, how much my argument owes to Dutch scholarship. This is perhaps not the place for me to recount my many intellectual debts, but I would like to make one exception. That is for Professor Wertheim who died recently. I believe that, twenty or thirty years from now, his works on Indonesian society and elite formation, his book 'Evolution or Revolution', and his concept of 'counterpoints' will still be consulted for their imaginativeness and their foresightedness.I came to visit him in 1974. We went for lunch and as we left, we came to an intersection at which the traffic lights were against us. But, since there was no traffic coming, I stepped out into the street and Wertheim said: 'James, you must wait'. Admonished by my elder, I stepped back onto the pavement and said: 'But, Doctor Wertheim, nothing is coming?'. And Wertheim said to me: 'James, it would be a bad example for the children'. I have always been impressed by the relationship between those two qualities of Dr Wertheim: a radical scholarship and a strong sense of correctness and civic spirit.
The talk I am giving today is essentially the talk I would have given five years ago if I had not taken the detour that led to my recent book: 'Seeing Like a State'. The question I ask is 'why has the state almost always been the enemy of people who move around?' Whether these people are Bedouins, Berbers, gypsies, wandering Jews, hunter-gatherers, nomadic pastoralists, slash-and-burn cultivators, the so-called sea-gypsies or 'Orang Laut' of the Malay world, masterless men, vagrants, sturdy beggars, the homeless. Why has it been the project of nearly every state to concentrate and fix its population in space?
My guess is that sedentarization is the oldest state project in the world. In many respects it is the prerequisite to perhaps the second oldest state project, which is taxation. This helps to explain, I think, the great social cleavage in much of mainland Southeast Asia, between hill peoples and valley kingdoms. The permanent settlement of population is a state project in Southeast Asia that spans pre-colonial, colonial, and post-colonial regimes.
The relationship between hill peoples, who are relatively mobile and valley peoples who are relatively fixed, is also in the Malay world the distinction between the upstream, 'the Hulu', and the downstream, 'the Hilir'. I suspect this distinction would also help us understand aspects of traditional state-craft and frontier people in China; the 'peripatetic Hakka', and perhaps many of the hill peoples of 'Yunaan' province bordering Southeast Asia. In India it might help us understand the relationship between the 'Mogul rulers', and later the colonial regime on the one hand, and hill peoples, or nomads, on the other.
It is worth noticing that this phenomenon is not a uniquely Asian phenomenon. Take for example, the gypsies, the 'Roma' and 'Sinti'. They are in constant movement. In the gypsy-holocaust, perhaps even more than that of the Jews, one can say that gypsies were seen as the original outsiders who belonged nowhere, who had no place of residence, and no social affiliation. Post-war efforts to sedentarize them make it clear how permanent the state-project tends to be.
Let me return to Southeast Asia and start with a demographic fact that in 1700 there were roughly six people per square kilometre in Southeast Asia and that this compared to something like thirty per square kilometre in India or China. The point that I wish to make is that rounding up people and concentrating them in a particular place was the central preoccupation of statecraft in Southeast Asia. No people, no state! The state then had a perfectly obvious preference for a certain demography and a certain geography, which it typically tried to promote and maintain. I like to think of these arrangements or situations, as what I shall describe as 'state-spaces'. The key to a state-space is that any state-making enterprise will require a ready supply of manpower and foodstuffs close at hand.
Given the logic of appropriation in pre-modern settings, such state-making required something like wet-rice cultivation on a sufficient scale to allow the production of a reasonably reliable surplus within a small radius of the court. The 'central place theory' elaborated by Von Tunen, Christaller, and G. William Skinner explains this in considerable detail. Within less than a hundred miles, for example, two oxen will have eaten the equivalent of the load of grain that they could pull in an oxcart. It does not pay to bring basic foodstuffs long distances, although water transport can modify this logic greatly. Other things being equal, of course, the greater the distance goods are carried, the greater their value per unit weight and volume. It follows then that mainland Southeast Asian states were located in areas of concentrated production, usually wet-rice agriculture. These places favoured state-formation, they raised the odds for state formation, but they did not guarantee it.
You could also get state formation without large irrigated wet-rice cores, for example Melaka, Srivijaya, Jambi, Palembang, and Pasai. This is a very special case which would seem to be against my thesis. These were all states that controlled 'choke points' on a river system or in a navigated strait. They were all trading kingdoms, concentrated in the Malay world, in these Hulu-Hilir systems. They were, I think, the exceptions that proved the rule. Their problem was always a lack of manpower and they were all, without exception, slaving states that ranged far and wide to grab manpower wherever they could find it. But the problem was most severe in the coastal kingdoms that did not have an irrigated wet-rice core. The Malay states resemble nothing so much as the Viking system of trading, raiding, and slaving. Several very important consequences follow for statecraft and for social structure.
The first is that Southeast Asian classical states had no interest in territory per se, especially territory far from the core area unless it was a valuable mine or an important pass controlling vital trade routes. As a court proverb from Siam says 'Yes, a soil, but no people. A soil without people is but a wilderness'. Both Burma and Thailand were organized by personal service and bondage, not territory.
A second characteristic, that has to do with the shortage of manpower, is that warfare was all about grabbing people and settling them back near the central core area. For example, by 1800 up to three hundred and fifty thousand people, perhaps one-fifth of the agricultural population of Upper Burma, consisted in military deportees or their descendants, concentrated on irrigated royal service lands. A kingdom defeated was a kingdom whose population was quickly swept up by its neighbours. After the 1569 Burmese sacking of Ayudhya and the deportation of tens of thousands of captives, the Khmers swept in, in turn, for the next two decades, sweeping up all the captives they could find. In 1767, after sacking Ayudhya, the Burmese left with at least thirty thousand captives, including much of the court and the nobility. In fact you could say that much of Eastern Indonesia was scoured clean of manpower.
A third characteristic of statecraft is that the ferocious desire for manpower meant generally that captives and slaves were relatively easily assimilated into the society of their captors. Manpower needs tended to reduce ethnic contempt, especially, but not only, for lowland peoples. Better put, systems of kinship and social organization were remarkably inclusive.
I believe it can be argued that Malayness, far from being an ethnic identity in some original, primordial sense, was rather simply the terms of cultural accommodation necessary to the creation of a cosmopolitan mini state at the coast. If you spoke Malay, professed Islam, and generally followed local dressing customs, you were very quickly a Malay, no matter where you came from and what your original tongue was.
The fourth characteristic of the manpower-starved state is that wars were not very bloody. Why destroy what is basically the prize or objective of war in the first place? Capture, not killing, was the objective of most military campaigns.
Finally and perhaps most important, flight was the basis of freedom in Southeast Asia. It was far more common than rebellion historically. I invite you, in this sense, to see statecraft as a never-ending struggle between the in-gathering of people through war, slaving, and resettlement on the one hand and the scattering and dispersion of people through movement away on the other.
The job of state-making is the creation of legible state spaces at the core of the kingdom. In pre-colonial eras this involved rather crude techniques: attempts to enumerate population listed by settlement and dues, very primitive forms of cadastral survey, corvée to expand the productive land concentrated at the core, and the like. Forced settlement and efforts to stop the leakage to the periphery (and the leakage to officials and nobles who were themselves trying to commandeer as much manpower as they could and to keep it out the hands of the crown) were common. But these techniques were all quite clumsy; they were capable of crude movements but nothing very subtle. Resources were either not mobilized or, more likely, campaigns of taxation and conscription and corvée drove people to flee from the centre.
The modern colonial state, of course, had many more exquisitely detailed and administratively competent techniques: the census, the cartographically correct cadastral survey, crop-cuttings, land-titles, licences, soil classifications, identity cards. These newer techniques were much more effective at pinning down manpower and production to make it available to the central state. And I believe without going into it in any detail today, that the newer techniques for concentrating population have the same general character. That is, if we think of the 'strategic hamlets' as a counter-insurgency technique, of planned transmigration villages in Indonesia, of plantation-agriculture generally, or even of Pol Pot's perfectly square fields and communes, these were all schemes designed to make population and production more legible, more controllable, and more appropriable.
Let me turn away from the goal of statecraft - the creation of legible state spaces - to its opposite: illegible, non-state spaces, the spaces towards which peasants were often fleeing. I know that this stark dichotomy will not hold up to close scrutiny. But I ask you to indulge my briefly to explore its value as a heuristic tool.
The geographical terrain is important, as is the ecology and demography that follow from it. If we were to think about non-state spaces generally, we could enumerate many examples, like swamps, marshes, deltas, bayous, mangrove swamps. The Orang Laut in Southeast Asia tended to live not on the strand which was exposed and where they were vulnerable to slave raids, but in the mangrove swamps which were in impenetrable.
Think also of hills and mountains that are particularly illegible. This calls to mind, in the context of the United States, Appalachia which was, in the eastern part of the United States, the main centre for illegal distillation of liquor during the Prohibition era and afterwards, and is today the major agricultural area for the production of marijuana, partly because it is an area in which state control has always been tenuous.
Think, thirdly, of deserts and wastes, Bedouins, Berbers... a different kind of illegibility in the desert. Think of frontiers and steppes, e.g. the Don Cossacks in Russia.
Here is the paradox, I think: the hill-valley social divide has, we know, been crossed in both directions by countless numbers of Southeast Asians. That is, hill peoples are becoming valley peoples all the time. Valley peoples are running away to the hills all the time. Nonetheless the cultural divide between the hill and valley is stunningly constant as an experienced and lived essentialism.
Let us concentrate for a moment on the perceived cultural differences. In Southeast Asia, there is a powerful cultural symbolism which the elite of court kingdoms use when they see these peripheral peoples and peripheral areas. They see them not just as people who are just out of reach. They see them instead as exemplars of all that is uncivilized, barbaric, and crude. Even when they are looked at with some sympathy, as they are by current 'developmental regimes', they are seen as benighted primitives, 'our living ancestors' who need to be developed, brought into modern life. They are thought of as what we were like before we discovered Islam or Buddhism, rice cultivation, sedentary life, and civilization.
The goal of my little enterprise then is to rework our understanding of the frontier, of the hills, of the Hulu, the periphery, our understanding 'of non-state spaces'. In short, my contention is that the hills are not the backward, not yet incorporated earlier, and ruder forms of social and productive organization, which will later be superseded by sedentary wet-rice hierarchical society. Instead, I think, the frontier is best seen as a place that is always being produced, as the social and economic consequence of state-making projects in valley kingdoms.
The hills and valleys, of course, have always been intimately joined. The Hulu and the Hilir are naturally complementary economic zones. The natural unit in the Malay world is this hill-valley symbiosis, with grain, often salt, fermented shrimp paste, manufactured goods, porcelain, prestige goods, moving up the river. Coming down the river are ores, livestock, charcoal, rattan, spices, exotic trade goods such as camphor and damar, wild game, opium.
Each Malay kingdom lived or died depending on whether it was able to monopolize this trade at the choke points on the river system. Coastal kingdoms were much more often competitors than allies. Coastal states had relatively little to sell one another. Hill groups, though to lesser extent, were also often natural competitors for the best patrons at the coast. But what is interesting is that this symbiotic, integrated effect of the natural patterns of exchange was, at the same time, perfectly compatible with consistent and persistent cultural differences. Despite constant personal movement across this boundary, it nevertheless persisted as a powerful cultural distinction'. The question then perhaps is why these differences are so historically persistent, so salient, and so productive of this civilizational discourse when the peoples in question have always been in close contact, when assimilation to valley culture, language, and status has always been a common process and when there are any number of social locations between the two, and when the social divide itself is more a continuum than a dichotomy.
The standard reasons one encounters to account for these cultural differences are essentially ecological. While true as far as they go, they are, in sum, unsatisfactory. Hill peoples are, according to this account, more dispersed. In turn, the nature of common property regimes typical of swiddening or horticulture, and hunting and gathering, and their technological simplicity, produces a relatively more egalitarian society in the hills. And, being unpromising settings for state formation, given the ease of flight, hill settlement lineages and families are relatively more autonomous. Thus, the process of community fission and breaking up is rather more common. Finally, even if hill peoples had started out indistinguishable from valley populations culturally, the ecological conditions of livelihood in the hills would have favoured a different social structure.
Now that is the ecological argument. I want, however, to add to this a very different argument to explain how the hills make the valleys and the valleys make the hills. We may begin with processes of social circulation: the equilibrium between in-gathering - forced and voluntary - and dispersion - also forced and voluntary. Focusing on this equilibrium, I hope, makes a little more sense of this civilizational discourse. So let me return finally to some of the ways in which the hills are constantly being produced by state-making projects in the valleys.
Think, for example, of flight from the ranks in armies. The Burmese invasion of Siam in the 1780s, began with two hundred thousand troops and the Burmese returned with only a hundred thousand, many of whom were captives. Very few of these losses were battlefield casualties. Most were certainly deserters, and most of these, I imagine, ended up in the hills, to remain there or eventually to drift home to their valley villages. Since wars were manpower affairs, there was as much reason to flee the approach of your own army as to flee the approach of the enemy.
Secondly, consider the flight of those ruined by taxes and corvée in the wet-rice valleys, not just by the crown and its appropriations, but by the exactions of lower and subordinate officials and nobles. Think for a moment of the long-run movement away, serfs running to the hills as a kind of centrifugal force, as opposed to the centripetal movement due to slaving and the resettlement of war captives.
Thirdly, think of the factions, the losers, and factional fights in towns and villages, both among civilians and among the Buddhist clergy. This is perhaps not only a result of the different ecology of such areas and the relative absence of close clerical attention to these areas. It is also a result of the fact that a whole series of what one might call 'traditional organic intellectuals', hermit monks, pretenders, magicians, dissident Buddhist sects, in effect, excommunicated from the valley kingdoms tend to drift to the hills. People who, in one way or the other, are likely to make a symbolic case against the centre and are likely to end up in the hills.
Fourth, even groups that were seen as the original inhabitants of the hills, we increasingly find, may have moved to the hills from the lowlands. Take for example the Orang Asli, the so-called aboriginal inhabitants of Malaysia. Many of these aboriginal groups, not all, are as near as we can make out the descendants of riverain or coastal peoples who rejected Islam, for factional reasons perhaps. Perhaps because they did not want to give up their gods, or their pigs, or because they were afraid of slaving raids, the most common case.
Finally, even the Punan in Sarawak, often the poster-children for the protection of old-growth forests, were in fact a sort of lowland people who moved into the hills and are described better by their ecological niche and their specialization in gathering primary forest products for international commerce, rather than by their ethnic identity.
We must imagine that a large number of those who lost their place in lowland villages because of factional fights, crimes, and social ostracization, drifted to the hills as individuals, households, kinship groups, and whole villages. If you grant me that valley kingdoms, in the pressure they exerted, especially on the core region, were always grinding out people in a thin stream and often in larger streams during periods of crisis, then it is plausible to imagine that many of these people ended up in the hills. Remember, at the same time, that there are always hill peoples assimilating into the valley kingdoms. Particularly in settled times and under stable dynasties, this process would have been paramount.
The process of assimilation to valley culture would be running ahead, would be preponderant in times of political stability and economic expansion. On the other hand, the process of dispersion and flight would be running ahead at times of economic crisis, epidemics, famines, civil wars, and the press of corvée labour. And, assuming that this is a process that continues over long periods of time, if there is time to percolate, time to cook, then one can imagine the incorporative hill societies both remaining distinct hill peoples and yet over long periods of time, and not paradoxically, absorbing newcomers, escapees, Maroons, and refugees. Given this time to cook, why cannot one take a radically constructionist view of hill ethnicity and culture, making due allowances for the effects of hill ecology and dispersion, mobility, economic organization, autonomy, and political organization. The expansion and contraction of non-state space is, I begin to believe, a privileged window on the history of state-making. In other words, non-state space is related to the state in much the way Maroon settlements are related to slave plantations or the way in which the Wild West was related to the Civil War, in the United States. It is the social product of evasion, flight, dissent, heterodoxy, desertion, social, political, and economic ruin.
Much of the periphery if we see it from this angle, is something like a shadow society or better yet, a reversed mirror image: in terms of its ecology, its religious practices, its social structure, its governance, and above all its fugitive dissident population. Emphasizing this aspect of the frontier and hill peoples together with their distinctive ecology and economy, allows us, I think, to understand the symbolic relation between lowland and hill peoples in Southeast Asia.
Professor James Scott can be reached at Yale University, e-mail: James.Scott@yale.edu.