¹ Moore, Barrington Jr, The Social Basis of Dictatorship and Democracy, (Boston: Beacon Press, 1966), p. 505Google Scholar.

² For some indication of this, see, for example, Harry Benda, “Peasant Movements in Colonial Southeast Asia,” Asian.Studies, 3 (December 1965), 422-24; The Glass Palace Chronicle, translated by Luce, H. and Aung, Maung Htin, (Oxford: Humphrey Milford, 1923), pp. 133, 159, 177Google Scholar; and Hall, D.G. E., A History of Southeast Asia, 3rd Edition (London: Macmillan, 1970), p. 273Google Scholar.

³ For Central Luzon we have the advantage of primary research while, for other areas, we have relied upon secondary sources.

⁴ In Cochinchina and Lower Burma, the complete transformation of social relations was partly due to the fact that such areas of pioneer settlement lacked the core of traditional communities which might offer more resistance to the colonial model.

⁵ This is, of course, in addition to the use of ritual and-magic to ensure good crops and to protect the kin group and community.

⁶ Note particularly the use of fictive kinship as a way of extending the kinship principle beyond its biological boundaries and the use of kinship terms of address between patrons and clients.

⁷ The strength of these pressures, as George Foster has argued, derive in part from the prevalent and usually realistic assumption among peasants that the social product is fixed and that, therefore, the gain of any one family comes necessarily at the expense of other families in the community. George M. Foster, “Peasant Society and the Image of Limited Good,” American Anthropologist, 67 (April, 1965), 293-315.

⁸ On Viet Nam, see Long, Ngo Vinh, Before the August Revolution: The Living Conditions of the Vietnamese Peasants under the French, manuscript xerox, (forthcoming, MIT Press)Google Scholar, who cites official French figures on communal land as 20 percent of the cultivated surface in Tonkin, 25 percent in Annam, and a mere 3 percent in Cochinchina. Gourou, Pierre, L'Utilisation du Sol en Indochine, (Paris: Centre d'Etudes de Politique Etrangere, 1940), p. 276Google Scholar, gives comparable figures. According to Long (p. 6), communal lands consisted of “salary land, ” given t o soldiers as part of their salaries; “tax assistance land, ” to help the poor pay their taxes; “study land, ” used for paying teachers and supplying students with educational materials; and “widows and orphans land, ” for poor widows and homeless children.

⁹ There is an extensive, literature, mostly anthropological, dealing with patron-client bonds which we have relied on. Some of the most useful include: Foster, George M., “The Dyadic Contract in Tzintzuntzan: Patron-Client Relationship,” American Anthropologist, 65 (1963), 1280–1294CrossRefGoogle Scholar; Wolf, Eric, “Kinship, Friendship and Patron-Client Relations,” in Banton, M., ed., The Social Anthropology of Complex Societies (New York: Praeger, 1966)Google Scholar; Campbell, J., Honour, Family and Patronage (Oxford: Clarendon Press, 1964)Google Scholar; Powell, John Duncan, “Peasant Society and Clientelist Politics,” American Political Science Review, LXIV, 2 (06, 1970) pp. 411–425CrossRefGoogle Scholar; Lande, Carl, Leaders, Factions and Parties: The Structure of Philippine Politics, Monograph No. 6 (New Haven: Yale University-Southeast Asian Studies, 1964)Google Scholar; and Weingrod, Alex, “Patrons, Patronage, and Political Parties,” Comparative Studies in Society and History, 19 (07, 1968), 1142–1158Google Scholar. See also Scott's, James “Patron-Client Politics and Political Change in Southeast Asia,” American Political Science Review, 66 (03, 1972), 91–113CrossRefGoogle Scholar. The central features of the patron-client tie serve to distinguish it from other vertical dyadic ties with which it is often confused. It differs in several important respects, for example, from the link joining the “cacique, ” the bandit leader, or a local “boss” to their “men.” While such power figures are also personal leaders with private followings, they generally are nouveaux arrives with little claim to higher status, their role is less culturally sanctioned - than the role of a patron, and their relationship to their men is less diffuse and relies more heavily on coercion and/or material rewards. Patron-client reciprocity must also be distinguished from the kinds of exchange that normally occur between, say, moneylenders and borrowers, officials and citizens, employers and employees. As the functionally specific role categories suggest, such exchange is typically restricted to a single category of reciprocity, it is less durable over time, and the terms of the exchange are governed in large part by impersonal regulations and legal contracts.

¹⁰ In Thailand and Burma, locally based aristocrats were each entitled to the service of a certain number of peasant households in arrangements of familial clientage that often lasted several generations. Appanage systems developed by the prijaji in Java amounted to much the same thing. Cf. Rabibhadana, Akin, The Organization of Thai Society in the Early Bangkok Period 1782-1873 (Ithaca: Cornell Southeast Asia Program, 1969)Google Scholar.

¹¹ In Cochinchina, the duties foisted on the head of the council of notables (thon-truong) were so onerous and expensive that it became increasingly difficult to find prominent villagers who would agree to serve. While the post lost prestige, as it involved an increasing tax and administrative burden on villagers, the prospects for personal profit seemed rather limited (at least until 1910) as compared with, for example, the higher post of canton chief (tong). Cf. Osborne, Milton E., The French Presence in Cochinchina and Cambodia: Rule and Response (1859-1905), (Ithaca: Cornell University Press, 1969)Google Scholar.

¹² See, for example, Benda, Harry J., “The Structure of Southeast Asian History: Some Preliminary Observations,” Journal of Southeast Asian History, 3 (03, 1962), 126CrossRefGoogle Scholar; and Kartodirdjo, Sartono, The Peasants' Revolt of Banten in 1888 ('S-Gravenhag Martinus Nijhoff, 1966), Ch. 3Google Scholar.

¹³ See Long, Ngo Vinh, Before the August Revolution, 82–95Google Scholar; Sansom, Robert L., The Economics of Insurgency in the Mekong Delta of Vietnam (Cambridge, Mass.: M.I.T. Press, 1970), pp. 37–39Google Scholar; and Furnivall, J.S., Colonial Policy and Practice, (New York: N.Y. University Press, 1960), pp. 103, 192–194Google Scholar.

¹⁴ During the late nineteenth and early twentieth century, there was a substantial flow of both seasonal and permanent migration from the Dry Zone of Burma into the Delta. When the rains occasionally failed in the Dry Zone, the migration became a torrent. Cf. Michael Adas's fine Ph.D. dissertation, Agrarian Development and the Plural Society in Lower Burma, (University of Wisconsin, 1971), Ch. 8Google Scholar. (Forthcoming University of Wisconsin Press). It appears that Cochinchina did not, however, substantially relieve the pressure on the land in Tonkin. Although there was an annual official migration under labor contracts of roughly 15, 000 (1919-1934) northerners (largely from the overpopulated provinces of Nam Dinh, Thai Binh, Ninh Binh, and Hai Duon g in Tonkin), the annual return rate left only a mean annual permanent migration rate of 3, 250. Unofficial immigration is hard to estimate but Robequain puts it at less than the official figure. He explains “… la maigreur des courants de migration annamites vers le Sud …” by the incidence of malaria in Cochinchina and the more protective (though poorer) social structure in the north. Robequain, Charles, L'evolution Economique de I'lndochine Franqaise, (Paris: Centre d'fitudes de Politique Etrangere, 1939), pp. 66–76Google Scholar.

¹⁵ Sansom, The Economics of Insurgency …, Chs. 2, 3; and Ngo Vinh Long, Before the August Revolution … passim.

¹⁶ Kerkvliet, Ben, Peasant Rebellion in the Philippines: The Origins and Growth of the HMB, (Ph.D. dissertation, University of Wisconsin, 1972), Chs. 1-2Google Scholar.

¹⁷ Adas, Michael, Agrarian Development …, 411Google Scholar.

¹⁸ Such categories are hardly mutually exclusive; a peasant might simultaneously be a smallholder and tenant or a tenant and wage laborer. Wage laborers might turn to petty trade or banditry during slack times.

¹⁹ The growth of private coercive force was conditioned both by the availability of public force to protect landed interests and the nature of the threat. In Central Luzon the constabulary was virtually at the disposal of the landed elite although local protection required an additional force of irregulars. See the discussion below on “peasant protest.” In Lower Burma the police seemed a less effective guardian of landed interests, but Indian irregulars could be mustered against recalcitrant Burman cultivators.

²⁰ Centers, Richard, The Psychology of Social Classes, (New York: Russell and Russell, 1949), Ch. 7Google Scholar, provides strong evidence for this argument. We might add here that since our use of “class consciousness” refers to the subjective dimension of class, it differs from Marx's use of class consciousness (class Jiir sich) to include class mobilization. Because the question of mobilization is so contingent on the coercion that can be brought against a class, it is reasonable to keep the question of awareness and allegiance separate from organized activity.

²¹ Organization, of course, influences consciousness as well; as we shall examine below. It is possible, nonetheless, for a substantial awareness of class identity and class allegiance to exist in the absence of much formal organization or activity. The sans-culottes of late 18th century Paris are a good example.

²² The situation in Lower Burma is interesting here since it produced a form of Burman inter-class solidarity against Indian laborers and moneylenders. Burmese landowners and moneylenders were going under to the Indian financiers who sat atop the Delta's financial structure while Burman laborers and tenants competed for tenancies and work with unemployed Indians moving out from the stricken urban economy. The anti-Indian nationalist riots of the 1930's expressed, for the peasantry, a class issue while, at the sametime, it singled out the Indian sector of the economy for attack.

²³ A careful assessment of such tendencies in the developing factory system may be found in Chapter 2 of Cole, G. D. H., Studies in Class Structure (London: Routledge and Kegan Paul, 1955)Google Scholar.

²⁴ In terms of his contributio n to the enterprise, the factory owner had a somewhat greater claim to legitimacy than the landowner since he had, in a sense, created the factory while the landlord most often simply monopolized a resource which he had neither created nor unproved. Underlying agrarian struggles in many different areas and times is the recurrent theme that the cultivator is entitled to the crop; any other claim must be based on real services to the cultivator and never simply legal title to the land.

²⁵ The term and its dynamics are from Geertz's, Clifford, Agricultural Involution (Berkeley: University of California, 1963)Google Scholar.

²⁶ Pitt-Rivers, J. A., The People of the Sierra, (Chicago: University of Chicago Press, Phoenix Books, 1961)Google Scholar.

²⁷ Hill, Frances R., “Millenarian Machines in South Vietnam,” Comparative Studies in Society and History, (10, 1971), 328Google Scholar.

²⁸ Maclntyre, Alasdiar, Secularization and Moral Change (London: Oxford University Press, 1967), pp. 42–43Google Scholar.

²⁹ The 19th century English working class analogues to rural Southeast Asia, in this respect, were the traditional monoclass weaving villages, which provided a structure of mutuality that worked to redistribute poverty much as the strongly sanctioned reciprocity in parts of village Southeast Asia.

³⁰ New associational activity among villagers seems, as one would expect, to have been more typical in recently settled villages or in villages economically dominated by elites who no longer served peasant subsistenc e needs (e.g., Lower Burma and th e Mekong Delta) than in older, more socially homogeneous villages (e.g., Burma's Dry Zone and Tonkin).

³¹ As in the West, there is great continuit y between local and seemingly apolitical self-help efforts and later political initiatives. Perhap s tracing the local lineage of peasan t unions and radical political parties would'revea l a similar continuit y with earlier self-help activities. For one striking case from England, see E. J. Hobsbawm and George Rude's discussion of the relationship between annual village feasts (and other rituals) an d the organization of rural protest movements. Captain Swing (New York: Pantheon, 1968), pp. 66–68Google Scholar.

³² Sansom, Robert, The Economics of Insurgency …, 39, 100-01Google Scholar.

³³ For the Samin movement, see Benda, Harry J. and Castles, Lance, “The Samin Movement,” Bijdragen tot de Taal-, land-, en Volkenkunde, 125, no. 2 (1969), 207–40CrossRefGoogle Scholar.

³⁴ The events of late 1965 showed that this adaptation had its demographic limits and that explosive tensions had developed in East and Central Java.

³⁵ Cf. Kartodirdjo, Sartono, The Peasants' Revolt of Banten in 1888 (S-Gravenhauge Martinus Nijhoff, 1966), Chs. 3-5, andGoogle ScholarMarr, David, Vietnamese Anti-Colonialism: 1885-1925 (Berkeley, University of California Press, 1971), pp. 61–76Google Scholar.

³⁶ Party leaders are thus often called upon to settle purely personal problems of their clients and to operate, in general, as diffuse patrons. Many villagers approached UMNO politicians in Malaysia for personal loans, help in arranging marriages, and settlement of personal disputes. In the Philippines a common complaint amon g politicians is that they must spend too much time listening to people's requests to solve personal problems.

³⁷ Wertheim, W.F., “From Aliran to Class Struggle in the Countryside of Java,” Pacific Viewpoints, 10 (09, 1969), 1–17Google Scholar.

³⁸ Roumasset, James, “Risk and Choice of Technique for Peasant Agricaltare,” (Social Systems Research Institute, University of Wisconsin, 1971)Google Scholar, demonstrates how peasant culti vators make decisions so as to avoid running the risk of falling below a certain food production level. The analysis seems more generally applicable to the social arrangements peasants must make for subsistence and security as well.

³⁹ Frances Hill, “Millenarian Machines …, ” 338-342. As we have noted, at the local level both the Cao Dai and Hoa Hao were partly sustained by self-help. See also Nguyen Tran Huan, “Histoire d'one secte religievse av Vietnam! Le Caodaisme, Ch. 7, Chesneaux, Jean, Tradition et Revolution av Vietnam, (Paris: Editions Anthropos, 1965)Google Scholar.

⁴⁰ For that reason one would expect such links to be more common in “open” villages withmany connections to the outside and in areas where elections increase the value of a following.

⁴¹ Peasants are similarly favored by the disappearance of national or provincial regimes which have, in the past, provided landlords with coercive power. The Japanese occupation, “for example was a boon to many Southeast Asian peasants as landlords fled the insecure countryside. The share of the harvest claimed as rent, declined in Central Luzon, Lower Burma, and the Mekong Delta during this period. In India, more recently, the election of left-wing state governments has meant the neutralization of the police power that had previously backed landowners. As a consequence, large owners, sensing a need for local good-will, have been more lenient to tenants and have been cautious abou t mechanization.

⁴² We do not mean to imply that electoral patronage alone alleviated the agrarian distress of the 1930's. In Central Luzon, the common use of the constabulary and private armies by landlords against peasants indicates the failure of electoral forms to address basic structural problems.

⁴³ There is strong evidence, however, that this elite route to agrarian peace has its limits. Its methods are most effective, as are those of city machines, with the most insecure sector of the subordinate population - in this case, marginal tenants, day laborers, and the agrarian lumpen elements. As in the city also, the groups most likely to resist such blandishments (since they have a margin of security and the habit of independent action) and most likely to initially voice class-based demands are the more secure lower-class elements such as smallholders. In effect the financial cost of agrarian peace grows steadily greater as structural problems remain untouched. A glance at the relationship between net government receipts and elections over the past 20 years reveals a sharp upward trend in electoral disbursements and increasingly shallow “recoveries in off years.” ( Averich, H.A., Denton, F.H., and Koehler, J.E., A Crisis of Ambiguity: Political and Economic Development in the Philippines, (Santa Monica, California: The Rand Corporation, R-473-AID, 1970), p. 162CrossRefGoogle Scholar. A sudden slump in government revenue would clearly jeopardize the tenuous rural peace that virtuoso government spending (and a growing tempo of coercion) has made possible.

⁴⁴ Similar features were in evidence in earlier periods of PKI history - for example, prior to the 1926-27 revolt. The appeal of the Persatuan Kaum Tani (a PKI linked peasant organization) in West Sumatra was at that time due in large part to its emphasis on “beneficial aims as mutual aid with regard to labor, funerals, celebrations, fire-fighting, the prospect of loans from the Association's chest for taxes or in lieu of forced labor formed another attraction in that propaganda.” “The course of the Communist Movement on the West Coast of Sumatra,” Part I (Political Section), Report of the Investigation Committee appointed under the Government Decree of February 13, 1927” in Benda, Harry J. and McVey, Ruth T., eds., The Community Uprisings of 1926-27 in Indonesia: Key Documents, Modern Indonesian Series, Translation Series (Ithaca, N.Y.: Cornell Southeast Asia Program, 1960), p. 146Google Scholar. There is additional evidence that such functions were, in effect, guaranteed by a single leader whose departure meant the disintegration of the local organization. Ibid., 146.

⁴⁵ Mortimer, Rex, “Class, Social Cleavage, and Indonesian Communism,” Indonesia, 8 (10, 1969), 6Google Scholar. We have also relied on the following sources for this account. Hindley, Donald, The Communist Party of Indonesia 1951-1963, (Berkeley: University of California Press, 1966)Google Scholar; Jay, Robert, Religion and Politics in Rural Central Java, Cultural Report Series, (New Haven: Yale University Southeast Asian Studies, 1963)Google Scholar, and W. F. Wertheim, “From Aliran to Class Struggle … ”

⁴⁶ The party did, of course, make efforts to create new horizontal links between peasants at the local level. It organized labor exchanges, work brigades, rotating credit associations, cooperative public works (schools, well, road repair), and producer cooperatives. It is unclear, however, whether these initiatives were largely dependent on continuous assistance and leadership from PKI cadre, or if they managed to become genuinely autonomous local entities independent of stimulation from above.

⁴⁷ , Hindley, The Communist Party …, 163Google Scholar.

⁴⁸ The PKI is not unique. The Communist Party of Italy, also operating under electoral pressures where peasant votes are important, has tended to reflect rather than change the clien-telist structure of rural Southern Italy. See the fine study of the PCI by Tarrow, Sidney, Peasant Communism in Southern Italy (New Haven: Yale University Press, 1967)Google Scholar.

⁴⁹ Drawing on Carl Lande's systematic comparisons between the behavior of categorical and dyadic groups a few main distinguishing characteristics can be isolated. “Groups and Networks in Southeast Asia, ” American Political Science Review, (forthcoming). (1) A class-based organization's membership will be more affected by policy shifts than by personnel shifts. A patron-client network's membership (within a formal organization) will be more affected by individual leaders joining or leaving than by policy shifts.

(2) A class based organization's membership will be more influenced by how effective the organization is in achieving collective goals than by the concrete rewards and opportunities it controls. A patron-client network's membership is more affected by the gain or loss of tangible rewards than by general policy achievements.

(3) Cleavages within class based organizations most often concern questions of policy and tactics while cleavages in patron-client organizations most often arise over issues of personnel and the distribution of control over the organization's resources.

⁵⁰ Bailey, F.G., “The Peasant View of the Bad Life,” Advancement of Science, 23 no. 114 (1966), 401Google Scholar. Fragments in brackets and emphasis added.

⁵¹ Cf. Tilly, Charles, “Food Supply and Public Order in Modern Europe,” manuscript, (03, 1971)Google Scholar. For other examples of collective bargaining by riot, see Rimlinger, Gaston V., “The Legitimation of Protest: A Comparative Study in Labor History,” Comparative Studies in Society and History, 2 (04, 1960), 329–343CrossRefGoogle Scholar. Chi the Luddites as another example of an institutionalized tradition of worker sabotage, see Thompson, E. P., The Making of the English Working Class, (New York: Vintage, 1966), pp. 521–602Google Scholar.

⁵² Hobsbawm an d Rude, Captain Swing, Introduction an d Chapter 1.

⁵³ Sansom, The Economics of Insurgency…, 60-61.

⁵⁴ 4 Akira Takahashi, Land and Peasants in Central Luzon, (Tokyo: Institute of Developing Economies, 1969), p. 77.

⁵⁵ 5 Traditional utopian-millenial movements, by contrast, might be expected t o develop where sudden overwhelming hardship s are thrust upon the peasantry, where the cause of these hardships is either very distant socially and very powerful (e.g., colonial regimes) or is simply unassailable (e.g., flood, disease), an d where beliefs in magic and divine agency ar e strong. One might even say that millenial movements bear the sam e relationship to protests aimed at restoring patronage customs in the traditional context as revolutionary movements bear to collective bargainin g strikes in th e moder n context. The former in each case strive for a thoroug h remaking of societ y and one outside the range of ‘normal’ politics. The latter are often conservative movements tha t accept given political forms and seek incremental changes.

⁵⁶ For an analysis of the peasant protest movement and rebellion in Centra l Luzon, see Kerkvliet, Peasant Rebellion in the Philippines …; also, his “Peasan t Society and Unres t Prior to the Huk Revolution in the Philippines,” Asian Studies (Quezon City), 9 (Aug. 1971), 164-213.

⁵⁷ Indeed these themes are evident in numerous peasant movements and rebellions elsewhere. Compare, for example, the peasant movements in Latin America analyzed in Lands-berger, Henry A. (ed.), Latin American Peasant Movements (Ithaca: Cornell University Press, 1969)Google Scholar; Tilly, The Vendee (especially Chapter 13); and Hobsbawn and Rudé, Captain Swing (especially Chapters 4 and 10).

⁵⁸ Marx, Karl, The 18th Brumaire of Louis Bonaparte, (New York: International Publishers, 1962), pp. 123–124Google Scholar. The fact that Marx was dealing here with a peasantry which lived in separate homesteads on the land rather than in villages may make his argument less applicable to the Russian mir or Southeast Asian village.

⁵⁹ This is one reason why conservative parties are often more socially heterogeneous than radical parties. When patron-client ties are no longer of much value to members of the community, the class of patrons and even the system of patronage itself may lose legitimacy. As Pitt-Rivers observes for Spain, “the system is, clearly, only to be judged good insofar as it ensures that people do not go hungry, that injustice is not done. Where the majority of the community can look to the patron in time of need, such a system reinforces the integration of the pueblo as a whole. Where those who enjoy the advantages of patronage are a minority, then they and their patrons are likely to be resented by the remainder.” (People of the Sierra, 204). Wertheim, “From Aliran to Class Struggle …,” 23, makes much the same argument about a declining patronage sector and a growing non-patronage sector in explaining santri-abangan violence after 1965. “… the followers of the rich santri, many of them poor sharecroppers, will have obeyed their patron rather than acting on a basis of class solidarity.”

⁶⁰ The traditional peasant subculture is not by any, means entirely village bound. Inter-village links through marriage patterns, trade, and religion can serve political ends too. Ethnic cleavages are another and more serious matter. Such cleavages were not important in Central Luzon though they were in lower Burma. The success of the Viet Minh among the Tho and the structure of the Pathet Lao may indicate that ethnicity can be recognized and accommodated within a revolutionary coalition.