Write a 5-7 page paper (double spaced, 12 point font, standard margins) on one of the following topics. Your title page and bibliography (and any other pages that are not writing) do not count towards

Production and Social Labor In formulating the mode of production concept, Marx began with two axiomatic understandings of the human condition. Both are also axioms of modern anthropology. The first sees the species Homo sapiens as a part of nature; the second defines Homo as a social species, its individual members always linked to others in social relationships. The human species is an outgrowth of natural processes; at the same time, the species is naturally social. The human species is, however. not merely a passive product of natural processes; it has also, in the course of evolution, acquired the ability to transform nature to human use. If humanity stands to nature as part to whole, then that part has acquired the ability to oppose the whole that encompasses it; or, as Marx phrased it, man "confronts the material of nature as one of her own forces .... [By] changing it, he at the same time changes his own nature" (quoted in Schmidt 1971: 77 -78). This active relation of the species to nature, while rooted in biological characteristics, is put into effect by the exosomatic means of technology, organization, and ideas. Man rises up against nature by means of what we would today call culture. Marx's second axiom emphasizes the sociality of humankind. Human beings exist in organized pluralities. Moreover, the way they are orga­ nized socially governs the way they confront and transform nature, and 74 CONNECTIONS nature thus transformed affects, in turn, the architecture of human social bonds. In Marx's words, "the restricted relation of men to nature determines their restricted relation to one another, and their restricted relation to one another determines men's restricted relation to nature" (quoted in Colletti 1973: 228).

Is there a concept that allows us to grasp this complex connection between a socially interrelated humanity and nature? Marx found such a concept in his notion of labor. Humankind adapts to nature and transforms it for its own use through labor. Thus, "the labor process .. .

is the general condition for the metabolism between men and nature; it is the everlasting nature-imposed condition of human existence" (quoted in Schmidt 1971: 136). Yet labor is always social, for it is always mobilized and deployed by an organized social plurality. Marx therefore drew a distinction between work and labor. Work represents the activi­ ties of individuals, singly or in groups, expending energy to produce energy. But labor and the labor process was for him a social phenome­ non, carried on by human beings bonded to one another in society.

This concept of labor as a social process carried on by an organized plurality could not be imagined as long as different kinds of work­ cultivating, spinning, praying-were thought of as qualitatively differ­ ent. Only when different kinds of work could be subsumed under the common denominator of money did "labor -in-general" become con­ ceivable. Marx credited Adam Smith with the first formulation of this concept, noting that this "immense step forward" occurred preCisely when different kinds of labor had become interchangeable (Gr . 1973: 104), that is, after the onset of capitalism . The utility of the concept, however, transcends its particular historical origins. Once one can talk about labor-in-general, one can begin to visualize how any organized human society activates this process and shares out its products . Understanding how humans transform nature to their use thus does not stop with the description and analysis of techno-environmental interaction. The laborer, the direct producer, is never an isolated Robin­ son Crusoe but is someone who always stands in relationship to others, as kinsman, serf, slave, or wage laborer. Similarly, the controllers of social labor are not to be thought of as technicians who guide the technical operations of work. They are assigned to their positions by the system of deploying social labor, which casts them in the role of elder kinsman, chief, seignoriallord, or capitalist. It is this conception of social mobilization, deployment, and allocation of labor that allows us to understand how the technical transformation of nature is conjoined with the organization of human sociality. Marx adopted the term production for this complex set of mutually dependent relations among nature, work, social labor, and social orga­ nization. We shall use the term in this sense in the present work.

Because modern usage often restricts it exclUSively to technology, it is important to be aware of the background that evoked it. The concept of Modes of Production 75 production was employed by Marx to contrast his perspective with Hegel's conception of Geist ("Spirit"). It thus carries overtones of Marx's confrontation with Hegelian idealism. For Hegel the various human transformations of nature represented successive concretizations of Spirit or Mind ("models of" and "models for"). Marx's use of production also contrasts with Feuerbach's contemplative materialism. Feuerbach had criticized Hegel for treating thought as transcendental rather than as an attribute of natural humankind. Yet he took account neither of human sociality nor of the human confrontation with nature. Marx, in contrast, stressed the activity of socially organized humankind in a double sense-active in changing nature, and in creating and re-creat­ ing the social ties that effect the transformation of the environment. The term production expressed for him both this active engagement with nature and the concomitant "reproduction" of social ties. It is also important to note that Marx's concept of production incorpo­ rates his insistence that the human species produces with both hand and head. In contrast to other animals, humans conceptualize and plan the labor process. Labor thus presupposes intentionality, and therefore information and meaning. Just as labor is always social labor, informa­ tion and meaning are always social. As Marx put it, thought does not descend from on high into the real world; thought and language "are only manifestations of actual life" (quoted in Coletti 1973: 225). Social labor with both hand and head is deployed to cope with nature; the deployment of social labor, in turn, reproduces both the material and the ideational ties of human sociality. Modes of Production The concept of social labor thus makes it possible to conceptualize the major ways in which human beings organize their production. Each major way of doing so constitutes a mode of production-a specific, historically occurring set of social relations through which labor is deployed to wrest energy from nature by means of tools, skills, organi­ zation, and knowledge. What modes of deploying social labor are there? Marx himself spoke of a number of different modes: an original, primitive, communitarian mode, conceived after Morgan's model of primitive communism; the slaveholding mode of classical European antiquity; a Germanic mode, supposedly characteristic of the Germanic peoples in their early migra­ tions; a Slavonic mode, said to characterize the early Slavs; a peasant mode; a feudal mode; an Asiatic mode; and a capitalist mode. Not all of these are based on equivalent criteria. Some may never have consti­ tuted primary modes in their own right, but may have been only accessory or supplementary modes; others represent extrapolations from historical interpretations now adjudged to have been erroneous. For the purposes of this book, it is immaterial whether Marx was right 76 CONNECTIONS or wrong-whether he should have postulated two or eight or fifteen modes of production, or whether other modes should be substituted for those suggested by him. The utility of the concept does not lie in classification but in its capacity to underline the strategic relationships involved in the deployment of social labor by organized human plurali­ ties. Since we want to deal with the spread of the capitalist mode and its impact on world areas where social labor was allocated differently. we shall construct only those modes that permit us to exhibit this encounter in the most parsimonious manner. For this purpose we shall define but three: a capitalist mode, a tributary mode, and a kin-ordered mode. No argument is presented here to the effect that this trinity exhausts all the possibilities. For other problems and issues it may be useful to construct other modes drawing further distinctions, or to group together differ­ ently the distinctions drawn here. Nor is there any intention, in the present context, to argue that these three modes represent an evolutionary sequence. While we shall ex­ plore certain historical relations between modes, it is a major argument of this book that most of the societies studied by anthropologists are an outgrowth of the expansion of Europe and not the pristine precipitates of past evolutionary stages. This position extends the caveats already introduced by other writers against the uncritical equation of the bands, tribes, or chiefdoms described by observers since 1400 with the societies existing before European expansion and even before the rise ofthe state (Service 1968: 167; Fried 1966, 1975). Fried has stated resolutely that the "tribe" is "a secondary sociopolitical phenomenon, brought about by the intercession of more complex ordered societies, states in particu- 1ar" (1975: 114). I believe that all human societies of which we have record are "secondary," indeed often tertiary, quaternary, or cen­ tenary. Cultural change or cultural evolution does not operate on iso­ lated societies but always on interconnected systems in which societies are variously linked within wider "social fields." One of the utilities of the concept of mode of production lies precisely in that it allows us to visualize intersystemic as well as intrasystemic relationships. We shall use the concept to reveal the changing ways in which one mode, capitalism, interacted with other modes to achieve its present domi­ nance. In this process Iroquois, Asante, Tamil, and Chinese are as much participants as Barbadians, New Englanders, and Poles. The process linked victims and beneficiaries, contenders and collaborators. The three modes that we employ should not be taken as schemes for pigeonholing societies. The two concepts-mode of production and society-pertain to different levels of abstraction. The concept of society takes its departure from real or imputed interactions among people . The concept of mode of production aims, rather, at revealing the political­ economic relationships that underlie, orient, and constrain interaction. Such key relationships may characterize only a part of the total range of interactions in a society; they may comprehend all of a society; or they Modes of Production 77 may transcend particular, historically constituted systems of social interaction. Used comparatively, the concept of mode of production calls attention to major variations in political-economic arrangements and allows us to visualize their effect. The use of the concept enables us, above all, to inquire into what happens in the encounters of differently constituted systems of interaction-societies-predicated upon differ­ ent modes of production. We shall begin our exposition with the capitalist mode, despite the fact that it developed later than the others, in the course of the eigh­ teenth century. It was in the analysis of this mode that Marx developed his general concepts, and we follow him in his conviction that an understanding of how this mode works provides the key to the under­ standing of others. The Capitalist Mode Marx spent most of his life on the analysis of the capitalist mode of production. He did so, of course, to understand it in such a way that he could help put an end to it. What, according to him, were its salient characteristics? For Marx, the capitalist mode came into being when monetary wealth was enabled to buy labor power. This specific capability is not an inherent attribute of wealth as such; it develops historically and re­ quires the installation of certain prerequisites. Labor power is not in itself a commodity created in order to be offered for sale in a market. It is an attribute of human beings, a capability of Homo sapiens . As long as people can lay their hands on the means of production (tools, resources, land) and use these to supply their own sustenance-under whatever social arrangements-there is no compelling reason for them to sell their capacity to work to someone else. For labor power to be offered for sale, the tie between producers and the means of production has to be severed for good. Thus, holders of wealth must be able to acquire the means of production and deny access, except on their own terms, to all who want to operate them. Conversely, people who are denied access to the means of production must come to those who now control the means and bargain for permission to operate them. In return they receive wages that will allow them to pay for what they need to sustain themselves. Indeed, in the capitalist mode production determines distribution . Those who detain the means of production can also detain the com­ modities produced. Those who produce the commodities must buy them back from the owners of the means of production. Means of production, in turn, circulate only among those with capital to acquire them. Those who lack capital and must sell their labor power also lack the means of production. Hence, the way in which the mode commits social labor to the transformation of nature also governs the way the 78 CONNECTIONS resources used and obtained are distributed among producers and non­ producers. Streams of resources, including income, are not-as an ecologically oriented anthropologist wrote recently (Love 1977: 32)­ the human analogue of the way biological organisms capture energy. Between people and resources stand the strategic relationships govern­ ing the mode of allocating social labor to nature. The holders of wealth who now detain the means of production, however. would have no reason to hire laborers if they produced only enough to cover the costs of their wage package. In the course of a working day, the laborers in fact produce more than the cost of their wages; they produce a surplus. This surplus, under the conditions of the capitalist mode, belongs to the holder of wealth, the capitalist, whose means of production the workers put into operation. The greater this surplus, the greater the rate of profit obtained by the capitalist when he measures it against his outlays for plant, resources, and labor. There are two ways in which capitalists can increase this surplus. One way is to keep wages low, or to reduce them to the lowest possible point that is biologically or socially feasible. The other way is to raise the level of the surplus produced, above and beyond the amount that has to be paid for labor power. through raising the output of workers during any given period of work. Such increases in productivity require improve­ ments in the technology and organization of production. These impera­ tives produce relentless pressures, spurring capitalists to ever-increased accumulation of capital and renewal of technology. The greater the capital at their command , the greater their ability to raise technological productivity; hence the greater their ability to accumulate additional surplus to further expand production, as well as to outproduce and undersell competitors who fail to invest in new technology and who attempt to meet competition through placing greater burdens on their laborers.

The capitalist mode thus shows three intertwined characteristics.

First, capitalists detain control of the means of production. Second, laborers are denied independent access to means of production and must sell their labor power to the capitalists. Third, the maximization of surplus produced by the laborers with the means of production owned by the capitalists entails "ceaseless accumulation accompanied by changes in methods of production" (Sweezy 1942: 94; Mandel 1978: 103-107).

These characteristics, however, must be understood not only syn­ chronically but also historically, as developing facets of a mode that had determinate origins in time and that develops over time. The point is crucial. Wealth in the hands of holders of wealth is not capital until it controls means of production, buys labor power, and puts it to work, continuously expanding surpluses by intensifying productivity through an ever-rising curve of technological inputs. To this end capitalism must lay hold of production, must invade the productive process and cease- Modes of Production 79 lessly alter the conditions of production themselves. As long as wealth remains external to the process of production, merely skimming off the products of the primary producers and making profits by selling them, that wealth is not capital. It may be wealth obtained and engrossed by overlords or merchants, but it has not yet entered what Marx called "the really revolutionary road" of appropriating and transforming the means of production themselves (Cap. III. 1967: 334). Only where wealth has laid hold of the conditions of production in the ways specified can we speak of the existence or dominance of a capitalist mode. There is no such thing as mercantile or merchant capitalism, therefore. There is only mercantile wealth. Capitalism, to be capitalism, must be capital­ ism-in-production. The capitalist mode of production, so conceived, is necessarily based on a division of classes. It initiates a division between segments of the population who produce surpluses and segments of the population who control the means of production, and it continuously re-creates that differentiation. At the same time, it differentiates each class internally. In the race for higher productivity, the owners of the means of produc­ tion are differentiated into victors and losers. In the continuous move­ ment between the genesis of new sources of surplus production and renewed recession, the labor force shuttles among full employment, underemployment, and unemployment. The two processes of differen­ tiation are in fact linked, as the shareholders in capital are continuously driven to seek new pools of cheap and tractable labor, or else to replace costly or intractable labor with machines. The growth of this capitalism-in-production is a historical. develop­ mental process, originating in certain areas of the European peninsula. It expanded from there to envelop areas beyond Europe. It grew through its own internal ability to reproduce itself on an ever-widening scale; it grew also by entering into working arrangements with other modes, siphoning off wealth and people and turning them into capital and labor power. The capitalist mode thus always exhibited a dual character: an ability to develop internally and branch out, implanting its strategic nexus of relations across the face of the globe; and an ability to enter into temporary and shifting relations of symbiosis and competi­ tion with other modes. These relations with other modes constitute part of its history and development. Indeed, as we shall see, the internal dynamic of the capitalist mode may predispose it to external expansion, and hence to interchanges with modes other than itself. The Tributary Mode In the world in 1400 the major agricultural areas traversed by our imaginary traveler were held by states based on the extraction of sur­ pluses from the primary producers by political or military rulers. These states represent a mode of production in which the primary producer, 80 CONNECTIONS whether cultivator or herdsman, is allowed access to the means of production, while tribute is exacted from him by political or military means. Marx characterized the key attributes of this mode as follows: It is furthermore evident that in all forms in which the direct laborer remains the "possessor" of the means of production and labor conditions necessary for the production of his own means of subsistence, the property relationship must simultaneously appear as a direct relation of lordship and servitude, so that the direct producer is not free; a lack of freedom which may be reduced from serfdom with enforced labor to a mere tributary relationship. The direct producer, according to our assumption, is to be found here in possession of his own means of production, the necessary material labor conditions required for the realization of his labor and the production of his own means of subsistence. He conducts his agricultural activity and the rural home industries connected with it independently .... Under such conditions the surplus-labor for the nomi­ nal owner of the land can only be extorted from them by other than economic pressure, whatever the form assumed may be. [Cap. III, 1967: 790-791] In other words, social labor is, under these conditions, mobilized and committed to the transformation of nature primarily through the exer­ cise of power and domination-through a political process. Hence, the deployment of social labor is, in this mode, a function of the locus of political power; it will differ as this locus shifts position. It is possible to envisage two polar situations: one in which power is concentrated strongly in the hands of a ruling elite standing at the apex of the power system; and another in which power is held largely by local overlords and the rule at the apex is fragile and weak. These two situations define a continuum of power distributions.

A ruling elite of surplus takers standing at the apex of the power system will be strongest when it controls , first, some strategic element in the process of production, such as waterworks (Wittfogel 1931), and second, some strategic element of coercion, such as a standing army of superior military capability. Rulers will then be able to deploy their own tribute gatherers without need of assistance from local power holders. They will be able to loosen the grip of local overlords over resources and hence over the primary producers of surplus, and render the overlords dependent on revenues tendered by the rulers. If the rulers are success­ ful in this, they can also induce the local overlords to fight among themselves for privileged positions at the source of revenue. Such rulers will also be able to curtail the power of traders, keeping them from access to the primary producers in the countryside and preventing them from financing potentially rebellious overlords on their own behalf.

Finally, such a strong central power will be able to place limits on translocal "grass-roots" organization, be they gUilds, estates, leagues, or religious sects. At the same time, strong central rule often finds support Modes of Production 81 among surplus-producing peasantries, since central rulers and peasants are linked by a common antagonism against power-holding and sur­ plus-taking intermediaries.

Conversely, the central power will be weak and local power holders strong where strategic elements of production as well as means of coercion are in the hands of local surplus takers. Under such conditions local figures can intercept the flow of tribute to the center, strengthen their grip over land and the population working it, and enter into local or regional alliances on their own. Such local alliances, however , are frequently directed not only against the center but also against members of their own class, with the result that factional struggles will ramify throughout the countryside, thus weakening their class position. Fac­ tional struggles, in turn, may allow the elite at the center to survive by stratagems of divide-and-rule. Paradoxically, internecine faction fights also weaken the position of the primary producers, since in the absence of strong central control they must seek protectors against unrest and predation . In broad terms, the two situations we have depicted correspond to the Marxian concepts of the" Asiatic mode of production" and the "feudal mode of production." These are usually treated as enduring and un­ changing opposites. One term is usually ascribed to Europe , the other to Asia. The preceding exposition should make clear, however, that we are dealing rather with variable outcomes of the competition between classes of nonproducers for power at the top. To the extent that these variable outcomes are all anchored in mechanisms exerting "other than economic pressure," they exhibit a family resemblance to each other (Vasiliev and Stuchevskii 1967; Tapfer 1967). This resemblance is best covered by a common term for this mode-tributary mode of production­ used by Samir Amin (1973b) .

Reification of "feudalism" into a separate mode of production merely converts a short period of European history into a type case against which all other "feudal-like" phenomena must be measured. The concept of the Asiatic mode of production, in which a centralized state bureaucracy dominates unchanging village communities of hapless peasants, Similarly suffers from an ahistorical and ideological reading of Asian history. It has long been customary in the West to counterpose Western freedom with Eastern despotism, whether this was done by Herodotus with reference to the Greek city-states in their struggles with Persia, or by Montaigne and Voltaire counterposing societies based on the social contract with societies characterized by multitudes groveling under despotic rule. Our portrayal should permit us, rather, to specify the politically relevant variables that distinguish one tributary situation from another . Thus China, with a strongly concentrated hydraulic component, clearly represents a set of tributary relationships different from those in India, with its reliance on dispersed "tank" irrigation, or in Iran , with its irrigation by means of underground wells and canals. 82 CONNECTIONS Moreover, strongly centralized "Asiatic" states frequently break down into political oligopolies resembling feudalism; and more feudal, dis­ persed controls by local power holders yield to more centralized and concentrated power over time. To reify the weak phases of the Sas­ sanian, Byzantine, or T' ang Chinese states into a feudal-like mode of production, and the strong phases of these same states into an Asiatic mode, wrongly separates into two different modes of production oscilla­ tions within the continuum of a single mode. If variation within the tributary mode depends on the organization of power in particular states, the operation of the mode is at least in part determined by whether that state is weak or strong in relation to other polities. Shifts of power within the states of North Africa and of western, central, and eastern Asia, for example, were intimately connected with the military and political expansion and contraction of pastoral­ nomadic populations, and with the widening and narrowing of surplus transfer through overland trade. If it is true that noncapitalist, class­ dependent modes utilize "other than economic means" for the extrac­ tion of surplus, it follows that successful surplus extraction cannot be understood in terms of an isolated society alone; rather, it is a function of the changing organizution of the wider field of power within which the particular tributary constellation is located.

Historical societies predicated upon the tributary mode may thus tend toward centralization or fragmentation, or oscillate between these two poles. They also exhibit variation in the ways tribute is gathered up, circulated, and distributed. Only in the rarest cases, where a surplus taker and his retinue consume all the surplus obtained in situ, is there no role for processes whereby surpluses are circulated either socially or geographically. Similarly rare are cases where all surpluses are si­ phoned upwards and redistributed downward through the echelons of a hierarchically organized elite without the participation of commer­ cial intermediaries or merchants .. The Andean Inca polity appears to have approximated this form most closely, but even there some evi­ dence exists for the operation of merchants in restricted areas of the Peruvian and Ecuadorian coasts. Much more frequently, surpluses are transferred and exchanged through the transactions of commercial in termediaries.

Civilizations The larger social fields constituted by the political and commercial interaction of tributary societies had their cultural counterparts in "civilizations"-cultural interaction zones pivoted upon a hegemonic tributary society central to each zone. Such hegemony usually involves the development of an ideological model by a successful centralizing elite of surplus takers, which is replicated by other elites within the wider political-economic orbit of interaction. Although one model may become dominant within a given orbit, as did the Confucian model carried by the Chinese scholar-gentry, the civilizational orbit is also an Modes of Production 83 arena in which a number of models coexist or compete within a multi­ ple array of symbols, which find their differential referents in the shifting relationships among the tributary' societies comprising the orbit.

A hallmark of these models is that they not only underline the status of the surplus takers and the social distance separating them from 'other people, but they also claim supernatural origins and validation. The Chinese emperor was the holder of the mandate of Heaven, ensuring the balance of Heaven and Earth; the Confucian sash-bearing scholars activated that mandate by enacting the proper hierarchical relation­ ships. The kraton, or palace, of the Southeast Asian ruler was more than the center of government; it was also the site of religious ritual carried on by the king-god and his noble entourage. The Islamic caliph is emir e/-mominin, the Commander of the Faithful, the guardian ofthe law:and the one who "orders Good and prohibits Evil" (Koran III: 106). Among the Shona the ancestral spirits of the royal clan of the Nembire link the clan with god, mwari . Elsewhere the relation of superordinate power to the supernaturally instituted order is less direct, and it may be mediated through priests. The Hindu raja follows arta, the principle of self-interest and utility, but he requires the services of the Brahmin to institute dharma, the principle of proper supernatural order. In Christendom, the king is ruler by divine right, but he shares rule with the other Coordi­ nate Power, the Church. Whether monolithic or bifurcate, domination is in all these cases inscribed into the structure of the universe. These ideological models paralleling the tributary mode have certain common characteristics. Typically they show a hierarchical representa­ tion of the cosmos, in which the dominant supernatural order, working through the major holders of power, encompasses and subjects human­ ity. At the same time, the ideological model displaces the real relation between power-wielding surplus takers and dominated producers onto the imagined relation between superior deity and inferior "subject" (see Feuchtwang 1975). The problem of public power is thus trans­ formed into a problem of private morality. and the "subject" is invited to win merit by maintaining order through the regulation of his own conduct. The displacement also embodies a contradiction. If public power falters and justice is not done, the ideological ties linking subject and supernatural are also called into question . The rulers lose legiti­ macy; the mandate of Heaven may pass to alternate contenders, or people may begin to assert the claims of their segmental morality against the official apparatus of mediation. Yet the arguments proffered in support of these claims will center upon the nature ofthe imaginary tie between subject and supernatural, not upon the nature of domination anchored in "other than economic means." Mercantile Wealth If the tributary mode points to key relationships through which sur­ pluses are extracted, one must also ask how these surpluses are distrib­ uted after extraction. In nearly all instances some part of the surplus is 84 CONNECTIONS placed into circulation and exchanged. Long before 1400 merchants were transferring commodities over wide areas by caravan and sailing ship, reaping profits from their sale and accumulating great stocks of wealth. Especially where tributary societies existed in a wider field created by competition or symbiosis among contending polities, long­ distance trade in elite goods or luxuries was a frequent and highly developed phenomenon. Such goods embodied the ideological models through which superiority was claimed, and therefore they had an important political referent. As Jane Schneider has phrased it:

The relationship of trade to social stratification was not just a matter of an elevated group distinguishing itself through the careful application of sumptuary laws and a monopoly of symbols of status; it further involved the direct and self-conscious manipulation of various semiperipheral and middle level groups through patronage, bestowals, and the calculated distribution of exotic and valued goods. [1977: 23] Yet this trade in lUxury goods often went hand in hand with long-dis­ tance transactions in bulky staples, especially where access to water ­ ways lowered tr.e energy costs of transport , as in the areas of the Mediterranean, the Black Sea, the Indian Ocean, and the China seas. When the European sea traders intruded into other continents, there­ fore, they often found long-standing networks of commercial relation­ ships that involved principles and operations with which they were wholly familiar. If tributary relationships and mercantile activity have long existed side by side, often to their mutual benefit. such mutualism also entailed conflicts. A merchant is a specialist in exchange, buying and selling goods to obtain a profit. To increase profits merchants strive to enlarge the sphere of exchange, drawing subsistence or prestige goods produced within the kin-ordered or tributary mode into the channels of commod­ ity exchange, the market. This transformation of use values into com­ modities, goods produced for exchange, is not neutral in its conse­ quences . It can seriously weaken tributary power if it commercializes the goods and services upon which that power rests. Granted too much latitude, it can render whole classes of tributary overlords dependent upon trade, and reshuffle social priorities to favor merchants over political or military chieftains . Thus, societies predicated on the tribu­ tary mode not only gave impetus to commerce but also repeatedly curtailed it when it grew too strong . Depending on time and circum­ stance, they have taught merchants to "keep their proper place" by subjecting them to political supervision or to enforced partnerships with overlords; by confiscating their assets, instituting special levies, or exacting high "protection" rents; by denigrating merchant status socially, supporting campaigns against commerce as sinful or evil , or even delegating mercantile activity to despised and powerless out- Modes of Production 85 groups. The position of merchants is thus always defined politically as well as economically, and is always dependent on the power and inter­ ests of other social classes. While defensive mechanisms were invoked whenever tributary power felt threatened by mercantile encroachment, it seems that the European polities that developed after A.D. 1000 granted merchants greater independence and privileges than did most other political sys­ tems. This may have been due to the very backwardness of the periph­ eral European peninsula, as compared to the stronger, wealthier, and more centralized tributary structures of the Middle East and the Orient. Sovereigns striving to consolidate power in the European core regions often needed the aid of merchants to gain access to funds, and they frequently lent support to merchant groups in order to check the demands of rival power seekers. Given the political fragmentation of the European periphery, moreover, merchants were better able to resist political and social pressures, through the creation of their own wide­ ranging networks of trade and finance. The European merchants also enjoyed locational and technological advantages over merchants on other continents. Europe's proximity to the sea permitted an early growth of river and ocean shipping. Water transport not only entailed lower energy costs than transportation overland; it also permitted a closer integration of local and translocal commerce, and avoided the heavy protection costs that burdened the transcontinental caravan trade. An expanding orbit of commercial transactions deployed over a widening grid of transportation, in turn, speeded up the turnover time of money-begetting money, allowing a given sum to earn repeated profits. Some scholars have seen in these medieval European merchants the direct ancestors of capitalism. In this view the change from merchant wealth to capital is continuous, linear, and quantitative; the develop­ ment of capitalism is thus envisioned merely as an expansion of pro­ cesses already at work in the tributary mode. This is essentially the position taken by Weber, Wallerstein, and Frank. If, however, the change from merchant wealth to capital is seen as entailing not merely quantitative growth but rather a major alteration in the determinant processes, then capitalism appears as a qualitatively new phenomenon, a new mode of mobilizing social labor in the transformation of nature. That was the position taken by Marx . From this viewpoint the history of money-begetting money was but "the prehistory of capital." Mercan­ tile wealth did not function as capital as long as production was domi­ nated by either kin-ordered or tributary relations. What was not con­ sumed by producers or tribute takers might be taken to market and exchanged for surplus products elsewhere, allowing the merchants to feed off the price differentials obtained in the carrying trade. The growth of trade after A.D. 1400 greatly enlarged the scale of the market, but it did not automatically lead to the installation of the 86 CONNECTIONS capitalist mode. The tributary mode remained dominant until the capi­ talist mode unfolded and began to threaten it from within in the course of the eighteenth century. During this long period, tributary surpluses continued to be the mainstay of a class of overlords, together with their retinues and servants. Tribute also continued to furnish the sinews of the state: it paid for its armies and navies, supplied its quartermaster corps, and remunerated its officials. The continued extraction of tribute thus set the terms under which mercantile activity could operate and thrive. Yet, by dint of its very success, mercantile wealth began to multiply the channels of commodity exchange, rendering tribute takers increasingly dependent upon it. It generated ever larger amounts of money-begetting money and invested that wealth so as to increase the flow of commodities to the market. In the process it drew producers in different parts of the world into a common web of exchanges, adjusting existing relations of production to embrace commodity exchange, or subsidizing coercive arrangements for the production of commodities. The European merchants engaged in overseas operations brought surpluses into mercantile exchange in a number of different ways. Sometimes they favored one alternative to the exclusion of the others; under certain circumstances they utilized all of them together. None of these ways of turning goods into commodities was new; all had ana­ logues in other tributary systems. They grew directly out of the opera­ tions of the tributary mode and long remained intertwined with it. One way involved the sale of tributary surpluses. Merchants bought stocks of surplus from tributary overlords and state agencies, and sup­ plied goods in return. Their commodities underwrote the life-style of the tributary class; their goods supplied the armies of the state and stocked its magazines. On occasion the merchants also participated in booty taking and plunder themselves, and then sold the spoils.

A second way in which merchants drew goods into the circuits of trade was to open exchanges with primary gatherers and producers. The merchants offered goods that were cheap for them yet desirable to the natives, in return for articles of little value to the producers yet capable of fetching high prices in distant markets. In the course of such ex­ changes, the native producers received use values that they treasured. If pursued over time, however, such exchanges rendered the target popu­ lations dependent upon the merchant. Intensified production of the strategic valuable usually entailed diminution or abandonment of other important economic activities. As producers grew more specialized in furnishing one kind of object, they came to rely increasingly upon the merchant for tools, household articles, prestige goods, and even food. Where the producers proved reluctant to enter or to continue the exchange, merchants sometimes had recourse to forcible sales of com­ modities , which the producers were then obligated to repay. At other times merchants smoothed the course of exchange with liquor or tobacco, which rendered the producer somatically dependent on the Modes of Production 87 donor, thus guaranteeing the resumption of exchange. In time, such unequal exchange, now extended temporally through a system of ad­ vances, could produce a kind of peonage in which the primary producer was constrained by his needs to commit himself to production of the same valuable in the future.

A similar process of increasing specialization and dependence marked the development of "putting-out" systems under mercantile control. Such systems usually originated with the production of specialized commodities by households, which then sold their product to mer­ chants for resale. Gradually, however, the merchants extended their control over the labor process by advancing tools or raw materials, receiving the finished product in repayment for the factors of produc­ tion advanced. Both commodity peonage and putting-out bordered on capitalism, but they were not yet governed by capitalist relations. Both forms of employing labor developed in mercantile terms, with the merchant as agent of exchange advancing subsistence and manufactured goods and receiving specialized commodities. By means of advances the merchant could develop a long-term lien on labor, be it the labol of a kin-ordered group or of a craft shop operating on the edges of a tributary domain. He might even take the further step of advancing tools and raw materials­ powder, shot, and traps, or looms and textile fibers-and thus outfit that labor with complex tools. Such a merchant, however, did not yet buy labor power in a market in which workers compete for available jobs, and he did not yet control the actual labor process. Surplus was not extracted as surplus value but through unequal exchange within the framework of monopolistic and quasi-tributary relationships. The pro­ cess of production was still governed from the demand end, from the reg uirements of merchants exchanging in a market, rather than flowing from the orchestration of labor power and machines within the process of production itself. As long as this was the case, merchants also remained limited in their ability to control the productive process and to alter it in the face of new demands. A third way in which merchants obtained surpluses for exchange was to expand slavery. Slave labor has never constituted a major indepen­ dent mode of production, but it has played a subsidiary role in providing labor under all modes-kin-ordered, tributary, and capitalist. Slavery has been employed repeatedly in large-scale agricultural and mineral production, where output is dependent on a maximization of labor, with minimal deployment of tools and skills. The use of slaves in such production has a continuous history in Europe since classical antiquity, and the option of using slaves to raise commodities for exchange was thus available from the beginnings ofthe European expansion overseas. The later American growth of slavery represents but an overseas rein­ carnation of a process already going on in Crete, Sicily, Madeira, the Azores, the Canaries, and the islands of the Gulf of Guinea. 88 CONNECTIONS As in the systems of advances initiated with primary producers and processors, slavery too required a heavy infusion of commodities to set it in motion. Merchants advanced commodities to African suppliers of slaves, thus placing slaves into the circuits of exchange as one kind of commodity among others. Merchants also advanced money and com­ modities to planters who purchased slaves for work on plantations. As a system of coerced labor, slavery entailed inherent costs, frequently covered through mercantile advances. Slaves had to be broken in and supervised; high productivity entailed high costs of coercion. Since most slave populations did not reproduce, new slaves had constantly to be acquired and paid for. Slaves had to be maintained by their owners, and the costs ate into plantation profits. If the slaves were allowed to supply their own needs on plantation "provision grounds," their increased autonomy decreased owner control. Effective control thus often depended on the importation of foodstuffs and other requisites. Mer­ chants were not the only participants in the system; planters often brought with them inherited tributary wealth and reinvested profits in their own plantations. Yet merchants played an ever-expanding role in financing slavery, in furnishing needed commodities, in providing product markets, and in repatriating profits to the home country. In the process of European expansion, mercantile wealth pioneered routes of circulation and opened up channels of exchange. Its source of gain lay in the maintenance of price differentials-enabling it to buy cheap and sell dear-and it defended itself against price leveling through alliances with any power that could impede the development of a "free" market. It relied on political and military power to seize zones of supply, to gain privileged access to suppliers, to bar interloping competitors in trade, and to ensure maximal profits through monopolis­ tic controls over sales. Aiming for power over persons in order to increase and to diversify output, it did not create a labor market. Thus, mercantile wealth did not alter the mode of mobilizing social labor and remained wedded to the tributary mode. That dependence would not be severed until new political and economic circumstances promoted the rise of industrial capitalism. The Kin-Ordered Mode of Production If the areas of intensive agricultural production in 1400 were occupied by societies predicated on the tributary mode, on the peripheries of these areas all around the world were social groups organized differ­ ently. Such populations are usually called "primitive" in the anthropo­ logical literature. The term is misleading if it leads one to think of Iroquois or Crow or Lunda as one's "contemporary ancestors," or as people who have not yet aspired to the heights of civilization. It is also analytically problematical, since it refers to a beginning which it does nothing to portray. Claude Meillassoux has rightly argued that to char- Modes of Produdion 89 acterize such populations by an absence of features, calling them "class­ less," "acephalous," or "stateless" tells us nothing about what they are. It is common to describe these populations as bound together by "kinship," but less common to inquire into what kinship is. Empirically, populations vary in the extent and intensity of their kinship ties. Some people have "a lot of kinship," others much less. Coresidence is often more significant than genealogy; many local groups include people who are relatives but also others who are not. Tasks may be carried out by teams of nonrelatives, and products of the hunt or of other activities may be shared out among nonkin as well as kinfolk. Indeed, many anthropologists have seen residence as more critical than kinship in understanding how people organize themselves. Thus, both Kroeber and Titiev have argued that coresidence underlay the formation of lineages (Kroeber 1952: 210; Titiev 1943). Leach, similarly, has en­ joined anthropologists to "start from a concrete reality-a local group of people-rather than from an abstract reality-such as the concept of lineage or the notion of kinship system" (1961: 104). Even Meyer Fortes, whose major contribution lies in the analysis of wider kinship systems and their jural and political implications, has noted: A lineage cannot easily act as a corporate group if its members can never get together for the conduct oftheir affairs. It is not surprising therefore to find lineage in African societies is generally locally anchored, but is is not necessarily territorially compact or exclusive. A compact nucleus may be enough to act as the local center for a group that is widely dispersed. [1953: 36] Particular populations also vary greatly in how far they "extend" patterns of kinship found within familial entities to more distantly related families. They differ, further, in the degree to which the ex­ tended or replicated patterns of familial kinship are made to bear the burden of jural and political obligations among groups. In other words, kinship rules may govern filiation (ties between individual parents and offspring) and marriage (ties between particular spouses), but little more. Such rules, moreover, may furnish people with only a vocabulary of kinship "names," without at the same time involving them in jural and political obligations. Among other populations, however, kinship looms large. Patterns of kinship may be used to expand the scope of social and ideological linkages, and such linkages may become major operative factors in the jural and political realm. Kinship can thus operate at two levels, that of the family or the domestic group and that of the political order. Yet such statements still suggest what kinship does and not what kinship is. Indeed, if we cannot define kinship, by the same token we cannot define nonkinship. It may come as something of a surprise to the non anthropological reader that anthropologists by no means agree about what kinship is. They divide, 90 CONNECTIONS generally, into three groups with respect to this issue . First, there are those analysts who assume that the facts of kinship are an outgrowth of human biology . Human beings are sexually dimorphous and engage in sexual relations; as a result, human females bear offspring. The biolog­ ical facts of sex relations and procreation are seen as basic to the human institutions of marriage and descent. In this view kinship is a matter of tracing pedigrees. Second, taking a stand against this position, other anthropologists have argued that kinship is not merely a matter of the social control of sex and procreation, but involves cultural definitions of the marriage bond and cultural constructs that allow offspring to be allocated to parental consorts. In this view kinship is a distinctive cul­ tural domain with its own content, which consists of symbolic con­ structs of descent and affinity. These symbols will vary from culture to culture. Finally, there are anthropologists who argue from a third posi­ tion, which holds that kinship is merely an "idiom " in which economic social, political, and ritual relations are discussed . In this perspective kinship is a metaphor; its real content lies elsewhere. The facts of kinship are explained when the relations it serves to "express" are explained. Those anthropologists who see kinship as the social regulation of biology (sex and procreation) put their emphasis on the way in which rights and obligations, including rights to resources and support, are shared out among biologically produced actors. In their view kinship forms or patterns are cultural epiphenomena serving the task of such allocation. Generally speaking, their concept of kinship has been pri­ marily jural: kinship serves to assign people born into the group to jural positions. The cultural symbolists, in contrast, see kinship as a domain of symbolic constructs connected with other symbolic constructs of the culture. The function of kinship constructs is seen as moral, as a contri­ bution to the ideological ordering of the symbolic universe of the culture bearers . In practice, the symbolists view the elementary family as a kind of storehouse of symbols of the culture and trust that inquiry into domains other than the familial will reveal identical or parallel symbolic constructs (see Schneider 1972). In a larger sense these two positions are complementary. Given that people-unlike snapping turtles-are not hatched from eggs, deposited in a safe spot, and then abandoned , but are born and socialized through the operations of the incest taboo , kinship names and categories are symbolic constructions ab ovo. The human institution ofthe incest taboo depends for its very operation upon a differentiation between those people with whom we share some kind of substance, symbolized by a commonality of blood or bone, with whom we may not mate, and those people with whom we can mate, who do not share our symbolic substantial heritage. Although explanations for the origins of the incest taboo are still incomplete , Claude Levi-Strauss has rightly made it the point of departure of his study of kinship . Just as the initial categories of Modes of Production 91 kinship set up by the taboo are symbolic constructs, so are all the other basic kinship categories , such as gender. absolute and relative age, descent , and affinity. Since symbolism thus enters into the very defini­ tion of human sociality , human beings everywhere have also brought these basic constructs of human "nature" into connection with their constructs of encompassing nature and supernature. (In view of this, the third position in kinship studies sketched above, which denies all but metaphorical status to kinship symbols , appears unsatisfactory , short-circuiting inquiry into phenomena with which it does not want to deal.) It is possible to combine these two approaches into an operational view of kinship that allows us to see kinship in the context of political economy. Kinship can then be understood as a way of committing social labor to the transformation of nature through appeals to filiation and marriage , and to consanguinity and affinity. Put simply, through kin­ ship social labor is "locked up ," or "embedded," in particular relations between people. This labor can be mobilized only through access to people, such access being defined symbolically . What is done unlocks social labor; how it is done involves symbolic definitions of kinsmen and affines . Kinship thus involves (a) symbolic constructs (,filiation/mar­ riage; consanguinitY/ , 1ffinity') that (b) continually place actors, born and recruited, (c) into social relations with one another. These social relations (d) permit people in variable ways to call on the share of social labor carried by each, in order to (e) effect the necessary transforma­ tions of nature . If kinship is a particular way of establishing rights in people and thus laying claim to shares of social labor, it is also true that the ways in which such rights and claims are established vary widely among different culture-bearing populations. Anthropologists have come to recognize tha t kinship works in basically different ways in two kinds of situations, those in which resources are widely available and open to anyone with the ability to obtain them, and those situations in which access to resources is restricted and available only to claimants with a "kinship license ." In the first case, the ties of kinship grow out of the give-and­ take of everyday life and link people who are in habitual interaction with one another. In the second case, the circle of kinship is drawn tightly around the resource base by means of stringent definitions of group membership. This contrast defines two variants of the kin-ordered mode, for social labor is deployed differently in the two . The first variant is best exempli­ fied in the anthropological literature by food-collecting "bands." Such populations do not transform nature, but gather up and concentrate for human use resources naturally available in the environment. The natural environment is not a means for humanly controlled organic transformations , as in cultivation or herding; it is "the object of labor" but not its "instrument " (Marx, Cap. I, 1977: 284-285). Under such 92 CONNECTIONS circumstances the aggregation or dispersion of people, each embodying a share of social labor, follows ecological constraints and opportunities . Upper limits to pooled social labor are set by the interaction of the technology with the local environment, as well as by the group's ability to manage conflict through consensus formation and informal sanc­ tions. Kinship then works primarily to create relations among persons­ partnerships among shareholders in social labor-through marriage and filiation. Such partnerships extend in reticulate fashion from par­ ticular participants to others. Having no defined boundary, they can attach newcomers or exclude them, as the interests of the interlinked partners permit or require. The deployment of social labor works differently in the second vari­ ant of the kin-ordered mode. Where nature is subject to transformation through social labor, the environment itself becomes a means of pro­ duction, an instrument on which labor is expended. A segment of nature is transformed by a set of people-equipped with tools, organiza­ tion, and ideas-so as to produce crops or livestock. In such a society, social labor is distributed in social clusters that expend labor cumula­ tively and transgenerationally upon a particular segment of the envi­ ronment, accumulating at the same time a transgenerational corpus of claims and counterclaims to social labor. Where conditions tend toward ecological closure, relations among these clusters need to be more closely defined and circumscribed, and the clusters readily become exclusive groups. Under these conditions the idiom of filiation and marriage is used to construct transgenerational pedigrees, real or fictitious. These serve to include or exclude people who can claim rights to social labor on the basis of privileged membership. Such groups are typically equipped with mythical charters defining culturally selected and certified lines of kin connection. These charters fulfill a number of functions. First, they allow groups to claim privileges on the basis of kinship. Second, they serve to permit or deny people access to strategic resources. Third, they organize the exchange of per­ sons between pedigreed groups through their definition of ties of affin­ ity; marriage, instead of being a relationship between bride and groom and their immediate relatives only, becomes a tie of political alliance between groups. And fourth, they allocate managerial functions to particular positions within the genealogy, thus distributing them un­ evenly over the political and jural field-whether this be as elders over juniors, as seniors over cadet lines, or as lines of higher over lower rank. In this process kinship on the jural-political level subsumes and orga­ nizes kinship on the familial-domestic level, making interpersonal rela­ tions subject to charters for categorical inclusion or exclusion. The "extension" of kinship is therefore not the same as kinship on the level of filiation and marriage; it is concerned with jural allocation of rights and claims, and hence with political relations between people. On Modes of Production 93 the level of filiation and marriage, kinship sets up individuated linkages among shareholders in social labor; extended kinship, in contrast, orga­ nizes social labor into labor pools and places controls over the transfer of labor from one pool to another. The persistence of the idiom of kinship in the jural-political realm, however, poses a problem. Kinship nomenclature always involves a symbolic process. In the escalation of kinship from a set of interpersonal relations to the political order, kinship becomes a governing ideological element in the allocation of political power. But why should the lan­ guage of kinship persist in this different setting? Meyer Fortes is one of the few to have dealt with this question.

Why descent rather than locality or some other principle forms the basis of these corporate groups is a question that needs more study. It will be remembered that Radcliffe-Brown [1935] related succession rules to the need fo r u neq u i vocal disc rim ina tion 0 f righ ts in rem and personam. Perhaps it is most closely connected with the fact that rights over the reproductive powers of women are easily regulated by a descent group system. But I believe that something deeper than this is involved; for in a homogeneous society there is nothing which could so precisely and incontrovertibly fix one's place in society as parentage . [1953: 30) While this explanation is not wholly satisfactory, Fortes does point to two major sources of power in the kin-ordered mode: control over the reproductive powers of women, and parentage. Both operate trans­ generationally; both allocate people differentially to positions of power and influence. The first grants rights over the social labor embodied in females, offspring, and affines; the second defines not only descent but also collaterality-the genealogical range of mobilizable allies. The terminology of marriage and filiation is thus used to convey information about differential capacities to mobilize labor for work and support­ information that is, about the shifting distribution of social labor among contending groups. Where the symbolic constructs of kinship are thus extended, the relations among the bearers of social labor in competition for resources are structured monopolistically or oligopolistically, with social groups vying for precedence and dominance. At the same time, the tendency to maximize external oppositions vis-a.-vis other groups goes hand in hand with a multiplication of internal oppositions . First, there are oppositions between men and women. Some complementary equilibrium between gender roles can perhaps be maintained as long as kinship is but one ordering element among others in a situation of open resources. With the emergence of pedigreed groups into the political field, however, affinal relations become political relations, and women lose status in relation to men as they become tokens of alliance. There is also the opposition of elders and juniors, with elders in characteristic positions of 94 CONNECTION~ managerial command inside and outside the group. Some juniors may come to be seniors and take their place; but others will never succeed to any position of importance. We know that this opposition can break out into open conflict. We shall see, for example, rebellion of "boys" against their elders in the expansion of horse pastoralism on the Great Plains (chapter 6) and in the formation of slave-raiding groups in Angola (chapter 7).

Finally, internal ranking creates oppositions between original settlers and newcomers, between senior and junior lines of descent from the same ancestor, and between lines rising to prominence and those in a state of decline. Oscillations of rise and decline may be due to demo­ graphic ascendancy or failure; to successful or unsuccessful manage ­ ment of alliances, people, or resources; to success or failure in war. Leaders effective in contracting politically strategic marriages or in the judicious redistribution of subsistence and lUXUry goods to followers gain at tht expense ofless apt contenders. With the passage of time, such gains can be translated into genealogical claims, and pedigrees modified to exhibit the change. The fact that leaders can rise to prominence in this way constitutes one of the Achilles' heels of the kinship mode , one of its diagnostic points of stress. For as a chief or other leader draws a following through judicious management of alliances and redistributive action, he reaches a limit that can only be surpassed by breaking through the bounds of the kinship order. He can manipulate bridewealth to acquire women who will produce offspring owing loyalty primarily to him; he can invite outsiders to settle in his group's domain, in the hope of attracting a personal following; he can acquire pawns and slaves to labor under his personal control. Yet as long as such strategies can be checked by his kinsmen and their allies, his radius of action is curtailed. One way in which a chief may try to expand his grasp on surpluses is by making war. Yet the fruits of warfare remain limited, since booty may be only occasional and impermanent, and it must be shared with others. To break through the limitations of the kin order, a chief must gain inde­ pendent access to reliable and renewable resources of his own. While kin ordering thus sets upper limits to internal differentiation, under conditions of closed resources it appears more likely to produce inequalities than an egalitarian distribution of life chances. Distinctions of gender, age, and prescriptive and acquired power work so as to create oppositions that disrupt the kinship order from within. In addition, disruptions are caused by conflicts between individuals or groups, by the assertion of conflicting claims over people on the part of different kinsmen, and by the nonperformance of normative kinship obligations among close or extended kin. All these forces and factors threaten the continuance of the kinship order. What, then, prevents its disintegra­ tion? How do kin-ordered units cohere at all over time? Modes of Production 95 The ability ofthe kin-ordered mode to regenerate itself may lie in the absence of any mechanism that can aggregate or mobilize social labor apart from the particular relations set up by kinship. The oppositions as they are normally played out are particulate, the conjunction of a particular elder with a particular junior of a particular lineage at a particular time and place, and not the general opposition of elder and junior as classes. In everyday life the kin-ordered mode contains its oppositions by particularizing tensions and conflicts. In myth and ritual, however, the very oppositions fraught with danger in everyday life are dramatized on the level of universality. Whereas in everyday life generality is dissolved into particularities, in myth and ritual particularities are dissolved into generalities, conveying messages about the nature ofthe universe. Explanations, if offered, take the form of universalized verities. One may speculate that such projec­ tion of particular conflicts upon the screen of universal mythic events and meanings can serve to defuse those conflicts. The effectiveness of such a mechanism would seem to depend on the degree to which real conflicts can be kept particularized and segmented. A continuous accumulation of conflicts of the same kind, in the same direction, may place the myth-ritual system under cumulative stress and diminish its efficacy. Conflicts within and between kin-ordered units may also be damp­ ened by a fear of the high cost of massing support. Seeking allies means calling in past promises of aid and pledging support to allies in the future. Any escalation of conflict thus threatens to extend the conflict temporally as well as spatially. Nevertheless, when the stakes are high enough, escalation may well become desirable, with an attendant increase in gift giving and in exchanges of women to cement alliances. The story of the North American fur trade can be read as the gradual extension of supportive alliances among "English Indians" against "French Indians." Such alliances may also be stabilized and reinforced through the elaboration of myth and ritual, as in the case of the Iroquois league, which attempted to curb internal conflicts by directing energies outward against common enemies. Yet conflict resolution in the kin-ordered mode encounters an ulti­ mate limit in the structural problems of the mode itself. Cumulative conflict often exceeds the capacity of kin-based mechanisms to cope with them. Groups will then break up and fission. Such occurrences are not only frequent but are, in fact, important sources of change . Because we have tended to conceptualize societies as if they existed in a timeless ethnographic present and in isolation from one another, we have been misled into seeing the breakup and fission of kin-ordered groups as merely replicating the ordering of the parent group. In reality, fissioning groups can rarely escape into unoccupied terrain to avoid competitors, and they are likely to experience pressures from societies in the tribu- 96 CONNECTIONS tary and capitalist modes. Replication is thus probably exceptional. It is more likely that fissioning groups began to change as soon as they encountered limits to free movement .

Social clusters built up on kinship, therefore, are in no way exempt from internal differentiation and external pressures for change. Differ­ ential allocation of shares of social labor can favor the emergence of influential managers; at the same time, contact with other groups can lend importance to persons able to deal with differences of interest and with possible conflict. These tendencies toward inequalities in function are greatly enhanced when kin-ordered groups enter into relationships with tributary or capitalist societies. Such relationships afford opportu­ nities for the seizure and transfer of surpluses beyond those available within the kin-ordered mode. Chiefs can then employ these external resources to immobilize the workings of the kinship order. This is why chiefs have proved to be notorious collaborators of European fur traders and slave hunters on two continents. Connection with the Europeans offered chiefs access to arms and valuables, and hence to a following outside of kinship and unencumbered by it. The Problem of Chiefdoms The term chiefhas come into common parlance to denote the recognized leader or head (from the Latin caput) of a SOcially organized population . In practice, the term was usually bestowed by Europeans upon any native person of influence who was in a position to forward or to hinder their interests. As such, references to chiefs cover different kinds of recruitment and degrees of authority, and are of little analytic utility. The actual ability of any such personage to command social labor and to influence the development of intergroup relations depends upon his assets in the game of power; the size and strength of the population under his jurisdiction; the nature of the resources held by that popula­ tion and their importance to outsiders; and his war-making potential, his capacity both to defend resources and to interfere with the opera­ tions of opponents. A Northwest Coast tais had less potential power than a Zulu induna, an induna less than a Mongol khan. These differences also affected the ability of a chief to break through the limitations of the kin-ordered mode of production and to become a partner in tributary or capitalist relations. Such variation among "chiefs" throws some light on the long­ standing anthropological problem of "the chiefdom." In efforts to estab­ lish an evolutionary ordering of cultures, the chiefdom was conceptual­ ized as a type of society intermediate between kin-ordered tribes and class-divided states. In this view of the chiefdom, status and power are allocated by differential rank within a common genealogy, yet without entailing differential access to the means of production. The chief and his high -ranking lineage are seen as acting on behalf of a social whole in coordinating specialized activities, planning and supervising public Modes of Produaion 97 works, managing redistribution, and leading in war. Chiefdoms are thus "redistributional societies with a permanent agency of coordination" (Service 1962: 144). While genealogical rank differentiates people by the functions they perform, the society as a whole appears to be laced together through common interests, common descent, and general redistribution. All are kinsmen, as it were; only some are more so than others. The concept of mode of production, however, shifts attention from the form and idiom of interaction between high-ranking chiefs and commoners in a given society to inquire instead into the ways social labor is deployed. In this perspective the societies classified as chiefdoms appear to be of two rather different kinds; those based on the kin­ ordered mode, in which the chief and his followers are still embedded in kinship arrangements and bound by them, and those in which the form and idiom of kinship may be maintained even as a dominant group transforms divisions of rank into divisions of class-in fact, using kin­ ship mechanisms to strengthen its own position. In this second kind of chiefdom, the chiefly lineage is in fact an incipient class of surplus takers in the tributary mode. The growth of such a class may involve a number of different pro­ cesses. Population increase can enhance the relative importance of the chiefly families. Such growth of the chiefly lineage allows it "numerous connections of different kinds with other lineages" (Service 1962: 149). The pursuit of affinal strategies requires that the chiefly lineage concen­ trate wealth from marriage exchanges in its own hands . This implies, in turn, control over exchangeable women and the interdiction of access to elite women by members of the lower ranks. Such control over women can be expanded downward, so as to widen elite controlover affinal exchanges in general. Affinal strategies, furthermore, entail strategies of inheritance. Who gets what is circumscribed by member­ ship in the privileged stratum; goods strategic in the matrimonial ex­ changes and in the inherited wealth of the aristocracy may therefore not pass into general redistribution.

At the same time, growing chiefly lineages may expand by a "budding off of families" (Service, 1962: 166), both within the habitual zone of interaction and beyond it. Such territorial proliferation of high-ranking personnel may create a plurality of power centers in place of a single apex of decision making. Members of the chiefly lineage can become contenders for the chiefship, or create new domains of their own by separating from the parent body. Competition for power feeds back, in turn, upon the processes of accumulation and redistribution. Con­ tenders for power must accumulate adequate "funds of power" and redistribute them selectively to gain followers, rather than open re­ sources to general redistribution. Seen in this light, therefore, redistribution appears as a set of strat­ egies in class formation, rather than as a general characteristic of chief- 98 CONNECTIONS doms as "redistributional societies." Polanyi, to whom anthropology owes the introduction of the concept of redistribution, allowed us to visualize mechanisms of exchange beyond those covered by "reciproc­ ity" or "market" exchange. It is, however, necessary to qualify the concept of redistribution in three ways. First, the different kinds and spheres of redistribution must be specified. Redistribution through feasting is not identical with the redistribution of supplies for public works or warfare, or with the redistribution of specialized resources through the agency of the chief. Second, it is important to be precise about what gets redistributed, how much, and-most importantly-to whom. Feasting with the general participation of all can go hand in hand with the privileged accumulation of strategic goods by the elite. Banquets for war veterans can honor the military contribution of the entire army even as captured people and resources are allocated differ­ entially to nobles and to commoners. Third, redistribution can also serve to "buy" allies and to pacify potential rivals by drawing them and their resources into the hierarchically managed flow of prestations. In this light, redistribution appears not as a kind of normative altruism charac­ teristic of a type of society, but rather as a recurrent strategy in a process of class formation.

In such chiefdoms of the second kind, therefore, the function of kinship changes from that of ordering similarly organized groups in relation to one another to that of drawing a major distinction between one stratum and another. There is now an aristocracy that utilizes and exhibits kin-ordered ties as a mark of its distinctiveness and separate­ ness, leaving to the commoner stratum only residual claims. The aristo­ cratic class thus constitutes itself by radically altering the bonds of kinship in order to promote social distance between rulers and ruled. They may claim differential descent from the gods, or privileged posses­ sion of mana; they may strive to subvert the kin ties of their subjects through the punishment of adultery and incest (see Cohen, 1969), even while setting themselves off as a separate stratum through the practice of class endogamy; and they may invoke special rights over the disposi­ tion of war booty, including conquered populations not included in their charter of kin relationships.

Aristocracies of the kind just described frequently bud off to conquer and rule foreign populations. In such fission and spread, the aristocracy characteristically maintains its separate kinship ties as a source of class solidarity and as a way of setting itself off from the body of the ruled.

This may conceivably happen peacefully, as it did when non-Alur ethnic groups invited the members of Alur chiefly lineages, who were bearers of rainmaking and conflict-resolving powers, to settle among them as their rulers (Southall, 1953). More frequently, however, war­ like and migratory aristocracies invoke supernatural entitlements to impose their models of domination upon subject populations. Examples Modes of Production 99 of such predatory aristocracies are the Toltecs who spread outward from Iula to the frontiers of Mesoamerica; the Luba and Lunda elites fanning out from their home in central Africa (see chapter 7); and the many Mongol, Turkic, and Arab aristocracies that imposed themselves on agricultural populations along the dry-belt corridor of the Old World. Our discussion should make it clear that the deployment of social labor has both an economic and a political dimension. The kin-ordered mode inhibits the institutionalization of political power, resting essen­ tially upon the management of consensus among clusters of partici­ pants. Moreover, the ties of kinship set limits to the amount of social labor that can be mobilized for collective purposes. Social labor can be aggregated through the temporary convergence of many separate ties, but it is dispersed again when changing conditions require a rearrange­ ment of commitments. At the same time, the extension and retraction of kin ties create open and shifting boundaries of such societies.

A chief can become a pivot of the power of his kinship group; but ifhe is sometimes able to incarnate the kin order , he is also its prisoner. Chiefs who want to break through the limitations of the kin order must lay hold of mechanisms that can guarantee them independent power over resources. Such chiefs must either allocate some of the labor under their control to another mode, or enter into the relations of that mode directly, be it as tributary overlords or as participants in capitalist pro­ duction. To effect such change requires new political instruments of domination , whether controlled directly by the chiefs or applied by others on their behalf. Failing this, the people they strive to mobilize may well rebel or secede, leaving them as chiefs only "over the pumpkins." In contrast to the kin-ordered mode, both the tributary and the capitalist modes divide the popUlation under their command into a class of surplus producers and a class of surplus takers. Both require mecha­ nisms of domination to ensure that surpluses are transferred on a predictable basis from one class to the other. Such domination may involve, at one time or another, a wide panoply of sanctions based on fear, hope, and charity; but it cannot be secured without the develop­ ment of an apparatus of coercion to maintain the basic division into classes and to defend the resulting structure against external attack. Both the tributary and the capitalist modes, therefore, are marked by the development and installation of such an apparatus, namely the state. In the case of the tributary mode, the mode itself is constituted by the mechanisms of domination that extract tribute from the producers by "other than economic pressure" (see p. 80). Politics in a tributary state may affect the concentration and distribution of tribute among con- 100 CONNECTIONS tending categories of surplus takers, but it remains anchored in the direct extractive relationship, no matter what the organizational form of the state.

The capitalist mode, in contrast. appears to be economically self­ regulating. As long as means of production are owned by capitalists and denied to laborers, the laborers are continuously forced back into the employment of capitalists after each cycle of production reaches com­ pletion, and the cycle starts anew . Yet the state has a strategic role both in the genesis of the mode and in its maintenance. To set the mode in motion it was first necessary to stockpile money-begetting money, to convert it into capital. and to create a class of laborers offering their labor power for sale as a commodity. In these twin processes of "original accumulation," the state played a vital part. Once the mode was in­ stalled , the state had to deploy its power further to maintain and guarantee the ownership of the means of production by the capitalist class, at home and abroad, and to support the regimes of work and labor discipline required by the mode. In addition , the state had to provide the infrastructure of technical services-such as transportation and com­ munication-required by the mode. Finally, it fell to the new state to arbitrate and manage conflicts between competing cohorts of capitalists within its jurisdiction, and to represent their interests in the competition between states-by diplomacy when possible, by war if required. The three modes of production [ have outlined constitute neither types into which human societies may be sorted nor stages in cultural evolution. They are put forth as constructs with which to envisage certain strategic relationships that shape the terms under which human lives are conducted. The three modes are instruments for thinking about the crucial connections built up among the expanding Europeans and the other inhabitants of the globe, so that we may grasp the conse­ quences of these connections.