PERIPHERY / CORE RELATIONS IN THE INCA EMPIRE CARROTS AND STICKS IN AN ANDEAN WORLD SYSTEvl

The Inca Empire exhibited labor exploitation and the rational extraction of resources from peripheral polities by a core polity. These characteristics fit the general definition of a world empire, although core/periphery relations were diverse. The nature of core/periphery relations depended on several attributes of the conquered polity including population size, political power, natural resources, and distance from the Inca core at Cuzco. A dynamic picture of core/periphery relations emerges as the outcome of Inca demands for labor and raw materials, and peripheral peoples' desire for control over their autonomy while seeking benefits from the Inca state.


Introduction
The Inca Empire (ca.AD. 1400 -A.D. 1532) wa-; the largest indigenous empire and state in the New World .It stretched 3000 km from modern Ecuador to central Andean Chile and Argentina (Figure 1 ).This empire had an army, a large bureaucracy of accountants and other functionaries, a system of taxation, a rigid cla-;s system, and ma-;sive public works including an extensive network of roads and storehouses (Figure 2).The Inca (1) were predatory and expansionist, conquering many different ethnic groups and appropriating goods from their conquered territories.I will explore the usefuln ess of world systems, or core/periphery, [Page 1] Journa I of World-Systems Research approaches for explaining these relations between the Inca and their conquered peoples, and detail some of the complexity in these core/periphery interactions.After summari zing d.cvclopments in antlu•opological world S)-'Stcms the01y.I will note how tl1e Inca Empire confo1ms to onr expectations of a hiernrchical world S)-'Stcm.1l1en.I will provide some details concerning how tl1e Inca elite maniigell their empire.and how local p,,pnlations confo1med to or resisted these machinations.Finally.I will condnde with statements about tl1e •nsefitlness ofworkl systems or corc/periphe1y approaches for Amlcim research.• ' •-•,

Pac ific
The collection of tribute is a central feature of an imperial world system, and theorists debate the nature of this uneven exchange.Wallerstein (1974) concentrates on what he a<;sumes are ba<;ic essential goods such a<; food, a<; opposed to sumptuous preciosities (often referred to as valuables in anthropology; see Dalton 1982) that are primarily trad ed among elites (Wallerstein 1974: 41).Schneider (1991Schneider ( [1977]]) provides a critique of this dichotomy, pointing out that trade in preciosities such a<; precious metals and spices, can be central to maintaining the hierarchies of core polities, and ther efore central to reproducing the economic structures that maintain an empire.Kardulia<; (1990) provid es an example and analysis of such a system, the development of the fur trade for beaver felt hats on the New World colonial periphery .Schneider (1991Schneider ( [1977]]: 54) also points out the power that gift giving can exert over dominated rulers, a<; the reception of preciosities then places these subsidiary rulers in debt to rulers in core polities.These modifications ofWallerstein's original formulation expand the scope of world system<; theory and provide clearer and more testable concepts for social scientists.
In sum, anthropologists have retained a concern with the designations of core, periphery and semi-periphery, and there is a general expectation that core polities exert a dominance over peripheries that ultimately results in the extraction of goo ds from the periphery.While the old designations core/periphery remain important, anthropologists do not typically a<;sume that peripheral populations are hapless victims of core progression, and there is considerable attention to the dynamics of periphery resistance and opportunism (Stein 1993;Gragson 1994;Jeske 1996;Shutes 1996;Wells 1996).
Anthropologists also are testing the limitations of world system<; approaches (Jeske 1996; Stein 1993).Finally, there is inerea<;ing attention to the myriad ways core/ periphery relations are manifest, including considerations of the different types of goods involved, a<; well a<; logistical problems that may limit core influ ences.Whether or not the Inca system can be understood from a core/periph ery perspective depends upon how well ba<;ic core/periphery concept<; fit our knowledge of the Inca Empire.An imp erial world system should exhibit the following features: -The empire should encompa<;s a large and bounded area, and be centraliz ed -The empire should be economica lly self-con tained -The empire should contain core, periphery and semi-periphery politi es -There should be a net economic flow of raw materials and wealth from periphery to core -Ethnicity should be used in the division oflabor Should the Inca empire fit these expectations, then further study may aid in understanding the dynamics of core/periphery interaction.

The Inca Empire as World System
The Inca Empire at the time of Spanish contact (AD.1532) conforms to most expectations for a core/periphery world system.Darrell La Lone (1991,1994) provided the first explicit world systems analysis of the Inca Empire.He argued that the Inca imposed a hegemony upon conquered pcrip hcral polities and transformed periph eral societies from kin-based to tributary-based economics, and he explained diversity in core/periphery relations from a historical, processural perspectiv e.While La Lone provided an intriguing analysis of the Inca Empire from a world systems perspective, further analysis can elaborate on the [Page 4] Journal of World-Systems Research diversity of core/periphery relations in the empire, and explain how people in peripheral and semi-peripheral regions adapted to, were oppressed by, and even profited from Inca domination.
The Inca Empire wa<; a large and economically self-contained polity.Despite the difficulty that many researchers have found in drawing boundaries around world system<; (see Cha<;c-Dunn and Hall 1991: 8-15;Hall and Cha<;e-Dunn 1993: 126), the Inca ca<;e provides clear-cut boundaries.The Empir e eventually stretch ed from the domains of powerful chiefdoms in modern-day Ecuador to the north, wa<; constrained by the Pacific to the west and the Amazonian lowlands to the ca<;t, and reached the Maule river of central Chile in the south (Figure 1).Their interactions with Amazonian and south Andean groups were probably not sustained enough to qualify a<; world-system interaction following Willdnson (1987).The Inca were economically self-contained because they did not have to import either essential foodstuff<; or min erals.The Inca incorporated any polity that possessed raw materials they desired .For example, Ncthcrly (1988) describes how the Inca took control of the middle Chill6n Valley near Lima, Peru, and set up Inca colonies to control the production of coca in this region (Figur e 2).Th e empir e had a very distinct geograph ical core that can be contra<;tcd with more peripheral regions, and also had semi-peripheral states.The Core -Cu:tco.C\izco was clearly the core of the Inca Empire.ideologically.economically.aml politically.Even in the region snmi-nmling C(tzeO, tliffercnces were dm°INn between trne.frmmling lineages of the Inca imd more peripheral people (Znidcma 1990;&ner 1991).Brian Bancr(l991: 115) argues thatthi~ corc-pcriphcry/Incaconqncred d:nalilyis embedded in Inca ideology.C\izco was the center of politics and religion as the residence of the Inca god-king and hi~ holy lineage.or /l(tr.t1"a;C(tzeO was considered the navel of the universe.The lnca brought the children of privileged (and controlled) foreign leaders to Cuzco for schooling (Silvcrblatt 1987: 62), and these children sometimes married into the lnca royal line as a means of solidifying alliances (Silvcrblatt 1987: 87-88).The lnca extended this benevolent hostage taking even to th e gods of conquered peoples a<; the lnca removed foreign idols from their home territories and placed them in the care of Cuzco residents or in the lnca temple of the sun (Silvcrblatt 1987: 94;Cobo 1990Cobo [1653]]: 48, 49) .These idol<; were simultaneously honored by being brought into the sphere of the one true god of the lnca while being subordinated to the lnca god.Finally, it wa<; in Cuzco that the Journa I of World-Systems Research descendants of dead lnca rulers propitiated their dead kings in a state-sponsored cult that was essential to lnca political and religious life (Conrad and Demarest 1984;Conrad 1992).Cuzco represented a core to which economic, political and even supernatural power flowed from the lnca periphery.
The Periphery.Peripheral polities would be regions where lnca rulers were able to appropriate raw materials, labor and wealth, and where the local people or their elites would have little power to alter the form of this extraction.The province of Chupachos is one example of such a peripheral polity (Grosboll 1993;Julien 1993).This province is largely a lowland area in central Peru at the headwaters of the Huallanga river system (Figure 2).A Spanish survey of the province in 1549 not ed that the lnca had substantially resettled the Chupacho province for the purposes of political control and economic extract ion .There were 4108 households recorded for the Chupachos province ( Julien 1993: 210), ofwhich the lncar emoved 1110 (27%) to Cuzco.Another 1498 hous eholds (36 .5%)resided out<;ide of the province either part or full-time.The lnca assigned 500 of these to military service.Of the remaining 1500 households, the lnca a<;signcd 500 to agricultural service, and relocat ed 1000 to specialized production communities (Julien 1993: 210) .ln sum, the lnca forcibly resettled 89% of the households of Chupachos for political, military, and economic purposes.The specialized communities set up by the lnca within Chupachos concentrated on the production of raw mat erials indi genous to that lowland region.These production communities specialized in pottery, woodworking, herding (probably llama<;), gold pann ing, feathcrworking, bird and honey coll ecting, and coca and maiz e product ion (Julien 1993: 206).Many other peripheries existed and current research not only detail<; their peripheral status, but also under scores the diversity in lnca core/periphery relations (Malpa<;s 1993;Morris and Thompson 1985;Earle ct al. 1987;D'Altroy 1994).
The Semi-Periphery.La Lone (1994: 34) notes that kin gdoms such a<; the Chimu and Lu paqa would be good candidates a<; semi-peripheral politics.The Aymara kingdom<; (such as the Lupaqa) to the south of Cuzco, by virtue of their large size, strong political hierarchy, and nearness to the Cuzco core arguably had semi-peripheral status.Th e Inca used pre-existing political structures in the administration of these regions and since many of the former elites were responsible for this administration (Rowe 1946; 272), a limited degree of autonomy and control could be exercised by thes e administrators, matching both Wallerstein's Julien (1993) provides evidence that the Aymara kingdom<; retain ed much of their indigenous political structure, although the Inca managed to extract resources from th em.
For instance, the Lupaqa (Figure 2) were organized into two huno, with each huno caraca lord responsible for 10,000 household<; (Julien 1993;188).Each huno caraca kept a census for the whole province that could be checked against the census of the other huno caraca."In the nearby Colla provinces, administrators of productiv e enclaves in one province resided in another" (Julien 1993;188).She suggest<; that the Inca institut ed this dual organization as a means of providing check<; and balances among the conquered rulers.
The Inca also appear to have sponsor ed specialized craft production in the Aymara kingdom<; (Julien 1993;189;Murra 1965).The Inca promoted a state ideology that laid claim to all land<; and beasts as ultimate property of the Inca state.Thi s fiction was th en used to set up state herd<; in the productive pastures of the Aymara kingdoms of Lupaq a and Qolla.The Inca established a, Pax Incaica; the local boundaries between the pa<;tures of ethnic groups were to be set up and enforced (Polo [1561(Polo [ ], 1940;;194;Falcon [1580Falcon [ ?], 1918;;149;Garcilaso de la Vega [1604], bk. 2. ch. 13;1960;61); many of the llamas were confiscated to form th e nucl eus of Inca state herd<; (Polo [1571], 19 16b; 62); others were granted as spoil<; to the Cuzco soldiers, indi viduall y (Murra 1965: 204).
However, these examples of extraction were not accomplished without the cooperation of local elites and a certain amount ofrec iprocity between Inca and Aymara lord<; that gave the Aymara kingdom<; a degree of control over production (Julien 1988;1993).
Another form of semi-periphery comes from the vicinity of the Inca capital, Cuzco.The Inca based their administrati ve organizati on in the Cuzco valley on a distinction between an Inca-by Blood core and a non-Inca, or Inca-by-Privilege periph ery (Zuidema 1990;12).This periphery wa<; subserv ient to the Inca core; "the inside wa<; considered superior and the [Page 7] Journal o.lWorld-Systems Research outside inferior" (Zuidema 1990: 54) in Inca society.Investigations of the Provinc e of Paruro, near Cuzco, indicates that these residents, known in Spanish a5 Incas de Privilegio, were "subservient to Cuzco, yet allied with it, the Inca de Privilegio represented a large, tribute-paying social stratum that support ed the ruling elite in Cuzco through direct produce, and by occupying low-level bureaucratic positions in state institutions" (Bauer 1992: 141).These administrators, while not true core members, nonetheless bcncfittcd from their administrative connections to the core, and so were not peripheral.Once again, this is evidence of an intermediate form of social organization between the core and periphery.
An uneven flow of goods from periphery to core is a fundam ental clement of world systems formulations (Wallcrstcin 1974;Schneider 1991Schneider [1977]] 1994).Inca exploitation of precious metals conforms to Schneider's ( 1991Schneider's ( [ 1977]]) argument that core politics dominate through their ability to control preciosities.
The Inca dominated many diverse ethnic groups, and this ethnicit y affecte d economic flow in two ways.First, there wa5 the all-important distinction between ethnic Inca by Blood and non-Inca (Zuidema 1990;Bauer 1992), with the attendant privileges of political and cultural dominance conferred upon the Inca by Blood.This distinction formed the ba5is for the Cuzco Core vs. non-Inc a periphery distinctions in economic flow .In their analysis oflnca manipulation of ethnicity in Huinuco, Morris and Thompson (1985: 165) state, "The Inca policy in Huanuco appears to have cmpha5izcd the maint enance and manipulation of diversity rath er than an attempt to integrate through the creation of cultural uniformity ."Ethnicity wa5 also an important factor in the Inca use of coloni es (Rowe 1946;1982;Murra 1980), described below.The uneven flows of wealth, the political domination of weak.erpolities, and the use of ethnicit y to define privilege were part of a system whose parts functioned to perpetuate these exp loitive relationshi ps.
[Page 8] Journal of World-Systems Research Cha..,e-Dunn and Hall (1991: 7) stress that in world systems, the forms of interaction of the system are important in reproducing the social structures of the component polities.Conrad (1981Conrad ( , 1992) ) and Conrad and Demarest (1984) provide an intriguing and controversial reconstruction of how Inca expansion became an essential ingredi ent in the maintenance of Inca social structure.Chroniclers (Poma 1990 [16151;Cobo 1990Cobo [1653]]: 39-43; Cieza de Leon 1959Leon [1553]]: 189) detailed the Inca worship of dead, mummified kings, and the cults that grew out of this worship.Ancestor worship cult.., were related to a system of split inheritance in which the heir to the Inca throne inherited the status of king, but the wealth acquired by his predecessor went to the other descendants.Those descendants who inherited the previous Inca's wealth used this wealth to sustain thc1rnclvcs and his cult.The new Inca, by necessity, had to find sources of revenue with which to build his own wealth.According to Conrad (1981Conrad ( , 1992), this ideological system of split inheritance wa.., the motivating factor in Inca expansion.Other Andean researchers (Paulsen 1976;W. Isbcll 1978;Carnicro 1992) suggest that material factors such a.., mitigating environmental stress, or plundering for personal aggrandizement, motivated the expansion of the Inca empire.Whether or not lofty ideo logy or cra..,s materialism were the motivators for Inca expansion, all researchers agree that the Inca core elite maintained their position through expansion and the extraction of resourc es throughout the realm.Furthermore, the descriptions of peripheral and semi-peripheral politics offered above demonstrate how Inca domination structured the economics of production throughout the empire.In some ca..,cs, this structur e wa.., essential to the maintenance of local elites.
The Inca state generally corresponds to a world-empire a.., defin ed by world systems theorists.As W allerstcin notes of all empires /economics, their ba..,ic purpose is to appropriate the wealth and resources of the periphery and to concentrate these in the core, and they accomplish this through the control of peripheral populations.Howev er, the empire wa.., not monolithic in its control, and used a combination of coercion and enticement to achieve this control.I propose the following hypothesis: the degr ee to which the Inca could coerc e resources wa.., related to the proximity of a polity to the Inca core, and to the size of the population and its pre-Inca political integration.Thos e polities that were small and nearby were ca..,ily coerced, those politics that already possessed strong her editary lead ers with a large population and/or that were distant from Cuzco required a softer touch .

Sticks: Maintaining Control By Coercion [Page 9] Journal of World-Systems Research
The Inca Empire was a large bureaucratic organization with a standing army.Th e maintenance of a costly war-machine indicates that coercion was part of imperial policy, and cthnohistorical data overwhelmingly attest to this.Chronicl ers (Poma 1990(Poma [1615]]; Cobo 1979Cobo [1653]]; Cicza de Leon 1959Leon [1553]]) largely focus on the hist orical a..:;ccndancy of the Inca over their Andean neighbors in battle after battle, and archaeological and iconographic evidence is replete with star-headed maces, defensiv e fortifications, violent injuries, and decapitated heads (Rowe 1946;Poma 1990 [16151).Not only wa..:; the Inca state an empire of conquest, but after a victory, the Inca also employed many means of coercion in order to maintain control.I have already detailed the conquest of new lands for the control of coca production, the forced schooling of conquered nobles' children, forced resettlement, extraction of precious metals, the eminent domain of the Inca state, and the general Pax Incaica.Furth er evidence exists in the form of uneven flows of goods, an Orwellian means of population administration, taxation, colonization, census taking, and the outright use of force.
The simple flow of goods from periphery to core docs not necessarily imply imp eria l economic hegemony a..:; Stein (1993) notes for Mesopotamia, Sinopoli (1994) notes for Medieval India, and Gragson ( 1994) notes for contemporary cattlemen and foragers in Venezuela.Evidenc e of the actual Inca control and appropriation of goods is required.This evidence comes from documentation and archaeological remains of th e Inca administra tion of Chupachos in the central Andes (Figure 2).This province lies in an intermediate location in the empire, only about 500 km from the Cuzco core, and therefore would have been more ca..:;ily coerced than more peripheral regions.Julien (1993: 209) notes the presence of severa l Inca tambos (storehouses) and fortresses .More importantly, the Inca established ethnic colonies of people from other parts of the empir e to garrison the province and to extract resources (Julien 1993: 208 -209), indi cating that the residents of Chupachos had little control over their territory.The radical resettlement of Chupachos within and without their province further suggest..:; a lack: of control over their own destiny.In fact, Julien (1993: 210) conclud es that Chupachos wa..:; an estat e that eventually served the state cult of the dead honoring the Inca Huayna Capac.
The Inca used a decimal system (Cobo 1979(Cobo [1693]: 198;]: 198;Vega 1966Vega [1609]]: 94 -95; Rowe 1946: 263; Julien 1982) to monitor the activities of the empire's subjects and to organize pea..:;ant labor.This system consisted of a hierarchy of foremen and state officials.It is best to consider the decimal system a..:; an ideali zed scheme, since there is debate amon g Andeanists over whether or not the exact figures presented reflected Inca administrative [Page 10] Journal of World-Systems Research realities (Murra 1980: 117 n. 48).I will use the spellings provided in Julien (1982: 123) for these varied decimal officials.Ideally, the system began with a chunka, or foreman, who was responsible for 10 tax payers.Above this official wa<; another local foreman, a pi ska chunka, who had 5 chunka, or 50 tax payers below him.Above this level were hereditary positions, subject to state approval.The pachaka wa<; responsible for two pi ska chunka, or 100 tax payers.The piska pachaka had 5 pachaka below him, or 100 tax payers.This system went up the line including a waranqa (chief of 1,000), a pi ska waranqa (chief of 5,000), and then a hunu (chief of 10,000).Above the hunu wa<; a provincial governor, or t'oqrikoq, who in turn reported to an apo, or prefect of one of the four geographical divisions of the empire (Rowe 1946: 263).Above the apo wa<; the Inca hitrnelf.Theoretically, officials could report dissidence at the lowest level to the Inca hitrnelf in just a few administrative steps (Vega 1966(Vega [1609]]: 94).
Another form of taxation involved the distribution ofland.The Inca divided land into three sections, allotting a portion for the local cotmnunity, a portion for the state, and a portion for the state-sponsored religion (Murra 1980: 31;Cobo 1979Cobo [1653]: 211 ;]: 211 ;Vega 1966Vega [1609]]: 242).These arc not necessarily equal divisions, but nonetheless represent substantial inroads upon a local community's resources.Every local community also owed a labor tax to be spent working these state fields (Rowe 1946: 265).Murra (1965) notes that a similar situation prevailed for state vs. local llama herds.
The Inca levied another de facto tax through the institution of nunneries.State officials chose girls from communities throughout the empire to populate its nunneries; these women were called aclla (Rowe 1946: 269 ;Silvcrblatt 1981). La Lone (1994:26-27) and Silverblatt (1987: 80) note the profound implication of the Inca's ability to dispos e of a polity's women a<; they saw fit, and they point out that the conscription of the aclla drained communities of much labor, and directed that labor toward the state.Rowe (1946: 269) states, "The I nca government controlled its women subjects a<; arbitrarily a<; its men" through this institution.These women, along with being instructed in th e official state religion, also

Journa I o.l World-Systems Research
Uprooting is related to colonization.Colonization is an ancient feature of Andean politics that predates the Inca (Murra 1968;1972, 1980;Salomon and Urioste 1991;Spaulding 1984: 37) , and archaeologists have located evidence of these ethnic archipelagos (Stanish 1989).Thes e colonies existed in several forms, but most involved the placement of ethnic groups in varied ecological zones to exploit indigenous resourc es.The Inca manipulated and expanded this system for state economic andpolitical needs.Row e (1982) details three types of rcsettlcment,yanaconas, camayos, and mitimas (mitmaq, mitimaes), and Murra (1980: 163) provides a similar ela<;sification.Afitmaq refer to colonies linked to specific provinces, the colonists retaining their kin and economic obligations to their own ethnic group (Rowe 1982: 105;Netherly 1988: 267).Yanaconas were personal retainers to Andean rulers, and the Inca maintained such retainers, sometime s recruiting th em from war captives (Rowe 1982: 100).Finally, the ca mayos were communiti es of craftsmen who were uprooted from their own groups, the effect being "to weaken the local loyalties of some ca mayos and create bonds to the Inca state" (Rowe 1982: 105).In the empire, mitmaq were colonies of ethnic Inca ( or other Inca-dominat ed groups) that the state set up in varied regions to control production and limit political dissent (Rowe 1946: 269; Cieza  de Leon 1959 [1553]: 166; Julien 1993).Even powerful, semi-peripheral states had some of their productive lands and resources appropriated by the Inca through the use of mitmaq.
As soon as one o.ltheselarge pro vinces [o.lbarbariamj were conquered , ten or twelve thousand o.lthe men and their wives, or six thousand, or the number decided upon, were ordered to leave and move themselves.fromit.These ivere tran:fferred to another tovvn or province o.lthe same climate and nature as that ivhich they left ... these were called mitimaes (Cieza de Leon 1959Leon [1553]]: 57).
Indians were also tran:fferredfor another reason.Whenever some warlike province had been conquered ivhich was distant.fromCuzco and peopl ed ivithf lerce and restless inhabitants and might therefore prove disloyal or unwilling to serve the Inca peacefitlly , part o.lthe population was moved mvayfrom the area -and o.fien the whole a.lit -and sent to some more docile region, where the neivcomers vvould.flndthemselves su rrounded by loyal and peaceable vassals and thus learn to be loyal themselves ... (Vega 1966(Vega [1609]]: 402-403).

Carrots: Maintaining Control by Enticement
The Inca maintained control with carrots a.., we ll a.., with sticks.These enticem ents included supporting local elites, constructing public works, and the provisioning of conquered peoples from state storehouses.Local leaders who consented to Inca ovcrlordship could benefit from the backing of th e powerful Inca administration, a.., documented for the large Aymara kingdoms of the Titicaca ba..,in (Murra 1965;Julien 1988;1993).The Aymara politics were populous states with strong hereditary rulers and so the Inca could not ca..,ily dominate them.The Inca supplied local lords with gifts and grants ofllatna.., from state herds; the local lords used these grants to provision Inca facilities along it.., highway to feed Inca armies (Murra 1965). Julien (1988;142) notes how local Lupaca rulers had their social position maintained by the Inca a.., heads of decimal systems for mit'a labor recruitment, enabling the Inca to limit the power of these local puppet ....In short, the Aymara lords retained their noble status by a..,sociation with the Inca state, provided they acquiesced to Inca domination and provided goods from their territory along with safe pa..,sagc for Inca armies and goods that pa..,scd to and from more peripheral parts of the empire (Figure 2).The political organization of the Incadominatcd Aymara kingdoms wa.., a mixture of Aymara and Inca organizational features, and in some ca..,cs, the Aymara lords' positions may have been strengthened with Inca support.The adoption of more intermediate forms of social organization between the strong centralized Inca core state, and smaller, more kin-ba..,cd Aymara kingdoms, along with their geographic proximity to Cuzco, matches Cha..,c-Dunn and Hall's (1991) expectations for a semi-peripheral polity.
Chase-Dunn and Hall (1991;8) point out that the "existence of exploitation, domination or unequal exchange should not be a matter of a ... sumption, but rathc r [Page 14] Journal of World-Systems Research investigation."The Inca provided benefits to their conquered va..,sals by expanding and creating many public works including canals and terraces ( essential for much of Andean agriculture), roads, and storehouses (Rowe 1946;229-233;Morris 1988;Ncthcrly 1988;Lynch 1993;Poma 1990Poma [1615]] (1946; 273-274) notes, "The government insured the individual against every sort of want, and, in return, demanded heavy tribute in labor, a very small part of which dir ectly bcncfittcd the people who paid it."The Inca manifested this control over their subjects through their extensiv e systems of food and goods stored in high -altitude colcas and tambos (Rowe 1946;231 ;Morris and Thom pson 1985;108).The Inca could use these stores to aid subjects in times of need, or to punish subjects by withholding them.In either way, the Inca harnessed the labor of conquered peoples to keep these stor es full.
Y asf el Inca, en este mes de marzo, tenfa mandado puesto un juez en cada pueblo de las sementeras, para que no las gastasen las indios ni las ac abasen presto las co midas; y queguardasenpara todo el aiio ... (Poma 1990. And so the Inca, in March, had placed a judge in each town of the agricultural lands, in order that no one wa<;tc nor cat up the food too fa<;t; and to guard the food all year ... Throughout the whole kingdom there were three sorts of storehouses to hold the harvest and tribute.Every village, whether large or small, had two storehouses: one wa<; used to hold the supplies kept for the use of the people in lean years, and the other wa<; used for the crops of the Sun and Inca.There were other storehouses at intervals of three leagues on the royal highway, and these came to serve the Spaniard<; a<; inns or taverns (Vega 1966(Vega [1609]]: 255).

[Page 15] Journal of World-Systems Research
These examples indicate that Inca administrators used enticements a<; well a<; punishments in maintaining control of its conquered subjects.Inca core/conquered periphery relations were varied; benefits a<; well a<; punishments could flow from core to periphery.Relations between the Inca and their subjects were hardly monolithic , and conquered people were able to exert their own powers to differing degrees in dealin g with the state.

Conclusion: The Inca Empire as a World System
Did the Inca Empire conform to a world system?Considering cthnohistorical accounts and emerging archaeological analyses, the answer is a qualifi ed yes.The empir e was a bounded, self-contained system with recognizable core, semi-periphery and periphery characterized by an uneven flow of goods and services from periphery to core that in turn structured economic and social organization of all polities involved.Recognition of th e Inca's use of indi genous social organization actually strengthens the case for a world system since we can see the systemat ic formation of a semi-periphery.La Lonc's appreciation for historical process unde rscores as well that the Inca Empire was not a.fait accompli, but rather an imperial world system in the making.This, too helps us to understand why the Inca were not an unqualified world system.La Lone (1994) suggests that the diversity we see in Inca core/conquered periphery relations is due to the fact that Inca domination was in process and incompl ete when the Spanish arrived in 1532 and

Journa I of World-Systems Research
Mesopotamian example).The only bca<;ts of burden were llama<;, which can carry only about 25 kg each, and people, who need to transport their own food (see Ha<;si g 1992 on the Aztec).As the Inca ventured farther from their Cuzco core, they were harder pressed to force their hegemony upon other politi cs.Finally, the Inca empir e wa<; still in a process of consolidation when it collapsed .
While studies of the Inca can aid our investigations of imperial world systems, Andean researchers can use world systems theory to reinforc e our und erstandin g of the empire's most salient characteristics.One beneficial influence of a world system<; approach for Inca research would be a reorientation of both archaeological and cthnohistori cal research to the empire's periphery, since knowl edge of the core is de facto incompl ete w ithout knowledge of the periphery.This reorientation is already underwa y (Malpa<;s 1993;Earle ct al. 1987;Morris and Thompson 1985) and it offers us both a richer understanding of the operation of the Inca Empire a<; well a<; information on previously littl e known or unknown regions and peoples.The comparative :framework of world systems theory will also benefit Andean research a<; it relat es the Inca to similar empires around the world and through time.Such comparison may enhance efforts at understanding the growth of the empir e and its eventual downfall.The approach also focusses Andc an researchers on the most salient characteristic of the Inca Empire a<; well a<; other Andean states -the control over the production of the varied goods necessary for existenc e.
A<; an example , many scholars (Murra 1972;Ba<;ticn 1978;Brush 1977;B. J. Isb ell 1978;Mujica ct al. 1983;Stanish 1989) have underscored the importanc e of vertica lity in Andean economics.The discussion above provides evidence that the Inca maintain ed such vertica l control either throug h tribute payments or through direct colonization.
World systems theory and its comparative :framework may also shed light on th e cause of empir es, and thereby help to resolve the current debate over the ideolo gical vs. mat erial foundations of Andean empires (sec Conrad 1981Conrad , 1992;;Carnicro 1992).Chase-Dunn and Hall (1991) call for empirical cross-cultural comparisons to enable a broader applicability of world systems theory concepts, and this description of the Inca Empire along with La Lonc's (1994) help to provide just such an empirical ba<;is for comparison . [

[Page 6 ]
Journa I of World-Systems Researchand Chase-Dunn and Hall's definitions.

k:h;:,qui Argentina
(Rowe 1946za deLeon   1959Leon     [1553]]: Vega 1966Vega [1609]]omplexes of knotted strings that quipu-camayocs used a<; mnemonic devices to remember accounts and historic events.Inca institutions ofland tenure, labor allocation, population organization, and accounting arc impressive mechanisms for resource extraction, that requir ed coercion to institute and maintain.The Inca coerced through the punishment of individuals, punishment of communities, military attack, and forced resettlement.Chroniclers noted that Inca punishment wa<; swift and severe.Individuals found guilty of an offens e received reprimand, exile to plantations, loss of office, torture, or death(Rowe 1946: 271, Cobo  1979 [1653]: 203-207;Vega 1966Vega [1609]]: 96; Cicza deLeon 1959Leon   [1553]]: 171).More importantly, dissident communities could receive harsh retribution from the state, being attacked, destroyed and/or uprooted (Vega disrupted this indigenous development.LaLone (1994: 21)states, "The incomplet eness of the transformation reflects also the still inchoate development of core/periphery hierarchy.Cuzco's status a<; core wa<; indisputable ... while the status of other regions a<; relative periphery wa<; still in contention. 11 offer the following hypothetical explanations for the variability in Inca core/ periphery relations.One is the use of indigenous political organization in the empire's administration of conquered politics, leading to different modes of extraction a<; well a<; to the creation of a semi-periphery among conquered, yet powerful kingdom s such a<; th e Lupaqa, Colla and Chimu (La Lone 1994).Another rca<;on is that Inca imperialist <;, despite their army, roads, and administrative organization, were still constrained by distance (secStein 1993 for a   [Page 16] Page 17] Journal of World-Systems Research Note l.The term Inca refers to several different things.It can refer to the Inca icing himself, th e core of Inca noble lineages, the ethnic Inca of the Cuzco region , or int egrated members of the Inca society.I will use it in all of these ways, indicating which definition prevail<; in each particular context.