Gose_diarchy, History and Divine Kingship

  • Upload
    iriepu

  • View
    228

  • Download
    1

Embed Size (px)

Citation preview

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    1/33

    History and Anthropology, 1996 1996 OP A (Ove rseas Pub lishers Association)Vol. 9, No . 4, pp . 383-414 Am sterda m B.V. Published in The Ne therlan dsPhotocopying permit ted by l icense only by Harwo od Academic Publ ishers Gm bHPrinted in M alaysia

    THE PAST IS A LOWER MOIETY:DIARCHY, HISTORY, AND DIVINE KINGSHIP IN THEINKA EMPIREPeter GoseUniversity of Regina

    IntroductionWhat k ind of government d id the Inkas have? From the Spanishchroniclers until the 1960's, most answ ers to this question assum ed thatthe Inkas were imperial monarchs comparable to those of early modernEurope. Debate about Inka political structure primarily focussed aroundthe degree of centralization of the empire. Some stressed the ongoingviability of the regional polities conquered by the Inkas, whereas othersdiscerned a mo re consolidated and "de spotic" m ode of sovereignty.Whatever their personal preference in this debate, most commentatorsassu m ed tha t Inka pol i t ics took place a long a co nt in uu m betw eenimperial centralization and regional autonomy, two poles that werepersonified by the Inka monarch and the provincial lords, respectively.They treated A ndean politics as an essentially secular matter, and thou ghtthat the main issue was the extent to which the central state monopolizedand institutionalized political power.With the pioneering study of Zuidema (1964: 127), however, a newinterpretation of Inka government began to emerge, one that claimed thatthe Inkas may have had two simultaneously reigning sovereigns. It hadlong been known that like many Andean peoples, the Inka elite of Cuzcodivided themselves into "upper" and "lower" moieties. Zuidema and hisfollowers argued that the duality of Andean social organization extendedinto government , such that each moiety s imultaneously provided asovere ign in what was e f fec t ive ly a d ia rchy (see Duvio ls 1979a ,Rostworowski 1983: ch. 5, Nethe rly 1993: 18). Suddenly, balance replacedhierarchical centricity as the paramount organizing principle of Inka

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    2/33

    384 PETER GOSEIn this paper, I will critically assess the evidence for diarchy in the pre-Columbian Andes, and conclude that although it existed in the provincial

    polities conquered by the Inkas, it was absent from the imperial level ofgovernment altogether. In the provinces, furthermore, diarchy was basedon hierarchy, not balance. It was one of several principles of politicalorganization that were all subordinated to a segmentary hierarchy thatculminated in the Inka sovereign. As in the provinces, the supreme Inkaruler (sapa inka)was from the upper division (Upper Cuzco), however thedifference was that his opposite number in the lower division (LowerCuzco) was no t another ruler, bu t the so-called high p riest or "priest w hoannounces" (willaq um u). The chroniclers simply do not mention anyins t i tu t iona l ized po l i t i ca l r ep resen ta t ive o f th i s moie ty , the re fo remonarchy, not diarchy, was the overarching feature of this politicalstructure.More serious is the problem of how to interpret the priest-king relationat the pinnacle of the Inka imperial hierarchy. Several Spanish chroniclers(e. g. Segovia 1553: 75-6, A non ym ous Jesuit 1590:161) saw it as analogou sto the relation between the pope and the king in Christendom, and morerecently Zuidema (1964: 246) suggested a model closer to the priest-kingrelation in Hinduism, where each figure is paramount in his respectivedomain. Both models assume precisely what needs to be proven in theA nde an case, nam ely the existence of a separation between transcendentalreligious authority and secular political power. I will argue that such adistinction w as foreign to the Inkas, w ho practiced a relatively mo nolithicversion of divine kingship. In life, the ruler was expected to prove hisdivinity through conquest, and after death, by providing fertility andoracular advice. The high priest of Lower Cuzco was an oracular m ediu mwho represented the Sun as the ancestral unity of previous Inka rulers.Thus, the king and the priest were opposed to each other as the presentsovereign to his predecessors, not as "politics" to "religion". I will showhow his distinction lay at the heart of Inka im perialism .A final aspect of this pa pe r is a cr i t ique of the roma nticism andexoticism that underwrites the search for diarchy. Although partisans ofthe diarchy thesis join m ost scholars in recognizing the asymm etric natu reof socio-political dualism in the Andes, in practice they seek to discover asymmetric sharing of power between social groups. Therefore, theydownplay the id ioms o f conques t and sub juga t ion tha t genera l lyorchestrate Andean moiety divisions. Instead, they argue that diarchy

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    3/33

    THE PAST IS A LOWER MOIETY 385Rostworowski (1983) oppose diarchy to monarchy, and present it as anarrangement that was so alien to the Spanish chroniclers that they couldnot understand or record i t properly. Not only did diarchy quicklybecome the only "Andean" polit ical principle, but the very lack ofevidence for it in the chronicles constituted the most definitive proof of itsexistence. As an antidote to this sort of argumentation, I propose a returnnot only to the evidence, but also to a broader comparative approach toInka political structure. Rather than assume the uniqueness of the Inkasas a point of method, I would prefer to discuss them from a comparativeperspective. In this way, it should be possible to reveal common groundwith other civil izations, but also the precise points where the Inkasdiverge and truly are uniq ue.

    Structure And HistoryTo und erstan d the mon archy-diarchy debate in Inka historiography, wemust begin with the l is t of Inka sovereigns given by the Spanishchroniclers. The names and deeds of each sovereign were recorded in

    paintings, epic songs, and knotted strands of yarn (kipus), and the Inkaspecialists in charge of each of these media were the primary sources forthe S pan ish chro nicle rs (see Betanz os 1551: 150, Polo 1561: 141-2 ,Sarm iento 1572:114). A lth ou gh t he Inka king lists w ere com piled ashistory by the Spanish chroniclers, the information on which they werebased came from the commemorative traditions associated with Inkakingship. However, these comm emorative traditions included m uch morethan a series of narratives and graphic representations of the life andtimes of dead kings: they also involved the mummification and worshipof the sovereign's dead body, and the extension of these ancestor cultsinto the social and spatial organization of the empire. As we will see,much of the controversy over the nature of Inka sovereignty result fromattempts to edit the full range and internal logic of Inka commemorativepractices to make them fit with a narrowly conceived version of western"history". Nonetheless, it is here that w e must beg in.There is a reasonable degree of consistency in the king lists publishedby the Spanish chroniclers, although agreement is not absolute, and manycontradictions arise when different accounts are compared in detail. Theversion most frequently expo und ed includes thirteen kings , starting with

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    4/33

    386 PETER GOS E( a ) . HistoryManco CapacSinchi RocaLloque YupanquiMayta CapacCapac YupanquiTca RocaYahuar HuacacViracocha Inca

    lowerupperlowerPachacuti Inca YupanquiTopa Inca YupanquiHuayna CapacHuscarAtahualpa

    upper

    time Lower Cuzco(b ) . Diarchy

    Manco CapacSinchi RocaCapac Yupanqu iLloque YupanquiMayta CapacTarco Huamanson of abovedon Juan Tambo Maytapanaca

    Upper Cuzco

    Inca RocaYahuar HuacacViracocha IncaPachacuti Inca YupanquiTopa Inca YupanquiHuayna CapacHuscar / Atahualpa

    Chima Panaqa(Manco Capac)

    Raura Panaqa(Sinchi Roca)

    Yauri Panaqa

    Auayni Panaqa(Lloque Yupanqui)

    Sauaseray Panaq

    Usca Mayta Panaqa(Mayta Capac)

    Masca Panaqa

    ( c ) . Slower

    Apo Mayta Panaqa(Capac Yupanqui)

    "Extra

    ructureupperUicaquiraa Panaqa(Inca Roca)

    1 Panaqas

    Aucaylli Panaqa(Yahuar Huacac)

    Suma Panaqa

    Socso Panaqa(Viracocha Inca)

    Iaca Panaca

    Hatun Ayll(Pachacut

    Capac Ayllu(Topa Inca)

    Tumi bamba Panaqa(Huayna Capac)

    Figure 1 From History to Structure

    immediately succeeded him are attributed to Lower Cuzco, whereas thelatter kings are attributed to Upper Cuzco. The first sovereign fromUpper Cuzco is usual ly ident i f ied as Inca Roca , but somet imes asPachacuti Inca Yupanqui. A more detailed discussion of this issue willfollow, but for now, the precise changeover point in the genealogy is lessimportant than the overarching pattern of the early sovereigns beinglower and the la ter ones being upper. Not only were these moietiesspatialized, then, but they are also temporalized, such that Lower Cuzcowas primordial and upper Cuzco was recent. The reasoning behind thispat tern will be extensively discussed below.These moiety affiliations of the Inka rulers are the point of departurefor the diarchy thesis. Zuidema argued that instead of treating the rulerschronologically, they could be seen as representatives of social groupsarranged in a synchronie structure. The moiety division between Upperand Lower Cuzco was the s imples t of these synchronie s t ruc tura lgro upin gs. Th us, Zu idem a (1964: 53,127 -8) paid particular attention to avarian t king list from Acosta (1590: Book 6, chs. 20,23) which h e interp retedas evidence of tw o sim ultaneous ruling d ynasties (see Figure lb ). W hereaspreviously, Sinchi Roca was th ou ght to be the second king an d Inca Rocathe sixth, now they were cast as contemporaries. In this way, the diarchy

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    5/33

    THE PAST IS A LOWER MOIETY 387Thus, the variou s Inka sovereigns were representatives of a group of 10simultaneously-existing descent groups.3 This effectively removed any

    remaining diachronic component from the list of rulers, and completedthe assimilation of Inka history into synchronie social structure (Zuidem a1964: 234). Thus, an early structuralist repudiatio n of history in the nam eof synchronie order has underwritten the controversy over the nature ofInka government.There are many technical objections that can be raised against thisformulation of Inka diarchy. First, Acosta nowhere specifies that thesetwo lineages were contemporaneous. Zuidema (1964: 128) infers that thelast ruler of Lower Cuzco , do n Juan Maytapanaca, w as a contem porary ofHuscar and Atahualpa because he had a Spanish surname. Otherwise,there is remarkably little information available on this individual and hisposition, in contrast to the other prominent Inka nobles who survived theconquest. How could so important a person be so thoroughly ignored?Acosta copied his genealogies from Polo, wh ich is why they includ e TarcoHuaman, who does not figure in the standard king lists given by otherchroniclers. Unlike Acosta, however, Polo (1585:10) does no t con tinue theline of Lower Cuzco on to don Juan Tambo Maytapanaca, althoughperhaps his lost account of 1559 did. This only underlines the historicalobscurity of Don Juan Tambo Maytapanaca. Second, neither Polo norAcosta assembled this information to show that the Inkas practiceddiarchic government. On the contrary, Acosta (1590: 418) categoricallystates that they had a monarchy. As for Polo, he appeared as a witnessbefore the inqu iry in to Inka history sponsored by Viceroy Toledo in 1572to corroborate accuracy of a series of written accounts and graphic panels(paos) that identify the first Inka rulers with Lower Cuzco, and the laterones wi th U pp er C uzco (see M onte s ino s 1882: 252-7) .4 Thus , theauthorities cited tum out to offer no support for a diarchic interpretationof Inka government. Even if we assume that Polo and Acostamisinterpreted the evidence available to them, we still have to ask whythere is no further evidence for two co-reigning dynasties, and so muchfor the m ore traditiona l view of the Inkas as a monarchy.Arguably, however, none of these objections really matter. From thebeginning, Zuidema held that the chroniclers' various versions of Inkahistory contain so many contradictions that they cannot be treated ashistory at all. Rather, they should be seen as a corpus of myths, whichreveal more about the categories of the Inka's social structure than their

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    6/33

    388 PETER GOS Eculture (see Netherly 1990: 461). In other words, these authors perceiveAndean societies as "cold" (Lvi-Strauss 1966: 234), and thus, resistant tohistory. It is difficult not to suspect that we are dealing with a romanticin v e r s io n o f s t a n d a rd w e s t e rn n o t io n s o f " p ro g re s s " h e re . O th e rexponents of the diarchy thesis, including Duviols (1973, 1979a) andRostworowski (1988a) have wisely taken a more moderate posit ion,which al lows that these tradit ions may express both the categoricalorganization and the history of Inka society. More recent advances instructuralist m ethod have revealed ho w right they w ere to do so.Sahlins (1985) showed the dichotomy between structure and history tobe ill-conceived because events are always created againts the backdropof a cultural order, which also invests itself in those events, and maychange because of them. Furthe rm ore, cultures have different (structural)mot ives for recount ing the pas t . Rather than cont inue to asser t aproblematic divide between "myth" and "history", it might be better tosc rap these ca te go r ie s a l tog e th e r in favo ur of a m ore de ta i le dinvestigation of particular cultural modes of temporal consciousness. Forexam ple, Rostw orow ski (1988a: 13, 53-9) argues that Inka dyn astictraditions were constantly marshalled and reworked in the struggles tosucceed each dead king. Their aim was to legitimate potential successors,and in so doing, they had to invoke established values and notions aboutthe past an d ap ply them to the present. Therefore, different poin ts of viewproduced different dynastic his tories . One hardly need add that inprinciple, at least, this does not distinguish Inka history from its westerncounterpart.H avi ng said this mu ch, I join Zu idem a in refusing to view "Inkahistory" as a precise record of what actually may have happened in theAndes before the arr ival of the Spaniards. Rather than see i t as anexpress ion of soc ia l s t ruc ture however , I see " Inka h is tory" as anem bodimen t of the values that unde rwro te divine kingship in the An des.In other w ord s, I follow Zu idem a in treating the Inka k ing lists as a veiledcommentary on Andean social arrangements, but differ from him inallowing that these arrangements were expressed in a local idiom centredon the past, not a synchronie notion of social structure of the sort thatprevailed in anthropology during the 1960's. What is at stake here is theintegrity of a form of cultural expression. The structuralists arg ue tha t thequasi-historical format of the Inka king lists was an Hispanic imposition,w hereas, I suggest that it also had an indige nou s basis.

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    7/33

    THE PAST IS A LOWER MOIETY 389past (1991: 83), bu t this does not fully explain the quasi-historical formatof their accounts. Even the most cursory reading of the chronicles showsthat Andean people themselves were vitally interested in the deeds ofthei:1 dead rulers, and went to great trouble to record them on quipus andin song s (Cieza 1553 II, ch. 11, Santilln 1553: 10) and to com m em orateand recount them sequentially in royal mortuary rites (see text below).The:;e narratives were part of a larger attempt to retain and conserve thelegacy of exemplary rulers by mum mifying their bodies and constructingstatues that contained their hair and nail clippings. But the Spaniards didnot perceive the continuity between these narratives, which they couldap p reciate and convert into "Inka history ", and the cult of mu m m ies andstatues, which they classified as "idolatry" and repressed without mercy(see Duviols 1971). Structuralist commentators make the same mistake inreverse when they fail to see the Andean "cult of the dead" at work in"Inka history", and focus only on the distortions introduced by Spanishh i s to r i ca l consc iousness . O n the con t ra ry , i t w as th rough the i rhistoriographie traditions, perhaps more than in any other aspect of theirculture, that the Span iards were m ost fully able to interact and harm onizewith the commemorative practices that were so important to Andeanpeo pie. Consequently, it would be far more p roductive to investigate theparallelism, and not just the sl ippage, between Andean and Spanishsensibilities in the joint creation of "Inka history".Civine kingship was an important motive for historical thinking. Likemost Andean peoples, the Inkas worshipped dead rulers who in life hadconquered new territories, or expanded the agricultural frontier throughterracing and irrigation (see Duviols 1987, 1979b). These actions wereconsidered so exem plary that up on d eath, the ruler w ho performed themwaf said to turn into stone and rem ain connected to his kingdom as ananc ?stor-deity w ho continu ed to provide his peop le with life, ag rarianfert lity, and oracular advice about the affairs of state. The descent groups(ayllus, panaqas) founded by dead rulers in turn sustained their deifiedprogenitor through sacrificial offerings. An additional aspect of thisworsh ip was the commemora t ion of the exemplary deeds o f the i rfounding ancestor. In this way, divine kingship rituals produced andenc Dmpassed a par ticu lar sort of history.The Inkas were imbued with these values of life-giving rulership, anddiff sred from other Andean people only in that they expected all of theirsovereigns to live up to their divine billing as "sons of the Sun", and

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    8/33

    390 PETER GOSE(P. Piza rro 1571: 70, Betanzos 1551:1 chs. 30-1 , Sarm iento 1571: 219). Thisaspect of mortuary ritual could not be completed for sovereigns like InkaYupanqui and Viracocha Inka who had no conquests to recount, thereforetheir sta tus as deities was prob lematic (Cieza 1553: II chs. 37, 46). The corenotion of Andean divine kingship was thus the development of livingwarriors into dead deities.Up on death, the bod y of an Inka sovereign w as mum m ified, a nd one ormore statues of him w ould be fashioned from gold or cloth. His surviv ingwives and children would from his descent group or panaqa, whichsubsisted on the produce form his country estate outside Cuzco, wherehis mummy and/or statue usually resided. Male and female retainers(yanakuna and mamakuna) attended to the mummy on his estate, and theirduties were both agricultural and priestly in nature. Only one of the deadsovereign's sons did not join his father's panaqa,and that was the onewho was chosen to inherit the kingly position of sapa inka. The successorwas expected to assemble a group of governors and military captainsaround him to enlarge the boundaries of the empire, to build his ownpalace in Cuzco, to marry many women and have many children, whowould ultimately form his panaqa on death. In turn, his successor wouldembark on a career of conquest, and leave behind a descent group of hisown, and so on. In this way, the social organization of Cuzco could beunde rstood as a series of descent groups founded by de ad rulers.The question Zuidema posed was whether the panaqas really werecreated in an additive, historical fashion, or whether they had alwaysexisted, and only later found a connection w ith particular sovereigns. Hisanswer has always been that the kings were little more than emblems ofpre-existing social groups (Zuidema 1964:122, 128, 1990a: 81). Part andparcel of this argument is that there were ten royal panaqas in Cuzco, justas nearby ethnic grou ps w ere composed of ten ayllus (Zuidem a 1964: 200,Sh erb on dy 1986: 52) . Ins tea d of be in g an acc re t io na l h is to r ica lphenom enon, the royal panaqas became just another expression of a socialstructure common to several ethnic groups in the Cuzco area.If Zuidema is right, then presumably there should only have been tenof these descent groups mentioned in the chronicles, and their identityought to have been at least as stable and well-known as the royalancestors nominally connected to them. In fact neither condition holds.The chroniclers generally attribute one panaqa to each of the thirteenrulers on the traditional king list, except for Huscar and Atahualpa,

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    9/33

    THE PAST IS A LOWER MOIETY 391has since shown that there are five additional panaqas mentioned byvarious chroniclers, two for Upper Cuzco and three for Lower Cuzco,which make a total of sixteen (see Figure lc).

    5Zuidema's solution to thisproblem has been to privilege those sources (e.g. Gutierrez de Santa Clara1630, Sarmiento 1572), and those aspects of Inka social organization (e.g.the ceque system) which include only ten panaqas, and to insist that therewere no more (see Zuidema 1990a: 81). Yet there is no reason why therecould not have been ten panaqas organized around the ceque system, andothers that were not.We know that the ceque system was created under the reign of eitherPachacuti or Topa Inca, and that it involved an exhaustive reallocation of

    land and water rights in the Valley of Cuzco, with their attendant ritualduities, to the ten panaqas that participated in it (see Sherbondy 1986:47-9). The "eleventh" panaqa, that of Hu ayn a C apac, held its land outsideof Cuzco and its ceque system (Sherbondy 1986: 59-60). There is ampleevidence that this exclusion was the source of considerable friction in thefinal years of the Inka empire. In the war of succession that followedHuayna Capac's death, Huscar threatened the established panaqas withanoth er land red istributio n (Betanzos 1551: 207, Piza rro 1571: 54), an dAtahuallpa's generals nearly annihilated Capac Ayllu, Topa Inca's panaqa(Sarm iento 1572: 269-70, M ur a 1613: 202-3) , w hic h acc ordin g toShe rbon dy (1986: 45-6), held th e m ost privileged position in the cequesystem. Th us, there w as a structure of ten panaqas in imperial Cuzco, onethat was fixed primarily by an allocation of land water rights. However,this structure did not necessarily prevent new panaqas from forming orold ones from disappearing. Rather, newly formed panaqas were eitherexpel led from the ceque sys tem, o r they had to e l imina te a l readyestablished ones within it, or push them downward through its ten-parthierarchy of rights. This is probably why there are alternate names forsome of the more prestigious panaqas, e.g. Hatun Ayllu and Iaca Panaca:instead of being synonyms, they probably refer to different groups thatoccupied the same position in the ceque system. Thus, it is possible toreconcile the ten-part structure of the ceque system with a processualunderstanding of how panaqas formed around each successive divineking.Ironically, Zuidema's own research into Inka kinship terminologysupports a king-centred "historical" perspective. Apparently, rank wasdefined from the egocentric perspective of each king, and would vary

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    10/33

    392 PETER GOSEwhat the latter involved was a constant redefinition of the social orderaround each succeeding king, which is precisely what was facilitated bythe usages Zuidema describes. Again, one suspects that an overly rigidcontrast is being drawn between the systemic and egocentric aspects ofhow the panaqas were formed.To summarize: when Zuidema decided to rewrite "Inka history" as asynchronie relation between ten social groups, he had to show that theroyal panaqas were not actually formed on the death of each successiveking, as nearly all the chroniclers claimed they were. Thus, he downplaysthe very rich information on royal mortuary practices to be found in thechronicles, and cites approvingly those who have directly challenged itsvalidity (e. g. Aranibar 1969-70). But there is overwhelming evidence forthe king-centred nature of the panaqas as social groups. They rarelyassembled in public without the mummified body of their founder, anddevoted most of their resources to maintaining his cult. Polo and otherSpaniards actually found most of the royal mummies, so they cannot betreated as purely mythical f igures . Although the ceque sys tem hasprovided new and interesting perspectives on Inka culture, it has hardlydealt a death-b low to this evidence.

    Whatever the ult imate fate of Zuidema's more radical attempts toreduce Inka history to social structure, there is no doubt that they havechanged Inka scholarship forever. At the very least, he introduced avariety of new ways (both critical and constructive) of looking at theavailable evidence. If the contrast between structure and history nowappe ars m isguided, at least it put structure on the agen da. The challengeis now to find ways of talking about structure that do not rule out localhistorical understandings, but instead make sense of the particular formthey tak e. In this light, the diarchy thesis becomes particu larly interesting,since it does no t presu m e to eliminate the chronology of kings altogether,only to truncate it into two contem porary dynasties.

    Andean DiarchyW hat do w e mea n by diarchy? This w ord has been used to characterizemany socio-political systems across the world, and presumably has some

    claim to generality.6

    Perhaps the most concrete definition comes fromHocart (1936: ch. 12), who identifies a pattern whereby a passive senior

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    11/33

    THE PAST IS A LOWER MOIETY 393overly restrictive and teleological, and prefer to treat diarchy moregenerically as an institutionalized du al division of pow ers . M any concretearrangements could conform to this minimal definition, and no usefulanalyt ic purpose would be served by creat ing an overly res tr ic t ivedefinition of diarchy that is just an abstract description of one particularcase. Rather, the point should be to identify, for comparative purposes, acertain range of variation among cases that display similar, but notnecessarily identical, morphologies and developmental tendencies. Forthe purposes of this discussion, I assume that there is always a tensionbetween s imilar i ty and complementary difference in the nature ofdiarchic powers, also a tension between symmetry and hierarchy in thewa y they are allocated. W hen these pow ers are nearly identical in nature,the inherently unstable condition of dual power results, but when theyare overly differentiated, the result is often no longer a division ofpowers, but an opposition between power and some other term, such as(religious) authority. The more similar these powers are, the more theirstable distribution requires hierarchy, whereas the more complementarythey are, the mo re symm etrically they can be allocated.The evidence for a loosely-conceived form of diarchy in the pre-Columbian Andes is overwhelming. Generally, diarchy was embedded inthe dual organizat ions or moiet ies that were so common in Andeansocieties. We have already encountered such a moiety division in thed is t inc t io n be tw een U pp er and Low er Cu zco . I w i l l a rg ue be low,however, that the Inka capital is a poor example of diarchy, and that thebest examples actually come from the so-called provinces, namely thepolities that the Inkas conquered in the creation of their empire. The non-Inka polities of the central A nd es frequently feature d du al territorialsubdivisions, each with their own paramount political office. So commonwere diarchic arrangements, and so central were they to the collection oftribute, that the Spaniards quickly came to a general working knowledgeof them:

    In each repartimiento or province there are two divisions: one which iscalled hanansaya [upper divis ion] , and the other hurinsaya [ lowerdivision]. Each division has a principal leader who rules the nobles andIndian s of his division, and does not interfere in the ruling of those of theother division, except the curaca [ruler] of the hanansaya division is thehead of the ent i re province, and the one whom the other curaca ofhurinsaya obeys in the things he says. The one from hanansaya has the

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    12/33

    394 PETER GOSEHe of hanansaya is the head of all, and he has authority over those ofhurinsaya. H e calls meetings and governs in general, even though he doesnot give particular orders. He collects tribute and pays it, because eventhough he does not personally collect from those of hurinsaya, he collectsit from their curaca or principal leader of the hurinsayas, who has collectedit from his ayllus. In this there is great cooperation, although sometimesthe most powerful, which is he of hanansaya takes some of the ayllos ofhurinsaya, and over this there used to be great dispute s, but those w ho arefrom the right understand well that each must have his ayllos, and thatthose of hurinsaya the head of hanansaya cannot have (Matienzo 1567:20-1, my translation).

    From this description w e see that the diarchy of the An dea n provincesinvolved a hierarchical division of similar powers between two rulers.The rulers of up pe r an d lower divisions had similar rights and d uties, butthe upp er ru ler outranked his lower counte rpart (see also Cobo 1653:112),and represented the unity of both divisions, or the polity as a whole, tothe outside world. Within the larger political unit, the ruler of the lowersubdivision would serve as the helper (yanapaq) or replacement {ranti) ofthe pre-eminent ruler of the upper subdivision. However, within theirow n respective subdivisions, each had h is own h elper or replacement, asRostworowski has shown (1983: ch. 5). Netherly (1993) shows that thesam e pattern prevailed on the northern coast. Thu s, du al sovereignty w assystematically an d recursively developed within m any And ean polities.These facts alone are enough to justify accepting diarchy as an aspect ofAndean pol i t ica l s t ructure . The quest ion is whether d iarchy was adominant principle, or whether it was subordinate to others. FollowingLvi-Strauss' (1963: 158) classic discussion of dual organizations, I willargue that in the Andean case, hierarchy lies behind the appearance ofdualism and symmetry.The first clue that diarchy was a subordinate principle can be found inits hierarchical nature. Indeed, diarchy was partially self-annullingbecause the ruler of the upper group encompassed the ruler of the lowergroup as the representative of the polity as a whole. In relation to otherpolities, these diarchies were one political unit and not two: from thisperspective their dualism was subordinated, and even suppressed. Pointof view determines whether or not Andean polities were diarchic. Froman internal perspective, diarchy was salient, and i t was possible tominimize the hierarchical nature of the division into upper and lower,

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    13/33

    THE PAST IS A LOWER MOIETY 395In the Cajatambo area of what is now north-central Peru, the divisionbe tween upper and lower d iv is ions was model led on the re la t ion

    between herders and agriculturalists. Broadly speaking, this occupationalcoding m akes sense, since herde rs occupy the up pe r ecological tiers of thealpine zone, whereas agriculturalists occupy the lower an d wa rm er valleyland. Tillers were called wari, after the tutelary ancestors in agriculture,whereas herders were cal led llacuaz, after a preferred technique ofsacrificing camelids (Duviols 1986: 500). The llacuaces were described asforeign conquerors who arrived more recently from Lake Titicaca,whereas the waris were depicted as conquered indigenous peoples, whowere nonetheless skilled in the arts of civilization, and settled theirnomadic overlords into a more sedentary way of life.7 The historicalvalidity of these traditions is questionable, as is the degree to which allmembers of each group actually engaged in their nominal occupation.However, it is undeniable that when people explained the origin of theirpolit ies through a mythic conquest of the lower group by the uppergroup, they also emphasized sectional hierarchy. For this conquestmythology took the hierarchy that otherwise might only be visible ininter-group relations, and implemented it as an image of the internalorganization of the political group. Although this did not negate therelations of ritual and economic complementarity that often prevailedbetween wari and llacuaz groups, it certainly must have coloured them.Gose (1993) has shown how the ritual complementarity of these groupsimplied the hierarchical superiority of the llacuaz pastoralist-foreigners,who were thought to control the water that wari tillers depended upon.But the main point is that these traditions of conquest, which present theu p p e r g r o u p a s c o n q u e r o r s a n d t h e l o w e r g r o u p a s c o n q u e r e d ,thoroughly dispel the notion of dual power at the core of the diarchythesis.Andean soc io-po l i t ica l dua l i sm was h ie ra rch ica l because i t wasenmeshed in broader segmentary processes that structured both theinternal organization of individual polities and their incorporation intoregional confederations and pan-Andean empires like that of the Inkas.Rather than being a formula that app lied only to the division of pow ers inthe polity, (asymmetric) dualism could exist at a level as minimal as themoieties of a hamlet, or expand to encompass the relation betweenm ountain peoples and coastal peoples within the An dean area as a whole(Zu ide m a 1962: 161) . Because it cou ld be ap p l ied a t va r io us

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    14/33

    396 PETER GOSEsystem of hierarchically embedded social groups (ayllus) in Cajatambo.Sometimes the two m oieties of a minimal ayllu segment w ere described aswari and llacuaz respectively, whereas at others, whole ayllu segmentswere identified as one or the other.8 In this way, the distinction betweenwari and llacuaz generated and cross-cut social groupings of differentmagnitudes. At any given level, the hierarchical encompassing of thelower half by the upper half already implied the next ascending level inthe segm entary system.It therefore seems likely that diarchy w as subord inate, both in logic andin practice, to segmentary hierarchy as a principle of socio-politicalorga nizat io n in the A nd es. Struc tural re la t iv i ty is the o uts tan di ngcharacteristic of segmentary organization. It could encompass diarchy,but also account for local socio-political organization (e. g. ayllu systems),and regiona l confederations of politices (e. g. the Ch ancas, the Collas, etc).During the late interm ediate period (1200-1400 A.D.), just before the Inkaem pire , ind iv id ua l An dea n po l it ies app ear to have expand ed andcontracted, mutated and disappeared at a bewildering rate. In such acontext, the flexibility of segmentary organization would have made it apar t icu la r ly usefu l id iom of conques t and a l l iance . Fur thermore ,segmentation took many overlapping forms: Before Inka conquest, it wasprobably expressed mainly through dual and tripartite territorial formsand hierarchies of local, regional, and pan-Andean shrines (wakas, seeDuviols 1978, Gose 1993). These idioms of segmentation were retainedunder the Inkas, who continued to encourage dual organization in theprovinces (Cobo 1653: 112, Sarmiento 1572: 110). However, the Inkassubsumed and rationalized these forms of segmentation under the so-called decimal system, which grouped the households of the empire intonesting units of 10, 50, 100, 500, 1,000, 5,000, 10,000 and 40,000. Ratherthan treat diarchy as something separate from these hierarchical forms oforganization, we should see it as a means by which they were realized.Such a conclusion should not be controversial, given the relations ofsegmentary encompassment of lower by upper described above. But it isprecisely as a romantic alternative to hierarchy that the notion of diarchyhas been elaborated in Andean studies. As a result, both Duviols (1973)and Rostworowski (1983) have tried to make segmentary dualism into amore sym metric phenom enon than it actually app ears to have been.9

    To further discuss wh ether diarchy w as a subordinate or superord inateprinciple of An dean governm ent, we mu st turn aw ay from the provincial

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    15/33

    THE PAST IS A LOWER MOIETY 397principle involved, then we might expect an essentially undivided formof sovereignty at the very " top" of the system. What w as the situation?

    Exponents of the diarchy thesis, particularly Duviols (1979a), haveargued that the Inka em pire had two co-rulers, each of wh om representedone of the mo iet ies of Cuzco. Alth oug h U ppe r Cuzco pro vide d theprimary or "unique" ruler (sapainka), he was always accompanied by a"subst i tu te" (ranti) or (as the Spanish put it) a "second person" whoshared with him the duties of imperial government. According to thed ia rchy thes is , th i s subs t i tu te was assoc ia ted wi th Lower Cuzco .Although the upper group outranked the lower group, this relation ofhierarchy was tempered by the notional symmetry of moiety division.Effectively, then, the diarchy thesis posits a link between the moietysystem of Cuzco and a kind of dual sovereignty represented by thecombined adm inistration of the uniq ue Inka and his substitute.I intend to refute the diarchy thesis by arguing (1) that the supremeInka was seen as the source of an undivided power that could bedelegated to a variety of subordinate substitutes, and (2) that the delegationof the supreme Inka's powers had little to do with the moiety system ofCuzco. Put differently, I challenge the existence of dual sovereignty forthe Inkas, and with it, the idea that imperial government was based onthe moiety system of Cuzco.

    Inka MonarchyIn fact there was only one supreme ruler, the sapa inka, who invariablycame from Upper Cuzco, at least during the proto-historical phase of Inkaimperial expansion. Although the supreme Inka had many "substitutes"who appear to have shared much of his administrative powers, they donot appear to have been associated with Lower Cuzco, as the diarchythesis would have it. Most accounts describe these "substitutes" as thegovernors of the four quarters of the empire or as the Inka's secretary andverbal intermediary.10 They were commonly recruited from among theInka's full or half brothers, and would therefore have been from UpperCuzco. Indeed, the only important member of the Inka eli te who isregularly associated with Lower Cuzco is the high priest of the Sun orwillaq umu, and th i s is p r i m ar i l y beca use the t em p le of the Sun(Coricancha) was located in Lower Cuzco. Cieza describes the willaq um u

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    16/33

    398 PETER GOSEthe high priest was generally a close relative of the Inka,11 and thereforethat he probably belonged to Upper Cuzco before assuming office.12Unlike other brothers of the sovereign, however, the ivillaq umu rarelyappears to have had persona l ambi t ions in wars o f success ion orusurpa t ion , bu t i s o f ten descr ibed as a tu te la ry f igure fo r youngsovereigns who were still unable to govern properly (Betanzos 1551:291-2, Sarmiento 1572: 253). Perhaps this is why Segovia (1553: 75-6)directly identifies the high priest as the Inka's "substitute". Far fromconfirming the diarchy thesis, however, this only confirms that the highpriest wa s not the sam e sort of figure as the Inka himself.The only evidence that suggest Lower Cuzco may have a polit icalr ep re sen ta t ive comparab le to tha t o f U ppe r Cuzco comes f romnegotiations over land rights around Cuzco in the early colonial period,wh ich Sherb ond y (1986: 51) describes w itho ut p rov idin g a reference.During these negotiations, the representative of Lower Cuzco was JuanTambo Uscamayta. Although this is not the do n Juan Tambo M aytapanacawho appears in Acosta's king list, Zuidema (1964: 87, 128n, 183) arguesthat the two were actually the same person. It is entirely possible thatAcosta could have mistakenly identif ied this individual with (Apo)Mayta Panaqa when he was actually a member of Uscamayta Panaqa.Alternatively, Zuid em a (1964: 209) argu es that a given pe rso n m ightrepresent different groups in different contexts. Even if we accept thisargument, it suggests that we are still dealing with a relatively minorfigure, a neighbourhood leader who was identified primarily with thesepanaqas, and not as a second sovereign. This inference is borne out in theappearance of "Don Johan Tambo Usca Mayta" as a 60 year old witnessbefore the Toledo inquiries into the histo ry of the Inkas in 1572 (Montesinos1882: 248). Here he appeared as one of two witnesses from the "Ayllo deMayta Ca pac", along with m any other w itnesses from other panaqas, all ofwhom ratified the monarchic view of Inka succession officialized bySarmiento (1572). It seems that at no point in these hearings did don JuanTambo identify himself as a second, co-reigning Inka, nor did othersidentify him as such. Thu s, the struc turalists ' prim e historical cand idatefor the second of the Inka diarchs not only missed a golden opp ortun ity toassert his claim to high office, he actually denied the very framework uponwhich such a claim could be based. It would be highly uncharacteristic ofthe Inka nobility to concede their privileges in this manner, given theirwell-documented willingness to resort to litigation in colonial courts.13

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    17/33

    THE PAST IS A LOWER MOIETY 399

    EMPIRECONFEDERACYETHNIC POLITYLOCALITY A

    ETHNIC POLITY 1

    C O N F E D E R A C Y 1 C O N F E D E R A C Y 2

    RULER E M P I R E

    \ = HIERARCHICAL PRE-EMINENCEFigure 2 Segmentary Dualism

    and apparently diarchic relationship typical of a provincial polity. In theprocess, Lower Cuzco may have found a neighbourhood representative,bu t contra ry to what Sh erbon dy (1986: 51) im plies, this hard ly confimsthat Lower Cuzco had an emperor figure comparable to the sapa inkaw hen it still formed the centre of an A ndean em pire.Moreover, there are good grounds for doubting that Lower Cuzcoought to have supplied a secondary imperial ruler according to the logicof A ndean diarchy. If diarchy functioned as Neth erly (1990: 464,1993:18)postulates, the paramount ruler of a polity was also associated with the"upp er" division of a series of lower segmentary levels, descend ing intoa particular locality. At each of these segmentary levels, the paramountruler would have a different replacement, who would be drawn from theopposite or "low er" division. As one mo ved u p the segm entary hierarchy,so the socio-political divisions involved became bigger. By the sametoken, the replacement was recruited from localities progressively moredistant from that with which the param ou nt ruler wa s associated. At thepenultimate level of the segmentary hierarchy, the replacement wouldcome from the opposite half of the realm, not the opposite half of thelocality from which the paramount ruler came (see Figure 2). This meansthat Lower Cuzco was the counterpart of Upper Cuzco only at the local,

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    18/33

    400 PETER GOSEWhen we leave behind the abstract logic of this system and turn to thehistorical record, however, it turns out that there was no significant dual

    division at the penultimate level of imperial organization. Rather, theempire was divided into four major territorial divisions, hence the titleTawantinsuyu (the unity of the four quarters) . Rostworowski (1983:176-7, 1988a: 229) has assembled the information from the chroniclers toshow that this territorial quad ripartition wa s echoed in governm ent. Eachqua rter of the realm had its representative in a governing council, andthey w ere each treated a s a substitute for the Inka within their jurisdictions.The evidence for this imperial council seems far more solid than that forInka diarchy. Yet i t has not spawned a quadrarchy thesis , probablybecause it is relatively clear that we are dealing with a case of delegated,not indep end ent po w er here. Still, it is entirely possible, as Guarnan Poma(1615: 365) suggests, that two of the representatives of each of the fourterritorial divisions may have been further identified with Upper Cuzco,and two with Lower Cuzco. It is, after all, very easy to nest quad ripartitioninto dualism. Nonetheless, to make such an argum ent, one wo uld have toshow that the Inkas themselves tried to make the division of Upper andLower Cuzco encompass that of the four quarters. It is possible that theydid so under certain circumstances, but in general, it appears that theydid not. When one looks at the empire's decimal system, it too featuresdualism at its lower levels but culm inates quad ripartition: two units of 50households make one of 100, two units of 500 households make one of1,000, and two units of 5,000 households make one of 10,000, yet fouruni ts of 10,000 make one of 40,000. I t seems that quadripar t i t ions u b s u m e s d u a l i s m , a n d n o t t h e r e v e r s e . F u r t h e r m o r e , th e r o a d sconnecting Cuzco to Tawantinsuyu were named after the four quarters ofthe em pire, not the tw o neighb ourho ods of the city. In short, dualism wa snot stressed in the maximal organization of territory and government inthe Inka em pire, which renders the diarchy thesis less probable.Because there is no direct evidence that Lower Cuzco provided asubstitute emperor, both Duviols (1979a: 78) and Rostworowski (1983:179) have blamed the Spanish chroniclers for failing to understand apolitical system that was so different from their own. Not only does thisargument beg the question of whether the Inkas really had a diarchy, itraises the far more serious question of why the chroniclers did recorddiarchy in the places where they saw it, namely the provincial polities.Had they not done so, then presumably we would not know enough to

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    19/33

    THE PAST IS A LOWER MOIETY 401government because i t was a key element to the indigenous tr ibute-collection system. The conclusion can only be that the chroniclers didun ders tand An dean diarchy in its broade r outlines, bu t failed to record itin imp erial Cuzco because they d id n ot encounter it there.Nonetheless, there are scattered references throughout the chronicles tovarious sovereigns contemplating a division of the realm into northernand southern domains. For example, Cieza (1553: 207, 220) suggests thatPachacut i enter ta ined th is idea , and that Huayna Capac consideredreviving it as a w ay of accom odating the kingly am bitions of both Hu scarand Atahuallpa. The proposed division would have distinguished themilitarily active northern periphery of the empire from its consolidatedsouthern core in a manner that accords well with the logic of Andeanmoiety systems. Thus, it seems that genuinely diarchic arrangements atthe imp erial level w ere conceivable, bu t that they took a different formthan that proposed by the structuralists, and ultimately did not carry theday.To recapitulate: The major representative of Lower Cuzco appears tohave been the high priest or willaq umu, but he was not necessarilyrecru i ted f rom among the people o f Lower Cuzco , though he wasassociated w ith them b y virtue of his occupation. Rather, he wa s a mem berof the elite of Upper Cuzco who changed his moiety affiliation uponoffice. Cobo claims that the willaq umu , like all lesser priests of the Sun,invariably came from the Tarpuntay Ayllu (1653: 224), which Sarmientolocates in Upper Cuzco (1572:119). In short, the claim of Lower Cuzco tosacerdotal supremacy was vaild only at a categorical level, and hadnothing to do with actually staffing the priesthood, where Upper Cuzcoonce again held de facto control. If the notion of diarchy is supposed toimply a duality or balance of power, then it clearly does not apply to thissituation. Everything indicates that the royal descent groups of UpperCuzco manipulated and controlled the only significant position allocatedto Lower Cuzco in the Inka court , that of the high priest . We musttherefore accept that Andean socio-political dualism was a veiled form ofsegm entary hierarchy that culm inated in m onarchy, not diarchy.

    Secula r Power And Tra nscendenta l Authority?The question now becomes how to characterize the relation betw een the

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    20/33

    402 PETER GOSEconsidered thus far, but it does not hold up under closer scrutiny. First,there was no c lear d ivis ion of labour a long secular / re l ig ious l inesbetween the sup rem e Inka and the high priest of the Sun. The chroniclesrepeatedly mention that the Inka performed important priestly activitiesin a variety of imperial rituals, during which he neglected the rest of hisgovernin g du ties (e.g. Betanzos 1551: 51-2, Cieza 1553: II, ch. 7). Like allof the Inka's substitutes, the high priest was essentially a stand-in whoacted on behalf of the Inka. It follows that the Inka could reclaim thepriestly pow ers he delegated to the willaq umu whenever he wished. Thusit w as said that w hen H uay na Capac took office, the first thing he did wa sto remove the high priest of the Sun and take the job on himself, vistingall the shrines, oracles and their estates (Sarmiento 1572: 238). Similar, ifslightly less dramatic, intrusions on the role of the willaq um u have beenreported for other sovereigns (Sarmiento 1572: 191-2). Even when hisrelations with the willaq um u were cooperative, the Inka would routinelyjoin or replace him in his priestly role, particularly when it came tooracular consultations of the Sun. Furthermore, the willaq um u couldassum e the non-priestly duties of the Inka un der certain circumstances, asdu ring the rebellion of Manco Inca, wh en he played an im portant m ilitaryrole (Betanzos 1551: 292, 299, cf. Sarm iento 1572: 253). In short, the re wasno entrenched distinction between secular and religious authority in thissystem, and the sovereign not only outranked the high priest in general,but also within his priestly domain. For these reasons, the Andean casedoes not conform to the separation between worldly power and spiritualauthority decribed for other civilizations by Dumont (1970) and Bloch(1977). Its political life was not secularized, nor was its religious lifetranscendentalist.

    To better understand the ruler-priest relation in the moiety system ofCuzco, we m ust retu rn to our previou s discussion of divine kingship . Wehave already seen that the core notion of An dean divine kingship w as thet ransformat ion of a l iv ing warr io r in to a dead de i ty . With in th isdevelopmental progression, the upper moiety represents the living phaseof a sovereign's career dedicated to military expansionism, whereas thelower m oiety represents his end urin g po sthu m ous influence as a deity, inwhich the fertility of his land and people was the primary concern. Asimilar progression existed in the lives of the Inka male elite, whoseinitiation into manhood, the warachikuy ritual, was decidedly military incharacter. Their most honourable option was to pursue a military career

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    21/33

    THE PAST IS A LOWER MOIETY 403whereas the more ancient kings of Lower Cuzco founded the solar cultbu t d id no t a t t e m p t to su bd ue the i r non - Inka ne i gh bo urs (cf.Ros tworo wsk i 1983: 161). The relation betw een the king an d the highpriest , as the respective representatives of Upper and Lower Cuzco,conformed to this same pattern. The king was expected to personify thevirtues of youthful vigour and military expansionism, at least initally,whereas the high priest was dedicated to consolidating the ancestral cultsand the fertility of the empire as it already existed.The military orientation of Upper Cuzco has been demonstrated byPease (1981): most of the "captains" of the Inka army were recruited fromthis division, and many were half-brothers of the sovereign. 14 Thesemilitary specialists appear to have been semi-charismatic leaders, aroundwhom congregated an entourage of lesser fighters recruited through thelab ou r-trib ute syste m (Segovia 1553: 74). Together w ith th e ruler, hisbrothers and sons, these "captains" formed a cohort with distinct butstrongly converging interests in pursuing imperialist warfare. For theruler, adding new territories and overcoming all rivals was the primarymeans of showing his divinity, and key to how he would be rememberedby posterity. More to the point, few Inka rulers could achieve election totheir office without a successful military career under their fathers, giventhe rule of succession by the "most able" of the incum ben t's sons.15 For theruler's brothers and other "captains", military prowess brought booty, aluxurious lifestyle, and perhaps the reward of a chosen woman (aqlla)from the Inka for especially m eritorious service (see Betanzos 1551: 43,Cobo 1653: 253). Since a military career m ad e these m en responsible forthe expansion of the Inka state, they also had the prospect of beingappointed governors or lesser administrators of the terri tories theysubdued. There was no c lear d is t inct ion between the army and theadministrative bureaucracy of the Inka state, primarily because of therelentless expansionism of the Inka elite and the chronic rebelliousness ofthe provinces, particularly on the death of Inka kings. Upper Cuzco'sclose association w ith the arm y th us became a significant m eans by w hichit monopolized imperial pow er, and negated the possibility of diarchy.The more quietist, ancestral orientation of Lower Cuzco is particularlyevident in its association with the Sun, supreme tutelary deity of theInkas, an d father of all their sovereigns. The Sun represented the u ltimatesource and repository of the political authority of each Inka ruler, thedeindividuated expression of the permanent influence of the Inkas as a

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    22/33

    404 PETER GOS E106). Some mummies even appear to have been stored with Punchao inthe temple of the Sun. During solar rituals, the royal mummies were notonly present, but received the same sacrifices as the image of the Sun inthe plaza of Cuzco (Betanzos 1551: 166, Cieza 1553: II, ch. 30, Segovia1553: 81-2). Althou gh the individua l exploits of the royal mum m ies w erenot forgotten, no r did they cease to be the focus of partic ular descen tg roups , the ove ra l l pa t te rn i s c lea r . Upon dea th , the aggress ive ,expansionist individuality of the sovereign was arrested, and thereafterhe becam e increasingly identified w ith the undifferentiated ancestralconcerns of the Inkas in general, of which the Sun was the primaryexpression. The Sun certainly was not indifferent to the politico-militarypow er of Upp er Cuzco, otherwise this deity wou ld no t have continued toact as king-maker through the oracular medium of the high priest .Nonetheless, it was primarily concerned with the priestly and agrarianactivities of Low er Cuzco, no t run nin g th e affairs of state.The association of dead sovereigns with agriculture is suggested in thevery semantics of the Quechua term mallki, which denoted not onlymummified bodies, but also trees, saplings, and crops. Once dead, itseems that these exem plary ancestors becam e increasingly identified withthe fertility of the land, as often happens in divine kingships. Thus, thepriestly ayllu from which the high priest of the Sun was recruited wasca l led Tarpuntay , which in Quechua , means "p lanter" or "sower" .Therefore, the priestly activities associated with the cult of the dead werealso represented as agricultural. In And ean terms, it is entirely logical thatthese agriculturally-oriented ancestors should also be associated with thelower territorial division. For in the du al organizations analyz ed byDuviols (1973), the lower group always is identified with agriculture byvirtue of its association with the warmer lowlands. It is primarily as asource of warmth that the Sun, too, came to be associated with the lowergroup. However, these lower groups are also depicted as having beenconquered by roving pastoral militarists of the upper division. The samepattern is repeated is repeated in Cuzco, where the lower terr i torialdivision, the ancestral rulers, and even the priests who maintained theircults, were often represented as conquered autochthons (Zuidema 1964:150-2, 161, 206, 1990a: 9). U nd ou bte dly the historica l reality b ehin d thesocial composition of Lower Cuzco w as far m ore complex th an this, sincewe know that the Tarpuntays were part of the Inka el i te , and weresom etimes classified as part of Up pe r Cuzco (see Sarm iento 1572:119). By

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    23/33

    THE PAST IS A LOWER MOIETY 405diachronic dimension of this opposition resolves itself into a process, inwhich the upper imperceptibly slides into the lower over time, much ashas been described for eastern Indonesia (Traube 1989).This same temporal passage from high to low is particularly well-expressed in the moiety affiliations of the succession of Inka rulers. Thechroniclers generally assign the first five Inka ru lers to Lower Cuzco , andthe remainder to Upper Cuzco . What th is means i s tha t when thedenizens of Upper Cuzco looked down the hillside into Lower Cuzco,they were also looking "back" (as we would say) in time. The moietydivision between Upper and Lower Cuzco was not merely a verticalgeographical distinction, but also a temporal distinction between twodifferent ages or phases of the Ink as' career as a people. In their accounts,the chroniclers make the substance of this temporal distinction clear. TheInkas were an in t rus ive people in the Cuzco a rea who, under theleadership of Manco Capac, took up residence around the site of whatwas to become the temple of the Sun in Lower Cuzco. There they livedamong the other ethnic groups of the area, practicing agriculture andpriestly service to the Sun. Although some accounts describe MancoCapac organizing Cuzco into upp er and lower m oieties at this stage, mostattribute the founding of Upper Cuzco to the latter reign of Inca Roca,who is said to have led Inka expansion into the higher reaches of thevalley that were to comprise that territorial division. Other accounts relatethe differentiation of Up per from Lower Cuzco to the subseq uen t reign ofPachacuti, who established the preeminence of the Inkas in the Andeanarea by winning a mythical war with the Chanca, dur ing which theincumbent rulers of Lower Cuzco neglected the defense of the city, anddedicated themselves to priestly and recreational activit ies on theirestates in the countryside outside Cuzco.16 What matters is not whichsovereign "really" founded Upper Cuzco. Far more significant is the factthat that the distinction between Upper and Lower Cuzco is consistentlyrelated to significant episodes in Inka imperialism and the ascendancy ofthe m ilitary in politics (cf. Pe ase 1981: ch. 1). The fou nd ing of U pp erCuzco, it seems, signifies the supremacy of the Inkas over other Andeanpeoples. The location of this event in "Inka history" shifts, and with it themoiety affiliation of the variou s sovereigns, as new thresho lds of conquestare attained.

    A s imi la r t ransfo rm at ion of the up pe r in to the lowe r m ay h aveoperated on a shorter cycle, with each succession to high office. Zuidema

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    24/33

    406 PETER GOSEinto the palace complex of Upper Cuzco. The overall effect of this cyclewould have been that mummies and their retinues circulated out of thepalaces of Upper Cuzco, while heirs and their factions circulated in.Elsewhere, Zuidem a (1989: 257) suggests that those w ho d escen ded frommarriages with non-Inka women lived in Lower Cuzco, whereas thosewho descended from marriages with Inka women lived in Upper Cuzco.The overlap between these two scenarios was great, since many of theInka's subsidiary wives were non-Inka, and those born of Inka womenremote from reigning sovereign were probably undergoing a subtle driftin to per ipheral s ta tus within the e l i te of Cuzco. In e i ther case , thesubs id ia ry wives and ch i ld ren , who were t rad i t iona l ly thought tocomprise the panaqa of each ruler, would always have lived in LowerCuzco , even dur ing the l i fe o f the sovere ign . In th is sense , theyrepresented the ult imate destiny of their king, a kind of downwardexfoliation that began in his life, and con tinued into dea th.In summary, the Inka moiety system encodes not only a geographicaland an occupa t iona l d i s t inc t ion be tw een mi l i t a r i s t ru l e r s andautochthonous priests, but a developmental cycle for the careers of Inkasovereigns, which is echoed in an historical mythology of the Inka peopleas a whole . Zuidema suggests that in Inka mythology there was aconsistent tendency to represent Lower Cuzco as the descendants oftimid, senile, original rulers who were incapable of defending theirdomain, and lost power to younger and more vigorous upstarts , whoformed the ruling moiety of Upper Cuzco (1964: 111-3, 138. 156-66). Thiswas not just a matter of the Inka conquest of the original inhabitants ofthe Valley of Cuzco, but of periodic dynastic renewal within the Inkaruling class itself, as in the mythical war with the Chanca (see Zuidema1964: 231). Above all, this mythology speak s to the w ars of succession thatusually followed the death of each Inka sovereign (see Rostworowski1960). These struggles were the outgrowth of intrigues among the royaldescent groups of Upper Cuzco, and an integral part of the Andeannotion of succession by the "most able" of the dead sovereign's potentialheir s (Rostworowski 1983:1549). In add ition, provinc ial rebellions on thedeath of a sovereign obliged the successor to reconquer various parts ofthe realm. Thus, every new ruler necessarily came to power through adisplay of force. Far from being opposed to legitimate authority, as inwestern social theory, this force was a sign of the incoming ruler's solarpedigree, and hence, his fitness to rule. Political conflict and instability

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    25/33

    THE PAST IS A LOWER MOIETY 407youth, vitality and power as the supreme good (cf. Valeri 1990a: 47-50).Unlike the Hawaiian and Tongan diarchies described by Valeri (1990a),however, the Inkas did not temper their admiration for these values byranking them below an immobile centre linked to the past. Thus, therewas not even an incipient sense in which the Inkas approached a Hindu-Christian distinction between secular politics and religious transcendence.Instead, the sa l ient d is t inct ion was between immediate and remotepha ses of sovereignty, in wh ich the former set the stand ard s for the latter,in keeping with the overriding in-worldly emphasis on vitali ty andpower. Dead sovereigns were wo rshippe d only to the extent that they hadconquered in life, and their divinity was expressed primarily throughoracular communication, the fertility of their land and of their descentgroup: signs that proved their exemplary abundance of life even in death.This was a system that celebrated its own kind of history, but with arelentlessly presentist orientation towards life and power in the here andnow. Hence, the current sovereign from Upper Cuzco was always thepara m oun t reference po int, even in the ranking system.Taken toge ther , the two moie t ies o f imper ia l Cuzco coheren t lyexpressed the expansionist character of Inka divine kingship. UpperCuzco represented the current growth phase of the empire and thosecompleted by recently dead kings, whereas Lower Cuzco representedprevious, more consolidated episodes of growth that took place in thedistant past. This process was very much like the growth of a tree, inwhich the living outer growth layer is constantly reabsorbed into thedead structural core in a series of growth rings, each of which wasequ ivalen t to the reign of an Inka k ing . Inevitably, the m oiety affiliation ofparticular kings had to be revised periodically as the growth of theempire proceeded. Because this revision was an ongoing process, like thegrowth that drove it, the chroniclers do not entirely agree on the moietyaffiliations of the series of sovereigns runn ing from Inca Roca to Pachacu tiInca Yupanqui. How ever the disagreement is not pa rticularly serious, andhardly threatens the basic credibility of our sources, as has sometimesbeen su gg es ted . The m a in i s sue a t s t ake is w he the r Inka mo ie tydistinction was established by the consolidation of their control over theupper reachers of Cuzco (Inca Roca), or the beginning of their expansioninto other areas (Pachacuti). In the first case, Inka moiety division wasliterally mapped onto the geography of Cuzco, whereas in the second, itw as detach ed from it, and linked to a spatially m ore abstracted process of

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    26/33

    408 PETER GOSEwell expressed by the ag rarian o rientation of Lower Cuzco. It w as as if animpulse to grow that originated in the agriculture and fertility rituals ofLower Cuzco somehow naturally culminated in the aggressive militaryexpansionism of Upper Cuzco. Here, Upper Cuzco became like a shootlead ing the g rowth of a p lan t , a lways a t the cu t t ing edge of i t sdevelopment. This metaphorical connection between military expansionand plan t grow th was fundam ental to An dean ideologies of d ivinekingship, in which a sovereign/ancestor typically achieved defication byexpanding the boundaries of the polity in war, or those of the agriculturalfrontier with in his do m ain. By a similar logic, the Inkas linked agricu ltureto wa r in the trium pha l m ilitarism of the haylli songs that they sung aftervicto ry in w ar (Santa cruz Pac hac uti Yam qui 1613: 193, 229), b ut alsodur ing the sowing , somet imes wi th the par t ic ipa t ion of the Inkasovereign himself (Guarnan Pom a 1615: 250-1). Argua bly it is w ar thatmetaphorically encompasses agriculture in these songs, just as the upperdivision generally encompasses the lower in Andean social thought. Yetthere can be little denying that here, the more pervasive image of growththat allows these domains to be linked comes from Lower Cuzco. FromLower Cuzco's point of view, Upper Cuzco was a teleological outgrowthof its own agrarian concerns. From Upper Cuzco's point of view, LowerCuzco w as a vestige of its own distant p re-military past. These interactingvantage p oints can be described in the language of complementarity, butperhaps more significant was the way they created transformationaldiachronic processes (cf. Valeri 1990a: 58), an d an ove rw helm ing im petu stowards imperial expansionism.

    ConclusionFrom the foregoing, it is clear that the Inkas d id not p ractice diarchic rule,i f by tha t we mean two s imul taneous ly re ign ing sovere igns , eachassociated with one of the dual divisions of imperial Cuzco. However, anasymmetric form of diarchy did prevail in the provinces as a veiled formof segmentation, in which upper divisions encompassed lower ones tocreate higher levels of organization. As the apex of this segmentarypolitical organization, Inka Cuzco was the one exception to the rule ofdiarchy, because i t was the one level of organizat ion that was notsurpassed by higher one. The fact that th is system culminated in a

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    27/33

    THE PAST IS A LOWER MOIETY 409The fact that the Inkas did not practice diarchy thus conceived does notmean that the dual organization of their capital was a hollow shell ,

    however. Rather, it embodied a number of important historical ideas anddiachron ic p rocesses . Upper Cuzco was assoc ia ted wi th the l iv ingsovereign, the military, and the growth of the empire, whereas LowerCuzco represented past sovereigns, priests, and conquered agriculturalists.Ove r time, rulers w ho had once been associated with the expansionism ofUpper Cuzco might be reclassified as belonging to Lower Cuzco, as theirdeeds were surpassed by subsequent sovereigns. Historical concernswere not alien to the Inkas, but they were constantly being assimilatedinto the dual o rganization of the capital, in which pa st epochs were equatedwith the lower moiety. This does not mean that the commemorativetraditions associated with each individual sovereign were necessarilyeffaced as the structural position of each ruler was modified over time.Rather , the s t ruc ture in to which they were ass imi la ted was i t se l fdiachronic, and therefore cannot be opposed to history in the manner ofearly structuralism. The moiety system of Cuzco was based on a growthmetaphor tha t worked i t se l f ou t in t ime and space th rough Inkaimperialism.Finally, this paper has argued that our understanding of the Inkas isbest advanced through a comparative perspective that includes moregeneral discussions of such issues as divine kingship, segmentation, anddual organization. Here, the attempt to turn diarchy into a regionalgatekeeping concept that defines a uniquely Andean form of governmenthas had the ironic effect of obscuring our understanding of Inka imperialgovernment, and how it was distinct from the provincial polit ies i tsubordinated. Like any other concept intended to monolithically typify agiven area , d iarchy fa i ls in the Andes. As par t of a more f lexibleconceptual repertoire, however, diarchy rem ains a useful notion, especiallyto the extent that it does not fully apply to the Andean case, or meanssomething slightly different there than it does in Polynesia or easternIndonesia. Ultimately, it may be possible to fix a m ean ing to this term , justas it may be possible to specify the peculiarities of Andean politicalorganization from a comparative perspective. In the meantime, it is bestto resist premature closure.

    Notes

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    28/33

    410 PETER GOSEAndean about this civilization. Following Murra (1980), many have argued that Inkapolitical economy w as based o n a uniq ue form of redistributive com plementarity thatwas neither tr ibutary or mercanti le. Although this part icularis t orientat ion hasoverseen a remarkable growth in our understading of the Inkas, it has long sinceacquired ideological pro portions . For exam ple, Murra (1987: 53) and Pease (1992: 11)arbitrarily claim that the Inkas lacked tribute because they did not collect tribute inkind but rather in labour. In the same apologetic vein, both Pease (1981: 63) andRostworowski (1988a: 15) deny that the word empire is applicable to the Inkaconquest s tate, and highlight i ts supposedly benign and organic character byreferring to it as Tahuantinsuyu, the unity of the four quarters. It is predictable thatthis emphasis on historical particularity would ultimately be reclaimed as a romanticcharter for modern nationalism. Ironically, however, Murra's initial attempts todiscover the Andean were based upon a relatively orthodox version of substantivisteconomic anthropology, whereas Zuidema's was based on Leiden structuralism. Asusual , there turns out to be a metropoli tan vision at the bottom of at tempts tocelebrate the un iqueness of the periphery.

    2. Similar arguments have recently been advanced as a general feature of socio-politicaldua lism on a w orld-w ide basis (see Maybury-Lew is 1989: 14).3. Zuidem a does not elaborate on the nature of this connection between sovereigns andpanaqas. One possibility is a system of circulating succession, comparable to thosewhich existed in Africa. Although Cieza (1553: 183) suggests a possible alternatingsuccession between Upper and Lower Cuzco, the improbability of this scenario in theInka case is suggested by Good y's general comm ents on circulating succession: that itrepresents an attempt to avoid wars of succession and prevent any one sectoralinterest group from monopolizing central office (1966: 142). We know that wars ofsuccession were the Inka norm, and that entire panaqas could be wiped out in them. Itis therefore extremely unlikely that anything so high-minded as a rule of circulatingsuccession existed among the Inkas. If such a rule existed, it was routinely breached.4. I am grateful to John Rowe for draw ing m y attention to these last two points.5. Note that Rostworowski treats Cusco Panaqa and Socso Panaqa as different groups,whereas they are probably different orthographies for the same group: she also omitsSuma Panaqa from her discussion.6. See va n W ou de n (1935), G ue rm on pre z (1990), Tra ub e (1989: 342), Valeri (1990a,1990b, 1991).7. See Arriaga (1621: 24, 117-8) an d Du viols (1986: 11, 52, 55, 60, 94, 120). There arecases, however, where wari mummies are described as "los primeros conquistadoresy fundadores" of a certain locality (see Duviols 1986: 59, 224, 428), which suggeststhat waris were considered to once have been conquering rulers, prior to their ownconquest by intrusive llacuaz groups.8. See Duviols (1986: 52, 59, 89-90, 202-3, 245, 343, 479-81, 486-8, 489-91, 497).9. Note tha t bo th Duviols (1979a) and Rostworowski (1988a: 183) implicitly revised theirearlier formulations, and later recognized the hierarchical nature of socio-politicaldualism in the pre-Columbian An des.10. See Santillan (1553: 15-6), Sarmiento (1572: 258-9), Guaman Poma (1613: 111 , 184),and Molina (1528: 280), also Betanzos (1551: 271) and Segovia (1553: 80-1).

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    29/33

    THE PAST IS A LOWER MOIETY 41113. Duviols (1979a: 78) claims that the Inka nobility had no interest in pressing suchclaims because the Spaniards did not understand diarchic government. However,there are early colonial law suites that successfully invoke Andean criteria (e.g. thecapacocha ritual) that would have been far m ore esoteric and problematic than diarchyfrom a Spanish point of view (see Rostworowski 1988b). Furthermore, one couldargue that those Inka nobles seeking self-interested accommodation with theSpaniards would have found considerable advantage in asserting the notion ofdiarchy during Manco Inca's revolt, when the Spaniards would have welcomed analternative 'traditional' sovereign.14. No te that Zu idem a (1964: 191-2) argu es that these military "ca ptain s" were in fact thesubsidiary sons of a ruler who did inherit high office, but instead, formed the ruler'spanaqa on death. At this stage in their lives, these captains became priestly figureswho were more or less identified with and dedicated to the service of their father'sm um m y (see Polo 1571: 124). This conforms to the transitio n from high to lowproposed below in the text.15. That a display of military might had to precede a successful administration is wellillustrated in Betanzos' account of the succession of Pachacuti Inca Yupan qui by TopaInca Yupanqui (1551: I chs. 33-6). Upon news of Pachacuti Inca Yupanqui's death,first Antisuyo, then Collao, rebelled against Inka rule, and Topa Inca personally ledthe armies that quelled these uprisings, leaving Yamque Yupanqui, his elder andmore administratively inclined brother, in Cuzco to govern. When Topa Inca defeatedthe rebels in Antisuyo, he brought their insignia to Yamque Yupanqui and beggedhim to tread on them, a prerogative that normally fell to the sovereign. Yamque

    Yupanq ui accepted this honou r, but died of old age shortly afterwards. N ot only doesthis recall Rostworowski's arguments about co-reign brothers (1960), it specificallydemonstrates the idea that the successor should prove himself militarily beforeassuming all the trappings of sovereignty, even if he had already been elected andinstalled, as had Topa Inca. By these standards, the antithesis of the good sovereignwas the one (like Inka Urcon or Huascar) who immediately dedicated himself todrink , wom anizing and fine dre ss up on installation in office (see Betanzos 1551: 207,238).16. Mura (1613: 52) writes that the differentiation of Upper from Lower Cuzco wasestablished under Manco Capac, but elsewhere he attributes the founding of Upper

    Cuzco to the latter reign of Inca Roca, who is said to have led Inka exp ansion in to thehigher reaches of the valley that w ere to com prise Upp er C uzco (M ura 1613: 69).Cobo (1653: 72) also reproduces these two different accounts of the founding ofUpper Cuzco, and notes explicitly that they contradict each other. Although (Polo1585: 10) and Sarmiento (1572: 145) also identify Inca Roca as the first ruler of U pp erCuzco, who founded it by virtue of territorial conquest, Sarmiento (1572: 182) laternotes that the division between Upper and Lower Cuzco was made, or perhapsremade, in the subsequent reign of Pachacuti. Betanzos (1551: 77) also states that thedistinction between Upper and Lower Cuzco was established by Pachacuti. In sodoin g, he implies that Inca Roca belonged to Lower Cuzco , in contrast to the pre viouschroniclers. This is clear in how Betanzos (1551: 77) assigns Vicaquirao (Inca Roca'spanaqa) to Lower Cuzco (Pumapchupa), whereas Sarmiento (1572: 145-6) puts it in

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    30/33

    412 PETER GOSEReferences CitedAcosta, Jos de (1590) [1987],Historia natural y moral de las indias. M adrid: Historia 16.Anonymous Jesuit c. (1590) [1968], Relacin de las costumbres antiguas de los naturales dePir, pp . 153-177 in Biblioteca de AutoresEspaoles, tomo 209. M adrid: Ediciones Atlas.Aranibar, C. (1969-70), Notas sobre la necropompa entre los Incas. Revista del MuseoNacional 36, 108-142.Arriaga. P . J. de (1621) [1968], Extirpacin de la idolatra en el Pir, p p. 191-277 in Bibliotecde AutoresEspaoles, tomo 209. M adrid: Ediciones Atlas.Betanzos, J. de 1551 [1987]. Suma y naracin de los Incas. M adrid: Ediciones Atlas.Bloch, M. 1977. The Disconnection between Power and Rank as a Process: An Outline ofthe Development of Kingdoms in Centra l Madagascar . Archives Europenes deSociologie 28, 107-148.Cieza d e Len, P. de (1553) [1984], La crnica del Per, Parts 1 and 2, in O bras Com pletas, 1.Madrid: Consejo Superior de Invest igaciones Cient if icas , Inst i tu to GonzaloFernndez de O viedo.Cobo, B. (1653) [1956],Historia del nuevo mundo. Madrid: Biblioteca de Autores Espaoles.Dumont, L. (1970), Homo Hierarchicus. Chicago: University of Chicago Press.Du viols, P . (1971), La lutte contre le s religions autochtones dans le Prou coloniale (L'extirpade l'idolatrie entre 1532 et 1660). Lima: Institut Franais d'Etud es And ines.Duviols, P. (1978), Un Symbolisme A ndin du Double: La Lithomorphose de l 'Anctre.Actes du XLIIe Congrs International des Amricannistes, 4, 359-64.Duviols, P. (1979a), La Dinasta de los Incas: Monarquia o Diarquia? Argumentosheursticos a favor de una tesis estructuralista. Journal de la Socit des AmricanisteLXVI: 67-83.Duv iols, P. (1979b), Un Sym bolisme d e l'Occupation, d e l'Am nagem ent et de l'Explotationde l'Espace: Le Monolithe 'Huanca' et sa Fonctione dans les Andes Prhispaniques.L'Homme 19/2: 7-31.Duviols, P. (1986), Cultura Andina y Represin: Procesos y Visitas de Idolatras y HechicerSiglo XVII.Cuzco: Centro de Estudios Rurales Andino s Bartolom de las Casas.Garcilaso de la Vega, I. (1609) [1960], Comentarios reales de los Incas. Madrid: Biblioteca deAutores E spaoles.Goody, J. (1966), Circulating Succession among the Gonja, pp. 142-176 in (ed). J. GoodySuccession to High Office. Cam bridge: Cam bridge University Press.Gose, P. (1993), Segmentary State Formation and the Ritual Control of Water under theIncas. Comparative Studies in Society and History 35 /3 : 480-514.Gu am an Poma de A yala, F. (1615) [1963]. Nueva Cornica y BuenGobierno. Paris: Universitde Paris.Guermonprez, J. F. (1990), Dual Sovereignty in Nineteenth-Century Bali, pp. 189-212 in(ed.) J. C. Galley Kingship and the Kings. New York: Harw ood A cademic Publishers.Gu tierrez d e S anta Cla ra, P. (1603) [1963-4]. Historia de las guerras civiles del Per. (1554-8Biblioteca de Autores Espaoles, tomos 165-7. Madrid: Ediciones Atlas.Lvi-S trauss, C. (1963), Structural Anthropology. N ew York: Basic Books.Lvi-S trauss, C . (1966), The Savage Mind.London: Nicholson and Weidenfeld.MacCormack, S. (1991), Religion in the Andes: Vision and Imagination in Early Colonial Per

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    31/33

    THE PAST IS A LOWER MOIETY 413Montesinos, F . de (1882), Memorias antiguas historiales y polticas del Per, por el licenciado DFernando Montesinos, seguidas de las informaciones acerca del seoro de losIncas hechamandado de D. Francisco de Toledo, Virey del Per. M adrid: Imp renta d e Miguel Ginesta.Murra, John (1980), TheEconomic Organization of the IncaState. Greenw ich: JAI Press.Murra, John (1987), Existieron el tributo y los mercados antes de la invasin europea?pp . 51-61 in (eds.) O. Harris, B. Larson and E. Tan deter La participacin indgena en losmercados sttrandinos. La Paz: Ed iciones CERES.M ura , M. de (1613) [1987], Historia general del Pe r, Madrid: H istoria 16.Netherly, P. (1990), Out of Many, One: The Organization of Rule in the North CoastPoli t ies , pp. 461-487 in (eds.) M. Moseley and A. Cordy-Coll ins The NorthernDynasties: Kingship and Statecraft in Chimor. W ashington D. C. Dum barton Oaks.Netherly, P. (1993), The Nature of the Andean State, pp. 11-35 in (eds.) J. Henderson andP. Netherly Configurations ofPower: Holistic Anthropology in Theory and Practice. IthacCornell University Press.Pease, F. (1981), Los ultimos Incas del Cuzco. Lima: Villanueva.Pease, F. (1992), Curacas, reciprocidad y riqueza. Lima: Pontificia Universidad Catlica delPer.Pizarro, P. (1571) [1978], Relacin del descubrimiento y conquista de los reinos del Per. Lima:Pontificia U niversidad Catlica del Per.Polo de O ndega rdo , J. (1585) [1916], De los errores y sup ersticiones de los indios, sacadasdel tratad o y averiguacin que h izo el Licenciado Polo, pp . 3-43 in Informaciones acercade la Religin y Gobierno de los Incas, 1. Lima: Imp renta y Libreria Sanmarti.Polo de Ondegardo, J. (1561) [1940]. Informe del Licenciado Juan Polo de Ondegardo alLicenciado Briviesca de M uatones sobre perpetuida d de las ecomiendas en el Per.Revista Histrica 13, 125-200.Rostwo rowski de D iez Canseco, M. (1960), Succession, Coop tation to K ingship, and RoyalIncest among the Inca. Southwest Journal ofAnthropology 16, 417-427.Rostworoski de Diez Canseco, M. (1983), Estructuras andinas del poder. Lima: I. E. P.Rostworoski de Diez Canseco , M. (1988a), La historia del Tahuantinsuyo. Lima: I. E. P.Rostworoski de Canseco, M. (1988b), Conflicts over Coca Fields in 15 th-century Per. AnnArbor: Museum of Anthropology, University of Michigan.Sah lins, M. (1985), Islands of H istory. Chicago : Unive rsity of Chicago Press.Santacruz Pachacu ti Yam qui, J. de [1927], Relacin de Antiguedades deste Reyno del Piru, pp.125-235 in (ed.) H . Urteaga Coleccin de Libros y Documentos Referentes a la Historia dPer,tomo 4. Lima: Imp renta y Librera Sanm arti y Ca.Santilln, Hernando de (1553) [1927], Relacin del origen, descendencia, politicay gobierno dlos Incas, in (ed.) H . Urteaga Historia de losIncas y Relacin de su Gobierno, Coleccin deLibros y documentos referentes a la historia del Per, 9. Lima: Imprenta y LibreriaSanmarti.Sarmiento de Gamboa, P. (1572) [1842], Historia de los Incas. Buenos Aires: BibliotecaEmec.Segovia, Bartolom de (1553) [1968], [formerly attributed to Cristbal de Molina (elAlmagrista)]. Relacin de muchas cosas acaescidas en el Per. pp. 56-95 in Biblioteca deAutores Espaoles 209. Madrid: Ediciones Atlas.Sherbondy, J. (1986), Los Ceques: Cdigo de canales en el Cusco Incaico. Allpanchis 27,

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    32/33

    414 PETER GOSEValeri, V. (1990b), Dea ths in Heaven: M yths and Rites of Kinship in Tongan Kingship, pp .213-251 in (ed.) J. C. Galley Kingship and the Kings. New York: Harwood AcademicPublishers.Valeri, V. (1991), Afterword to S. Lansing Priests and Programmers: Technologies of Powethe Engineered Landscape of Bali. Princeton: Princeton U niversity Press.van Wouden, F. A. E. (1935) [1968], Types of So cial Structure in Eastern Indonesia. TheHague: M artinus Hijhoff.Zuidema, R. T. (1962), The Relationship Between Mountains and Coast in Ancient Peru,p p . 156-165 in The Won der of Man 's Ingenu ity. Leiden: Publication of the StateM useum of Ethnology.Zu idem a, R. T. (1964), The Ceque System of Cuzco. Leiden: Brill.Zu idem a, R. T. (1989), The Moieties of Cuzco , pp . 255-257 in (eds.) D . Maybury-Lewis andU. Almagor The Attraction of O pposites: Though t and S ociety in a Dualistic Mode. AnnArbor: Unive rsity of Michigan Press.Zu idem a, R. T. (1990a), Inca Civilization in Cuzco. Au stin: University of Texas Press.Zuidema, R. T. (1990b), Dynastic Structures in Andean Culture, pp. 489-505 in (eds.) M.Moseley and Cordy-Collins The NorthernDynasties: Kingship and Statecraft in ChimoWashington D . C : Dum barton Oaks.

  • 7/28/2019 Gose_diarchy, History and Divine Kingship

    33/33

    Copyright of History & Anthropology is the property of Routledge and its content may not be copied or emailed

    to multiple sites or posted to a listserv without the copyright holder's express written permission. However,

    users may print, download, or email articles for individual use.