A Calendrical and Demographic Tomb Text from Northern Peru.

Society for American Archaeology $&DOHQGULFDODQG'HPRJUDSKLF7RPE7H[WIURP1RUWKHUQ3HUX $XWKRU V *DU\8UWRQ 6RXUFH/DWLQ$PHULFDQ$QWLTXLW\9RO1R -XQ SS 3XEOLVKHGE\Society for American Archaeology 6WDEOH85/http://www.jstor.org/stable/972052 . $FFHVVHG Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=sam. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Society for American Archaeology is collaborating with JSTOR to digitize, preserve and extend access to Latin American Antiquity. http://www.jstor.org A CALENDRICALAND DEMOGRAPHICTOMBTEXT FROMNOt1 HERNPERU Gary Urton Thefocus of this study is on a khipu-a knotted-stringrecordingdevice-from the Chachapoyaregion of the northeasternAndes of Peru. The khipu was one of 32 khipus discovered, along with some 220 mummybundles, in 1996 in a half-dozen chullpas (burial houses) built into a rock-overhangoverlooking a lake, called Laguna de los Condores, near the town of Leymebamba (Departmentof Amazonas). The cultural materialsfound with the mummiesand khipus date from the pre-lnkaic Chachapoya culture (ca. A.D. 800-1450), throughthe Inka occupation of the region and on into the early colonial era. It is argued that the khipus stored with the dead representedtomb texts, which contained informationpertaining to the history of the mummiesand the social groups descendedfrom them. One of the khipu samples (UR6) is interpretedas a combined biennial calendar and census of the tributepayers in Chachapoya territory,around Laguna de los Condores, in late Prehispanic times. It is argued that khipu UR6 was the source of informationfrom which thefirst colonial census in the region was drawn up by the Spanish administrators,in 1535. El eje del presente estudio es un analisis de un khipu-un instrumentodonde se registrainformacioncon hilos anudados-grande y excepcionalmentecomplejoprocedentede la regionde Chachapoyas,en los Andes nororientalesdel Peru.El khipuen cuestion fue uno de un grupo de 32 descubiertosen 1996, conjuntamentecon 220fardosfunerarios, en seis chullpas (edificios mortuorios) de piedra y morteroconstruidas en la repisa rocosa de un acantilado que mira a la Laguna de los Condores (departamentode San MartEn),cerca del pueblo de Leymebamba(provinciade Chachapoyas,departmentode Amazonas).Los materiales encontrados con las momiasy khipus, entre los que se encuentran ceramica, textiles, matesp irograbadosy cuentas europeas, datan de la cultura chachapoya (c. 800-1450 d.C.), la ocupacion incaica de la region (1450-1532 d.C.) y la tempranaera colonial. Se argumentaque de modo muyparecido a las inscripcionesde las tumbasde mayasy egipcios, los khipusguardadoscon los difuntos representaban"textosfunerarios";en cuantotales, dichosdocumentospodrfanhabercontenidoimportanteinformacionacerca de temas como la historia genealogica de las momiasy los grupos sociales (esto es, los ayllus) que descendfande ellas, ademas de otros tipos de informacion(calendarica y demografica,por ejemplo) que podrfan haber tenido un intere'sduraderopara el grupo en cuestio'n.En conformidadcon esta hipotesis, se argumentaque uno de los khipus(UR6) de Lagunade los Condoresfue construidocomo un calendario bianual y un censo conjuntode los tributariosde la parte sur-centraldel territoriochachapoya, alrededordel lago, en la e'pocaprehispanicatardfa.Asimismose argumentaque el khipu UR6no solo es lafuente con la cual los funcionarios espanoles realizaronel primercenso colonial de la regionen 1535, sino que adema's,los restos delfuncionario local que segu'nlos documentoshistoricos dio la informacioncensal a los espanoles-un hombrellamado Francisco Guama'n-podrfa, en realidad,estar entre los fardos funerarios recuperadosen las chullpas de Lagunade los Condores. The khipus the knotted-stringdevices used for record-keepingin the Inkaempire have longrepresenteda majorconundrumforthose who would attemptto interpretInkaculturehistory. One of the centralproblemsconfrontingus in the -studyof theseancientrecordsis thefactthatwe have so far been unable to decipher anything but the numericalvalues encoded on many (but not all) of thesedevices.However,evenin thecase of thosekhipus containingnumerical(quantitative)data,we are generallyunable to state convincinglywhat object is being specified,or counted,by any given number recordedon a particularkhipu. That is, while we may be able to say, for instance,thata certainpendantstringon a given khipucontainsthe numerical value253, in mostcases we cannotanswerthe question: "253 (of) what?" There are two notable exceptions to the statements made above. The first concerns calendrical counts;for instance,if an arrangementof knotson a Gary Urton * Departmentof Sociology and Anthropology,Colgate University, 13 Oak Drive, Hamilton,New York 13346 LatinAmericanAntiquity, 12(2), 2001, pp. 127-147 CopyrightC)2001 by the Society for AmericanArchaeology 127 128 LATINAMERICAN ANTIQUITY khiputotals365, we cansuggestthateachunitin such a recordreferredto one "day"(ornight)in an annual calendarcount. The second areafor which we can make strong argumentsfor the meaningof counts recordedon the khipusis in termsof the accounting unitsused in the decimal-basedadministrativesystem of the Inkaempire.In this case, recordedkhipu values of lOs, lOOs,and l,OOOs,for instance,may havehad significancein relationto the recordingof informationon tributarygroupingsin the empire. Laterin this article,we will examineone extraordinary khiputhat was recoveredfrom cliff tombs in the northeastern Andes of Peru,in which, it will be argued,both of these types of valuesalendrical andadministrative areencoded. The initialfocus of this study,however,will not beon theinterpretation of numericalvaluesrecorded in the khipus,but ratheron another,perhapsrelated conundrumfacing students of the knotted-string records.This is the problemof whatwe areto make ofthe fact thatthe vast majorityof survivingkhipus havebeenrecoveredfromgravesitesortombs.VVhat couldhave been the significanceof a culturalpracticein which a record-keepingdevice, presumably fullof information,was depositedin a graveor burialchamberwith a dead body? We may be sure, I wouldsay,thatthekhipuwouldhavebeen of importanceto theindividualwithwhomit was interredduoringhis/herlifetime;perhapsthe personin question hadbeen a khipukamayuq("knotmaker/keeper"), oneof the staterecord-keepers, andtheplacementof thekhipuwithinthe tomb was meantto providethe deadwith a mementorepresentinga life's work.But what, if any,significancemighttheinformationin that khipu have had for the living afFerit was placed in thetomb?Was the significanceof the information recorded on such a khipunegated,or in some other waymadeirrelevant,by virtueof its being placedin atomb,awayfromthe living?As we will see below withregardto the intermentof mummifiedremains ofancestorsin theAndes,theseobjects themummy bundles becameobjectsof frequentvisitationsand -worship by the living. In such circumstances,could thedisposalof these khiputexts with the dead have signaled anevengreaterreligiousorcosmologicalsignificancefor such records, rendering them, for instance, asAndeanequivalentsof thevarietyof mortuary texts found in otherancientsocieties, such as thehieroglyphictexts commonly inscribedwithin Mayan andEgyptiantombs?These are some of the [Vol. 12, No. 2, 2001 questionsthatI will explorein this article. A spectacularexampleof the practiceof placing recordedtexts with the deadis the recentdiscovery, in thenortheasternAndes of Peru,of a groupof some 220 mummybundlesthatwerefoundin a half-dozen rooms,orchullpas(burialchambers),builtintoa rock overhangin a cliff face high above a lake, called Lagunade los C6ndores("Lakeof the Condors"). The site in questionis locatedalong the northwesternborderof theDepartmentof SanMartin(Province of Huallaga), near the town of Leymebamba (Province of Chachapoyas,Departmentof Amazonas; Figure 1). In prehispanictimes, this region was includedwithinthe territoryof the Chachapoya chiefdom,ormulti-ethnicconfederation.Along with theextraordinarilywell-preservedmummiesfound inthecliff tombsatLagunade los C6ndores,theburial goods included an equally well-preservedcollectionof some thirty-twokhipus. In the courseof this articleI will, first,providean overviewof thecontextin whichthe khipusfromthe siteof Lagunade los C6ndoreswere found;second, Iwill discussthepossiblesignificanceforkhipustudies of the disposal of these recordingdevices with mummiesin open (i.e., accessible)burialchambers; third,I will analyzethe numericalcontentsand the possiblecalendricalandsociopoliticalsignificanceof oneespeciallyinterestingandcomplexkhipusample fromthis collection; and finally, I will suggest an attribution of this particularkhiputo a specific individual,a nativeChachapoyan,namedGuaman.We knowfromvarioushistoricalsources(see below)that Guaman was a headmanof an aylluin thisregion,as wellas a local administratorin service to the Inka overlords in lateprehispanictimes.Inaddition,immediatelyfollowingthe Spanishconquestof theregion, Guaman (who now borethe nameFranciscoPizarro Guaman) becameanincreasinglyimportantfigurein thecolonialadministration of thelocalencomendero, Alonsode Alvarado.The suggestedattribution made hereinof proprietorshipof a khipufrom the site of Laguna de los C6ndoresto the local lord (Francisco Pizarro) Guamanis the firstattemptin the historyof thestudyof the khipusto attacha specific archaeologicalsampleof one of thesedevicesto a particular, named historicalindividual. The Region and Culture of Society: Chachapoyals Chachapoyas is todaythe nameof a provincein the A CALENDRICAL ANDDEMOGRAPHIC TOMBTEXTFROMNORTHERN PERU Urton] l l Ut T l u 129 , l AMAZONAS s - ' s-z*W i: w0: CAJAMARCA s" tt} ) Celendfn f Cochot * Cajamarca 0 10 20 1 ' o 24 30miles t- 48 km ,>s - , t -_t ,+ I > _ s J LA LIBERTAD g * CitiQS and towns t Chuchupoya culture archoeological sites ---depsrtmental boundaries Figure 1. Map of Chachapoya Territory and the Location of the Site of Laguna de los Condores. Departmentof Amazonas,in northeastern Peru(Figure 1). However,in bothprehispanicandpre-Inkaic times (i.e., from about A.D. 800 to 1470), ''Chachapoya''l referredto a largerchiefdom,orethnic confederacy-like polity whose borders were formedto the northandwest by the MaranonRiver, to the east by the dense montaneforests above the Huallagariverfloodplains,whilethesouthernextent of Chachapoyaterritorystretcheddown the narrow stripof landbetweenthe middleMaranonandHuallaga riversas far south as the town of Pias (ca. 8° south latitude). Chachapoyas is a rugged, rain- drenchedterritorythatwas as difficultfor the Inkas to penetrateand controlas it has been for succeeding states to integrateeffectively into their transportation and communications systems (see KauffmannDoig 1993;Lerche1995, 1996;Muscutt 1998; Savoy 1970).2 As elsewhere throughoutthe Andes, the preInkaicChachapoyapopulacelived in dispersedkinship, land-holdingand ritualgroupings,which are known throughoutmuch of the Andes by the term ayllu ("lineage,family").Inthehistoricaldocuments fromthe northernPeruvianAndes, these groupings LATINAMERICAN ANTIQUITY [Vol. 12, No. 2, 2001 130 arealso sometimesreferredto by the termpachaqa ("one-hundred"),as well asparcialidad("part";see EspinozaSoriano1981;Remy 1992:72-79;andRostworowski de Diez Canseco 1981 :40). The Chachapoyaregionwas broughtunderInkacontrol less than a half-centurybefore the arrivalof the Spaniardsin 1532 (Espinoza Soriano 1967). The demographicmake-upof Chachapoyawas significantlyaffectedby Inkaimperialpolicies of population control. This was specifically the case with regardto the institutionof the mitimae,which was the termused for the Inkapracticeof moving people fromone placeto anotherwithinthe empire,primarilyfor economicandpoliticalmotives.In regard to theChachapoyaarea,theInkasstationedgarrisons mannedby foreigners forexample,byWankapeoples from the southcentral highlands of Peru andthey (Schjellerup1997:69) withintheterritory, populaChachapoya shippedportionsof the native the tionto some 18differentlocationsaround empire, includinga largecontingentthatwas sent to live in the Inka capital city, Cusco (Schjellerup 1997: 66-69). In additionto the institutionof the mitimae, the Inkas appear to have introduced into the Chachapoyaregionthe full complementof imperial accoutrements,includingthe worshipof the sun,the decimal administrationof the population,and the recordingdevice,thekhipu(Espinoza knotted-string Soriano1967:233-239). ThefirstSpanishentryintoChachapoyaterritory occurredunderAlonso de Alvarado,in 1535. The Chachapoyansquickly allied themselves with the SpaniardsagainsttheirformerInkaoverlords.The institutionof theencomienda(aroyalgrantof patronage over a particulargroup of nativepeoples) was introducedat the beginningof the colonialeraas the principal institution for the administration and exploitation of native Andean peoples. In 1538, Alvaradoreceivedthe encomiendaof the threeparcialidadesof Cochabamba, Leymebamba, andChilchos;thesewerethreeethnicgroups,eachcomposed of multiplepachaqas, or ayllus, locatedin the middle and upper portions of the Utcubamba River drainage,as well as in the cloud forestbetweenthe upperUtcubambaand the Huallagarivers.The latterregion,which includesLagunade los Condores, was the home territoryof the nine ayllus thatmade up the Chilchosethnicgroup. Numerousvisitas(fact-findingvisits by colonial administrators) were made to the area of Cochabamba,Leymebamba,andChilchosfromthe 1540sthroughthelate 1570s.By the 1570s,theSpanhadundertakenthe reorish colonialadministration ganizationof the populationof the region, moving scatteredayllu groupings into central towns, thWe called reducciones (Lerche 1995:67; Schjellerup 1997:83-84). Iwhevisitas and other administrative processesundertakenat this time resultedin a veritableexplosionof written(Spanish)documentation pertainingto censuses, land-boundarydisputes,as well as accountsof the organizationof this region undertheInkas.As elsewherethroughouttheAndes, the Spanish administratorsin Chachapoya drew heavilyon Inkaadministrativerecordscontainedin the khipus, as interpretedfor them by the native record-keepers,the khipukamayuqs.3 In general terms, khipusf which some 600 samples,mostlyfromcoastalPeru,survivetodayarecomposedof a main,rope-likecord(the"primary cord")to which are attacheda variablenumberof spunandplied strings,called "pendantstrings"(for anexcellentoverviewof the structuresof khipus,see Conklin2001). Mostpendantstringshaveknotstied into them in an arrangementthatin manycases can be shown to have recordednumericalvalues in the decimal place system of notationused by the Inka (and presumablyby the Chachapoya)for adminisKhipusweresaidby theSpantrativerecord-keeping. ish chroniclersto have been used to record both informationas well as units of numerical/statistical information recordedin some mannerthatwe do notas yet entirelyunderstand(seeAscherandAscher 1997; Quilter and Urton 2001; and Urton 1998, 2002) that were consulted in reciting histories, genealogies, and otherforms of narrativeaccounts. It appearsto havebeencommonpracticein Inkaand earlycolonial times to place khipusamongthe burial remains of the groups and/or of the khipukamayuqsto which they pertained. The Chullpas of Laguna de los Condores In 1996,a groupof haciendaworkmen,cuttingtrees in the heavily forested area aroundLagunade los Condores, spotted a painting on a cliff face high above the lake. Makingtheir way to the site, they found a half-dozenstone and mortarconstructions, commonlyreferredto as chullpas(burialchambers), builtinto the overhang(Figure2). Eachof the chullpas containednumerousmummybundles,totaling some 220 in all, as well as a wealthof gravegoods, Urton] PERU TOMBTEXTFROMNORTHERN AND DEMOGRAPHIC A CALENDRICAL 131 Figure 2. The Chullpasat Laguna de los Condores. including pottery,textiles, pyro-engravedgourds, wooden sculptures,and a collection of some thirtytwo khipus.4Althoughthe men looted the site for a time whichincludedtheirhackingopennumerous mummybundleswiththeirmachetes thanksto the quick action of PeterLerche, a local historianand guide,mostof thematerialfromthetombswas eventuallyrecovered(see Lerche1999).Thematerialwas then placed in the care of the Peruvianarchaeologist, Dr. Sonia Guillen, and her colleague,Adriana von Hagen. All materials mummies and grave goods were soon movedinto a housein the nearby town of Leymebamba(Guillen 1999). The house was quickly transformedinto a researchfacility, see ("CenteroftheAncestor"; namedCe7atroMallwui vonHagenandGuillen1998;andWilford,NewYork Times,Dec. 16, 1997:F3). Due to thefactthatthemummybundlesandother gravegoods originallyfoundat Lagunade los Condores were partiallyplunderedand badly disturbed at the time of theirdiscovery,we do not have good informationon the precise relationshipsbetween mummiesand khipus.The loss of this information meansthatwe haveprobablylost foreverthe opportunityto do a close analysisof the social organization of the disposal of the khipus that is, their distributionto specific individuals and groups of mummy bundles.Nonetheless, as one of only two archaeologicaldiscoveriesof khipusfor which we have(relatively)goodprovenience,its virtuallycomplete aboriginalmaterialculturalcontext,as well as ethnohistoricaldocumentsfrom the region written up by Spanishadministratorsfrom khipureadings thesiteof Laguna providedby nativerecord-keepers, de los Condoresrepresentsone of the most important archaeologicaldiscoveries for advancingthe studyof thekhipussincethetimeof theSpanishconquest. ThechullpasatLagunade los Condoreswereutilized for disposalof the mummifiedremainsof the dead duringthe severalcenturies-longperiodleading up to the Inka conquestof the region, through period (ca. A.D. 145s1532) the Chachapoya-Inka andon into the earlycolonial era.As for the prehispanicmaterials,therichiconographyadorningespecially the textiles and pyro-engravedgourdsfound 132 LATINAMERICAN ANTIQUITY atthesite suggestconnectionsto culturesof thenorth Peruvianhighlandsand coast beginningas early as theEarlyIntermediateperiod(2000 B.C.-A.D. 700; von Hagen 2000). The continueduse of the site for disposalof the deadduringthe earlycolonialperiod is confirTned by thepresenceamongtheburialgoods of such items as a small wooden Latin cross and Spanishglazedpottery,as well as a varietyof European glass tradebeads. Concerningthe tradebeads, duringthe summer of 1999, severaldifferenttypes of Europeantrade beadswerefoundinsideone of themummybundles, which had been openedfor conservationpurposes. The beads were suspendedon a spun cotton string thatwas attachedto a braidedcord;also suspended from the cord were several other items, including seven small, worked shell ornaments,two metal shawl pins (tupus),and an unidentifiedseed pod. The glass beads includedsphericalred-white-bluegreen beads, as well as tubularbeads of the types known as Nueva Cadiz Plain and Nueva Cadiz Twisted.The lattertwo types of beads were introduced into Peruduringthe earliestyears following the Spaxiishconquest(if not, like the germs of the conquerors,even somewhatearlier)until the early seventeenthcentury(Smithand Good 1982:1>11, 31-33). Thistimeperiodaccordswell withtherange of radiocarbon dates from several khipusfrom Lagunade los Condores.6 Inprehispanictimes,thelivingdescendantsof the mummiesdisposedof in the chullpasat Lagunade los Condoreswould probablyhave includedmembers of the local Chilchos ethnic group,as well as membersof otherneighboringethnic groups.Following the entryof the Inkasinto the region,the site may also have been used for the disposal of the remainsof mitimae,people who were moved into the region from elsewherearoundthe Inka empire (Schjellerup1997:69-70), as well as Inka bureaucratsfromthe capitalcity of Cusco. The local Chilchospopulationprobablylivedin a settlementacross the lake, at a site called Llaqtacocha.The mitimae dndInkabureaucrats mayhavelivedatthe samesite, or perhapsin one of the nearbyInkaadministrative centers,suchas Leimebamba(Leymebamba)and/or Cochabamba(Schjellerup1997:7S73). Oneof themainproblemsthatconfrontsus in our studiesof thekhipusfoundin thechullpasof Laguna de los Condoresis explainingwhy the inhabitantsof theregionwouldhavekepttheseknotted-stringdoc- [Vol. 12, No. 2, 2001 umentsin the hard-to-reachtombs of the(ir)dead. Indeed,it is for most of us, at least an arduous walkof an hourortwo fromLlaqtacochaaroundthe lake to the foot of the cliff in which the tombs are located.Onceatthefoot of thecliff, one is facedwith a difficult climb up the side of the cliff and then acrossa narrowtrailto reachthe tombsthemselves. The difficultiesof this excursionnotwithstanding,it is nonethelessclear from the large numberof dead buriedin these tombs,as well as fromthe recordof burialgoods that includes both Precolumbianand colonialartifacts,thattheinhabitantsof Llaqtacocha andperhapsotherneighboringsettlementsdidin fact make this journey innumerabletimes over several centuries.Whatcan we hypothesizeaboutthe possible significanceand uses of the khipusfor these peoplewho weredepositingandvisitingtheirancestorsin the cliff face aboveLagunade los Condores? Khipus and the Dead Fromthenumerouspublishedreportsof previousdiscoveries of khipus,it appearsthatthe final phasein the life history of the majorityof these recording devices was burial, or interment with the dead. Reportsof this form of disposalarefound scattered throughoutthe archaeologicaland ethnohistorical literaturesin the Andes. Mackey (1970) in particular has described several such cases from coastal Peru.In addition,the catalogsin the variousmuseums where we today encounterlargecollections of khipus such as the Museum fur Volkerkunde,in Berlin(300 samples),andtheAmericanMuseumof NaturalHistory,in New York (100 samples) are replete with references to khipus recovered from grave sites along the Peruviancoast. The problem thatis raisedby this practice-particularlyin those cases (like Lagunade los Condores)in which the remainsof the dead remainaccessible, ratherthan being buriedin the ground is that of understanding and explainingthe curious,indeedcounterintuitive, habitof linking an active signifying object (a khipu)with a (passive)deadbody. Thereis always some element of mysteryassociated with any object interredwith a dead body. Tomb goods have significancein relationto their attachmentto the life and deeds of the deceased,as well as to the interests,sentiments,andneeds of the living (Rowe 1995:31-32;for anexcellentoverview of burialbeliefs andpracticesin theprehispanicand colonialAndes, see Dillehay 1995). The practiceof Urton] PERU TOMBTEXTFROMNORTHERN ANDDEMOGRAPHIC ACALENDRICAL placingitems of value with the dead may be relativelyunproblematicfor utilitarianobjects, such as food vessels, huntingand artisanalequipment,etc. As with all items of materialcultureused daily by the membersof a given society, such grave goods wouldrepresentthe objects to which sentimental values were attached that were used by the deceasedwhen living and that would continue to representimportantobjectsof thememoriesanddiscourseof the living membersof thatsociety.These powerfulin theircapacobjectswouldbe particularly ity to evoke memoriesof the living on those occasionswhendescendantsof themummiesvisitedtheir ancestors,which was common practicein prehispanic and early colonial Peru (Doyle 1988; Isbell 1997;Rowe 1995;Salomon1995).Butwhat,in these terms,can we concludeaboutthe significanceof a device like a khipu,which not only had meaningas an object in the local system of remembrancesand discourse,but which also had potentialuse as an active "signifier,"a recordthat was susceptibleof being read,interpreted,andeven potentiallyof provokingactionon the partof the living?Whatwould havebeenthe statusof sucha device in the local system of values and meanings,as well as in the systems of controland authorityamongthe living? To state in a direct way the position that I will adopthereinon the questionof the disposalof khipus in open or accessible tombs, I thinkthatin the prehispanicand early colonialAndes, the mummified remains of the dead were given custody of recordswhich still hadthe capacityto communicate withthe living becausethese were notjust anydead bodies;rather,they were ancestors.In Andeanideology and cosmology, ancestors whose mummified remains were referred to by the term mallki were objects of greatvenerationand worship(see Doyle 1988;Salomon1995).Mallkiswere often kept in caves or in otherbuilt structuresthat affordedaccess to the ancestralmummiesby theliving (Bonnier1993;Isbell 1997).As Doyle has noted in her excellent studyof burialritualand the ancescentralcult in seventeenthand eighteenth-century tralPeru, From the documentarytestimonies it is apparent that cave machays [burial chambers] were almost always sealed with irregularly shaped, uncut stones, while the doorways of the other types of machays [e.g., subterraneanand aboveground built structures] are rarely described. 133 The doorways were not designed to be permanently sealed because periodic access to the interior was necessary for the performance of ceremonies honoring the dead and the placement of new burials [Doyle 1988:110]. In prehispanicandearlycolonial times, the mallkiof anyparticularsocial group,suchas an ayllu or pachaqa,wasregularlyvisited,givenfood anddrink, redressed,and was askedfor aid and guidance.We know a considerableamountabout such activities, primarilybecausetheyformsome of thecentralpreoccupationsof the "idolatry"campaigns,the investigationsof idolatrous,"pagan"beliefs andpractices thatwere carriedout by Catholicclergyin the Peruvian countrysidethroughoutthe seventeenth and early eighteenth centuries. The objectives of the clergywerenot only to investigate,butalso to stamp out suchpracticesandto destroythe objectsof veneration,such as the ancestralmummies (see Mills 1997).The mallkiwere consideredto be the owners andprovidersof all food (plantandanimal)andthe fields,irrigationcanals, firstownersof all agricultural terraces,etc.; as such, ritualsand ceremonieswere held to honorandworshipthe ancestorsin the caves (mach'ay) where they were interred (Doyle 1988:68). In one very interestingdocument, it is reportedthatchicha(cornbeer)was producedforritual consumptionat a mach'ayby the membersof a group of ayllus that were consideredto have been descended from a particularmallki. It is said that everyhouseholdwithineach of the ayllus gave one ear of corn for makingthe chicha;a khipuaccount was createdto see to it that everyone contributed (Doyle 1988:151). Each mallki communicatedwith his/her living descendants through an intermediary,usually a memberof thedeceased'sayllu who servedthemallki as its oracle (Doyle 1988:61, 117, 135-137). Mallkis were the most powerfulobjects validating theexistence,history,socialidentity,as well as overseeing the well-being of the group of people descended from that ancestor.Whatevera mallki "said" throughits diviner/oracle was considered to be sacredand true. By the same token, I would arguethatwhateverthegravegoods (suchas a khipu) thatwere buriedwith a mallkiindicatedwould similarlyhavebeen takenby the living as the true,final, or at least the original/ancestralword on a particular matter. Given the powerful position occupied by the 134 LATINAMERICAN ANTIQUITY mallkis in the life and thoughtof prehispanicand early colonialAndeanpeoples, how betterthan by worshipingthe mummifiedremainsof one's ancestors could one realize and validatethe connection betweenan individualor a groupandits past?I suggest thatat the site of Lagunade los C6ndores(and no doubtat otherchullpasites as well) the interaction between past and present, realized in the encounterbetweena personandhis/hermummified ancestors,was mediatedin some cases by, among otherthings,the khipus.If this was the case, then I thinkthis gives us groundsto arguefor some rather specifictraitsthatmay have characterizedthe kinds of informationencodedin these khipu"tombtexts" thatwere depositedwith the dead. In particular,I hypothesizethat the information inscribedin such texts would probablyhave concerned,or been relevantfor, both the living and the dead thatis, for the pastandpresent.Suchrecords probablywould not have incorporatedinformation thatwould lose its relevanceover a single lifetime. Rather,theinformationwouldprobablyhavebeenof a more enduringnature;for instance,it might have concernedthegenealogicalhistoryof thegroup(e.g., an ayllu,orpachaqa)andits relationswithits neighbors,as well as withsupernaturals. Tothedegreethat chronological,demographic,andothertypesof measurementsmay have been recordedon such khipus, theywouldprobablyhaveregisteredaboriginal,standardized,or ideal values or units (e.g., whole numbers, full decimalvalues, or powers of ten, etc; see Urton1997).In these ways, khipurecords,or whatI amreferringto hereas tombtexts,wouldhaveserved thelivingas standardsagainstwhichto evaluatepresent conditions and from which to measure the changes that had takenplace in the population,its organization, andperhapsits understanding of its own historyandidentityovertime. In regardto the latterpoint, I thinkwe are warrantedin makinga comparisonbetweenthe khipus placedin tombslike thoseatLagunade los Condores withtombtextsfoundin otherancientsocieties,most notablyamong the ancientMayansand Egyptians. Intheformercase,forinstance,we havesuchremarkable examplesas thatof the Templeof the Inscriptions at Palenquein which the corpse in this case, LordPacal was placed in a tomb on whose walls were inscribedhis dates of ascensionand death,as well as otherimportantdates of his lineage history (Lounsbury1974;Robertson1983:55ff.).As for the [Vol. 12, No. 2, 2001 Egyptiantombtexts, we know thata whole class of such documents the so-called PyramidTextswereinscribedon the walls of pyramidtombsfor the purposeof helping the deceased royaltyascend to heaven and continue to rule there in the afterlife (David 2000:18-19; Hare 1999:65). The question raisedby these comparativeexamplesof tombtexts elsewherein the ancientworldis whetheror not the disposalof khipuswith the dead in the prehispanic Andes shouldbe regardedas an indicationthatthese documentshadsimilarprofoundhistorical,religious, andperhapsotherformsof significancefor the people who placedthemin the tombs.My hypothesisis that,in the case we areexamining,the khipusdid, in fact,havesuchsignificancefortheChilchosandother peopleswho lived aroundLagunade los Condores. While my studyof the collection of khipusfrom Lagunade los Condoresis stillin a preliminarystage, I will examinebelowtheinformationrecordedonjust one khipufromthiscollectionthat,I believe,contains in its structure,organization,and apparentcontents thekindof ancestral,regulatorytextdiscussedabove. A Possible Biennial Tributary Khipu from Laguna de los Condores ThekhipusfromLagunade los Condores,whichare overwhelminglymadeof cotton,rangein size from sampleshavingonly a few pendantstrings(one set of which was found inside the mummybundleof a young woman) to several samples containing 60>800 pendantstrings.In termsof its size, state of preservation,andthe organizationof information thatwas spun,strung,dyed, and knottedinto it, the khipualludedto at the end of the previoussectionis one of the most spectacularsamplesrecoveredfrom the chullpasat Lagunade los Condores(see Figure 3). I will referto this khipuin the following discussion as khipuUR6.7 The Organizationof CalendricalInformationin khipu UR6 KhipuUR6 containsa total of 762 pendantstrings made of Z-spun, S-ply cotton fibers. Comparedto otherkhipusin thiscollection,the pendantstringsof khipuUR6 areattachedto theprimarycordandorganizedin an unusual,highly systematicmanner.That is, with only a few exceptions,the basicunitof organizationof the pendantstringsis a paired set (see Figures4 and 5). One memberof these pairedsets is composedof either20, 21, or 22 pendantstrings; Urton] PERU TOMBTEXTFROMNORTHERN AND DEMOGRAPHIC ACALENDRICAL Figure 3. KhipuUR6(circular thesependantstringsareattacheddirectlyto theprimarycord.The othermemberof each pairedset is composedof 8, 9, or 10 pendantstrings.These latter pendants,however,are not attacheddirectlyto theprimarycord;rather,theyareattachedto a string whose two ends arethemselvesattachedto the pristrings(i.e., as they marycord.These"intermediary" standbetween,andconnect,the pendantstringsand theprimarycord)hangdownbelowtheprimarycord in a slight arc,or loop. I will referto these intermediary stringsand the pendantsattachedto them as "looppendants."Thereare24 groupsof thesepaired sets of pendantstringsand loop pendantson khipu UR6 (see Figures6 and 7). Before proceeding,I shouldnote thatin the calendricalanalysisof khipuUR6 thatI will presentin this section, I will disregardthe 32 pendantstrings on this sample that do not conformto the regular arrangementof paired-that is: [20, 21, 22] + [8, 9, 10] -- sets of pendantstringsas outlinedabove.The excludedpendantstringsare those that are circled in Figure 7. The knot counts on these excluded 135 lay-out; summer, 2000). stringswill come into play later in my analysis. I believe thatthese strings,which do not conformto the patternof paired sets, may have been used, in somemannerthatI do notentirelyunderstandatpresent, to provideflexibility in relationto one of the principaluses of this khipu-that is, as a biennial calendar.8 Therearetwo clearindicationsof the calendrical significanceof the organizationof pendantsstrings on khipuUR6. The firstindicationis the numberof stringsin each of the pairedsets; thatis, if we take the mean value of the numberof pendantstringsin the 24 pairedsets, we arriveat a (theoretical)repetitivecombinationof 21 + 9 (= 30) pendantstrings(see Figure5). This count is, of course,the whole number valueclosest to the numberof days in a synodic lunarmonth (= 29.53 days). If each one of the 24 pairedsets in khipuUR6 containedthe meun number of strings(i.e., 21 + 9 = 30), this would produce a totalof 720 pendantstrings(30 x 24 = 720). Onehalf of this totalwould give us 360 pendantstrings, a numberwhichis close to an annualcalendarcount LATINAMERICAN ANTIQUITY 136 [Vol. 12, No. 2, 2001 Figure4. Detail of Primary Cord and Pendant Strings on Khipu UR6. | Illoop ' 30 strings | Figure 5. The Structure of Pendant and "Loop Pendant" String Pairs on KhipuUR6. Urton] PERU TOMBTEXTFROMNORTHERN AND DEMOGRAPHIC ACALENDRICAL Figure 6. KhipuUR6(linear 137 lay-out; summer, 1999). less thanthecountof the numberof dayscoveredby However,when we count the actual of 365 cWays. aphelion,whichis theperiodof the sun's(apparthe the in strings pendant numbersof pendantandloop movementfromthe Marchequinoxthroughthe ent) numthe at arrive we 24 pairedsets (see Figure7), solstice and back to the Septemberequinox June number a 365, is count ber730; one-halfof thistotal 1980:66).Thus,althoughsuch an interpreta(Aveni is what in days of number that coincides with the resultin a biennialcalendarof disconwould tion number whole the (i.e., year" knownas the "vague quarterly(i.e., four half-year) interlocking tinuous, khipu Thus, days). of daysin a solaryearof 365.242 given the sums of pendant nonetheless, periods, cala represent to UR6 appearson a superficiallook half-years la and 2c, we the in contained strings perilunar synodic endarintegrating24 essentially thatthebiennialcalpossibility the for allow should respective The ods into two (vague) year counts. in thepairings:1a reckoned was UR6 khipu in endar periods half-year pendantstringcountsforthesefour ratherthanof the 366), (= 2d + lb and 364) (= 2c + in summarized are illustrated in Figure 8 and + 2d (=368). 2c and (=362) lb + la groupings: 1, Table and 8 Figure schematicformin Table1. In respectto this with appreciate to point main The loop and pendant we note thatthe 12 pairedsets of to representa appears UR6 khipu that is discussion of count total a pendantstringsin YearOne contain pause for a should We count. calendar biennial count a contain Two 362; the 12 pairedsets forYear circumstance unusual rather the consider to moment am I what of count average of 368 strings.Thus,the callinghereinthe two yearsrecordedin khipuUR6 of findinga khipurecordinga double,ratherthana single,yearcount.It is relevantto notein this regard is 365 (i.e.,362 + 368 = 730 *.2 = 365). the Inkasexhibiteda strongemphasison pairing that we if emerges fact in It is interestingto notewhat in manyaspectsof theirmaterialculture dualism and of sets paired 12 of divideeach ofthe two groupings organization(for generaldiscussociopolitical and what producing pendant strings in half, thereby of Inkaculture,see Duviols 1973; aspect this of sions interpretation should, accordingto the calendrical andvande Guchte1996).For 1997; Urton 1986; Platt (see periods offered here, representfour half-year culture,Cumminshas material of terms in instance, halfthe 1, especiallyTable1). As we note in Table of Inkakeros,the studies his in convincingly shown pendant 179 year labeled Year One (a) contains vesselswere these that cups, drinking ceremonial Inka con(c) Two Year stringsand the half-yearlabeled 1988:124). (Cummins pairs in produced always more day one-half is tains 185 pendants.Now 179 Althoughwe do not have extensivedocumentathanthe countof the numberof days coveredby the foranequalemphasison dualismamongthepretion sun's the of time of perihelion; this is the period Chachapoya,there are certaindata that do Inkaic equinox September (apparent)movementfrom the the importanceof this organizationalprinsupport March the to back throughtheDecembersolsticeand regionandin its materialcultureas well. this in ciple days one-half and equinox.The number185 is one LATINAMERICAN ANTIQUITY [Vol. 12, No. 2, 2001 138 (top string) 20 9 (t) 21 9 21 9 21 9 21 9 21 9 21 9 21 10 (top strings) (top strings) 22 9 22 8 21 9 22 9 21 9 22 9 22 9 22 9 22 9 22 9 22 9 22 9 22 8 22 9 22 8 21 9 2 5 2 W Figure 7. The Organization of String Groupings on Khipu UR6 (top of this figure is at right-hand side of Figure 6). For instance,we find a notableemphasison paired artifactsin the materialrecoveredfromthe chullpas at Lagunade los Condores.Thereare severalpairs of pyro-engraved drinkinggourds,eachpairof which isrdecoratedwith virtuallyidenticaldesigns, found amongthe gravegoods. I wouldnote the very interesting ethnographicdatumreportedby Schjellerup in which she reportsthat,priorto 1960, villages in the region were paired for the purposes of maintainingthe roadsystem,the bridges,and the canals (1997:46). This informationis particularlyinteresting in relationto the interpretationI will develop belowregardingthetwo-yearcalendarof khipuUR6 as a recordof laborservice in the region in the late prehispanicperiod.Finally,Lerchedetectsnot only a strong dualistic organizational principle in Chachapoyasociopoliticalorganization,buthe also arguesthatwhenpolitiesin thepastweredividedinto dualgroupings-one exampleof which is the dualism "Chilcho/Llaja" these entitieswere each further subdivided into nine subdivisions (Lerche 1995:5841). Thus, whetherthe calendricalkhipu underconsideration(i.e., UR6) was manufactured andused by the pre-InkaicChachapoyansor by the Inkaadministrators aftertheirentryintoChachapoya territory,we ought not, in fact, be surprisedto find thatthecalendarspecialist(s)constructeda two-year calendarcount. Therefore,at this first level of analysis, we can concludethat one of the messages the people from Laqtacocha (or elsewhere) would have been remindedof when they visited the tombs at Laguna de los Condoresandtook up khipuUR6 was a powerfulrepresentationof what I would termthe ideal, or proper,structureand organizationof time i.e., as a unit composed of two complementary,probably interconnectedvague year counts. As we will see below,theaboveinterpretation has only scratchedthe surfaceof the calendricalinformationencodedin khipuUR6. However,in orderto understand the additionaltemporalinformationthat wasencodedintothe stringsof khipuUR6, we must shift our attentionto the knot counts that are containedon thependantandloop pendantstringsof this khipu.In the courseof this analysis,we will identify anotherpossible use of this knotted-stringrecordthatis, as an accountof the organizationof tributepayersin the region of Lagunade los Condoresin lateprehispanictimes. \> Urton] A CALENDRICAL AND DEMOGRAPHIC TOMB TEXT FROM NORTHERN PERU Pandan1/Loop 139 Pondan1 Paln 20 9 (top) 6) 21 9 21 (a) > 21 9 21 9 21 9 =179 \ / \ Y*ar Ono =362 21 9 21 10 (b) (top) v_ _ .. . = 183 > 22 9 22 9 (c) Thar IWO 22 9 22 8 21 9 22 9 (top) = 185 22 9 22 9 22 9 _ 22 9 22 9 22 22 9 22 8 21 9 =368 \ > = 183 / / 2 5 2 Figure 8. The Calendrical Organization of Strings on Khipu UR6. / = 730 . 2 = 365 140 LATINAMERICAN ANTIQUITY [Vol. 12, No. 2, 2001 Table 1. The Schematic Organizationof the Biennial CalendarCount in Khipu UR6. (a) (b) Totals: Year 1 29 30 30 30 30 30 30 31 31 30 30 31 179 183 362 The Significance of Knot Values in kkipu UR6 As is well-known,many(butby no meansall)9khipus incorporated a hierarchicaldecimalorganization in the arrangement of knotstied intopendantstrings such thattherewere differenttypes of knots tied in clustersondifferentlevelsalongthelengthof thependantstringsasplace-valueindicatorsof ls, lOs,lOOs, l,OOOs,and lO,OOOs (see Ascher and Ascher 1997 [1981]; Garcilaso de la Vega 1966:330-331 [1609-17]; Locke 1923;Urton1994). In Figure9, I providea readingof thenumericalvaluesof theknots tiedintothependantandlooppendantstringsof khipu UR6.Intermsof theconstructionof knotsinthissample, I would note thatthe majorityof knotsin khipu UR6 are tied as, what I have termedin an earlier study(see Urton1994),Z-knots.However,thereare also several scatteredexamples of S-knotson this sample(see below).l° In the interpretationof knot-countvalues pre*sentedin Figure9, I have dividedthe khipuat the following places: a) the non-pairedgroupings of stringsat thebottomof the khipuareset off fromthe pendantsand loop pendantscomposing what was interpretedabove as the biennialcalendarcount;b) anotherdivision is made between the two annual periodsof the biennialcalendarcount;and c) sub.divisionsof the two yearly counts are made (as in Figure8 andTable1) to producefourhalf-yearperiods. I will firstconsiderthe calendricalinformation encodedin the knot countsillustratedin Figure9. The CalendricalInformationin the KnotValuesof KhipuUR6 When we study carefullythe knot values tied into 185 183 Year 2 30 31 31 31 31 31 31 31 30 31 30 30 (c) (=364) (=366) (d) 368 the pendantandloop pendantstringsof khipuUR6, we encountera ratherstunningamountof calendrical information(Table2). We areawarefromearlier studies of Inka calendrics(see especially Zuidema 1977 and 1989) thatInkacalendarspecialistshad a particularinterestin observingboththesidereallunar cycle as well as the synodiclunarcycle, andin correlatingthese lunarperiodicitieswith the solaryear of 365.242 days.Thesecalendricalperiodicities,and the apparentinterestin correlatingthem, are also evident in the Chachapoya-Inkacalendarin khipu UR6. For example, we find in the knot values of khipu UR6 that the calendar specialists in Chachapoyashad recordednumbersthatwere very close numericalequivalentsof sets of sidereallunar cycles (= Ll in Table2) composed of groupingsof W 110 (-0.4 day), 112 (-1.1 days), and 113 (-2.3 days) sidereallunarcycles; these values are encoded in khipu UR6 by the values (respectively)of 3,005, 3,059, and 3,085. Thiskhipualsodisplaysremarkably close approximationsof multiplesolar-yearcounts(= S in Table 2). These areseen in the periodsof: a) 1,826, which is two-tenthsof a day less thanfive solaryears, and b) 730, whichis one-halfof a day less thantwo solar years.What is perhapsmost strikingaboutthe calendricalperiodicitiesencoded into khipuUR6 are those cases in which therearecorrelationsof different lunar cycles with the solar year count(s). For example, the total knot count on the pendantsand loop pendantsof YearTwo (= 977) is two and onehalf daysmorethan33 synodiclunarmonthsand/or six-tenthsof a day more than two solar years plus nine sidereallunarcycles. One sees a similarcorrelation of a lunar(sidereal)periodicityand the solar A CALENDRICAL AND DEMOGRAPHIC TOMBTEXTFROMNORTHERN PERU Urton] Pondant Per.Onts Sts Loop Pandants (non-palrod Per.dun1s) (O) 75 20 9 (a) Yeor 135 (31) (]) (top) 21 9 21 9 21 9 21 9 21 9 58 19 60 12 52 595 58 150 44 10 1 21 9 21 45 ll 49 10 103 22 (b) 83 9 22 8 21 9 22 9 (top) 148 67 13 71 12 71 Total 21 (c) 121 52 12 55 86 30 76 26 8 28 22 9 22 9 22 8 22 9 22 8 21 9 33 64 19 22 5 39 66 26 66 Total rotals 13 (23) 57 Year 2 6rand (=2028) 59 9 22 9 22 9 22 9 22 9 22 9 (top) (d) 87 1295 733 446 531 1 179 1826 (=977) (=3005) (80) (=308S) Figure 9. Knot Counts on Khipu UR6. 141 LATINAMERICAN ANTIQUITY 142 [Vol. 12, No. 2, 2001 Table 2. A Summaryof CalendricalValues Encoded in KhipuUR6. Value in Khipu UR6 3,085 3,059 3,005 2,963 1,826 1,179 977 733 730 531 446 185 179 122 53.5 = = = Calendrical Correlation 113 L] (-2.3 days) 112L] (-l.ldays) 110 L] (-0.4 day) = 100 L2 (+9 9 days) = = 5 S (-0.2 day) = 40 L2(-2.2 days) = 33 L2 (+2.5 days) 25 +9L,(+0.6day) = = 2 S (+2.5 days) = 2 S (-0.5 day) = 18 L2(-0.6 day) 1 S + 3 L] (-1.2 days) = aphelion (-1.5 days) = perihelion(+0.5 day) = = 4 L2(+3-9 days) = 2L(-l.ldays) Source of Value in Khipu UR6 total knot count on all pendants total knot count on all pendantsminus the knot count on the bottom five pendant groupings(=26) knot count on all pendants& loop pendantscomposing Yr. 1 & Yr. 2 biennial calendar value of all knots tied as Z-knotsa total knot count on loop pendantsof Yr. 1 & Year2 pairedsets total knot count on pendantsof Yrs. 1 & 2 pairedsets total knot count on Yr. 2 pendantsand loop pendants total knot count on Yr. 1 pendants(of pairedsets) total # of pendants& loop pendantscomposing biennial calendar total knot count on Yr. 2 loop pendants knot count on Yr. 2 pendants total pendantstrings in Yr. 2 (a) total pendantstrings in Yr. 1 (a) value of all knots tied as S-knotsa average# of loop pendantsin l/2-yr. periods Key: S = 365.242 = the numberof days in a solaryear L] = 27.321 = the numberof days in a sidereallunarmonth(i.e., the monthlycycle of the movementof the moon from a fixed point in the sky back to that point) L2= 29.531 = the numberof days in a synodiclunarmonth(i.e., the cycle of the phases of the moon; see Aveni 1980:98-100) a See note #10. regionencodedinto khipuUR6 is an accountingof tribute-payers thatbelongedto theayllusorpachaqas in this region.This informationis includedprimarily in the knot values of khipuUR6. That is, as we see in the knot counts(i.e., the decimalvalues)presentedin Figure9, theknotcountforYearOne totals 2,059 (i.e., 2,028 + 31 = 2,059), while theYearTwo knot-count totals exactly 1,000 (i.e., 977 + 23 = 1,000).These two valuesgive a total(biennial)knot count of 3,059. However,if we eliminatethe three groupsof pendantstringscircledin Figure9 (i.e., 31 + 23 = 54), which are those stringgroupsnot organized accordingto the pairingof pendantand loop pendantstringswithinthe two-yearcounts,we end up with a total,biennialknotcountof 3,005. As we will see below, this knot count is very close to the total numberof tribute-payersthatwere countedin the first census of the population(i.e., the descendantsof themallkisatLagunade los Condores)made Khipuand DocumentaryAccountsof Tributein this regionin earlycolonial times. PayersaroundLagunade los Condores Althoughtherehas not been extensiveethnohisThe othertype of informationthatit appearslikely toricalwork carriedout concerningthe early colothe khipukamayuqsof the southern Chachapoya nial populationin the southernpartof Chachapoya year count in the knot count on pendantstringsin YearTwo(= 446; see Table2 for the calendricalcorrelation). The organizationof knotvaluestied intothe pendantsand loop pendantsof this khipusample provide the information for very complex and sophisticatedcalendricalreckoningandcorrelations. The calendricalinterpretationof khipu UR6 presentedhereshouldbe consideredprovisional.As furtherstudiesof additionalkhipusamplesfromLaguna de los Condoresarecompleted,andas these studies arebroughtinto relationshipwith studiesof the historicaldocumentsfrom the region, it is hoped that we will arriveatfullerandmorecontextualizedviews on how and why certainastronomicalobservations were made, and calendricalcalculationsand correlations were performed,by the Chachapoya-Inka khipukamayuqs. Urton] PERU TOMBTEXTFROMNORTHERN AND DEMOGRAPHIC ACALENDRICAL aroundLagunade los Condores,important territory colonialcensusdatafromtheregionaregivenin studiesby EspinozaSoriano(1967), Lerche(1995), and Schjellerup(1997). We know that,duringthe brief period of Inka domination of the region, the Chachapoyapopulationwas organizedinto decimal unitsof tributepayers.Administrativeoversightof thesouthernChachapoyaregionwas conductedfrom theInka site of Cochabamba,located to the southwest of Lagunade los Condores(see Figure 1; see Schjellerup1997:64). In the Inkadecimaladministrativeorganization, the principal accounting units among the Chachapoya(as elsewherethroughoutthe empire; see Julien1988 andMurra1982) were groupingsof chunka("10"),pachaqa("100"),waranqa("1000"), andhunu("10,000")tribute-payers.Subjectsof the empirewererequiredto paytributewithinthesedecimal units of accounting.As is well-known,in the Inka empire, the mannerof "payingtribute"was throughthe performanceof labor service for the state.This was undertakenin mit'a, "turns"of labor groupsworkserviceperformedby differenttributary we learn Now, 1982). Murra (see ing in succession thatdur1572, to dates which document, from one polittop (the principal cacique the times, ing Inka large the of official) administrative ical and of west and south the to just region administrative Guaman. named man a was Condores los de Laguna Guamanis describedin this documentas the "lord" (senor) of the threewarangas(= 3000 tributepayers) of Cajamarquilla,Condormarca,and Bambamarca(Schjellerup1997:315-316). Soon afterthe Spanishenteredthe Chachapoya region, in 1535, Alonso de Alvaradowas awarded theencomiendaof Cochabamba,Leymebamba,and Chilchos (Espinoza Soriano 1967:299); this encomiendagrantincludedthe ayllus of Chilchos Indianswho were removedto colonial reducciones ("towns")fromthe areaaroundLagunade los Condores.Whatis criticalto note is that,at the time of theestablishmentof thisencomiendagrant,Alvarado is saidto havetakencensusinformationfroma local lordnamedFranciscoPizarroGuaman.As we learn fromreadingthecolonialdocumentsfromthisregion carefully,FranciscoPizarroGuamanwas the same man identifiedearlier (i.e., in the documentfrom 1572; see above) as "Guaman,"the Chachapoyain Inkacaciqueprincipalof the3,000 tribute-payers prohave to said is Guaman Francisco region. this 143 withcensusdatafromkhipuaccounts videdAlvarado inhis (Guaman's)possession (BibliotecaNacional Lima,A585 f93r; published in Espinoza Soriano 1967:299).Inthatcensusaccount,madein 1535,the totalnumberof tribute-payers,or mit'a laborers, countedwas given as threewarangas i.e., 3,000 (see Schjellerup1997:40,318). may have The number3,000, or threewarangas, been a rounded-off,or "idealized,"number,as we often find in colonial documentationpertainingto censusesandtributerecords(see Remy 1992:72-79; andUrton 1997). Nonetheless,I find it remarkable andquite suggestivehow closely the initial census count of 3,000 mit'a laborers approximates the numericalvaluesencodedintothebiennialcalendar countin khipuUR6 from Lagunade los Condores. To recapitulate,when we subtractthe knot values tied into the circledpendantgroupswithineach of the two "yearcounts"thatdo not conformto the calendricalorganizationin khipuUR6, we arriveat the totalof 3,005 (= 2,028 + 977). I hypothesizethat this value, and thereforekhipuUR6 itself, referred in the regionof to the totalnumberof tribute-payers Laguna de los Condores, Leymebamba, and Cochabambain late prehispanic(i.e., ChachapoyaInka)times. Conclusions and Questions for Future Studies Theimportantpointto stressfromtheanalysesgiven aboveis the coincidencebetweenwhatI have interpretedas a) calendricalvalues,andb) censusfigures (for tribute-payers)on khipuUR6. The crux of my argumentis that this was, in fact, a coincidenceof numericalvalue thatis, theywerecomplementary, WhatI meanby thisis thatthecalnotcontradictory. endricalorganizationof (FranciscoPizarro)Guaman's khipuwould have provided the temporal patternfortheorganizationof mit'alaborserviceprovidedto theInkastatein theregionpriorto the Spanish conquest. The knot count pertaining to each pairedset of pendantandloop pendantstringswould have indicated the number of tributepayers that would have been responsiblefor performingstate laborservicein the regionduringone synodiclunar monthoverthe two-yearaccountingperiodencoded in the stringsandknots of khipuUR6. offered Althoughit is hopedthattheinterpretation above in the form of a hypothesishas at least gone some distancetowardexplaininghow andwhy khipu UR6 was constructed as it is, it is also clearly 144 LATINAMERICAN ANTIQUITY recognizedby the authorthatthereare,in fact,many more new questions raised by this interpretation. Thesequestionsincludethefollowing:If khipuUR6 is, indeed, a biennial tributaryrecord,then where does the firstyearbegin?At the top or at the bottom of the khipu?At whatpointin the annualcycle does the calendarbegin (e.g., at one or the otherof the solstices?At one or the other of the equinoxes?)? Why, if this is a two-yearcalendarof state service thatwas performedby peoplein thisregion,does the recordshow that twice as many people workedin YearOneas inYearTwo?Andfinally,why werethere such greatdifferencesin the numbersof tributaries who workedforthe statefromone monthto thenext? I cannotprovideconvincinganswersto any of these questionsatthepresenttime.However,we areatleast now asking questions of khipu records that have neverbeenaskedbefore,andhopefullysomeof these questionscan helpguideourfutureinvestigationsof the knotted-stringrecordsin productivedirections and1ncreat1veways. Finally,andas a corollaryto the hypothesisarticulatedabove with regardto the integratednatureof the calendrical and demographic information encoded into khipuUR6, I would furtherhypothesize thatthiskhipumayhavebeentheactualaccounting khipu from which Lord (Francisco Pizarro*) Guaman supplied census figures to Alonso de Alvarado,in 1535.Thislatterhypothesisimpliesthat the mummifiedremainsof FranciscoPizarroGuaman may be amongthe 220 or so mummybundles recoveredfromLagunade los Condores.1l When descendantsof the mallki(s)at Lagunade los Condoresvisitedtheancestor-custodian of khipu UR6, they could havetakenup the khipuin orderto remindthemselvesof theproper,or "standard," organization of time and the calendar,as well as the "appropriate," traditionaldivision and organization of laborserviceto the stateperformedin the pastby the populace in the region. During precolumbian times,these"messagesfromthemallkis"wouldhave been of contemporaiyrelevance,as a recordof their standinglabor service obligationto the Inka state. Afterthe Spanishconquest,the accountingof time and sociopoliticalorganizationcontainedin khipu UR6 would have servedthe people of the regionas a historicaldocumentagainstwhich to measureand evaluatecertainchanges that had occurredin their worldsince the beginningof Spanishdominationin the region.As in the readingof all texts, the reading . . [Vol. 12, No. 2, 2001 of the informationencoded into khipuUR6 would have been subjectto the (re-)interpretations of each successivereader,as well as eachgenerationof readers. Each new readingof the tomb text, whetherits substancewas the same as or differentfromthe last reading,wouldhavebeensanctionedby thepresence of the mallki. Acknowledgments.I would like to thankthe following people for commentson earlierversionsof this paper:AnthonyAveni, Sonia Guillen,DarrellGundrum,PeterLerche,JuliaMeyerson, Keith Muscutt,FrankSalomon, Inge Schjellerup,Adrianavon Hagen, and Tom Zuidema.I am especially gratefulto Anthony Aveni for the time he spent talkingto me aboutpossible calendrical correlationsin the string and knot values on the khipu sample discussed herein. I also benefitted from helpful commentaries and critiques provided by five readers for Latin AmericanAntiquity.I alone am responsiblefor any errorsthat remain in the article. I gratefully acknowledge the supportof the following institutionsor organizationsfor supportof my researchon the various topics touched on in this article. For support to carry out research at the Centro Mallqui, in Leymebamba,Peru, I thank the ResearchCouncil of Colgate University (summer, 1998), and the Wenner-GrenFoundation for AnthropologicalResearch (summer, 1999). In particular,I thankSonia Guillen andAdrianavon Hagenfor theirsupportof my researchon the khipushoused in the CentroMallqui. For supportof ethnohistoricalresearchin Seville, Spain, I thankthe AmericanPhilosophicalSociety (summer,2000). For support to analyze and write up the resultsof my research,I gratefully acknowledgethe NationalEndowmentfor the Humanities,for a post-doctoralfellowship (2000), andI gratefullyacknowledge the JohnD. andCatherineT. MacArthurFoundationfor its support of my researchover the period 2001-2005. References Cited Arellano,Carmen 1999 Quipuy tocapu.Sistemasde comunicacionincas.InLos Incas. Artey simbolos,edited by FranklinPease et al., pp. 215-261. Coleccion Arte y Tesoros del Peru. Banco de Creditodel Peru,Lima. Ascher,Marcia,andRobertAscher 1978 Codeof theQuipu:Databook.Unpublishedmanuscript, Archivist,CornellUniversityLibrary,Ithaca,New York. 1997 [1981] Code of the Quipu:A Studyin Media, Mathematics,and Culture.Dover Publications,New York. Aveni,AnthonyF. 1980 Skywatchersof Ancient Mexico. Universityof Texas Press,Austin. Bonnier,Elisabeth 1993 Architectureof the Living andthe Dead in Tantamayo, Peru.Manuscripton file, LongyearMuseumof Anthropology, ColgateUniversity,Hamilton,New York. Brokaw,Galen 1999 TransculturalIntertextualityand Quipu Literacy in Felipe GuamanPoma de Ayala's Nueva Coronicay Buen Govierno.Ph.D.dissertation,IndianaUniversity.University Microfilms,Ann Arbor,Michigan. Conklin,WilliamJ. 1982 The InformationSystem of the MiddleHorizonQui- Urton] PERU TOMBTEXTFROMNORTHERN AND DEMOGRAPHIC ACALENDRICAL in theAmerpus.InEthnoastronomyandArchaeoastronomy ican Tropics,editedby AnthonyF. Aveni and GaryUrton, pp.261-281.Annals oftheNewYorkAcademyof Sciences, Vol. 385, New York. 2001 A khipu Information String Theory. In Narrative Threads:Explorationsof Narrativityin Andean KnottedStringRecords,edited by JeffreyQuilterand Gary Urton. Universityof TexasPress,Austin,in press. Cummins,ThomasB. F. 1988 AbstractiontoNarration:KeroImageryof Peruand the ColonialAlterationof Native Identity.Ph.D. dissertation, Universityof California,Los Angeles. UniversityMicrofilms,Ann Arbor,Michigan. David,Rosalie 2000 TheExperienceof AncientEgypt.Routledge,London andNew York. Dillehay,TomD. (editor) 1995 Tombsfor the Living: Andean MortuaryPractices. DumbartonOaks ResearchLibraryand Collection,Washington,D.C. Doyle, MaryE. 1988 TheAncestorCultandBurialRitualin Seventeenthand CentralPeru. Ph.D. dissertation,UniEighteenth-Century versityof California,Los Angeles. UniversityMicrofilms, Ann Arbor,Michigan. Duviols, Pierre 1973 Huariy Llacuaz.Agricultoresy pastores:un dualismo Revistadel prehispanicode oposiciony complementaridad. MuseoNacional, Lima39:153- 191. EspinozaSoriano,Waldemar 1967 Los senoriosetnicos de Chachapoyasy la alianzahispano-chacha.RevistaHistorica30:22$322. 1971-2 Los huancasaliadosde la conquista.Tresinformaciones ineditas sobre la participacionindigena en la conquista del Peru. 1558, 1560 y 1561. Anales Cientificos 1:9407. 1981 El fundamentoterritorialdel ayllu serrano.Siglos XV y XVI. In Etnohistoriay AntropologiaAndina, edited by Amalia Castelli, Marcia Koth de Paredes, and Mariana Mould de Pease, pp. 93-130. SegundaJornadadel Museo Nacional de Historia. Centro de Proyeccion Cristiana, Aguarico,Lima. Garcilasode la Vega,El Inca 1966 [1609] Royal Commentariesof the Incas. Part One. by HaroldV.Livermore. TranslatedandwithanIntroduction Universityof TexasPress,Austin. Guillen,Sonia E. 1999 Arqueologiade emergencia:inventario,catalogaci6ny conservaci6nde los materiales arqueol6gicos de los mausoleos de la Lagunade los C6ndores.Finalreportto theInstitutoNacionalde Cultura,Lima. Hare,Tom 1999 RememberingOsiris:Number,Genderand the Wordin AncientEgyptianRepresentationalSystems.StanfordUniversityPress,Stanford,California. Horkheimer,Hans 1958 Algunasconsideracionesacercade la arqueologiaen el valle del Utcubamba.Actas y Trabajosdel II Congreso Nacional de Historiadel Peru'1:71-101. Lima. Isbell,WilliamH. Postprocessual 1997 MummiesandMortuaryMonuments:A Prehistoryof CentralAndeanSocial Organization.University of TexasPress,Austin. Julien,CatherineJ. 1988 How Inca Decimal AdministrationWorked.Ethnohistory35(3):257-279. 145 KauffmannDoig, Federico 1988 Investigacionesarqueologicasen los AndesAmazonicos, 1980-1988.InstitutodeArqueologiaAmazonica,Lima. Langlois,Louis 1939-40 Utcubamba.Investigacionesarqueologicasen este valle del departamentode Amazonas. Revista del Museo Nacional 9(1):33-72, 9(2):191-249. Lima. Lerche,Peter 1995 Los Chachapoyasy los simbolosde su historia.Cesar Gayoso,Lima. 1999 A GraveCase of Robbery.Geographical71(5):18-23. Locke,L. Leland 1923 TheAncientQuipu,or PeruvianKnotRecord.American Museumof NaturalHistory,New York. Lounsbury,Floyd G. 1974 The Inscriptionof the SarcophagusLid at Palenque.In PrimeraMesaRedondade Palenque.PartII,editedby Merle GreeneRobertson,pp. 5-20. The RobertLouis Stevenson School, PebbleBeach, California. Loza, CarmenBeatriz 1998 Du bon usage des quipusface a l' administrationcolonialeEspagnole,1553-1599.Population53, No.2: 139-160. Mackey,Carol 1970 KnotRecordsin Ancientand ModernPeru.Ph.D. dissertation,University of California, Berkeley. University Microfilms,Ann Arbor,Michigan. 1990 Comparacionentre quipu Inca y quipu modernos.In Quipuy Yupana:Coleccion de Escritos, edited by Carol Mackeyet al., pp.135-155. ConsejoNacionalde Cienciay Tecnologia,Lima. Mackey, Carol, Hugo Pereyra, Carlos Radicati, Humberto Rodriguez,andOscarValverde(editors) 1990 Quipu y yupana: coleccion de escritos. Consejo Nacionalde Cienciay Tecnologia,Lima. Mills, Kenneth 1997 Idolatryand Its Enemies: Colonial Andean Religion and Extirpations,164S1750. PrincetonUniversityPress, Princeton,New Jersey. Murra,JohnV. 1975 Lasetno-categoriasde unkhipuestatal.In Formaciones economicasy polfticas en el mundoandino, pp. 243-254. Institutode EstudiosAndinos,Lima. 1982 The Mit'a Obligationsof Ethnic Groupsto the Inka State. In TheInca and Aztec States, 1400-1800, edited by GeorgeA. Collier,RenatoL. Rosaldo, and John D. Wirth, pp. 237-262. AcademicPress,New York. Muscutt,Keith 1998 Warriorsof theClouds:ALostCivilizationin the Upper Amazonof Peru.TheUniversityof New MexicoPress,Albuquerque. Nordenskiold,Erland 1925 Calculationswith Yearsand Months in the Peruvian Quipus.ComparativeEthnologicalStudies,Vol. 6, Part2. Goteborg. Parssinen,Martti TheInca Stateand Its PoliticalOrgani1992 Tawantinsuyu: zation. Societas HistoricaFinlandiae,StudiaHistorica43, Helsinki. Platt,Tristan 1986 MirrorsandMaize:TheConceptof Yanantinamongthe Macha of Bolivia. In AnthropologicalHistory of Andean Polities, edited by John V. Murra,Nathan Watchel, and JacquesRevel, pp. 228-259. CambridgeUniversityPress, Cambridge. Quilter,Jeffrey,and GaryUrton(editors) 2001 Narrative Threads: Explorations of Narrativity in 146 LATINAMERICAN ANTIQUITY AndeanKnotted-String Records.Universityof TexasPress, Austin,in press. Reichlen,Henry,andPauleReichlen 1950 Recherches archeologiquesdans les Andes du haut Utcubamba. Journal de la Socie'te'des Ame'ricanistes 39:219-251. Remy,Pilar 1992 El Documento. In Las Visitas a Cajamarca. 1571-72/1578, compiledandeditedby MariaRostworowski and Pilar Remy, vol. 1, pp. 37-108. Institutode Estudios Peruanos,Lima. RiveraSerna,Raul 1958 Libroprimerode cabildosde la ciudadde San Juande la Fronterade Chachapoyas.Revista "Fenix,"Biblioteca Nacional, Nos. 1l & 12. Lima. Robertson,MerleGreene 1983 The Sculpture of Palenque. Volume1. The Temple Inscriptions.PrincetonUniversityPress,Princeton,New Jersey. Rostworowskide Diez Canseco,Maria 1981 La voz parcialidaden su contextoen los siglos XVI y XVII. In Etnohistoriay AntropologfaAndina, edited by Amalia Castelli, Marcia Koth de Paredes, and Mariana Mould de Pease, pp. 3548. SegundaJornadadel Museo Nacional de Historia. Centro de Proyecci6n Cristiana, Aguarico,Lima. Rowe, JohnH. 1995 BehaviorandBelief inAncientPeruvianMortuaryPractice. In Tombsfor the Living:AndeanMortuaryPractices, edited by Tom D. Dillehay; pp. 27X2. DumbartonOaks ResearchLibraryandCollection,Washington,D.C. RuizEstrada,Arturo 1970 Exploracionesarqueologicasen el valledel Utcubamba. Culturay Pueblo 6, nos. 19-20. Salomon,Frank 1995 'TheBeautifulGrandparents': AndeanAncestorShrines andMortuaryRitualas Seen ThroughColonialRecords.In Tombsfor the Living:AndeanMortuaryPractices, edited by Tom D. Dillehay, pp. 315-354. Dumbarton Oaks ResearchLibraryandCollection,Washington,D.C. 2001 PatrimonialKhipusin a ModernPeruvianVillage:An Introduction to the 'Quipocamayos' of Tupicocha, Huarochirl'. In NarrativeThreads:Explorationsof Narrativityin AndeanKnotted-StringRecords,edited by Jeffrey Quilterand GaryUrton.Universityof TexasPress,Austin, in press. Savoy,Gene 1970 Antisuyo:TheSearchfor the Lost Cities of the Andes. Simon & Schuster,New York. Schjellerup, Inge R. 1997 IncasandSpaniardsin theConquestoftheChachapoyas. GOTARCSeriesB. GothenburgArchaeologicalTheses,#7. NationalMuseumof Denmark,Gothenburg. Smith,MarvinT., andMaryE. Good 1982 Early SixteenthCenturyGlass Beads in the Spanish ColonialTrade.CottonlandiaMuseumPublications,Greenwood, Mississippi. Urton, Gary 1994 A New Twistin an OldYarn:Variationin Knot Directionality in the Inka khipus.Baessler-ArchivNeue Folge, BandXLII:271-305. 1997 The Social Life of Numbers:A QuechuaOntologyof Numbersand PhilosophyofArithmetic.Universityof Texas Press,Austin. 1998 From Knots to Narratives:Reconstructingthe Art of HistoricalRecord-Keepingin theAndesfromSpanishTran- [Vol. 12, No. 2, 2001 scriptionsof InkaKhipus.Ethnohistory45(3):409438. 2001 Recording Signs in Narrative-Accountingkhipus. In NarrativeThreads:Explorationsof Narrativityin Andean Knotted-StringRecords,editedby JeffreyQuilterandGary Urton.Universityof TexasPress,Austin,in press. 2002 BinaryCoding in khipuRecordkeeping.In First Writing, edited by StephenD. Houston.CambridgeUniversity Press,Cambridge,in press. van de Guchte,Maarten 1996 SculptureandtheConceptofthe DoubleamongtheInca Kings. Res 29-30:256-268. von Hagen,Adriana 2000 NuevaiconografiaChachapoyas.Iconos:Revistaperuana de conservacio'n,artey arqueologia4:8-17. von Hagen,Adriana,and Sonia Guillen 1998 Tombswith aView. Archaeology51, No. 2:48-54. Wilford,JohnNoble 1997 MummiesMay Be of Incan Elite, After Conquestof 'CloudPeople.'New York7imes December16:F3. Zuidema,R. Tom 1977 The Inca Calendar.In Native AmericanAstronomy, edited by Anthony F. Aveni, pp. 219-259. University of TexasPress,Austin. 1989 A QuipuCalendarfromIca, Peru,with a Comparison to the Ceque Calendarfrom Cuzco. In WorldArchaeoastronomy,edited by AnthonyF. Aveni, pp. 341-351. CambridgeUniversityPress,Cambridge. Notes 1. Following Lerche (1995), I will observe the following orthographicconventionsregardingthe spelling of the name of thecultureandgeographicalregionunderdiscussionin this article. Chachapoyawill be used to referto the people, or the ethnic group,as well as the prehispanicpoliticalunity (probablya chiefdom),and the cultureunderreview here; Chachapoyas,as thename is spelled on most contemporarymaps of the region, will be used for the name of the geographicalregion within which the prehispanicand early colonial Chachapoyaethnic groupand culturewere centered. 2. For archaeological studies of Chachapoya and Chachapoya-Inkasites in the region of Leymebamba and Lagunade los C6ndores, see Horkheimer 1958; Langlois 193940; Reichlenand Reichlen 1950; Ruiz Estrada1970; and Schjellerup1997. 3. For general studies of the khipus, see Arellano 1999; Ascherand Ascher 1997; Brokaw 1999; Conklin 1982; Locke 1923;Loza 1998; Mackey 1970; Mackey et al. 1990; Murra 1975,1982;Parssinen1992;Quilterand Urton2001; and Urton 1994,1998. 4. Concerningthe conditionof the khipusupontheirdiscovery,accordingto Adrianavon Hagen (personalcommunication, 1999),the khipus were found scatteredamong the debris left behindafter the site was disturbedand looted. Several khipus thatare now storedas individualsampleswere originallyfound tiedtogether into linked bundles of khipus. These groupings wereuntiedby CentroMallquistaffas an essentialstep in cleaningand conservingthe khipus.In addition,one khipu a (once) magnificient sample composed of some 266 dyed cotton pendantstringsaffixed to a carved wooden stick was washed in detergent by the wife of the hacendado whose workmen discovered(and plundered) the burial site at Laguna de los Urton] A CALENDRICAL AND DEMOGRAPHIC TOMBTEXTFROMNORTHERN PERU C6ndores.This latterkhipuis now virtuallycompletelywhite! I am currentlyworkingto constructan overviewof the historyof the khipusfollowing their recovery.This account will be published in my complete study of the khipusfrom Lagunade los C6ndores(in preparation). 5. Anotherarchaeologicalcollection of khipusfor which we have good proveniencedata and archaeologicalcontext is from the central coastal Peruvian site of Puruchuco(see Mackey 1990). Anothercollection of khipusof exceptionalimportance for theirhistoricalandethnographiccontexts,which is currently the object of intense and importantinvestigationsby Frank Salomon, is that which is currentlyin use in the village of Tupicocha, in the central Peruvian highlands. The khipusin Tupicochaare broughtout for ritualdisplay on the occasion of public ceremoniesconductedby the headmenof the ayllus.As to their date of manufacture,four samples taken by Salomon from the Tupicochankhipusyielded acceleratordates ranging between 1650 and 1950 (Salomon2001). 6. While this article was in press, the authorreceived the results of the radiocarbondating of samples from three of the khipusfromLagunade los C6ndoresandfroma textilefragment foundassociatedwith one additionalkhipu(table3). These data, providedby the NSF ArizonaAMS Facilityat the Universityof Arizona,are presentedhere for the firsttime. From the this data, we see that the mean age (= 399 + 35 years B.P) of three of the samples (T12818A, T128819A, and T128822A) falls within the calibrated (2<s) date range: 1426-1630. This dating is consistentwith the interpretationof the khipusgiven in this article,in which theiruse is seen as covering the period from late pre-Hispanicthroughearly colonial times. The fourthdate shown above (T12821A) is a bit of an outlierdate in this group of samples and will be consideredin future publications along with additionalradiocarbonassays thatare still in process at this writing. 7. The current museum accession number at the Centro Mallqui,in Leymebamba,Peru,of the khipuwhich I have designatedherein as khipuUR6 is: CMA 625/LC1-254. However, I would note that in the summerof 1999, this same khipuwas designated as CMA 1889/LC1.052 (see Figure 4) in the museum inventory.In designatingthe khipusin this collection with labels beginningUR (=Urton),I am following the convention establishedby the pioneers of khipudocumentarystudies, Marcia and RobertAscher (see especially Ascher and Ascher 147 1978). Each khipustudiedby the Aschers is designatedin their reportby an AS (=Ascher) number,therebyindicatingto later researchersthe source(s) of measurementsand observations made on the khipusamplesin question. 8. For previous studies of possible calendrical values encodedinto khipus,see Nordenskiold1925 andZuidema1989. 9. From my close studies of about 375 khipu samples in museumcollections in Peru,Europe(Germany),and the U.S., I havefoundthatapproximatelytwo-thirdsof the samplesin most collections have their knots tied in the decimal place notation system, thus stronglysuggestingthatthe dataon these samples were of a numerical-statistical nature.However,aboutone-third of the samplesstudieddo not exhibitknotstied in a hierarchical, decimal-place notation manner; that is, these samples have "units"knots (i.e., figure-eightand long knots)tied above single knots,the latterof which, dependingon theirplacement,usually designate increasinglyhigher powers of 10. In addition,there are examples of the so-called long-knots,which normallyhave a maximumof nine turnsto the string(any greatervaluebeyond which the long-knotshouldbe replacedby a single knot, in the lOs position)thathave up to 16 turns.Khipusof this lattertype, which I have referredto as "anomalous"and"narrative-accounting" khipus,are describedand analyzedin Urton2001. 10. As I have reportedin a studyof the constructionfeatures of khipus in the American Archaeology collections of the Museum fur Volkerkunde,in Berlin (Urton 1994), khipuknots aretied in two differentways to produceeitherS-knots(in which the dominantaxis of the knotrunsfromupper-leftto lower-right (= \) or Z-knots(in which the dominantaxis runs from upperrightto lower-left(= /). 11. While admittingthat it is possible, my colleague Keith Muscutt, who has spent many years exploring in the Chachapoyasregion (see Muscutt 1998), doubts that Guaman wouldhavebeen buriedin the tombsat Lagunade los C6ndores. Instead,he favorsas a final restingplace for Guamaneitherthe cliff-tomb burial site of La Petaca (which is nearer to Cochabamba)or othercliff tombslocatedaroundthe lakes of La Sierpeand Mishacocha,at Atuen (Muscutt,personalcommunication, 2001). ReceivedNovember3, 2000; acceptedJanuary13, 2001; revisedFebruary14, 2001. Table 3. Radiocarbondating of samples from three of the khipus from Lagunade los C6ndoresand from a textile fragment found associated with one additionalkhipu. Lab ID T12818A T12819A T12821A T12822A Date B.P. 414 + 35 403 + 35 .834 + 35 379 + 34 Date cal. A.D. 2<s 613C 1426-1624 -26 1433-1626 -21.9 1068-1278 -24.5 1443-1630 -22.5 Material cotton cotton cotton cotton Context khipufrom siteLC2 khipufrom siteLC1 khipufrom siteLCl textile from site LC1 (assocd. w/ khipu CMA480)