Matrilocality, Corporate Strategy, and the Organization of Production In

Home , Matrilineality, Matrilocal residence, Neolocal residence

MATRILOCALITY, CORPORATE STRATEGY,AND THE ORGANIZATION OF PRODUCTION IN THE CHACOAN WORLD

Peter N. Peregrine

Kinship is central to an understandingof sociopolitical organization and the organization of production in Chaco Canyon. Between A.D. 700 and 900, lifeways in the Chacoan wor7ldunderwent a transformationthat reflects the evolution of matrilocal residence. Matrilocal groups became the foundation of a polity based on a corporate political strategy.Matrilocality pro- vided the peoples of the Chaco region a social structure in which women were able to form stable agricultural communities while men werefreed to take part in long-distance resourceprocurement and trade. Leaders mobilized goods from across the polity for corporate-affir-mingactivities, such as construction of great houses. This corporate strategy is evident in an exami- nation of turquoiseproduction. The production of turquoise ornaments was diffuse and uncontrolled, but the consumption of turquoise,at least in Chaco Canyon, was highly centralized. This is what we would expect to see in a polity operating within a corporate strategy.

Elparentesco es centralpara entenderla organizaci6nsocio-politico y la organizaci6nde produci6n en Cafi6n Chaco. Entred.C. 700 y 900 la manera de vida en el mundo Chaquefiose cambi6 reflejandola evoluci6n de la residencia matrilocal. Los grupos matrilocales convenieronla fundaci6n de una regi6n politica basada en una estrategiapolitica corporativa.La matrilocalidad suministr6a la gente Chaquefiasuna estructurasocial en que las mujeresfueroncapaz aformar comunidadesagriculturales esta- bles mientraslos hombresfueronlibres a conseguir recursosy hacer neg6cios de larga-distancia.Los lider-esmobilizar-on recur- sos al trave'sde la regi6n politica para actividades corportiva-afirmando,como la construci6n de las Casas Granzdes.Esta estrategia es evidente en una investigaci6nde la produci6n turquesa.Los ornamentosturquesos fuei-on fabricados por familias quien especializaronen la produci6nartistica pero la produci6nfue difusay incontrolada.Sin embai-gola consunci6nde turquesa fue altamenteespecializada, al menos entre Can6n Chaco. Eso es precisamenteque espe7rar-iamosver en una region politica fun- cionando entre una estrategia corporativa.

A mong the most striking monuments of the evolution of matrilocal residence and, more ancient North America are the Chacoan specifically,of a polityrooted in the possibilitiesand greathouses. Generations of archaeologists constraintssuch residentialgroups create. I suggest have marveledat theirscale and craftsmanshipand emergentleaders in the Chacoanpolity used matrilo- wonderedwhy they were built, what purposesthey cal groupsas the foundationof a corporatepolitical served,what kind of society fosteredthe construc- strategy.As my colleagues and I have explainedin tion of such impressivestructures. These questions a series of recentarticles (Blanton 1998; Blantonet are made only more enigmaticwith knowledge of al. 1996; Feinman 2000), leaders following a cor- the archaeologicalrecord for the erapreceding great porate political strategy attemptto build a power house construction.BetweenA.D. 700 andA.D. 900 base by developing and promoting activities that lifeways in the Chacoanworld underwenta trans- reinforcethe corporatebonds that tie members of formation.Single-family pit structures,a basic res- the polity together.A commoncorporate strategy is, identialform since as early as 200 B.C., were given for example,to mobilize goods from across a polity up for multifamilypueblos. What accountsfor this for large public ritualsor constructionprojects that transition?What does it meanfor ourunderstanding bring membersof the polity togetherin corporate- of the organizationof productionin the Chacoan affirmingactivities (activities that seem reflectedin world? the Chacoan great houses and pre-Chacoangreat I suggestthe developmentof pueblosmay reflect kivas).

Peter N. Peregrine * Departmentof Anthropology,Lawrence University, 515 E. College Ave., Appleton, WI 54911

AmericanAntiquity, 66(1), 2001, pp. 36-46 Copyright( 2001 by the Society for AmericanArchaeology

36 MATRILOCALITY,CORPORATE STRATEGY, AND ORGANIZATION 37

My colleagues and I have arguedthat corporate or status,receive no tributeor compensationfor their strategyshould be seen as one end of a continuum work,and are powerlessin termsof judicialauthor- of politicalstrategy with the oppositeend being what ity. Yet Leach (1954:206) argues that in terms of we have called an exclusionaryor "network"strat- politicalauthority, "in practice a gumlao 'headman' egy. Leadersfollowing a networkstrategy attempt maybe hardlydistinguishable from a gumsa'chief."' to build a power base by controllingaccess to net- How do such gumlao leaders gain authority?The worksof exchangeand alliance both within and out- answerLeach (1954:263, 275) gives is thatthey do side the polity (Blantonet al. 1996:4-5; Peregrine so by appealingto corporateideology and support- 1992).Thus a networkstrategy is one in which lead- ive myths andprinciples of behavior.It is important ers attemptto monopolize sources of power,while to note thatpolities with a gumlaoorientation are not a corporatestrategy is one in which leadersattempt hinderedin what they can accomplish-they build to sharepower across differentgroups and sectors andmaintain agricultural terraces, defend themselves of a polity. against enemies, engage in trade with other poli- It is importantto note thatcorporate and network ties-and yet they do so within a frameworkthat strategies lie on a continuum and are not "pure" downplaysstatus differences and individual aggran- strategiesin any given case. Rather,leaders tend to dizement(Leach 1954:231-238). promotemore corporate-oriented strategies or more While the idea of corporatepolitical strategy was network-orientedones, but it is unlikely any leader initiallyput forwardas partof a theoryfor the evo- will promotea "purely"corporate or networkstrat- lution of complex societies in Mesoamerica(Blan- egy. It is also importantto make clear that these ton et al. 1996), severalof the papersin this issue of strategiesare used in societies with varyingdegrees AmericanAntiquity use it as a model of political of political centralization,scale, and technological behavior to provide insight into the structureand complexity.They do not define societal "types"nor operationof the ancient Chacoanpolity. Timothy do they define a unilinealevolutionary trend. Earle,for example,explains that the focus of polit- While archaeologistshave tendedto ignore cor- ical strategyin the Chacoanpolity appearsto be "sta- poratelyoriented polities in favorof the often more ple finance"-the mobilization of goods for colorfulnetwork-oriented ones, severalgood exam- corporateactivities. Similarly,Colin Renfrew sug- ples do exist, perhaps the most prominentbeing gests thatgreat houses were used for largepublic rit- Teotihuacanand the Induscivilization. In bothcases uals,rituals coordinated and funded by leaderswhose we find few of the typical indicatorsof powerful activitiesmaintained the corporateorientation of the rulers-there are no royal tombs, few or no repre- polity.While not directlyemploying the ideaof a cor- sentationsof rulers, and where writing is present porate-orientedpolity, he suggests thatthese rituals rulers are rarely mentioned-such polities seem also servedto legitimatethe authorityof the leaders "faceless"(Blanton 1998:149). And yet these "face- who hosted them,just as the mobilizationof mate- less" polities were able to constructcities, conquer rials used in them did.Thus both Earle and Renfrew largeregions, and integrate tens of thousandsof peo- suggestthat a basic politicalstrategy in the Chacoan ple. SimilarlyColin Renfrew (1974) describedwhat world was the coordinationof corporaterituals. he called "group-oriented"chiefdoms in thirdmil- Whateffects might a corporateorientation have lenniumB.C. Europethat were capableof organiz- on the organizationof production?We would expect ing laborto producelarge public works suitablefoi productionin corporate-orientedpolities to be dif- communal ritual within an apparentlyegalitarian fuse, with no clear centersof control,and to be dis- frameworkwhere political authoritywas "faceless persed across the polity. This is what we appearto and anonymous"(1974:79). see in the Chacoan world. We would also expect The ethnographicliterature provides another pic- productionto be directedtoward group rather than ture of such "faceless" polities. Leach (1954) individual consumption. Again, this is what we describesgumlao chiefdoms organized around egal- appear to see in the Chacoan world. We would itarian principles among the Kachin of highland expect consumption, on the other hand, to be Burma.Leaders in gumlaopolities (who Leach terms focused at points of grouparticulation. In the Cha- "headmen"to distinguish them from the gumst coan worldthose points of articulationappear to be chiefs of otherKachin polities) have no specialrights greathouses andgreat house middens.Thus the idea 38 AMERICAN ANTIQUITY [Vol. 66, No. 1, 2001 that the Chacoan polity was corporatelyoriented -xample,the dataon small sites presentedby Peter seems both a reasonableand useful one. It appears M\1cKenna(1986) suggest that during the period to fit the patternof the materialrecord and provides betweenA.D. 500 and 700, pit structuresaveraged useful insights into how productionmay have been about15 squaremeters in floor area,and this aver- organized. age size remained fairly constant in the period between A.D. 700 and 900, although variationin Matrilocal Residence in the Chacoan World size seems to have increased. When the pueblos Ratherthan furtherdiscuss the natureof corporate beganto be constructed,in the periodbetween A.D. political strategiesin the Chacoanworld (which is 900 and 1100, individualrooms averagedabout 7 discussed in more detail by Earle and Renfrew),I squaremeters in floor area,but each site consists of would like to focus on the context in which those an agglomerationof theserooms, giving overall sizes strategieswere employed-the context of matrilo- of the room blocks a rangefrom roughly70 square cal residence.As I statedabove, I believe thatthe pit- meters (29-SJ-629) to over 300 squaremeters (29- structure-pueblotransition reflects a transitionto SJ-633). Clearlydwelling size increasedin a man- matrilocalresidence (also see James 1994), and I ner consistent with the evolution of matrilocality, have two fundamentalreasons to accept this asser- and, in fact, the conclusion reached by McKenna tion.The first is thata numberof themost likely mod- (1986:38), thatthis change marksa transitionfrom ern descendants of the Chacoan people are single-familyto multifamilydwellings, is also con- matrilocal.Could the inhabitants of the Chacoregion sistent. havebeen matrilocalfrom very early in theirhistory? If we accept that the pit structure-pueblotransi- It seems unlikely.The Archaic hunter-gatherers who tion marks the evolution of matrilocal residence populatedthe AmericanSouthwest were probably amongthe inhabitantsof the Chacoregion, it is rea- patrilocalor neolocal, as patrilocalresidence is by sonableto ask why matrilocalityevolved andwhy it farthe most commonform of post-maritalresidence evolved when it did. The answersto these questions in the world (about60 percentof known societies), go well beyond the purposeof this paper,but there andneolocal residence appears to be a commonadap- are several general conditionsthat appearto foster tationof hunter-gatherersin marginalenvironments. matrilocality,each of which may have played a role Matrilocalresidence only occursin about10 percent in the Chacoancase. Matrilocalresidence has been of known societies, and thereis an unusualcluster- shownto evolve whereeither (1) a groupof migrants ing of matrilocal groups among likely Chacoan has entered a previously inhabitedregion (Divale descendants(including the modernHopi andZuni). 1984); (2) there is frequentextra-community war- It is very likely, therefore, that this "pocket" of fare (Emberet al. 1974); (3) significantlong-dis- matrilocalityhas relatively recent roots (Martin 1950; tance traderelations have recentlybeen introduced Steward1955). (Peregrine1994); or (4) a group'scultural practices My secondreason for suggestingthe adoptionof are understress due to contactwith more dominant pueblo dwellings reflects the evolutionof matrilo- societies (Helms 1970). These four explanations cal residence comes from Melvin Ember's (1973; share a common element: competitive interaction replicated by Divale 1977; also see James with othersocieties. 1994:120-124)finding that matrilocal societies have Why does competitiveinteraction foster matrilo- significantlylarger dwellings than patrilocalones. cality?Two answershave been given: (1) matrilocal The dwellingsare larger because they house the fam- residence creates a culturally stable core group ilies of severalrelated women who togetherform a throughwhich children can be raised in a "tradi- singleeconomic (and residential) unit under matrilo- tional"manner (Helms 1970; Peregrine1994); and cal residence.Ember's (1973) findingssuggest that (2) matrilocalresidence allows men to be absent societies with patrilocal residence tend to have fromthe core groupfor long periodsof time for trad- dwellingsthat are less than60 squaremeters in floor ing, raiding,or diplomacy(Ember et al. 1974;Pere- area, while matrilocal societies tend to have grine 1994). Both of these explanationsare really dwellingslarger than 100 squaremeters in totalfloor oppositesides of the same coin. On the one side, the area.Such a shiftis clearlyevident in ChacoCanyon matrilocalhome groupallows domesticlife to con- dwellings in the periodfrom A.D. 700 to 100. For tinue with stabilityeven in the face of political and Peregrine] MATRILOCALITY,CORPORATE STRATEGY, AND ORGANIZATION 39 economicactivities taking men awayfrom the home matrilocalresidence (see Driver1956 versusDivale groupfor long periodsof time. On the otherside, the 1984), the datado clearlysupportAberle's argument fact thatmen marryout of theirnatal group appears for NorthAmerican societies (Divale 1974; Ember to foster the developmentof fraternalassociations and Ember 1971). thatcan coordinatethe laborof largegroups of men Subsistenceproduction in a matrilocalChacoan for activitiesof pan-societal,rather than natal group world would be organizedaround women's work or lineage, importance. groups.These workgroups, in turn,would likely be Competitiveinteraction has been routinely put composed of membersof individualmatrilineages forwardas an explanationfor both the development who controlaccess to the areasof agriculturalland of pueblodwellings and of politicalcentralization in they work. The matrilineageswould similarlycon- Chaco Canyonitself. Wilcox (1993), for example, trolaccess to the agriculturalproducts produced from arguedthat the Chacoanpolity beganas a refugefor that land, and women's work groupswould handle displacedpopulations from the northernSan Juan the tasks of processingand storingthose products. basin,and became centralized as a militaristicpolity Thus we can envision Chaco Canyonbeing divided with an eye to expansionand conquest.Such situa- up into areasof landcontrolled by local matrilineages tions are ones that have repeatedlybeen suggested residentat nearby pueblos. Groups of relatedwomen, as fosteringmatrilocal residence (see Ember et al. althoughlikely with some help fromhusbands, sons, 1974; Helms 1970). Similarly, Weigand (1992; or brothers,would prepareand sow fields of corn, Weigandand Harbottle 1993) arguedthat an expand- beans, and squash,rotating those fields on a regular ing tradenetwork with Mesoamericawas centralto basis aroundthe areas of land they control.!These sociopolitical developmentsin Chaco Canyon. In groups would harvestthe crops, and preparethem this scenario,one can easily envisionmatrilocal res- for storageby drying the beans and squashes,and idence being fostered as a way to allow groups of grindingthe corn.Women would have been the most malesto leave for long periodsof time on tradeexpe- likely producersof basic storage utensils; that is, ditions (see Peregrine1994). Thus, while I am not ceramics.Ceramic production was likely organized able to put forward a specific argumentfor why in muchthe sameway as subsistence,through groups matrilocalityevolved in the Chacoanworld when it of relatedwomen. These groupswould have formed, did, the sociopoliticalconditions of the region dur- slipped,fired, and paintedthe ceramicsas a group, ing the period from roughlyA.D. 700 to A.D. 900 with the whole processtaking several days andper- (and perhapseven earlier) seem ripe for fostering hapsbeing repeatedonce every few months,as new matrilocalresidence. ceramicswere needed. Whatwould men contributeto subsistence?The Matrilocal Residence and the simple answer is meat. Hunting,especially during Organization of Production the Pueblo Bonito phase and later,would have been If we acceptthe idea that the pit structure-pueblotran- a time-consumingactivity, likely requiring the hunter sition reflectsthe evolutionof matrilocalresidence, to travelsome distancefrom Chaco Canyonto find what are the implicationsfor our understandingof game. This is precisely where matrilocalresidence Chacoanpolity and, in particular,the organization providesgreat benefits, for it allows groupsof men of production?A good startingpoint, I suggest, is to be absent for long periods of time without dis- the ideaput forward by DavidAberle(1961:661) that rupting the core community (Ember et al. 1974; "Matrilinealgroups arise ... in connection with Helms 1970; Peregrine1994). In terms of utensils, women'swork groups and the resourcebases which men wouldlikely havemade their own huntinguten- these groups control."In his cross-culturalexami- sils; thatis, chippedstone tools. Unlike ceramics,it nation of matrilineal societies (and virtually all is unlikelythat men would have done so in groups, known matrilocal societies are also matrilineal), but ratheras individualsproducing tools for them- Aberle(1961:670) also arguesthat "matriliny is most selves. Particularmen withunique skill in flint-knap- likelyto developon a horticulturalbase, with women ping might producetools for others,but likely not doing the agriculturallabor." WVhile there has been on a large scale. Raw materialswould be procured some debateover the importanceof women's con- either by directly obtainingthem from the source, tributions to subsistence on the development of perhapsin the courseof huntingexcursions or while 40 AMERICAN ANTIQUITY [Vol. 66, No. 1, 2001 visitinghome communities.However, the samefree- thatalienation are often found.One form is the kind dom that matrilocalresidence would have allowed of fraternal associations mentioned above, but men to be absent for long periods of time hunting another,related, form might be cross-community wouldhave also allowedthem to spendtime actively associations.A thirdmight stem from createdrela- procuringraw materials from long distancesor trad- tions of debt throughgifts and reciprocalfeasting. ing withdistant communities. Indeed, we mightenvi- In all cases these would requirethe communitiesto sion a large, formal networkof traderelationships become increasinglyinteractive. This leads, in turn, spreadacross the Chacoanworld much like the Kula to the notionthat "isolated communities (or smaller ring of the TrobriandIslanders (who are also matri- groups)consisting of matrilinealcore andin-marry- lineal-see Malinowski1922). ing spouses are extremely difficult to maintain" Men would also be the most likely producersof (Schneider1961:27). Thus, the increasedinteraction special craft work and, indeed, of the very pueblos so evident archaeologically in the Chaco region I am suggestinggive evidence of Chacoanmatrilo- would be a logical, perhaps necessary, result of cality.The rawmaterials for thepueblos and for craft matrilocalresidence. items would be obtainedeither directlyor through trade,but in eithercase men would have to be able Matrilocal Residence and Corporate Strategy to leave the home groupto obtainthese items, and The ethnographicrecord suggests that matrilocal res- in large work groups, particularly if the work idence often encouragesleaders to adopta network involved transportingbeams, or mining stone and ratherthan corporate strategy (see Peregrine1992 for minerals.Production of craftitems would likely be examples).This makes some sense as the ease with doneby individuals,but matrilocality could promote which matrilocalityallows large groups of men to some form of fraternalcraft "co-ops"at individual leave for long periodsof time to tradeor raidseems pueblos, and certainlythe constructionof pueblos to fit nicely with a networkstrategy. The Chacoan themselves would requirea group of coordinated polity, then, may be somewhatunusual in combin- workers(Lekson 1986:257-273; Wills 2000). Such ing matrilocalresidence with a corporatestrategy. workor craftgroups would be createdin partto inte- Why did Chacoanleaders pursue a corporatestrat- gratemen intotheir marriage community (much like egy? An importantfactor may have been the mar- quilting societies did for ruralAmerican women in ginalnature of the Chacoanenvironment, at least for the nineteenth century). Such integrativemecha- an agriculturaleconomy (Judge1979, 1989). Places nisms are importantin matrilocalsocieties, for in- where network strategiesthrive, such as highland marryingmales have few formalmeans to tie them New Guinea,western Polynesia, and the northwest to the community-they are not importantin the coast of NorthAmerica, tend to have temperatecli- domesticeconomy, and they have no role in thepolit- matesand extremely rich environments.One factor, ical life of the community.Thus fraternalassocia- indeed, that may promotea networkorientation is tions and interestgroups often develop as means to thefact that virtually any family or lineagegroup can create bonds among in-marryingmen, as well as survive,at least in termsof food, shelter,and other bonds between these men and the community. necessities of life, in isolationfrom the rest of soci- Such fraternalassociations and interest groups ety.What network-oriented leaders control is access would, indeed, extend to other communities. As to socially relevantgoods, knowledge, and events, David Schneider (1961:20) explained this issue, and not to things needed for survival (Peregrine "Matrilinealdescent groupshave special problems 1992). A networkstrategy may not be very success- in the organization of in-marrying affines with ful in a situationwhere families or lineages cannot respect to each other" and, I would add, to the easily survive independentof others. Leaderswho matrilocalgroup as a whole. In-marryingmales may stresscooperative activities aimed at mutualsupport carrypolitical authority and high statusin theirnatal and groupsurvival may be more successfulin gain- community,but have none in the communityof their ing followers than those who stress individual wife. There,they may be subjectto the rule of other advancementand the exclusion of othersfrom net- males to whom they feel superioror contentious.In works of power. short,men become alienatedin matrilinealsocieties How did Chacoanleaders implementtheir cor- (Schneider1961:20), and some meansof alleviating porate-orientedpolitical strategies?I suggest there Peregrine] MATRILOCALITY,CORPORATE STRATEGY, AND ORGANIZATION 41

Table 1. Counts of GeneralAbraders, Lapidary Abraders, and Ceramic Polishers by Location.

Observed (expected) GeneralAbraders Lapidary Abraders Ceramic Polishers Total Room 211 (182) 28 (36) 3 (11) 242 Pithouse 131 (141) 24 (29) 32 (18) 187 Ramada/Plaza 34 (41) 16 (8) 5 (5) 55 Kiva 75 (86) 21 (17) 18 (11) 114 451 89 58 598 %2= 4067, df= 6,p< .001 Source Data adoptedfrom Akins (1997): general abraders= types 10 and 20; apidaryabraders = types 12 and 22; ceramic polishers = type 30. are at least three areas where a corporatestrategy house communities. Thus I agree with Windes seems manifestedin the archaeologicalrecord. First, (1992:162) that, at least from a community per- it seems clearthat Chacoan leaders promoted events spective, "participationin turquoisejewelry manu- thatbrought matrilineages together for commonpur- facturewas universal,or nearlyso." poses at the greathouses. This idea is discussed in The availabledata also suggestthat turquoise pro- some detail by Earle and Renfrew in their papers, ductionwas not a spatiallylocalized activitywithin and I will not elaboratehere. Second, I suggest that communities,but rather took place in a varietyof con- the sharedstylistic featuresin ceramics and archi- texts. Mathien's (1984, 1997) identification of tecture, features that also undergo coordinated turquoise"workshops" demonstrates that they were changethough time, may reflect the sanctioningby locatedin kivas,pithouses, rooms, andplazas. Thus Chacoanleaders of efforts to downplaydifference there is no single specialized location where and division between matrilineages(see Peregrine turquoisebead production took place. Is therea loca- 1991 for similarethnographic cases). Chacoanlead- tion where turquoisework was more likely to take ers may have used their influence to promote an place than other activities?It appearsnot, as evi- emphasison commonbonds of relationshipamong dencedby the datapresented in Table 1. matrilineagesthough shared stylistic patterns, ritual Table 1 shows the cross-tabulationof general practices,and the like. Third, I suggestChacoan lead- abraders,lapidary abraders, and ceramic polishers ers encourageddiffuse production of importantreli- by theirlocation in excavatedcontexts within Chaco gious or ceremonialgoods thatwould be pooledinto Canyon.General abraders should tend to be located common activities.This would be in starkcontrast whereunspecialized daily activities take place, while to network-oriented strategies in which leaders ceramicpolishers should be locatedwhere ceramic tightly manage such goods through the use of productiontakes place. By comparingthe locations attachedspecialists and sumptuary rules (Blantonet where lapidaryabraders are found (Figure 1), one al. 1996:5;Peregrine 1992). should be able to determinewhere turquoisebead Turquoiseappears to be a goodexample of anitem productiontook place.The valueof chi-squaredsug- with symbolicimportance in the Chacoanworld that gests that the distributionof these items is not ran- is produceddiffusely but consumedwithin limited dom, and if we examinethis table more closely, we contexts. Indeed, the available data suggest that see thatgeneral abraders are much morelikely to be turquoiseproduction is an exampleof whatFlannery foundin rooms thanchance would suggest,but that and Winter(1976:38-39) term "householdspecial- ceramicpolishers are much less likely to be found ization";that is, activities that take place in every there. Ceramicpolishers appearmuch more likely communitybut not in every household.While not than would be suggestedby chance to be found in every Chacoancommunity has in situ evidence for pithouses or kivas, and lapidaryabraders are more turquoisebead production, Windes's (1992) studyof likely to be found in ramada or plaza areas than turquoisedebris in foragingant moundsfound that chance would suggest. This patternholds up if we 94 percent of Pueblo II period sites had turquoise examine only lapidaryand general abraders(X2 debris.Roughly half of all excavatedsites in the val- 15.25, df = 3, p < .01), but changes slightly if we ley do have in-situ evidence of turquoisebead pro- examineonly lapidaryabraders and ceramic polish- duction,and such evidenceis foundin every areaof ers. In this case, lapidaryabraders are more likely to Chaco Canyon and in both great house and small be foundin roomsthan would be expectedby chance 42 AMERICANANTIQUITY [Vol. 66, No.1, 2001

Arl

410 4w 'b 46

EX - to 0, .j - I F S'. V,* 4A 7, 041 A t- vo, 'P it PMV " yo -i - rVej 'k __1T_

wi-W

10 W-WO-Z.

A .0. - v gw

4-1

lk_W4

ALAW W-11-MON"------

1. abrader recovered from Pueblo Bonito. From I Figure0 O Lapi'dary Judd (1954), figure Institution, Nati'onalAnthropolo cal Archives.

(x2 = 21.29, df = 3,p < .001). If we take the presence bead seems reasonable. If this is the case, then it took of lapidary abraders as an indicator of turquoise bead roughly 200,000 person-hours of labor to produce production (Akins 1997:733), it appears that there all the known turquoise beads from Pueblo II Chaco were no special locations where turquoise bead pro- Canyon. This may seem like quite a lot, but if we duction took place; rather, turquoise beads were pro- divide this by the 150 years of the Pueblo II period, duced in a variety of locations, a preferred one then we are talking about something less than 1,500 perhaps being outside, in the shade of a ramada or beads per year. Using the low estimate of Chacoan on a plaza. population, 2,500, and suggesting that only adults The image we get from this conclusion is that made beads, this figure would only require a couple turquoise bead production was not very intensive, but of beads a year from each manufacturer. Clearly this rather was done in favored locations in peoples' is misleading, because not all produced beads have "spare time." This seems in stark contrast with some been found, and not all adults were likely making scholars suggestions that beads were produced by them, but it does demonstrate that turquoise bead pro- elite-influenced (and perhaps controlled) specialists duction could have easily been done in the spare (Weigand 1992:174), or in factory-like conditions for time of a handful of people in each community-it market exchange (Kelly and Kelly 1975:204). It also need not have been an intense or specialized pro- seems impossible that the vast number of turquoise duction regime. beads found in Pueblo II contexts, minimally esti- While turquoise bead production appears to be mated at some 200,000, could have been produced completely decentralized, the pattern of consump- as a leisure activity. tion is totally different. Of all the beads found in exca- Experiments on turquoise bead production by vated contexts in Chaco Canyon, roughly one quarter Haury (1931) and Jernigan (1978), informedby more come from two rooms in Pueblo Bonito, and the detailed experiments on shell bead production by majority of them from two individual burials (Color Francis (1988), suggests that a rough estimate of 45 Photo 5, p. 67). Thus, while production is diffuse and minutes to 1 hour of work would be required to pro- uncontrolled, consumption, at least within Chaco duce a single turquoise bead (Figure 2). This includes Canyon itself, was highly centralized. This is pre- roughly 15 minutes to shape the bead, 15 minutes to cisely what we would expect to see in a polity oper- drill the bead, 5 to 10 minutes to round and polish ating within a corporate strategy. Turquoise, a the bead, and 5 to 20 minutes for raw material acqui- relatively "exotic" material with rich symbolic mean- sition and initial processing (including rough shap- ing in the Puebloan world, was produced by every- ing and heat treating).2 Thus, an estimate of the time one capable or interested in doing so. It was involved in making turquoise beads of one hour per consumed, however, by particular individuals in spe- Peregrine] MATRILOCALITY,CORPORATE STRATEGY, AND ORGANIZATION 43

cific contextsand only afterbeing collected into mas- sive amounts.It is interestingthat the PuebloBonito burials that were accompanied by most of the turquoisefound in the canyonwere locatedadjacent to otherapparently ritual facilities or storagerooms. It may be that these are not burialswhose accom- panyingriches display conspicuous consumption by a wealthyelite as has oftenbeen suggested,but rather burials of corporateleaders containing gifts from individualsthroughout the polity as a means of dis- playingthe corporatebonds manifestedthrough the deceased.

Conclusions Chaco was a corporate-orientedpolity, but it was, perhapsmore importantly,a matrilocalpolity. Cor- poratestrategies at once fosteredand were fostered by the evolution of matrilocalresidence. Matrilo- cality providedthe peoples of the Chaco region a social structurein which women were able to form stable agriculturalcommunities while men were freedto takepart in long-distanceresource procure- ment and trade.Political leaderstook advantageof this situation and developed strategies to build a powerbase from them. Out of this matrilocalstruc- tureand the associatedcorporate political strategies arose the large communitiesthat came to dominate the Chacoanworld, the intenseinteraction between those communities,and the rise of influentialpan- societal fraternalgroups. To understandthe "Chaco phenomena,"the organizationof productionin the Chacoanworld, the intenseinteraction between Cha- coan communities, or indeed any other aspect of Chacoansociety, I arguethat one must approachit throughthe lens of matrilocalresidence. To argue that kinship is central to Chacoan sociopoliticalorganization, and to the organization of productionin particular,should not be surprising to us as anthropologists.And yet in the literatureon the Chacoanworld we find little discussionof kinship, and we even find scholarssuggesting that the Chacoanarchaeological record "indicates a degree of centralizationand political specialization beyond the capacityof societies whose decision making is carriedout within the frameworkof kinship . . ." Figure 2. Turquoise beads recovered from Pueblo Bonito (Sebastian1991:119). This is unfortunate,for it con- averaged 4 mm in diameter, most were discoidal in shape, tradictswhat we know aboutsocieties from all over and are flat-sided. Variations in form are shown here. From Judd (1954), figure 12, drawn by Irvin E. Alleman. the world, even highly politically centralizedones Courtesy of the Smithsonian Institution, National Anthro- like Natchez, Tonga, and Kongo. I have suggested pological Archives. elsewhere that archaeologistsoften confuse ethno- 44 AMERICAN ANTIQUITY [Vol. 66, No. 1, 2001 graphicanalogy with ethnology (Peregrine1996). system dominating the Chacoan world (Lekson Here I will go fartherto suggest that archaeologists 1997), I suggest thatwhat we see is an overarching oftenforget the lessons ethnology taught us as anthro- social system, a social system based on matrilocal pologists. To argue,as an anthropologist,that kin- residenceand the demandsand opportunitiesit cre- ship is not importantto understandingany given ated. society is problematic;to suggest it is unimportant Acknowledgments.I want to thankall theparticipants in the Chaco to understandingthe organizationof a non-statesoci- Organizationof Production conference-Tim Earle, Melissa ety is ridiculous.Kinship in all non-statesocieties Hagstrum,Steve Lekson, Peter McKenna,Joan Mathien,Colin structuressocial relations.It defines who a person Renfew, Tom Windes, and, in particular,Cathy Cameron and is, who their leaders are, what resourcesthey have Wolky Toll for inviting me to participate.They providedmany useful suggestionsfor improvingthe ideaspresented in thispaper, access to, whom they may marry,where they may and they were exceptionallygracious and helpful in bringingan live, what occupations are available to them, and "outsider"up to speed on currentdata and thinkingin Chacoan spiritualpractices they will follow. Kinship influ- archaeology.My deepestgratitude goes outto all of them.I would ences, bounds, and shapes all aspects of life. To also like to thankLynne Goldsteinand the four externalreview- understanda non-statesociety, like Chacoan,in the ers who providedsuggestions and critiquesthat have markedly improvedthis paper. absenceof understandingthe relationsof kinshipis an impossibletask. References Cited Onereason archaeologists may shy awayfrom the Aberle,D. discussion of kinship in the societies we study is 1961 MatrilinealDescent in Cross-CulturalPerspective. In because kinship is not materialin nature.It is not MatrilinealKinship, edited by D. Schneiderand K. Gough, pp. 655-729. Universityof CaliforniaPress, Berkeley. somethingwe can readilymeasure like the volume Akins, N. J. of chertdebris or the numberof ceramics.3As I hope 1997 The Abradersof Chaco Canyon:An Analysis of their the discussionof the pit structure-pueblotransition Formand Function. In Ceramics,Lithics, and Ornamentsof Chaco Canyon, edited by F. J. Mathien, pp. 701-946. suggests, kinshipdoes have materialcorrelates that NationalPark Service, SantaFe. can be identifiedthrough ethnology. Indeed, ethno- Blanton,R. E. logical studies have demonstratedstrong relation- 1998 BeyondCentralization:StepsTowardsaTheoryofEgal- itarianBehavior in ArchaicStates. In ArchaicStates, edited ships between social organizationand the material by G. Feinmanand J. Marcus,pp. 135-172. School of Amer- record(see Emberand Ember 1995 for an overview). ican ResearchPress, SantaFe. But even if such materialcorrelates are unavailable, Blanton,R. E., G. M. Feinman,S. A. Kowalewski,and P. N. Pere- grine we must still ask the questionof whatkinship struc- 1996 A Dual-Processual Theory for the Evolution of tureswere presentin the ancientsocieties we study MesoamericanCivilization. CurrentAnthropology 37:1-14. andhow theymay have affectedthings like the orga- Divale, W. T. 1977 Living Floors and MaritalResidence: A Replication. nization of production.If we consider only those BehaviorScience Research 12:109-115. processes we can see clearly in the archaeological 1984 Matrilocal Residence in Pre-literate Society. UMI record,we become like the proverbialdrunk seek- ResearchPress, Ann Arbor. Driver,H. ing his lost wallet underthe lamppost,even though 1956 An Integrationof Functional,Evolutionary, and Histor- he did not lose it there,because that is the only place ical Theoryby Meansof Correlations.Indiana Publications whereit is light enoughto see. Whatthe drunkreally in Anthropologyand Linguistics,Memoir 12:1-35. Ember,M. needs is a flashlight,and I suggest that archaeolo- 1973 AnArchaeologicallndicatorofMatrilocalversusPatrilo- gists need the flashlightof ethnologyto find kinship cal Residence.American Antiquity 38:177-182. in the archaeologicalrecord. Ember,M., and C. R. Ember 1995 WorldwideCross-Cultural Studies and theirRelevance Productionin the Chacoanworld was organized for Archaeology. Journal of Archaeological Research throughkinship. I have arguedthat Chacoan kinship 3:87-111. includedmatrilocal residence as one of its features. Ember,M., C. R. Ember,and B. Pasternak 1974 On the Developmentof Unilineal Descent. Journal of I might furthersuggest, based on its common asso- AnthropologicalResearch 30:69-94. ciationwith matrilocalresidence, that the Chacoans Feinman,G. M. also practiced matrilinealdescent, and that local 2000 Dual-ProcessualTheory and Social Formationsin the Southwest.In AlternativeLeadership Strategies in the Pre- matrilinealgroups formed the basic structureof Cha- hispanic Southwest,edited by B. Mills, pp. 207-224. Uni- coan society. Ratherthan an overarchingpolitical versity of ArizonaPress, Tuscon. Peregrine] MATRILOCALITY,CORPORATE STRATEGY,AND ORGANIZATION 45

Flannery,K. V., and M. Winter 1984 Social and Economic Implicationsof JewelryItems of 1976 Analyzing Household Activities. In The Early the Chaco Anasazi. In Recent Research on Chaco Prehis- MesoamericanVillage, edited by K. V. Flannery,pp. 34-47. tory,editedby W. J. JudgeandJ. D. Schelberg,pp. 173-186. AcademicPress, New York. Reportsof the ChacoCenter, Number 8. NationalPark Ser- Francis,P. vice, SantaFe. 1988 The Manufactureof Beads from Shell. In Proceedings 1997 Ornaments.In Ceramics, Lithics, and Ornamentsof of the 1986 Shell Bead Conference,edited by C. F. Hayes, Chaco Canyon, edited by F. J. Mathien, pp. 1119-1220. pp. 25-36. Rochester Museum and Science Center, Publicationsin Archaeology 18G, Chaco Canyon Studies. Rochester,NY. NationalPark Service, SantaFe. Haury,E. W. McKenna,P. J. 1931 Minute Beads from Prehistoric Pueblos. American 1986 A Summaryof the Chaco Center'sSmall Site Excava- Anthropologist33:80-87. tions: 1973-1978. In Small Site Architecture of Chaco Helms, M. Canyon, edited by P. J. McKenna and M. L. Truell, pp. 1970 Matrilocality,Social Solidarity,and CultureContact: 5-114. Publicationsin Archaeology 18D, Chaco Canyon Three Case Histories.Southwestern Journal of Anthropol- Studies.National Park Service, SantaFe. ogy 26:197-212. Peregnrne,P. N. Hill, J. N. 1991 Some Political Aspects of CraftSpecialization. World 1970 BrokenK. Pueblo: PrehistoricSocial Organizationin Archaeology23(1): 1-1 1. theSouthwest. University of ArizonaAnthropologicalPapers 1992 MississippianEvolution: A World-SystemPerspective. 18. Universityof ArizonaPress, Tucson. PrehistoryPress, Madison,WI. Jernigan,E. W. 1994 Tradeand Matrilineality: A HypothesisBased onWorld- 1978 Jewelryofthe PrehistoricSouthwest. University of New Systems Theory.Cross-Cultural Research 28:99-110. Mexico Press,Albuquerque. 1996 Ethnology versus EthnographicAnalogy: A Common James,S. R. Confusionin ArchaeologicalInterpretation. Cross-Cultural 1994 RegionalVariation inPPrehistoricPuebloHousehold and Research30:316-329. Social Organization:A QuantitativeApproach. Ph.D. dis- Renfrew,C. sertation,Department of Anthropology,Arizona State Uni- 1974 Beyond SubsistenceEconomy: The Evolutionof Social versity,Tucson. Organizationin PrehistoricEurope. In ReconstructingCom- Judd,Neil M. plex Societies:AnArchaeological Colloquium, edited by C. 1954 The Material Cultureof Pueblo Bonito. Smithsonian B. Moore, pp. 69-95. Bulletin of the AmericanSchool of MiscellaneousCollections, vol. 124. SmithsonianInstitution, OrientalResearch, Supplement 20, Cambridge,MA. Washington,DC. Schneider,D. Judge,W. J. 1961 Introduction:The Distinctive Features of Matrilineal 1979 The Developmentof a ComplexCultural Ecosystem in Descent.In MatrilinealKinship, edited by D. Schneiderand the Chaco Basin, New Mexico. In Proceedingsof the First K. Gough,pp. 1-3 1. Universityof CaliforniaPress, Berke- Conferenceon ScientificResearch in theNational Parks, Vol. ley. 2, editedby R. M Linn,pp. 901-905. Transactionsand Pro- Sebastian,L. ceedings Series 5, NationalPark Service, Washington, DC. 1991 Sociopolitical Complexity and the Chaco System. In 1989 Chaco Canyon-SanJuan basin. In Dynamics of South- Chacoand Hohokam, edited by P.L. Crownand W. J. Judge, westernPrehistory, edited by L. S. Cordelland G. J. Gum- pp. 109-134. School of AmericanResearch, Santa Fe. merman,pp. 209-261. SmithsonianInstitution, Washington, Steward,J. DC. 1955 Theoryof Social Change.University of Illinois,Urbana. Kelley,J. C., and E. A. Kelley Toll, H. W. 1975 AnAlternativeHypothesis fortheExplanationofAnasazi 1984 Trends in CeramicImport and Distributionin Chaco CultureHistory. In CollectedPapers in Honor of Florence Canyon.In RecentResearch on ChacoPrehistory, edited by Hawley Ellis, edited by T. R. Frisbie,pp. 178-223. Papers W. J. Judgeand J. D. Schelberg,pp. 115-135. Reportsof the of the ArchaeologicalSociety of New Mexico, Number2, Chaco Center, Number 8. National Park Service, Albu- Albuquerque. querque. Leach,E. Weigand,P. C. 1954 PoliticalSystemsofHighlandBurma:A Study ofKachin 1992 TheMacroeconomic Role of Turquoisewithin the Chaco SocialStructure. Harvard University Press, Cambridge, MA. CanyonSystem. In AnasaziRegional Organizationand the Lekson, S. Chaco System,edited by D. E. Doyel, pp. 169-173. Anthro- 1986 GreatPuebloArchitectureof Chaco Canyon,New Mex- pological Papers,Number 5. Maxwell Museumof Anthro- ico. Universityof New Mexico Press,Albuquerque. pology, Albuquerque. 1997 Rewriting Southwestern Prehistory. Archaeology Weigand,P. C., and G. Harbottle 50:52-55. 1993 The Role of Turquoisesin the Ancient Mesoamerican Longacre,W. TradeStructure. In TheAmericanSouthwest and Mesoamer- 1970 ArchaeologyasAnthropology:A Case Study.University ica: Systemsof PrehistoricExchange, edited by J. E. Eric- of ArizonaAnthropologicalPapers17. Universityof Arizona son andT. G. Baugh, pp. 159-177. Plenum,New York. Press,Tucson. Wilcox, D. R. Malinowski,B. 1993 The Evolutionof the ChacoanPolity. In The Chimney 1922 Argonauts of the WesternPacific. Waveland Press RockArchaeological Symposium, edited by J. M. Malville Chicago. and G. Matlock, pp. 76-90. Technical Report RM-227. Martin,P. S. USDA ForestService, Fort Collins, CO. 1950 ConjecturesConcerning the Social Organizationof the Wills, W. H. Mogollon Indians.Fieldiana Anthropology 38:556-569. 2000 Political Leadershipand the Constructionof Chacoan Mathien,F J. Great Houses, AD 1020-1140. In AlternativeLeadership 46 AMERICANANTIQUITY [Vol. 66, No. 1, 2001

Strategiesin the PrehispanicSouthwest, edited by B. Mills, group in question (the Chacoans in this case) were not unusu- pp. 19-44. Universityof ArizonaPress, Tuscon. al in their practices, but ratherto first determine if common Windes,T. C. practices are evidenced in the archaeological record before 1992 Blue Notes: The ChacoanTurquoise Industry in the San seeking evidence for more unusual ones. JuanBasin. InAnasaziRegional Organization and theChaco 2. This value is per bead; that is, is calculated by estimat- System,edited by D.E. Doyel, pp. 159-168. Anthropologi- ing the entire time required to travel to a turquoise source, cal Papers,Number 5. Maxwell Museumof Anthropology, Albuquerque. mine the raw material, do rough processing, heat treating, and shape blanks, and then dividing that time by the total Notes number of beads produced from the raw material. 3. The "drubbing"taken by scholars such as Hill (1970) 1. This is not the case in some modern puebloan groups, and Longacre (1970) who attemptedsuch analyses has prob- but it is importantto note that the idea of women agricultur- ably not encouraged others to pursue kinship in the archaeo- al groups comes not from ethnographic analogy, but from logical record of the Southwest. ethnology; that is, in matrilocal societies these kinds of work groups are the norm, as are all the features I am putting forward here. That modern (or ancient) groups may not have followed these practices is understood-there is a range of variation in all human behaviors. However, it seems most Received August 10, 1999; Accepted December 16, 1999; reasonableto begin an analysis under the assumptionthat the RevisedApril 14, 2000