You are on page 1of 21

Specialization, Market Exchange, and the Aztec State: A View From Huexotla [and Comments

and Reply]
Author(s): Elizabeth M. Brumfiel, Kenneth L. Brown, Pedro Carrasco, Robert Chadwick,
Thomas H. Charlton, Tom D. Dillehay, Connie L. Gordon, Roger D. Mason, Dennis E. Lewarch,
Hattula Moholy-Nagy, Jeffrey R. Parsons, David A. Peterson, Hanns J. Prem, Barbara J.
Price, Frances Rothstein and William T. Sanders
Source: Current Anthropology, Vol. 21, No. 4 (Aug., 1980), pp. 459-478
Published by: The University of Chicago Press on behalf of Wenner-Gren Foundation for
Anthropological Research

Stable URL: http://www.jstor.org/stable/2742060 .


Accessed: 25/11/2014 13:56
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
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.

The University of Chicago Press and Wenner-Gren Foundation for Anthropological Research are collaborating
with JSTOR to digitize, preserve and extend access to Current Anthropology.

http://www.jstor.org

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

CURRENT ANTHROPOLOGY

Vol. 21, No. 4, August 1980

FoundationforAnthropological
Research0011-3204/80/2104-0001$02.25
1980by The Wenner-Gren

Specialization,MarketExchange,and the
Aztec State:A View fromHuexotla'
byElizabeth M. Brumfiel

AND EXCHANGE make good economic sense.


Specializationallows forthe exploitationof differences
in the
naturalabilitiesof individualsand in the naturalresourcesof
geographicregions.It permitseconomiesofscale and minimizes
investment
in duplicatingthe toolsof production(Smith 1937
[1776];Marx 1933[1867]:355-59; Samuelson1964:49-50). Exchangeprovidesforthetransfer
ofgoodsand servicesfromthe
specialistswho produce them to consumers.It is therefore
essentialifthe benefitsofspecializationare to be realized.2
At thesame time,however,specializationand exchangemay
produceeconomicinsecurity.
If any segmentofproducersfails
to supplyits customaryoutput,or if the exchangenetworkis
disrupted,thosewho have come to relyupon the systemwill
lack access to the productstheynormallyconsume.If these
productsare basic subsistencegoods,a breakdownin the system will have serious consequences.It might be supposed,
therefore,
that highlydevelopedsystemsof specializationand
exchangewill mostfrequently
occurin associationwithstatelevelpolities-territorially
extensive,administratively
complex
politicalsystemsin whichgovernmental
institutionsmononoSPECIALIZATION

1 Fieldwork
at Huexotla was supported by a Ford Foundation
Grant for Overseas Archaeological Research and by a National
Science Foundation Dissertation ImprovementGrant (GS-38470).
This supportis gratefullyacknowledged.I also wish to thankJeffrey
R. Parsons for providingsuggestionsand encouragementat each
stage of the collectionand analysis of the Huexotla data. Mary
Hodge, Elinor Melville, and Henry Wrightall read an earlierdraft
of thispaper and offeredveryusefulcommentson it.
2 Of course,in any givencase, the economicutilityofspecialization
and exchangemay be surpassed by theirimportancein buttressing
social and politicalrelationships(see Mauss 1954[1925],Levi-Strauss
1969,Sahlins 1972,Wright1977).

M. BRUMFIELis AssistantProfessorof Anthropology


at Albion College (Albion, Mich. 49224, U.S.A.) and Adjunct
AssistantResearch Scientistat the Museum of Anthropologyof
theUniversityofMichigan.Born in 1945,she was educatedat the
Universityof California,Los Angeles (M.A., 1969) and the Universityof Michigan (B.A., 1965; Ph.D., 1976). She taught at
Eastern Michigan University1971-77. Her researchinterestsare
the political economyof agrarian states, cultural ecology,and
Aztec ethnohistory.
She is now at workon a studyof the political
implicationsof marriageamong the Aztec elite 1350-1521. Her
publicationsinclude"Regional Growthin the Eastern Valley of
Mexico: A Test of the Population PressureHypothesis,"in The
Early MesoamericanVillage,edited by Kent V. Flannery (New
York: Academic Press, 1976), and "Ethnohistoryof the ChalcoAmecameca Region," in Prehispanic Settlement
Patternsin the
SouthernValleyofMexico,by J. Parsons,E. Brumfiel,
M. Parsons,
and D. Wilson (UniversityofMichiganMuseum ofAnthropology
Memoir 13, in press).
The presentpaper was submittedin finalform22 viii 79.
ELIZABETH

comadministrative
lize the use of legal force.Givensufficient
plexity,theoutputsofdifferent
segmentsoftheproductivesystem can be monitored,and, should these outputsfall below
theirrequisitelevels,correctiveaction can be initiated(Flancontrolofpolice and military
nery1972). Givengovernmental
institutions,
theoutbreakofwarfarewithintheexchangeregion
can be forestalledand disruptionof the exchange system
systemof
avoided.Conversely,themaintenanceofan effective
specializationand exchangewill confersubstantialeconomic
benefitsupon citizensof the state, winningtheirloyaltyand
thuscontributing
to its persistence(Service1975,1978).
The purposeof thispaper is to investigatethe interrelationships of specialization,marketingactivity,and state-level
polities.It drawsinspirationfromthe workof Sanders (1956,
1965, 1968; Sandersand Price 1968), who has suggestedthat
specializationand exchange were importantfactorsin the
emergenceof pre-Hispanicstates in highlandCentralMexico.
Sanders has argued that the environmentaldiversityand
of raw materialsin highlandMexico enlocalizeddistribution
specializationand
couragedthe developmentof intraregional
systems.In his view,specializationfavored
regionalmarketing
of ecopoliticalcentralizationbecause "as the diversification
nomicinterestbecomesgreater,the greateris the need forthe
centralizedregulationof thoseinterests.Specialistsin control
of menand of resources-politicalspecialists-are increasingly
required"(Sandersand Price 1968:232-33). The formationof
regionalmarketingsystemsfavoredthe emergenceof territoriallyextensivepolitical units because marketingsystems
"are difficult
to maintainexceptwhenthe communities[participatingin the system]are part of some largersociopolitical
structuresuch as a regionalstate" (Sanders 1965:6).
To evaluate the hypothesisthat Mexican states arose and
thenexpandedto facilitatespecializationand marketexchange,
a programofarchaeologicalresearchwas carriedout at Huexotla, a site in the easternValley of Mexico datingto the Aztec
period(ca. A.D. 1150-1519).For severalreasons,data fromthe
Aztecperiodoughtto be appropriateto thistask.
thatenvironmental
diversity
First,historicalsourcesconfirm
withinthe Valley of Mexico was associated witha degreeof
economicspecializationat and beforethe date of Spanishconquest. Duran (1964[1581]:32),for example,recordsthat the
firstinhabitantsofTenochtitlanexchangedtheproductsof the
lake that surroundedthem for constructionmaterialswhich
werenot locallyavailable. Cortes(1970[1519-26]:51)observed
a briskcommercein salt centeredaround threecommunities
just south of Tenochtitlan,and Diaz del Castillo (1908-16
[1632]:34) noted that lime kilnswereheavilyconcentratedin
the northern
end of the valley.Later 16th-century
documents

Vol. 21 * No. 4 * August1980

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

(especiallythe RelacionesGeogrdficas;
see Cline 1972) indicate
that manycommunities
in the Valley of Mexico weresupplementingsubsistenceagricultureby specializingin one or more
activitiesthatyieldedproductsformarketsale. Many of these
activitiescould have persistedfrompre-Hispanictimes.
Second, the existenceof a regionalexchangesystemwithin
theValleyofMexicoduringtheAztecperiodcannotreasonably
be questioned.The largestof the Aztec settlements,
Tenochwas servedby a daily marketcontainingan astonishtitllan,
ingvarietyofgoodsand attendedby thousandsofpeople (Narrative1917[1556]:65-67;Cortes1970[1519-26]:62-64;Diaz del
Castillo 1956[1632]:215-17;also see Calnek 1975 and Parsons
1976). Markets were also held periodicallyin many of the
valley's smaller cities and towns ("Carta" 1870[1563]:296;
Chimalpahin1965[ca.1606-31]:206;Duran 1971[1570]:213-19;
Motolinia1950[ca. 1536-43]:59). By the date of Spanish conquest, marketexchangehad existedin the Valley of Mexico
forat least two centuries(Duran 1964[1581]:32;Tezozomoc
1975[1598]:231).3
duringtheAztec
Third,thepatternofpoliticaldevelopment
moreextensive
periodwas markedby thegrowthofterritorially
units.DuringEarlyAztectimes(ca. A.D. 1150-1350),thevalley
was dividedinto small,localized,politicallyautonomouscitystates (Gibson 1964:32-36; Sanders and Price 1968:157-59;
Soustelle 1961 xxii). After 1350, these city-stateswere incorporatedinto a large regionalempire(Barlow 1949, Gibson
1971). Althoughthe empirewas stillonlylooselyintegratedat
the time of Spanish conquest,there had already been considerable movementtoward greaterpolitical centralization
(Cortes 1970[1519-26]:65;Ixtlilxochitl1952[ca. 1600-40]:17678; Motolinia1950[ca.1536-43]:210, 215), greaterbureaucratic
complexity(Chimalpahin1965[ca. 1606-31]:194,219, 228-29;
Tezozomoc1975[1598]:268-69),and morerigidsocialstratification (Duran 1964[1581]:131-32,223-24).
oftheAztececonomicsystem
Whilehistoricaldocumentation
is extensive(see Berdan 1975), it is not detailed enough to
elucidateseveralimportantquestionsconcerning
specialization,
exchange,and Aztec politicalhistory.For example,how complex werethe economiesof the small Early Aztec city-states?
Sandersand Price (1968:152) have suggestedthat theywere
marked by considerablediversity.Differentkinds of craft
specialists,residingin the central towns, exchanged their
productsboth with other craftspecialistsand with agriculeach town.The
turalistsresidingin theruralareas surrounding
were also specialiststo a degree; theircrops
agriculturalists
varied withrainfall,drainage,and soil depth,and they were
able to combineagricultural
productionwithone or moreother
reextractiveindustries,dependingupon the nonagricultural
sourcesavailable to them.If thissuggestionis correct(and it is
highlyplausible),one mightargue (as Sanders[1968]appears
to) that the existenceof smallcity-statesin Early Aztec times
was foundedupon the need to stabilizespecializationand exis, howchangeat thelocal level.The historicaldocumentation
to resolvethe issue.
ever,insufficient
of CentralMexico duringLate
Was the politicalunification
Aztec times,with the ensuinggrowthin centralizationand
bureaucraticcomplexity,accompaniedby the intensification
of specializationand exchangeon a regionalscale? The functionalrelationships
exchange,
postulatedbetweenspecialization,
and politicalcomplexity
suggestthatthisshouldhave occurred,
but, again, verificationof this suggestionthroughhistorical
researchhas not beenpossible.
oftheAztecmarketsystem?
What was theprimaryfunction
Specializationand exchangecan operateas adaptive mechaof a pronisms,providingmeans forenhancingthe efficiency
ductivesystemat theregionallevel.Thus theprimaryfunction
of the Aztecmarketsystemmighthave been to distributethe
3 The existenceof a large,well-integrated
regionalmarketingsystem in the Valley of Mexico duringAztec timesis also indicatedby
the settlementpattern(Smith 1979).
460

theregion.
productsoflocal specialiststo consumersthroughout
Thereare,however,alternativepossibilities.Many marketsystems are orientedtoward the extractionof produce froma
regionalhinterlandeitherforexportor forurbanconsumption
(see Smith 1976 for numerousexamples). These alternatives
is notplentifulin
but relevantinformation
meritinvestigation,
publishedhistoricalaccountsof the Aztec economy.
Finally, to what extentand in what ways was the Aztec
marketsystemaffectedby othermodes of exchangesuch as
tributecollectionand long-distancetrade? The documentary
evidencecompiledby Berdan (1975, 1977) suggeststhat these
but it does not permit
threesphereswerecloselyinterrelated,
upon the
one to gauge the impact of these interrelationships
structureand intensityof marketexchange.
The archaeologicaldata collectedfromHuexotlaprovidethe
on thesematters:
information
following
1. The local economyofat least thissmallEarly Aztec citystate was not verycomplex.Verylittleevidenceof specialization on the local level was obtained at Huexotla. Thus, dependenceupon specializationand marketexchangeat thelocal
level does not appear to account for the pattern of small,
localizedcity-states
typicalofEarly Aztecpoliticalstructure.
2. The political unificationof CentralMexico duringLate
of regional
Aztec timeswas accompaniedby an intensification
exchange.Greaterquantitiesof salt, spindlewhorls,obsidian,
and probablyclothfromnonlocalsourceswereprocuredby the
inhabitantsofHuexotladuringLate Aztectimes.Concurrently,
the productionof maize and magueysyrupby-productswas
thatimportedproductswerebeing"paid
intensified,
suggesting
for"withfoodstuffs.
3. The Early Aztec marketsystemappears to have been
the productsof local specialists
orientedtowarddistributing
to a regionalpopulation of consumers.During Late Aztec
activityshiftedtowardproviding
times,thefocusofmarketing
food for burgeoningurban populations.The intensityof exchangebetweenHuexotla and the regionalcapitals of Tenochtitlanand Texcoco increasedwhile the intensityof exchange
local centersremained
betweenHuexotlaand otherlower-level
unchanged.This suggeststhat the Late Aztec intensification
as resulting
of regionalexchangeis probablybest interpreted
fromurbanizationratherthanfroma generalizedmovementof
the population toward a betteradaptive adjustmentto its
environment.
diversified
4. Marketexchangeand tributeextractionwereveryclosely
linkedduringLate Aztec times.Much of the obsidianand the
cloth procuredby Huexotla's inhabitantsduringLate Aztec
timeshad probablybeen producedas itemsof tributerather
than as marketcommodities.Introducedinto the marketsystem in the marketplacesof regionalcapitals, these goods enby the
couragedspecializationin the productionof foodstuffs
ratherthancraftspecializaruralpopulace.Tributeextraction,
tion, was the methodused by the urban populationof the
Valleyof Mexico to pay forthe foodit consumed.
The data uponwhichtheseconclusionsare basedare reported
a generalrebelow.The concludingsectionof thispaper offers
of specialization,market
examinationof the interrelationships
exchange,and the existenceof the Aztec state.
THE VALLEY OF MEXICO AND THE HUEXOTLA
CITY-STATE
The Valley of Mexico is a large internaldrainagebasin lying
morethan 2,240m above sea level at the southernedge of the
Central Mexican plateau (fig. 1). Measuring 180km northsouthand 60 km east-west,the valleyencompassesan area of
roughly8,000km2.In pre-Hispanictimes,muchof the valley
floorwas coveredby a chainofveryshallowlakes. The central
segmentof the lake systemwas quite salty,the northernand
southernsegmentsmuchless so. By the date of Spanish conCURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

quest,the entiresurfaceof the southernlake systemhad been


of ridged
broughtundercultivationthroughthe construction
fields(chinampas)(Palerm 1973,Parsons 1976).
Because of the highelevation,frostsare commonduringthe
wintermonthsand can occur as early as Septemberand as
of the
late as May. Precipitationis seasonal,about two-thirds
The seasonsof
rainfallingbetweenlate May and mid-October.
days and heavyprecipitationthus coincideto define
frost-free
growingseason. Maize, beans,amaranth,
a four-or five-month
squash,and two speciesof cactus(magueyand nopal) werethe
dominantagriculturalcropsduringpre-Hispanictimes.
The city-stateofHuexotlalay on theeasternside of theValley of Mexico. Its boundariesencompasseda stripof land approximately12 km east-west and 5 km north-south.Its
westernlimitswereprobablyset by the shoreof Lake Texcoco
and its easternlimitsby the peaks of the SierraNevada, the
mountainchain thatformsthe easternboundaryof the Valley
ofMexico.To thesouth,Huexotla'slandslay adjacent to those
of Coatlich'an,a city-stateofmuchthesame size and antiquity
as Huexotla. Fromthe 14thcenturyonward,Huexotla'sneighbor to thenorthwas the city-stateofTexcoco. As a memberof
theAztecTripleAlliancefrom1430untiltheSpanishconquest,
Texcoco came to dominatemuchof the easternvalley,including Huexotla.
controlledby Huexotla was not exAlthoughthe territory
diverse,providingthepotential
tensive,it was environmentally
forinternaleconomicspecializationand exchange(fig.2). Because thelandsadjoiningLake Texcocoweresubjectto periodic
flooding,theywereimpregnatedby salt and generallyuseless
foragricultural
purposes.However,thelake did yielda variety
of aquatic foods such as insect larvae and eggs, algae, fish,
It also providedreeds
waterfowl.
frogs,and migratory
crayfish,
formatting,and a usable salt could be extractedfromthe soil
washedby its waters.The lightlyerodedlowerpiedmontslopes
were ideally suited to maize cultivation,generallyreceiving
more rainfalland being less subject to frostthan the valley
floor.The lowerpiedmontsoils are comparativelydeep, with
capacities. The soils of the more heavily
good water-storage

Mexico

Brumfiel:SPECIALIZATION,

EXCHANGE, AND THE STATE

eroded upper piedmontslopes are shallow and not ideally


suited to maize production.However, native crops more
tolerantofshallowsoilswithpoorwater-storage
capacitieswere
available for cultivation,among them beans, maguey,and
nopal. The huntingof deer and small game animals and the
fromtheforestsoftheSierra
extractionoflumberand firewood
Nevada could also have contributedto the subsistenceof the
occupantsofHuexotla'supperpiedmontsector.
A large proportionof Huexotla's populationresidedwithin
a small urban settlementthat coveredan area of about 300
hectaresand was locatedat thejunctureof thelowerpiedmont
slopes withthe flatlakeshoreplain. Most of Huexotla's civicceremonialarchitecture,
includinga walledprecinctthatprobably containedthe residenceof its ruler,lay withinthisarea.
The existenceof this small urban centerwould have added
anotherdimensionof variabilityto the culturalgeographyof
the city-state.If it containeda marketplace,specializationin
the processingof nonlocalraw materials(such as obsidianand
rawcotton)mighthave supporteda numberofurbanresidents.
Craftproductionwithintheurbancentermightalso have been
subsidizedby the patronageof residenturban elites ("Carta"
1870[1563]:296;Tezozomoc 1975[1598]:668-69).
Data pertainingto thestructure
ofHuexotla'seconomywere
obtained througha programof intensive,systematicsurface
collectioncarriedout at the site duringthe summersof 1972
and 1973. The debrisof Aztec occupationwas collectedfrom
1,243 5-m X 5-m squares scatteredover the territory
encompassed by the city-state.Almosthalf the units (560 squares)
were situatedwithina square-kilometer
area in the heart of
Huexotla's urban center(providinga roughly1% sample of
the debrisin that area). The restof the unitsweresituatedin
thelakeshoreand the lowerand upperpiedmontsectorsofthe
site. In theseareas, each localizedconcentration
ofprehistoric
debris(representing
a prehistoric
ruralhousehold)was sampled
at a 6% level or better(see Brumfiel1976:50-56).
The materialscollectedfromeach unitincludedall fragments

LAKE ZUMPANGO

AX

Teotihuacan

Otumba

The Valley of Mexico


Ecatepec

Texcoco
LAKE TEXCOCO

*Huexotlo
H
uexotia

Azcopotzalco

* Coatlichan

Tlocopan
8

_
10 km

~~~~~~~~~Tenochtitlan

\LAKE XOCHIMILCO
\

>

t~~~~~~~~~~~~~~~~~Chalco

Xochimico

lake systemand the locationof Huexotla. Late Aztec regionalcapitals are starred.
FIG. 1. The Valley of Mexico, showingthe prehistoric

Vol. 21 * No. 4 * August1980

461

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

of decoratedceramicvessels,vesselrims,stonetoolsand chipceramicmolds,and shell. The


page, spindlewhorls,figurines,
in thecollections
functions
ofmanykindsofartifactsoccurring
eitheron the basis of previousarchaeological
can be inferred,
analogy
researchin the Valley of Mexico or by ethnographic
(table 1). The variationin the relativefrequenciesof these
artifactsfromone part of thesite to anotherprovidesinformation concerningthe divisionof labor at the local level. Other
about Huexotla's place in the
materialsprovideinformation
broadersystemof regionalexchange(table 2).
either
Most collectionunitscan be regardedas representing
Early Aztec or Late Aztec occupation,dependingupon the
decoratedceramicmaterialsthat they contain (see Brumfiel
1976:57-93) 4 A comparisonof Early Aztec with Late Aztec
4 Dating the Huexotla collections involved a two-step process.
scaling was used to assess the temporalsigFirst,multidimensional
nificanceof 35 decorated variants of Late Postclassic pottery.The
resultsol thisanalysismade it possibleto regardeach of the variants
ofeitheran earlieror a laterphase ofLate
as generallyrepresentative
Postclassic occupationat Huexotla. Next, each collectionwas cateeitheran earlieror a laterphase ofoccupation,
gorizedas representing
dependingupon whetherearlyor late decoratedvariantswerebetter
representedin it.
The 19 decoratedceramicvariantsjudged representativeof Early
Aztec occupationincludedfivevariantsof B]ack-on-Orangevessels,
all of which would fall within the category"Tenayuca Black-on-

collectionsmakes it possible to detect temporalchanges in


Huexotla'seconomicstructure.
EarlyAztecoccupationis representedby someunitsfromtheurbancenterand by unitsfrom
thewesternmost
fringeofthelowerpiedmont.Late Aztecoccuand Espejo (1950:17-28); Plain Red and BlackOrange" of Griffin
on-Red copas; Black-on-Redbowls and incised bowls as described
by Parsons (1971:298-99); Black-on-Red bowls bearing interior
decoration(illustratedby Noguera 1935:pl. 15, 6 and Tolstoy 1958:
fig. 10a); seven variants of Black-and-White-on-Redbowls (see
Parsons 1971:309-12); Polychromeplates and dishescoveredwith a
thickwhitepaint bearingcomplexdesignsin red,orange,and black;
censervariant (illusand a thick-walled,
roughlyfinishedRed-on-Buff
tratedby Noguera 1935:pl. 17, 1 and 2).
The 16 variants judged representativeof Late Aztec occupation
includedsix variantsofBlack-on-Orangevessels,fiveofwhichwould
fall within the category"Tenochtitlan Black-on-Orange"and the
sixth of which correspondsto "Tlatelolco Black-on-Orange,"both
definedby Griffin
and Espejo (1950:29-52); two variants of Blackon-Orangeslab supports associated with "Tenochtitlan Black-onOrange" molcajetesand dishes; two variants of Plain Red bowls;
Black-on-Redbowls bearingclustersof black verticalparallel lines
on theirexteriorsurfaces(see Parsons 1971:309); Polychromebowls
covered with a deep red paint (sometimesoverlain by a coat of
streaky orange or black paint) bearing complex designs in black,
orange,and/orred; Fabric-markedpottery(see Tolstoy 1958:51-54;
Parsons 1971:312-13); Texcoco Molded and Texcoco Filletedvessels
(see Tolstoy 1958:49-50); and unpainted, poorly finishedCrossHatched vessels (illustratedby Noguera 1935:pl. 34, 4).

Urb~~~~6r

-N-

2. Aztec settlementat Huexotla. Dotted linesencloseareas ofdispersedruralsettlement;hatchedline indicatesthe approximateshoreof


Lake Texcoco.

FIG.

TABLE 1
CATEGORIES

OF TOOLS

AND WASTE MATERIALS ANALYZED IN THE HUEXOTLA


THEIR INFERRED INDUSTRIAL ASSOCIATIONS

ARTIFACT

ASSOCIATED

STUDY AND

INDUSTRY

Cultivationof annual domesticatessuch as maize,


beans, and squash"
Extractionof magueyfibers
Felsite blades............................
Extractionof magueysyrup
Heavy scrapers..........................
........ Extractionof magueysyrup
Thick-walledvessels (jars) ........
Deer huntingwithbow and arrow
Projectilepoints.........................
Fired clay balls..........................
Small-gamehuntingwithblowgun
Salt extraction
Fabric-markedpottery
....................
Spinningof magueyfibers
Large spindlewhorls.....................
Small spindlewhorls......................
Spinningof cottonfibers
. Productionof prismaticblades
Obsidian coresand waste flakes.........
Productionof figurines,
Ceramic molds...........................
spindlewhorls,urnadornments,
and ceramicstamps
Productionof shell ornaments
Shell fragments
..........................
Unspecifiedcraftproduction
Drills-reamers
...........................
Unifaciallyretouchedprismaticblades......

SOURCE: Brumfiel(1976:102-30, 136-50).


a Tentative
interpretation.

462

CURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

pationis represented
by otherunitsfromtheurbancenterand
by virtuallyall the unitsin the moreeasterlylowerand upper
piedmontsectors.All unitsfromthelakeshoresectorcontained
an even mixof Early and Late Aztec materials.
My analysisbeginswithan examinationof theinternalcomplexityofproductionat HuexotladuringEarly and Late Aztec
times.
INTERNAL SPECIALIZATION

AND EXCHANGE

The materialsfromHuexotla indicatethat internalspecialization and, presumably,marketexchangedid exist withinthe


city-statebut were not well developed. Evidence for the
existenceof an internaldivisionof labor was foundin three
patternsof artifactdistribution.
First,artifactsassociatedwithagriculturalproduction(felsite and prismaticblades, scrapers,thick-walledvessels) were
less commonin thelakeshoresectorthanelsewhere
consistently
(tables 3 and 4). Evidently,the salt-impregnated
soils of
Huexotla's lakeshorewerean impedimentto agriculture,and
its inhabitantsmusthave supportedthemselvesthroughsuch
activitiesas fishing,
and manufacturing
fowling,
reedmatsand
baskets.However,thelakeshorewas neverinhabitedby a large
numberof people; Parsons's (1971:116, 139) estimatesforthe
Huexotlaregionsuggestthatless than5% ofthepopulationof
the city-stateresidedthere.
Second, therewere some indicationsof a differentiation
of
labor betweenthe inhabitantsof the urbansectorand the inhabitantsof the piedmontzone. In the collectionsof Early
Aztec materials,thisdifference
was seen in the morefrequent
occurrence
ofunifacially
retouchedbladesin thelowerpiedmont
sector; evidentlymaize productionwas somewhatmore intensivetherethan withinthe urbanzone. In the collectionsof
Late Aztec materials,the difference
was more pronounced;
artifactsrelatedto maizeproduction,theprocessingofmaguey
and deerhuntingwereall significantly
by-products,
morecommonin the lowerand upperpiedmontthanin the urbanzone.
These data suggestthat the inhabitantsof the piedmontwere
moreintensivelyengagedin food-producing
activities;a portionofthisproducewas probablydestinedto supportsomesegmentsof Huexotla's urbanpopulace.
Does thisreflectan increasingly
importantdivisionof labor
betweenruralfood-producers
and urbancraftspecialists?Probably not, fortherewerefewindicationsthat craftproduction
was especiallyintensivewithinthe urban sector.Large and
small spindlewhorls,used in the productionof magueyfiber
and cottoncloth,werenot significantly
morecommonin the
urban zone than in the rural sectors. The distributionof
prismaticcores closelyparalleled the distributionof utilized

Brumfiel:SPECIALIZATION,

EXCHANGE, AND THE STATE

prismaticblades across all sectors,indicatingthat the inhabitantsof each sectorproducedthe blades used withinit rather
specialistsresidingin
than dependingupon obsidian-working
of ceramicmolds sugthe urban zone. Only the distribution
gested more intensivecraftproductionby Huexotla's urban
populace. Of the 13 moldscollectedin the Huexotla area, 12
came fromtheurbanzone. Sincetheywerescatteredovermost
of thatzone, however,theyprovidedno evidenceof especially
intensivecraftproductionby particularurban householdsor
wards.
Apparently,the economicrelationsbetweenthe urban and
More intenpiedmontsectorsof Huexotla wereasymmetrical.
sive foodproductionin thepiedmontsectorswas not countered
by moreintensivecraftproductionin the urban sector.Relationsbetweenthesetwoareas wereprobablynotbased uponthe
goods withina marketsetting.
exchangeof complementary
Perhaps the recipientsof the food surpluses produced in
Huexotla'spiedmontwereurbanelitesratherthanurbancraft
these surpluses
specialistsand the mechanismof transferring
was tributeextractionratherthanmarketexchange.
between
Finally,thereseem to have been some differences
upperand lowerpiedmontsectorsin the natureofproduction.
were
Artifacts
relatedto theprocessingofmagueyby-products
more commonin the upper piedmont,while unifaciallyretouchedblades (relatedto maize cultivation)weremorecommonin thelower.As I have said, magueyis moretolerantthan
soil conditions.Evidentlythis was
maize of upper-piedmont
recognizedand exploitedto a degree,moremaize beingcultivated in the lowerpiedmontsectorand moremagueyin the
upper.However,judgingby the ceramicmaterials,the upper
piedmontsectorwas occupied only duringLate Aztec times.
Therefore,the incipientdivisionof labor betweenmaize producersand magueyproducerswas a Late Aztec phenomenon.
offood
It was probablyassociatedmorewiththeintensification
inproductiorn
forexportthan with an increasein efficiency
tended to meet local needs. This point will be discussedat
greaterlengthlater.
In summary,the data fromHuexotla indicatethat the internaleconomyof this Aztec city-statewas notcharacterized
by a complexdivisionof labor dependentupon a stable local
marketsystem.When specializationwas trulyqualitative in
nature(as it evidentlywas in the lakeshoresector),it engaged
of thepopulation.Wheregreaternumonlya smallproportion
activities
bersofpeoplewereinvolved(as in thefood-producing
of the inhabitantsof the piedmont),productionwas probably
linkedto mechanismsof exchangebeyondthose of the local
marketsystem.
If specializationand local marketexchangewere not im-

TABLE 2
NONLOCAL PRODUCTS COLLECTED AT HUEXOTLA AND THEIR POINTS OF ORIGIN
PRODUCT

Fabric-markedpottery....................
Large black or buffspindlewhorlswith
inciseddecoration......................
Large red-and-blackspindlewhorlswith
incisedor moldeddecoration............

POINT OF ORIGIN

Salt-makinglocales east of Texcoco or south ofTenochtitlan(see Brumfiel1976:182-83)


Tenochtitlan,Zumpango, and/or Teotihuacan regions,
Valley of Mexico (see Parsons 1975)

Chalco-Amecamecaregion,Valley of Mexico (see Parsons 1975)


Greenobsidian...........
... Cempoala-Pachuca region,Hidalgo (see Spence and Parsons 1972)
TA-79 and othersites in the Otumba region,Valley of
Greyobsidian...........
Mexico (see Spence and Parsons 1972)
Cotton (indicatedby the presenceof small
spindlewhorls)....
Various warmerlowland regionsof CentralMexico
.......
Shell...........
Gulf and/orPacificcoasts

Vol. 21 * No. 4 * August1980

463

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

sherds,associatedwiththe importation
of salt fromValley of
Mexico sources,weremuchmoreabundantin Late Aztec collections.5Importedlarge spindlewhorls(Type I; see Parsons
1975:213) came to predominateover locally produced ones
(Types IIA, B, and D; see Parsons 1975:215). Whileobsidian
declinedin relativeabundance in the Late Aztec collections,
the absolutequantityof obsidianimportedinto the city-state
increased.(Huexotla's populationgrewan estimated50% in

portantaspects of the internaleconomy,why did markets


exist?Probablytheyfacilitatedexchangeon the regionallevel.
The collectionsfromHuexotla indicatethat manyproductsof
ofthesitethroughnonlocaloriginwereprocuredbyinhabitants
out theAztecperiod.These productsincludedobsidian,cotton,
salt, spindlewhorls,lime,and smallamountsof shelland jade.
The Huexotla data clearlyindicatethat participationin the
over time.
regionalexchangesystemintensified
INTENSIFICATION

I Because ofthedifficulty
forfabric-marked
withthefigures
pottery
indicated in the note to table 5, increased importationof salt at
Huexotla duringLate Aztec timesis morelegitimatelyinferredfrom
the multidimensionalscaling analysis, which clearly indicates the
association of fabric-markedpottery with the other Late Aztec
ceramicvariants.

OF REGIONAL EXCHANGE

participaThereare severalindicationsofHuexotla'sincreasing
tion in a regionalexchangesystem(table 5). Fabric-marked

TABLE 3
DISTRIBUTION

OF ARTIFACTS BY SECTOR, EARLY AZTEC TIMES

Felsite blades..........
..........
Heavy scrapers..................
Thick-walledvessels..............
Unifaciallyretouchedprismatic
blades........................
Prismaticcores..................
..........
Projectilepoints.......
Fired clay balls.........
........
Large spindlewhorls.....
........
Small spindlewhorls..............

LOWER
PIEDMONT

URBAN

LAKESHORE

1 (0.17)
2 (0.34)
207 (35.0)

11 (2.7)
3 (0.73)
186 (45.0)

5 (3.0)
1 (0.6)
75 (45.0)

108 (18.0)
1 (0.44)
3 (0.5)
2 (0.34)
3 (0.5)
5 (0.84)

155(38.0)
9 (1.2)
6 (1 .5)
6 (1.5)
7 (1 .7)
6 (1.5)

111 (66.0)
6 (1.4)
2 (1 .2)
0
4 (2.4)
4 (2.4)

ofonecase,frequencies
Figuresin parentheses
are,withtheexception
per100comalrims.
function
common
toall Aztechouseholds
Sincecomals(ceramic
griddles)
serveda well-defined
and
I haveusedthelatteras a measure
intheHuexotlacollections,
sincetheirrimsareverycommon
of artifactabundance.For theprismatic
obsidiancores,thestandardmeasureused is per 100
bladefragments.
prismatic
NOTE:

TABLE 4
DISTRIBUTION

OF ARTIFACTS BY SECTOR, LATE AZTEC TIMES

URBAN

LOWER
PIEDMONT

UPPER
PIEDMONT

13 (1.1)
9 (0.76)
471 (40.0)
317 (27.0)
24 (1.0)
3 (0.25)
14 (1.2)
6 (0.51)
12 (1.0)

89 (3.1)
84 (3.0)
1,461(52.0)
2,835 (101.0)
77 (1 .0)
32 (1.1)
16(0.57)
30 (1. 1)
17 (0.6)

54 (4.8)
42 (3.8)
975 (87.0)
838 (75.0)
14 (0.7)
14 (1.2)
2 (0.18)
5 (0.45)
3 (0.27)

LAKESHORE

Felsite blades.......................
Heavy scrapers.........
............
Thick-walledvessels.................
Unifaciallyretouchedprismaticblades.
Prismaticcores.........
............
Projectilepoints........
............
Fired clay balls .........
............
..........
Large spindlewhorls......
Small spindlewhorls.................

1 (0.17)

2 (0.34)

207 (35.0)
108 (18.0)
1 (0.44)
3 (0.5)
2 (0.34)
3 (0.5)
5(0.84)

ofonecase,frequencies
Figuresinparentheses
are,withtheexception
per100comalrims(seenoteto table3); for
prismatic
cores,thefrequencies
areper100prismatic
bladefragments.
NOTE:

TABLE
DISTRIBUTION

OF LOCAL AND NONLOCAL

GOODS BY TIME

EARLY AZTEC

Fabric-markedsherds.30
Large spindlewhorls
Type I (Nonlocal).3
Type IIA, B, D (Local).
Obsidian
Grey (fromOtumba).
Green (fromCempoala-Pachuca)....

PERIOD
LATE AZTEC

(5.2)

792 (14.0)

(0.52)

21 (0.37)
19 (0.33)

8 (1.4)

561 g (97.0 g)
2,419 g (417.0 g)

3,882 g (66. 0 g)
21,385 g (374.0 g)

arefrequencies
Figuresin parentheses
per100 comalrims.
Sincefabric-marked
was usedto distinguish
Late AztecfromEarlyAzteccollections,
pottery
thecountsand frequencies
recorded
heremayunderstate
theirabundancefortheearlierperiod
and overstate
it forthelater.

NOTE:
a

464

CURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

Brumfiel:
SPECIALIZATION,

TABLE 6
DISTRIBUTION

OF ARTIFACTS BY TIME
EARLY

Agriculturalimplements
Felsite blades......
............
Heavy scrapers.................
Thick-walledvessels.......
......
Unifaciallyretouchedprismatic
blades .......................
Huntingimplements
Projectilepoints................
Fired clay balls .................
Spinningimplements
Large spindlewhorls......
......
Small spindlewhorls.............
NOTE:

PERIOD

AZTEC

LATE
AZTEC

16 (2.8)
4 (0.69)
261 (45.0)

157 (2.7)
137 (2.4)
3,114 (54.0)

266 (46.0)

4,098 (73.0)

8 (1.4)
6 (1 .0)

52 (0.91)
34 (0.59)

11 (1 .9)
10 (1.7)

44 (0.77)
37 (0.65)

Figuresin parentheses
arefrequencies
per100comalrims.

the transitionfromEarly to Late Aztec times;the amountof


obsidianalso increased,by an estimated27.5%, but did not
keep pace withthe rate ofpopulationgrowth.)
The collectionsalso indicatecertainchangesin theintensity
ofproductionat Huexotla (table 6). Heavy scrapersand thickwalledvessels,bothrelatedto the collectionof magueysyrup,
weresignificantly
morecommonin the Late Aztec collections.
Also morecommonwereunifaciallyretouchedblades, related
to maize production.This suggeststhat the greaterquantities
of importswerebeing"paid for"by a moreintensiveproducat the local level. On the otherhand, both
tion of foodstuffs
less common
large and small spindlewhorlsweresignificantly
in the Late Aztec collections,indicatinga declinein the local
manufacture
of magueyfiberand cottoncloth.Less intensive
of regional
clothproductionmay itselfreflectan intensification
exchange;it is possiblethat moreclothwas enteringthe citya decline
statethroughtheregionalexchangesystem,permitting
in local production.
oftheValleyofMexico during
Thus, thepoliticalunification
Late Aztec timesseems to coincidewithan intensification
of
regionalexchange.Certainfeaturesof the Huexotla data suggest,however,that not all linkagesin the regionalexchange
networkwereaffectedto the same degree-that exchangesbetweenHuexotla and the two dominantcentersof the Aztec
while
Triple Alliance,Tenochtitlanand Texcoco, intensified
local
exchangesbetweenHuexotlaand similarsemiautonomous
historical
politicalcentersdid not. For example,16th-century
sourcesindicatethatresidentsof the easternValleyof Mexico
eitherpurchasedsalt in the major marketof Tenochtitlanor
reliedupon producersin the immediateTexcoco area (Cortes
1970[1519-26]:51;Papeles 1906-48:62, 67, 86; Pomar 1941
[1582]:62; Parsons 1971:115-16). Similarly,the nonlocalspindle whorlsmorecommonlyfoundin the Late Aztec collections
in the TenochtifromHuexotla wereprobablymanufactured
area (Parsons 1975:213).
tlan-Tlatelolco-Tlacop'an
Two typesof obsidianoccur in the Huexotla collections:a
greyobsidianobtainedfromoutcropseast of Otumba in the
northeasterncornerof the Valley of Mexico and a higherqualitygreenobsidianfrommoredistantquarriesin the Cempoala-Pachuca area northof the valley (Spence and Parsons
1972).OnlytheCempoala-Pachucatypewas obtainedingreater
quantitiesin the Late Aztec period. Greaterreliance upon
Cempoala-Pachucaobsidian cannot be attributedto technologicalfactors;the Huexotladata revealno suddenchangesin
the utilizationof obsidian which would have increasedthe
demandforthe higher-quality
type.A morelikelyexplanation
is that the Late Aztec inhabitantsof Huexotla made greater
use of Cempoala-Pachucaobsidian because it could be procured in the nearby centersof Tenochtitlanand Texcoco.
There is documentaryevidence that both Tenochtitlain
and
Texcoco obtainedobsidianfromCempoala-Pachucaas an item

EXCHANGE, AND THE STATE

of tribute(Relacion 1949[1580]:30,35). I would suggestthat


a part of the obsidian tributeextractedby the Aztec Triple
Alliance was made available to the residentsof other communitiesin the Valleyof Mexico.
Circulationof tributereceiptscould also have suppliedthe
residentsof HuexotlawithclothduringLate Aztec times.The
importanceofclothas an itemoftributeto theTripleAlliance
is well documented(Barlow 1949). It mightbe suggestedthat
as these tributereceiptsgrew,Tenochtitlanand Texcoco beforthe inhabitantsof Huecame centersof clothprocurement
xotla.As suppliesof clothat thesetwo centersexpanded,cloth
productionat Huexotladeclined.
To summarize,the data pertainingto salt, spindlewhorls,
Cempoala-Pachucaobsidian,and clothall indicatean intensificationof exchangebetweenHuexotlaand theTriple Alliance
capitalsduringLate Aztectimes.In contrast,thedata pertainof
ing to Otumba obsidiansuggestno parallel intensification
exchangebetweenHuexotla and similarlower-levellocal centers.I inferfromthese data that the political unificationof
no
CentralMexico underthe Aztec TripleAllianceperformed
moreintensivesysin stabilizinga progressively
directfunction
temof regionalspecializationand exchange.Instead it appears
that as the Triple Alliance grew in strength,the regional
Exchange
economicsystemunderwenta basic reorientation.
betweenthe Triple Alliancecapitals and local centerssuch as
Huexotla came to dominatethe system; exchange between
Huexotla and otherlocal centersbecame relativelyless important.
URBANIZATION AND THE LATE AZTEC ECONOMY
In lightof thesizes of theLate AztecpopulationsofTenochtitl'anand Texcoco,theirdominantrolein the Late Aztec economy of the Valley of Mexico is not surprising.Accordingto
Calnek (1973:192), Tenochtitlan'spopulationrangedbetween
150,000and 200,000in 1519.Texcoco is estimatedto have had
20,000to 30,000inhabitants(Parsons 1971:120; Sanders1971:
in the Valley of Mexico (Ameca449). Five othersettlements
meca, Ixtapalapa, Tlacop'an, Tlalmanalco, and Xochimilco)
containedabout 15,000each (Sanders 1971:449-50). Any of
thefirsttwo,mayhave been too
thesecenters,but particularly
largeto have been fedby theproduceof theirimmediateagrilocated on a seriesof
culturalhinterlands.For Tenochtitl'an,
islandsin westernLake Texcoco,theproblemof foodprocurementwouldhave been especiallysevere.
For Tenochtitlanand Texcoco, the problemof food supply
was complicatedby a secondfactor,thesociopoliticalstatusof
the residentpopulations.Both communitiescontained particularlyheavy concentrationsof political elites. This was
of empireand partlya consepartlya resultof the prosperity
quence of the strategyexercisedby Triple Alliance rulersto
allies: the rulersof
insurethe loyaltyof theirsemiautonomous
subordinatecity-stateswererequiredto residein Triple Alliance capitals formuchof the year (Cortes 1970[1519-26]:68;
Ixtlilxochitl1952 [ca. 1600-40]:176-77; Motolinia 1950 [ca.
1536-43]:210,215). Cortesdescribedthesituationin Tenochtitlan as follows(p. 65): "There are in thisgreatcitymanyfine
and largehouses,thecause beingthatall therulersoftheland,
city
vassals ofMontezuma,have housesin theaforementioned
and residetherea certainportionof the yearand, in addition
to this,thereare many rich citizensthat also have veryfine
the ratio of
houses." In Tenochtitlanand Texcoco, therefore,
foodproducersto foodconsumerscould have beenparticularly
low (also see Calnek 1972).
The organizationof foodsuppliesforTenochtitlanhas been
analyzedby Calnek (1975) and by Parsons(1976). They agree
fromsubject provincesand "rent"
that tributein foodstuffs
465

Vol. 21 * No. 4 * August1980

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

fromlanded estates assigned to high-ranking


administrative
to feedTenochtitl'an's
personnelwouldnothave sufficed
population.Bothpostulatethatmarketexchangeplayedan important
rolein the food-supply
system.
To acquire foodstuffs
throughthe marketsystem,the residents of Tenochtitl'anand Texcoco could have focusedupon
the productionof nonfooditemsthat could be offeredforsale
to rural food-producers.
The more abundant suppliesof salt
and the growingpredominanceof Tenochtitlan/Tlacopanderivedspindlewhorlsin theHuexotla collectionssuggestthat
thisstrategywas pursuedto a degree.At least one analystof
the Aztec economy(Katz 1966:50) has argued,however,that
urbancraftproductionwas "an artisanproductionof luxury"
supplyinghigh-status
goodsforconsumption
by local elitesand
foruse as itemsofexchangeby long-distance
traders(the pochteca).Sjoberg(1960:198) has goneso faras to arguethatin preindustrialcitiesgenerallythe outputof urbancraftspecialists
is primarily
intendedforconsumption
by urbanites:"Although
someofthe city'sproductsare conveyedto the countryside
by
travellingmerchants,theseconstitutebut a small fractionof
urbanmanufactures."
The growingquantitiesof Cempoala-Pachuca obsidian at
Huexotla and the strikingdeclineof clothproductionsuggest
a secondstrategyforacquiringfoodstuffs
throughthe market
system.A portionof the massivequantitiesof nonfoodtribute
flowingintotheTripleAlliancecapitalswas used by theurban
elitesto pay forfoodoffered
forsale in theurbanmarketplaces.
Historical sources indicate that great quantitiesof tribute
goods were directlyredistributedby Triple Alliance rulers
to attendant nobles, valiant warriors,skilled artisans, disabled veterans, widows, and orphaned children (Cortes
1970[1519-26]:56; Duran 1971[1570]:195; Ixtlilxochitl1952
[ca. 1600-40]:234; Tezozomoc 1975[1598]:301,323, 430, 66669). Since manyof thesepersonslived in the cities,a sizable
proportionof thesegoods may have been carriedto the urban
marketplacesto be exchangedforfood.(Sahagun [1956(1577):
156]describesthe use of clothto purchasefoodin the marketplace as ifit werea customaryactivity.)
The purchaseof foodwithredistributed
tributegoods could
account forthe reorientation
of the Valley of Mexico market
systemwhich, I have argued, occurredduring Late Aztec
in urbanmarketplaces,
times.The heavydemandforfoodstuffs
coupled with the abundant supply of nonfoodtributeitems,
wouldhave resultedin morefavorabletermsof trade in those
marketplacesfromtheperspectiveofruralfood-producers.
The
marketsin Triple Alliancecapitals would have gained a competitiveadvantage over the marketplacesof hinterlandcommunitiesand would have been able to commandmore than
theirfairshareof marketing
activitiy.
A second consequenceof introducing
nonfoodtributeitems
into the marketsystemwouldhave been to depressthe value
of rurallabor devotedto theproductionofnonfooditemsrelative to that of labor devotedto foodproduction.Rural communitiesall over the Valley of Mexico would have responded
by decreasingtheirproductionof nonfooditems (either for
householdconsumption
or formarketsale) and expandingtheir
productionof food.The less intensiveproductionof clothand
more intensiveproductionof maize and maguey syrup byproductsat Late AztecHuexotlasuggeststhatthisdid, in fact,
occur.6The outcomeof thistrendwouldhave been the transformationof the rural populace into a populationof foodproducingspecialists,exploitingthe environmental
diversity
6 The width of the alluvial plain to the west of Huexotla and the
gentleslope of the piedmontzone to the east create conditionsthat
are exceptionallyfavorableto agriculturalproduction.Hence, it is
possibleto argue that theintensification
offoodproductionat Huexotla duringLate Aztec times is not representativeof a general reorientationof ruralproductionin the Valley of Mexico, but a consequence of the competitiveadvantage Huexotla enjoyed. This issue
can and shouldbe resolvedthrougharchaeologicalresearchat Aztec
siteslocated in areas less favorableto agriculturalproduction.

466

of the valley insofaras thisdiversityaffectedfoodproduction


(as in the differential
productionof maize and magueyin the
lowerand upper piedmontsectorsof Huexotla) but ignoring
environmental
variationthatmightotherwisehave servedas a
basis forspecializedcraftproduction.The conditionsof the
of
Late Aztecmarketsystemweresuch thattheintensification
of some
regionalexchangecould generatean intensification
typesof specializationbut not of others.
CONCLUSIONS
of
The data fromHuexotla suggestthat the intensification
marketactivityduringLate Aztectimeswas closelyrelatedto
seems
theexpansionoftheAztecstate.In fact,thisrelationship
to have been moreintimatethan I firstanticipated.
I originallyenvisionedthe state as playinga passive, facilitative role in commercialexpansion.I hypothesizedthat the
state,by preventingthe outbreakof warfarewithinthe marketingregionand by mediatingconflictsbetweenbuyersand
sellersand specialistsof various types,createda situationin
whichtheintensification
ofspecializationand marketexchange
The majorimpetus
couldproceedfreeofpoliticalimpediments.
forcommercialintensification,
I believed,was the greatereffidiverse
ciencyof specializedproductionin an environmentally
habitat.However,the data fromHuexotla have led me to suspect thatstate expansionplayedan active rolein the intensificationoftheAztecmarketeconomy.It seemsto have provided
a stimulusforintensification
and to have determined
how more
intensiveparticipationin the marketsystemalteredruralproductionduringLate Aztec times.
was
I would now suggestthat commercialintensification
during
generatedby the emergenceof large urbansettlements
Late Aztec times.It seemsappropriateto add the Late Aztec
example to the list of cases in which marketsystemshave
(see
grownin responseto urbandemandforpeasant foodstuffs
Smith 1977:128-29). I suspectthat in the case of the Aztecs
urbanizationwas stimulatedby state expansion.The populationof the Triple Alliancecapitals increasedas largerbureaucratic staffswere needed to administera growingempire,as
greaternumbersof conqueredrulerswererequiredto take up
and domestics
residencein the capitals,and as morecraftsmen
foundemploymentin the serviceof the elite population.In
otherwords,commercialintensification
was a resultof urban
growth,and urban growthoccurredbecause of the expansion
of theTripleAlliancesphereof conquest.
Tributegoods,a secondresultof Triple Allianceexpansion,
werealso importantin the processof commercialintensification.The Huexotla data suggestthat tributein nonperishable
rawmaterialsand craftgoodsenabledthecitiesto pay forfoodstuffsoffered
throughthe marketsystem.The passage of nonperishabletributegoodsintothe marketsystemseemsto have
had a noticeableimpact upon ruralproduction.By lowering
the value of most nonfoodcommoditiesin relationto foodin favorof theproducstuffs,
it induceda reallocationof effort
theabilityof the marketto provide
tionoffood;by improving
a steadysupplyofnonfooditemsto ruralhouseholdsat reasonable prices, it created conditionsin which peasants would
sacrificetheireconomicautonomyforgreaterdependenceupon
commercialactivity.It seems likelythat the greatervolume
of urban-ruralexchangeduringLate Aztec timeswas due as
muchto the circulationofimperialtributethroughthe market
systemas to an elaborationof the divisionof labor between
urbanand ruralproducers.
Thus, by stimulatingthe processof urban growthand by
intothemargenerating
a flowoftributegoodsforintroduction
ket system,the expansionof the Aztec state resultedin an
intensification
ofthe Aztec marketsystem.
To what extentdid commercialexpansioncontributeto the
formationand persistenceof the Aztec state? Certainly,the
CURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

Late Aztec marketsystemlent supportto the processof urbanization,and urbanizationcontributedto persistenceof the
Aztec state in at least two ways: it facilitatedcommunication
betweenspecializedsegmentsof the state's bureaucracy,proand it enabled the
vidingfor more effectiveadministration,
state to requirethe residenceof conqueredrulersin the Triple
Alliancecapitals,wheretheiractivitiescould be monitored.In
addition,commercialintensification
probablyresultedin an
increasein prosperityfor the Late Aztec populationin the
ValleyofMexico,insuringtheirloyaltyto the state. I suggest,
was foundedupon the passage of
however,thatthisprosperity
tributegoodsintothemarketsystemand thelowercost ofnonfood commoditiesthat this entailed.This contrastswith my
in whichprosperitywas attributedto the
initialformulation,
efficiencies
in productionachieved througha more complex
divisionof labor.
It is my initialemphasisupon the relationshipof a complex
that seemsmost
divisionoflabor to commercialintensification
in need of revision.The Early Aztec materialsfromHuexotla
suggestthat the Aztecmarketsystemdid originallyserveas a
means of distributingthe productsof local specialists to a
regionalpopulationofconsumers.
Hence,specializationand the
environmental
diversityupon which it was based were imoftheAztecmarketsysportantfactorsin theinitialformation
tem. Rather than developingthroughan elaborationof this
systemof local specializationand regionalexchange,however,
the Late Aztec marketsystemintensified
because its original
functionwas supersededby a second: the exchangeof tribute
forfoodbetweenurbanand ruralpopulations.The exploitation
of local environmental
diversitythrougha complexdivisionof
laborwas probablyless crucialaftertheformation
of theAztec
Triple Alliance than before.As the Triple Alliance sphereof
conquestgrew,it createda networkof economicrelationships
thatextendedbeyondtheconfines
ofthevalley.Tribute-paying
populationsin the moreremoteprovincesof the empirecould
be coerced into assumingresponsibility
for certain types of
productiveactivitiesthat had previouslyfallento the inhabitants of the valley. The resultwas an economicsystemthat
made the best use of the resourcesof Central Mexico as a
whole (so faras the valley's inhabitantswere concerned)but
one that was attuned to environmental
diversitywithinthe
valleyonlyas thisdiversityaffectedtheproductionoffood.7
Thus, I tend to reject the suggestionthat the Aztec state
aroseto facilitateexchangebetweenspecialistsofdiversetypes.
I would argue that marketexchangein the Valley of Mexico
duringLate Aztec times was dominatedby the exchangeof
tributeforfoodbetweenurbanand ruralpopulationsand not
by theexchangeofproductsbetweengroupsofspecialists.The
success of the Aztec Triple Alliancerestedupon its abilityto
intervenein the marketsystemin such a way as to insurethe
supportof its administrative
centers.
This paperhas focusedupontherelationship
ofspecialization
and exchangeto the state withinthe specificcontextof Aztec
culture,but its conclusionsmayhave widerrelevance.Political
centralization
has frequently
entaileda degreeof demographic
nucleation(see Sjoberg1960),and therulersofmanyemerging
states have had to concernthemselveswith the problemof
feedingurban populations(see Tilly 1975). Solutionsto this
problemwillhave been foundin variousformsof intervention
in existingsystemsofproductionand exchange.Thus, I would
expectthatarchaeologicalinvestigation
ofemergingstate-level
politieswill often reveal evidence of extensiveeconomicreorganization(see Johnson1973). Reorganizationwill neverbe
7Thus, while I would agree with Sanders, Parsons, and Logan
(1972:177-78) that the extremelydense populationof the Valley of
Mexico during the Aztec period was supportedby intensive,specialized food productioncoupled with an effectiveregionalmarket
system,I would also emphasizethe contributionofmilitaryconquest
and tributeextractionto thesubsistencebase in thecenturypreceding
European contact.

Brumfiel:SPECIALIZATION,

EXCHANGE, AND THE STATE

entirelydivorcedfromconsiderations
of adaptive effectiveness
or economicefficiency
(Price 1979:299-300),and in somecases
theseconsiderations
willthemselveshave providedtheimpetus
to change. In other cases, however,the impetus will have
stemmedfromnew patternsof administrativecontrolintroduced by the emergingstate to secureand extendits power.

Comiments
byKENNETH L. BROWN
Department
ofAnthropology,
University
ofHouston,Houston,
Tex. 77004, U.S.A. 16 i 80
This paper representsan interesting
discussionof the Aztecperiodeconomicsystemof the Basin of Mexico. Unlikeother
it viewsthe systemfromtheperspectiveof a rural
treatments,
secondary political/economiccenter. While this procedure
offers
new insights,thepaper has a numberofproblems.
Brumfielestablishesas her centralthemean evaluationof
Sanders's arguments (1956, 1965, 1968) "concerningthe
economic functionsof Central Mexican states." However,
as Brumfielnotes,referto the originof
Sanders's comments,
CentralMexican states and not to theirhighestlevel of comthe Aztec system.In
plexityand centralizedadministration,
is the highdegreeof
fact,what Brumfielneatlydemonstrates
centralizedadministrationof the economic system by the
Aztec rulers.She neverdirectlyreturnsto the questionraised
by Sanders.She does statethatthe Aztececonomicsystemwas
primarilypolitical rather than an economic-maximization
strategy,but it is preciselythispointthatmay be herweakest.
to determinesolely on the basis of the
While it is difficult
Huexotla data givenby Brumfiel,her discussionof the other
information
available on the Aztec economysuggeststhat it
was an attemptat economicmaximization.
Brumfieldemonstratesthe changes in behaviorwhich occurred at Huexotla as the complex, territorially
expansive
Aztecstateevolved.Alongwiththesechanges,shedemonstrates
theAztecmanipulationofthetributesystemto introducenonfooditemsinto the economy.Such action,as Brumfiel
argues,
wouldhave reducedthe value of nonfoodgoods,since moreof
thesegoods (e.g., cloth) would have enteredthe market.This
would have forcedmanypeople to turnto primaryfoodproductionin orderto survive.So far,no debate. However,the
Aztec administrators
actually had two alternativesin their
of the economicsystemto meetthe demandsof
intensification
as tribute,
an everincreasingpopulation:to extractfoodstuffs
thusallowingthe Basin economicsystemto continuemuchas
it had before,or to importnonfooditemsand forcea shiftin
the internalBasin economytowardincreasedlocal foodproduction.They selected the second. The questions,then,are
a
whytheyselectedthisalternativeand whetherit represents
maximizationstrategy.
In the answerto thesequestions,thereare two criticalfactors: the use rate of the variousitems(foodvs. nonfood)and
the transportation/manufacture
costs of theseitems.Use rate
can be definedas therate at whichitemsare totallyconsumed
(the time fromproductionto completeeliminationfromthe
is highcomparedto that
system).The use rate forfoodstuffs
of cottoncloth or obsidian tools. Thus, by selectingthe first
alternative,the Aztecswouldhave been forcedto keep renewNonfooditemsare consumedat a
ing theirstockof foodstuffs.
slowerrate. In termsof the transportation/manufacture
costs,
one mustconsidernot only the weightbut the bulk of items.
Outsideof the Basin, goods wereprimarilymoved by human
labor-an inherentlyexpensiveprocess. Basic foodstuffs
are
both heavy and bulkyin comparisonwithcloth.Manufactur-

Vol. 21 * No. 4 * August-1980

467

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ing costs includethe energyemployedin production.Not so


withinthis equation. Raising
obvious is the energytrade-off
cottonor magueyforfibersremovesland fromfood producthe individualmustpurchasethe lost food.By
tion; therefore
clothand othernonfoodcraftitems,theAztecsystem
importing
promotedthe developmentofhighlyintensivefoodproduction
systems.
The Aztecs established a highly productive-maximally
productive,given theirtechnology-economicsystemwithin
theBasin. Forcingtheproductionofbulky,heavy,and rapidly
consumedfooditemswithinthe Basin made ecologicalsense.
It reducedtransportation
costsbecause thefoodwas consumed
in the local systemin whichit was producedor, if carriedto
one of the major urban markets,had to move only relatively
shortdistancesand could take advantage of the lake system.
Importingitemswhichwereconsumedslowly,wereless expenand tookenergyaway fromfoodproduction
sive to transport,
oftheBasin agricultural
wouldpermitincreasedintensification
Basin
make use of the highlydiversified
systemand therefore
ecology.
The selectionofthesecondalternativeis in factan excellent
maximizationstrategygiven the diverseecologicalsettingof
theBasin. The Aztecmayhave gainedpoliticaldominanceand
held it because theireconomicsystemworkedbetterthan the
previousones. The large and dense populationit supported
atteststo its adaptability.WhileBrumfielmay want to debate
the ecologicaladaptabilityof the Aztec system,her own data
demonstratethat the debatewilllikelybe a shortone.
byPEDRO CARRASCO
ofNew Yorkat
State University
ofAnthropology,
Department
StonyBrook,StonyBrook,N.Y. 11794,U.S.A. 12 ii 80
interpretation
In general,I am inclinedto agreewithBrumfiel's
ofherarchaeologicaldata fromHuexotla,since,as I have said
elsewhere,I believethat the economicorganizationof ancient
and thatmarketorganization
Mexicowas politicallyintegrated
played a subordinaterole (Carrasco 1978). Her paper shows
how archaeologycan providedetailedquantifiabledata about
thetemporaland spatial distribution
oftheproductionand use
is unavailable,as, forinof goods whenhistoricalinformation
stance, in her data on maguey scrapers,salt pots, spindle
whorls,or obsidian.Yet theseobjectsrecordonlya fewof the
manycraftsand goods that existedin ancientMexico; many
goods and productiontools leave no materialremains.Obviously, we need to compare systematicallyall the available
evidenceon materialculturefrombothhistoricaland archaeological sources. Historical documentationsuggeststhe existencein Huexotla of a largernumberof specializationsthan is
treated here. Informationfor preconquestor early colonial
timeson the divisionof labor in Tenochtitlan(Sachse 1966),
Huexotzinco(Prem 1974, Carrasco 1974a), and a few other
places such as Xochimilco,Coyoacan,and Tlaxcala raises the
questionas to what the picturewas in Huexotla. From the
Huexotzinco material one would expect such major craft
mat makers,and
groupsas potters,carpenters,stoneworkers,
basket makers to be representedin a city-statethe size of
Huexotla. One would also expect certaincraftsto have been
presentin the centralurban settlementof Huexotla. Crafts
practiced in the palaces, as reported elsewhere,included
wood carving,sculpture,fine stoneworking,
featherworking,
and painting.What archaeologicalevidence,if
goldsmithing,
any, can one expectto findof the presenceand importanceof
all theseothercrafts?And whatwas therelativeimportanceof
theactivitiesdiscussedby Brumfielwithinthe total divisionof
labor in Huexotla?
also suggestsadditionalquestions.
Historicaldocumentation
The parochialarchivesof Huexotla showthat one of its major
wards was that of Tlailotlacan,whosepeople were knownas
paintersand made historicalbooks (Ixtlilxochitl1952:69-70)
468

and wereprobablyalso merchants(Acosta Saignes 1945:32).


Sahagiun(1959 [1577],Book 9:48) includesHuexotlain thelist
of townswhosemerchantswenton expeditionsto Tochtepec.
We also knowthatHuexotlawas one of the townsin the Acolhuacan whose rulers held lands in each other's territories
(Guzman 1938:95). This would imply a certain amount of
movementof goods, not necessarilychannelled
interregional
throughthe capital at Tetzcoco and not always organizedby
the market.Historicaldocumentationthusraisesbroad issues
is not available. Brumfiel's
even when detailed information
materialshows that archaeologycan contributetowardsthe
ofsomeproblems.It is to be hopedthatmoresuch
clarification
studies,whenavailable fora largearea, willincreaseourknowledge not only of the technology,but also of the economic
organizationofancientCentralMexico.
A small point about the use of the term"Aztec": Ethnohistoriansuse it rarely.It is usuallyapplied to thepeoplewho
lived at or migratedfromAztlan,accordingto historicaltraditions. It is also oftenused in the phrase "Aztec Empire" to
denote the Triple Alliance of Tenochtitlan,Tetzcoco, and
along withotherrecentwriters,seems to
Tlacopan. Brumfiel,
apply it to all post-Toltecpeoplesin the Basin of Mexico. The
problemwiththisusage is thatit suggestsan ethnicor political
unityof the Basin whennone existed.Ethnicityand political
and fragmentation
showeda greatdeal ofdiversity
organization
withintheBasin, whileat thesame timevariouspeoplesin the
Basin had ethnicand/orpoliticalties with othersbeyondits
mountainrim. It is betternot to use labels that tendto distortthisbasic fact.
byROBERT CHADWICK
N.J. 08903,
New Brunswick,
University,
Transaction-Rutgers
U.S.A. 5 II 80
It is unfortunatethat Brumfielhas not made more use of
Aztec sourceson the marketand tributesystemto document
herideas about specialization,marketexchange,and theAztec
stateas viewedfromHuexotla.The SpanishversionofSahagiun
is listed, but not the more informativeEnglish edition by
Andersonand Dibble of the FlorentineCodex, whichcontains
moreextensivedata on Aztecmerchants(Sahag(un1959[1577]).
is AcostaSaignes's
fromherlistofreferences
Missingcompletely
(1945) classic workon Aztec merchants(see Chadwick 1966
for summaryand discussion).Perhaps a study of this sort
study than
should rely moreheavily on documentary-source
on an archaeologicalsurfacesurveythat is likelyto presenta
skewedpictureofa verycomplexsubject.
byTHOMAS H. CHARLTON
of Iowa, Iowa City,
University
of Anthropology,
Department
Iowa 52242, U.S.A. 10 II 80
Brumfiel'spaper and the dissertationupon whichit is based
are welcomeadditionsto approachesused in the studyof the
evolutionof secondarystatesin Mesoamerica.Brumfielgeneralizes fromthe data and resultsof her Huexotla studyto the
evolutionof city-statesthroughoutthe Basin of Mexico. She
suggestsalternativemodelsand hypothesesto accountforthe
apparent anomalies in her data fromHuexotla and general
trendsin Late Aztec regionaleconomy.As admirableas this
of untestedmodels and hypotheses
intricatesuperstructure
it is necessaryto directattentionto
may be in its formulation,
the questionof the validityof the Huexotla data and their
applicabilityto the entireBasin of Mexico.
It is not at all clear that Huexotla was a suitablechoiceof
city-stateunit withwhichto examinequestionsof local-level
specialization,exchange,and politicalevolution.Texcoco's influence,althoughcited forthe Late Aztec period,is nowhere
mentionedforthe Early Aztecperiod,a timeduringwhichthe
Acolhuaquewere expanding(cf. Alden 1979:173). The proxCURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

imity of Huexotla to Texcoco undoubtedlyresultedin its


into a sphereof influencewhichaffected
early incorporation
and distortedany local-leveldevelopments.In addition,althoughBrumfieldescribesthe Huexotla city-stateas an area
with considerableenvironmentaldiversity,she has sampled
onlyone section,a sectioncharacterizedby limiteddiversity.
It is also necessaryto raise questionsabout samplingprocedures,includingthe level of sampling,the relevanceof the
materialscollected,and the neglectof the postconquestoccupation.Withintheurbansectorthe 1% level of samplingis so
low that evidence for craftspecializationcould easily have
been missed.CharlesS. Fletcherand I in 1966locateda heavy
concentration
of spindlewhorlsindicatinga workshop.A reexaminationof this area in 1977 indicated its continued
presence.Brumfiel'ssamplingmissedit. Severalof theartifact
categoriescited by Brumfielin table 1 as relevantto testing
forpast behaviorpatternsrelated to symbioticrelationships
need furtherstudy beforethe postulatedbehavioralimplications are reasonablysecure. In collectingartifactsBrumfiel
excludedglazed sherds,hopingto excludepostconquestoccucurrentevidencesuggestsquitestrongly
pation.Unfortunately,
that Aztec unglazedceramicspersistedforat least 100 years
afterthe conquest.Thus the "Late Aztec" ceramicsand assoa postconquest
ciatedartifactsat Huexotlaprobablyrepresent
chapel,
One structure(Ti. 31) is a colonial-period
distribution.
indicatingpersistenceof occupation.
Even if the data generatedin Brumfiel'sHuexotla study
werereliable,therewouldstillremaintheproblemofhow well
in theBasin ofMexico. Studies
othercity-states
theyrepresent
of theOtumbaarea in theTeotihuacanValley (Charlton1971)
indicate that Otumba did possess significantlocal-levelspecialization and exchange,unlike Huexotla as describedby
Brumfiel.This suggeststhat Brunfielhas overextendedthe
applicationof herdata.
weaknessesin
In summary,I suggestthereare significant
the data base fromwhichBrumfielgeneralizesto the Basin of
of her economicmodels.The apMexico in the construction
parentdiscrepanciesbetweenthe Huexotla data and the predicted occurrenceaccordingto Sanders's models may result
fromHuexotla's relationship
to Texcoco,Huexotla's ecological
situation, and the sampling proceduresused. Data from
OtumbaindicatethatresultsfromHuexotlashouldnot be uncriticallyapplied to the entireBasin of Mexico.
byTOM D. DILLEHAY
LexingofKentucky,
ofAnthropology,
University
Department
ton,Ky. 40506, U.S.A. 17 ii 80
Brumfielshouldbe complimented
to reexamine
upon hereffort
theAztececonomyfromtheviewpointoftheimpactofpolitical
decisionsupon specializationand exchangeat both the local
and theregionallevel.This is a formidable
task thatchallenges
the axiom that highly centralizedadministrativenetworks
developed,in part, in responseto the need to stabilizelarge,
expansiveeconomies.Her articlehas the virtueof combining
There are,
references.
archeologicaldata and ethnohistorical
however,some seriousweaknessesin substancethat make it
difficult
to appreciatethe ideas that she is tryingto convey.
The major weaknessrelatesto the failureto clarifycertain
positions.For example,Brumfielrefersto the "distortion"of
the economyduringLate Aztec timeswithoutever explicitly
theresult
showinghowtheeconomywas distorted.Is distortion
of the increasingroleof state politicsin regionaleconomiesor
of thefailureto strengthen
competi"the forcesoffree-market
tion"? The argumentseems to blend formalexchangetheory
withsubstantivism
of the
as representedby acknowledgment
importanceof Aztec politicsin economicdecisionmaking.In
addition,thepoliticalvariablesthataffectedthe patterningof
regionalexchangeare neverdiscussed.The readermustassume
thattheyare tiedup withdecisionmaking.Further,thereis no

Brumfiel:
SPECIALIZATION,

EXCHANGE, AND THE STATE

explanationof the decisionto use tributeto purchasefoodfor


the urban populations.More extensiveuse of ethnohistorical
documentswouldprobablyhave clarifiedsome of thesepoints
the argument.
in additionto strengthening
I am puzzled also about the relationshipbetweenthe nonproducingurbanitesand the ruralproducers.Brumfielstates
that in orderto obtain food throughthe marketsystem,the
and Texcoco could have focused
inhabitantsof Tenochtitlcan
on the productionof nonfooditemsthat could be offeredfor
"sale" to rural food producers.Presumablythese nonfood
itemswerereceivedin the formof tributefroma ruralpopulation otherthan the one that providedfoodforthe market,
but this is unclear.If a givenpeasant populationwas paying
tributeas well as providingfood for exchange,was it not
being overtaxed?Or were theredistinctgrades of specialist
among the peasants? It is also unclear whetherthe food
providedat the marketconstitutedanotherformof tribute.
In any case, the questionis raised what the urbaniteswere
in return."Know-how" in the formof technocontributing
logicalinnovation?Protection?
Brumfielstates that one intentof the paper is to evaluate
Sanders's convictionthat higherformsof politicalunification
emergein responseto the growingcomplexityofan expanding
she does not adequately
regional economy. Unfortunately,
assess Sanders'sideas in lightof the data presented.Such an
assessmentwould requireinterpretation
of the emergenceof
the Late Aztec political systemas it relates to economy.
Brumfieldoes not treatthis emergence,but ratherdeals with
the means and end resultsof economicpolicies which have
been impactedby politicaldecisionmaking.
There are also some seriousproblemswiththe data acquisimaterials,while
tion technique employed. Surface-collected
probably very reliable primarysource material for chronomust be consideredinlogical and functionalinterpretation,
of socioculturalchange.Moreforthe interpretation
sufficient
over, thereis substantialsocioculturalvariationin the late
of the Valley of Mexico. One mustwonderwhether
prehistory
thetotalAztececonomyin thearea can be interpreted
through
a single archeologicalsite and its relationshipswith other
settlements.
byCONNIE L. GORDON
New York,
of Anthropology,
ColumbiaUniversity,
Department
N.Y. 10027,U.S.A. 1 II 80
Beginningwith the premisethat an article's title correlates
stronglywith the material that follows,I was puzzled by
article.Both thedata used and theuncleartheoretiBrumfiel's
cal approach taken obscureratherthan clarifythe economic
and politicalissuesrelatingto theAztecsystem.What Brumfiel
presentsresemblesa site reportaccompaniedby an attemptto
comprehendartifactualdata in termsof state-levelprocesses.
betweenthe environment,
Considerationof the interrelation
the productionbase, population distributionsand market
and
location,and the characterof sociopoliticalstratification
flowswithina givensystemis essential
its effecton commodity
in economicanalysis.The marketsystemcomprisedonly one
sector of the Aztec economy;there was at least one other
one
sector,a state-organizedand -dominatedredistributive
(Carrasco 1978, Gibson 1964). Both mustbe consideredwhen
system.
examiningthe dynamicsof the Aztec socioeconomic
For historicalreasons,documentationof the sociopolitical
and economic processes and the archaeologicalremains is
abundant for the late preconquestBasin of Mexico. In this
area, the geopoliticaland demographiccenterof the imperial
Aztecstate,economicprocesses-includingthoseofintensifying
production,specialization,and exchange-are conspicuous.
subYet, takenout ofits propercontext,a smallgeo-economic

Vol. 21 * No. 4 * August1980

469

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

sectionof the Basin such as Huexotla inadequatelyrepresents


boththestructure
oftheAztecpolityand thedynamiceconomic
processesunderinvestigation.
On the state level, specializationand marketexchangecan
onlybe understoodin relationto the productionbase and the
political constraintsplaced on access to goods and services
withina givensystem.For example,duringthe Aztec period
on the
the productionbase was expandingand intensifying
regionallevel (Parsons 1971, Sanders 1957, Sanders,Parsons,
and Santley1979),withshiftsin themodeofproductionoccurringin the southernlake chinamparegionand in the waters
surroundingthe Aztec capital (Armillas 1971, Coe 1964,
Palerm 1973). By late Aztec timessettlementand agricultural
productionin the Texcoco region(whichincludedHuexotla)
had great time depth (Parsons 1971), and the production
regimetherehad undergonea shiftfromrainfall-dependent
agricultureto irrigatedagriculture.Both of theseagricultural
strategiessupportedthe singlecroppingof food staples. Hydraulicagriculture,however,was moresecure and permitted
theintenseuse ofgreaterareas ofproductiveland: organization
and controlof thesewaterworksrestedwiththe state. In contrast, state-controlleddikes installed relativelylate in the
pre-Hispanicsequence(Armillas1971,Palerm 1973) protected
the southernlakes fromsaline inundation,and labor-intensive
densesettlementthereand
agriculturaltechniquesunderwrote
permittedthe multicroppingof staples and truck-garden
products (Armillas 1971, Palerm 1973, Sanders 1957), the
bulk of the latter destinedfor sale in Tenochtitlan'slarge
daily markets.
Further,in relation to the production base, part- and
full-timeeconomic specializationarose from local resource
extraction(e.g., lime,salt, fishing,
hunting,and the collection
ofsmallplantsand animals,rushweaving,potteryproduction,
etc.). It appears that thelimitedsampleand the natureof the
artifactsexaminedby Brumfielprovideevidenceforregional
specializationin theproductionofgrainin theHuexotlaregion.
zone, in termsof
Analysisof specializationby environmental
or vis-a-vis"cottage-type"
legallydefinedextractiveprivileges,
on thesettlement
levelis lackingin Brumfiel's
workindustries
a furtherindicationthat her conclusionsshould be regarded
withthe utmostcaution.
By late Aztec times,marketexchangewas controlledby the
state (Cortes 1962 [1519-26],Diaz 1963 [1568],Sahag(un1959
[1577],Books 8, 9). The numberof marketsin the Basin of
Mexico was fixedby the state; theirlocation coincidedwith
long-distancetrade routes(Gordon 1980),and provincialcenterswereeconomicand politicaldependentsof the two major
regionalcenters,Tenochtitlanand Texcoco (Gibson 1964). The
geographicproximityof Aztec settlementsto marketcenters
and
representedleast-costsolutionsto specificenvironmental
sociopoliticalproblems(see Gordon 1980 for a central-place
analysisofAztecmarketsin theBasin ofMexico). Considering
the nature of Aztec transportsystems-paleotechnicfoot
on landand bioenergetically
moreefficient
waterborne
transport
on canals maintainedby thestate(Cortes1962[1519transport
26]; Gibson 1964:62)-markets werelocated withina reasonable distance of most settlements.This led to the multiple
patronageofdailyand periodicmarketsto obtaincommodities
of limitedgeographicaldistribution.
Brumfiel
overlooksthefactthatperiodicmarketssuchas the
one in Huexotlareflectcertainregularities
in marketexchange.
Peasant communities
thereand in otherportionsof the Basin,
forexample,were small and, in productiveterms,relatively
uniforminternally.A numberof such communitiesvarying
zone or by differenwithinthemselves(e.g., by environmental
tial resourceextractionactivities) and supplementedby the
presence of full- or part-timespecialists in urban centers
createdsuficientdemandto makemarketsprofitable
(McBryde
1947).
The Aztec state strictlydefinedaccess to mostcommodities

470

withinthe economy.The basic economic"wants" of peasants


were satisfied:by and large, they produced, collected, or
barteredfor their minimal "needs" in market centers,for
formaltrade outside market centerswas prohibitedby the
the milistate (Duran 1867:179). Nobles, state functionaries,
tary,and the clergyweresupportedby the tribute/redistributive system,which included staple, nonstaple,and "exotic"
long-distancetrade goods. The incipientmiddleclass of merto the "wants" ofthe
chantsand craftsmen
catered,primarily,
Aztec elite. Therefore,the largest,most diversemarket/craft
not
centerscoincidedwiththe largestelite urbansettlements,
prowithruralprovincialcenters.Huexotla's undistinguished
vincial marketmust thereforebe viewed in relation to the
patternwithinthe Basin. Clearly,the nearby
largermarketing
outletforthe sale
marketin Texcoco provideda moreefficient
of specialized goods which may have been producedin the
Huexotla zone.
Space limitationspreventan in-depthdiscussionof the complex interplayofthevariablesoutlinedhereand ofthepolitical
and economictensionstheycreatedwithinthe Aztec political
economy.Yet, whathas been presentedindicatesthe direction
that Brumfiel'spaper mighthave takenin usingthe Huexotla
data to provide a fullerunderstandingof the processes in
whithshe is interested.
byROGER D. MASON and DENNIS E. LEWARCH
Universityof Nebraska-LincolnCannon ReservoirHuman
EcologyProject,Box 68, Perry,Mo. 63462, U.S.A. 9 ii 80
Brumfiel'sgeneralizations,based on archaeologicalevidence
fromHuexotla,providea new perspectiveon the organization
of Aztec economicsystemswithinthe Basin of Mexico. When
interpretations
viewedin conjunctionwithrecentethnohistoric
(Carrasco 1978; Calnek 1975; Katz 1972:218), the data indicate that a model that emphasizes a hierarchyof central
places, each havingspecializedeconomicfunctionsin proportion to its size (as originallysuggestedby Sandersand Price
[1968:159] and morerecentlyby Smith[1979]),is an unwarrantedprojectionof Westerneconomicpatterns.Interestingly,
Sanders,influencedby Brumfiel'swork,has recentlychanged
his viewpointto conformmorecloselyto the Huexotla results
(Sanders, Parsons, and Santley 1979:179-80). Similarly,
of documentarydata
Carrasco's (1978) recentinterpretation
suggeststhat the Aztec economicsystemwas politicallyadministered,which would imply centralizationof specialized
economicfunctions
at the centerof politicalpower(Tenochtitlan-Tlatelolco)as Brumfielhas suggested.
Brumfiel'sresults demonstratethe potential of intensive
about
surfacecollectionforprovidingfunctionalinformation
complex,spatiallyextensiveeconomicsystems.Whilethework
overmuch
an improvement
carriedout at Huexotla represents
oftheprevioussurfacearchaeologicalresearchon Aztecsystems
moreor
fromdeficiencies
in the Basin ofMexico,it stillsuffers
less commonto extantstudieswithregardto (1) treatmentof
archaeologicalformationprocesses,(2) samplingdesign, (3)
and (4) the quality of the data used to
artifactclassification,
inferences.
supporthigher-level
There can be littledoubt that intensive,systematicsurface
field technique for providing
collection is a cost-effective
ofartifactsin space. Little considerquantitativedistributions
ation has been given, however,to the formationprocesses
whichconditionarchaeologicalmaterialand the consequences
of theseprocessesforfieldperceptionsof artifactdensityand
samplesderivedfromsurfacecollection.Samplingand formationprocessesare ofparticularimportance
here,sinceBrumfiel's
of specialargumentsrest on the quantitativedetermination
ization. Althoughthe sampling programat Huexotla is a
marked improvementover the grab samples and subjective
assessmentof sherd densitycommonin otherprojects,it is
still inadequate to controlforthe spatial structureof artifact
CURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

withinthe site. Determinationof the degreeof


distributions
economicspecializationrequiresnot only populationsof artiof funcfact classes,but also spatiallydefinedconcentrations
tional artifacttypesat a level below that of the site sectors
employedby Brumfiel.The sample size at Huexotla may be
adequate forreliableestimationof the populationparameters
are not discussed),but
of site sectors(samplingconsiderations
forstudyoftheinternalstructure
it is mostlikelynotsufficient
of the site. It is beginningto be recognized(e.g., Jermannand
Dunnell 1979) that delineationof spatial pattern calls for
highersamplingrates thanpopulationestimators.
There are numerousproblems in the classificationand
statisticalanalysisofportableartifacts.Most seriousis theuse
of ethnographicanalogy, rather than explicit definitionof
functionalclasses and analysis of wear-relatedvariables, to
inferartifactuse. Virtuallyabsent fromthe discussionare
ceramics (other than thick-walledvessels and comals) and
groundstone. Certainly,these classes of artifactsplayed as
importanta role in economicactivitiesas spindlewhorlsand
chipped stone. Carrasco (1978:34) mentionspotters,stonecutters,brickmakers,carpenters,basket makers, and mat
makers as specialists commonlyfound in towns. Although
someof thesemay not have archaeologicalcorrelates,the possibilityof theirexistenceat Huexotla should be entertained,
ratherthan dismissingspecializationin the urban zone solely
on the basis of frequenciesof spindlewhorls,figurinemolds,
and prismaticobsidiancores.
These deficienciesare compoundedwhenapplied to higherlevel generalizations.For example, Brumfiel apparently
assumes only productionwithinhouseholds,since no samples
were taken betweenhouses in the rural area. This precludes
investigationof the possibilityof productionorganizedat a
level above that of the household,employingworkersin a
specializedarea away fromtheirindividualresidences.In the
urban area, if one assumes householdproduction,it would
appear thata collectionunitwouldhave fallenwithinless than
10% of the houses.Data fromthis small proportionof urban
houses are being comparedwithsamples fromall or most of
the ruralresidences.
Thus, while Brumfiel'spropositionsabout Aztec economic
the severelimitationsof the suporganizationare interesting,
to assess them.This criticism
portingevidencemake it difficult
is intendedsimplyto point out how far researchin Mesoamericamustprogressbeforean adequate archaeologicaldata
inferences.
base can be generatedforthe testingofhigher-level

byHATTULA

MOHOLY-NAGY

1204 Gardner,Ann Arbor,Mich. 48104, U.S.A. 7 ii 80


I am glad to see the directapplicationof archaeologicaldata
to the complex relationshipbetween exchange systemsand
artipoliticalcomplexity.Using selectedtypesof cooccurring
materialsinstead of focusingupon just one
factsof different
industrysuch as ceramicsor obsidian gives a betteroverall
idea of the Aztec exchangesystem.I also like the attention
paid to the differentspatial contextsof selected artifacts
duringEarly and Late Aztec times.
My onlycriticismofthisexcellentpaper concernsthespatial
distributiontables (3 and 4) upon whichBrumfielbases her
in whichartifactcountsand frequenciesare both
inferences,
presented.The use of frequenciesis quite important,but I am
to see them expressedin terms of prismatic
uncomfortable
blade fragmentsforprismaticcores and comal rims for the
other artifacts.Prismatic blade fragmentsand comal rim
Being
sherdsare quite variablein size and spatial distribution.
breakable,theytend to increasein numbereven withcareful
handling,which means that frequencycalculationsmay not
be duplicablelater. If artifactsmustbe compared,weightis a
more meaningfulattribute.I hope eventuallyarchaeologists
will routinelyweigh-as well as count-excavated materials.

Brumfiel:SPECIALIZATION,

EXCHANGE, AND THE STATE

However,in this instance,whereBrumfielhad exact spatial


control,expressingartifactfrequenciesin termsof a constant
such as area wouldhave been even better.
byJEFFREY R. PARSONS
ofMichigan,Ann Arbor,
University
MuseumofAnthropology,
Mich. 48109, U.S.A. 4 ii 80
Archaeologistshave generallyshown only limitedinterestin
the last fewpre-Hispaniccenturiesin the Valley of Mexico.
This neglectmay have produceda seriousblurringof manyof
our viewsof Aztec society,not the least of whichis the structureof polityand economyat local levels. Brumfiel'sworkat
Huexotla is virtuallythe only systematicattempt to test
archaeologicallysome long-heldpropositionsabout economic
organizationin the Valley of Mexico duringthe threeor four
centuriesimmediatelyprecedingEuropean contact. Her conclusions,whilenecessarilytentativeand hypothetical,constitute a major contribution:they convincinglychallengesome
about specializationand exchange
traditionalinterpretations
someprovocativenewinsightsinto
in Aztec economy;theyoffer
economicstructureand urbanizationin archaic-statesociety;
and theystimulatesomeexcitingnew researchdirections.
The principal weaknesses of Brumfiel'sstudy are easy
enough to identify:the samplingfraction(only 1% of the
ancient urban core at Huexotla) should ideally have been
associationsbetweencategoriesof artifacts
larger;the inferred
and productivefunctions,while reasonable at this stage of
research,are still inadequately demonstrated;chronological
controlsare less than ideal; there is almost no comparable
archaeologicaldata fromotherparts of the Valley of Mexico;
in Aztec
and thereis too littleconcernwiththe ramifications
in Central
timesoftheratherprofoundsocietaltransformations
precedingthe 12thcentury
Mexico duringthe half-millennium
A.D. Nevertheless,she has produced,at low cost and through
nondestructivetechniques(surface collectionas opposed to
excavation),a sound and reasonablestudy whichprovidesa
modelupon whichan expandeddata base can be generatedin
the formof extendedinvestigationsat Huexotla itselfand
comparableresearchat otherknownAztec-periodsites in the
Valley of Mexico.
Brumfiel'sinferencesdovetail nicely with other lines of
evidenceto suggestratherstronglythat the Valley of Mexico,
duringat least the finalcenturyof the pre-Hispanicera, was
composedof twoprincipalsymbioticunits: (1) foodproducers,
the great bulk of whomresidedin provincialtownsand dispersedruralsettlementsthroughoutthe Valley of Mexico and
its immediateenvirons,and (2) others,thegreatbulk ofwhom
If the huge Aztec capital (whichmay
residedat Tenochtitl'an.
have containedas manyas 200,000people and whichproduced
verylittleof its own food) was to be adequatelyprovisioned,
mostof the remainingpopulationin its sustainingarea could
foodproducers.
onlyhave been full-time
It wouldappear that all other15th-and early 16th-century
in the Valley of Mexico weretinyrelative
urbancommunities
to Tenochtitl'an.Texcoco, its closest rival, was probablyless
than a sixthas large,and mostothertownswereprobablyno
morethanhalfthesize ofTexcoco (e.g., Sanders,Parsons,and
Santley 1979). Such a top-heavysystemrendersfutileefforts
economydirectly
to apply classicmodelsofWesternmarketing
to thelate pre-Hispanicperiod(e.g., Smith1979),especiallyin
view of recent archival research which demonstratesthe
rapiditywith whichSpanish economicpatternsand concepts
were imposedin the Valley of Mexico duringthe 1520s and
1530s (Elinor Melville,personalcommunication).I would go
fatherthan Brumfieland argue that mostlate pre-Hispanic
withinthe Valley of Mexico was state-managed,
redistribution
I suspect that what
administereddirectlyby Tenochtitla~n.
471

Vol. 21 * No. 4 * August1980

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

Hernan Cortesand Bemal Diaz del Castillo describedas the


ofsomething
great"market"at Tlatelolcowas a manifestation
more akin to the administeredexchange that characterized
contemporaryAndean society under Inca domination(e.g.,
Murra 1972, LaLone and LaLone 1979) than to the 16thcenturyEuropean markets with which the newly arrived
Spaniardswerefamiliar.
byDAVID A. PETERSON
Department
of Anthropology,
State University
of New York
at Albany,Albany,N.Y. 12222,U.S.A. 3 II 80
On the basis of surfaceevidencefroma singlearcheological
sitein theValleyofMexico,Brumfielbelievesthatan increase
in tributeto a major marketcenterled to increasedexchange
contactwiththatcenter,Tenochtitlan,
and smallersurrounding
centers.The resultofall thiswas thatsmallercommunities
had
less contact with nearbycentersand became agriculturalists
ratherthanspecialistsin othercrafts.Thoughshe doesn'tseem
to argue this,the resultof such a processcould have been the
rise of a primatecenter.Beyond this,I am haunted by the
questionof whetherthe data base supportsher conclusions.
More oftenthan not in Mexico thereare surprisesbelow the
surface,and a combinationof surfaceand excavation data
wouldhave been mostwelcomein thiscase.
I am especiallycuriousabout the roleof salt. The lakeshore
sector of the site could have produced salt. Why would
Huexotla have purchasedit at a distance if it had ample
quantitiesnearby? However,I understandand applaud the
suggestionthat importedproducts at Huexotla were being
exchangedformaize and magueysyrup,since I employeda
site in
similaridea in explainingexchangeat a salt-producing
the Valley of Oaxaca (Peterson 1976). The problem with
to
exchangein the archeologicalrecordis that it is difficult
demonstrate
thatit was bidirectional.
Because ofthedifferential
survivalof artifacts,some exchangemay appear to be uniratherthanmarket-like.
directional,or tribute-like,
As fortheeconomicsinvolved,we shouldnote thata formal
economicsystemwitha militarypresenceis as advantageous
as one without,if not moreso. That the latter is moreprecariousthantheformer
is wellillustratedby thecontemporary
economicsituationsof Japan and the Federal Republic of
whilethe
Germany,whichmustresortto economicenticement
Soviet Union and the United States may use force,or the
threatof force,to get what theywant.
We shouldalso note that if tributewas actuallysold in the
Tenochtitlanarea, thenit was not givenaway. Thus thezerosum utilitysuffered
in obtainingthe tributewas put to a final
marketuse, and thosewho suffered
theloss ofutilitydid so to
the marketbenefitof the populationof Huexotla. If goods
weresold-in otherwords,if therewas competitionforgoods
in theAztecmarketon thepart ofthesurrounding
populations
-how can anyone say therewas no marketcompetitionin
thesystem?
What Brumfielseemsto be sayingis that theAztec had two
choices:to producecraftspecialtiesto exchangeforfoodor to
go out and steal them and exchange them for food. This
ignoresa third option: why didn't they simplyforcelocal
peoples to give themfoodas tributeinstead of goingto such
lengthsto get tributetheycould use to buy foodwith?

and/orthe TripleAlliance.This not only bewildersthe reader


familiarwith the Central Mexican Postclassic,but at times
the
may also have confusedthe authorherself.Furthermore,
startingpointof her "Early Aztec" (A.D. 1150) goes back well
into the timebeforethe Aztec enteredthe scene,approaching
dated timeof Ce Acatl Topiltzin
the somewhathypothetically
of Tula (Davies 1977:464; fordivergingchronologicalplacementsofAztecsubphases,see Charlton1978:1229and Parsons
1974:98). The startingpointofher "Late Aztec" at A.D. 1350
has the same flaw.It was not at this early time that small,
more or less autonomouscity-stateslike Huexotla "were incorporatedinto a large regionalempire"-more precisely,the
Triple Alliance-but only more than half a centurylater.
divisionnecessarilyhas adverse
Such a distortedchronological
of the data.
effects
on theinterpretation
The primaryaim of the paper is to test a considerably
versionof the hypothesisconcerningthe interaction
simplified
betweenorganizedexchangeof goods and productsand the
developmentof states and political and social institutions.
Brumfiel'sconclusionthat thishypothesisis not supportedby
herarchaeologicaldata had alreadybeen anticipatedby Flannery(1972:407), and in thisrespectone may wonderwhyshe
made it thefocusof herstudy.
of economicconditionsin
Her archaeologicalinterpretations
with
theValleyofMexicoin late pre-Hispanictimescorrespond
sourcesand the many
the details knownfromethnohistorical
does not,however,
studiesbased on them.This correspondence
because they are
serve to corroborateher interpretations,
fromwrittensources.
based fromthebeginningon information
Thereare also someweak pointsin details.For example,her
notionthat "the processing... of raw cottonmighthave supported a numberof urban residents"cannot be accepted.
Spinningand weavingweregenerallywomen'swork(Sahagiun
1959[1577],Book 10:51-52). Theremayhave beensome intenbut specializationin the senseof moreor less exclusification,
sive occupationwithspinningand weavingwas absent.
In summary,thispaper showsthat duringthelast fource nturiesbeforethe conquest,populationand exchangeincreased
in Huexotla and commercialand politicaltieswiththe centers
of political power were intensified-notmuch more, a nd
nothingnew.

byBARBARAJ. PRICE
New York,N.Y., U.S.A. 2 II 80
Of the many debatable issues raised in this paper, these
remarkscan address only a few. One wondersinitiallythat
thereseemsno consistentmodel,implicitorexplicit,ofexpected
in a
scale or patterningof internalsocioeconomicdifferences
communityof the size, density,and political status of a
Huexotla. Neither the range of variation observed in contemporarymunicipiocabeceras-the most probable counterpart of such communities-nor the ample ethnohistoric
documentationof, e.g., Huexotzinco (Dyckerhoffand Prem
1976) seems to have influencedthis attemptat retrodiction.
Withoutany stated expectations,there is no possibilityof
relevant
any,norcan an appropriateor demonstrably
falsifying
investigativestrategybe developed. It is hardly surprising,
that the methodology
employedis incapableof distherefore,
tinguishingamong possible or probable options. Sampling
techniques,howeverfashionable,are unproductivewhen apof only300 ha., are unlikelyto reveal the
plied to a settlement
byHANNS J. PREM
most probable extentand organizationof specializationin a
Institut fur Vo5lkerkunde
und Afrikanistik,Universitdt small ruraltownsuch as this,and represent,
moreover,merely
Miinchen,Schellingstrasse
33, Munich 40, Federal Republic the investmentof methodologyforwhat seems to be its own
ofGermany.
8 II 80
obsersake. Finally,I questionthe legitimacyof extrapolating
To beginwith,a remarkon terminology
seems necessary.In
vations taken in a politicallydependentrural backwaterto
Brumfiel'spaper the term"Aztec" has severaldifferent
and
meanrevealprocessesofproductionand exchange,centralization
ings.It is used to designatea long chronological
periodas well
power in an imperialpolitical economy.Treatmentof these
as a culturaland politicalaffiliation
withthe Mexican capital
throughoutthe discussionand
questionsis self-contradictory
472

CURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

ignores the most powerfulexplanationof the dynamicsof


expansionof the Aztec state.
Two conditions,capable of archaeologicalrecognition,
could
singly or in combinationgeneratea predictionof densely
settledwards of craftsmen:productionforexportoutsidethe
community(Jacobs 1969) and/ora large,dense,permanently
residenturbanpopulation."Lateral exchange"as used in this
paper can referto eitherconditionand thus give rise to two
distinctmodels.In theformer(cf.McBryde1947),everyonein
a communitywitha nonsubsistence
specialty,agriculturalor
has the same one; thispattern,intensified
nonagricultural,
by
land pressuresand increasinginabilityof subsistenceproducersto breakeven,entailslittleexchangewithinthe village,
in a regional,intercommunity
resulting
symbiosis.Even surface
samplingcould easily reveal thispatternif enoughindividual
componentswere investigated:each village will yield an assemblagedistinctiveto its exportwork.The degreeto which
exchangesare egalitarianor are hierarchially
organizedshould
be reflectedin the relativesize and complexityof the settlementcomponentsof the system.For Huexotla theprobability
of specialized exportproductionis low; the communityexploiteda varietyoflands withdifferent
agriculturalpotentials
(includingirrigatedgrainland), and its residentswere probably generalizedproducersof subsistencestaples in quantities
above immediatesubsistenceneeds. Such a communitydoes
not normallysupporta large,diverse,regularmarket.
"Lateral exchange"of a second type,urban in formif not
necessarily in scale, is predictably weakly developed in
Huexotla,as it is todayin cabecerasservingresidentand surroundingpopulationsoffarmers
largelyself-sufficient
in staples
and havinglow purchasingpower.This secondmodelindicates
thedegreeto whichcommunity
residentsare socioeconomically
differentiated
and the communityinternallysymbiotic.Its
materialexpressiontakes the formof the rowof small tiendas
at or adjacent to the townplaza, perhaps with small miscelaneas at barrionodes. The butcher,baker,and candlestickmaker are all cheek-by-jowlon the same block; because
demand is low thereis probablyonly one of each. Although
this is the moreprobablemodel,Brumfiel'ssamplingmethod
is likelyto overlookmuch of whateverspecializationof this
typeis present,and seemsindeedto have doneso. Insufficiently
refinedin scale, it does not even directconcentratedattention
to what the model indicatesas the most probable parts of
town.
Because thispaper showsonly the mostnaive grasp of the
relationof basic economicsto marketcompositionand structure (principleswell knownto Adam Smith),it cannotassess
eitherthe benefitsconferred
by marketsor the costs incurred
by them,cannot evaluate the probabilitiesof alternativeor
competingoptions. The technoeconomicand demographic
context,includingchangingman/resourceratios, acts as a
majordeterminant
ofrelativeefficiency
ofone or anotherinstitutionalpatternbut is not systemicallyrelatedin this treatment to problemsof productionand distribution,degree of
marketdependence,marketsize, or extentand organizationof
local specialization.At both inter-and intracommunity
levels
the numberof specialiststhat can be supportedis limitedby
the caloric surplusesproducedby those who patronizethem
and by the extentof demandfortheirservices.Huexotla was
situatedon the veritabledoorstepof Texcoco, a large urban
centerin whicha daily marketconcentrated
large numbersof
producersand consumersand could easily underselleithera
rotatingcircuitof rural marketsor a restricted,lower-order
local market. Dependence of Huexotla residentsupon the
Texcoco market,givenits loweroverheadin relationto volume
of businessand the low transportcosts,is cost-effective
and
should inhibit-as it is here observedto do- developmentof
local exchangein subordinatecommunities.Unless archaeological communities
are exemptedfromthe operationof these
systemicprocesses,it mighthave been preferableto devise a

Brumfiel:
SPECIALIZATION,

EXCHANGE, AND THE STATE

researchstrategysomewhatmoresensitiveto theprobabilities
theydetermine.
in population numbersand distribution
Cited differences
betweenEarly and Late Aztec-the materialindicatorsof an
alteredpoliticaleconomy-are moreparsimoniously
explained
withreference
not to trade but to the developingshiftin the
mode of productionto chinampa cultivation.Tenochtitlan's
geopoliticalposition,admirablysuitedto controlling
the zone
in which this reclamationrequiredleast labor per unit of
returnand in whichthe shiftthus took place earliest,underwroteits regionaland supraregional
dominance.This last is recontrolnotonlyoftradeat all levels,but
flectedin itsincreasing
is equivprobablyof muchoftherestof theeconomy.Brumfiel
of theAztec Emocal concerningthe degreeof centralization
pire withinits own heartland;by the end it was probablyfar
higherthansheseemsto credit.But agriculture,
nottrade,continuedto accountforthebulkofthegrossnationalproductand
therefore
explainsthe abilityof the state to move,withlittle
into increasingcontrolof
competitionfromotherinstitutions,
sectorsas trade(cf.Wittfogel
suchancillaryrevenue-producing
1959).The energy-richer
activitygovernsthepatterningofthe
one.
energy-poorer
Steward (1950) criticizedthe "microcosm"view of the
peasant communityin the wider, stratifiedsociety-the
tendencyto treat local units as though they were isolated
tribeswith institutionsresponsiveto purelylocal conditions.
Reiteratingthese strictures,Wolf (1956) suggestedthat the
at thebottomofan economic
articulationofsucha community
is a majordeterminant
and politicalhierarchy
ofitsway oflife.
The obverseis that processesat the nationalor imperiallevel
at the local levelmay be only palely or obliquelyreflected
may be unrecoverablevia the worm's-eyeview. Hardy's
is hardlyrepresentative
of theinstitutional
Casterbridge
range
or structureof VictorianBritain.Justas the nationalpolitical
economyis not the village communitywritlarge,the village
community
or rural small town is not the national political
economywritsmall.
byFRANCES ROTHSTEIN
Departmentof Sociologyand Anthropology,
Towson State
University,
Towson,Md. 21204, U.S.A. 6 II 80
Not beingan archeologist,
I was reluctantat firstto comment
on Brumfiel'sarticle. I cannot evaluate its methodologyor
how it fitswithrelatedarcheologicalstudies.I am, however,
enthusiasticabout its findingsand the relevance of such
concerns.It
archeologicalresearchto broaderanthropological
is, in fact,because of the significant
implicationsof the paper
that I mustraise two criticalcomments.First,I would have
liked morediscussionof how Brumfiel'sfindingsof increasing
inequalityrelate to otherrecentarcheologicalstudiesof the
state. Second,I wouldhave likedto see herfollowthroughon
theimplications
ofherdata, and I thinkshe has failedto do so
because she implicitlyadopts a view of the state that stresses
benefitsratherthanclass and conflict.
Brumfiel'sanalysis shows that political centralizationand
increasingspecializationin the Aztec empirewere associated
withan increasein asymmetrical
relationsbothwithinHuexotla
and in the region.She recognizesthat herdata raise questions
about the viewthat the state emergedbecause ofits functions
in stabilizingregionalsystemsof economicspecializationand
exchange,a variationof integrativeor benefittheoriesof the
state. She concludes,however,that the integrativehypothesis
does notapplybecausepremodern
statesdid notoperateunder
free-market
conditions.She neglectsto mentionWallerstein's
(1979:134) argumentthat modernstates do not operatein a
freemarketeitherbecause the factorsof productionare only
partiallyfree.Had she begunwithstratification
as the central

Vol. 21 * No. 4 * August1980

473

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

tribute,and redistributedto political clients,who in turn


homoexchangedsome of themforfoodwithan increasingly
geneousruralpopulation.
None of these conclusionsis warranted,certainlynot by
ethnohistoric
data and probablynot even by herown archaeological data. First, with respect to the homogeneityof the
Huexotla population,I doubt that Brumfiel'ssamplingwas
adequate to documentfullythepresenceofnon-food-producing
specialistsin the urban core,consideringthe dispersednature
of craftactivities(see Dyckerhoff
and Prem 1976 withrespect
toHuexotzinco).In fact,evenwithan inadequatesampleshehas
but she choosesto downmanagedto discernthispatterning,
byWILLIAM T. SANDERS
My hunchis thata carefulvisual survey
Department
of Anthropology,
PennsylvaniaState University, gradeits significance.
of each gridunit would have revealedmuchmoreevidenceof
University
Park, Pa. 16802,U.S.A. 14 ii 80
specializedactivitiesthana 1% surfacesample,no matterhow
My majorcriticismof thispaper is that statisticalsamplingis
between
rigorouslyconducted.With respectto differentiation
uselessas a methodofanalysisunlessone has a well-conceived
urbanand ruraland amongruralcomponents,I am surprised
modelto test. In the case of specializationof productionand
thathersamplingstrategies
have revealedas muchas theyhave.
distributionof goods in Aztec times, there is a wealth of
She findsspecializationin the lakeshoresector,as the 16thethnohistoric
data fromwhichone could designvery specific
centurydocumentsindicate,but she considersit insignificant
models(a resourcetheauthordoes use but in a highlyselective
ofthepopulation."
because"it engagedonlya smallproportion
manner). Specifically,Brumfielattemptsto test, using priMinimally,a modelof the consumptionof lakeshoreproducts
marilyarchaeologicaldata, my modelof the economyof the
by the inlandpopulationof the Huexotla domainmustbe deBasin of Mexico in 1519, whichwas reconstructed
primarily
veloped beforeone can determinewhetherthe proportionof
fromethnohistoricdata. The model includes the following
or not.
the populationengagedin providingthemis significant
components:
The factthatthedifference
betweenupperpiedmontand lower
1. Tenochtitlanand Texcoco were top-levelmarketcenters
piedmontfarmersis quantitativeratherthan qualitativeis not
that providedcraftgoods and food forthe urban population
surprising.Even withher small sample,she has detectedthe
residentthereas well as forthe restof the populationof the
I suggestedon thebasis ofthekindsofspecialization
patterning
Basin.
foundin the area in the 16thcenturyand today.
2. At some(in myfirstformulation
all) politicallydependent
Withrespectto regionalspecialization,Huexotla was a poor
smallertowns(whichhad been centersof independentstates
choicefora testofmymodel-thereis no regionalspecialization
prior to theirconquest) therewere lower-levelmarketsthat
references
for
in thearea today,and I knowofno ethnohistoric
rural
had twofunctions-tosupplytheirimmediatesupporting
the 16thcentury.I agree withBrumfiel'sconclusionthat this
populationwith food and craftgoods and, in some cases, to
mymodel,
and, following
was an area ofbasic farmproduction,
providelargerareas with the productsof unusual local spedocuthis was Huexotla's specialty.Aside fromthe definitive
cialization(forexample,thedog marketat Acolman).
mentaryevidenceofsuchspecialization,thereare also archaeo3. Tenochtitlanand Texcoco wereoccupiedalmostentirely
logical data. For example,we located a continuousband of
by non-food-producing
political,religious,and economicspesalt-makersettlementsfromEcatepec to Tenayuca-a lakecialists,the last of whomproducedgoods forexchangein the
shorestrip20 km long; thereare severalobsidianmanufacturmarket.
above Otumba and around Pachuca, several
ing communities
mostor all) of the
4. A minority(in my firstformulation
communitiesin the northwestcornerof the
lime-processing
residentsof the small townswerenon-food-producers.
near Culhuacan
Basin, and numerouschinampacommunities
5. Because of the greatenvironmental
varietyin the Basin
that,fromthe small size of the plots,seem not to have been
and thetendencyofsomeresourcesto be highlylocalized,there
sites (documentarydata confirmthat
basic grain-producing
was a strongtendencytowardregionalpart-timespecialization.
communities).
theyweretruck-garden
Much of this specialist productionundoubtedlycirculated
Finally, Brumfiel'sreconstructionof the nature of the
throughthe two higher-order
markets,but it probablymoved
Tlatelolco market at Tenochtitlanis justifiedneither by
fromone small marketto anotheras well (as it does today in
data. I do not doubtthata
archaeologicalnorby ethnohistoric
highlandmarketsystems).
portionof the imperialtributefoundits way intothe market6. In the two major centers,craftsmenof the same trade
place, but the vast majorityof the goods were,in fact,proresidedin particularbarrios;in the smallercenterstheywere
duced by urban craftsmenor regionalpart-timespecialists.
eitherdispersedor groupedin lineage-sizebarrios.
Many basic goods sold therewere eithernot collectedas imBrumfielattemptsto test theseconclusionsby an intensive perial tribute(such as obsidian,groundstone tools, wooden
surveyand rigoroussurfacesamplingof the small town of
farmimplements)or quantitativelyinsignificant
(such as potHuexotla and rural sites withinits domain. On the basis of
teryand mats). Documentaryevidenceof urbancraftspecialthesedata and, as I have said, a highlyselectiveuse of docuizationis abundant (see Calnek 1978,Carrasco 1978). On the
mentarysources,she comesto the following
conclusions:
archaeologicalside,does the authorbelieveAztec potterywas
1. There was littleeconomicspecializationand fewmarket producedoutsidethe Basin of Mexico?
functions
at Huexotla,and at smallAztectownsgenerally,
and
In summary,this paper is not very convincingeven when
the residentialpopulation was relativelyhomogeneousboth
dealingwiththepoliticaldomainofHuexotla,and its attempt
withinthe townand withinthepoliticaldomain.
to reconstruct
theeconomyoftheBasin ofMexico on thebasis
2. There was not much regionalspecializationwithinthe
of the Huexotla situationalone has little merit.Like many
Basin as a whole.(Her statementson thispointare notentirely archaeologists,
Brumfielfails to recognizefullythe equivocal
consistent.)
natureof much archaeologicalevidence;she rarelyor at best
3. Even at Tenochtitlan,craftspecializationwas of little
inconsistently
uses and does not fullyappreciatethe value of
significance,
and the bulk of the goods foundin the market- ethnohistorical
sources as a means of testingarchaeological
place were manufacturedoutside of the Basin, collectedas
data and as a sourceof models.
featureof the state,whethermoderncapitalistor premodern,
she mighthave examinedmorefullytheincreasingasymmetry
of specializationand trade.For example,one wonderswhether
theurbanzone in Huexotlain theLate Aztecperiodhad more
nonlocalgoodsthantheotherzonesand whetherthedisparity,
if any, increasedbetweenEarly and Late Aztec times.More
systematicattentionto inequalitiescould have given further
supportto hercontentionthat the residentsof the urban zone
in Huexotlawerean urbaneliteratherthancraftspecialists.

474

CURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

Brumfiel:
SPECIALIZATION,

Reply
byELIZABETH M. BRUMFIEL
Albion,Mich., U.S.A. 17 III 80
The responsesto my paper are quite diverse.I am glad to
acquirereferences
to recentworksin Mesoamericaneconomics
by Carrasco, Gordon,and Peterson,and I am pleased that
Brown and Rothsteinfind the Huexotla data pertinentto
theoreticalissues beyondthe immediatescope of my analysis
-substantivist vs. formalist
approachesto precapitalisteconomies,on the one hand,and benefitvs. conflictapproachesto
nonegalitarianpolities,on the other.I findmyselfstraddling
the fenceon bothissues,not fromfearof alienatingone camp
or the other(mypaper seemsto have smoothedfewfeathers),
but because each side appears to have settled for partial
truths.My emphasisupon theactive role that state expansion
played in the intensification
of the Valley of Mexico regional
economyplaces me in basic sympathywith substantivist
economics,yet once the institutional
contextof an economyis
definedand once the goals of participantsare understoodwe
can probablyexpectto findparticipantspursuingtheirgoals
accordingto somesortofleast-cost/maximum-return
principle;
hence,the mix of formalexchangetheoryand substantivism
observedby Dillehay. Similarly,althoughI sympathizewith
the conflictapproachto nonegalitarianpoliticalsystems,I do
notbelievethatthisexcludesconsideration
of thebenefitsthat
mightbe gained by elite administration.
As Price (1979:300)
veryaptly observes,"For the elite the questionis essentially
one ofstrategy:Whichcombinationofforceand benefits(both
of which,of course, cost something)will have the highest
payoffand underwhat circumstances?"Considerationof this
questionmightprovidean answer to Peterson'squery as to
whyTriple Alliancerulersdidn'tsimplyforcelocal peoples to
give themfoodas tributeinsteadof goingto such lengthsto
get tributethey could use to buy food with. Perhaps the
potentialcostsof a tax revoltin the TripleAllianceheartland
led by disaffected
local elites but mannedby oppressedcommonerswere great enough to discourageimplementation
of
this ratherstraightforward
option. These questionsaside, I
wouldlike to devotetherestofmyreplyto a discussionoftwo
issues:thevalidityoftheHuexotladata as a sourceofinformation about the Late Postclassic economyin the Valley of
Mexicoand theadequacy of mymodelof theeconomyin light
of theadditionalinformation
and alternativeformulations
presentedby therespondents.
Several respondentshave questionedthe adequacy of the
sample taken fromHuexotla,particularlyits abilityto gauge
the intensityof craftspecializationwithinthe city-state.One
questionposedby Mason and Lewarchconcernsthepossibility
of my having missed evidence of craftspecializationin the
rural areas because "no samples were taken betweenhouses
in theruralarea." Evidently,thereis a misunderstanding
about
mysamplingprocedure.In theruralareas, everyconcentration
of prehistoric
debriswas sampled,withoutpriorevaluationof
whetherthese concentrations
representedworkshopsor residences. If workshopdebrishad been presentin surfaceconcentrationslocated away fromthe individualresidences,those
concentrations
wouldhave been sampled,and noneof the collectionsfromtheruralareas indicatetheexistenceofworkshop
locales.
Charlton,Mason and Lewarch,Parsons,Price,and Sanders
all question whethera 1% sample drawn fromHuexotla's
urbancore was largeenoughto providean accuratepictureof
the intensityof craftspecializationby the town's residents.
For many typesof statisticalquestions,it is the absolutesize
of the sample drawnfroma populationratherthan the perofthepopulationsampledthatdetermines
theaccuracy
centaige
of statisticalinference,
and the questionof whatproportionof

EXCHANGE, AND THE STATE

urbanhouseholdsengagedin any typeofspecializedproduction


is a case in point.Each of the 560 unitscollectedin the urban
core can be viewedas an independenttrialforthe presenceof
specializationofsometype.None ofthesecollectionsproduced
the superabundanceof tools or waste materialto be expected
visiblecraftsas potfromthepracticeofsucharchaeologically
tery making (misfiredsherds), stonecutting,carpentry,and
woodcutting(groundstone chisels,adzes, axes), salt making,
cloth production,obsidian working,and the productionof
sherds,spindle
mold-madeceramic artifacts(fabric-marked
whorls,obsidian debris,and ceramicmolds were all present,
see Brumfiel[1976:chap. 5]
but neverin veryhighfrequencies;
of artifactclasses withinsite
fordiscussionof the distribution
sectors).Even if we assume that only 100 of our trialswere
"fair" (in the sense that artifactabundancewas greatenough
to have provideda chance for very high frequenciesto be
observedand that the distancebetweencollectionunits was
great enoughto insurethat all of the collectionswere taken
households),it can be stated at a betterthan
fromdifferent
that fewerthan 5% of the households
99% level of confidence
wereengagedin such activities(calculatedfroma cumulative
This is entirelyconsistentwithmystatePoissondistribution).
mentthat specializationengaged"only a small proportionof
the city-state'spopulation."
Charltonreportsthat he and Fletcherlocated a heavy concentrationof spindlewhorlsin Huexotla's urban core which
my samplingprocedurefailed to detect.My firstreactionto
this is to wonder about Charlton and Fletcher's sampling
"strategy."It is possible that, having found two or three
spindlewhorlsin a relativelysmallarea, Charltonand Fletcher
became sensitizedto the presenceof spindle whorlsin that
area and wereable to collectquitea numberofthem.Had they
stumbledacross two or threespindlewhorlselsewhereon the
as a workshoplocale.
site,this too mighthave been identified
On the otherhand, I wouldnot rule out the possibilityof my
havingfailedto detect the presenceof some workshopareas.
My data permitme to inferonlythat craftspecializationwas
relativelyrare,not that it was absent,and thisis what I have
said.
Both Carrascoand Mason and Lewarchpointout thatmany
typesof craftspracticedby the pre-Hispanicpeoples of highland Central Mexico would not be representedin materials
collectedfroma site surface.They are of coursecorrect,and I
have posited the existenceof matmaking,basketry,fishing,
and fowlingin Huexotla's lakeshoresectorin the absence of
finestoneworking,
goldanypositiveevidence.Featherworking,
and paintingprobablywerecarriedout at Huexotla
smithing,
under the patronageof the town's elite. But since the conto memsumptionof thisclass of goodstendedto be restricted
bersof theelitestratum,thiskindofspecializationwouldprovide little basis for reciprocalexchangeand economicinterdependenceamong large segmentsof Huexotla's population.
Such specialistsmightalso have providedHuexotla's resident
populationofmerchantswithgoodsto be exchangedin distant
marketplaces,but both Acosta Saignes (1945) and Berdan
(1975) notethatthegoodscarriedback to theValleyof Mexico
raw materialsand
fromdistantmarketswere predominantly
by the elite.Thus,
productsdestinedforconsumption
finished
fine
again, it would seem that the presenceof featherworkers,
goldsmiths,etc., would not imply dependence
stoneworkers,
upon internalexchangeby large segmentsof the city-state's
population.
If the archaeologicaldata do providean adequate measure
of internal specializationand exchange at Huexotla, how
are thesedata? Was Huexotla an appropriatesite
meaningful
Charltonsugforresearch,giventhegoals oftheinvestigation?

Vol. 21 * No. 4 * August1980

475

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

gests that the paucity of evidencefora complexdivisionof


labor duringEarly Aztec times mightbe a consequenceof
Huexotla's early incorporation
into Texcoco's sphereof influence,makingit atypicalofthevalley'searlycity-states.
I reject
this notionbecause the ethnohistorical
literatureclearlyindicates that Huexotla was a more powerfulcommunitythan
Texcoco until sometimeafter Techotlalatzin'ssuccessionto
rulein Texcoco,ca. A.D. 1332(Chimalpahin1965[ca. 1606-31]:
78). This is indicatedby Sahag(un's(1959[1577],Book 8:14)
referenceto Huexotla as the "seat" of Acolhua rule,by the
territorial
extentof Huexotla's politicaldominance(Guzman
1938:91-92),and by the statusof Huexotla'srulinglineageas
"wife-givers"to both Texcoco and Coatlichan (Ixtlilxochitl
1952[ca. 1600-40];see Carrasco1974bforthepoliticalimplicationsof marriagewithinthe Acolhua domainat a somewhat
later date). If thepowerof rulersin earlycity-stateswas sustainedby theiradministration
of complexlocal economies,the
local economyat Huexotlashouldhave been more(ratherthan
less) complexthanthenormforValleyofMexico communities
duringEarly Aztectimes.
Price questions the legitimacyof using the data froma
"politicallydependentruralbackwater"such as Huexotla to
make any inferencesat all about economicprocessesin the
Valley of Mexico duringLate Aztec times.I believe this endeavoris legitimate.Huexotla's economyclearlydoes notprovide a view of the Aztec imperialeconomyin "microcosm,"
but it doesprovidea way ofgaugingtheimpactofhigher-order
economicand politicalprocessesupon the local componentsof
a regionalsystem.Many discussionsof the originsand persistenceofthesehigher-order
processeshave focusedupon their
consequencesforlocal populations.Therefore,it seems both
appropriateand necessaryto focusupon one or several such
unitsand attemptto gatherdata that permitus to verifythe
presenceor absenceofthesepostulatedconsequences.
Charlton,Dillehay,Gordon,Price,and Sandersall pointto
the geographicvariables that make Huexotla different
from
some,but by no meansall, of the secondaryand tertiarycommunitiesin the Valley of Mexico. It is certainlytrue that
Huexotla was situated on the doorstepof Texcoco, a large
urbansettlement
and an importantmarketcenterduringLate
Aztec times.It is also true that Huexotla was endowedwith
unusuallyrichagricultural
lands.This does not,however,automaticallyinvalidate the conclusionswhich mightbe drawn
fromthe Huexotla data concerningthe structureand consequences of the Late Aztec regionalsystem,forHuexotla was
a part of that system.What we need at this point are additionalstudiesofsecondaryand tertiary
Late Azteccommunities
whoseattributescontrastwiththoseof Huexotla so that the
inferencesdrawn from the Huexotla study can be tested.
Charlton'sreferenceto Otumba suggeststhat he has completedone such study,and I urgehim to presenthis findings
in publishedform.
Turningfromissuesof methodology
to issues of interpretation, I have no argumentwith the model of the Valley of
Mexico economyin 1519 presentedhere by Sanders (though
I would want to add to it; see below). However,Sandershas
not accuratelypresentedmyownposition:
1. I do argue that therewas littleinternalspecializationat
Huexotla, and at small valley townsgenerally,but I do not
argue that marketfunctionsat thesetownswerefewand unimportant.The abundance of nonlocalproductsin the Huexotla collectionssuggestsdependenceupon regionalexchange
duringbothEarly and Late Aztec times,and I explicitlystate
mybeliefthatmarketing
servedas themechanismofexchange.
2. I do notarguethattherewas notmuchregionalspecialization withinthe valley as a whole.Rather,I argue that some
typesof regionalspecializationwouldhave declinedin relative
importanceas tributeextractionfromremoteprovincesprovided an alternativeto productionwithinthe valley. I base
this argumentupon the reducedimportanceof Otumba as a
476

sourceof obsidianforthe Late Aztec inhabitantsof Huexotla


(obsidianwas an itemof tribute;see Relacion 1949[1580]),as
wellas thedeclinein clothproductionat Huexotla.Othertypes
fromthe
of regionalspecializationwhichmighthave suffered
matflowoftributeintothevalleyincludeceramicproduction,
making,and perhaps the collectionof firewoodand pine
torches(all itemsoftributeformorethanhalfofthe36 tributepaying provinces recorded in Informaci6n1957[1554],but
omittedfromthe "Matricula de Tributos" [Barlow 1949]
probablybecause the Spaniardshad littleinterestin goods of
low per itemvalue). There is no doubt that I wentoverboard
variation
in statingthat eventuallyall kindsof environmental
except those affectingfood productionwould have been ignored.Many regionalspecializations,such as lime processing
and mano-metatemaking,would have persisted,since goods
ofthesekindswerenotbeingreplacedby itemsoftribute.However,the Otumbaobsidiandata make tenablean argumentof
less intensiveregionalcraftspecialization(did Otumba also
enjoy a competitiveadvantage in agriculturalproduction?),
testing,I wouldcontinueto argue that
and, subjectto further
of marketexchangewas not a consequence
the intensification
of the elaborationof an existingsystemof regionalspecialization.
3. I do notarguethatthebulk (i.e., majority)ofgoodsin the
outside
Tlatelolco marketat Tenochtitlanweremanufactured
the valley. I argue that Late Aztec marketexchangein the
by the exchangeof tributeforfoodbevalley was dominated
tween urban and rural populations-dominatedin the sense
thatthistypeofexchangewas decisivein termsofthestructure
of the marketsystem(the predominanceof urban-ruralexchangeover exchangebetweenruralcenters)and in termsof
the impact of marketexchangeupon regionalproduction.I
wouldalso arguethatfewofthegoodspurchasedby ruralfoodproducersin the Tlatelolco marketwereproducedby the resident craftspecialistsof Tenochtitlanbecause most of these
specialistsseem to have been engagedin the production of
luxurygoods,consumedby thepoliticaleliteand used as items
of exchangeby the pochtleca
(see Monzon 1949:50-51 fora list
of thekindsof specialistsresidingin Tenochtitlan).
sectionsof my
and interpretive
4. Since both introductory
to ethnohistorical
sources,
paper containnumerousreferences
I fail to see why Sanders concludesthat I make little use
literature.
of and do not fullyappreciatethe ethnohistorical
WhileI wouldnot deny that archaeologicalevidenceis someevidence
timesequivocal,I wouldpointout thatethnohistorical
sources
containslimitationsof its own. First,the ethnohistoric
do not providefullcoverageof the entirerangeof topicsthat
Second, most of the literature
may interestanthropologists.
was generated30 to 100 yearsafterconquestand may or may
not accurately describe conditions that obtained in preHispanictimes.Third,muchoftheliteraturewas producedby
Europeans(or nativeseducatedin theEuropeantradition)and
was intendedto communicatewithEuropeansand may,thereor interpretations
of native
fore,providedistorteddescriptions
culturalinstitutions.Clearly,both archaeologicaland ethnohistoricaldata are neededto generateaccuratemodelsof preHispanic cultures.If both typesof data are used, the distortionsand gaps in one typecan be correctedand supplemented
by the other.The Huexotla studywas conductedwith these
in mind.
considerations
In general,I am disappointedthat Sanders discussesthe
valley'seconomyin suchstaticterms,becausemypaperis concernedwitheconomicchange and its relationshipto ongoing
politicalprocesses.However,I would findSanders'smodelof
the Valley of Mexico economyin 1519 morecompleteif two
pointswereadded:
7. Few of the goods producedby the residentcraftspecialists of Tenochtitlanand Texcoco wereintendedforsale to the
populace of thevalley.
food-producing
8. The status of Tenochtitlanand Texcoco as top-level
CURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

marketcenterswas a consequenceof theirpoliticaldominance


of theValleyofMexico and a broad regionbeyond;it was not
a consequenceof theirimportanceas centersof production.
Gordonand Price observethat the importanceof Tenochtitlanand Texcoco in the marketsystemwas a functionof their
size and prosperity.
They also arguethat thesetwo citieswere
large and prosperousbecause they were able to capture and
resourcesin theirimmediatehinterlands.
develop energy-rich
These statementsare certainlytrue, at least for the initial
stages of urban growth,and they should have been given
greateremphasisin my paper. However,theydo not explain
how Tenochtitlanwas able to supportitselfonce its growing
population requiredmore food than could be gained from
tributeand landed estates. Nor do they account for the increased importanceof Cempoala-Pachuca as a supplier of
obsidian to Huexotla and the decreased importanceof the
Otumba source. Far fromprovidinga moreparsimoniousexplanationof thesetwo phenomena,an exclusiveappeal to the
developmentof chinampacultivationprovidesno explanation
at all. But in general,I findtheirdiscussionscompatiblewith
my own line of argument-urbanization
servedpoliticalends,
it was accomplishedby politicalmeans,and it had a decisive
impactupon the structureand intensityof marketingwithin
theValleyof Mexico duringLate Aztec times.

References
Cited
ACOSTA SAIGNES, MIGUEL. 1945. Los Pochteca.Acta Antropol6gica1

(1).

[PC, RC]

ALDEN, JOHN R. 1979. "A reconstruction


of Toltec period political

units in the Valley of Mexico," in Transformations:


Mathematical
approachesto culturechange. Edited bv C. Renfrewand K. L.
Cooke, pp. 169-200. New York: Academic Press.
[THC]
ARMILLAS, PEDRO. 1971. Gardens in swamps. Science 174:653-61.
[CLG]
BARLOW, ROBERT H. 1949. The extentof the empireof the Culhua
Mexica. Ibero-Americana28.
BERDAN, FRANCES F. 1975. Trade, tribute,and marketin the Aztec
Empire. UnpublishedPh.D. dissertation,Universityof Texas at
Austin,Austin,Tex.
--.
1977. "Distributive mechanismsin the Aztec economy,"in
Peasant livelihood.Edited by R. Halperin and J. Dow, pp. 91-101.
New York: St. Martin's Press.
BRUMFIEL, ELIZABETH M. 1976. Specialization and exchangeat the
Late Postclassic (Aztec) communityof Huexotla, Mexico. UnpublishedPh.D. dissertation,Universityof Michigan,Ann Arbor,
Mich.
CALNEK, EDWARD E. 1972. Settlementpatternand chinampa agricultureat Tenochtitlan.AmericanAntiquity37:104-15.
1973. The localization of the 16th-century
--.
map called the
Maguey Plan. AmericanAntiquity38:190-95.
1975." Organizaci6nde los sistemasde abastecimientourbano
de alimentos:El caso de Tenochtitlan,"in Las ciudades
deAmerica
Latina y sus areas de inflencia a travesde la historia.Edited by
J. E. Hardoy and R. P. Schaedel, pp. 41-60. Buenos Aires:
Ediciones A.I.A.P.
1978." El sistemade mercadode Tenochtitlan,"
in
Econom,a
politicay ideologiaen el Mexico prehispanico.
Edited by P. Carrasco
and J. Broda. Mexico: Nueva Imagen.
[WTS]
CARRASCO, PEDRO. 1974a. "Introduccion La Matricula de Huexotzinco como fuente sociol6oica," in Matricula de Huexotzinco.
Edited by Hanns J. Prem, pp. 1-16. Graz: AkademischeDrucku. Verlagsanstalt. [PC]
1974b. Sucecion y alianzas matrimonialesen la dinastia
Teotihuacana. Estudiosde CulturaNahuatl 11:235-41.
1978. "La economiadel M'xico prehispanico,"in Economia
politica e ideologiaen el Mexico prehispdnico.Edited by P. Carrasco and J. Broda, pp. 15-76. Mexico: Nueva Imagen.
[PC,
CLG, RDM, DEL]
CARRASCO,PEDRO, and JOHANNABRODA. Editors. 1976.Estratif
caci6n
social en la Mesoamerica.Mexico: Nueva Imagen.
[WTS]
--.
Editors. 1978. Economia politica y ideologia en el Mexico
prehispdnico.
Mexico. Nueva Imagen.
[WTS]
"Carta de los caciques e'indiosnaturalesde Suchimilcoa Su Magestad
... (2 de Mayo de 1563)." 1870 (1563). In Coiecci6nde documentos
ineditos... de Indias, vol. 13, pp. 293-301. Madrid: Jose Maria
Perez.
CHADWICK, ROBERT E. L. 1966. The " Olmeca-Xicallanca" of

Brumfiel:
SPECIALIZATION,

EXCHANGE, AND THE STATE

Teotihuacan: A preliminarystudy.MesoamericanNotes7-8:1-23.
[RC]
H. 1971. Informesobre trabajos del laboratorio,
CHARLTON, THOMAS
enero a mayo, 1971. MS, Instituto Nacional de Antropologiae
[THC]
Historia,Mexico.
-. 1978. Teotihuacan, Tepeapulco, and obsidian exploitation.
[HJP]
Science200:1227-36.

CHIMALPAHIN QUAUHTLEHUANITZIN,DOMINGO FRANCISCO DE SAN


ANTON MUNON.1965 (ca. 1606-31). Relacionesoriginalesde Chalco

Amaquemecan.Translated by S. Rend6n. Mexico, D.F.: Fondo de


Cultura Econ6mica.
of
CLINE, HOWARDF. 1972. "A census of the Relaciones
geograficas
New Spain, 1579-1612,"in HandbookofMiddle AmericanIndians,
vol. 12. Edited by R. Wauchope,pp. 324-69. Austin:Universityof
Texas Press.
American
COE, MICHAEL.1964. The chinampasof Mexico. Scientific
[CLG]
211(1): 90-98.
CORTES, H. 1962 (1519-26). ConquestdispatchesfromCortesin the
New World.Translated by J. R. Blacker and H. M. Rosen. New
[CLG]
York: Grossettand Dunlap.
1970 (1519-26). Cartasde relaci6n.Mexico: EditorialPorrua.
DAVIES, NIGEL. 1977. The Toltecs.Norman: Universityof Oklahoma
Press.
[HJP]
of the
DIAZ DEL CASTILLO, BERNAL. 1908-16 (1632). Thetruehistory
conquestof New Spain. Vol. 3. Translated by A. P. Maudslay.
London: Hakluyt Society.
and conquestofMexico. Translated
--. 1956 (1632). Thediscovery
by A. P. Maudslay. New York: Noonday Press.
--. 1963 (t568). The conquestofNewSpain. Translatedby J. M.
[CLG]
Cohen. Baltimore:Penguin.
DURAN, DIEGO. 1867 (1581). Historiade las Indias de Nueva Espania
[CLG]
e islas de TierraFirme.Mexico: Porrua.
1964 (1581). The Aztecs: The historyof the Indies of New
--.
Spain. Translated by D. Heyden and F. Horcasitas. New York:
OrionPress.
1971 (1570, 1579). Book ofthegodsand ritesand theancient
--.
calendar.Translated by F. Horcasitas and D. Heyden. Norman:
Universityof Oklahoma Press.
DYCKERHOFF, URSULA, and HANNS S. PREM. 1976."ILa estratificaci6n
social en la Mesoamerica
social en Huexotzinco,"in Estratificaci6n
Edited by P. Carrasco and J. Broda, pp. 157-80.
prehiispdnica.
Mexico: Nueva Imagen.
[BJP, WTS]
FLANNERY, KENT V. 1972. The cultural evolutionof civilizations.
3:399-426.
Annual ReviewofEcologyand Systematics
GIBSON, CHARLES. 1964. The Aztecs underSpanish rutle.Stanford:
StanfordUniversityPress.
. 1971. "Structure of the Aztec empire," in Handbook of
Middle AmericanIndians, vol. 10. Edited by R. Wauchope, pp.
376-94. Austin: Universityof Texas Press.
view of an econGORDON, CONNIE L. 1980. An anthropogeographic
omy: The late preconquestBasin of Mexico. UnpublishedPh.D.
[CLG]
dissertation,Columbia Universitv,New York, N.Y.
GRIFFIN, JAMES B., and ANTONJETAESPEJO. 1950. La alfareria
correspondienteal ultimo periodo de ocupacion nahua del Valle
de Mexico. 2. Tlatelolcoa travesde los Tiempos11: 15-66.
EULALIA. 1938. Un manuscritode la Coleccion Boturini
GUZMAJN,
3:89-103.
que tratade los antiguosSefioresde Teotihuacan.Ethnos
Informaci6nsobrelos tributosque los Indios pagaban a Moctezuma.
Aniode 1514. 1957. Edited by F. V. Scholes and E. B. Adams.
Documentospara la Historiadel Mexico Colonial 4.
IXTLILXOCHITL, FERNANDO DE ALVA. 1952 (ca. 1600-40). Obras
hist6ricas.Vol. 2. Edited bv AlfredoChavero.M6xico: Nacional.
JACOBS, JANE. 1969. The economyof cities. New York: Random
House.
[BJP]
JERMANN,JERRY V., and ROBERT C. DUNNELL. 1979. "Some limitain
tionsofisoplethmappingin archaeology,"in Computer
graphics
Statisticalcartographic
applicationsto spatial analysis
archaeology:
Edited by S. Upham,pp. 31-60. Arizona
in archaeological
contexts.
[RDM,
State UniversityAnthropologicalResearch Papers 15.
DEL]
JOHNSON, GREGORY A. 1973. Local exchangeand earlystatedevelopIran. Universityof Michigan Museum of
mentin southwestern
AnthropologicalPapers 51.
Anthropology
KATZ, FRIEDRICII. 1966. Situaci6n social y econ6micade los A ztecas
durantelos siglosXV y XVI. M6xico: Institutode Investigaciones
Hist6ricas,UniversidadNacional Autonomade Mexico.
1972. AncientAmericancivilizations.New York: Praeger.
--.
rRDM, DEL]
LALONE, D., and M. B. LALONE. 1979.Trade and marketplacein the
Inca realm. Paper presentedat the 43d meeting of the International Congress of Americanists,Vancouver, B.C., Canada,
August.
[JRP]
of kinship.
structures
LEVI-STRAUSS, CLAUDE. 1969. The elementary
Translated by R. Needham. Boston: Beacon Press.
y Tepecoacuilco:Provincias
LITVAK KING, JAIME. 1971. Cihzuatldn

Vol. 21 * No. 4 * August1980

477

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

de Mexico en el sigloXVI. Mexico: Institutode Investitributarias


gacionesHist6ricas,UniversidadNacional Autonomade Mexico.
McBRYDE, FELIX W. 1947. Cultural and historicalgeographyof
SouthwestGuatemala.SmithsonianInstitutionInstitute of Social
[CLG, BJP]
Publication4.
Anthropology,
MARX, KARL. 1933 (1867). Capital. London: Everyman'sLibrary.
MAUSS, MARCEL.1954 (1925). The gift.Translated by I. Cunnison.
London: Cohen and West.
MONZON, ARTURO. 1949. El calpulli en la organizaci6nsocial de los
Tenochca.Institutode HistoriaPublicaci6n 14.
TORIBIO DE. 1950 (ca. 1536-43).Historyof
MOTOLINIA (BENAVENTE),
the Indians of New Spain. Translated by E. A. Foster. Cort6s
SocietyDocumentsand NarrativesConcerningthe Discoveryand
Conquest of Latin America,n.s., 4.
MURRA,J. V. 1972. "El 'control vertical' de un maximo de pisos
ecologicosen la economiade las sociedades andinas," in Visita de
la provinciade Le6n de Hudnucoen 1562,by IfiigoOrtizde Ziufniga,
pp. 427-76. Huanuco, Peru: Universidad Nacional Hermilio
Valdizan.
[JRP]
NarrativeofsomethingsofNewSpain and ofthegreatcityofTemestitlan
Mexico. 1917 (1556). Translated by M. H. Saville. New York:
Cortes Society.
de Tenayuca y las excavaEDUARDO.1935."La cer6amica
NOGUERA,
in Tenayuca,pp. 141-201.Mexico: Departacionesestratigraficas,"
mentode MonumentosArtisticos,Arqueol6gicose Historicos.
en el sistema
PALERM, ANGEL. 1973. Obras hidrdulicasprehispdnicas
lacustredel vallede Mexico. Mexico: InstitutoNacional de Antropologia e Historia.
Papeles de Nueva Espana. 1906-48. Vol. 7. Edited by F. del Paso y
Troncoso.Mexico- Sucesoresde Rivadeneyra.
patternsin theTexsettlement
PARSONS, JEFFREY R. 1971. Prehistoric
coco region,Mexico. Universityof Michigan Museum of AnthropologyMemoir3.
. 1974. The developmentof a prehistoriccomplexsociety: A
regionalperspectivefromthe Valley of Mexico. Journalof Field
[HJP]
1: 81-108.
Archaeology
1976. "The role of chinampa agriculturein the food supply
of Aztec Tenochtitlan,"in Culture changeand continuity:Essays
Edited by C. Cleland,pp. 233-62. New
in honorofJamesB. Griffin.
York: AcademicPress.
ofLate Postclassicspindle
PARSONS, MARY H. 1975.The distribution
whorlsin the Valley of Mexico. AmericanAntiquity40:207-15.
PETERSON, DAVID A. 1976. Ancientcommerce.UnpublishedPh.D.
dissertation,State Universityof New York at Albany, Albany,
N.Y.
[DAP]
POMAR, JUAN BAUTISTA. 1941 (1582). "Relacion de Tezcoco," in
para la historiade Mexico. Edited by
Nueva colecci6nde documentos
J. Garcia Icazbalceta, pp. 1-64. Mexico: Salvador Chavez Hoyhoe.
PREM, HANNS J. Editor. 1974. Matricula de Huexotzinco(Ms. mex.
387 der BibliothequeNationaleParis). Graz: AkademischeDrucku. Verlagsanstalt. [PC]
PRICE, BARBARA J. 1979."Turning state's evidence:Problemsin the
in politicaleconomy.
theoryof state formation,"in New directions
Edited by M. B. Leons and F. Rothstein,pp. 269-306. Westport:
GreenwoodPress.
Relacion de Zempoala y su partido, 1580. 1949 (1580). Tlalocan
3:29-41.
SACHSE, URSULA. 1966. "Acerca del problemade la segundadivisi6n
social del trabajo entrelos aztecas," in Traduccionesmesoamericanistas, vol. 1, pp. 73-145. Mexico: Sociedad Mexicana de Antro[PC]
pologla.
BERNARDINO DE. 1956 (1577). Historiageneralde las cosas
SAHAGUIN,
de Nueva Espanta.Vol. 2. Edited by A. M. Garibay K. Mexico:
Porrua.
ofthethingsof
1959 (1577). FlorentineCodex:Generalhistory
--.
New Spain. Translated by ArthurJ. 0. Andersonand Charles E.
Dibble. Santa Fe, N.M.: School of AmericanResearch/University
of Utah.
[PC, CLG, HJP]
SAHLINS, MARSHALL D. 1972. StoneAge economics.Chicago: Aldine.
analysis.
SAMUELSON, PAUL A. 1964. Economics: An introductory
New York: McGraw-Hill.

478

T. 1956." The CentralMexican symbioticregion,'


in theNew World.Edited by G. R
settlement
patterns
in Prehistoric
23
Willey,pp. 115-27. VikingFund Publicationsin Anthropology
. 1957. Tierra y agua. Unpublished Ph.D. dissertation
[CLG]
Harvard University,Cambridge,Mass.
1965. Cultural ecology of the Teotihuacan Valley. MS
PennsylvaniaState University.
Departmentof Anthropology,
. 1968. " Hydraulic agriculture,economicsymbiosis,and the
arche
evolutionof states in Central Mexico," in Anthropological
ologyin the Americas. Edited bv B. J. Meggers, pp. 88-107
Washington,D.C.: AnthropologicalSocietyof Washington.
-. 1971."The populationof the Teotihuacan Valley, the Basii
of Mexico, and the Central Mexican symbioticregionin the 16t1
century,"in The TeotihuacdnValley project:Final report.Vol. 1
pp. 385-457. PennsylvaniaState UniversityOccasional Papers ir
3.
Anthropology
H
SANDERS, WILLIAM T., JEFFREY R. PARSONS, and MICHAEL
LOGAN. 1972. "Summary and conclusions," in The Valley o
Mexico. Edited by E. R. Wolf, pp. 161-78. Albuquerque: Uni
versityof New Mexico Press.
SANDERS, W. T., J. R. PARSONS, and R. S. SANTLEY. 1979. The Basil
ofa civilization.Nev
ofMexico: Ecologicalprocessesin theevolution
[CLG, RDM, DEL, JRP]
York: Academic Press.
SANDERS, WILLIAM T., and BARBARA J. PRICE. 1968. Mesoamerica
The evolutionofa civilization.New York: Random House.
SERVICE, ELMAN R. 1975. Originsof thestateand civilization.Nev
York: Norton.
. 1978." Classical and moderntheoriesofthe originsofgovern
ment," in Origins of the state. Edited by R. Cohen and E. R
Service,pp. 21-34. Philadelphia:Instituteforthe Studyof Humai
Issues.
1960. The preindustrialcity. New York: Frei
SJOBERG, GIDEON.
Press.
SMITH, ADAM. 1937 (1776). An inquiryinto thenatureand causes o
thewealthofnations.New York: Modern Library.
SMITH, CAROL A. Editor. 1976. Regionalanalysis.2 vols. New York
Academic Press.
1977. "How marketingsystemsaffecteconomicopportunit,
in agrariansocieties,"in Peasant livelihood.Edited by R. Halperii
and J. Dow, pp. 117-46. New York: St. Martin's Press.
SMITH, MICHAEL E. 1979. The Aztec marketingsystemand settle
ment patternin the Valley of Mexico: A central-placeanalysis
AmericanAntiquity44:110-25.
JACQUES. 1961. Daily life of the Aztecs. Translated bi
SOUSTELLE,
P. O'Brian. Stanford:StanfordUniversityPress.
SPENCE, MICHAEL W., and JEFFREY R. PARSONS. 1972."Prehispani,
obsidian exploitationin Central Mexico," in Miscellaneousstudie,
in Mexican prehistory,
pp. 1-43. Universityof Michigan Museun
of Anthropology
AnthropologicalPapers 45.
STEWARD, JULIAN H. 1950. Area research:Theoryand practice.Nev
[BJP]
York: Social Science Research Council.
TEzOZOMOC, HERNANDO ALVARADO. 1941 (1609). Cr6nicamexicdyotl
Translatedby A. Le6n. Institutode Historia,UniversidadNaciona
Autonomade Mexico, PrimeraSerie, 10.
1975 (1598). Cr6nicamexicana.Edited bv D. Manuel Orozcc
y Berra. Mexico: Porrua.
TILLY, CHARLES. 1975. "Food supply and public orderin moderi
Europe," in The formationof national statesin WesternEurope
Edited by C. Tilly, pp. 380-455. Princeton:PrincetonUniversity
Press.
PAUL. 1958. Surface survey of the northernValley o
TOLSTOY,
Mexico: The Classic and Post-Classic periods. Transactionsofth,
AmericanPhilosophicalSociety48, pt. 5: 1-101.
Cam
IMMANUEL. 1979. The capitalistwvorld-economy.
WALLERSTEIN,
[FR]
bridge:CambridgeUniversityPress.
WITTFOGEL, KARL A. 1957. Orientaldespotism.New Haven: Yale
UniversityPress.
[BJP]
WOLF, ERIC R. 1956. Aspectsofgrouprelationsin a complexsociety
58:1065-78.
Mexico. AmericanAnthropologist
[BJP,
WRIGHT, HENRY T. 1977. Recent researchon the originof the state
6:379-97.
Annual ReviewofAnthropology
SANDERS, WILLIAM

CURRENT

This content downloaded from 132.248.9.8 on Tue, 25 Nov 2014 13:56:34 PM


All use subject to JSTOR Terms and Conditions

ANTHROPOLOGY

You might also like