Domestic Architecture, Ethnicity,
and Complementarity in the South-Central Andes
Edited by Mark 8. Aldenderfer
University ofTowa Press YP towa CityDomestic Architecture, Household Archaeology,
and the Past in the South-Central Andes
Mark §, Aldenderfer and Charles Stanish
Resumen
La arquitectura doméstica es una categoria de informacién,
‘muy til; se puede usar para investigar problemas de gran
interés en la arqueologia de los Andes sud-centrales, tales
como la identificacin de la complementaridad vertical, el
Aesarrollo de grupos éites y los origenes de grupos étnicos.
ste informe define conceptos importantes como la arqut-
rectura doméstica, la organizaci6n de la familia y etnicidad,
Joa infor a nq
volumen.
This book is about the relationship between domestic
architecture, ethnicity, households, and economic com-
plementarity. The setting is the south-central Andes, a
region defined as stretching north-south from the Rio
Majes in southern Pers to Taltal in northern Chile and
cast-west from the Pacific coast to the vallwna, or selva,
on the eastern flanks of the Andes in Bolivia (Aldenderfer
19894; fig. 1.1). More specifically, the papers in this vol-
ume focus on two areas: the western valleys ofthe Pacific
watershed and the Titicaca basin. Two valleys figure
prominently in the papers—the Moquegua (or Ostore)
of southern Peru, and the Azapa of northern Chile
(fig. 1.2). Both have been the focus of intense archaco-
logical studies for more than a decade (see, for example,
Hidalgo ct al. 1989 and Rice, Stanish, and Scarr 1989).
In both, research has been conducted from the coast to
the puna and across all major time periods, thus offer-
ing an unprecedented opportunity to examine cultural
change and transformation in this important environ-
rental transect. Likewise, two portions of the Titicaca
basin are tepresented—the south shore of the lake
around the modern city of Juli in southern Peru and the
area surrounding the major site of Tiwanaku in Bolivia
(fig. 1.3). Like the Moquegua and Azapa valleys, the Ti-
ticaca basin has long been a focus of archaeological re-
search, with new, major research projects initiated inthe
1980s (see Kolata 1986, 1991),
‘As most Andeanists know, the south-central Andes
as Been The scene oF Consicerabte ert -
search, and among the many important scholars who
have worked in the region, probably the most influential
has been John Murra. In a series of papers published
from the 1960s to the present, Murra has argued that
the late pre-Hispanic and immediately post-Conquestin-
digenous societies in this region practiced what has been
She eli ozo emma. His model
cation of
aitemapied 16 Broaden thelr subsistence and
fac by obtaining aczeseto prodversffam other produc
i Macea's model, puna-based societies reliant
iy upon tuber production and camelid pastoral-
‘ism attempted fo control, Vir the establishment of eth-
nic enclaves, or colonies, the subsistence production of
2, jy Wid other lowland products in the western
alle of HEN “the Andean slopes. Ie shouldbe clear that, itr
“Dart, the success or failure of this model as_an explana-
tion of che past rests upon ihe ability to demonstrate the
existence of thse enclaves.
“Traditionally, this model has been tested through the
use of antifacts and high-status goods found cither in2 Domestic Architecture, Household Archaeology, and the Past
BoLivia
ARGENTINA
Figure 1.1. The south-central Andes.
public architecture or burial contexts. However, as one
‘of us has argued, this approach to the problem suffers
from a number of defects, the most telling of which is
simply that artifacts are highly mobile (Stanish 19898).
‘They are traded, exchanged, heirloomed, and lost, and
thus, as clear and unambiguous measures of ethnicity,
they leave much to be desired. Itis not that the artifacts
themselves lack indicators of ethnicity, since ethnoar-
chaeologieal research bas abundantly demonstrated that
“style” can serve as a social or ethnic marker, but rather
that because they are’mobile they are freqently not
found in the zone of production and therefore do not
necessarily represent some “local,” ethnic entity. This
problem is exacerbated by many archaeologists who fail
to distinguish or uncritically examine as a coherent
assemblage artifacts obtained from different contexts.
Mortuary contexts in the Andes frequently contain spe-
cial, often exotic and fonlocal ceramic types that are not
found in kitchen middens or other refuse (Lumbreras
1974b; Stanis ithout the use of proper cau-
tions, these artifacts can give a false reading of the ethnic
affiliations of the dead whom they accompany.
‘We argue thar the study of domestic architecture, com-
bined with traditional archaeological practice, is a po-
tentially powerful means of testing not only the hypothe-
sis of zonal complementarity in the south-central Andes
but also a number of other problems of current interest,
such as the identification of ethnic groups, the causes of
stability and change in household composition and size,
and much more. As a number of scholars have noted,
there are many distinct forms of complementarity (Salo-
mon 1985}, and while Andeanists have a special interest
in the concept, itis best to think of complementarity as a
broad, anthropological principle that refers to the at-
tempt of societies as simple a6, tobile foragersor as
complex as empires stence risk, expand
their productive capacity aad political hegemony, or
seck marriage partners (Aldenderfer 19894; Spielmann
1986). While the Specific dimensions of complementarity
vary with sociopolitical and soci
many of its forms are concemed
number of Andean societies practice a form of comple-
méntarity that is based solely on kin relationships and
the ability to make mastiages, ard thus economic tes, to
groups living in other productive zones, Although this
may or may not involve distinct “ethnic” groups, it may
well affect the size or composition of the household itself.
New individuals will be added to the household, and
new space or structures may be required to house them.
Changes may also occur in the structure of activity per-
formance as well or in its spatial location. In short, there
is good reason to be concerned with the houschold as a
possible means of accessing other forms of complemen-
‘arity relationships.
~ Domestic, or residential, architecture is here defined as
those stmuctures,-facilities, activity and work areas, and
re associated, with the anth pola
“itis clear that other individuals not a part of the house-
hold may live there, or that hoilsehold members may
iain residence elsewhere (Ashmore and Wilk 1988:
5 see Rice, this volume). At the minimum, domestic
architecture is simply-skélter, but it is obvious that it can
bbe much more, depending on the size and range of eco-
nomic functions performed by that Household. What is
important to stress, however, is that domestic architec-
ture is an empirical concept (arr indirect observable,Honsebold Archaeology, and she Past 3
Figure 1.2. ‘The Osmore (or Moquegea) drainage, showing the location ofthe sites discussed in this4 Domestic Architecture, Household Archaeology, and the Past
BOLIVIA
\
\ Titicoca
\& 7
wuaduera
riwallaxy
Figure 1.3. The Titicaca basin, showing the location of the sites discussed in this volume.Dontestic Architecture,
sensu Kaplan 1964), and the variability in the use of
space, building materials, facility placement and loéa-
‘ions, anid associated aftifacts within and.near the com-
plex of domestic structures is used to infer household or-
ganization and activity. Wilk and’Rathje (1982: 620)
ave drawn a similar distinction. Our perspective is also
comparable to the definition of the “household unit” by
Flannery (1983: 45)—"the complex of structures and
features resulting from a typical houschold’”—and the
“household elustes” by Winter (1976: 25).
It should be obvious that the definition of what com-
prises domestic architecture and the way in which house-
holds are inferred once domestic architecture has been
identified are not necessarily straightforward, but the
comparison of context and argument by analogy seems
to be critical to the process. For instance, Winter (1976:
25) describes the “typical” household cluster of the For-
ative Period in the Valley of Oaxaca as consisting of
“one house, va to six large storage pits, one to three
graves, and various additional features separated from
the nearest contemporary cluster by an open arca of
20~40 meters.” Context here is defined through exten-
sive excavation, and comparison is made with similarly
excavated areas through the resolution of key empirical
questions such as the form and content of the dwelling,
the range_of functions performed within and near it,
the spatial extent of che dwelling and associated activity
areas ad BEFUCTURES, ane tHe pron -
hold clusters or differently configured clusters of struc
itares, such as public architecture. Analogy, employed
‘through whatever paradigm is operative in the research
design, is used to evaluate the working hypothesis that
the cluster in question reflects a household or instead is
better thought of as some other analytical unit.
Domestic architecture is a distinct and independent
class of data useful in testing archaeological models as,
fineware pottery, decorated textiles, and the like. Domes
tic architectural styles ean be analyzed in a similar man-
ner. Styles, for instance, can be grouped into types, sta-
tistically manipulated, and compared with other classes
of data for goodness of fir with proposed models or
hypotheses
“The use of domestic architectural style in addressing
problemsin Andean archaeology has a number of advan-
tages over other classes of data. First, domestic architec
ture is not portable, unlike most artifacts encountered by
the archaeologist. The style of construction, then, bar-
ring abandonment and subsequent reuse, therefore un:
equivocally represents the resident population, Smaller
_parisons of diff
Housebold Archaeology, and the Past S
objects, such as ceramic vessels, metal objects, wooden
artifacts, of textiles, may be the product of complex ex
change relationships, and their occurrence in sites does
not necessarily reflect local styles of.production but in-
stead the operation of an exchange system with “for.
ign” groups. The style of domestic architecture, how-
ever, represents local cultural preference because there is
litle dove rat the resident population constiticted the
“dwelling: Cilltiiral or “ethnic” differences between soci-
ties iiay well be reflected in the nature, size, compos
tion, ind, most importantly, material fearures of the
structure and its facilites.
‘second advantage of this class of data is its scale. As
a unit of analysis, domestic architecture is intermediate
in scale between settlement data’and excavation data.
“The household infersed from the analysis of the struc
tures represents the primary socioeconomic group in An.
dean society (Stanish 1989). At an analytical level, itis
a corporate group of less complexity than the aylluyor
village, bit is obviodsly-‘more comiplex than the indi-
‘vidual artisan’ and areas OF Specialized activity within
sites.
‘A third advantage is comparability. The consistent use
of the household as an analytical unit standardizes com-
ent sites and different time periods. The
‘corporate groups responsible for the manufacture and
distribution of actifacts found in archaeological contexts,
‘i ¢ time The
only exception to thi isthe househola
‘A final advantage of using domestic architecture as an
independent class of data is the. accessibility of surface _
architecture in. many areas of the Andes, Sufface map-
ping of structure foundations and broad exposures of
excavation units are usualy sufficient to define the range
of variability that characterizes architectural pattems.
Tn this manner, “ideal” architectural plans can be iden-
tified and then compared both within and between
settlements,
‘The limitations of the_use of domestic architecture
mast also be recognized. To benefit fully from the ap-
proach, extensive excavation is frequently required, es-
pecially ifthe household and its activities and thei trans-
formation through time are the topic of interest, The
spatial extent of the Early and Middle Formative house
hold cluster in the Valley of Oaxaca has been estimated
6 be 300 m: (Winter 1976: 25}; to excavate simply this
area and, more importantly, to excavate a number of
these areas so that effective comparisons of context can
be made may be prohibitively expensive for most proj-6 Domestic Architecture, Household Archaeology, and the Past
‘ects, While this problem is exacerbated by the increas
ingly large scale of villages, towns, and major centers
iypical of more complex societies, it cannot be avoided
even in simpler foraging societies, in which contempora-
neous household clusters ay be separated by hundreds
of meters (Whitelaw 1991). However, if ethnic afflia-
tion is the focus of investigation, extensive excavation
ay not be required (see below). Likewise, idiosyncratic
variation in consteuction.ox conscious efforts to imitate
nonlocal architectural styles_to. enhance social “status
‘may complicate matters and muddy the emergence 6f
patteming, but these problems can be generally over-
come through effective sampling designs.
Domestic architectural data, thei, can be used, just
like ceramics, to discover meaningful, variation and
change in both space and time. To make this case,
however, we miist fist define two important terms—
household‘and etbnicity—and then relate them to the
themes ofthis book. :
Houscholds
The household is a concept that has been defined and
redefined by ethnographers, ethnologists, sociologists,
istorians, ad archaeologists. While we do not propose
“to yet again define the tec, a brief review of the salient
characteristics ofthe household isin order, especially for
the archaeologist interested in defining them on the basis
of material culture.
Households share three common characteristics: do-
mestic functions, co-residentiality, and some form of fa-
milial relationship (Bender 1967). Of the three, only the”
figst two are easily accessible to the archaeologist. Do-
«itestic funetions, such as cooking, storing, processing,
sleeping, and such, are relatively easy to “‘see” archaeo-
logically, since they leave durable remains or involve the
construction of facilities to perform them. Co-residence,
however, is somewhat less tractable but can be inferred
through the repeated, regular co-occurrence within sites
of structures with domestic features. Tt can also be in-
ferred through the identification of artifacts that pertain
to the sexual division of labor or those things that are
clearly toys and belong to children. The ability to discern
family relationships depends on the existence of mortu
ary remains.and.the.careful study of ostcological evi-
dence to demonstrate the presence of genetically related
individuals. While archacologists have ihade «-numbef
of attempts 0 define the sociological structure of ar-
chaeological households, primarily through the analysis
of stylistic motifs on ceramics, most have fallen short of
their goal
Four primary household funetions have been identi-
fied: production, through which people obtain resources;
distribution, in which that which is produced is either
distributed within the household or to a larger sociopo-
litical entity; transmission, involving the transference of
property, rights, or roles; and finally eproduction, which
centers upon child cate and socialization, (Wilk and
Rathje 1982: 622=631). Onily production and distribu
tion are generally visible in the archaeological record.
‘While obvious, itis important to stress that house-
holds within any society will vary across a number of sig-
nificant social dimensions, such as wealth, status, and
degree of economic diversification, In simpler societies,
siich’ a8 mobile foragers, most household units should
look the same. However, with the advent of sedentariza
tion and the integration of the houschold into village-
and regional-scale economies, we can expect to find
variation in household size and composition, an increase
in domestic structure size or in numbers of structures
used by the household, and building of special-purpose
structures to house craft activities or surplus accumula-
tion. Despite this potential variability, however, the basic
‘esidential plan will remain more or less the same, and
ani‘ideal” Forna of domestic residence can usually be de-
fined (Bawden 1982)
“Therefore, itis necessary to carefully identify the em-
pirical indicarors of houscholds; that is, the characteris
tics of domestic architecture that are used to infer house-
hhold structure and its variability. Perhaps the most
important are (1) the numbers and sizes of structures in
use, (2) the use.of space both within structures and out-
side them, (3) types of facilities employed and their spa
at distribution, and (4) facility or artifactural indicators
of the Fange Of activities performed. While other indica-
tors of wealth status, size, and economic diversity can be
posited, many of these, such &s decorationyare perishable
(Rapoport 1969, 1980) and thus are generally of little
value to the archaeologist.
‘Whatever the cultural context, however, its lear that
most of the primary determinants of houschold strac~
ture, composition, and form can be inferred from mate-
rial culture. While some archaeologists, such as Hodder
{1982), have argued that because of the lability of sym-
bolic communication itis essentially impossible to infer
sociological entities such as the household with any ac-
curacy, Wylie (1989: 105-107) and others, using the no-
tion of “qualified abjectivism,” have shown that such"Domestic Architecture, Household Archaeology, and the Past 7
definitions are indeed possible and plausible when care is
taken to specify the constraints and the context of the
problem at hand.
Ethnicity
Everyone knows what ethnicity is, but there is litle
agreement on just what a definition of the term should
include. Rather than focusing upon a potentially sterile
definition, ie is more valuable to look for critical features
of ethnic groups and then to make some effort to deter-
rine to what extent ethnicity can be read and interpreted
from the archaeological record. Most authors agree that
c groups are, social groups, often based. on situa
83), thereby having a potentially
lage field, anid; further, they are defined by ascription
(Barth 1970: 14), meaning that they ate defined both
within groups as well as berween groups. That iy in
viduals see thernselves as distinct from other individuals
of different “ethnic” groups. This emphasis on ascription
by the individuals themselves is highly important, be-
cause it means that there RO hevessairy equivalence be-
tween ethnic units and cultural differences. While Barth
(1970: 14) suggests that there are, at least analytically,
two distinct eypes of cultural contents that serve to define
“ethnic differences (overt signs, including dress, house
fori {emphasis added), and language, and basic value
‘Orientations), there is nothing in theory or From first
principles that defines just how ethnic groups necessarily
define themselves and are in turn defined by others. What
this means, simply, i that there ig no such thing as an
ethnic group “out there” to be precisely and unambigu-
ously defined.
The problem for the archaeologist is obvious: if eth-
nicity is_a_ self-referential, situational concept, we are
joing to have great difficulty finding it represented in
the archaeological record. The vehicle of the “archaco~
logical culture,” thought by practitioners of the cultural
historical “approach to archacology characteristic of
much of the twentieth century (Dunnell 1986: 33-34) to
be analogous to ethnographically defined and observed
cultures, has been found to be wanting both theoretically
and empirically and therefore is not a useful pursuit for
the archaeologist interested in ethnicity (Shennan 1989
5-14). What may be of greater value is the search for
ethnicity inthe stylistic variation of material cultare, par-
ticularly by the definition of so-called emblemic style
(Wiessner 1983: 257; Shennan 1989: 17-22). Emblemie
_styleis variation in material culture that is desig dt
__send a message to other individuals about the social or
cultural identity of the makers of the object—in other
iid, it defines boundaries and separates, a least con-
ceptually, individuals of different social entities from one
another. Emblem style, if itis indeed amenable to ar-
chacological analysis, would appear to be an ideal way
to define potential ethnic differences in the archaeologi-
cal tecord, but unfortunately-the, situation is not so
simple. There is algo assertive style)or that variation in
‘material culture that eihcamicjously or unconsciously
describes individual identity (Wiessner 1983: 258), and
finally isochrestie variation (Sackett 1982), which is the
ne ase af particular forms of repre-
fation, one that is said to carry considerable
symbolism.” What is most important about isochrestic
“yaFiatiON aS a marker. of ethnicity is.that it arises from
tional, automatic ways.of making or decorating
‘things that arise primarily through enculturation (Shen-
nan 1989: 19)
‘The key to disencangling these types of “style,” then,
is through the analysis of context—how objects are
in fact used and the kinds of messages sent (Wiessner
£983). While this type of analysis is exceedingly dificult,
it is nevertheless possible, especially if Wyle’s (1989)
saialified objectivisin i recalled and pressed into service.
Since it has been demonstrated with ethnoarchaeological
data that itis possible so.identify emblemic style {Wiess-
er 1988), the careful anc considered se of the-con=
cept in an archaeological context is both feasible and
worthwhile
Architecture and Fi
icity
The use of domestic architecture as a major method-
ological tool in south-central Andean studies obviously
hinges upon the degree co, which domestic architecture
can be reliably said to reflect ethnicity. Therefore, we
rust explore the degree to which, domestic architecture
ccan express emblemic or assertive style or more simply
reflects isochrestic variation. By expressing this question
as.a matter of atyle, we are attempting to explore the po-
tential theoretical basis for the'use of domestic architec
ture’as a reflection of ethnicity. However, we mist stress
that this theoretical assertion must be tested as an em=
pirical matter, one that must be investigated within the
context of the spatial variation in all forms of material
culture, not just architecture. This conforms with the no-
tion of identification by comparison developed by Wiess-
ner (1989). z8 — Domestic Architecture, Household Archaeology, and the Past
AAs we have argued above, domestic architecture may
well have stylistic features that can be used to determine
the ethnic or cultural affiliations of its builders. Under
what sorts of contexts would this assertion be valid? To,
answer this question, we must frst discuss other factors
that may influence the form of domestic architecture, Be
fore so doing, it is important to specify what exactly we
mean by form. We choose, atleast for simplicity’s sake,
to focus on the empirical dimensions of form. Aspects of
this discussion overlap with Rapoport’ (1980: 291
296) contention that the form of domestic architecture
has four dimensions: the organization of communica
tion, time, space, and meaning, Form includes the shape
df the'sifticture and its dimensions, the organization of
space (number of structures or roms, the placement of
facilities, the kinds of activities performed within differ-
ent kinds of spaces), ratios of room sizes, and the connec
tivity and orientation of structures of rooms (ie are
there déorways, halls, etc., and ifso, where do they lead).
Each of these is easily observed in the archaeological rec
ord, and the composite of their expression is the empici
cal statement of the form of domestic architecture.
‘One of the most obyious determinants of. form is cli
‘mate, but as Rapoport (1969) and others have noted,
ite does not seem determinative of thespecific forms of
‘houses, although it may play some modifying role. Like
-wise, thexange of building matefials preient are best seen
‘as modifiers, rather than determiners, of house shape
(Rapoport 1969: 26), Similarly, status or social factors
are, probably. also seen-as‘modifying, rather than deter-
‘mining, thé form of domestic architecture, While large
households are likely to have lager domestic structures
than smaller ones, household size or organization cannot
predict the form of that structure. In short, while many
factors can be seen as modifying the form of domestic
architecture, only ethnicity or cultural afiliation can be
said to influence structural form in a direct way. Conrad
(this volume} summarizes this position nicely in his dis
cussion of Estuquifia Phase domestic architecture.
If ethnicity. (ox “culture” is determinate, what are the
contexts in south-central Andean prehistory in which we
can expect ste to reflect ethnic distinctions? What we
are asking here is the degsse to.sehich domestic archi-
tecture willbe used asiemblemic style, rather than as a-
sertive style of isochrestic.variation. One of thé most
obvious contexts is the establishment of classic zonal
complementarity ia the western valleys of the Andes (sce
below). Colonists or mere setters are likely to bring their
architectural canons with them as an expression of eth-
nic or cultural difference. Similarly, if state organizations
impose colonists in other regions, we may also expect
domestic architecture to. become emblemic in che mes-
sage it sends. That Andean states attempted, at least in
some instances, to maintain ethnic distinctions in colo-
nized or otherwise conquered regions is testified to in the
ethnohistorical record, Julien (1983: 4445}, in describ
“ing ethnic variation in the Titicaca basin, notes that the
Inca encouraged colonists to. maintain their styles of
dress further, she notes that sanctions were applied to
individuals who broke these codes. While the example
she uses is Colla headdress and head deformation, it is
also possible to hypothesize a similar role for domestic
architecture. Under these circumstances, distinct archi-
tectural forms most probably can be read as signaling
ethnic differences and accordingly analyzed.
Applications in Andean Archacology
‘We identify five sets of archaeological problems in the
south-central Andes which the analysis of domestic ar-
chitecture can be particularly useful in resolving. These
problems center on (1) testing models of zonal comple
mentarity, (2) defining the processes of core-periphery re
ips in imperial contexts, (3) defining the proc-
i386 Of elite group formation in pre-Hispanic contexts,
(4) analyzing. changes.in the structure and composition
cof houscholds that may reflect suprahouschold organ-
iaational changes, and (5) examining the processes of
ethnic group formation. It should be obvious that these
pidblem areas are to some extent overlapping, and that
further they complement some of the problem do-
mains proposed by Rathje (1983) for the study of Maya
households.
Zonal Complementarity
‘Models of zonal complementarity or.vecticality comprise
some of the most successful theoretical tools in Andean
studies, and a huge literature exists concerning these
‘models (for an excellent synthetic review of the concept
of veticality, see Salomon 1985). For our purposes here,
‘we wish to emphasize the central role that ethnicity plays
in zonal covmpleniestarity models. Classic direct comple-
sientarity (Muza 1964) involves the maintenance of dis-
_Gfete ethnic populations intermixed with others over a
of complementary ecological zones. The resul's a
hnic énclayes over a landscape. In some
cases, control is ¢Hected through the establishment of
colonists from one productive zone into another. The de-
fee to which these colonists maintain their original eth-
nic identity can be measured, we believe, in great partDomestic Architecture, Household Archaeology, and the Past 9
by. the kinds of domestic strucmres they build. What is
clearly cequiced for the success ofthe test ofthis hypothe-
sis is to identify clear, contemporaneous differences in
domestic architecture within one productive zone, and
then to discover the source of one of these residential pat-
terns in some other zone. This type of research isthe fo-
cus of Goldstein, Stanish, and Conrad, Goldstein is con-
cerned with defining the nature of Tiwanaku settlement
in the middle Moquegua deainage and has employed
studies of both domestic architecture and house content
to argue that the site of Omo reflects the presence of al=
tiplano colonists in the valley. In contrast, Stanish and his
colleagues, working in the Titicaca basin near Juli are
concerned with defining the typical residential patterns
‘of: domestic structures of Late Intermediate Period sites,
In previous research, he has rejected the notion that Lu-
paqa colonists penetrated portions of the upper Moque~
gua drainage during the Late Intermediace Period (Stan
ish 1989a, 1989b). While his hypothesis is plausible, ix
lacks one of the components necessary for a complete
test of the model—the definition of the source of the dif-
ferent pattern of domestic architecture. This paper re
pots on his recent efforts to define che highland sources
of “foreign” domestic. architecture-in the Moquegua
drainage. Conrad, in his analysis of the contemporary
Late Inteémediate sites of Estuquifia and San Antonio,
provides additional local support for the rejection of the
hypothesis of Late Intermediate Lupaqa coVonists in the
Moquegua drainage.
Models of indircet complementarity, such as those that
posit the eistence of trade or exchange. relationships
that may'Fitiay not be controlled by some centralized
authority, also rely upon definitions of ethnic groups,
‘While some models may require trade relations between
either afinal or consanguifiea kin living in diferent pro-
ductive zones, proposed that such trade is in
‘the hands of different “peoples.” In this case, itis neces-
sary to define the existeicé OF at least two distinct ethnic
‘groups, each of which ive in different zones, This isthe
focus of Mufioz’s research in the Azapa valley of north-
cern Chile. Using domestic architecture as well as artifac-
-tual and subsistence evidence, he has argued for the long
term development of two distinct peoples—lowlanders,
who utilized the littoral and inland valleys, aiid high-
landers, who lived in the high sierra and puna.
Core-Periphery Relationships
A related methodological problem centers on defining,
the relationships between core imperial territories and
their colonial citiers:in their provinces. In the south-
central Andes, the Tiwanaku state had expanded its
influence to an area of more than 400,000 km? by
‘Ap, 800. The nature of that “influence” remains highly
Tiwanaku expansion are
panisms of imperial control range from the
gious integration to complex forms
ie exchange (Berenguer, Castro, and
1981).
‘The methodological problem of modeling core-periph-
cry relationships therefore centers oi the ability to dis-
‘riminate between various integrative mechanisms. The
usé of domestic architectural data can help to resolve
some of these problems. In the case of colonies, for in
stance, we would expect to see distinct and intrusive do-
restic architectoral styles that contrast with those char-
jose models in which
other integrative mechanisms are of paramount impor-
tance, domestic architecture should remain more or less
constant through time, although it may become more
complex if household functions or composition become
more complex.
The use of domestic architectural data is superior to
other classes of data such as ceramic Finewares and high-
status items. This is because both of these artifact types
will be cizculated to peripheral zones in either a colonial
cor exchange-based model of imperial politcal economy.
These high-status items will be found in both models ei-
TEE TAGiTeCY CHTOUBN exCHANgE BOWER aMOTTON
polities or through complex redistribution and/or recip-
+rocal relationships between an administrative site and lo-
‘al setlements. In contrast, domestic architectural styles
vill vary accordingly to che nacure ofthe relationship be-
tween core and periphery.
‘The most powerful methodological strategy is a con-
textual one that centers.on the household and includes
ail relevant classes of data. That is, the houschold
becomes the basic unit of analysis, and other data (ccr-
amie finewares, textiles, etc.) are evaluated within this
context. The significance of these objects for testing cote-
periphery models therefore varies according to the con-
text in which they are found. Tomb contexts tend co
represent interregional exchange mechanisms whereas
id to contain fa fewer exotic items
“unless their inhabitants are of high status or are wealthy
(Stanish 19892: 13). Those culturally distinct objects
found in domestic. contexts are therefore more likely
to represent local stylistic traditions. While specific
applications of these contextual approaches must be
farther refined, we feel that without question 2 con-
textual methodology based around the household is10
superior to-current ones. employed in. the south-central
Andes.
Of the essaysin the volume, Goldstein and Van Buren,
Biirgi, and Rice approach this theme most closely. Gold-
stein, as noted above, has argued strongly that Omo isin
fact a Tiwanaku colony and presents data supporting
this asvertion, His feld research closely approximates the
prescriptions offered above for contextual research based
upon the household, Van Buren and her associates look
_at residential patterns at Torata Alta, a Late Horizon site
in the upper Moquegua drainage. While the dating of the
it@ remains Somewhat problematic, itis clear that the
site is a planned settlement and was established either
theoiigh Tika of Spanish influence. Whatever the mech-
aiism-of-formatién, the authors conclude that the site
was occupied most probably by Lupaga settlers late in
the Late Horizon. The degree to which this site can be
said to be a colony, however, remains unresolved,
‘The Formation of Elite Groups
A virtually unused methodological strategy for under-
standing the complex processes surrounding the evolu-
tion of hierarchy in the pré-Hispanic Andes isa researcli
design based upon an analysis of domestic architecture:
In theiz work in the upper Mantaro valley, Costin and
Earle she impact the Jak: thad on lox
cal sociopolitical structures. In particular, the strength-
«ening ofthe local cite groups relative to their community
‘ava whole by identification with the Inka state is clearly
‘manifest in changes in the domestic architectural plan on
the loca sites:
“The “Inkancization of loca elites is seen farther ina shift
in archtectusal style. In Wanka Il (the presInka period),
weit simple, gular etre
‘typical Differences berween lite
and commoner housing were primarily in the fineness of
che masonry and che size ofthe patio group. fo, Wonka
), while commoner housing continued
iy unchanged in form, elies began building some
rectangular structure with inteiog trapezoidal niches
Ther
se of local
cs became marked not dvrough us
tymbob bs bysslene to he oka sates perhaps
to stretch the point, became legitimized
oy deni iri torre
ania
This lengthy quote dramatically illustrates the sensi-
tivity of domestic architectural patterns to some larger
Domestic Architecture, Household Archavology, and the Past
cultural process of enormous importance to contempo- _,
rary archaeological theory. The nature of Wankai elite
control ‘with their in-
ships outside their society was destroyed under Inka rule,
their inten appears to have actualy increased
“With the backing of Inka state authorities. This classic
instance OF a Secondary state formation process can be
documented by a research design based upon the analysis
of changing structure form,
Stability and Change in Household Organization
Households of course, do not exist within a vacuum but
are instead embedded ina sociocultural milieu of varying
omiplexity: Oe of the key questions in south-central
“Kndean axchaeology, or for that matter the archaeology
‘of whatever region, is the degree to which changes in the
broader fabric of society are reflected at the household
level, While this question is hardly news, the emphasis on
domestic a cone of che means of understand-
ing this problem represents a new direction in research.
“The empirical indicators of form of domestic architec-
5 of haeologist a wealth of new possibilities
t the
“for te exploration of this problem. For instance, Rathje
“den? Rath
“TIBBBY questions to what degree the rulers of classic
Maya society were functions!” or ¢“fungal.” Such a
“Guestion betrays the fa¥ larger problem concerning the
utility of functionalist versus structuralise explanations
in prehistory. That is, did the elite provide benefits to
society, or did they simply sponge off
‘and act as.a bur-
oses to fest this dichotomy with mea-
“sures of household well-being: Although he does not spe-
Gfically include afehitectté, itis obvious that changes in
Size aid range of activities pérformed within households
say well ellect the relationship of that household to the
wider world.
A similag question is concerned with the ise of craft
specialization, usually considered to be a hallmark of
oitplex societies, Specialization itself has a number of
‘dimensions, some of which are relevant toa concern with
_ houschold archaeology and domestic architecture. These
are the locus of productioi aid the scale of that produc-
tion (Brumfiel and Earle 1987: 5). Are crafts organized
within the household? If so, are many housebolds active
in craft production? What is the nature of local and
regional variability in craft production? Are specialist
households wealthier than their nonspecialist counter-
parts, or have specialists become specialists simply due
toa lack of other options?
t-Domestic Architecture, Household Archaeology, and the Past
The degree to which houscholds are articulated with
the surrounding world is an important topic. Some
households, due to differences in wealth, size, or oppor-
tunity, are more connected to the broader world, and it
js obvious that some of the measures of household well-
being proposed by Rathje are relevant here. Bermann, in
his discussion of household archaeology at Lukurmata,
demonstrates that households remain remarkably stable
in terms of functions despite significant sociopolitical
changes taking place around them. That is, while the
Fiouschold may be integeated into the wider sphere of ex
change and tribute, the effects of changes in the forces
cannot easily be deduced from domestic architecture or
the debris found within it
‘Changes in social organization itself may be reflected
by changes in domestic architecture. Many Andeanists
have argued that dual organization's characteristic of so-
cial structures in the Andes aid, further, that this form
of (Netherly and
Dillehay 1986). Alehough individial households may
not reflect this type of organization, it is clear that the
study of aggregates of houscholds and structures offers,
{great potential for answering this question. Bawden, in
his analysis of changes in domestic architecture at the
_Tumilaca stein the Moquegua drainage, presents & com-
“Pelling case for the role of both ethnicity and the devel-
opment of dual orga alin changes in the
Tuchitecture from the late Tiwanakw occupation of The
site to the ithmediate post-Tiwariaka period. Similarly,
Aldenderfer argues that domestic dithiternuxe: monitors
highland and lowland Torag
rity relationships between
sin the upper.
Sonata .of.comple-
mmentaity. However,
change in this system to one of balanced reciprosiry, a
change which domestic architecture reflects very well.
The Formation of Ethnic Groups
The Spaniards, entering the Titicaca basin in the six-
teenth century, were impressed by, and conseqiieily ré*
corded, what they perceived as ethnic variation in the re-
gion (Julien 1983). Among the groups said to exist were
Colla, Lupaga, and Pacajes, among others, While it is
*poisible to argue about the degree to which these distinc-
tions were truly ethnic, it can nevertheless be asserted
that, there was considerable cultural diversity in the re-
¢gion, atleast part of which was ethnic in character.
“dwellers i th
a
Ichas only been within the past decade, however, that
scholars have
gins of ethnic groups, and there isa growing belief that
ethnic groups arise ont of the of.
Some, such as Shennan (1989: 17), have argued that
there is no such thing as ethnicity “outside the orbic of
early states.” While it is certainly clear that there are
likely to hiave been ethnic distinctions in the orbits of
carly states (Smith 1986), itis equally clear that the ques-
tion is far from being settled. Most of these recent argu
ments focus upon an important facet of the definition of
ethnicity—the self-ceferencing recognition of personal
difference. Shennan (1989: 16), citing Bentley (1987),
notes that when powerful external forces begin to break
down traditional modes af organization, such as kinship
and faiily”siructure, individuals. begin to develop a
clearer sense of their own individuality,that is, an assess-
ment of similatties and differences based on personal
characteristics. :
‘Whatever the strengths or shortcomings of this theo-
retical position, it is possible to test the model, and
domestic architecture is of major importance in this
process. If this model is correct, there should be no de-
monstrable.cthnic differences in south-central Andean
reopls pris to the oeaiaton of the
polities (ca. 200 x
tecture may not be homogeneous across this
EGIOW Ta PRET iwanilew ies, Shout HEvETTIESS WOT
‘be used as an ethnic marker—that is, whatever variation
~“Giisis should Be ire isochrestic variation in form and
‘kes indicaitve oF caleures.
The essay by Wise most closely reflects this theme.
‘Through a detailed analysis of domestic architecture and
household refusé, she makes a strong claim that during
Late Intermediate Period times there existed a group of
lacustrine-focused peoples living on the lake margins in
{he Southern Titicaca basin. The style of construction of
the domestic architect of these peoples strongly re-
sembles the houses of thé Uru and Chipaya, modern Jake
“Although more datarare necessary,
ihe iimplication is strong that we are looking at ethnic
variation, What is even more Tnteresting is that there is
“the possibility that the lacustrine-focused economy and
the circular domestic architecture existed int pre-
sak times as well.
While domestic architecture is not a methodological
panaces for the south-central Andes or any other region,
it nevertheless offers distinct advantages to the archae-
cologist working with a variety of problems, ranging from
the organization of the household itself to the relation:12 Domestic Architecture, Household Archaeology, and the Past
ship of that household tothe politcal entities with which
itis afliated. There i good theoretical reason to believe
that domestic archirectute in fact represents ethnic varia~
tion; however, it must be stressed again that despite this
‘theoretical possibility, architecture must e empirically
demonstrated as monitoring eihiiciy aid not simply
assumed.
We believe that the essays assembled in this volume
demonstrate the promise of the use of domestic architec-
ture, and we hope they will stimulate other Andeanists to
consider this powerful category of data in their own
studies.