Nineteenth-century Apache wickiups

Nineteenth-century Apache wickiups:
historically documented models for
archaeological signatures of the
dwellings of mobile people
Highly mobile people must have sheltered in structures of some kind; but these are notoriously
difficult to find. The author uses nineteenth-century photographs of an occupied Apache settlement
to show how such shelters may have been made, comparing them with their archaeological remains
in the present day. This suggests a ‘signature’ for the temporary shelters used by mobile groups in
any period.
Keywords: American Southwest, Palaeolithic, Apache, wickiup, shelter, settlement, mobile
groups
Introduction
An important but challenging task of archaeology is the identification of ephemeral
structures, especially those temporary shelters erected by highly mobile people. Since the
first Late Palaeolithic structures were defined in the Châtelperronian Layers X-IX by André
Leroi-Gourhan in the Grotte du Renne, France (Leroi-Gourhan & Leroi-Gourhan 1964)
questions have arisen about their interpretation as cultural rather than natural features
and their classification as houses rather than living surfaces (Gamble 1986; Kolen 1999).
The identification of these shelters remains an important area of debate because of the
implications of intentionally modified space and the ‘making of place’ by early humans, and
for the origins of behavioural modernity (Zilhao 2006).
There are substantial differences in housing types between sedentary or semi-sedentary and
highly mobile groups. Theoretical studies have examined the relationship between the character of structural remains and type and degree of mobility (Sahlins 1972: 33; McGuire &
Schiffer 1983; Binford 1990; Diehl 1992; Smith 2003) – in general more mobile huntergatherers invest in less elaborate shelters (Upham 1994; Smith 2003: 163). Mobile people
tend to build domed circular or semicircular rather than rectilinear houses (Robbins 1966;
Whiting & Ayres 1968; Schiffer & McGuire 1983: 284; Binford 1990: 123; Diehl 1992).
Houses reflecting high mobility also tend to be situated on the surface and have informal
∗
The Southwest Center, 1052 North Highland Avenue, University of Arizona, Tucson, Arizona 85721, USA
(Email: [email protected])
Received: 30 October 2007; Revised: 8 September 2008; Accepted: 25 October 2008
ANTIQUITY
83 (2009): 157–164
157
Method
Deni J. Seymour∗
Nineteenth-century Apache wickiups
superstructures made of unmodified, easily obtained local materials (Binford 1990: 123,
126; Smith 2003: 172).
Highly mobile groups are known to have resided in the American Southwest, but until
recently there has been a general failure to identify evidence of their residential sites and
housing features. Even seventeenth-century Europeans in western North America noted
that the natives had ‘no houses, but only huts of branches’ (Ayer 1965: 13), illustrating the
way in which mobile group structures are devalued. Consequently, there has been either
a general expectation that no structures will be found or a focus on inapplicable material
signatures. The result has been that mobile group residential sites and their hut remains have
defied recognition. The structures are difficult to distinguish from natural clearings or rock
alignments because they represent the barest modification to the ground surface. Referred
to widely by the adjectives ‘enigmatic’, ‘subtle’, ‘unobtrusive’, and ‘ephemeral’ such features
often go unnoticed. Yet, as Binford (1990: 121-2) has observed: ‘There are no known cases
among modern hunter-gatherers where shelter is not fabricated in residential sites (anywhere
that hunter-gatherers plan to sleep), regardless of the expected occupational duration, and only
in rare instances are sites of any kind produced by hunter-gatherers where no shelter is provided
by the occupants.’ For this reason shelters should not only be expected on mobile group
residential sites but they should be widely distributed. Moreover, these types of structures
are commonplace and a routinely accepted feature type in the far western deserts in America
(Rogers 1939:8) where archaeologists have learned how to distinguish them.
One reason for the failure to identify the shelters is that researchers have relied on
historical analogies that are too broad. Housing correlates in one area (such as tipis on the
Great Plains) are inappropriately applied in another (the mountainous Southwest) where
social, natural and historical conditions differ and where differing types of mobility prevailed.
Although Apache hut signatures in the American Southwest are difficult to recognise, do not
meet accepted standards of classifications for ‘dwellings’, and often fall below the detection
threshold, hundreds have now been identified (e.g. Seymour 2002, 2004). One of the
most efficient ways to identify the archaeological traces of these structures, and so to begin
addressing how mobility affects construction, is to use historical examples with specific
known contexts. In this paper I enlist the help of illustrations and descriptions of structures
originally seen and photographed during Geronimo’s penultimate surrender at Cañon de
los Embudos (Embudos), Sonora, Mexico, which occurred on 25-27 March 1886. The
location of one of his last camps was photographically documented at the time by Camillus
Sidney Fly (Figure 1). Modern researchers have matched the distinctive characteristics of the
terrain shown in the photographs to locations on the ground (Hayes & Hayes 1991; Van
Orden 1991), and in this way the location of the dated camp has been precisely identified.
More importantly for our discussion, the structures in the encampment can be pinpointed
and studied.
The structures
These photographically documented, historically described features correspond to three
‘types’ of archaeological features and to four ethnographic ‘types’. These features match
the attributes typical of features previously referred to as ‘structural clearings’ or ‘sleeping
158
Method
Deni J. Seymour
Figure 1. Fly’s 1886 photograph no. 170 of an Apache wickiup, entitled ‘The captive white boy, Santiago McKinn’.
circles’, as well as ‘rock-ring huts’ and ‘modified boulder areas’, for example at the Cerro
Rojo Site, a Peloncillo Mountain site, and the Cochise-Howard Treaty Site (Seymour 2002,
2004, 2008b; Seymour & Robertson 2008). The ‘structural clearings’ or ‘sleeping circles’
consist of flat clearings on an otherwise rocky surface, where rocks are pushed out to form
the clearing, and may be slightly mounded or clustered around the perimeter. Except for
this light human touch no other evidence of modification is apparent, other than perhaps
artefacts and, rarely, an associated feature. Similarly subtle are the natural boulder outcrops
that have been slightly modified by the movement of stones or supplemented with cobbles
placed in intervening spaces forming ‘boulder-rimmed circles’ or ‘modified boulder areas’.
‘Rock-ringed hut’ circles are often the most ‘obvious’ with cobbles purposefully placed to
form a structure outline (Figure 2).
A documentary source specific to the Embudos site provides further details: ‘We moved
from one “jacal” to another, all being constructed alike of the stalks of the Spanish bayonet and
mescal and amole, covered with shreds of blankets, canvas, and other textiles’ (Bourke 1980: 476).
Close inspection of the images shows that these are not ‘jacals’ but rather are huts and that
ocotillo branches were also incorporated into some of the wickiup frameworks at Embudos
(Figures 1 and 3). Ocotillo, yucca and agave are also shown growing naturally and each is
present today, but of these only ocotillo branches are sufficiently supple to bend, as shown
in the framework of the wickiup in Figure 1. Some of the ocotillo branches so employed
were still rooted and were simply bent over to form one side of the structure, alleviating the
need for rock supports and accounting for breaks in rock alignments (Figure 3). In at least
two instances the head and leaves of dried agaves and yuccas were used, along with blankets
and other fabrics, to make the side walls. Some structures incorporated stiff robust yucca or
agave stalks to form conical tipi-like constructions. In most instances, untrimmed poles and
159
Nineteenth-century Apache wickiups
Figure 2. Archaeological hut traces from the Cañon de los Embudos site: a) structural clearing; b) cobble-rimmed area with
breaks in rock-ring structure; c) structural clearing at the location of the wickiup shown in Figure 1; d) modified boulder
area.
branches were incorporated as found, as were other materials at hand. In the structure shown
in Figure 1 the alignment of cobbles and boulders were leant upright against the ocotillo to
hold blankets and ocotillo branch bases in place. Some of these rocks were dispersed when
the blankets were removed and the house dismantled, thus rolling away from the feature
boundaries.
From an ethnographic standpoint these structures are not all constructed alike, but instead
show considerable variation, as is consistent with Diehl’s (1992) worldwide sample. All are
curvilinear rather than rectilinear, consistent with expectations for highly mobile groups that
have not been substantially influenced by modern technologies and materials (e.g. as have
Alyawara structures; O’Connell 1987). In this single encampment, house forms include
long elongated dome-shaped constructions, conical tipi-like constructions, small circular
squat conical tents and pup-tent-like shelters (if including the scout camp).
Clearly, there were many different hut forms that archaeologists might be inclined to
classify as different house types. The morphological variations (structural clearing, boulderrimmed circles and rock-ringed hut outlines; Figure 2) visible at Embudos and the region
in general were determined by the nature of the terrain itself and the prevailing conditions:
season of use including weather and temperature, anticipated duration of stay, group identity
or ethnicity, group size and function (Seymour 2002, 2008a, in press; see also O’Connell
1987). Rocky areas were sought to facilitate wickiup construction, provide additional shelter
160
Method
Deni J. Seymour
Figure 3. Fly’s 1886 photograph no. 173, entitled ‘Geronimo’s camp with sentinel’.
from the elements, and camouflage the encampment. If a surface was especially rocky, an area
would be cleared. If fewer rocks were present, rocks would be stacked or pushed together
around the perimeter. Under warm conditions, a less substantial superstructure would
be prepared, or in some cases no superstructure would be constructed (Polzer & Burrus
1971: 266, 273; Seymour 2002, 2004). In cold weather a more ample covering would be
constructed if natural rockshelters were not used instead, and in windy conditions the walls
were prepared more substantially to withstand gusts. Rocks might be placed on-end to lean
against the branches, holding blankets (formerly hides) in place. When conditions changed,
house form changed. Regardless of location and conditions, structures were improvised
using immediately available materials, circumstances and settings, which explains why these
features are most effectively identified in contradistinction to the nature of rocky surfaces
in their surroundings. Boulder-rimmed sleeping circles and those without boulders are
contemporaneous and, as Rogers (1966: 45) says the ‘construction and distribution of the two
forms are governed entirely by the nature of the terrain upon which they are found. They appear on
rocky land forms where coarse stones and even small boulders had to be removed from the surface
in order to produce the circular clearings. These stones were then used to build the confining rims,
or windbreaks.’ The wickiups at Embudos are represented today as rock-ring structures,
boulder-rimmed circles and structural clearings. Rock-ring structures are a feature type that
is more widely recognised for the Apache and other groups than are structural clearings
and boulder-rimmed circles (Goodwin n.d.: 37; Pillis 1981; Donaldson & Welch 1991;
Seymour 2002).
The similarity of these Embudos structural footprints today (Figure 2a-d) – nearly a
century and a quarter after this historical event – to those encountered in mobile Apache
and non-Apache sites throughout the southern Southwest many centuries old is striking.
161
Nineteenth-century Apache wickiups
Character of the settlement
Social and terrain-specific factors were influential in the layout of this residential site.
Contemporaneous wickiups in the same encampment, including at Embudos, may be widely
dispersed (several hundred metres apart), resisting archaeological boundary definitions.
Several reasons for this can be contemplated, including safety in dispersal under conditions
of impending attack, mother-in-law avoidance, separation for purification, or even rifts
between occupants (Seymour 2002). At Embudos and on other Apache sites in this region,
most wickiups are situated in or immediately adjacent to the rockiest areas showing the
operative effects of terrain. A simple undifferentiated distributional pattern is represented
at Embudos where structure location is determined by the terrain itself, rather than by
other factors, such as food sharing (e.g. O’Connell 1987: 100-2, 105). Food preparation
and consumption activities were most often situated apart from structures and there is
no discernable spatial correspondence between hearth and hut. In some instances paired
clearings are visible suggesting side-by-side huts or a wickiup and adjacent work area, as has
been noted on a number of other ancestral Apache sites (Seymour 2002). In other cases, one
wickiup is about twice the size of others, perhaps providing an alternative to preparation of
two individual features. In still other instances a single wickiup is visible but many outdoor
areas were used.
In environments with favourable weather many activities and temporary storage areas were
positioned outside, around the structure, and in distant locations that provided suitable space
for tasks, also accounting for the lack of hearths inside. Placement of wickiups among the
rocks allowed enlistment of the maximal functionality of the location so that equipment
could be scattered on the rocky surface and draped on rocks (as shown in Figure 3). The
common absence of hearths in the shelters and the separation of work areas from them
provide some indication of activity organisation. While shelters provided secluded space for
sleeping, privacy, storage, corralling children, or protection from the elements, they do not
seem to have contained the full range of household activities commonly associated with
‘dwellings’. Many of these same activities and others, like food preparation, were probably
carried out in areas distant from the houses, in the shade, in areas designated for visitors, on
flats adjacent to the hut, and in locations inherently suitable for the desired tasks.
Artefacts include flaked and ground stone and non-native materials (glass and metal)
that were used and modified by the Apache. The association of artefacts is one of the
most widely accepted criteria for designating a clearing or alignment as a cultural feature.
Although artefacts are scattered lightly across the Embudos site, they are found in and near
only some structures, indicating that this criterion for feature designation is far too limiting
for mobile people who may select work areas that are many tens or hundreds of metres
from their huts. Other indices are more reliable, such as consistent and verifiable evidence
of modification (surface flattening; clearing, stacking, arranging or removal of rocks).
Conclusions
It is widely recognised that high mobility results in inconspicuous material culture imprints,
little accumulation and low archaeological visibility (Binford 1980: 7). This essay has
162
Deni J. Seymour
References
LEROI-GOURHAN, A. & A. LEROI-GOURHAN. 1964.
Chronologie des grottes d’Arcy-sur-Cure (Yonne).
Gallia Préhistoire 7:1-64.
MCGUIRE, R.H. & M.B. SCHIFFER. 1983. A theory of
architectural design. Journal of Anthropological
Archaeology 2: 277-303.
O’CONNELL, J.F. 1987. Alyawara site structure and its
archaeological implications. American Antiquity 52:
74-108.
PILLIS, P.J., JR. 1981. A review of Yavapai archaeology,
in D.R. Wilcox & W.B. Masse (ed.) The
protohistoric period in the American Southwest AD
1450-1700 (Arizona State University
Anthropological Research Papers 24): 163-82.
Tempe (AZ): Arizona State University.
POLZER, C.W. & E.J. BURRUS. 1971. Kino’s biography of
Francisco Javier Saeta, S.J. (Sources and Studies for
the History of the Americas 9). St. Louis (MO):
Jesuit Historical Institute, St. Louis University.
ROBBINS, M.C. 1966. House types and settlement
patterns: an application of ethnology to
archaeological interpretation. Minnesota
Archaeologist 28(1): 3-26.
ROGERS, M.J. 1939. Early lithic industries of the Lower
Basin of the Colorado River and adjacent desert areas
(San Diego Museum Papers 3). San Diego (CA):
San Diego Museum.
–1966. Ancient hunters of the far west. San Diego (CA):
Union-Tribune Publishing Company.
SAHLINS, M. 1972. Stone Age economics. Chicago &
New York: Aldine-Atherton.
SEYMOUR, D.J. 2002. Conquest and concealment: after
the El Paso Phase on Fort Bliss. Report prepared for
Conservation Division, Directorate of
Environment, Fort Bliss, Texas. El Paso: Lone
Mountain Report 525/528.
AYER, MRS. EDWARD E. (trans.) 1965. The memorial of
Fray Alonso de Benavides. Albuquerque (NM): Horn
& Wallace.
BINFORD, L. 1980. Willow smoke and dog’s tails:
hunter-gatherer settlement systems and
archaeological site formation. American Antiquity
45: 4-20.
–1990. Mobility, housing, and environment:
a comparative study. Journal of Anthropological
Research 46: 119-152.
BOURKE, J.G. 1980 [1891]. On the border with Crook.
New York: Charles Scribner’s Sons.
DIEHL, M.W. 1992. Architecture as a material correlate
of mobility strategies: some implications for
archaeological interpretation. Cross-Cultural
Research 26: 1-35.
DONALDSON, B.R. & J.R. WELCH. 1991. Western
Apache dwellings and their archaeological
correlates, in P.H. Beckett (ed.) Mogollon V :
93-105. Las Cruces: Coast Publishing &
Research.
GAMBLE, C. 1986. The Paleolithic settlement of Europe.
Cambridge: Cambridge University Press.
GOODWIN, G. n.d. Unpublished materials on
western Apache dwellings and shelters, Ms. 17,
Folder 54. Tucson (AZ): Arizona State Museum
Archives.
HAYES, A.C. & K. HAYES. 1991. Sierra Madre revisited.
Journal of Arizona History 32(1): 125-52.
KOLEN, J. 1999. Hominids without homes: on the
nature of Middle Palaeolithic settlement in Europe,
in W. Roebroeks & C. Gamble (ed.) The Middle
Palaeolithic occupation of Europe: 139-75. Leiden:
University of Leiden.
163
Method
explored what these types of constructions can look like archaeologically, drawing on
the analogies provided by nineteenth-century wickiups. Highly mobile groups in warmer
climates and varied physiographic environments (such as the Southwestern basin-and-range
province) tend to seek terrain sectors (such as elevated rocky projections and rocky slopes)
that facilitate successful improvisation, including shelter construction, while maximising
security and other factors. This means that no two houses will be exactly alike (in a typological
sense), but they will be recognisable as archaeological features because they exhibit consistent
and verifiable forms of terrain modification. Highly mobile groups that do not transport
their shelters focus on terrain characteristics that allow them to shelter themselves with the
least effort, using materials at hand rather than importing materials from afar. Residentially
mobile groups, from Palaeolithic people to the historic hunter-gatherer-raider Apache, use
the terrain as it is, rather than inflicting substantial modifications. This means their houses
are difficult to see, but real nonetheless.
Nineteenth-century Apache wickiups
–2004. A Rancherı́a in the Gran Apacherı́a: evidence of
intercultural interaction at the Cerro Rojo site.
Plains Anthropologist 49(190): 153-92.
–2008a. Surfing behind the wave: a counterpoint
discussion relating to ‘A Rancherı́a in the Gran
Apacherı́a.’ Plains Anthropologist 53(206): 241-62.
–2008b. Despoblado or Athapaskan heartland:
a methodological perspective on ancestral Apache
landscape use in the Safford area, in D.E. Purcell
(ed.) Crossroads of the Southwest: culture, ethnicity,
and migration in Arizona’s Safford Basin: 121-62.
New York: Cambridge Scholars Press.
–In press. Distinctive places, suitable spaces:
conceptualizing mobile group occupational
duration. International Journal of Historical
Archaeology.
SEYMOUR, D.J. & G. ROBERTSON. 2008. A pledge of
peace: evidence of the Cochise-Howard Treaty
campsite. Historical Archaeology 42(4): 154-79.
SMITH, C.S. 2003. Hunter-gatherer mobility, storage,
and houses in a marginal environment: an example
from the mid-Holocene of Wyoming. Journal of
Anthropological Archaeology 22: 162-89.
UPHAM, S. 1994. Nomads of the desert West: a shifting
continuum in prehistory. Journal of World Prehistory
8(2): 113-67.
VAN ORDEN, J. 1991. Geronimo’s surrender: the 1886
C.S. Fly photographs (Museum Monograph 8).
Tucson (AZ): Arizona Historical Society.
WHITING, J.W. & B. AYERS. 1968. Inferences from the
shape of dwellings, in K.C. Chang (ed.) Settlement
archaeology: 117-33. Palo Alto (CA): National Press.
ZILHAO, J. 2006. Neanderthals in the Levant:
behavioral organization and the beginnings of
human modernity edited by Donald O. Henry.
Journal of Field Archaeology 31(1): 117-20.
164