Western Snake River Prehistory

Daniel S. Meatte

(taken from the Prehistory of the Western Snake River Basin (1990) pp.63-70)



The archaeological sequence described in the preceding chapter is collapsed into a three-stage developmental model of settlement and subsistence systems as first proposed by Randall Schalk for the southern Plateau (1980:25-41), and later applied to most of the Columbia Plateau (Schalk and Cleveland 1983:23-44). The model is formulated to identify the basic evolutionary changes of adaptive systems as evidenced in the archaeological record (Schalk and Cleveland 1983:44-45).

This is accomplished by identifying relatively homogeneous subsistence and settlement patterns over broad geographic areas (Schalk and Cleveland 1983:45). For the Columbia Plateau, Schalk and Cleveland identified three major adaptive systems: (1) Broad Spectrum Foraging (11,500-4,200 years B.P.), characterized by mobile foragers who used simple tool inventories and exploited a wide variety of food resources; (2) Semisedentary Foraging (4,200-250 years H.P.), characterized by foragers who were able to extend residential stays during the winter months by storing foods. Archaeological evidence of these extended residential stays is characterized by the presence of housepits, storage facilities, diverse artifact assemblages, cemeteries, and increased reliance upon fish resources; (3) Equestrian Foraging (250-100 years H.P.), characterized by heavy reliance on horses as an efficient transportation method which permitted either fall and winter hunting in the upland ranges or the formation of large coordinated horse-mounted groups that pursued bison outside the local foraging range for extended periods of time (Schalk and Cleveland 1983:23-39).

Schalk and Cleveland have cautioned that this model is "imposed" on the archaeological record, causing some distortion of "long-term trends and processes" (1983:44). They further note that it (the model) does not imply complete regional uniformity; it simply focuses on homogeneous qualities discernible over wide geographic areas (Schalk and Cleveland 1983:45). This model was selected over numerous others (Butler 1986; Cressman et al. 1960; Jennings 1957; Madsen 1982; Nelson 1969, 1973; Sanger 1967; Thomas 1982; Zilverberg 1983) because it is designed for a specific regional system, the Columbia-Frasier Plateau, which includes the western Snake River Basin. In applying this model to the study area, all three adaptive systems were easily distinguished in the archaeological record. Further, the temporal developments of these adaptive systems were relatively synchronic with those described by Schalk and Cleveland for the Columbia Plateau.

BROAD SPECTRUM FORAGING (11,500-4,204) YEARS B.P. OR 9,550-2,250 B.C.)

Evidence for the arrival and subsequent "settling in" of the first inhabitants in the study area is difficult to characterize because of the sample size. The archaeological evidence itself is often subject to highly variable definitions and interpretations. Conservative and liberal interpretations of the archaeological evidence are commonly seen. The conservative interpretation is best exemplified by the work of C. Vance Haynes, who strongly argues for a minimal set of criteria by which all potential early man sites must be judged (Haynes 1969). When measured against these criteria, Haynes has found no evidence for man's presence in North America dating earlier than 11,500 years H.P. (Haynes 1980). In contrast, a number of proponents have marshalled evidence for man's presence in North America ranging from 19,000 years H.P. at Meadoweroft Rockshelter in southwestern Pennsylvania (Adovasio et al. 1983) to over 100,000 years B.P. for such sites as Calico in southern California (Simpson 1978).

A conservative interpretation of the cultural materials associated with the two early dates from Wilson Butte Cave would find the assemblage not acceptable as firm evidence of early peoples' presence in the study area at ca. 14,500 years H.P. The paucity of the artifactual remains, the pooling of the bone collection for a radiocarbon date, together with the difficulty in reconciling a temporal gap of nearly 4,000 years between the two early dates at Wilson Butte Cave (ca. 14,500 and 15,000 years B.P.) and the next oldest radiocarbon dates in the study area (ca. 10,000 years B.P. at Deer Creek Cave and 9,500 years H.P. at Dirty Shame Rockshelter), challenge the early age assessment. And comparing other early dated sites from the surrounding region (Fort Rock Cave ca. 13,000 years B.P.; Wasden Site ca. 10,000 to 12,000 years B.P. and Jaguar Cave ca. 11,580 250 years B.P.) still leaves a significant temporal gap to explain.

The assemblage from Wilson Butte Cave simply does not present a strong case for man's presence in the study area at such an early time. Until further evidence from other sites in the region dating ca. 14,000 to 15,000 years H.P. can be brought forth, together with evidence for the later intervening temporal gap (14,000 to 10,000 years H.P.), the early age assessments at Wilson Butte Cave will remain unfounded.

In general, this period is marked by an increasing diversity of diagnostic point styles through time: Clovis, Folsom, Windust, Haskett, Cascade, and Northern Side-notched. There are temporal overlaps and variable degrees of persistence in several of the styles, but at present, the archaeological chronology is too coarse to precisely stipulate the temporal arrangements of these point styles. Whether each style persists briefly and then gives way to another or there are a synchronic plurality of styles is not clear at this time.

Equally meager is our understanding of regional settlement and subsistence systems. It is assumed that peoples responsible for the manufacture of the distinctive Clovis, Folsom, Windust, and Haskett type points pursued a highly mobile lifestyle while procuring an assortment of food resources. The hunting of big-game animals such as elephant (Mammuthus sp), bison (Bison antiquus), mountain sheep (Ovis sp), elk (Cervus sp), camel (Camelops sp), horse (Equus sp), and deer (Odocoileus sp), could presumably have been complimented by a variety of small game animals, roots, tubers, and berries. These assumptions are based upon evidence from adjoining regions where similar cultural assemblages are better documented (Butler 1978b, 1986; Dort and Miller 1977; Frison 1978, 1983; Miller 1982, 1983).

More detailed evidence for the remaining portion of this time period is available from Dirty Shame Rockshelter, in southeastern Oregon. The rockshelter is situated at an elevation of 1,433 m asl in an arid, steppe-like upland setting characterized by expansive tablelands heavily dissected by deep, stream-cut canyons (Kittleman 1977:1-2). The early occupational record documents a time range from 9,500 to 5,900 years H.P. which attests to a light, but steadily increasing occupational use of the rockshelter. A generalized pattern of exploiting a broad range of locally available plant and animal food resources is evident. Mountain sheep, (Ovis canadensis), antelope (Antilopcapra americana), deer (Odocoileus cf hemionus), rabbits (Sylvilagus idahoensis and nuttallii), crayfish (Astacus), mussels (?), sagebrush (Artemisia tridenta), wild rose (Rosa fendleri), and wild cherry (Prunus emarginata) are just a few of the food stuffs that were consumed at this site (Hall 1977).

The inhabitants of Dirty Shame Rockshelter used a simple assemblage of tool types, including projectile points, bifaces, and various flaked tools (Richard Hanes, personal communication 1987). Ground stone milling slabs are added to the assemblage inventory at approximately 7,500 years B.P. Residential features are absent until 6,800 years B.P., when shallow grass-lined pits first appear. Contents of these lined pits indicates their use for caching of personal articles and tools, presumably for later reuse, and the building of latrines. Evidence of bulk food storage or differentiated processing or cooking areas is not present. The occupational record at Dirty Shame Rockshelter ceases at ca. 5,900 years B.P., when an occupational hiatus of 3,200 years occurred, and resumes at ca. 2,700 years B.P. (Aikens et al. 1977:21).

A similar assemblage composition for this time period is found at Wilson Butte Cave (Gruhn 1961a:4-6). The Wilson Butte II assemblage, estimated to date at approximately 8,000 years H.P., is composed of several lanceolate points, a number of simple flaked tools, and one groundstone mano (Gruhn 1961a:l18-119). Occupational density is light with no cultural features present (Gruhn l96la:118-119). Faunal remains are sparse and indicate bison (Bison bison), and camel (?) were the utilized food resources (Gruhn 1961a:127).

SEMISEDENTARY FORAGING (4,200 B.P. - 250 YEARS B.P. OR 2,250 B.C.-A.D. 1700)

The earliest evidence for the emergence of Semisedentary Foraging can be seen in a distinct artifactual assemblage termed the Midvale Complex (Warren et al. 1971), found throughout the northern portion of the study area. This complex, which is dated at 4,500 to 2,500 years H.P., displays a diversified artifactual assemblage profoundly different from the generalized artifact assemblages seen in the Broad Spectrum Foraging period. The assemblage is characterized by a broad array of functionally discreet tools, which are employed in a variety of specialized site types including quarries, workshops, hunting stations, campsites, and plant processing sites (Warren et al. 1971). Just a few of the diagnostic artifacts that typify this diversified artifactual assemblage include: large side-notched points, an assortment of scrapers including the distinctive, unifacial "elongate" scraper; drills; punches; gravers; conical mortars; elliptical mortars; cylindrical pestles; manos; awl sharpening stones, and edge-ground cobbles (Ruebleman 1973; Warren, et al. 1971;). An impressive mortuary complex termed the Western Idaho Archaic Burial Complex (Pavesic 1985) was recently added to the Midvale Complex. This complex attests to the use of large burial cemeteries in which enormous amounts of exotic material wealth (burial furniture) were placed (Pavesic 1985).

At the Givens Hot Springs locality (10-OE-1689), a large pithouse (House #2) dates well within the temporal span of the Midvale Complex at 4,200 years H.P. (F. Green 1982a:41, 1988:Table 1). The pithouse measures six and one half meters in diameter, has steep side walls, a central roof support, a central hearth, and is associated with hopper-mortar bases, Northern Side-notch, and Humbolt series points (T. Green 1982a:41). Green also reports the presence, at the Givens Hot Springs locality, of several other pithouses of similar size and configuration dating at 2,200 years H.P. (10-OE-1689:House #4); 1,270 years H.P. (l0-OE-60:House #3); 1,150 years B.P. (10-OE-60:House #1) and 1,100 years H.P. (10-OE-1691 :no house # reported) demonstrating the persistence of this style until relatively recent times (Davis and Green 1988; T. Green 1982a:41-43, 1988:Table 1). Similar house forms have been found at Montour Valley (10-GM-61) and date to approximately 3,118 years B.P. (Artz 1983:110).

Other house forms from the study area are documented from this time span as well. At the Givens Hot Springs locality, Green also found circular, saucer-shaped house floors dated at 4,300 years B.P., and post 2,400 years B.P. (T. Green 1982a:42-43). Similar house forms dating to the past 1,500 years B.P. have been reported at the Swan Falls site (Ames 1982b, 1983), Bancroft Springs site (Butler and Murphey 1982b; 1983), Three Island Crossing (Meyer and Gould 1987), Crutchfield site (Murphey and Crutchfield 1985), Hagerman National Fish Hatchery (Pavesic and Meatte 1980), and at Big Foot Bar (Plew 1980c). In general, these houses do not have excavated floors (just shallow, saucer-shaped profiles), lack a central roof support, and exhibit lighter weight construction.


The introduction of the horse to the Snake River Plain at approximately A.D. 1700 (Haines 1938) marks the advent of the final stage of the developmental model described here. Horse mounted Shoshone and Northern Paiute families were better able to extend their foraging range in pursuit of food and material resources. As Symmes Oliver has correctly noted, "The introduction of the horse created a different ecological situation which required new sociological arrangements" (Oliver 1974 :304).

These "arrangements" are best characterized as new sociocultural arrangements manifested themselves in the formation of composite bands of equestrian groups. Often ranging in size from 20 to more than 100 members these bands undertook communal hunting trips in pursuit of bison (Liljeblad 1957:40-43; Liljeblad 1972:14-15; Steward 1938:235). Long trips, in excess of a year, to the more productive grasslands on the eastern flanks of the Rocky Mountains were common. The communal hunts also allowed the bands to obtain and accumulate large quantities of material wealth. The greatly expanded foraging range, facilitated by the use of horses, enabled individuals to accumulate or trade for a broad spectrum of food and material goods. By the early 1800s, wide-ranging predatory bands were preying on unmounted families and the less protected Euroamericans (Layton 1978:135; Malouf and Findlay 1986:514; Shimkin 1986:519-520). Regional trade centers assumed a more dynamic appearance, allowing extensive trade of food and material resources (Anastasio 1972; Layton 1978, 1981; Sven Liljeblad, personal communication 1985).

Historic narratives and ethnographic surveys have provided a wealth of descriptive information about this time period. But archaeological evidence for the Equestrian Foragers within the study area is relatively rare, with very few sites dated to this period (Bonnichsen 1964; Crabtree 1968; Plew and Meyer 1987; Webster and Peterson 1974). Characteristic of these archaeological assemblages are artifacts of Euroamerican manufacture.

Back  Next