[NPS Arrowhead] U.S. Dept. of Interior National Park Service Archeology Program
Quick Menu Features
* Sitemap * Home
  Kennewick Home Cultural Affiliation Report  

Chapter 2
Review of the Archaeological Data
Kenneth M. Ames

The Southern (Columbia) Plateau: Background


Many of the topics of importance to this study are covered in detail in the papers in The Handbook of North American Indians, Volume 12, The Plateau, edited by Deward E. Walker Jr. (Walker 1998). The reader of this report is referred to that volume if they wish more background than given here.

Environments (cf. Chatters 1998 and references)

The Columbia Plateau (or the Southern Plateau [Figures 1 & 2]) is flanked on the west by the eastern slopes of the Cascade Mountain range, on the east by the Northern Rocky Mountains, on the north by the Okanogan Highland of northern Washington State and southern British Columbia, and on the south by the Southern Uplands, the Blue – Ochoco Mountain system of central Oregon. The central feature of the Plateau is the Columbia Basin, a depression in the center of the region filled with vast sheets of Miocene basalt, the results of multiple, volcanic eruptions. This lava plain tilts down to the west and is lowest along the eastern flanks of the Cascades. It rises gradually to the east to more than 600 meters (2000 ft). The region's rivers, including the Columbia River, the Snake River and their tributaries, are entrenched in deep canyons in the northern and eastern portions of the Plateau, but flow through relatively low terrain in some parts of the southwestern portions. The Columbia River is pinned against the Cascades by the basalt flows.

The Plateau has a continental climate, with cold winters and hot summers. Rainfall comes from the west, off the Pacific Ocean. Thus, the very low western portions of the Plateau are the driest because they lie in the Cascade's rain shadow (Figure 3). Rainfall increases to the east. The uplands flanking the basin are forested, while the basin itself is covered with a shrub and bunch grass steppe. This bald description masks great ecological diversity, particularly in the wetter and higher portions of the region (Figure 4).

Mammals of economic importance in the past include elk (Cervus elaphas), mule deer (Odocoileus hemionus), white-tailed deer (Odocoileus virginianus), bison (Bison bison), mountain sheep (Ovis canadensis), mountain goat (Oreamnos americanus), and pronghorn (Antilocapra americana). The region is most famous, of course, for its once prolific salmon runs. Three of the Northwest's five salmonid species were present in the Columbia and Snake River systems: the Chinook (Oncorhynchus tschawytscha), sockeye (O. nerka), and steelhead (O. mykiss). Salmon are anadromous, born in fresh water, travel to the ocean after a time, grow to adulthood in the North Pacific, and then return to the stream where they were born to spawn. Unlike Atlantic salmon, Pacific salmon die upon spawning. Other economically important fish include sturgeon (Acipenser transmontanus), lamprey eels (Entosphenus tridentatus), suckers (Catostomus sp.), and other resident fish. A wide range of plant material was harvested for an equally wide array of purposes. Some of this is discussed separately below (Section 6.4.)

Archaeologists on the Plateau have been considerably interested in reconstructions of paleoenvironments. Despite that, I do not attempt one here, nor do I offer a summary. This is in part due to the time constraints of this study. It also does not seem necessary for the purposes of this document. Chatters (1994, 1998) offers a recent one.

Anthropological and Archaeological Research

Several good accounts of the history of anthropological and archaeological research on the Plateau have been recently published (e.g. Campbell 1989, 1991; Reid 1991a, Lohse and Sprague 1998). The reader is referred to those works for detail, analysis1, and sources on that history. Only a few points need to made here. The vast majority of all archaeological work on the Plateau has been CRM–related, even before there was a concept of CRM. While fieldwork on the Plateau began well before World War II, its real impetus was post-war dam construction, and the resulting River Basin Surveys of the 1950s. The great majority of projects since the 1950s has been related to dams and reservoirs. Within the last 25 years work has expanded out of the canyons and river bottoms. Virtually all of this work is also CRM related in the form of Forest Service projects, pipeline projects, etc. An impressive body of evidence has built up, but it has limitations. Excavations in the canyons, for example, focus on pithouse sites, and on the house pits themselves. We have, therefore, far more information about the contents of the structures than we do for exterior activity areas. While the earlier work often involved extensive excavations of a few major sites, more recent work has tended to be survey and tests of a wider range of sites. Clearly that corrected an earlier bias in the data, but creates one of its own if we do not have comparable samples.

While work has been extensive, four major projects have profoundly shaped what we currently know about Plateau archaeology. In the early 1960s, faculty and graduate students from Washington State University began work in proposed reservoirs in the Lower Snake River Region. This work continued into the mid-1970s along the Lower Snake and Lower Clearwater Rivers. It produced the longest and for a time, best-documented, multi-site chronology in the Southern Plateau. In the late 1970s, a massive, three season project based at the University of Washington excavated 18 sites behind the Chief Joseph Dam on the Upper Columbia, developing what is still probably the region's best, large modern data set. It is certainly among the best documented. Within a year's time of the completion of the Chief Joseph Project, an additional 13 sites in the Wells Reservoir were excavated by a consortium including Central Washington University, Washington State University, and University of Washington. These sites were located near the confluence of the Okanogan River and the Columbia River. Although methods differed, these two projects produced a data set of 31 excavated sites in the same general region. The fourth and most recent project was performed by INFOTEC, Inc, along a pipeline expansion route that ran through California, central Oregon, eastern Washington, and northern Idaho. The Oregon data in particular have been extremely important. This is not to suggest that there have no been other projects, including surveys and excavations of single and multiple sites to have made major contributions to Plateau archaeology. There have been, many of high caliber. Among these are a series of excavations and surveys across the Plateau accomplished by Eastern Washington University. However, these four data sets are the major, large blocks of data to which everyone has recourse.

The focus of early work on the Plateau (beginning in the 1940s through the end of the 1960s) was on establishing the time depth of human occupation of the Plateau, and when the "Plateau Pattern" – the archaeological manifestation of the ethnographic reconstruction (Ray 1939) of Plateau culture – began. After the 1960s, research shifted away from explicitly historical questions and focused much more heavily on questions about the evolution of hunter-gatherer adaptations, in line with broader changes in the discipline. What is important here, however, is that neither approach produced data that lend themselves to an inquiry such as this one. The earlier work depended primarily on looking for artifacts thought to indicate the Plateau Pattern. These included pithouses and some artifact types. However, these were treated as "indicator fossils." Only one was needed to document the pattern. Thus, there was not as much concern with quantification as we now think basic, nor was there much interest in artifacts not thought to be culturally sensitive. For example, Nelson (1966), in his report on the Tucannon site, only lists ground stone tools generally, and does not give counts. The later, hunter-gatherer oriented work, was little interested in historical relationships and so the taxonomic and stylistic studies of artifacts, which might yield such data, did not occur.

One last issue to be touched on here is simply the availability of the database. As someone who has a final report awaiting completion now going on 17 years, I cannot complain too loudly. I would merely support Lyman's point (Lyman 1985b, 1997) that the literature of the Plateau is very difficult to get, because so much of it is published (or not) in limited distribution reports, or in venues which are obscure, or difficult to even to find out about. Therefore, this study is limited by our success in tracking reports down rapidly, and getting them from the SHPO libraries or interlibrary loans. In many cases, we could not locate reports.

Archaeological Background

The known archaeological record for the southern Columbia Plateau spans a period of perhaps 13,500 calendar years. Archaeologists have approached this record over the past 50 years or more from a variety of theoretical standpoints and basic assumptions. During that time, they constructed a number of cultural chronologies for the Plateau, most specific to a particular site, reservoir, or sub-drainage. There is no single, unifying sequence (but see Chatters and Pokotylo 1998). This in part reflects the piecemeal nature of archaeological research on the Plateau, but also the considerable archaeological and environmental variability across the region. In any case, it is far beyond the scope of this study to review these sequences in detail. A recent synthesis of sequences for the Southern Plateau (Ames et al. 1998) is summarized in Figure 5 and a somewhat more detailed synthesis (Galm et al. 1981) is presented in Figure 6.

It is important to understand that these phases and periods were often developed using very different methodologies and samples. For example, the construction of the Lower Snake River sequence was based on quite explicit typological procedures (Leonhardy and D. Rice 1970, D. Rice 1972, Bense 1972) that included, among other things, not using archaeological assemblages that were thought to represent mixes of assemblages from different phases. This had the effect of eliminating potential temporal and spatial variability. Additionally, the phases were defined based on geological context, and the presence and absence of certain artifacts (e.g. Cascade points), lithic technology, and other attributes. In contrast, the phases for the Chief Joseph Reservoir sequence (Kartar, Hudnut, Coyote Creek) in the South-central Plateau were defined, at least in part, on the basis of breaks in the distributions of radiocarbon dates (Salo 1985). A sequence proposed by Schalk and Cleveland (1983) was based entirely on subsistence and mobility strategies.

Despite this variation in sequences, workers on the Plateau generally see three broad periods. The divisions between the periods may be placed at somewhat different times (e.g. Figure 5) depending on the scope of the sequence (in Figure 5, the Chatters and Pokotylo sequence is for the entire Plateau, while the Ames et al. sequence is only for the Southern Plateau), and the interests of the researcher.

A Summary Chronology (based on Ames et al. 1998, Figure 5)
Period I (11,500 BC to 5000/4400 BC):
This period is divided into a Period IA and IB. IA is Clovis2, which is weakly represented and outside the temporal parameters of this study. Subperiod IB was described as being "post-Clovis" although more recent evidence discussed below suggests that the earliest IB cultures of the region may be contemporary with Clovis. This period is discussed in greater detail in Section 3. Some of the characteristics of the period are:

  • Very low population densities;
  • High levels of mobility;
  • Subsistence orientation emphasizing relatively mesic environments;
  • Early artifact assemblages (pre–7000 BC) marked by the presence of stemmed and shouldered lanceolate projectile points. These points display:
    1. Wide bases relative to blade size;
    2. Edge grinding of the stems;
    3. Highly variable blade shape because of resharpening and reworking;
  • Later assemblages (post–7000 BC) are dominated by foliate, or leaf shaped points (Cascade points), although these forms were present in small numbers earlier;
  • Lithic assemblages dominated by cherts;
  • During the later portions of the period, tool stone sometimes includes significant frequencies of fine–grained basalts;
  • Lithic reduction includes manufacture of macroblades and flakes from prepared cores;
  • Lithic technology during later portions of the period includes some instances of the Levallois prepared core technique;
  • The presence of burins;
  • A bone tool technology that includes small bone needles and antler wedges, as well as barbed points;
  • Fishing gear includes very rare net weights;
  • The presence of "bola" stones, small, girdled pebbles that some (e.g. Carlson 1996) suggest might be a form of netweight;
  • Hunting of a range of large (including bison) and medium (including rabbits) mammals; some evidence for salmon fishing, no evidence for storage;
  • Plant exploitation is suggested by the presence of small milling and hand stones, and, particularly after c.7000 BC, edge ground cobbles;
  • Evidence for temporary shelters, including windbreaks and huts;
  • After the Mazama ash fall, assemblages contain large side–notched projectile points (Northern Side Notched) that predate the ash fall in sites to the south;
  • After the Mazama ash fall, some assemblages in the South-central Plateau contain microblades and microblade cores. Assemblages elsewhere in the southern Plateau do not;
  • At some time during this period, the central Columbia Basin was abandoned.

Period II (5000/4400 – 1900 BC):
Ames et al. (1998) note in that in some portions of the Southern Plateau, particularly the southwest, this period differs little from the preceding Period I. However, in other areas, there is considerable change:

  • Pithouses are present in the Southeastern and South–central Plateau by c. 4000 BC, if not earlier;
  • These structures occur both in the river canyons and the southern uplands;
  • The houses are associated with substantial deposits, indicating rather long periods of occupation;
  • Mortars and pestles are present, some are massive in size. They are sometimes present in large numbers;
  • Projectile points include a variety of stemmed, and corner and side-notched forms;
  • Chipped stone technology sometimes lacks the investment of time and skill evident in previous periods; reduction techniques are opportunistic, although bifacial cores occur;
  • There is a variety of well–made bone tools, including large needles and leister parts.
  • Decorated bone objects are present;
  • Mobility strategies associated with these structures are not clear and subject to debate;
  • There is no obvious evidence for storage;
  • While a range of mammals was taken, medium–sized mammals (e.g. rabbits) were not. The degree to which the subsistence economy focused on fish and/or roots is a matter of debate. What seems clear is that subsistence was significantly different than during previous periods.
  • Settlements seem to have been small, with few contemporaneous houses.
  • There appears to have been sporadic use of the central Basin. Upland areas were used for a wide range of activities.
  • By the end of the period, there appears to have been a brief but virtual cessation in the construction of pithouses across the region.
  • Presence of the Western Idaho Burial Complex in the far Southeastern Plateau and perhaps in other areas.

Period III (1900 BC – AD 1720):
This period is marked by a number of changes:

  • The widespread presence of pithouses (which had virtually disappeared before the end of Period II), with increased variation in size;
  • The apparent appearance of mat lodges after AD 500;
  • Intensive exploitation of camas and probably other roots;
  • Collector mobility strategies that show continuity into the Historic Period;
  • Large settlements, and concentrations of houses after AD 500;
  • Ubiquitous evidence for fishing, particularly with nets;
  • Widespread evidence for storage, including storage pits and storage caves;
  • Evidence for intensive use of salmon;
  • Evidence for increased populations;
  • Use of the central Columbia Basin, and expanded use of other portions of the Plateau;
  • Presence of basketry, fiber, and wood artifacts in the record;
  • Small projectile points indicate the presence of the bow and arrow. However, atlatls continue in use for a considerable period, until about AD 1000;
  • Appearance of cemeteries associated with house pit villages at c. 500 BC (and the disappearance of the Western Idaho Burial Complex).

Modern Period (c. AD 1720 – present):
This is not a period used in most previous discussions of Plateau prehistory (but see Ames 1991, Ames and Maschner 1999). The Modern Period extends from the appearance of the horse to the present. It is divided into two subperiods: the Early Modern, which spans the time from the appearance of the horse (c. AD 1720) to the establishment of reservations (c. 1850), and the Late Modern, which extends from the beginning of the reservation era to the present day. Characteristics of the Early Modern period are discussed below in the section entitled "Archaeology of the Early Modern Period."

Much of this sequence is discussed in detail in the sections that follow. Here I briefly review some of the broader, cultural historical frameworks that have been proposed to explain the Plateau's archaeological record. I also will point out a few places where archaeologists have seen continuities, discontinuities and gaps in the record, so as to keep them in mind as we review that record.

The Intermontane Western Tradition (Daughtery 1962)
Daugherty presented a model of cultural development for interior western North America (between the Rocky Mountains and the Sierra Nevada–Cascades) that postulated a Desert Culture–like (Jennings 1957) basal culture for the region's initial inhabitants. Subsequent culture changes were gradual and proceeded as traits were added or subtracted from the repertoire. As time passed, the basal culture gradually differentiated into the cultures of the Plateau, Great Basin, and Southwest in response to environmental differences and the influences of adjacent culture areas. Daugherty saw no discontinuities in this record.

The Old Cordilleran Culture (B. R. Butler 1961, 1962, 1965)
B. R. Butler postulated that the Pacific Northwest was initially occupied by a cultural tradition whose subsistence focus was on the foothill/mountain region (Cascades), hence the name "Old Cordilleran." The type artifact was the foliate Cascade point. He subsequently added edge–ground cobbles to the original definition. As originally defined, the Old Cordilleran extended into northern South America, but later discussions effectively limited it to the Pacific Northwest. B. R. Butler was not concerned with how the Old Cordilleran was related to subsequent or the Early Modern cultures in the region. A number of workers have recently started applying the term to a variety of early cultures which are marked by the presence of a range of foliate–shaped points and bifaces, bone tools, and a generalized hunter-gatherer subsistence base (e.g. Bense 1972, Matson and Coupland 1995, Dixon 1999). Daugherty (1962) saw it as an area co-tradition with the Intermontane Western Tradition that was supplanted or absorbed by the latter.

"The Emergence of the Plateau Pattern" (Swanson 1962a, Nelson 1969, 1973; Smith 1977)
Swanson postulated that Plateau culture, as described by Ray (1939) did not emerge until AD 1300 or so, having been preceded by what he described as a forest–hunting culture. The emergence of Plateau culture was fueled by "a quickening pulse" which included increased trade and contacts with the coast and by climate change. His thinking stressed the differences between the preceding forest hunting culture and Plateau culture. Nelson refined and reworked Swanson's ideas, using Nelson's excavations of the Sunset Creek site. Nelson postulated that the key event in the development of Plateau culture, which he dated to about AD 1, with the beginning of his Cayuse phase (Figure 6), was the expansion of Salish speakers from the southern Northwest Coast across the Cascades to their present position straddling the international boundary (Nelson 1973). While the evidentiary basis for Nelson's hypothesis (the relatively older ages of pithouses in British Columbia than on the southern Plateau) has been refuted, his Salish–expansion hypothesis has been recently revived on other grounds by Smith (1977). His evidence will be reviewed below.

"The View from Wenas" (Warren 1968)
Warren described Plateau prehistory as a series of patterns which evolved one into the other, based on the direction from which cultural influences flowed into the Plateau. His theory, then, was essentially diffusionist, in which the influences of other regions spread into the Plateau, and either mixed with, or replaced the traits of preceding periods. His model has had little impact, except that Browman and Munsell (1969) combined it with Daugherty's Intermontane Tradition to produce the only synthesis of Plateau archaeology to be published, until recently, in a venue of wide distribution. Browman and Munsell's formulation of Plateau culture history has also had virtually no subsequent effect on the region. It was published at a time when the interests of archaeologists were shifting from traditional culture history to more processual forms of archaeology, and only one year before the publication of Leonhardy and D. Rice 's sequence for the Lower Snake (Leonhardy and D. Rice 1970) that almost by default became the master sequence for the southern Plateau for many years.

The Lower Snake River Sequence (Leonhardy and D. Rice 1970) (Figures 5 & 6)
The Lower Snake River sequence developed by Leonhardy and D. Rice included the Snake River between its confluence with the Clearwater River and its confluence with the Columbia River. The sequence covers the entire known chronology for that region, and it is still the only sequence for the southern Columbia Plateau that spans the entire Holocene. After its initial publication, a number of doctoral dissertations and master's theses have been done to examine particular phases in detail, and to flesh out the sequence (Leonhardy 1970, D. Rice 1972, Bense 1972, Kennedy 1976, Brauner 1976, Hammatt 1976, Yent 1976, Lucas 1994, Harder 1998). Sappington (1994) essentially extended the sequence up the Clearwater, with some modifications, and workers in Hells Canyon (e.g. Reid 1991a) have used it as a master sequence.

The chronology was derived from Daugherty's Intermontane Western Tradition, but with some modifications. Leonhardy and D. Rice (1970) divided their sequence into three traditions: the Pioneer (including Windust [11,000 – 7000 BC]3 and Cascade [7000 –4500 BC] phases), the Initial Snake River (Tucannon phase [4500 – 500 BC]), and the Snake River (Harder [500 BC – AD 1000], Piqunnin [ AD 1000 – 1720] phases). Their Ethnographic Tradition (Numipu phase) was a continuation of the Snake River tradition. They saw a break between the Pioneer and Initial Snake River traditions. This break was marked by:

  • A change in lithic technology, with a decline in quality in Tucannon times;
  • A shift away from the use of basalt as tool stone;
  • The appearance of a range of small stemmed and large side and corner notched points and other developments.

They saw cultural continuity from the Harder (500 BC – AD 1000) phase on. In fact, after their initial proposal, they subsequently suggested that the Piqunnin phase (AD 1000 – 1720) be dropped (Leonhardy and D. Rice 1980) because they saw no significant changes from c. AD 500 to the appearance of the horse. The Early Modern Numipu phase (post – AD 1720) was essentially Harder phase people with the horse: i.e. Nez Perce. Considerable subsequent work (Kennedy 1976, Ames 1984, Lucas, 1994) has focused on the Tucannon phase (4500 – 500 BC), since it was, and remains, the fuzziest of these phases.

Recent Cultural Historical Frameworks: Carlson (e.g. 1983,1996,1998)
More recently, Carlson has argued that the Pacific Northwest, including the Plateau and the Northwest Coast, were initially occupied by three different cultural traditions: the Microblade Tradition on the northern Northwest Coast, the Stemmed Point tradition on the Plateau (Lind Coulee/Windust) and the Pebble Tool tradition. An earlier Fluted Point tradition (Clovis) may also have been present. These traditions are seen as distinct before 7000 years ago, though some sites have overlapping characteristics of both. Of interest here is that the Pebble Tool tradition includes what B. R Butler (1961). called Old Cordilleran, and includes foliate points and bifaces, cobble tools and net weights. It is thus what workers on the Plateau call Cascade. His division of the early materials from the interior into a pebble tool tradition and the stemmed point tradition would imply different origins for Windust and for Cascade. Dixon (1999) provides a somewhat different organization of these materials.

Recent Settlement and Subsistence models
Schalk and Cleveland (1983) presented a sequence of what they termed hunter-gatherer land use strategies in the Pacific Northwest. They recognized three periods, based on land-use practices and subsistence activities: broad–spectrum foraging, semi–sedentary foraging, and equestrian foragers. The first spans the period from the earliest occupants to the appearance of pit houses, the second from the earliest pit houses until the introduction of the horse, and the last, the post-horse period until the reservation period. Although Reid (1991a) has questioned the utility of these broad periods, this represents the first attempt to order Plateau archaeology on some basis other than temporally sensitive artifacts and perceived similarity to the "Plateau Pattern." Chatters (1995) has developed a second such model, and it is the most germane here. Both Chatters (1989, 1995) and Ames (1988a, 1991) have observed that there were gaps in the radiocarbon dates for pit houses on the Plateau, particularly between 2400 B.C. and 1600 B.C. There are a very few dated houses in this period, but far fewer than previously or subsequently. There are also contrasts in house form, size, variability and associated mobility patterns before and after those dates. Chatters (1989) proposed that these differences reflect the establishment of a form of sedentism around 3000 B.C. marked by the presence of pithouses (Pithouse 1). This pattern lasted until about 2400 B.C. when it was widely abandoned (both events results of climate changes). A different form of sedentism (Pithouse 2), accompanied by collector mobility strategies was established around c. 1600 B.C., a pattern that then lasted with some changes until contact. Ames (1991) suggests more fluctuations in mobility patterns over the past several thousand years than does Chatters. In any case, Chatters posits three abrupt, region-wide shifts in settlement patterns between 3000 B.C. and 1600 B.C.

Different researchers have seen different patterns in the region's archaeological record. However, what emerges from the summary chronology and the review of cultural-historical frameworks is a series of general issues that need to be addressed in this report:

  • Relationship of Cascade to earlier/subsequent manifestations;
  • Relationship of Tucannon to earlier and later manifestations;
  • Change to Pithouse 1 and to Pithouse 2;
  • Microblades in South–central Plateau;
  • Evidence for warfare and conflict (Smith 1977/Chatters 1988);
  • Possibilities of abrupt subsistence changes;
  • Changes in material culture (shifting projectile point styles, disappearance of edge-ground cobbles, appearance of large grinding tools, etc.).

This list is not exhaustive, and other shifts need to be addressed. However, these do provide a framework within which to examine the record. The next section also is intended to provide a framework for this study, but instead of looking specifically at the Plateau, issues of continuity/discontinuity and gaps in the record are addressed more generally.

1Campbell 1991 is particularly useful for this study.
2Recently, several researchers, including Grayson (1993), Beck and Jones (1997) and Dixon (1999) have applied the term "Western Fluted" to materials west of the Rocky Mountains that has previously been termed Clovis. In the Northwest, the materials recovered from the Wenatchee (or Richey-Roberts) cache appear to be classic Clovis, while the Dietz site materials from southcentral Oregon would be termed Western Fluted.
3Non–archaeologist readers of the final rough draft of this report found it difficult to keep track of all the phases names and their ages. Therefore, phase names will be followed by their age range throughout the rest of the report. The age ranges are those used in Figure 5, which are based on calibrated dates.

Kennewick Home | Ames Home | Next Section >>