Prior to initiation of the survey projects that are the subject of this report, only 25 prehistoric sites and 33 historic sites and trails had been identified in the Drewsey Resource Area. No sites within the area have been the subject of intensive excavations, but seven sites have received minor testing. These and other sites have been surface collected with some analysis of the data. (BLM 1979, Fagan 1974 and Pettigrew 1979.)
One of the sites examined was part of a larger research project by Fagan (1974). The other six sites that were tested, along with four additional sites from which surface collections were taken, were analyzed by Pettigrew (1979) as part of the Stinkingwater Pass Highway relocation project. Both of these projects are discussed in further detail below.
In order to understand Fagan's (1974) or Pettigrew's (1979) perspective, we will need to examine some of the earlier regional frameworks that have been developed in nearby areas. From a synthesis of these approaches we may develop a framework from which to examine the data presented here.
The Drewsey Resource Area lies on the northeastern fringes of the Great Basin Cultural Area, which in this area includes the northern portions of the Basin and Range physiographic province, the High Lava plains, and parts of the Blue Mountain and Owyhee Upland physiographic provinces. To the north and east of the Drewsey Resource Area is the Columbia Plateau cultural area. The Drewsey area then, especially in the northern part, is peripheral to both cultural areas and may have been occupied by people from either area for as long as this cultural boundary has existed.
A variety of cultural chronologies have been developed both for the northern Great Basin and the southern Plateau area. Figure 2 and Figure 3 show some of the major developmental schemes for the Great Basin and Plateau regions respectively. Although it would appear from these charts that we are dealing with a bewildering array of localized specialization, a close look at the various approaches shows a certain amount of similarity and continuity. Rather than discuss each of these frameworks in detail, I will refer the reader to the references cited in Figures 2 and 3, or for a summary see Toepel et. al. (1979) and Minor et. al. (1979). Here I will attempt, instead, to summarize and synthesize the common elements found in these approaches and sketch a broad outline of the regional prehistory. Naturally, there are going to be local variations in this broad picture, some of which we will discuss Later in terms of Fagan's (1974) and Pettigrew's (1979) models.
For both the Great Basin and the Columbia Plateau, the earliest periods are identical. The first period sometimes hypothesized is a Pre-projectile Point Horizon for which there is little reliable evidence. The strongest argument for this period is the simple logic that it is followed by a period in which a sophisticated form of projectile point, the fluted point, appears. Fluted points are not found in the Old World, so this type must have developed in the New World out of some antecedent. I would suggest that it is more logical to suppose that the fluted forms were preceded by unfluted lanceolate types, as are found in Old World cultures, rather than some "Pre-projectile Point" phase. The problem with this idea is that lanceolate forms are found subsequent to (and also contemporaneously with) fluted forms, so they are difficult to place chronologically on a purely typological basis.
The Fluted Point Horizon has not been dated in the northern Great Basin or Plateau regions, but is assumed to date from about 10,000 to 12,000 years ago as it does in the Southwest. The widespread distribution of fluted points over most of North America suggests that some temporal variation may be found on a regional basis, but these dates are probably at least close. The Fluted Point Horizon apparently represents a period of nomadic big game hunting.
The Fluted Point Horizon was followed by what Bedwell (1973) calls the Western Pluvial Lakes Tradition. Aikens (n.d.) has noted that other variously-named complexes of similar forms overlap and continue subsequent to the Fluted Point Horizon in the West. He has termed this the San Dieguito- Windust-Milliken Horizon. Under whichever name it is referred to, this horizon probably developed out of the Fluted Point tradition as there are a number of technological similarities (Aikens 1978). This horizon is characterized by large lanceolate and leaf-shaped projectile point forms, some with stems or shoulders, as well as crescents, various forms of scrapers, and in some instances manos and metates. Generally, this horizon dates from 8,000 to 11,000 years ago.
Between about 4,000 and 8,000 years ago, large side-notched points make their appearance in both the northern Great Basin and southern Columbia Plateau regions. Willow-leaf and lanceolate forms, with or without stems or shoulders, continue to be found throughout this period as well. Some theorists characterize this period as being drier and warmer than either the preceding or subsequent periods; subsistence shifts to higher elevations with possible population declines are hypothesized. Some distinctions in material culture are becoming evident between the Basin and Plateau regions during this period, but no clear-cut division is yet possible.
Subsequent to 4,000 years ago, however, the Basin and Plateau regions become quite distinct and substantial differences become evident. Within each region, local diversity increases and the record becomes more complex. Plateau subsistence becomes organized around large, semi-permanent village sites with easy access to riparian resources, especially anadromous fish, with upland hunting and gathering activities taking on secondary importance. In the Great Basin, meanwhile, subsistance characteristic of the Desert Culture Tradition is dominant, with broad based resource utilization by small nomadic bands most likely.
The material culture of the Great Basin over the past 4,000 yeats is characterized by a wide variety of specialized lithic tools. Projectile points, during the early part of this period, include some of the large side-notched type found in earlier times, as well as a variety of triangular forms with corner notches and stems. During the latter part of this period, points become much smaller, presumably with the introduction of the bow and arrow, and are found in various notched and stemmed forms.
In the Columbia Plateau area, the material culture includes scrapers, net sinkers, pestle and mortars, utilized cobble spalls and various projectile point types. Here, the large side-notched points seem to persist later than in the Great Basin, and triangular, stemmed and corner-notched forms are also found. Towards the end of this period snraller points again become apparent with the introduction of the bow and arrow, though the larger points seem to persist for some time alongside the smaller ones of similar forms.
Returning now to Fagan's Study Altithermal Occupation of Spring Sites in the Northern Great Basin (1974), we can determine how his model integrates with the overall picture and its possible applications to this study. Fagan set out to test Bedwell's (1970) theory that the altithermal period was marked by a subsistence shift from lowlands to uplands in the Great Basin. The altithermal is a climatic period hypothesized by Antevs (1948) from about 4,500 to 7,000 years ago, during which there was much less precipitation and generally warmer temperatures in the Basin area. The validity, severity and importance of this climatic change has been questioned by a number of authors (Jennings 1957, 1964; Bryan and Gruhn 1964; Aikens 1970; and Swanson 1962), while Antevs (1948) and Baumhoff and Heizer (1965) have suggested that this warm, dry spell led to abandonment, or near abandonment of the Great Basin.
Fagan investigated twelve sites in southeastern Oregon, one of which falls within the Drewsey Resource Area, and all of which are close enough to supply comparative material of relevance to this report. The sites were divided into higher elevation locations (5100 to 6000 ft.) and tower elevations (4200 to 4300 ft.), with nine sites falling in the former category and only three in the latter. Pagan excavated "several" one-by-two meter test pits at each site. Over 11,000 artifacts were recovered; the bulk of the report consists of analysis and categorization of these artifacts. Only two sites yielded sufficient organic rnaterial for carbon-14 dating, and in only one of these were there enough dates to establish a tentative obsidian hydration rate through cross-dating.
Fagan divided his material into four time periods, based on the carbon-14 samples where possible, and also on typological and stratigraphic evidence. The earliest times recorded were Period IV, which spanned from 7,000 to 11,000 years ago. Period IV is characterized by large leaf-shaped, stemmed and shouldered projectile point forms, and a paucity of grinding implements. Period IV materials were found at four of the twelve sites, all of them at higher elevations.
Period III approximately corresponds to the Altithermal Period, from 5,000 to 7,000 years ago. This time span is characterized by large side-notched, ELko Eared and Elko Corner-notched projectile points, and grinding tools are still relatively scarce. Evidence for occupation during this period was found at ten of the twelve sites investigated. Fagan states in his conclusion that "eight of the ten Period III cultural deposits were located at springs situated between 5,000 and 6,000 feet in elevation" (Fagan 1974:102). Remembering that only three Lower elevation sites were investigated, we could restate these results as: two out of the three Lower elevation sites (4200-4300 ft.) showed evidence of Altithermal occupation. Since the one Low elevation site that did not show evidence of Period III occupation also lacked Period IV deposits, it hardly represents "abandonment" of the site during the Altithermal. Looked at objectively, 67% of the lower elevation sites and 89% of the higher elevation sites showed evidence of Altithermal occupation. This difference is based on too small of a sample to be statistically significant.
For Period I (150-3,000 BP) and Period II (3,000-5,000 BP) Fagan notes an abundance of milling stones at low elevation sites. Period II is charac- terized by a continuation of Elko type projectile points and the introduction of Basal-notched points with expanding stems and flaring barbs. Period I includes small, corner-notched, Desert side-notched, Cottonwood Triangular and Eastgate Expanding Stem projectile points.
Overall then, Fagan fails to present compelling evidence in support of Bedwell's model of a subsistence shift to upland resources during the Altithermal, but he does identify what appears to be a functional difference between higher and lower altitude sites during the last two periods of occupation. Subsequent to 5,000 years ago his lower altitude sites show an abundance of milling stones, implying greater reliance on plant foods, while higher sites tend to have relatively greater proportions of projectile points, implying an orientation toward hunting as a primary pursuit at these sites.
Pettigrew's study Archaeological Investigations at Stinkingwater Pass, Harney County, Oregon is of particular relevance to this report because the entire investigation falls within our study area, and some of the seeding areas that were surveyed are in close proximity to Pettigrew's project area. Unfortunately, there are a number of prolems with the analysis and conclusions in the Pettigrew report.
Pettigrew investigated ten sites that were to be impacted by road construction. Systematic surface collection samples were taken from each site, and additional tools were collected from the unsampled portions of four of the sites. Six sites were tested for subsurface materials, and additional excavations carried out on two of them. A total of 875 tools and 15,814 unmodified flakes were recovered from the ten sites. Tools were divided into fourteen categories and analyzed in terms of distribution and constituent material. The category "projectile points" was further subdivided into twelve types which were typologically divided into chronological periods. Neck widths were measured on projectile points for which that measurement was possible, and a bimodal distribution thought to have temporal significance was discerned. Points with necks less than eight millimeters wide were defined as narrow necks and are thought to represent recent periods. Wide necks were defined as equal to or greater than eight millimeters, and are thought to represent older periods.
Based on his analysis of the data, Pettigrew concludes that the Stinkingwater Pass area was utilized by two distinct groups of people, one group from the Harney Basin and the other from the Malheur River valley. The area was not heavily utilized by either group prior to about 4,000 years ago. Beginning around 4,000 years ago, the Harney Basin people began to make regular and relatively intensive use of the area while the Malheur River people used the area sporadically through this time period.
Pettigrew sees the Harney Basin people as living a nomadic hunting and gathering lifestyle, characteristic of the Great Basin. Their use of the Stinkingwater Pass area would consist of a temporary base camp established as part of the annual round, and outlying hunting camps. The Malheur River people, on the other hand, he perceives as living a Plateau-like lifestyle. These people would have established large winter villages and summer fishing stations outside of the Stinkingwater Pass area. Utilization of the upland game resources would have been a secondary activity for the Malheur River people, so they would have left evidence of less intensive occupation, from small, temporary hunting camps.
Pettigrew goes on to hypothesize that changing climatic conditions were responsible for the Harney Basin people's shift to greater utilization of the Stinkingwater Pass area about 4,000 years ago. Prior to the Altithermal period, Pettigrew suggests that faunal resources were more abundant and more concentrated in the Basin area, so there was no need to exploit the uplands. During the Altithermal, Pettigrew suggests that the Stinkingwater Pass area was not utilized much because it was too dry to support a very large game population, and hunters from the Harney Basin would have sought even higher elevations. With the end of the Altithermal, sufficient moisture would have been available for animals in the Stinkingwater Pass area. However, resources in the Harney Basin would have been less concentrated than prior to the Altithermal, so more intensive utilization of the Pass area would have begun about then.
One problem with Pettigrew's analysis is his rather cavalier treatment of typology and the subsequent dating scheme derived from projectile points. It is not clear where Pettigrew gets his dates for established projectile point types. Hauck and Weder (1978) and Hester and Heizer (1973) each reference a variety of published chronologies for various point types, but the Pettigrew dates match none of them. Hauck and Weder note that certain point types occur later in the western Great Basin than further east and suggest a chronology that may be tentatively used for southeastern Oregon until further research classifies the dates. Table 1 gives the most extreme ranges cited by Hauck and Weder (1978) and Nester and Heizer (1973) for various point types, as well as Pettigrew's dates and suggested dates from Hauck and Weder (1978). In most cases, the oldest of the extreme dates have been discounted and the early dates for Elko and Northern Side-notched are questionable: the Elko because it's uncharacteristically recent and the Northern Side-notched because the later dates seem to apply to the Plateau area rather than the Great Basin. In any case, Pettigrew's dates seem slightly out of line, especially for Rose Spring, Eastgate and Northern Side-notched types. This discrepancy is not so serious until it is compounded by grouping point types into three broad categories and assigning temporal periods. Thus, Pettigrew's late period, which he dates from historic times to 1950 B.P., might more appropriately date from historic times to 1500 B.P., based on the diagnostic point types he associates with it. His middle period, which he dates from 1950 B.P. to 3950 B.P., may span from 1500 R.P. to 5000 B.P. The early period (3950 B.P. - 8950 B.P.) may span from 5000 B.P. to 8000 B.P., based on the Northern Side-notched type, but Pettigrew also includes Humboldt type points in this time period, although they probably date from the middle period.
All of this dating confusion is further compounded by Pettigrew's unusual interpretation of typological categories. His "type B" projectile points, for example, are treated as if they were Cottonwood Triangular points for dating purposes, yet his illustrations in this category show large, broad, thick bifaces that are most likely blanks, preforms or knives. His "type J" points, which he describes as "leaf shaped" and dates to his early period, include six illustrated specimens, all short basal fragments. "Type H" points are described as "shouldered" with "non-diverging stems," though shoulders are not evident on all specimens illustrated. Also, a base broken off any of the "type H" points would easily fall into the "type J" category, although Pettigrew considers the former representative of his middle period and the latter from the early period.
Much of Pettigrew's subsequent analysis is thrown into question by these irregularities in his typological and chronological treatment. Consider, for example, the following:
Considering all sites taken together, the sample of projectile points is dominated by the middle period types (F, G, H, and I), which make up 43.7% of the total. These are folloved by the late period types (A, B, C, D, and E), with 37.5%. The Least proportion, 18.7% is made up by the early period types (J, K and L).
Let us assume a constant rate of deposition of projectile points in the area of Stinkingwater Pass. Let us assume also that the middle period extended from 2000 B.C. to A.D. 1 and the late period from then until A.D. 1850 (see Appendix). We would then expect 1850/3850, or 48.1% of the points deposited since 2000 B.C. to be late period types, and 2000/3850, or 51.9Z, to be middle period types. These percentages are very close to the actual figures from the sample. Excluding the early period types, there are 91 classifiable points, 42 (46.2%) of which are from the late period, and 49 (53.8%) from the middle period. The rate of deposition of projectile points under these assumptions is 2.4 points per hundred years. Early period types account for only 21 points; if these were spread throughout the 5000 years of the early period (7000-2000 B.C.), then the rate of deposition was only 0.4 points per hundred years. If the rate of deposition of points in the early period was the same as in the middle and late periods, then the early period occupation could have lasted only 875 years, e.g., from 2875 to 2000 B.C.
It becomes clear, then, that the sites at Stinkingwater Pass were used either very little or for only a short time in the early period, and that intensive occupation began around 2000 B.C. (Pettigrew 1979:47).
Ignoring for a moment the weakness inherent in both the explicit and implicit assumptions in this statement, our adjusted chronology gives aslightly different picture. If the late period only lasted 1400 years (assuming a historic cut-off date of 1850 A.D.), and the middle period 2000 years, we would expect 41.1% late period points and 58.9% middle period, suggesting that late points are over-represented with 13.5% more points than predicted by Pettigrew's analysis. Nor are early points as grossly under-represented as he suggests if the early period was only 3000 years long instead of 5000.
An even more critical problem with this analysis lies in the assumptions made. The explicit assumption that a constant rate of deposition might be expected from a consistent level of use of the area by prehistoric inhabitants is rather unlikely. Most models of Great Basin prehistory incorporate a change in hunting technology when the bow and arrow were introduced to supplant javelin and atlatl darts. A hunter could presumably carry and expel more of the small arrows than larger darts or spears in a given amount of time. Differing prey and resource utilization through time might also affect the rate of deposition. Certain projectile forms may be more commonly broken during manufacture, leaving fragments easily mistaken for completed points. (See Pettigrew's Figure 19e for an example of a probable preform classified as a projectile point.)
Pettigrew's implicit assumption that his collection accurately reflects the types and relative number of points deposited is even less defensible. Only 20 of the 112 points classified by Pettigrew came from excavations, the rest being surface finds. Therefore, more recent points would be expected to be over-represented in the collection. Both natural and cultural forces may also be expected to cause attrition to projectile points through physical destruction or reshaping beyond recognition. It stands to reason that any such effects are cumulative and would therefore impact earlier forms greater. Projectile points can be seen to decrease in size through time, so earlier points would be more attractive to later inhabitants for use as knives and scraping tools. Later in this report we will examine the frequency distri- bution of projectile point types from the seeding projects and compare them to Pettigrew's.
Given that Pettigrew's chronological analysis and conclusions about intensity of occupation are somewhat questionable, let's examine the other major component of his model for occupation of the Stinkingwater Pass area: namely, his contention that two distinct groups of people with different subsistence regimes utilized the area. Pettigrew bases this conclusion primarily on an observed difference in the distribution of two lithic materials, crypto- chrystalline silicates (CCS) and fine-grained basalt.
One interesting observation that Pettigrew fails to make concerns the different ages of the sites he examined. His Table 4 (Pettigrew 1979:25) lists the sites roughly in chronological order based on projectile point typology. Assuming that this order is at least approximately correct (without considering what absolute dates might apply), we can see that the Table fairly accurately discriminates between the sites Pettigrew associated with the "Malheur" people and the sites he attributes to the "Harney" people. In general, the sites attributed to the "Harney" people tend to be later sites while those of the "Malheur" people are earlier sites. So one alternative explanation for the observed differences in the archaeological record is that those differences reflect changes through time rather than the activities of two hypothetically distinct groups of people.
There are also some very basic problems with Pettigrew's analysis of the evidence that leads him to postulate two distinct groups. Consider, for example, his statement that:
It has been proposed that the Stinkingwater sites were used by two different groups of people coming from different directions into the area. A corrolary to this hypothesis is that those from the southwest were the principal users of the basalt from HA 66, and those from the northeast the principal users of the CCS from HA 70/1. If this is true, then we would expect CCS and basalt detritus to cluster separately on the surfaces of the sites more often than would CCS and obsidian or basalt and obsidian, since the basalt- users and CCS-users would have occupied these sites at different times. (Pettigrew 1379:53).
Three basic materials were evident at the sites: obsidian, basalt and CCS. Both hypothetical groups presumably used obsidian, one group basalt, and the other CCS. Thus, they would leave scatters of obsidian, basalt, CCS, obsidian and basalt, or obsidian and CCS. The occurrence of basalt and CCS together represents the chance juxtaposition of materials due to the two groups using the same location at different times. This is exactly the opposite from what Pettigrew claims in the above paragraph. Obsidian and basalt or obsidian and CCS should cluster separately more often than CCS and basalt. Thus, if we accept his analysis that "CCS and obsidian cluster separately in four cases, and obsidian and basalt in five cases; but CCS and basalt are separately clustered in eight cases" (Pettigrew 1979:54), he has mustered evidence against his own hypothesis. Actually, the quoted figures themselves are also questionable, since an examination of the raw data (Pettigrew 1979:34-43) shows that Pettigrew's use of the term "cluster" is as inconsistent as his typological categories.
Overall then, there seems to be no reason to conclude that the intensity of use of the Stinkingwater Pass area has changed much through time, nor is there any reliable evidence that the area constitutes a border area between culturally distinct groups. There does seem to be a real difference in distribution between basalt and CCS at the sites investigated. However, it is more likely that this distribution can be explained by either the proximity of sites to lithic sources, changes in lithic resource utilization through time, or some combination of these factors.
The history of the Drewsey Resource Area has been well documented as part of historical summaries for Harney County or Southeastern Oregon (Bright 1981, Brimlow 1951, LoPiccolo 1962 and Simpson 1973). Although historical resources were considered as an integral part of this project's surveys, only three historical resources were encountered (tin can scatters), none of any particularly great significance. In view of the few sites involved, I will refer the reader to the above mentioned sources rather than recount the project area's history here.