Unlike Swanson's model, the Vantage Phase shall not be subdivided in this report. Although the two Vantage Phase components at 45KT28 display apparent cultural differences, they have not been characterized as subphases because the sample from each is far too small and incomplete. However, each component will be described separately. [8]

[9] CULTURAL COMPONENT I

Stratigraphy. Cultural Component I, encountered in the excavations at House Pit 15 (Fig. 4), was contained in two lenses of coarse sand located in the upper portion of Stratum 3 just above a small, alluvial-filled stream channel associated with the development of the Sunset Canyon Fan (Figs. 5 and 6). Although both lenses were culture bearing, the deeper of the two, where it intersected and crossed the channel fill, was particularly productive. This fact is easily explained by the mode of deposition. During a sheet flood or period of high water, a culture-bearing deposit was eroded somewhere to the north and west of the excavations, whisking the smaller artifacts, flakes, pebbles, and other cultural debris down the inclined bank of the river. The debris moved freely along this surface, except at the place where the (gravel) channel fill had been exhumed during a period of erosion. At this point the gravel of the channel fill lessened the slope and provided an irregular surface on which flakes and artifacts might naturally accumulate.

Two test excavations, beneath House Pits 7 and 28 (Fig. 4), failed to reveal the source area for the cultural materials of Cultural Component I. Lack of time and insufficient funds prevented further testing.

   To keep within the final publication budget, it was necessary to shorten the report by removing a number of figures and the accompanying text. This figure, which shows the excavation units in Cultural Component I, was one such figure. During the final three years of excavation, the river partially flooded the excavation, so each year a new excavation had to be established. The smaller sounding represents the first year of excavation in CCI, while the larger represents the second year. The filled area is the top of the distributary channel gravel deposited by Sunset Creek and which created the irregular surface on which most of the CCI artifacts were recovered.

The top of the channel, together with overlying and marginal deposits, are seen in this composite photograph. Compare with Figure 5 for the wider context.

Artifact Assemblage. Although small, the assemblage of 62 artifacts contained several fragments of leaf-shaped points and knives. Two of the point bases which were recovered are slightly shouldered (Fig. 7, c-d). They resemble specimens recovered by Borden from the early levels of DjRi3, near Yale, British Columbia (Borden 1957), and occasional specimens which occur in Olcott assemblages along the Puget Sound littoral (Butler 1961). Other characteristic artifacts are illustrated in Figures 7 and 8.

The following is a complete catalogue of the artifacts recovered from Cultural Component I. Bone artifacts were not present because bone was not preserved. Stone detritus was common, several hundred flakes of cryptocrystalline silica having been recovered. Ten basalt flakes were also recovered, suggesting that cobble implements may have been used. In Cultural Component III, where cobble implements were recovered, the percentage of basalt flakes was also quite low.

   Bifaces from Cultural Component 1. Vantage Phase. Drawn by Delmar Nordquist. Catalog Nos.: a, 2239; b, 4561; c, 3616; d, 3633; e, 2241; f, 3645; g, 3095; h, 3102.

Artifact Catalogue.

   Representative artifacts from Cultural Components I and II, Vantage Phase.

CULTURAL COMPONENT II

Stratigraphy. Cultural Component II is situated in the lowest portion of Stratum 4 between six inches to one foot above the uppermost lens of Cultural Component I (Figs. 5 and 6), and is characterized by scattered, discontinuous lenses of compact yellow-brown silt or clay. Though shown in the profiles as directly underlying Cultural Component IV, this component exists else¬where, where it underlies Cultural Component V. Because the lenses are water deposited, we are faced with the problem of redeposition, probably from a nearby source.

Artifact Assemblage. Although only eight artifacts were recovered, several of these proved to be diagnostic (Fig. 8, f-j). These specimens included three leaf-shaped points, a rudimentarily stemmed point, and a large, leaf-shaped knife whose base is stemmed. Since simply stemmed points are present in Cultural Component III, it is likely that Cultural Component II represents the end-point of the Vantage Phase. This is also indicated by one of the leaf-shaped points (Fig. 8, h), which is much smaller than the other points from the Vantage Phase, and correspondingly like those of the Frenchman Springs Phase.

The following is a complete catalogue of the artifacts recovered from Cultural Component II. In addition, only a handful of cryptocrystalline flakes and a few basalt pebbles were recovered.

Artifact Catalogue.

Total number of artifacts (8)

FLAKES AND ARTIFACTS FROM BENEATH CULTURAL COMPONENT I

A total of 30 artifacts, 80 cryptocrystalline silica flakes, and 3 "bolas stones" were found resting on the lower sand lens of Cultural Component I. The enamel of a beaver or muskrat tooth and two river cobbles were also recovered. The diverse nature of these specimens, ranging from minute flakes to a core weighing three pounds (Fig. 9), the survival of the tooth enamel, and the fact that flakes struck from the same core were often found side by side all suggest that these are the earliest in situ deposits in the House Pit 15 area.

Artifact Assemblage. Unfortunately, the most diagnostic artifacts recovered from these deposits were cobble scraping planes (Fig. 65, a-c) similar to those of Cultural Component III. Thus it is difficult to assign the assemblage to either the Vantage or Frenchman Springs phases.

Artifact Catalogue.

[12] Total number of artifacts (30)

DISCUSSION

Because the assemblages from both of the vantage phase components have been redeposited, it is not possible to estimate accurately their age on the basis of geological information alone. The strata in which they occur are estimated to have been deposited between 1000 and 3000 B.C., a minimal date for the age of the cultural assemblages themselves.

Indeed, based on the few comparisons of material culture that can be made. Cultural Component I is probably no more recent than ca. 4000 B.C. and very possibly as old as 7000 B.C. The large leaf-shaped points which characterize this component are comparable to specimens which Borden has reported from the early components at DjRi3, a site on the Fraser River in British Columbia (Borden 1957). Based on geological estimates and a series of Carbon 14 dates, these components at DjRi3 span the period between 5000 and 7000 B.C. Although there are no other dated occurrences, such points are characteristic of many early assemblages from the Puget Sound littoral (Butler 1961; Thomson 1961). Moreover, the large, slightly shouldered leaf-shaped points from Cultural Component I also find counterparts in both these assemblages.

Based on the assemblages at hand, Cultural Component II appears to be substantially more recent than Cultural Component I, possibly in the neighborhood of 2000 to 3000 B.C. This estimate is based on the occurrence of a Type 1 stemmed point and a Style 1 leaf-shaped point, both common in the early Frenchman Springs period.

Although the slightly shouldered leaf-shaped projectile points and the large leaf-shaped knives from Cultural Components I and II suggest that there is a close typological link between the Vantage Phase and the earliest assemblages from the Puget Sound littoral and the Fraser River val¬ley, the flaking techniques employed in tool muanufacture are basically like those recorded from other Anathermal and Altithermal sites in the Plateau.

The flaking detritus from Cultural Components I and II reveals that low angle cores with faceted striking platforms were used to produce many of the flakes from which tools were sub¬sequently manufactured. Although smaller in size, these flakes and flake tools are otherwise very similar to those from the Lind Coulee site (Daugherty 1956a; 1965: personal communication) and the Lind-Coulee-like assemblage at Thorn Thicket (45WT36), a site on the Lower Snake River (Roderick Sprague 1966: personal communication). Platform-faceted flakes and blades of a slightly different type also have been reported from more recent Altithermal components along the Lower Snake River (C. M. Nelson 1965a; 1966) and the Middle Columbia River (Butler 1959:13), and evidence for various kinds of platform faceting exists from every component at the Sunset Creek site.

Although flakes with faceted platforms exist in surface collections from the Fraser River valley, they are more recent than the Anathermal and Altithermal specimens from the Columbia Plateau. Moreover, platform faceting has not yet been reported from the early materials in the Fraser River valley or the Puget Sound littoral.

THE VANTAGE PHASE AND THE OLD CORDILLERAN CULTURE

Introduction. The Vantage Phase components are two of the many early occurrences of leaf-shaped projectile points and knives in the Pacific Northwest. Recently, Butler (1958c; 1961; [12/13] 1962a; 1965) has used the distribution of such projectile points, associated artifacts, and ecological information to postulate the existence of an Old Cordilleran Culture.

As Butler (1965) himself has pointed out, the Old Cordilleran Culture concept was developed as a tool for visualizing Northwest prehistory in terms of itself rather than an extension of Great Basin or Plains patterns of prehistoric development. The fact that Butler recognized the fun¬damental ecological unity of early archaeological assemblages in the Pacific Northwest represents a major contribution to our understanding of archaeological sequences in the area. However, accumulating archaeological data now suggest that the Old Cordilleran Culture concept may not account for these fundamental similarities. It is my purpose to review the relationship between the archaeological content of the Old Cordilleran Culture concept and the known archaeological data, and to demonstrate that alternative hypotheses can be developed which are simpler and account for the known data more completely.

In 1961 (64-65), Butler stated that the only highly diagnostic tool type characteristic of the Old Cordilleran Culture was the leaf-shaped Cascade projectile point. Relying heavily on information obtained from a single site near The Dalles (Butler 1959:13), he further suggested that these projectile points were made on blades struck from large conical cores. In addition to Cascade Points, the Old Cordilleran Culture was characterized by the more ubiquitous leaf-shaped knives, unifacially flakes cobble tools, and unspecialized scrapers and flake tools.

Using information from the Fraser River valley (Borden 1957), the Puget Sound Littoral (Butler 1961), The Dalles (Cressman 1960; Butler 1958c, 1959, 1962), the northern Great Basin (Cowles 1959), and the Lower Snake River (Butler 1958b), Butler postulated that the Old Cordilleran Culture had developed along the foothills of the Cascade Mountains between northern Puget Sound and the northern Great Basin between ca. 12,000 and 13,000 B.P. (Butler 1961:63-64). He also speculated that it might have spread westward into the Puget Sound lowlands between ca. 7000 and 8000 B.P. and eastward across the southern Columbia Plateau after its emer¬gence in The Dalles locale between 10,000 and 11,000 B.P.

In 1962a (77-78), Butler expanded his original thesis, enumerating the following characteristics for the Early Dalles Phase of the Old Cordilleran Culture: "a well developed bone and antler industry; blade implements, including the finely made ‘willow-leaf’ Cascade points and end-of-blade scrapers; the edge-ground cobble complex; hunting of such large mammals as deer and elk, and, apparently, large numbers of birds; and intensive, probably seasonal, exploitation of such riverine resources as fish." He also suggested that the Old Cordilleran Culture spread eastward from The Dalles locale into Idaho between 7000 and 8000 B.P.

In 1965 (1127), Butler extended these characteristics to the entire Old Cordilleran Culture and added unifacially beveled antler wedges to his list of diagnostic traits.

If Butler's treatment of the data is accepted without critical analysis, then the Vantage Phase may be seen as an early extension of the Old Cordilleran Culture from the foothills of the Cascade Mountains into the western margin of the Columbia Plateau prior to 7000 or 8000 B.P. Such an extension would be considered analogous to the occupation of The Dalles area. Although this point of view is superficially supported by the typological similarity between the projectile points from the western Cascades and the Sunset Creek site, a critical appraisal of all the comparative data suggests that the entire Old Cordilleran Culture concept must be revised. [13]

[14] Previous reviewers of the Old Cordilleran Culture concept have pointed out that there is considerable ambiguity about the specific nature of the Old Cordilleran Culture because its constituent elements and regional phases have not been clearly defined (e.g., Carlson 1962;Gruhn 1962). Although Butler's (1965) most recent reply to these criticisms has helped to clarify some of the issues, it has only complicated others. For example, prior to the reply, the Early Dalles Phase was the only well defined regional component in Butler's hypothesis (Butler 1962a). When Butler (1965) extended all of the characteristics of the Early Dalles Phase to the entire Old Cordilleran Culture, he limited the possibility of defining regional phases while adding a considerable amount of detail to the archaeological content of the Old Cordilleran Culture, itself. On the one hand, this expansion of the overall construct is clearly useful since it embodies a more specific and more easily tested hypothesis. But, on the other hand, the apparent reversal of Butler's earlier concept of regional phases makes it increasingly difficult to assess his view of the areal relationships between the northern and southern Cascades, the northern Great Basin, and the various parts of the Plateau, areas which may have been occupied by the Old Cordilleran Culture.

In the absence of large amounts of hard primary data and detailed hypotheses, arguments for and against the Old Cordilleran Culture concept have tended to revolve around each particular individual's personal ideas about which specific archaeological occurrences are considered as representative of the Old Cordilleran Culture and the applicability of the word "culture', to archaeological data (e.g., Daugherty 1962:148-49; Butler 1962a:7-9). Although these arguments are interesting in the context of developing a meaningful research framework for archaeology, they have been far too general in scope to provide a meaningful review of the phenomena to which the caption "Old Cordilleran Culture" actually refers. Therefore, let us begin with a review of the regional sequences for areas in which the Old Cordilleran Culture is said to occur.

Regional Evidence. There are four major Anathermal-early-Altithermal archaeological sequences in the southern Columbia Plateau. The one which Butler relies most heavily upon has been constructed around the Five Mile Rapids site at The Dalles, Oregon. As Butler notes (1961:24), the typological descriptions presented in the published report of the Five Mile Rapids site (Cressman 1960) are so general that it is impossible to make adequate comparisons, even with other assemblages in The Dalles locale. However, on the basis of personal knowledge. Butler has equated the earliest portion of the sequence at Five Mile Rapids with the earliest component at Indian Well. Since this equation forms the only basis for hard information on the age and many of the characteristics of the Old Cordilleran Culture in The Dalles locale, it is central to any evaluation of the Old Cordilleran Culture. David L. Cole (1964: personal communication) and others who have personally worked with the collections from Five Mile Rapids and other early sites on the Middle Columbia insist that Cascade projectile points, possibly the most important single criterion for identifying the Old Cordilleran Culture, only occur during the Altithermal. These conflicts will never be completely resolved until the assemblages from The Dalles locale have been properly documented and re-evaluated.

From conversations with individuals who have worked with the collections and from the published report (Cressman 1960), I would provisionally reconstruct the early part of the sequence at Five Mile Rapids as follows. The earliest component (ca. 10,000-8000 B.P.) is characterized by a variety of leaf-shaped and unstemmed lanceolate projectile points and knives; there is also one stemmed lanceolate projectile point resembling the specimens from Lind Coulee. The leaf-shaped specimens are broad and short with wide, shallow flake scars and little or no retouching; they are not Cascade points. In addition, edge-ground cobbles, girdled bolas, and antler splitting wedges are also present. Later components contain a wide variety of stemmed lanceolate, lanceolate, and leaf-shaped projectile points some of which may be Cascade points. [14]

[15] The three remaining early sequences occur along the Lower Snake River, at Windust Caves (H. S. Rice 1965), Marmes Rockshelter (Fryxell and Daugherty 1962; Fryxell 1963), and at Thorn Thicket (Sprague 1966). They reveal the following sequence.

1. The earliest components, which appear to be roughly coextensive with the Anathermal (ca. 10,000-8000 B.P.), and which include the Lind Coulee site (Daugherty 1956a), are heavily dominated by a wide variety of stemmed-lanceolate projectile points in association with smaller numbers of unstemmed lanceolate and leaf-shaped projectile points (Rice 1965: Figs. 11-13, 17; Fryxell and Daugherty 1962: Fig. 6; Daugherty 1956: Figs. 18-20). Flake tools are made almost completely of cryptocrystalline silica; the use of basalt is uncommon, and specimens of obsidian and quartz crystal are very rare. Information from Lind Coulee and Thorn Thicket (Daugherty and Sprague 1966: personal communication) suggests that many flake tools were manufactured on large flakes and blades detached from exceedingly low angle, conical or biconical cores with faceted and frequently edge-ground striking platforms. These flakes are typically triangular or diamond-shaped in outline, with a narrow striking platform remnant in relationship to overall body width. Both unifacial and bifacial cobble implements occur in varying quantities, and edge-ground cobbles are present at some sites.

2. The second period roughly corresponds to the first half of the Altithermal (ca. 8000-6500 B.P.), and is represented both in the major sequences and at sites which possess no Ana-thermal component, including Indian Well (Butler 1959; 1961), Hat Creek (Shiner 1953; 1961), Ash Cave (Butler 1958b; 1962a), and Weis Rockshelter (Butler 1962a). These assemblages are dominated by leaf-shaped projectile points including those most frequently referred to as Cascade Projec¬tile points, but small numbers of stemmed- and unstemmed-lanceolate projectile points also occur. Cryptocrystalline silica is still the dominant material, but the use of basalt is more common and may dominate at some sites; though still rare, obsidian is far more abundant. Many flake tools are manufactured on blades struck from conical or biconical cores with faceted and frequently edge-ground striking platforms (C. M. Nelson 1965a; 1966; Butler 1959:13). These blades tend to be parallel sided or slightly diamond-shaped in outline. Unlike the specimens manufactured during the preceding period, they are frequently (and sometimes dominantly) of basalt, possess a less acute angle at the striking platform, are much smaller, and have physically larger striking platform rem¬nants which more closely approximate body width. Predominantly unifacially flaked cobble imple¬ments occur at all sites, and edge ground cobbles are common at some site (e.g., Bryan et al., 1963: Goldendale; Butler 1962a: Weis Rockshelter) and absent or very rare at others (e.g., Marmes Rock-shelter, Daugherty 1965: personal communication).

3. The latter part of the Altithermal (ca. 6500-4000 B.P.) is characterized by the Cold Springs Phase along the Lower Snake River and the upper part of the Middle Columbia River. This phase is represented at Cold Springs (Osborne and Shiner 1949; Shiner 1961), Three Springs Bar (Daugherty 1965: personal communication), Marmes Rockshelter (Fryxell and Daugherty 1962), 45C01 (C. M. Nelson 1965b; 1966), 45Ga3 (C. M. Nelson 1965a), Thorn Thicket (Roderick Sprague 1966: personal communication), and Weis Rockshelter (Butler 1962a), where it may occur at a significantly later time. It is characterized by Cold Springs Side-Notched projectile points in associa¬tion with lesser numbers of finely made leaf-shaped projectile points usually referred to as Cascade projectile points; unstemmed lanceolate projectile points occasionally occur in small numbers (e.g., C. M. Nelson 1966: Fig. 11). Basalt is commonly used in the manufacture of flake tools and over¬whelmingly dominates the assemblages in which there is very little variation in stem form and no trend towards corner removing or corner notching. Elsewhere, cryptocrystalline silica is the most common material; obsidian may account for as much as ten percent of the projectile points in some assemblages. Many flake tools are still manufactured from blades struck from conical or biconical cores with faceted and sometimes edge-ground striking platforms. In contrast to the preceding [15/16] period, the blades are slightly smaller and less frequently edge-ground. Although usually absent, edge-ground cobbles are found at some sites (e.g., Thorn Thicket, Roderick Sprague 1966: personal communication). Predominantly unifacially flaked cobble implements are common at all sites, and manos, grinding slabs, and pestles occur at some sites (Butler 1962a; Nelson 1966; Daugherty 1965: personal communication).

4. In the succeeding period on the Lower Snake River (ca. 4000-2000 B.P.), the large side-notched and leaf-shaped projectile points are replaced by a wide variety of corner-notched, corner-removed, contracting-stemmed, and rectangular-stemmed projectile points. In the extreme southeast corner of the Columbia Plateau, large side-notched projectile points persist in small quantities (e.g., C. M. Nelson 1966), and leaf-shaped points and knives occur as a minor part of the assemblage everywhere in the Plateau. These leaf-shaped specimens are typologically quite different from earlier types. The use of the faceted platform technique appears to be absent or very rare along the Lower Snake River, but continues to persist along the Upper Columbia River.

In the northwestern Columbia Plateau, along the Upper Columbia River, the only major early sequence occurs at Sunset Creek. Using comparative data the following periods can be defined.

1. The earliest period, the Vantage Phase, is characterized by large and small leaf-shaped projectile points and knives typologically most similar to specimens from the early assemblages in the Fraser River valley and along the Puget Sound littoral. Flake tools are frequently made on flakes and blades similar to those from Lind Coulee and Thorn Thicket; they are only smaller in size and less frequently occur with edge-ground striking platforms. Cobble implements do not occur in the selectively sorted sediments of 45KT28, but the flaking detritus and comparative data suggest that they were probably present during the Vantage Phase. Edge-ground cobbles are probably also present (Swanson 1962a).

The relationship between the Vantage Phase and the early phases along the Lower Snake River is not yet completely clear. The most likely reconstruction would place it successive to Lind Coulee and other late Wisconsin or Anathermal sites, and contemporary with the early Altithermal sites in the southern Columbia Plateau.

2. The second clearly represented period is the Cold Springs Phase (ca. 4500-3800 B.P.). it is represented at 45YK5 and Meyer Caves (Bryan 1955), and its presence is indicated by a transitional component at Sunset Creek (Cultural Component III). It is coeval with the terminal portion of the Cold Springs Phase in the southern Columbia Plateau and is characterized by a set of similar traits, including finely made leaf-shaped and large side-notched projectile points.

3. The Cold Springs Phase is succeeded by the Rabbit Island Phase which is represented at such sites as Sunset Creek and Schaake Village (Swanson 1962a; 1962b). It is characterized by rectangular-stemmed projectile points in association with a variety of leaf-shaped points and knives similar to those from post-Cold Springs components in the southern Columbia Plateau. Unifacially flaked cobble implements are rare or absent, but there is some evidence to suggest that implements similar to edge-ground cobbles still persist (Appendix A: Edge-Ground Basalt Spall, Daugherty 1952:382).

In the western Cascade Mountains, the sole reliable sequence comes from the Fraser River valley at DjRi3 (Borden 1957; 1961; 1962). Here components with C-14 dates ranging from 7500 to 9000 radiocarbon years B.P. have yielded large leaf-shaped projectile points and knives which are typologically similar to the leaf-shaped specimens from the Vantage Phase. These are associated with flake scrapers and unifacially flaked cobble implements; edge-ground cobbles do not occur, [16/17] and the faceted platform technique has not been reported.

The early components at DjRi3 are followed by a hiatus between ca. 7500 and 4000 B.P. Succeeding components contain contracting stemmed projectile points similar to the Rabbit Island Stemmed and long, narrow leaf-shaped projectile points. These are associated with ground slate knives, ground chisels, and a variety of flake tools. Personal knowledge of the flaking detritus suggests that a faceted platform technique somewhat different than those used in the Columbia Plateau may be in use.

Further to the south, along the Puget Sound littoral, the oldest known assemblages are not yet dated. They are broadly similar to the early materials from DjRi3, being characterized by a variety of leaf-shaped projectile points, flake scrapers, and unifacially flaked cobble tools (Butler 1961; Thomson 1961). The leaf-shaped projectile points vary greatly in size and include some small specimens with serrated edges. Edge ground cobbles do not occur and there is no reported occurrence of blades with faceted striking platform remnants.

After a hiatus of unknown duration, the earliest known foothills components contain radically different assemblages. The earliest assemblage, from the Marymoor Site (Greengo 1965), contains side-notched projectile points similar to those from the Cold Springs Phase in the Columbia Plateau and leaf-shaped projectile points very similar to Cascade points. Of 25 illustrated leaf-shaped projectile points, none are serrated and possess easily identifiable bases. As a group they tend to be slightly larger than Cascade points from the Plateau and range from 3.48 to 6.70 cm. in length; average, 4.75 cm. They are also wider than Cascade points, ranging from 1.46 to 2.18 cm. in width; average 1.92 cm. Thickness ranges from 0.64 to 0.89 cm., and averages 0.72 cm. Length/width indices tend to be lower than those of true Cascade points, ranging from 1.81 to 3.25; average 2.47. These ranges overlap with those for true Cascade projectile points and open the possibility of expanding the type definition. Unfortunately, there is no information on base edge-retouching or on the possible association of blades with faceted striking platforms. In his commentary, Greengo (1966: 12) mentions that some of the specimens have striking platform remnants intact at their bases, a fact which suggests they may have been based on blades of some type. If these leaf-shaped and side-notched projectile points are related to their counterparts in the Columbia Plateau, it is probable that there was diffusion from the Plateau to the western Cascades during the Cold Springs Phase.

The Marymoor points are associated with a wide variety of other flake tools, including stemmed and corner-notched projectile points, and microblades. C-14 dates indicate that the site may have been occupied between ca. 3000 and 1500 B.P., but the association of these dates and the side-notched and leaf-shaped projectile points is not yet completely clear (Greengo 1966).

The second major assemblage, from the Duval site (C. M. Nelson 1962b), is characterized by large corner-notched projectile points in association with lesser numbers of stemmed and triangular projectile points; a few leaf-shaped knives also occur, but they do not have any of the characteristics of Cascade points. Associated tools include unifacially flaked cobble implements, adzes, gravers, scrapers, drills, microblades, and side scrapers based on large blades. Where platform remnants are left intact on the large blade tools, they are unfaceted; large, unworked blades have not yet been recovered. Among the flakes and flake tools there are a few specimens with faceted striking platform remnants, but none are true blades and none display edge-ground striking platforms (author's note). However, two of the microblades, both quartz crystal, do have finely faceted striking platforms (author's note 1966: specimens 799 and 801). These platforms are not edge-ground. [17]

[18] Typological comparisons suggest that the Duval assemblage is more recent than the bulk of the assemblage from the Marymoor site.

The earliest marine components, from Rosario Beach (Bryan 1963) and Comet Bay (Bryan 1963; C. M. Nelson 1962a), also contain a variety of leaf-shaped projectile points which are found in association with triangular and rectangular-stemmed projectile points, barbed and unbarbed bone points, composite harpoons, adzes, wedges, and abrasive stones. Unifacially talked cobble tools occur, but are rare. Edge-ground cobbles and the faceted platform technique are absent. These com¬ponents are believed to date from ca. 1500 to 3500 B.P.

This is what we generally know of the early sequences from the Columbia Plateau and the Cascade cordillera, and it is against these facts that the Old Cordilleran Culture must be evaluated. But what is the Old Cordilleran Culture? Recently Butler (1965: 1127) reviewed the content of the Old Cordilleran Culture concept and enumerated the following tangibles which are said to be characteristic of the Old Cordilleran Culture: (1) the Cascade projectile point, (2) a variety of oval knives, (3) generally nondistinctive cutting, chopping, and scraping implements, (4) blade tools, such as end-of-blade scrapers, (5) unifacially beveled antler wedges, (6) other unspecified tools of bone and antler, (7) the edge-ground cobble complex, (8) hunting of modern species of artiodactyla, including deer, elk, antelope, and sheep, (9) snaring or trapping of birds, (10) use of root crops such as camas and kouse, (11) fishing, and (12) use of fresh water mussels and snails. Since these are the characteristics of the Old Cordilleran Culture, we can reasonably expect them all to be found associated with one another in each of the major geographic areas in which the Old Cordilleran Culture is said to exist. These areas include the western Cascade Mountains, the southern Columbia Plateau, and the adjacent portion of Idaho (Butler 1965:1126-27). In 1961, Butler strongly inferred that portions of the northern Great Basin might also be characterized by the Old Cordilleran Culture. However, since he does no mention this area in subsequent publications, it is by no means clear what relationship it bears to the Old Cordilleran Culture concept. Let us now review the distribution of the 12 traits listed above to see if they are found in association with one another in the areas in which the Old Cordilleran Culture is supposed to have existed. We will begin with the Cascade projectile point since it is considered the single most important characteristic of the Old Cordilleran Culture (see Butler 1961; 1962a:8).

Definition of the Cascade Projectile Point Type. But what is the Cascade projectile point? In 1961 Butler stated (28-29):

"When I first examined these points, they reminded me of the Lerma type illustrated in the Handbook of Texas Archaeology as well as the Type 11 A points in the Upper Klamath Lake Region. Early in 1958 photographs of 58 of the Indian Well I specimens were sent to Krieger for identification as possible Lerma Points. Krieger recommended sending the material to MacNeish, who had originally established the type. MacNeish wrote in reply that all but the serrated specimens were "A-1" Lerma points, and he noted that points of this type were found at scattered locations along the cordilleras of the New World.... Because of the problems inherent in our limited knowledge of the distribution of these forms, Krieger has suggested that the Pacific Northwest specimens be given a separate type designation. Hereafter, these will be referred to as Cascade points. In addition to the specimens from Indian Well I, the Cascade type includes the Type 11A points from Kawumkan Springs, the leaf-shaped points from Fort Rock and Cougar Mountain Caves, and similar specimens from the Five-Mile Rapids site."

Butler (1961; 1962a) subsequently included the leaf-shaped projectile points from Hat Creek, Ash Cave, Cold Springs, Lind Coulee, Haller Lake, the Olcott site, DjRi3, and Weis Rock-shelter in the newly proposed Cascade projectile point type. The type assemblage for all these [18/19] specimens is Indian Well I (Butler 1961; 1962a:36), so the fundamental question is: Are all of the projectile points which have been called Cascade points similar enough to be considered typologically identical?

A close inspection of the comparative data from the southern Columbia Plateau discloses that the assemblage from Indian Well I is beyond doubt typologically identical with the early Altithermal assemblages from Hat Creek (Shiner 1953; 1961), Ash Cave (Butler 1958b; 1961; 1962a), Windust Caves (H. S. Rice 1965), Marmes Rockshelter (Daugherty 1965: personal communication), Thorn Thicket (C. M. Nelson 1963; Roderick Sprague 1966: personal communication), and Weis Rockshelter (Butler 1962a). Each of these assemblages contains projectile points with the following characteristics. (1) They are leaf-shaped in outline and possess easily definable, retouched basal areas. (2) They are finely pressure flaked. (3) They are manufactured from blades with faceted and frequently edge-round striking platforms (Butler 1959:13; C. M. Nelson 1965a: Fig. 1). These blades tend to be parallel sided or roughly diamond-shaped in outline, and have wide platform remnants in relation to overall blade width. (4) Many are finely serrated. (5) The widest part of the point may occur at the junction of the body and the base, or anywhere within the lower one-third of the body. (6) The transverse cross section commonly varies from lenticular to rhomboidal, but plano-convex examples occasionally occur. (7) The longitudinal cross section is usually lenticular, though plano-convex specimens also occur. (8) Length most commonly ranges from 3.0 to 5.5 cm., but specimens as small as 2.5 cm. and as large as 6.7 cm. occasionally occur. (9) Width commonly ranges from 1.1 to 1.8 cm., but specimens as narrow as 0.9 cm. and as wide as 2.2 cm. are known to occur. (10) Thickness commonly ranges from 0.4 to 0.8 cm., but specimens as thin as 0.3 cm. and as thick as 0.9 cm. also occur; the average thickness falls between 0.5 and 0.6 cm. (11) Length/width indices commonly range from 2.4 to 4.0, although specimens occasionally have length/width indices as low as 2.2 and as high as 4.4.

Using even the rudimentary definition outlined above, it is obvious that the Cascade projectile point type embraces a highly distinctive set of morphological characteristics and technological practices. It is in no way a catch-all category for "non-diagnostic" leaf-shaped knives and projectile points or a term which should be applied indiscriminately to ancient assemblages dominated by leaf-shaped forms. With this in mind, let us consider those leaf-shaped projectile points to which the type name has been applied.

Distribution of Cascade Projectile Points. Besides the early Altithermal components which are unquestionably dominated by true Cascade points, Butler mentions three other definite occurrences in the Columbia Plateau: (1) a single projectile point from Lind Coulee (Daugherty 1956a), (2) the early leaf-shaped points from the Five Mile Rapids site (Cressman 1960), and (3) the leaf-shaped projectile points from Cold Springs (Shiner 1961). The specimens from Cold Springs as well as other Cold Springs Phase components, including those at Windust Caves, Three Springs Bar, 45C01, Marmes Rockshelter, and Weis Rockshelter, possess all of the characteristics of the Cascade type.

The possible Cascade point from Lind Coulee (Daugherty 1956a: Fig. 20, 3) is a fragmentary specimen which might be the stem of a large stemmed-lanceolate point or the base of a leaf-shaped projectile point. It is slightly edge-ground and possesses broad, shallow flake scars, traits which are not characteristic of Cascade points. Moreover, it possesses no definable retouched base and is not serrated. Like the other projectile points from Lind Coulee, it probably was manufactured from a blade with a faceted and possibly edge-ground striking platform remnant. But unlike the small blades associated with known Cascade components, this blade would have been large, diamond or triangular shaped, with a physically small striking platform remnant in relation to blade width. Conclusion: this specimen is not a Cascade projectile point. [19]

[20] It is impossible to tell from the published data what the distribution of Cascade projectile points is in the Five Mile Rapids site at The Dalles. Butler cited Cascade points in the earliest component, but David L. Cole (1964: personal communication) maintains that they occur much later in the sequence. The earliest component is dominated by lanceolate and leaf-shaped projectile points which are short and broad with wide, shallow flake scars and little or no retouching. Such projectile points do not fall within the range of variation of Cascade points.

There is also a large Anathermal assemblage of stemmed-lanceolate, lanceolate, and leaf-shaped projectile points from Windust Caves (H. S. Rice 1965). One of the illustrated specimens from this assemblage (Fig. 13, d) appears to be serrated and may be a Cascade point. The Windust assemblage should be reexamined in order to determine if Cascade points are present in association with the lanceolate projectile points in the earliest components.

Butler cites the leaf-shaped projectile points from Kawumkan Springs (Cressman 1956), Fort Rock Cave (Cressman et al., 1940; Cressman 1942), and Cougar Mountain Cave (Cowles 1959), as evidence for the presence of Cascade projectile points in the northern Great Basin. The specimens from the Kawumkan Springs site vary tremendously in size and morphological form, but they do not include any which fully satisfy the definition of the Cascade point type (see Cressman 1956: Fig. 45). Moreover, there is no evidence to suggest that they are associated with the faceted platform technique. The specimens from Cougar Mountain Cave (Cowles 1959: Pl. 1) do not have readily identifiable bases and they are not serrated. The flaking is observably different from that of Cascade points and there is no report that the specimens were manufactured on blades with faceted striking platforms. Moreover, the Cougar Mountain points range from 7.6 to 13.9 cm. in length, measurements well above the range for known Cascade assemblages.

Butler (1962a:8-9) states that 44 leaf-shaped projectile points were recovered from beneath the volcanic ash at Fort Rock Cave. These specimens are described as carefully made, doubly pointed, thick, frequently serrated, between 3.8 and 8.9 cm. long and between 1.3 and 2.3 cm. wide. For his description of Fort Rock Cave, Butler cites Cressman, Williams, and Krieger (1940) and Cressman (1942). A search of these and other works (Cressman 1939; 1940a; 1940b; 1947) reveals no adequate illustrations or descriptions of these 44 specimens, so it is not possible to ascertain if any true Cascade points are present. However, there is no report of the faceted platform technique; several flake tools are illustrated which have unfaceted platforms.

Conclusion: Cascade points do not occur at the Kawumkan Springs site or Cougar Mountain Cave and their presence is doubtful at Fort Rock Cave. The presence of Cascade points in the northern Great Basin cannot be demonstrated.

Along the western Cascade Mountains, Butler cites the early material from DjRi3 and the Olcott site as evidence for the early occurrence of Cascade points along the Cascade cordillera. The specimens from DjRi3 (Borden 1961: Pl. 2), Zones E, G, and I, possess none of the characteristics of Cascade projectile points. They tend to be larger and wider; they probably are percussion flaked; they are not serrated; they do not possess the Cascade type of point base; and there is no reported evidence which suggests that they were manufactured on blades or are associated with a faceted platform technique. Conclusion: the specimens from DjRi3 are not Cascade points.

Just down river from DjRi3, at Yale, there is an early component with quite sophisticated technology executed on local volcanics, including large, serrated leaf-shaped points that might be considered Cascade points.

Assemblages from the Olcott and associated sites (see Thomson 1961) are dominated by projectile points very similar to those from DjRi3. However, small numbers of serrated specimens also occur which are similar in outline and size to true Cascade points. Unfortunately, there is no information of technique of manufacture or the occurrence of the faceted platform technique. [20]

[21] In addition, Cascade-like projectile points occur in a later assemblage from the Marymoor site (Greengo 1966: Fig. 6) where they are associated with side-notched projectile points very similar to the Cold Springs Side-Notched type. It is not known if these leaf-shaped points satisfy the technological associations definitive of true Cascade points, but it is quite possible that they do. In any case, the Marymoor assemblage strongly suggests diffusion from the Plateau to the western Cascades between 5000 and 3000 B.P.

Conclusion: small numbers of possible Cascade points are found at some sites in the Western Cascade Mountains, but further work is needed to plainly demonstrate the presence or absence of the type.

In summation. Cascade projectile points are known to occur only in the southern Columbia Plateau and the adjacent portion of Idaho during the Altithermal. Unconfirmed or dubious reports exist for some Anathermal components in the Southern Columbia Plateau and components of unknown age in the western Cascade Mountains of Washington. Reports of Cascade points from the northern Great Basin are completely unwarranted on the basis of our present knowledge.

The Age of Cascade Components. Up to the present, I have referred to Cascade-point-dominated components in the Southern Columbia Plateau as Altithermal in age. What is the basis for the assignment in age, particularly in light of the evidence adduced by Butler (1962a:77-80) which suggests that such components are much older in The Dalles locale? Butler dates the Cascade point-bearing components at The Dalles at ca. 8000 to 11000 B.P., the component at Ash Cave between 8000 and 9000 B.P., and the component at Weis Rockshelter at ca. 7500 to 3500 B.P., and uses these estimates to support the idea that the Old Cordilleran Culture spread "eastward across the Columbia Plateau in early post-glacial times" (1962a:80). Let us examine these age estimates.

The two major occurrences of Cascade projectile points in The Dalles locale are at Indian Well and Five Mile Rapids. There are no C-14 or geological age estimates for the component at Indian Well, so Butler's age estimate of ca. 8000 to 11000 B.P. hinges on his interpretation of the Five Mile Rapids sequence, for which he claims that Cascade points occur in the earliest component. Cole (1964: personal communication), however, states that Cascade points do not appear in the earliest component, which is dominated by leaf-shaped and lanceolate projectile points similar to those from Lind Coulee, Windust Caves, and Marmes Rocksehlter. Since such projectile points do not occur at Indian Well I, the Cascade component at this site is certainly not comparable with the earliest assemblage from the Five Mile Rapids site, where very small numbers of Cascade points may or may not occur in the context of a different tool assemblage.

The age estimate of 8000 to 11000 B.P. is based on the earliest C-14 date from the Five Mile Rapids site. This date was obtained from a composite sample taken from Stratum I and is 9785 ± 220 radiocarbon years B.P. (Cressman 1960). Since it is a composite sample, Cressman believes that the bottom of Stratum I must date from 11000 to 11,500 B.P. (1960:66). There is no way of substantiating this opinion. Thus, an age estimate as early as 11,000 years is not only undemonstrable, it is applicable only to a pre-Cascade component and does not date the association of tools which identifies the Early Dalles Phase of the Old Cordilleran Culture.

Ash Cave has a single associated C-14 date of 7940 ± 150 radiocarbon years B.P. (Butler 1962a:71). The dated sample directly underlies volcanic ash believed to be from the eruption of Mount Masama about 6500 B.P. (cf. Fryxell 1965). Since the date comes from the top of the Cas¬cade component, Butler believes that the bulk of the component antedates 8000 B.P. [21]

[22] At Weis Rockshelter the lower part of the Cascade sequence is dated at 7340 ± 140 radiocarbon years B.P. It has led Butler to the belief that such components date from as early as 7500 B.P. in west-central Idaho. In fact, the date only provides information on the occupation of Weis Rockshelter; older Cascade components may exist elsewhere in the region.

Since 1962 a number of dates have become available from the Cascade component at Marmes Rockshelter. They underlie an identified deposit of Mount Mazama volcanic ash and range from 6200 to 8000 B.P. (Fryxell and Daugherty 1962; 1965: personal communication).

The available C-14 dates suggest that Cascade components first appear between 7500 and 8000 B.P., but there are not enough dated occurrences to demonstrate where they first appear or the direction(s) in which they spread. In the southern Columbia Plateau they persist until about 6500 B.P. and in adjacent Idaho the evidence suggests that they may have persisted until as late as about 3500 B.P. (Butler 1962a).

The Occurrence of Cascade Projectile Points in the Early Phases of the Southern Columbia Plateau. Although Butler treats The Dalles Cascade components as comtemporaries with the stemmed-lanceolate dominated component at Lind Coulee (1961), there is now abundant evidence to show that these components represent two successive phases in the southern Columbia Plateau. The earlier phase, properly called the Lind Coulee Phase after its type site, is characterized by stemmed-lanceolate and lanceolate projectile points which frequently have ground edges. Lesser numbers of leaf-shaped projectile which may or may not include a few true Cascade points also occur. This phase is represented at Five Mile Rapids, Lind Coulee, Windust Caves, Marmes Rock-shelter, and Thorn Thicket, and corresponds roughly with the Anathermal (ca. 8000-11,000 B.P.). The second phase, properly called the Indian Well Phase after its type site, is characterized by Cascade projectile points in association with small numbers of lanceolate and other leaf-shaped forms. It corresponds roughly to the first part of the Altithermal (about 8000-6500 B.P.) in the southern Columbia Plateau, and it is represented at Indian Well, Hat Creek, Windust Caves, Ash Cave, Marmes Rockshelter, Thorn Thicket, and Weis Rockshelter.

The Indian Well Phase is followed by the Cold Springs Phase (see below for details), in which Cascade projectile points are found in small quantities in association with large Cold Springs Side-Notched projectile points. In the southern Columbia Plateau, the Cold Springs Phase roughly corresponds to the end of the Altithermal (about 6500-4000 B.P.). Later phases do not contain Cascade points.

Leaf-Shaped Projectile Points in the Pacific Northwest. In the proceeding paragraphs it has been demonstrated that the Cascade projectile point is restricted to the southern Columbia Plateau, that it is probably not the earliest projectile point type in the Columbia Plateau, and that there is no conclusive evidence which demonstrates where it originated or how it diffused. Proponents of the Old Cordilleran Culture concept may chide me for this, saying that even if these statements are true, they do not effect the Old Cordilleran Culture hypothesis because they ignore and cloud the broader issue of the distribution and meaning of leaf-shaped projectile points in the Pacific Northwest. But facts are not capable of clouding anything, and the facts are that the leaf-shaped projectile points from the southern Columbia Plateau, the northwestern Columbia Plateau, the lower Fraser River valley, the Puget Sound littoral, and the northern Great Basin are obviously different from one another and are associated with different techniques of manufacture, different tool assemblages, and ecologically different environments in which differing economic and social adaptations were almost certainly made. For these reasons, the Old Cordilleran Culture hypothesis does not provide a model that can satisfy even the limited amount of data which is available. [22]]

[23] Gruhn (1962) states that in actual fact there was “evidence only for the existence of a projectile point tradition, which was probably held in common by a number of distinctive cultural groups...” Butler (1962a:8) replied by asking "what is the cultural significance of a projectile point tradition? Such a tradition would seem to indicate cultural continuity, historical relationships, and shared patterns of behavior. This definition can be extended to include an 'archaeological' culture... An archaeological culture represents a temporal segment of a complex set of shared be¬havioral patterns expressed in an assemblage of cultural traits occurring in a geographic continuum.”

In fact, however, no projectile point tradition, which embraces shape, size, and method of manufacture, has been demonstrated outside of the southern Columbia Plateau. Even if such a tradition did exist, there is every indication there is no "complex set of shared behavioral patterns" in the large area to which Butler attributes the Old Cordilleran Culture simply because there is no complex set of cultural traits archaeologically present in that area.

At best, the only reasonable hypothesis which can be constructed is that the idea of making projectile points which had a leaf-shaped outline diffused through adjacent portions of the Great Basin, Columbia Plateau, and Cascade cordillera at an early date. And that is the broader significnace of the distribution of leaf-shaped projectile points in the Pacific Northwest.

Oval Knives. Butler briefly refers to oval knives in 1959 (13) and 1961 (90: Fig. 4, Ovate knives), and describes a series of "broad oval points" from Weis Rockshelter in the following terms:

"This is a highly variable group comprising 41 complete and fragmentary examples. All are generally oval in outline and most are bifacially flaked. Some are well made and rather thin; others are crudely chipped and are often quite thick... The largest specimen... although bifacially flaked and elliptical in outline, may have been some sort of scraper rather than a knife. All of the specimens are made of silica minerals. Size range: 3.7-13.2 cms. in length; 2.0-6.1 cms. in width; and 0.6-2.3 cms. in thickness. Distribution: 14 in Substratum 5e; 15 in 5c; 8 in 5a; 4 in Stratum 3; and 1 in Stratum 1..." (Butler 1962a:38-39).

The internal variation in size, shape, cross section, and technique of manufacture which exists in this type render it of little use as a criterion for defining phases or archaeological cultures. As Butler indicates, oval knives are distributed throughout the sequence at Weis Rockshelter. They are distributed throughout the Plateau where they occur at every time level. They are also widely distributed in the western United States.

Nondistinctive Cutting, Chopping, and Scraping Tools. The tools which Butler refers to include a wide variety of unifacial and bifacial cobble and flake implements. Since tools of this general sort have an almost universal distribution at an early time level and are commonly associated with more recent hunting and gathering societies, they can hardly be considered diagnostic in themselves. Since there are no adequate descriptive studies of such implements in the Pacific Northwest, the distribution of specific types cannot be used either to support or refute the Old Cordilleran Culture concept.

Blade Tools. Blade tools associated with Indian Well Phase components include Cascade projectile points, end scrapers, side scrapers, and utilized flakes. The blades, themselves, are large, tend to be roughly rectangular or slightly diamond-shaped in outline, and possess faceted and frequently edge-ground striking platform remnants (C. M. Nelson 1965a). [23]

[24] These blades are known to occur in components of the Indian Well and Cold Springs phases in the southern Columbia Plateau. They do not occur in the sequence from 45KT28 and they have not been reported from other sites in the northwestern Columbia Plateau, the western Cascades, or the northern Great Basin. The fact that they are a part of a sequence of blade forms in the Columbia Plateau strongly suggests a local development rather than transmission from surrounding areas such as the Cascade cordillera.

Unifacially Beveled Antler Wedges. Butler bases this trait on specimens from Weis Rock-shelter and Five Mile Rapids, which include specimens of problematical use as well as true wedges. Some of the specimens from Five Mile Rapids (Cressman 1960) may be associated with the Lind Coulee Phase component there. Antler splitting wedges are also known from the Cold Springs Phase (e.g. C. M. Nelson 1966), as well as the Rabbit Island, Quilomene Bar, and Cayuse phases. Although bone and antler are not commonly preserved in the foothills sites of the western Cascade Mountains, antler splitting wedges in the earliest known coastal components suggest that they may have been important at an earlier period in the foothills. The only published evidence for antler splitting wedges in the northern Great Basin comes from sites in the Klamath area such as Kawumkan Springs (Cressman 1956). They do not occur at Fort Rock or Cougar Mountain caves.

Conclusion: although antler splitting wedges are wide spread in the Pacific Northwest at almost all time levels, they are absent in the more arid portions of the northern Great Basin and their existence cannot be conclusively demonstrated or refuted for the early period in the western Cascades due to the preservation problem.

Edge-Ground Cobble Complex. Edge-ground cobbles occur in the southern Columbia Plateau in association with Lind Coulee, Indian Well, and Cold Springs phase components such as those at Five Mile Rapids (Cressman 1960), Goldendale (Warren et al., 1963), Marines Rockshelter (Daugherty 1965: personal communication), 45C01 (C. M. Nelson 1966), Thorn Thicket (Sprague 1966: personal communication), and Weis Rockshelter (Butler 1962a). Edge-ground cobbles also occur on the Upper Columbia River at an early date (Swanson 1962a). However, there are no known occurrences in the western Cascade Mountains or in the northern Great Basin.

The edge-ground cobble complex of which Butler (1962a:44-46; 1963; 1965) speaks consists of the association of edge-ground cobbles with circular, well rounded river cobbles which have slightly convex, highly polished faces. This association occurs at Weis Rockshelter, the only site at which the cobble bases are known to occur. However, since the bases are easily mistaken for unaltered cobbles, it is possible that they have been overlooked at many other sites.

Use of Root Crops. Butler's comment that roots such as kouse and camas were an important part of the subsistence pattern is based on the equation of the edge-ground cobble complex with the grinding of soft food crops. The following facts are pertinent to this hypothesis. (1) The association of edge-ground cobbles with polished cobble bases is known from only one site. In order to establish a functional relationship between these tools it will be necessary to establish their archaeological relationship at many more sites. (2) Butler notes that the distribution of edge-ground cobbles is limited to areas in which camas and kouse are present and to sites which could have been used as camas or kouse gathering bases. Almost every site which exists within the geographic area from which edge-ground cobbles are known is sufficiently close to camas and kouse producing areas to qualify as a possible root gathering station. Moreover, there are areas, in which camas and kouse occur and were ethnographically utilized where edge-ground cobbles do not occur, including the Puget Sound littoral and the Fraser River valley. (3) The highly polished, nearly flat facets which occur on edge-ground cobbles and the highly polished, convex surfaces of the possible cobble bases are functionally poorly suited to the preparation of roots which are easily pulverized but ground only with difficulty. [24]

[25] The function of edge-ground cobbles and the use of root crops during the Indian Well Phase cannot be demonstrated with the existing evidence.

Hunting of Modern Artiodactyla. The hunting of animals such as deer and elk was undertaken by hunters and gatherers and formative horticulturalists wherever such animals occurred. In itself, this trait cannot be considered distinctive of any phase, period, or archaeological culture in the Plateau.

Snaring or Trapping of Birds. Butler bases this characteristic primarily on the occurrence of bird bones and girdled "bolas" from the Five Mile Rapids site (Cressman 1960). Since the presence of an Indian Well Phase component is in question at this site, the extension of bird trapping from it to the Old Cordilleran Culture is a very doubtful procedure. Moreover, there is no evidence for extensive bird trapping at any other Indian Well Phase component, though occasional bird bones do occur. It is also uncertain that the birds represented in the fauna from Five Mile Rapids were trapped instead of hunted since the associated girdled pebbles might have been weights for fishing nets or atlatls.

As bone is not preserved in the acid soils of the western Cascades, there is no faunal evidence to substantiate or refute bird hunting at DjRi3 or the Olcott assemblage sites, but girdled "bolas" do not occur. Bird bones are relatively common in Great Basin assemblages, but girdled "bolas" are not found.

Fishing. Like the trapping of birds, Butler relies heavily on the Five Mile Rapids sequence to demonstrate the importance of fishing in the Old Cordilleran Culture. Since Butler's basic interpretation of the sequence is questionable, the importance of fishing in the Old Cordilleran Culture is an open question.

Elsewhere, at Indian Well Phase components, fish bones are absent or uncommon.

Utilization of Fresh Water Mussels. The remains of fresh water mussels occur throughout the Columbia Plateau during every period of prehistory. Poor preservation makes it impossible to know whether they were in use at an early time period in the western Cascades. Cressman (1956:387) reports that they were commonly used in the Klamath area but are rare in the Great Basin proper.

Conclusion. Of the twelve activities and items of material culture considered characteristic of the Old Cordilleran Culture (Butler 1965:1127). the snaring or trapping of birds, the use of roots such as kouse and camas, the edge-ground cobble complex, and the association of the edge-ground cobble complex or of edge-ground cobbles with food grinding are conjectural, speculative traits which cannot be demonstrated for any early phase or component in the Pacific Northwest, including the Lind Coulee and Indian Well phases in the southern Columbia Plateau. The first clear evidence which can be used to support the utilization of root crops is the introduction of mortars and pestles during the Cold Springs Phase.

Oval knives, nondistinctive cutting, chopping, and scraping tools, unspecified tools of bone and antler, and the utilization of modern artiodactyla, birds, fish, and fresh water mussel shells are characteristics so general that they apply to virtually all phases of prehistory in Columbia Plateau and adjacent areas. Although important for reconstructing economic organization, they cannot be used to define archaeological cultures or phases on the simple basis of presence or absence.

This leaves Cascade projectile points, a specific type of blade with a faceted platform, and edge-ground cobbles, three items of material culture whose association can be traced in time and [25/26] space. And these traits occur together only in the southern Columbia Plateau and the adjacent portion of Idaho in the Indian Well and Cold Springs phases. However, the Cold Springs Phase is also noted for Cold Springs or Bitterroot Side-Notched projectile points, manos, grinding slabs, hopper mortars, and pestles, items which do not occur in Indian Well Phase components and are not said to be characteristic of the Old Cordilleran Culture. Therefore, if the Old Cordilleran Culture corresponds to any set of hard data which actually exists in reality, it is the assemblages and compo¬nents of the Indian Well Phase. It follows that the Old Cordilleran Culture concept can be used as a hypothetical model to explain the Indian Well Phase and to propose meaningful research which will help to clarify the meaning of the Indian Well Phase. Obviously, such research should be conducted in the eastern and western foothills of the Cascade Mountains adjacent to the southern Columbia Plateau. If work in these areas reveals Indian Well Phase components which are demonstrably older than their counterparts in the Columbia Plateau, there will be important evidence to support the Old Cordilleran Culture concept as it applied to the Indian Well Phase.

Since the archaeological content of the Old Cordilleran Culture does not correspond to any set of archaeological data in the Northwest except the Indian Well Phase components, it cannot be used as an explanation or model for the prehistory of the northern Cascades, the northern Great Basin, or the northern Columbia Plateau.

Alternative Hypothesis. The Indian Well Phase may be explained in terms of an internal development within the southern Columbia Plateau. With the exception of the Cascade projectile point and the associated blades with faceted striking platforms, archaeological sequences at Five Mile Rapids, Windust Caves, Marmes Rockshelter, and Thorn Thicket demonstrate that the characteristics of the Indian Well Phase were present in the southern Columbia Plateau prior to its emer¬gence. Also a variety of leaf-shaped points was present out of which Cascade projectile points could have developed. The blades from which Cascade projectile points were made also have antecedents in the blades and flakes of the Lind Coulee Phase. In addition, lanceolate projectile point types characteristic of the Lind Coulee Phase survive in small quantities.

The continuity in the Anathermal-early-Altithermal archaeological sequence in the southern Columbia Plateau strongly suggests an internal development, an hypothesis which would be supported by work in areas adjacent to the southern Columbia Plateau which revealed no Indian Well Phase components or the presence of such components at a more recent date.

Addenda. Newman (1966) reports the existence of an Indian Well Phase component on the western slopes of the Cascade Mountains in the Willamette Valley of Oregon. A date of 7910 ± 280 radiocarbon years B.P. is associated with the earliest occupation of the site and indicates that the Indian Well Phase was well established west of the Cascades at a time comparable to its expression in the southern Columbia Plateau. As the date is neither older nor younger than comparable dates from the Columbia Plateau, neither hypothesis reviewed above is confirmed. [26]

TOP

LAST REVISED: 14 OCT 2018