Environment, Prehistory & Archaeology of Mount Rainier National Park, Washington
Greg C. Burtchard

Chapter 5:
PREHISTORIC SITE DISTRIBUTION & HOLOCENE LAND-USE PATTERNS ON MOUNT RAINIER & THE SOUTHERN WASHINGTON CASCADES

MOUNT RAINIER SITE TYPES AND SITE DISTRIBUTION PATTERNS

Within the maze of tables, figures and associated text that summarize prehistoric lithic data in Chapter 4, are two analyses that group sites by quantitative distinctions in raw material frequency (see Table 4.6, Figure 4.3 and accompanying text entitled Material Variability and Site Function, and Table 4.8 and associated text entitled Assemblage-based Site Groups). While use of numerical data such as these provide a sheen of quantitative objectivity to site groups so derived, it is important to recognize that Mount Rainier data presently are quite thin. Site types do not have an inherent, easily recognizable reality that can be used uncritically as units of analysis. Rather, different site types are the culmination of processes by which empirically observable archaeological phenomena are grouped into like categories for other descriptive and research purposes (cf., Leonard and Reed 1993). Such taxonomic models are useful to the extent that they subsume the bulk of variability in the data and help us generate meaningful–and ultimately falsifiable–ideas about past processes of organized human use of an area.

Sample size limitations, in terms of both site number and low surface artifact density, presently prevent site taxonomies from achieving the level of quantitative rigor necessary for truly powerful arguments based on empirical grounds alone. Such concerns notwithstanding, modeling functional site distinctions for an area such as Mount Rainier (under the rubric of site type variability) provides a useful tool to investigate subsistence strategies. Furthermore, overlapping patterns apparent in the Mount Rainier data set suggest presence of genuine functional variation between the Mount Rainier archaeological assemblages. These apparent patterns are the subject of this section, suggestions for refining the system's quantitative rigor are offered in the final chapter of this volume.

Here, site and isolated find information is reconsidered in light of environment and land-use arguments developed in Chapter 2 to build a working model of site types and distribution across Mount Rainier National Park. Even with present data limitations, I suggest that: 1) deductive arguments predicting long-term redundant use of immature, upper elevation landscapes on ecological/resource grounds are theoretically sound; 2) the site taxonomy proposed below subsumes the widest range of location, assemblage and environmental information presently available in and near the Park; and 3) current site data are generally consistent with the site distinctions proposed. The model assumes these considerations to be true. As a working model, however, the scheme should not be considered immutable, but rather taken as a starting point to be refined or changed as additional archaeological data become available.

The 10 part taxonomic model outlined below distinguishes nine basic site types and an isolated artifact category. It offers functional, content and location expectations for each. To the extent possible, I have used terminology compatible with Binford's (1980, 1983) model for hunter-gatherer settlement systems. Consideration also has been given to Ubelacker's (1986:150, 198-200) and Benson and Lewarch's (1989) site type distinctions for the eastern slope of the southern Washington Cascades. In cases where neither system was appropriate, I have tried to keep terms simple and descriptive. Please refer to Burtchard and Hamilton (1998) for site specific detail relevant to prehistoric localities included here.

Type 1: Multi-task, Mixed Group, Residential Base Camps or Residential Field Camps

Predicted Site Function
Residential locations are expected to be base-camp sites repeatedly used by mixed age and sex groups exploiting multiple floral and faunal resources. Because these groups may consist of only a portion of a larger lowland based group (at least in the mid to late Holocene) they may qualify as residential field camps in Binford's (1983:346) sense of the term. Occupation timing and duration should have been linked to late summer/early autumn availability of key upland resources; principally elk, deer, bear, marmot, game birds (ptarmigan and grouse), and huckleberries (and perhaps alpine lilies). Goats are available over a longer season, but are expected to have been exploited concurrently with other game.

Expected Assemblage Characteristics
Lithic assemblages should be varied with a broad mix of both light and heavy tools. Debitage density and raw material variety should be high. Features should include hearths, a possible mix of small shelter depressions and/or post-molds, and plant and animal processing features such as huckleberry drying pits (see below).

Location
Residential base/field camps are expected to be located in upper forest or lower sub-alpine ecotonal settings in order to compromise access to varied upland resources, minimize distance to lowland villages (late Holocene), and to moderate susceptibility to unpredictable high-elevation weather patterns.

Current Representation in the Park
Five of the presently documented Mount Rainier sites are classified as possible residential base camps. Three of these meet basic assemblage expectations noted above–the Sunrise Ridge Borrow Pit Site (FS 90-01) and Little Sunrise Lake Site (FS 95-11) in the Park's northeast quadrant, and Forgotten Creek Site (FS 95-10) in the southwest quadrant. Although heavy tools are not present in the surface assemblage, other characteristics tentatively justify inclusion of the Buck Lake Site (FS 71-01) in the northeast quadrant and Tipsoo Lake One (FS 88-01) on the boundary between the Park's northeast and southeast quadrant. All five sites are located immediately below or at upper forest/subalpine boundaries. The location of plausibly residential sites on various sides of the mountain is consistent with use by varied, socially distinct populations as suggested by Smith (1964) in his Ethnographic Guide to the Archaeology of Mount Rainier.

Type 2: Limited-task Field or Hunting Camps

Predicted Site Function
Prehistoric sites in this category are expected to have been places of short-term residence used by small, predominantly adult male hunting groups. Tasks should have been limited to those directly or indirectly associated with hunting and overnight residence; including tool maintenance, repair, and late stage manufacture. Associated uses may include moderate butchering and cooking activities, involving a low frequency of early stage core reduction of locally available materials. Apparent high use intensity is expected to be a function of repeated use events. Sites may occupy open or rockshelter settings.

Expected Assemblage Characteristics
Lithic assemblages are expected to be dominated by late stage debitage and light tools. Core manufacture and flake blank production, if any, should be limited largely to local materials near source localities. Heavy tools are expected to be absent or present in very limited number. Type 2 field camps should be associated with a moderate light tool to debitage ratio (present data suggest between 10% and 50%, and moderate raw material variety. Hearth features may or may not be present. Locations in alpine settings may be associated with stacked stone windbreak features or blinds (none currently documented).

Location
Site location should be biased heavily toward subalpine context. Because they provide construction-free shelter, rock overhangs and shallow caves should have been particularly desirable short-term camp locations. Rockshelter locations may have ranged more extensively from upper forest to lower alpine settings as dictated by geological, rather than floral, characteristics.

Current Representation in the Park
Eight currently documented sites are included as potential repeated use, short-term hunting camps. These include Fryingpan Rockshelter (FS 63-01), Berkeley Rockshelter (FS 86-02), Upper White River Trail Site (FS 95-03) and Yakama Park Rim Site (FS 95-04) in the northeast quadrant; and Mt. Pleasant Rockshelter (FS 72-02), Vernal Park Rockshelter (FS 74-01), Middle Spunkwush Lake (FS 95-08) and Mist Park Overlook (FS95-05) in the northwest quadrant. The absence of hunting camp localities in the southeast and southwest quadrants is believed to reflect greater subalpine landmass in the northern half of the Park, accentuated by small sample size. Early stage core reduction presently is represented only in rockshelters.

Type 3: Low Redundancy, Low Intensity Hunting Locations

Predicted Site Function
Type 3 hunting locations are expected to have functioned similar to short-term hunting camps noted above, but with very low intensity, limited task and/or single event use (or very limited reuse). Hunting stops may not have involved overnight stay.

Expected Assemblage Characteristics
Low intensity hunting locations are expected to be associated with overall low lithic debitage density and raw material diversity. Light tools should be present; heavy tools absent. In depositional environments like Mount Rainier, isolated finds may represent low intensity sites, particularly when represented as debitage or dual item combinations of tools and debitage. [31]

Location
A wider range of environmental zones (forest to alpine) is expected compared with moderate intensity residential hunting camps (Type 2). Overall, locations should continue to express a bias toward subalpine and secondarily to alpine settings.

Current Representation in the Park
Twelve localities are classified as Type 3 low use intensity sites. Eleven of these currently are documented as isolated finds. One is recorded as a site. These localities include Sunrise Creek (IF 01-72), Deadwood Lakes Pass (IF 01-75), Lower Deadwood Lake (IF 01-95), Upper Berkeley Park (IF 05-95), Grand Park One (IF 03-95), Grand Park Two (IF 04-95), Yakama Park One (IF 06-95), and Yakama Park Two (IF 07-95) isolated finds in the Park's northeast quadrant; the Windy Gap One Site (FS 90-03), and Yellowstone Cliffs (IF 01-68) and Mirror Lakes (IF 11-95) isolates in the northwest quadrant; and the Success Cleaver (IF 01-70) isolated points in the Park's southwest quadrant. Again, the tendency toward north and northeast settings is clear, increasing probability that distribution reflects genuine prehistoric use patterns.

Type 4: Butchering Locations

Predicted Site Function
Butchering sites are expected to be located near primary kill locations and used predominantly for initial game processing (hide removal, disarticulation and partial drying).

Expected Assemblage Characteristics
Butchering tool assemblages are expected to have a high fraction of light cutting and piercing tools to debitage (over 70% in the current sample). Heavy tools are not expected to be present. Flakes or flake tools may function as cutting and scraping implements. If located near lithic source material, a relatively large fraction of early stage flakes should be present, and should exhibit cutting/scraping wear and potentially retain blood residues on the cutting and scraping edges.

Location
Location should be biased toward kill sites in subalpine to alpine settings. If meat drying and flake production are important concerns at such sites, then locations may be expected to optimize distance to hunting areas and lithic sources. If meat drying is a concern, then sites also may be situated in exposed, windy settings (e.g., Frozen Lake).

Current Representation in the Park
Two localities currently are classified as butchering sites; primarily by virtue of assemblage characteristics and setting. These include the Frozen Lake Site (FS 86-01) and the Sarvant Glaciers Site (FS 90-02). Both are found in exposed, alpine contexts. The Frozen Lake site offers the most inferentially compelling assemblage of butchering-related tools and early-stage reduction flakes currently documented in the Park.

Type 5: Lithic Procurement and Lithic Reduction Locations

Predicted Site Function
Grouped in this designation are direct lithic procurement (quarry) sites located at the point of extraction, plus early stage reduction locations expected to be situated a short distance from source locations. Lithic procurement sites serve to reduce the need to transport heavy, complex tool kits into the mountains. Lithic reduction activities at these sites are expected to be limited to very early stage removal of cortical material and generation of debitage incidental to preparation of curated cores and implement preforms. Residence and hunting maintenance activities are not expected.

Expected Assemblage Characteristics
Quarry and reduction station assemblages are expected to be dominated by early stage flakes and shatter of single material type. Source material may or may not be immediately present. Finished tools should be absent or present in very low frequency.

Location
Quarries and reduction stations are expected to be situated near fine-grained lithic material sources, particularly in close spatial association with subalpine to alpine hunting areas.

Current Representation in the Park
Tum Tum Quarry (90-04) in the Park's southwest quadrant is currently the only lithic procurement or quarry site documented in the Park. Early stage, single event reduction stations are considered to be represented by a small lithic concentration at Tipsoo Lake (FS 95-01) in the northeast quadrant; and at Mother Mountain Lake 5554 (FS 95-06) and Windy Gap (FS 95-07) in the northwest quadrant.

Type 6: Stacked Rock and Talus Feature Locations (not included above)

To date, no stacked rock or talus pit features have been identified in the Park. They are common, however, in the Cascades and have been identified on upper elevation landscapes of Mt. Hood (see Winthrop et al. 1995, and Burtchard and Keeler 1991), in the North Cascades (Mierendorf pers. com. 1990), in the southern Washington Cascades east of the Park (Burtchard and Miss 1998) and at a number of other places in the Cascades and beyond. The category is retained here to accommodate the high probability that such features eventually will be documented at Mount Rainier as well.

Predicted Site Function
Functional interpretations of stacked rock and talus pit features are variable but tend to center on ceremonial or vision quest functions, hunting blinds, and territorial or travel markers. I generally favor explanations emphasizing direct material/ functional relationships. Accordingly, for stacked rock features observed on Mt. Hood, I emphasized 1) hunting-related use of alignments, semi-enclosures and enclosures situated on upper elevation scree slopes in ungulate habitat; and 2) travel and territorial markers (often historical) for cairns (Burtchard and Keeler 1991). Similar functions may be anticipated for Mount Rainier. Low stacked walls in exposed alpine settings may have functioned as temporary wind break shelters; simple rock alignments to anchor more ephemeral temporary shelters (e.g., historic period canvas tents). Citing informant interviews, Winthrop et al. (1995) emphasize ceremonial functions. Such use cannot be discounted. Indeed, multi-functional use is consistent with the structural variability characteristic of these features (see below). Specific functions are difficult to establish empirically. Expected Assemblage Characteristics Stacked rock and talus features are morphologically varied; typically appearing as piled and stacked stone alignments, circular enclosures, talus pits with low mounded edges, cairns and other similar features. The primary range of features anticipated for Mount Rainier are expected to be hunting blinds, temporary wind shelters and storage pits. Except for storage, features are expected to be associated with very low density, late stage lithic debitage and fractured projectile points associated with implement maintenance and repair.

Location
Stacked rock features should be biased toward upper elevation glacial rubble and scree slopes, particularly in association with alpine to subalpine elk and goat habitat.

Current Representation in the Park
None.

Type 7: Culturally Modified Tree Locations

Predicted Site Function
Culturally modified trees typically include trees blazed during the historic period to mark trail routes; bark stripped mature pines and Douglas fir used to create an insulation layer in earth ovens; and most important for present purposes, bark stripped (or peeled) cedar. Historically, cedar bark filled a variety of functions including use for rain-repellant clothing, woven baskets and mats, and durable expedient containers (see Stewart 1984). At Mount Rainier, Alaska yellow-cedar bark may have been striped for any of these functions or to manufacture low-investment berry containers for transport away from the mountain, thereby minimizing transportation weight and bulk during ingress.

Expected Assemblage Characteristics
Culturally modified trees (here emphasizing wholly or partially peeled cedars) typically appear in small groves of ax cut rectangular scars or as delta (M) shaped scars on the upslope side–with or without cut marks (see Burtchard et al. 1993:49-64 and Mack 1996). Other durable cultural debris generally are absent.

Location
The location of peeled cedars is constrained by habitat. Western red cedar (Thuja plicata) is most common in low elevation river valleys. Alaska yellow cedar (Chamaecyparis nootkatensis) habitat extends higher through mid to upper elevation forests to circa 5,500 ft. Alaska cedar effectively is limited to well watered valleys with northerly exposures. Because of plausible association with huckleberry collection and denser representation in the Park, peeled Alaska cedars are expected to be the most common.

Current Representation in the Park
No modified tree sites have been formally documented within Park boundaries. However, Carl Fabiani (pers. com. 1995) reports a small grove of peeled Alaska cedars on the banks of Shaw Creek in the Park's northeast quadrant. Working for the Gifford Pinchot National Forest, J.E.D. Garoutte located a grove of eight peeled Alaska cedars on the Park/Forest boundary on the north-facing slope of Laughingwater Creek at the edge of the Park's southeast quadrant (Lake Beverly Site, FS #15N10E-36/01 [McClure pers. com. 1996]). Finally, Janet Liddle (pers. com. 1996) reports a peeled cedar locality on the Park's southern boundary north of Skate Creek Road, west of Bear Prairie on the north shore of the Nisqually River.

Type 8: Plant Processing Locations

No unequivocal evidence of exclusively plant processing locations has been found in Mount Rainier National Park. However, there is no doubt that huckleberries were collected on the mountain in the ethnohistoric past, and were reportedly "...dried, packed in bags, and brought back to the valley homes for winter consumption." (Curtis 1911 in Smith 1964:150). Historically, huckleberry drying involved use of fire. Accordingly, drying features should leave distinct characteristics in the archaeological record. A number of probable huckleberry drying features have been identified on the flanks of Mt. Adams in the Gifford Pinchot National Forest (Mack and McClure 1996). An excavation report from a site north of Mount Rainier (Miss and Nelson 1995) claims to have identified a functionally similar feature at Mule Spring in the (appropriately enough) Huckleberry Mountains. Given abundant huckleberry habitat on Mount Rainier, ethnographic reference to huckleberry collection, and documented huckleberry features in the general vicinity, there is ample reason to expect that presence of such features ultimately will be documented within Park boundaries.

Predicted Site Function
In Mount Rainier National Park, plant processing localities are expected to be limited to huckleberry drying features; probably in association with residential sites as discussed below.

Expected Assemblage Characteristics
Huckleberry processing features investigated historically and archaeologically on Mt. Adams suggest two basic feature types: 1) an elongated trench with huckleberry laden mats on one slope facing a felled and fired log opposite; and 2) pole raised huckleberry (and meat) laden mats over an elongated fire pit fueled with scavenged wood. Rocks may be used to secure the mats and to store and transmit heat (Mack and McClure 1996). The remnant archaeological signature of such features should consist of these elongated, charcoal rich features, fire cracked rock and presence of charred huckleberry in macrobotanical samples. Chipped stone tools and debitage are not anticipated unless multifunctional activities, such as meat processing and drying and/or general residential activities also are taking place in the immediate vicinity of the plant processing feature(s).

Location
Huckleberry processing features should be linked to the distribution of most productive huckleberry habitat–blue and black huckleberries (V. ovalifolium and V. membranaceum) in midelevation forest burns, and dwarf huckleberry (V. deliciosum) in subalpine habitats. Because huckleberry drying is a time-consuming process requiring periodic attention and constant scavenger protection, processing may be most effectively carried out at or near residential base camps at the upper forest/lower subalpine ecotone.

Current Representation in the Park
None.

Type 9: Prehistoric to Early Historic Period Trails

Predicted Site Function
Foot and later equestrian trials can be expected to have linked resource zones to lowland communities, assuming that ingress and egress from Mount Rainier was a sufficiently regular occurrence to create and maintain established routes. Trails also provided passage across the mountain to points on the east and west. It is plausible that population densities were high enough to stimulate regular travel routes by mid to late Holocene times. Various references allude to the presence of routine travel into the Park from several sides (Meany 1916; Bjarke 1949; and especially Smith 1964).

Expected Assemblage Characteristics
Prehistoric trails are difficult to identify unambiguously in the field. Where present, trails should appear as faint linear terrain irregularities. In practice, we typically are forced to rely on ethnographic accounts, early maps (e.g., USGS 1915), and assumptions regarding least effort access routes between points–usually without concrete indications of an actual route.

Location
Relying on ethnographic and historical accounts, Smith (1964:229-238) outlines a series of possible trail routes centering on mountain passes. These include Chinook Pass linking Mount Rainier to the American River drainage and Yakama territory on the east; Naches Pass in the northeastern corner of the Park reportedly used by Yakama and Puget Sound groups in the early historic period; Carlton Pass south of Chinook Pass; Cowlitz or Packwood Pass southeast of the Park reported used by the Nisqually to access the Ohanapecosh River area from the southwest; Yakama and Snoqualmie Passes north of the Park; White and Tieton Passes south of the Cowlitz; and Cayuse Pass trending north and south trough the eastern margin of the Park. Smith also discusses direct mountain access trails with specific reference to 1) a side trail from the main Nisqually River-Cowlitz Pass trail up into Indian Henry's Hunting Ground in the Park's southwest quadrant; 2) a trail from Taidnapam by Lookout Mountain in the Tatoosh Range and Reflection Lake to the Paradise Glacier area; and 3) a trail via the Carbon River and Tolmie Creek into the Mowich Lake-Mist Park-Spray Park region in the northwest quadrant (Smith 1964:240).

Because it summarizes both potential trail locations, the probable ephemeral character of prehistoric and early historic routes, and general Mount Rainier land-use patterns, the following indirect citation is offered from Smith's (1964:241) ethnography:

Speaking of trails in the Rainier area in general, Plummer (1900:89), the author of an early forest survey in the Park and adjacent territory to the south, emphasizes... "The routes of travel ...are few. Most of the trails shown upon the map are hardly deserving of the name, but indicate blazed lines where better progress can be made than taking a course through the timber and brush. The Indian's policy was to go only where his pony could take him, ...; therefore his lines of travel were along the sparsely timbered ridges, where feed was generally plenty, where game abounded, or where huckleberries grew."

Allen agrees. He (Allen 1916:56) writes: "Every summer parties of hunters and berry pickers from the sage-brush plains crossed the Cascades with their horses. They followed the high divides and open summits of the secondary ridges until the came around to the open parks about Mount Rainier where they turned their horses out to graze and made their summer camp." ...

Please note that horse transport only became possible in the very late prehistoric and early historic periods. Use of horses probably affected the specific routes selected. They certainly extended the distance that could have been traveled to access Mount Rainier and increased the weight that could have been carried to and from the Park. With the onset of equestrian transportation, lower value, bulky resources such as huckleberries may have assumed greater importance than when resources had to be packed out of the mountain on foot. Even so, the passages above allude to cultural features worthy of note, point to the resource importance of the uplands, and plausibly draw our attention to the best access routes to these resource areas for the prehistoric past.

Current Representation in the Park
No trail locations have been formally documented in Mount Rainier National Park. In addition to references noted above, however, popular lore alludes to an early historic trail linking Sunrise Ridge to the Yakama area via the southeastern ridge slope in the vicinity of site FS 90-01 (and presumably on across Chinook Pass as discussed by Smith). Sections of this trail are reported to be visible (C. Fabiani pers. com. 1995; J. Morrison pers. com. 1995; R. Drake pers. com. 1996) but had not been recorded at the time of writing. The general area and outline description are included in Figure 3.5 and Table 3.6 to draw attention to the trail segment as a high probability early historic/late prehistoric cultural feature.

Type IF: Isolated Lost Artifacts

Predicted Site Function
Truly isolated artifacts are individual lost projectile points and tools, isolated transported exotic materials (manuports), or individual broken and discarded implements. They are assumed to be unaffiliated with a broader assemblage array, and hence unassociated with base camps or other sustained activity sites. Isolated tools function as part of a spatially extensive, generally hunting related, use of the landscape. As such, they can help inform us as to the distribution of those land-use practices, and (less reliably) the character (principally size) of the animals sought, and general temporal range of those practices. As with the broader lithic array, raw material sourcing and comparative stylistic attributes may provide information of use in inferring originating areas for human populations using the Park.

Expected Assemblage Characteristics
In this classification, artifacts considered most likely to be genuine isolates (as opposed to surface visible representatives of limited use activity areas) consist of single, whole or broken finished tools and manuports not associated with debitage or other lithic remains. Please note that these criteria are more limiting than those used during the present reconnaissance in which two or less surface visible artifacts, regardless of type, were recorded as isolated finds. Given the forested, depositional nature of Park landscapes, I suggest that, in the absence of subsurface discovery techniques (which are highly recommended), future inventories adopt the more stringent isolate criteria applied here.

Location
Because isolates are assumed to be part of extensive, hunting-related use of the landscape, they are expected to be biased toward most productive hunting areas. Accordingly, in the Cascades and at Mount Rainier, most isolated artifacts should be found in subalpine to alpine contexts.

Current Representation in the Park
Seven currently documented finds meet the criteria noted above. These include the Upper Palisades Trail (IF 01-84) and Upper Summerland (IF 02-95) isolates in the Park's northeast quadrant; Spray Park Shatter (IF 08-95) and Spray Park Slab (IF 09-95) in the NW quadrant; Tokaloo Trail (IF 01-87) and the twin Copper Mountain Cobbles (IF 10-95) in the SW quadrant; and the Bench Lake Trailhead isolate (IF 02-63) in the southeast quadrant.

Table 5.1 below summarizes site type distinctions, and brings together several of the more pertinent lithic and environmental attributes emphasized above and in the preceding lithic section. Only documented or well located sites and isolates are included. Site locations can be seen on fold out site distribution map Figure 4.2 in the previous chapter, or on color fold out Park quadrant maps in Chapter 2 (use Park Quad to locate the proper quadrant map). Except for Berkeley Rockshelter (FS 86-02) and Fryingpan Rockshelter (FS 63-01), artifact counts are limited to surface observation. Fryingpan Rockshelter material (see Rice 1965) is now housed with the Park's museum collections at Longmire, but were not available at the time present analyses were completed. Material counts for this site are drawn from Rice's report. I emphasize again that site type distinctions outlined above and tabulated below constitute a working model. Refinements and modifications are expected and encouraged.

Table 5.1 Mount Rainier Site Types, Sites and Surface Remains

Table Notes:
^aFryingpan rockshelter totals are extrapolated from results of an unscreened test unit excavated in 1964 and reported by Rice (1965).
^bBerkeley Rockshelter test results (indicated by brackets) are extrapolated from Bergland's (1988) work at the site.
^cCounts extracted from Bergland's 1990 site form.