North Cascades Ethnography of The North Cascades

The ethnographic information compiled in this chapter has been drawn primarily from June McCormick Collins' (1974a) field study of the Upper Skagit carried out in the 1940s. These data are especially full in the areas of religious and social life, less complete in describing the traditional subsistence base and material culture of the tribe, doubtless because these latter segments of the old ways were both less sharp in the minds of her informants and of a lesser interest to Collins herself.

These data are supplemented by information fragments culled from published ethnographic and ethnohistorical sources. This is particularly the case for those subject areas not dealt with by Collins or considered only cursorily by her.

In this chapter a number of tribes of northwestern Washington and the adjacent British Columbia coast are mentioned. The maps that follow (Figure 1-1 and 1-2) are reproduced here to assist the reader unfamiliar with the local ethnography in placing most of these groups.

Figure 1-1. Fragment of a tribal distribution map of Washington by Spier (1936:42-43), slightly modified. The two Skagit division are shown, but not differentiated by name (see arrows). This and Figure 1-2 locate most of the tribes referred to in this chapter. Approximate boundaries of the North Cascades National Park Complex are marked by a dot-dash line and the area is designated by hatching.

Figure 1-2. One of two complementary tribal distribution maps by Suttles (1957:158,159). This map covers the lower Fraser and Strait of Georgia areas, showing the location of the Lower and Upper Skagit (#19, 20) and their western and northern neighbors. Tribes designated by the missing numbers apear on the companion map.

In this section four matters of a general ../background nature are briefly considered: the "native" terms by which the Skagit were known, the geographical boundaries of the group's homeland, the cultural and linguistic relationships of the tribe, and such early population data as are available for the group.

Ethnonymy

It is customary to think of the Skagit as comprised of two major population units: (a) the Skagit proper who occupied the area about the mouth of the Skagit River and the offshore islands, and (b) the Upper Skagit whose territory began a short distance inland from the salt water and extended eastward up the Skagit drainage into the high mountains that formed the western flank of the Cascades.

The Upper Skagit are sometimes considered to have had two component segments: the Upper Skagit proper on the upper Skagit River whose subsistence country reached well into the Park Complex, and the Suiattle-Sauk who ranged the Suiattle and Sauk river region to the south of the territory of the Upper Skagit proper. This latter group, however, is at times considered as a third major division of the Skagit or even as a separate tribal entity comparable to the Skagit.

In this study the Upper Skagit and Suiattle Sauk are, for the sake of simplicity, regarded as a single population group. However, because the territory of the latter lay some distance from the Park boundaries, lesser attention is given to it where the data differentiate between it and the Upper Skagit proper.

Whether these three Skagit population components were thought to be individually cohesive units and distinct from one another by the Skagit Indians themselves and possessed their own distinctive names I do not know. Certainly the term "Skagit" has been used in several different ways by Whites as a population unit designator: for example, to refer to (a) the Skagit proper (or Lower Skagit), (b) the Upper Skagit alone, (c) the Lower and Upper Skagit together, and (d) the Lower Skagit and a few of the Upper Skagit villages (Indian Claims Commission 1974:315).

According to Hodge (1910:585) and to Swanton (1952:441) quoting Hodge, the Skagit name for them selves was Hum-a-I uh, "the people," information third-hand by the time it reached Hodge. It is unclear whether this designation applied to the Lower Skagit, Upper Skagit, and Suiattle-Sauk or only to some part of this larger unit. All other terms for the Skagit listed by Hodge are obvious variants of "Skagit": e.g., Sachet, Scad-jat, Skagats, etc.

Tribal Territory

There seems to be only a single study based on field research that attempts to define in detail the traditional tribal boundaries of the Upper Skagit. This is a map (Collins 1974c:facing 760) printed as an annex to Collins' (1974b) report of 1946 concerned with the nineteenth century shift of the Skagit from their aboriginal religion to Shakerism (Figure 1-3). It shows the territorial limits, settlement locations, and major subsistence use areas of the Upper Skagit as defined in this study, as well as these data for two neighboring tribes. The authorship and date of this cartographic document are not identified with the map, but in the 1960 Indian Claims Commission (1974:317) findings of fact concerning the Upper Skagit land claims litigation, the map is attributed to Collins. As printed, this map is unfortunately a small-scale, reduced, and therefore cluttered cartographic effort. Nevertheless, it must be used as the primary source from which to define the aboriginal homeland limits of the Upper Skagit.

Figure 1-3. Tribal boundaries, settlements, and subsistence areas of the two Upper Skagit divisions (Upper Skagit proper and Suiattle-Sauk) and of the Skykomish and Snoqualmie, comprising a second major grouping. This map, printed anonymously with Collins' (1974c) Upper Skagit religion research, is apparently extracted from her still unpublished study submitted in connection with Upper Skagit early land claims litigation.

The Upper Skagit, according to this map, were divided into the Upper Skagit proper and the Suiattle-Sauk.

The eastern limits of the country of the Upper Skagit proper, following the line between Whatcom-Skagit counties on the west and Okanogan-Chelan counties on the east, coincided with the sinuous crest of the

Cascades southward from the Canadian border to the vicinity of Lizard Mountain (USGS Concrete quadrangle 1955). In this distance the boundary was broken by a number of roughly east-west passes, among them Castle, Woody (just west of Barron), Harts, McBee, Methow, and Cascade, to mention only those marked on the USGS Concrete quadrangle (1955). Just below Cascade Pass the Upper Skagit border left the North Cascades National Park and moved on south to the Lizard Mountain region.

From the Lizard Mountain area the territory of the Suiattle-Sauk division continued down the Cascades watershed over Ross Pass, now following the Snohomish-Chelan county boundary, and Suiattle Pass, soon bending to the southwest to run just south of Glacier Peak to White Pass, where it veered sharply to the south again until it reached a point about due east of Monte Cristo. There it turned abruptly to the west immediately south of the 48th parallel. Now at last well south of the Park Complex, we need not trace its limits beyond this point.

The northern border of Upper Skagit country, as defined by this map, proceeded along the international boundary to the watershed between the upper Skagit River (now Ross Lake) and the Chilliwack drainage. Obviously to some uncertain degree this following of the Canadian border is painting with a broad brush, for the straight line political separation along the 49th parallel bears no appreciable relationship to the topographic realities that were major considerations in determining the limits of tribal use areas. Leaving the Canadian boundary it proceeded to the southwest along the water-directing crests to circle the headwaters of Baker River, to climb over Mount Shuksan, and then to move southwesterly above the western source streams of Baker River and to pass just east of Mount Baker. Here we are significantly west of the Park limits, making further plotting of the Upper Skagit northwestern boundary, as indicated in this map, irrelevant for this study.

Similarly, the dividing line between the Upper Skagit and Skagit tribes lay well down the Skagit River to the west and so well beyond our Park concerns.

The border between the traditional territory of the Upper Skagit proper and the Suiattle-Sauk ran roughly west and slightly northwest from the general Lizard Mountain environs to reach the Skagit River just east of the mouth of the Sauk River. This means, as already indicated, that the entire country of this small Suiattle-Sauk group lay some distance outside the southwestern perimeter of the North Cascades Complex and that it and the people themselves are of lesser interest to this report.

Especially notable is the fact that there is a considerable territorial overlap between the northeastern country claimed by the Upper Skagit as plotted here and the traditional country of the Lower Thompson as described by Teit (1900:166) (Figure 3-1). As remarked in greater detail below, this overlap region was to some degree subsistence resource terrain of both the Upper Skagit and the Lower Thompson. Yet it is surely not without significance that on Collins' map (Figure 1-3) no Upper Skagit game and fish resources are specifically marked for the region southeast, east, and west of Ross Lake, a portion of which Teit claims for the Lower Thompson. From these map data it is likewise clear that the Upper Skagit maintained no settlements on the Skagit River above Newhalem, some 11 miles downstream from Ross Dam (see Residential and Task Sites section below).

The Upper Skagit boundaries as described by Collins (1974a:5) very sketchily and in a slightly muddled fashion down in the Sauk country do not differ materially from the above delineation. She reports only that the eastern border followed the Cascade divide, that the northern line abutted the territory of the Thompson, who hunted in the upper Skagit River country in Canada, and followed the ridges between the Skagit and Nooksack basins, and that the southern boundary fell along the ridges separating the waters of the Skagit from those of the Stillaguamish.

The boundary of old Upper Skagit territory is also specified in the tribe's 1958 claims petition against the federal government. That portion relating to the back country in and near the Park borders reads as follows:

... [from] the crest of Mount Baker and along the divide separating the waters of Chilliwack River and Skagit River, to the International Boundary line ...; thence eastward on said ... Line to the crest of the Cascades Mountains; thence southward along the crest of said mountains to the most eastern point of the divide separating the waters of Sauk River from the headwaters of Skykomish River; thence westward, northward, and again, westward along said divide, said area including the watershed of the Sauk River and the Skagit River, to a point on said divide approximately due south of the present town of Ehrlich, Washington....

(Indian Claims Commission 1974:318-319)

This line, when plotted, follows exactly or as nearly as I can determine the boundary as defined by Collins' (1974c) map (Figure 1-3). This is not surprising since the map was a petitioner's exhibit in this litigation.

Cultural and Linguistic Affiliation

Culturally the Upper Skagit were at root a Northwest Coast group. Still, much of their territory, extending eastward to the very crest of the Cascades, was excessively mountainous. Accordingly, they also possessed traits that were in keeping with the biophysical properties of such interior terrain. Collins (1974a:4, 6-7) summarizes this Janus-like situation as follows.

Like coastal peoples, the Upper Skagit relied on salmon as a staple; used red cedar in making their large winter dwellings, canoes, containers, and other objects; resided in permanent winter villages and in summer camps with smaller movable shelters; carved and painted the inner posts of their houses to symbolize guardian spirits and placed some carved wooden figures in and near their villages; were structured into upper, middle, and slave classes; possessed little political organization; practiced the potlatch; divided the year into a winter season of religious ceremonies and a summer period when the food quest was the paramount concern; and possessed the Raven and Mink as important mythological figures.

Like the people of the nearby Plateau, the Upper Skagit made extensive use of deer, elk, and other land mammals as food and of baskets as containers, and they told myths with Coyote as a principal figure. As with typical Plateau groups, they possessed no secret societies and held few slaves. Especially the people of the villages farthest upriver made coiled watertight baskets, wore skin clothing including moccasins, used snow shoes, and relied strongly on individual guardian spirits though these were inherited within the family.

This special combination of cultural traits, some facing west and some representing high, backcountry adaptations and practices, was recognized as early as the mid-i 800s. In 1854 George Gibbs (1855a:432) wrote:

. . . besides the tribes, or bands, inhabiting the shores and the lower part of the rivers, there are on the headwaters of the latter, along the whole course of the Cascades mountains, another range of tribes, generally independent of the former, who rarely descend from their recesses, but are intermediate in their habits between the coast and mountain tribes; except the Taitinapam [who were Sahaptian speakers related linguistically to the Plateau tribes east of the Cascades], however, they all belong to the general [Salishan linguistic] family upon whose borders they live. Those in the neighborhood of the passes own a few horses, which subsist in the small prairies skirting the base of the mountains.

These cultural differences, comprehended at a general level by Gibbs and specifically for the Upper and Lower Skagit by Collins as already noted, formed the base upon which Smith (1941:205-206) formulated a cultural subarea -- the Foothill Culture -- distinct from the subareas on the salt-water shores. This proposed subarea consisted of a tier of small tribes against and on the western face of the Cascade Mountains, of the Upper Skagit and their inland neighbors to the north and south. It also had its counterpart among the Plateau Indians on the eastern flank of the Cascades. This formulation is discussed more fully in the later Chilliwack chapter of this North Cascades study. It is sufficient to note here that the Upper Skagit shared with the other foothill people a negligible use of clams, seals, and other salt water products, a hunting pattern almost equal in importance to fishing, cross-country travel by land as well as movement by canoe, social gatherings without the potlatch exaggeration, a prestige pattern based on behavior and achievement rather than on inherited class, and individualized guardian spirit power (Smith 1941:206).

The important point is that the Upper Skagit had, in a way, two lines of cultural affiliation. One was down the Skagit River to the Lower Skagit and other nearby salt-water groups, with whom they shared one block of cultural traits. The second, running north and south, was with other tribes possessing this backcountry "foothill" culture. This foothill proposition has been criticized by Suttles (1957) as poorly documented, as it surely was, and even illusory, as it definitely was as a tightly structured cultural subarea. But certainly these high-area tribes on the western slopes of the mountains shared some special environmental adaptations that brought them into a common cultural fold, and, as a group, lacked some particular maritime adaptations of their western neighbors.

According to Collins (1974a:8), both traditions and linguistic evidence in the form of place names suggest that Interior Salishan speakers once occupied the Skagit Valley. Other traditions tell of recent migrations into the Skagit Valley from the Washington Coast and Vancouver Island as well as from the Plateau. However, linguistic evidence and especially traditions of these kinds should not necessarily be accepted as proof of an historical fact.

Upper Skagit speech was a dialect of Looshootseed (Puget Sound Salish), a language that included the dialects of the Snohomish, Snoqualmie, and other neighboring groups. Looshootseed, in turn, belonged to the large Coast Division of the Salishan language family, one with member languages from southern Vancouver Island south to the Oregon coast. Upper Skagit was mutually unintelligible with Nooksack and with Chilliwack, and substantially more different from the Interior Salishan speech of the Lower Thompson and Chelan (Suttles 1957:163, 1985; Thompson 1979:693). Bilingualism, however, was common in the area (Indian Claims Com mission 1974:323).

Slight dialectic differences occurred from one Skagit extended village to another (Indian Claims Commission 1974:323).

Demography

Population estimates for the Upper Skagit made in precensus times must be considered only rough approximations. As George Gibbs (1877:181-182) so perceptively points out for the area tribes in general:

I am well satisfied that none . . . [of the estimates made to my time] can be taken as . . . accurate calculations. . . . In arriving at any [demographic] conclusion, it is necessary to regard not merely the actual facts of increase or mortality known to us, but the capacity of the country to furnish subsistence, the modes of obtaining it followed by the Indians, their general character and habits, their fecundity, their wars, and various circumstances directly or indirectly bearing upon life.

With these cautions in mind, Gibbs (1855a:433) in 1854 estimated the Skagit (evidently wholly or principally the Lower Skagit) at 'about three hundred all told; and there are other bands upon the headwaters of their river plainly wholly or in large part the Upper Skagit, amounting probably to as many more" (cf. also Gibbs 1855a:436). Wilkes' (1841; in Stevens 1855:435) earlier 1841 estimate for the Scatchae (Skagit) tribe was 650. Whether this includes both the Lower and Upper Skagit or only the former is in question, conceivably exclusively the former with the difference between 650 persons and Gibbs' 300 (assuming that both figures were reasonably accurate) being the consequence of population decimation in the 13 year period. At any rate, the figure of about 300 persons has generally been taken as close to the population of the Upper Skagit in the mid-nineteenth century.

Other population estimates for about 1780 and later have been made in which the Skagit -- Upper and Lower divisions together -- are submerged in still larger tribal groupings, e.g., by Mooney (Swanton 1952:442) in the early 1900s and by Taylor (1963:161-162, 163). There is, however, no way to extricate from these approximations the figures for the Skagit specifically, let alone the Upper Skagit.

The effects of Western epidemic and other diseases -- smallpox, respiratory ailments, etc. -- upon the Upper Skagit in the late 1700s and early 1800s must have been great. The tribe, incidentally, termed smallpox Nooksack sickness," because it spread to them from that group (Smith 1950a:332). In 1961 Taylor (1963:162) calculated "that the population of the Puget Sound area declined about 50 percent in the period from 1780 to 1840." In his careful study of the influence of these diseases in the late 1700-early 1800 time span, Carroll Riley arrived at a population reduction estimate for the Puget Sound tribes of between 70 and 90 percent. Based on Gibbs' Upper Skagit approximation of 1854 and Riley's depopulation figure, the Indian Claims Commission (1974:326-327) concluded that the "Upper Skagits may have numbered 880 members at the beginning of the nineteenth century."

Consistent with the focus of this study on the aboriginal and early postcontact Upper Skagit, population estimates and census figures for the period after 1860 are not considered here.

The Upper Skagit food quest involved an extensive exploitation of the fish, game, and plant resources of their territory (cf. Figure 1-3). The data, however, are unfortunately very thin. Especially regrettable is the fact that virtually nothing is said except obliquely about the subsistence use of the higher terrain and nothing at all specifically concerning the streams and mountains within the Park boundaries. I have yet to ascertain how far up the Skagit River anadromous fish were able to negotiate in aboriginal and early postcontact times. We may be certain that they reached at least the Newhalem area since the uppermost Upper Skagit village was sited there. But whether they ascended farther into the present Ross Lake region I do not know. Chapman and Turner (1956) designate the Skagit as a salmon river only up to about the Newhalem sector, but I am unsure whether this is a pre- or post-dam plotting.

On the evidence of the purportedly factual tradition referred to elsewhere in this study, salmon ascended to the headwaters of the Cascade River (Collins 1974a:14).

The subsistence-quest strategy of the Upper Skagit was patterned in a very special way -- quite differently, for example, from that of the Chilliwack -- that had its effect upon the group social structure. As Smith (1941:201) explains the situation:

[The] houses of the local units [of the "extended village"] were erected on suitable sites near important salmon fisheries; the buildings were at the same time homes and smoke houses; and if the population cared to live only on its food staple it needed never leave its head quarters. When the units dispersed, families went in opposite directions and at.. . berry and root fields often joined families from units which belonged to other extended villages. Socially, these people who met during the summer were as closely knit as were those of the contiguous winter quarters of the extended village.

These data seem to underscore the political insignificance of the extended village and also of its component individual villages.

Apparently food was not invariably available in necessary amounts to the Upper Skagit, as it is said to have always been to the Lower division. According to John Fornsby (in Collins 1949:302), an Upper Skagit born in 1855, the Upper Skagit "used to have hard times." He recounted one supposedly specific instance in his grandmother's day when the downriver region was visited by a number of upriver Skagit who were so weak from hunger that they could hardly paddle their canoe. They had come down for clams, because "they had nothing to eat" back home. Fornsby then added:

[S]ome of those [Upper Skagit] people never did much hunting, and they couldn't get any fish. There weren't many fish up-river. They couldn't catch them. . . . They got berries in the mountains. They generally got meat: they made a trap and caught beaver, bear, deer meat. . . . Just the up-river Indians were poor. They had a hard time to get blankets up there before the White people came. It was pretty hard to catch goats. You had to have a good gun. (Fornsby in Collins 1949:302-303)

Though Upper Skagit born, Fornsby had spent his life among the Lower Skagit. It is hard to know the extent to which he exaggerates the difficulties and country-cousin status of the upstream people. But there is little doubt that their food supply was less secure than that of the downriver Lower Skagit.

Fishing

Of the three principal food sources, fish contributed most to the Upper Skagit diet. This is evidenced, Collins (1974a:45) reports, by the location of most villages "where . . . the maximum amount of fish could be secured during a run." Certain difficulties in understanding this statement are discussed at a later point (see Residential and Task Sites section).

Three fishing areas are plotted by Collins on her subsistence-marked map (Figure 1-3): one on the lower length of Baker River, one on the upper Skagit River in the Marblemount sector, and the third -- for "sockeye and other salmon" -- in the mountains between these two which I fail to comprehend.

The only fish mentioned ethnographically as having been important to the Upper Skagit are anadromous types: specifically, five kinds of salmon (chinook, pink, sockeye, chum, and coho) and steelhead trout, each with its distinct native name. Of these, the chinook and coho were preferred. Salmon ran from May or June to into the fall; steelhead trout ran as late as November. Trout were also caught, but seem to have been comparatively insignificant in the native diet.

Salmon were netted, taken in weirs and with hook and line, and were speared, all techniques used when the fish were swimming upriver. Three kinds of nets are described.

(a) A seine, with ends attached to poles and long enough to reach nearly across a stream, was stretched vertically by men from two poled or paddled canoes in the light of torches. This was regarded as the most productive fishing method.
(b) A net with a closed end was set for some hours in shallow water near a river bank.
(c) A woven willow-bark or grass (Indian hemp?) dip net was "used to scoop fish from a canoe, a platform, stones in the river, or from the bank." The bag net was suspended from a vine maple hoop attached at right angles to two long pole handles that were tied together at their upper ends. Between these poles immediately above the bag net was stretched a net baffle to prevent fish from swimming between the handles.

Apparently several weir types were in use. One variety, employed in July and August, was constructed by as many as 20 men, under the supervision of an expert, across a tributary stream where it was narrow and shallow and the riverbed rocks were small. The frame consisted of a row of single fir poles firmly set vertically into the stream bottom and tied together by two or three rows of horizontal cedar poles. The top one, above the water level, served also as a walkway to the weir platform. Between these vertical poles, tripods were placed at intervals, two legs of which were placed so as to assist the vertical poles in supporting the fence. This fence was fashioned on land, and was rolled out against and tied to the vertical poles and tripods. Above an opening in this fence a platform was constructed; on this the fishermen stood with their dip nets. All men who, as volunteers, had helped build the weir were allowed to fish from it. According to Collins (1974a:50), the fishing continued all night, but whether it was exclusively a night activity is not specified. At the close of the season the fence was removed to be used the following year when the weir, carried away in the late spring floods, was rebuilt.

At riffles a watersoaked log was placed diagonally across part of a stream. Salmon, leaping this barrier, fell on an upstream stranding platform from which they could be easily taken.

Other fishing methods involved spearing salmon with a leister, consisting of an ironwood pole and three prongs of ironwood, deer antler, or mountain goat horn, and angling for them with a line twisted of women's hair and baited with salmon eggs.

Although fish were eaten fresh, most were smoke-dried for winter. To be prepared for eating, salmon (a) were covered with wet moss and steam-baked whole in an ash pit; (b) were cut into pieces and stone-boiled in a box or basket; (c) were split, spread with small splints, skewered onto sticks, and roasted in a slanting row around a fire; and (d) prepared in small slices, were grilled over a fire. Salmon eggs were wrapped in alder bark and either roasted in hot ashes or hung over a small fire to dry for later consumption.

Shellfish -- clams, oysters, mussels, barnacles, and crabs -- were secured on the coast when the Upper Skagit traveled there to visit relatives or by exchanging dried venison for them smoked.

There is no ethnographic information reporting Upper Skagit fishing sites within the Park Complex bound aries. Mentioned incidentally by Collins, however, are dip net sites on the Cascade River and on the Skagit near the town of Concrete, a chinook salmon fence-weir locality on the Sauk River, and a pink salmon spearing site on the Skagit above Marblemount.

Special guardian spirit power played an important role in fishing activities. Power was required, for example, in making effective dip nets, and in supervising and carrying out specialized tasks in weir construction. (Collins 1974a:45-52, 76)

Hunting

Collins (1974a:52) states:

Second in importance in the diet were land animals. The Upper Skagit hunted deer, elk, mountain goat, black bear . . , grizzly bear, beaver, snowshoe rabbit, fisher, raccoon, and land otter. Deer and elk were the most highly sought . . . . Among the Indians of their area, the Upper Skagit maintain that they tended to specialize in hunting.

For what it is worth, on Collins' subsistence map (Figure 1-3) eight hunting areas are shown. The four deer areas are on the lower Skagit River in the Burlington sector, in the stream valley above Big Lake, on the upper Cascade River, and on the middle Suiattle River. The two elk sectors are the mountains southeast of Newhalem and the middle Suiattle. The four mountain goat regions are all in the high mountains: one northwest and one northeast of Baker Lake, one southeast of Newhalem, and one on the upper Cascade River. The areas where bear were killed are shown to have been in the country above Big Lake, in the upper Cascade sector, and once again on the middle Suiattle vicinity. Ducks were taken in the Big Lake district. It will be noted (Figure 1-3) that in four cases more than one animal is shown to have been killed; hence the total number of only eight separate hunting localities.

Game was taken with bow and arrow, though the detailed techniques followed are not reported. Bows were of yew or alder, had recurved ends, and were sometimes sinew backed; they varied in length up to about five feet depending on the size of the game for which they were to be used. Bow strings were of sinew. Arrows had an ironwood shaft, fitted with a two-shouldered point of white stone "obtained from a rock slide at . . . [a] place on the main Skagit" (Collins 1974a:52-53).

Deer and bear were caught in log deadfalls, set in their trails to water. Animals were also taken in baited spring traps. Deer were driven over cliffs or into ravines, both by single hunters and by hunting parties.

Meat was cooked by being cut into pieces and then stone-boiled in coiled baskets made watertight by an overnight soaking; sometimes roots (e.g., wild onions) were boiled with the meat. Meat was also roasted and broiled. Deer and other meat was dried for winter use, an important cold season food.

Many birds -- especially waterfowl including geese and ducks -- were hunted. Ducks were shot with bow and arrow and caught at dusk in nets stretched between two vertical posts. Duck grease, particularly valued, was caught in a clamshell as it dripped from a bird cooking on a spit over a fire.

Upriver Upper Skagit rarely if ever hunted sea mammals. But downriver villagers, when visiting relatives near the Skagit mouth, hunted seal and porpoise (but not whales) with harpoons with their shore-dwelling kinsmen.

Mammals and birds were hunted not only as food but also for some of their parts -- like deer skins and bird bones -- that served other purposes. These will be mentioned in later sections. (Collins 1974a:52-55)

Plant Collecting

Many plant varieties were gathered for food as well as for other purposes. As dietary constituents these were important, providing the starch and sugar otherwise lacking in their high protein foods. Still, vegetable eatables were not available in sufficient quantities for them to be consumed even once daily through the year.

Two root plants -- the tiger lily and sweet wild carrot -- were to judge from Collins' (1974a:55) data, the most significant food plants. They were semicultivated on the prairie at bakálb village on the Sauk River as well as on two meadows in the lower Skagit River area; on Collins' subsistence map (Figure 1-3) tiger lily root grounds are shown on the middle Suiattle River but none on the Sauk. The land was divided into marked plots -- each three or four acres in size at bakálb -- that were owned by women and passed down to their daughters irrespective of where the latter might later live. During "the late summer, women with . . . [harvest] rights came from widely distant villages to the plot of their mother" (Collins 1974a:55). With a three-foot vine maple digging stick sharpened and fire-hardened at both ends, women dug the roots in August when the blossoms died, replanted the stems, and weeded the plots.

Camas also grew in the tiger lily and carrot plots and were collected, but whether these plants were also semicultivated is uncertain.

Roots of several kinds of ferns were also eaten. These plants grew near Birdsview in the Skagit Valley, as well as in lower Skagit River country. Collins' food resource map (Figure 1-3) places a fern root area on the middle Suiattle River. For these plants, however, no individual plots were recognized: any woman with kin at the Birdsview village, for example, was allowed to dig these roots during her summer visit.

Wild onions were dug for food, for they are mentioned by Collins (1974a:53) in relation to their being cooked with meat. They are, however, not otherwise referred to. The wapato root (Sagittaria spp., arrowroot [Hitchcock and Cronquist 1981:559]) grew in only one Upper Skagit locality, this near the mouth of the Skagit. Evidently it made essentially no contribution to the Upper Skagit diet. Roots "could be found in certain mountain meadows; [but] women preferred not to dig them there if they had a choice" (Collins 1974a:56). This is a revealing observation in terms of the special interests of this report, but would be far more meaningful if we were told the basis for this statement.

Roots were baked in a deep pit or in ashes of the hearth or were stone boiled with meat in a water-filled basket. "Roots were stored for winter use in a special type of basket which women buried in the ground in September" (Collins 1974a:56).

Tree lichen was gathered and prepared by being placed under ashes all night.

Berries used as food included the blackberry, wild strawberry, salmonberry, blackcaps, thimbleberry, serviceberry, huckleberry, blue elderberry, red elderberry, gooseberry, trailing currant, mountain blueberry, and salal, the list given by Collins (1974a:56-57). Wild crabapples and plums added fruit to the diet. Forest areas were carefully burned to make berries grow more prolifically. Guardian spirit assistance was also sought to encourage berry growth and to control forest burns to improve berry yields. (Collins 1974a:55-57) The only berry ground plotted on Collins' food source map (Figure 1-3) lies on the north side of the Skagit River in the Birdsview district.

Concerning the preparation of berries and fruits for eating we are told very little. All berries were eaten fresh. Elderberries and blackberries were also boiled until they were nearly dry. They were kept by being wrapped in maple leaves and placed in a hole dug in wet sand. Sun drying was regarded as the best method of berry preservation.

The green sprouts of some berries were enjoyed as spring greens.

The cultivation of the white potato among the Puget Sound Indians began early in the 1800s, the seed potatoes having been secured by salt-water groups from trading vessels. As Suttles (1951) has demonstrated, they spread very rapidly among these coastal tribes. Interestingly, the earliest mention of potato cultivation on the part of these groups appears to be a report of Father Blanchet, the first priest among the Skagit, of having stumbled "onto a Skagit potato patch" in 1840. The following year Wilkes found the Skagit growing these tubers. In the 1850s potatoes were again reported for the Skagit. (Gibbs 1855a:433; Suttles 1951:274, 276, 283) The speed with which this root, together with its simple cultivation techniques, diffused has suggested to some -- like Smith (in Rivera 1949) -- that the wild root gathering experience of these native groups, the importance of roots in their diet, and their traditional pattern of planting, tending, and harvesting a few native roots, on a small scale, made them especially sensitive to the food potential of the cultivated potato. Moreover, with their traditional concept of family ownership of certain especially valuable wild root grounds, the cultural base existed for ownership of potato patches (Suttles 1951:281).

Food Preparation

Some fragments of information relating to food preparation are reported in the preceding sections. As a supplement to them, the following general data are of interest: fish (salmon and steelhead), game (deer, elk, bear, and other animals), roots (wild carrot, tiger lily, and other forms of the root class), and many different kinds of berries were all dried for winter use (Collins 1977:2).

The manufacture of canoes and building of the large cedar dwelling are described elsewhere in this chapter. Hunting weapons and devices as well as fishing and plant collecting implements are reported in those sections. In this unit we are concerned with lesser technological processes and the associated items of material culture.

Rigid Containers

Although much less common than baskets used for similar purposes, wooden boxes were fashioned by steaming and bending a flat cedar piece to form four sides; tying the corner joint together with thongs through bored holes; fitting a separate bottom piece in place with corner pegs; and soaking the box until it swelled and the thongs tightened. Water buckets were made by hollowing out a section of a cedar trunk. Rectangular wooden dishes with sloping ends, made of hard wood, were used in serving meals. A canoe answered the purpose of a giant cooking vessel in preparing for feasts. Wooden or cedar bark items were made for religious ceremonial purposes.

Spoons were of hard wood or horn. The horn was softened by soaking in heated water and molded into shape. The handle was sometimes carved into human or animal form (Collins 1974a: 66 -- 67, Figure 4). Horn dishes with "carvings and carved lines traced on them, some of which represent faces," were probably made by the Skagit, "who roamed in the Cascade mountains," Eells (1985:231) notes, writing in the late 1800s.

Basketry, Matting, and Weaving

Baskets were very important to the Upper Skagit. Though they were coiled, twilled, twined, and woven in a checkerboard design, coiled cooking, burden, and storage basketry was a speciality of Upper Skagit women. Of particular importance, it appears, and distinctive to the tribe in some ways were the coiled, watertight baskets of cedar or spruce roots that were used in stone-boiling and storage. These roots were not dug but were commonly collected from eroded river banks, sometimes by men who chanced to discern them when in their canoes. The inner root part served as the horizontal element and the outer part as the stitching element. The latter was passed through openings made in the coil below with deer or elk antler awls. The bottoms of these baskets were always round; cooking baskets were round in cross section.

Basket ornamentation was by imbrication: i.e., by doubling a separate colored element under each sewing stitch on the basket exterior (cf. Underhill 1944:104) (Figure 2-3). Used for this purpose was a cream-colored grass not native to Upper Skagit territory but purchased from the Upper Snoqualmie or people living east of the Cascades: it grew abundantly around Snoqualmie Pass and near a lake east of the divide. Also used was the black top of horsetail root and brown bark of the wild sour cherry. Favorite designs were barbecued salmon back fins, fish gills, fish nets, frogs, butterflies, and a little fly. In this listing by Collins the prominence of water-related motifs seems revealing.

An open-work, stiff, twined basket was used in clam digging on visits to the coast, the water draining through the interstices.

Carrying baskets were provided with holes below the top border to receive small buckskin loops to which were attached the braided wool or buckskin carrying band. This band could be arranged to allow the container to hang from the forehead as a back basket or from the side of the waist as in berrying. Soft cedar bark bags, twilled or made in checkerwork pattern, were made and used (e.g., as a baby bag inside the cradleboard) though they were less common in the Skagit Valley than on the coast.

The Sauk River people were reported not to have manufactured baskets until comparatively recently, but to have used deer and mountain goat stomachs, turned inside out, as water containers. (Collins 1974a:67-69) This remark smacks of a folk tradition of doubtful authenticity.

Cedar bark and cattail mats were fabricated by some women but the skill was regarded as a specialty of saltwater people. These mats served as camp shelter walls, as inner wall hangings in permanent dwellings, as occasional partitions in dwellings, and as bedding, canoe kneeling pads, rain capes, and caps.

For cattail mats the reeds were pulled in September, dried, trimmed of their ends, and sewed together side by side. This sewing was accomplished with a bone or wooden needle drawing a nettle string through the reeds individually. While the needle was still in place, the reeds were pressed down crosswise above the needle with a special wooden creaser (cf. Underhill 1944:107-108). (Collins 1 974a:69-70)

Even though weaving was considered a "principal craft" of the Upper Skagit, not all women practiced the skill. The basic material was mountain goat wool obtained from skins brought home by hunters. Since goats occurred only in the upriver country, weaving was essentially limited to women of the interior villages. The hair of a special variety of dog, bred for their hair and sometimes kept on river islands to maintain breeding isolation, or goose down or even bear fur in the case of blankets was sometimes mixed with the goat wool in fabricating the yarn. The wool was carded with a wooden comb, spun with a spindle consisting of a cedar disk and stick, and twisted into yarn on the woman's thigh as it came from the spindle. If down was used, it was placed loose in a basket and the yarn was drawn "through two holes in the basket so that it passed through the down. When the yarn emerged from the basket it was covered with down which then became part of the completed fabric."

The loom for making large pieces consisted of two wooden uprights and two horizontal bars. The warps ran around these bars and the wefts were woven in place horizontally to create a tubular fabric that was cut to make the flat cloak or blanket (cf. Underhill 1944:110-113). These blankets, which were not plentiful, were valuable as gifts.

A small belt loom -- one end suspended from a tree and the other attached to the weaver's waist -- was used in weaving narrow belts and shoulder pads from which baskets were hung. (Collins 1974a:70-71)

Clothing and Body Ornamentation

In dressing skins, a woman's task, the hide was soaked for a week in a solution of cooked and mashed deer brains, twisted until pliable, and hung to dry. It was then stretched beside a fire by being pulled this way and that. Finally it was hung over a stick frame and smoked to a golden brown over a small fire of bark from rotten logs.

In the clothing fashioning process, likewise a woman's activity, holes were punched with an awl of a deer ulna split and sharpened on a stone. The sewing was accomplished with an eyed bone needle and thread of deer sinew that had been dried, pounded, and stripped.

Two types of clothing were worn, one a coastal style of cedar bark, furs, or wool and the other a tailored skin style like that of the Plateau. The temporal relationships between these two modes and the extent and specific form of the early post-White influence on them are not wholly clear.

When men wore clothing -- they frequently went naked -- their garments consisted as appropriate of a sleeveless "vest," a cloak of bearskin or woven goat or dog wool fastened over one shoulder with a bone or wooden pin, and a cattail matting rain cape. Women wore a thick, ankle-length skirt of cedar bark strips, hanging from a woven cedar bark girdle. In summer women preferred a "kind of black apron, presumably woven of bark." Like the men, they protected themselves from rain with the matting cape. Woven cedar bark hats were sometimes worn, but whether by one or both sexes is not reported by Collins. Basketry hats were not made but were occasionally received from kinsmen in the Plateau.

Both sexes also wore skin clothing -- buckskin in winter and doeskin in summer -- when the weather was cold in the mountains or on trips into the forest. Both sexes wore breechclouts and shirts, the latter (sometime of bearskin) open down the front and sewed up the sides with the sleeves left unsewed, sometimes trimmed with mink and marten. In winter the hair side was against the body. Both sexes wore mink hats; men also affected peaked caps with the animal tail hanging behind. In upriver villages both men and women wore gloves. Moccasins, one-piece and extending up nearly to the knee, were worn only in the woods or mountains. Clothing of some sort was likewise made of fox fur, for men were reported to have captured these animals and kept them for their peltries. (Collins 1974a:71-73)

Upper-class persons had their heads flattened so they "rose from the forehead to a peak at the back." This deformation was achieved by placing a cedar bark pad on the infant's forehead and binding it to the cradle- board. (Collins 1 974a:73-74)

Both men and women wore their hair parted in the center, men with theirs tied in the back with a beaver skin thong and women with theirs braided on each side and bound with a thong. Combs were of wood or fish backbone. (Collins 1974a:73)

Body ornamentation was fairly extensive. Red earth mixed with grease and "cooked" served as a facial paint, both for decoration and as protection from the cold. White paint was applied by women to their faces and to their infants. During guardian spirit dances red paint was put by women in their hair. Black paint was used only by men on raids. While labrets, ornaments worn in the lower lip, were not the custom, a bone was worn through the nasal septum. Women had small tattooed designs pricked into their faces, arms just above the wrist, legs, and ankles. The patterns were made with a bone needle and black scrapings from coal. To their person women applied perfumes of fern and other ingredients secured from groups east of the Cascades. (Collins 1974a:73-74)

As used in this study, the "settlement" consists of four principal elements: (a) the physical units -- dwellings and ancillary structures -- that comprise the settlement; (b) the grouping of these into villages or communities (winter or summer, fishing or otherwise, and so on), the location of these communities, and the clustering of them into larger "extended villages"; (c) the placement and relationships of the component physical units within the boundaries of the single settlement; and (d) the domestic and tamed animals held within or near the village as living but nonhuman components of the settlement. These matters are reviewed in the pages that follow.

Dwellings and Associated Structures

Of the structural entities of the Upper Skagit only the permanent winter/summer dwelling is reasonably well described. The winter ceremonial house, the small temporary summer shelters erected at certain fisheries both for resident families and for visiting kinfolk, the warm season temporary lean-to, the sweat house, and the girl's puberty shelter are hardly more than referred to. Such common settlement components as food drying racks, food and article caches if these existed apart from the drying and storage areas within the dwelling, and other lesser outside structures of the living community receive no ethnographic attention whatever.

Upper Skagit structures were of three types in terms of the materials from which they were fashioned: wooden houses, mat shelters, and brush sweat houses. Subterranean and semisubterranean houses were not made, though known among other tribes both on the Coast and in the Plateau.

The structures of wood included "permanent" winter and summer dwellings, small temporary summer shelters for relatives visiting to fish, and structures used exclusively for winter dance ceremonials.

Dwellings of wood -- at least the largest of them -- were built, under the supervision of a hired specialist, by the men whose related families were to reside in them and own them, sometimes assisted by hired workers. Huge cedar trees near a river were felled, trimmed, and floated down to the site. Selected trees were squared for the six posts and their tops grooved to support the roof beams. The two tallest were slipped into holes in the ground about 18 feet apart and fitted with the ridgepole. Two shorter posts were placed about 10 feet in front of the main pair and the two rear posts were slid into their holes about 40 feet behind the main two. These six posts created a rectangular ground plan and a double-pitched roof, a short section in front and a long one sloping from the ridgepole toward the back posts. Although not mentioned by Collins, a cross beam must have joined the two front posts and another the two back posts. Stringers connecting the three post pairs are not mentioned and I gather were not used.

The walls were of wide cedar planks lashed either horizontally or vertically, overlapping, both to one another and to the house frame. Cracks were stuffed with moss. Vertical side planks were kept in place by heavy stones set against the base of each plank. The roof was of similar overlapping planks, laid on the ridgepole and cross beams. Near the ridgepole was a smoke opening that could be closed by a board moveable with a pole from inside the house. A small, round entrance was cut in a plank at one end of the structure, generally facing a river or lake. It was raised slightly above the ground and sometimes closed by a small plank.

The interior of the house was normally without partitions. Fires were on the hard-packed earthen floor and were differently arranged depending on the number of occupant families: e.g., down a center line or one toward each corner if there were four families. A sleeping platform, covered with skin robes, mountain goat wool blankets, and matting, was built against the wall around the open center. Above this platform was a continuous storage shelf and still higher up pegs for hanging clothing. On and hanging from the beams were dried meat and other articles.

House posts were carved and painted with the sacred guardian spirit designs of the dwelling owners. These posts were sometimes inherited as a special honor, the transfer formalized by a potlatch. Occasionally the house front around the entrance was painted to represent an owner's guardian spirit. Outside, self-standing totem poles were not a feature of Upper Skagit culture.

Larger dwellings of this type, as much as 120 feet in length, were provided with more than six supporting posts. Some were built in 40 foot sections with three ridgepoles.

Houses of the above kind were "permanent," the wall planks being moved -- by canoe -- only to a new "permanent" location. (Collins 1974a:59-63, 83, Figure 3)

Similar but much smaller shelters of this variety but no more than 15 or 20 feet long were erected in two or three days at the best fisheries along the Skagit and its tributaries. Relatives "who came to visit and fish at the height of fish runs were partly accommodated in such houses" (Collins 1974a:64). I gather that this is the small, detached plank house that was, according to Smith (1947:258; cf. Duff 1952:49), the only dwelling "type used on the Sauk River and the upper reaches of the Skagit River. . . . [O]ne of these occasionally held four families, many of them were built for two families, and others were so small as to house but one family."

A second kind of post, beam, and plank house is described by Collins (1 974a:63) as "built of separate wooden sections joined together." It is not otherwise detailed and in any event appears to have been rarely in use among upriver villagers. Presumably this was the dwelling, described for the middle Fraser, consisting essentially of small, independent houses erected side by side and against each other.

Lean-tos covered with cattail mats were used as temporary shelters in summer. When traveling to places where they planned to stay for a time, as when going to the salt water to clam, people took folded mats with them. (Collins 1974a:64)

Sweat houses were far less important and common than in the Plateau. Some Upper Skagit, however, made and used for this purpose small, circular structures consisting of a pole framework covered with brush. (Collins 1974a:64)

Special, small, well-built shelters were constructed for a girl to occupy during a six-week period at the time of her first menses and the three to follow. Typically they were constructed near water where she could dive morning and night. (Collins 1974a:225-226)

Residential and Task Sites

(omitted from the on-line edition)

Village and Camp Arrangement

As already noted, cold season communities consisted most notably of one or more large plank dwellings and a ceremonial plank structure. Warm season settlements were principally made up of these same dwellings if they were located at good salmon fisheries, or of smaller wooden houses if the site was not occupied also during the winter. Often in summer villages were small temporary dwellings for visiting relatives. Associated with settlements of both seasons were sweat lodges and sometimes women's puberty shelters. Presumably in all communities whatever the time of year were food drying racks, food and artifact processing areas, perhaps caches for food and articles, and canoe beaching areas. But how these structures and task centers were spatially positioned winter and summer into integrated, functional living complexes we are not told.

Even less is known of the typical temporary hunting, plant collecting, and other task-focused camps of the Upper Skagit. Even the shelters and other structures that stood in these camps are undescribed. How these and their related work areas were spatially organized into cohesive social and economic configurations remains likewise undisclosed.

Domestic and Tamed Animals

In the traditional period, the Upper Skagit possessed only two domestic animals: the horse and the dog.

A few horses were held by groups with sufficient pasturage (Gibbs 1855a:432, Smith 1941:199). They were of particular value as pack animals on cross-Cascades trading trips to the Plateau.

Dogs of two varieties were kept in far greater numbers. Hunting dogs wre of great importance in the constant search and killing of game. The wool-bearing dog was in its way as important, for it yielded, though being periodically shorn, a wool-like hair that was widely employed in weaving blankets and clothing. These dogs, it is reported, were sometimes kept on river islands so they could not interbreed with hunting and camp dogs.

Foxes were captured by Upper Skagit men and kept until killed to convert their pelts into clothing (Collines 1974a:73). Whether other wild animals were likewise held captive in this fashion or were tended, at least for a time, in tamed status is not reported.

Residential and Task Sites

The Upper Skagit occupied a number of permanent villages, most on the Skagit River near the confluence of tributaries. A few, however, were located on the larger lateral streams (e.g., the Sauk River) and a very few were on lakes draining into the Skagit. Villages changed their location only as a river shifted its course. (Collins 1974a:3, 15-16)

Villages were grouped by the Upper Skagit into named "village clusters" (Collins' "extended villages") occupying a stretch of the Skagit River several miles long and in some cases also areas along tributary streams. The section of the drainage system made use of by the people of such a village was ordinarily upriver from the village site (Smith 1941:206). According to Collins, ten of these clusters were recognized by the Upper Skagit. Typically a cluster consisted of one or more winter villages of varying sizes and, in the case of four of the ten clusters, also of one or more summer villages. These summer settlements -- more strictly, I presume, warmer weather villages -- evidently served as bases for exploiting seasonal subsistence resources in their vicinity. (Collins 1974a:3, 15-20)

Three maps are known to me which plot the settlements of the Upper Skagit, each compiled by Collins: one (1974a:17 map 2) in her Upper Skagit ethnography (Figure 1-4); a second (1974b:622) accompanying her Shakerism paper written in 1946 (Figure 1-7); and the third (1974c:facing 760) evidently prepared as a petitioner's exhibit in the Upper Skagit land claims case of the 1950 period (Figure 1-3). Of these three maps, that reproduced here as Figure 1-3 can safely be assumed to be the most accurate. It evidently incorporates research conducted subsequent to her Shaker study and is the most informative as a cartographic product. On the other hand, the geographically more casual (and even inaccurate) map of Figure 1-4 is accompanied by settlement descriptions, rough demographic data, and seasonal use details. These materials form the basis of the village discussion that follows. All three maps, however, are on so small a scale, at least in their published form, that it would be impossible to locate the individual occupation sites except in the most general manner were it not for Collins' (1974a) textual data.

Figure 1-3. Tribal boundaries, settlements, and subsistence areas of the two Upper Skagit divisions (Upper Skagit proper and Suiattle-Sauk) and of the Skykomish and Snoqualmie, comprising a second major grouping. This map, printed anonymously with Collins' (1974c) Upper Skagit religion research, is apparently extracted from her still unpublished study submitted in connection with Upper Skagit early land claims litigation.

Figure 1-4. Upper Skagit "extended village" ("village clusters") as reported by Collins (1974a:17 map 2 [modified]). Six stream identifications have been added: (a) Jackman Creek, (b) Diobsud Creek, (c) Bacon Creek, (d) Thunder Creek, (e) Cascade River, (f) Illabot Creek. "Cascade River" has been corrected. Village clusters have been numbered and encircled for clarity. The two uppwermost clusters of "sk'wix^w and k'wabacáb are of particular interest. (Base mape reproduced with permission of the University of Washington Press).

A total of 47 villages, grouped into the 10 clusters, is enumerated and located by Collins (1974:16-20). An additional village is shown on her map (Figure 1-4) for the Sauk River Valley but not mentioned in her text and so is not included in the considerations that follow. Of these 47, 35 were winter villages; nine were summer villages; and three were settlements for which the season of occupancy is not indicated. Two of the summer communities are reported to have occupied the sites of winter villages, reducing the total number of occupation localities to 45. These winter and summer data raise an interesting problem. Inasmuch as there were so many more winter than summer villages and since at least 25 of the winter settlements were large or very large, as classified by Collins or deduced by me from her descriptive data, and only one of the summer settlements was certainly large -- five at most for there are four summer communities the size of which is not noted -- one may ask where many of the Upper Skagit were following the seasonal withdrawal from the winter quarters? Further attention will be directed to this question at a later point.

Each village -- winter or summer -- consisted of one or more wooden dwellings. In fact, 26 of the 35 winter settlements and five of the nine summer villages were made up of only a single structure. Thus 31 of the 44 villages for which season of use is reported were one-dwelling communities, as, indeed, were all three of the settlements for which season of occupancy information is lacking. In sum, 34 of the 47 known Upper Skagit communities were single house villages. Of the winter villages, five consisted of two structures, two of three dwellings, one of four, and one of five. Of the summer settlements, one is reported to have contained three houses, one four dwellings, and two five houses. Obviously the typical Upper Skagit village was one of a very limited number of living structures. From these data it is apparent that village size as discussed in the preceding paragraph relates to population numbers, not to number of constituent residential units. Collins (1974a:19) mentions specifically two dwellings that measured 120 feet in length and 40 feet in width, one of which housed on occasion as many as 20 families, plainly in each case a very considerable number of people.

With these basic settlement data before us, we may now turn to those village clusters and individual communities that were located in and near the North Cascades Complex.

According to Collins (1974a:17 map 2,19), the easternmost Upper Skagit village cluster -- that termed kwabacabs -- lay "deep in the Cascade Mountains" (Figure 1-4). This settlement complex, her data indicate, was wholly within the perimeters of the Park Complex, specifically in the southwestern end of the Ross Lake Recreation Area. It consisted of three villages, all on the north side of the Skagit. In downstream order:

(1) A large winter house at Newhalem.

(2) A "small winter house with three families to the east of Thornton Creek."

(3) A large winter dwelling "at Portage, west of Damnation Creek."

On the basis of these brief data, the locations of the three settlements are plotted as closely as possible in Figure 1-5 to this point I have been unable to locate Portage. A river distance of approximately five miles separated villages (1) and (3) above. No summer settlements are noted as parts of this village cluster. This fact -- assuming that there were in fact none such -- returns us to the question posed above: where were the occupants of these three winter communities during the spring to fall period? At least part of the time scattered in temporary shelters in small fishing camps, too insignificant and mobile to qualify in the Upper Skagit view as villages, along the Skagit and perhaps lower reaches of tributaries? And at times in the mountains pressing their hunting and gathering activities? If so, they must surely have passed some of these months exploiting the subsistence resources of the nearby North Cascades Complex. We return to this interesting question following a review of other village clusters farther downriver.

The next village cluster on the Skagit downriver from k'wabacáb was that of sk'wix^w (Collins 1974a:17 map 2) or bsix^wix^w as Collins (1974a:14, 19) transcribes the term in her text. Again in downriver order, three separate villages were positioned on the Skagit:

(4) "A large winter house holding ten families.., on the south side of the Skagit across from the mouth of Bacon Creek."

(5) Two small winter houses with two families each, "a mile below Diobsud Creek on the west side of the Skagit."

(6) A winter village of several families near Marblemount.

On Cascade River, which empties into the Skagit in the vicinity of Marblemount, four villages were remembered by Collins' informants:

(7) A winter house at bsix^wix^w on the north side of the Cascade River about one mile west of Marblemount; surely Collins' means "east" -- not her only apparent east-west mixup. This structure was 120 feet long and 40 feet wide and housed 20 families at one time.

(8) A summer structure on the south side of Cascade River opposite village (7).

(9) A summer dwelling on "the south side of the Cascade River at Boulder Creek."

(10) A large summer house "on the north side of the Cascade River to the west side of Monogram Creek."

The preceding seven villages are located in Figure 1-5 as accurately as the above data and the markers on Collins' map (Figure 1-4) will allow. Monogram Creek drains Monogram Lake (Green Trails, Marblemount n.d.), a stream shown but unnamed on USGS North Cascades National Park (1974).

It is notable that this sk'wix^w cluster included three summer communities, all on the Cascade River, and that all four of its winter villages were either on the Skagit or not more than a mile distant from it. All seven villages of this complex were outside the Cascade Park boundary; however, the northeasternmost winter settlement (4) was less than a mile beyond the point where the southwestern line of the Ross Lake area crosses the Skagit. Whatever may have been the case with the more downriver Upper Skagit village complexes, the people of this sk'wix^w cluster surely must have made some use of the resources within the North Cascades Complex.

Below the sk'wix^w cluster was the village grouping known as the sk'axax&uactue;cid with three winter villages (Collins 1974a:18-19) (Figure 1-4):

(11) A winter village (púk^walicu) near the mouth of Rocky Creek. (Whether to the east or west of this side stream remains unstated.)

(12) A winter settlement (saxípp) on the south bank of the Skagit opposite Rocky Creek.

(13) A winter house (six^wucid) at the river bend on the north side of the Skagit "west" of Rockport. (Her village plotting [Figure 1-4] makes it appear that she means "east" of Rockport.) This structure was the largest in Upper Skagit country, 120 feet long and 40 feet wide.

These approximate locations are indicated in Figure 1-5.

Immediately downriver from the village cluster termed sk'axaxúcid was the two village complex known as s?ilayucid, differing from most in that the first village consisted of a split site (Collins 1974a:18) (Figure 1-4):

(14) Two large winter houses, one east of the town of Rockport and the other west of the town.

(15) A small winter dwelling "at the town of Sauk west of Rockport."

These sites are marked in Figure 1-5.

At this point, roughly 15 miles downriver from the southwestern corner of the Ross Lake Recreation Area, we are sufficiently distant from the Park Complex to make irrelevant an accounting of the numerous Upper Skagit traditional settlements farther downstream. Further, this is a convenient geographical break because we are here at the junction of the Sauk River with the Skagit. The Sauk, with its principal tributary the Suiattle, and the upper Skagit itself served, it appears, as the two main routes from Upper Skagit country eastward over the Cascades.

The southeastern corner of Upper Skagit country includes the drainage basin of the Sauk River and its major tributary, the Suiattle. In the Sauk Valley, Collins (1974a:19, 55) reports, was a village cluster -- sák^wbix^w, "people of digging roots" -- with houses in three villages (or four according to her map [Figure 1-4]) along the Sauk banks:

(16) A large winter village located "at the mouth of the Sauk on the south bank of the Skagit very close to a major concentration of villages of other clusters that ran along the Skagit Valley from the mouth of Illabot Creek downstream for miles (Figures 1-4, 1-5, and 1-6).

(17) Four winter houses forming a single village (bakálb) situated on Sauk Prairie, where the Upper Skagit folk "semicultivated wild roots," on the Sauk above the entrance of the Suiattle (Figure 1-6). Sauk Prairie lies between the Sauk and the North Fork of the Stillaguamish River.

(18) "One summer house... at Bedal near the joining of the North and South Forks of the Sauk River" (Figure 1-6). Bedal fails to appear on contemporary maps (e.g., USGS Concrete quadrangle 1955), though Bedal Peak and Bedal Creek are marked. This village in the vicinity of Bedal was on the order of 37 airmiles southwest of the lower limits of Lake Chelan Recreation Area and is presumed to be too distant to be of much interest here.

According to Collins (1974a:19), no winter villages were maintained by the Upper Skagit on the Suiattle River before the 1890 period, although summer camps were made along the stream. However, to how great an extent the region was utilized by hunting, fishing, and foraging expeditions is unclear, except that it is designated a tiger lily and fern root area and a hunting district for deer, bear, and elk on Collins' food resource map (Figure 1-3). In any event, this country, like the Sauk Valley farther toward the coast, was substantially west of the North Cascades Complex border.

Both Suiattle and Sauk were fed from headwater streams on the western slopes of the Cascades crest. At the source of the Suiattle were, for example, Ross and Suiattle passes leading over to the feeders of Lake Chelan, and Buck Creek Pass and High Pass giving access to the upper Wenatchee country (USGS Concrete quadrangle 1955). The North Fork of the Sauk, in turn, has its origin in the White Pass sector, which likewise points southeastward to the Wenatchee basin. Native traders and travelers surely passed through some of these cols, if not all of them, in moving east and west between the Plateau and Upper Skagit territory.

With the eastern and southeastern sectors of aboriginal Skagit territory considered, we may now turn to the northern region. According to Collins' (1974a:17 map 2,18) (Figure 1-4) data the village cluster known as s.báliuq^w occupied the Skagit Valley both upstream and downriver from the confluence of Baker River, settlements of no essential moment to this study. But one winter village (19) comprised of a single small dwelling "was located at Baker Lake connected by a tributary to Baker River." On its face this is a perplexing statement if taken literally, since Baker River flowed through Baker Lake. But Collins' village distribution map makes clear what is intended (Figure 1-4). The village in question was, in fact, situated slightly northwest of Baker Lake on the west side of Swift Creek, which in earlier years flowed into Baker River immediately downstream from Baker Lake (USGS Concrete quadrangle 1955; for the contemporary post-dam construction topography see USGS North Cascades National Park 1974). This village was only about four miles west of south from the southwestern corner of the Park jog around Mount Shuksan; it was approximately 5.5 miles east of the north-south line of the Park east of Baker Lake. More to the point, the people of this rather isolated winter settlement had only to ascend Baker River Valley to reach the Park or, alternatively, to follow the Swift Creek and then Shuksan Creek courses to reach the Mount Shuksan vicinity of the Park from a southerly direction. It would be curious indeed if the occupants of this winter community failed to visit the Shuksan region of the Park Complex.

What were the factors governing the division of the individual winter and summer villages into village clusters, Collins' "extended villages"? A glance at the situation for the second cluster -- the sk'wix^w -- suggests an answer. This village complex extended along the Skagit for some four miles and up the Cascade River for approximately 6.5 miles. Its uprivermost Skagit River village, that opposite Bacon Creek (4), was about four miles below the southwesternmost village of the k'wabacáb cluster, that near Damnation Creek (3), and roughly 2.5 miles upstream from its nearest fellow sk'wix^w village below Diobsud Creek (5). Phrased differently, the river distance between villages (3) and (4) was somewhat greater than the river mileage between any neighboring pair of villages within the kwabacabs cluster and also between similar pairs within the sk'wix^w complex. Similarly, the uprivermost settlement of the next village cluster -- the sk'axexúcid -- below sk'wix^w was about four stream miles below the latter's Marblemount village (6), a river gap rather greater than any separating the individual communities of the sk'wix^w cluster and, as inspection of Figure 1-4 will reveal, than any dividing the settlements of the sk'axaxúcid. All of this suggests the possibility that a proximity factor may have played a role in determining the boundaries of the village complex. Except for the village clusters from the sk'axaxúcid down to the Birdsview area, a river stretch in which villages formed an all but continuous chain of communities, this hypothesis seems to be generally supported by the spatial separation of the individual downriver complexes.

Figure 1-4. Upper Skagit "extended village" ("village clusters") as reported by Collins (1974a:17 map 2 [modified]). Six stream identifications have been added : (a) Jackman Creek, (b) Diobsud Creek, (c) Bacon Creek, (d) Thunder Creek, (e) Cascade River, (f) Illabot Creek. "Cascade River" has been corrected. Village clusters have been numbered and encircled for clarity. The two uppermost clusters of sk'wix^w and k'wabacáb are of particular interest. (Base map reproduced with permission of the University of Washington Press.)

To what extent do the other two village distribution maps support or offer data differing from the village location information of Collins summarized above? First, Collins' (1974b:622) map (Figure 1-7), printed with her 1946 Skagit religious change paper, is patently a rough effort. It is difficult to read, at least in its published form, mixes unclearly drainage, village cluster (her "extended village") and individual settlement terms, and is without accompanying text discussion. To some degree, however, it is possible to match village locations with those in her later map (Figure 1-4) and her terms with the village cluster and individual settlement designations provided in her Upper Skagit ethnography (Collins 1974a:17-20).

In this 1946 sketch map no villages of the k'wabacáb cluster are shown. Indeed, the easternmost Skagit River village is marked at Marblemount. This settlement is said to have been termed basq'éqwiuq^u (Figure 1-7, village [13]), which, in this early and less skillful transcription, is surely equivalent to her later basq'éqwiuq^u (Figure 1-5, village [7]), which was on the Cascade River about one mile "west" (Collins means east) of Marblemount. Even she plots it on the west bank of the Skagit northeast of the present town. This is the only settlement reported on this 1946 map for the sk'wix^w village cluster of Figure 1-4, doubtless called to Collins' early attention because of its size (see above).

The next village (Figure 1-7, village [12]) downstream is shown to be on the east side of the Sauk River just above its mouth. It appears from Collins' map key to be given the name tcáqwaiq. No village with a designation approximating this transcription is found in Collins' (1974a) settlement roster in this general area. Nonetheless, it is placed in about the same location as village (16) (Figure 1-6) of the sák^wbix^w cluster. Another settlement (Figure 1-7, village [14]) is located on the east side of the Sauk perhaps 2.5 miles north of Darrington (USGS Concrete quadrangle 1955); its name is given as sábkiuq^u. The later Collins' (1974a) study fails to report the native designation of the individual Sauk Valley communities, but this term looks suspiciously like a somewhat mangled cluster term sák^wbix^w (see above) with a transposition of the -k^w- and -b- sound sequence. I suggest that it is to be equated with village (17) of Figure 1-6, unless the settlement placed by Collins (1974b:622) (Figure 1-7, village [15]) on the lower Suiattle and called suyátl'biuq^u is intended to be this village (17). At any rate, one community (Figure 1-7, village [14]) appears to be located on the upper Sauk -- but not nearly as far upstream as village (18) in Figure 1-6 -- and one (Figure 1-7, village [15]) on the lower Suiattle River. No traditional village is shown on the Skagit between the mouth of the Sauk and the present town of Concrete, where we are clearly beyond our geographical area of interest. And no settlement is marked in the Baker Lake region.

We now turn to the Collins' (1974c:facing 760) map that was printed as an adjunct to the 1960 report of the Indian Claims Commission (1974). Although the map (Figure 1-8) identifies no authorship, Commission com ment credits Collins. In addition, circumstantial evidence points strongly to a Collins source. This is essentially proved by the data that follow, the perfect matching of the villages of this map, at least in the geographical region of concern here, with those reported by Collins (1974a) in her Upper Skagit ethnography.

Although this map is unaccompanied by descriptive material, it plainly identifies 44 "villages and house sites," to follow the phrasing of the Commission's (1974:317) findings of fact, in Upper Skagit territory together with several other sites that remain for some reason unnumbered. Of these, 41 (settlements 1-41) were in the country of the Upper Skagit proper and three (42-44) were located in the Suiattle-Sauk drainage (Figure 1-8).

As in the case of the second Collins' (1974b) map, we find of interest here only those villages in the eastern segment of the Skagit River region (those numbered 28-41), in the Suiattle-Sauk area (42-44), and in the Baker Lake country (23), three portions of the aboriginal Upper Skagit territory nearest the borders of the North Cascades National Park Complex. The simplest way of examining the details of this map, none too clear in its reproduced and doubtless reduced form, is to compare its village locations with those marked on Collins' 1974a map and to tabulate the correspondences (Table 1-1). The two sets prove to be in a simple one-to-one relationship.

Table 1-1. Upper Skagit Village Locations Reported in Collins (1974c) (Figure 1-3) Compared with those indicated in Collins' (1974a:16-20, 17 map 2) Upper Skagit Ethnographic Study (Figures 1-5 and 1-6).

Village Clusters per Collins (1974a)	Villages per Collins (1974c)	Villages per Collins (1974a)*
k'wabacáb	41 40 39	1 2 3
sk'wixw	38 37 3? 36 35 34 33	4 5 7 10 9 8 6
sk'axaxúcid	32 31 30	12 11 13
s?ilayucid	29 29 28	14 14 15
sákwbixx	42 43 44	16 17 18
s.báliuqw	23	19

*Number as in my text.

To the results of this matching exercise, only three particular observations seem called for:

(a) the locational dot for village site (41) appears on the south side of the Skagit, not on the north bank as with the identical settlement described by Collins (1974a:19) in her general ethnography and shown on her 1974a map (Figure 1-4, village [1] in my text). In addition, it is placed slightly upstream from the town of Newhalem not "at Newhelem" (Collins 1974a:19). How meaningful these minor discrepancies are is uncertain.

(b) Village (30) is sited more precisely than in Collins' 1974a description of village (13) as numbered in my text. Here it is placed at or close to the mouth of Barr Creek, while in 1974a Collins wrote simply of its being "at the bend in the river on the north side of the Skagit west of Rockport." The position of this settlement marker confirms my earlier supposition that Collins' "west" should be read "east."

(c) Village site (23) appears to be placed on Baker River just below the pre-dam lake, not on the small northern affluent of Baker River as shown for site (19) in Figure 1-4 (village [19] in my text) and as described in Collins' (1974a:18) Upper Skagit ethnography.

On the basis of her Upper Skagit village roster, Collins (1974a:20) states that:

... part of the pattern of the seasonal round of activities can be identified. Other than the winter houses at Sauk, the permanent winter houses were nearly all situated at fishing sites near the mouth of tributaries of the Skagit with a few on the main branches of the river and a few on lakes linked by rivers or creeks to the Skagit. Small wooden houses were built at sites where the fishing was particularly good, to house the visiting relatives who came when the fish were running.

In her village roster, however, Collins makes a point of explicitly locating at or very near creek entrances only 13 of the 49 villages. Perhaps she considered her generalization sufficient on the point to make specific geographic information largely unnecessary. More oddly, she calls attention in her village listing to only one winter village (wáal?l) at a fishing site and two summer communities (súsutiya and scácuks) at fisheries, all three in the downrivermost village cluster in the Mount Vernon-Sedro Woolley neighborhood. This is difficult to explain unless again, having made her statement regarding village locations and fisheries, she thought it redundant to report such information for most specific villages. Further, only one winter village (q^wq^wq^ww) of the 37 total is specifically reported to have been occupied as well during the summer months, the fishing season; and only one summer village location (kabálah) of the nine total is said to have been lived in during the winter. This apparent noncorrespondence between winter and summer village sites requires some further examination.

Figure 1-4. Upper Skagit "extended village" ("village clusters") as reported by Collins (1974a:17 map 2 [modified]). Six stream identifications have been added: (a) Jackman Creek, (b) Diobsud Creek, (c) Bacon Creek, (d) Thunder Creek, (e) Cascade River, (f) Illabot Creek. "Cascade River" has been corrected. Village clusters have been numbered and encircled for clarity. The two uppwermost clusters of "sk'wix^w and k'wabacáb are of particular interest. (Base mape reproduced with permission of the University of Washington Press).

Attention has already been called to the absence of summer settlements in Collins' record for the eastern most Upper Skagit village complex (k'wabacáb) and likewise to their relatively small numbers in comparison to winter villages farther downstream. The question has been raised as to the possible significance of this fact for a comprehension of the warm season subsistence and population distributional patterns of the Upper Skagit. In this light, it is of interest to review Collins' (1974a:15-20) "house" data more fully and to attempt some rough population estimates for the entire Upper Skagit group.

Table 1-2 presents a compilation of dwelling data extracted from Collins' descriptive text. For the purpose of this present analysis, the five village clusters down the Skagit through the s?ilayucid (Figure 1-4; Table 1-2, 4) and the sák^wbix^w cluster on the Sauk River (Figure 1-4; Table 1-2; 5) may be rather arbitrarily regarded as the upriver Upper Skagit. The five clusters beginning with the s.báIiuq^w cluster (Figure 1-4; Table 1-2, 6), located from Faber Ferry (USGS Concrete quadrangle 1955) westward down the Skagit, maybe considered to comprise the downstream division of the Upper tribe. This arrangement not only divides the village clusters into two groups of five each, but also places the division line near the geographical midpoint in the cluster distribution along the Skagit.

From the data of Table 1-2, it is apparent that, assuming the essential accuracy and completeness of Collins' information, there were 19 winter dwellings -- large, small, and of unspecified size -- in the upriver area and 33 in the lower. Likewise that there were relatively more large winter dwellings than small -- in a 2:1 ratio -- in the eastern region than in the downriver sector (virtually a 1:1 ratio). The situation in regard to the summer dwellings was different: only one summer "house," a large one, in the entire upriver area (plus three of undefined size) as against 17 small and no large summer dwellings in the downriver division (plus one of undescribed dimensions).

Table 1-2. Number of Dwellings by Season and Size, Compiled from Data Provided by Collins (1974a:15-20

Village Clusters*	Winter			Summer			Season Unspecified			Total
	Small	Large	Size Unstated	Small	Large	Size Unstated	Small	Large	Size Unstated
1 k'wabacáb	1	2								3
**2 sk'wixw	2	2	1		1	2				8
3 sk'axaxúcid		1	2							3
4 s?ilayucid	1	2								3
5 sakwbixw		1	4			1				6
Totals	4	8	7		1	3				23
6 s.báliuqw	12	5		7			2		1	27
7 basla?álux		1								1
8 ub?áb		2								2
9 sk'ikwigwilc	3	4								7
10 dqwab's	1	5		10		1				17
Total	16	17		17		1	2		1	54
Subtotals:	20	25	7	17	1	4	2		1
Totals:		52			22			3		77

In summary, this dwelling catalog tells us that there were only about one-fifth as many upriver summer dwellings as winter. And approximately one-half as many summer houses as winter among the western segment of the Upper Skagit, but practically all of these summer structures were small, whereas only half of the winter dwellings were of small size. Probably some upriver people moved downstream to take advantage of summer fisheries in the lower river sector. At least according to Collins (1974a:20) as noted above, some Upper Skagit visited their tribal kinfolk "where the fishing was particularly good" and "small wooden houses were built at [these] sites . . . to house the visiting relatives who came when the fish were running." I would suppose that the downriver fisheries were, on the average, more productive than those upstream, causing in summer a predominately downriver population movement, though I have no hard evidence that this was the case. If this fisheries supposition is correct, it would also mean that any upriver flow of people from the lower Upper Skagit villages would probably have gone back into the high country and so not show up to a significant extent in the summer settlements in the eastern Upper Skagit territory. At any rate it is obvious that there were not nearly enough summer villages and dwellings to take care of the entire upper Skagit winter population in the lower segment of the tribal homeland. This would surely seem to be so even considering the temporary housing constructed to accommodate the visiting kinfolk mentioned above. Certainly there is nothing to argue for any considerable downstream summer flow of the Upper Skagit population.

It is also probable that some Upper Skagit from the eastern villages left their tribal territory during at least part of the warmer months for the country of other nearby tribes. For example, Upper Skagit with wives from the Lower Skagit and from the Nooksack, Swinomish, Upper Samish, Stillaguamish, and other coastal tribes both to the north and the south and those who were partly from these tribes in descent, visited their relatives in those groups in the summer. Others went to these nearby tribes to trade, especially for marine foods, but also in some cases to dig roots (Curtis 1974a:9-11, 32). Presumably people from the upriver Upper Skagit villages were at times involved in these journeys. But the downstream attraction of better fishing localities and visits to neighboring tribes could hardly have drawn away from the eastern Upper Skagit winter villages a large enough segment of the population to account for their small number of summer settlements. And as already observed, even the lower reaches of the Skagit River had too few summer settlements to account for its winter population.

One of the most obvious explanations for the facts revealed by Table 1-2 is that among both divisions people spent a considerable part of the summer season away from the river and that this was notably more so for the eastern group. This seems consistent with what might be supposed to have been a relatively rich supply of anadromous fish downriver and the relative abundance of game upriver. The people simply were not congregating during the fishing period at a few especially favorable fishing sites.

This conclusion is made still more obvious by converting the dwelling data to population approximations (Table 1-3). Unfortunately, Collins describes houses only as "small or "large" and provides no average or range numbers for families or persons occupying these two classes of dwellings. Nor does she give average person/family data. The best we can do is to examine the following specific family/dwelling information, some of which is used to illustrate social organizational patterns (Collins 1974a:83-84) and so must be matched with her separate village roster information (Collins 1974a:16-20).

Table 1-3. Population Estimates for Large and Small Dwellling Aggregates and Aggregates of Dwellings of Indeterminate Size by Season, Divided into Upriver and Downriver Groups*

*Three dwellings of unknown season are excluded.
**Formula A: small dwelling = 15 persons; large dwelling = 70 persons; dwelling of unstated sized = 42 persons.
Formula B: small dwelling = 25 persons; large dwelling = 125 persons; dwelling of unstated size = 75 persons.

"Large" winter dwellings by Collins' definition:

(1) qweqwqweqw: four families. (Collins 1974a:18, 83-84)

(2) besqixwixw: house 120 x 40 ft; once 20 families. (Collins 1974a:19)

(3) besqizwixw: to judge from the location given by Collins (1974a:84), she is again referring to village (2) above, though she here terms it kwebacabs, which was considerably farther up the Skagit. Five families, which I take to be the occupants in more recent years (Collins 1974a:84).

(4) sqixwucid> house 120 x 40 ft; therefore presumably about 20 families as in (2) above. (Collins 1974a:19)

(5) Native name ?. Opposite mouth of Bacon Creek; 10 families. (Collins 1974a:19)

"Small" winter dwellings by Collins' definition:

(6) Native name ?. Near Diobsud Creek; two houses with two families each. (Collins 1974a:19)

(7) Native name ?. East of Thornton Creek; three families. Collins 1974a:19)

(8) Apparently wawalah: three dwellings sheltering one family, three families, and four families respectively. (Collins 1974a:16, 84)

No comparable data are given for "large" and "small" summer dwellings.

From these sparse and not entirely unambiguous data we conclude that "large" dwellings housed from about five to 20 families and "small" dwellings from one to four families. These data translate into approximately 30 to 120 (average 70) people and five to 25 (average 15) persons respectively. Probably these figures are smaller than they were in earlier times before population reduction occurred. But if both estimates are too low, the results of the rough computations that follow and their implications would not be appreciably altered by adjusting them proportionately upward. Where the dwelling size is not reported by Collins, an intermediate population average of 42 people is estimated. Lacking any evidence in the matter, summer large and small dwellings and those of undescribed size are tentatively assumed to have housed populations about equal to those of winter villages. The three dwellings of undesignated seasonal occupation are omitted in this compilation; since at least two of them were small, their exclusion from the data base can hardly affect the results significantly. It turns out that, using this relative size formula (Table 1-3, formula A), approximately 21.4 percent of the winter population (196 divided by 914 persons) among the eastern Upper Skagit is accounted for during the summer months and about 20.8 percent of the winter population (297 divided by 1,430 persons) among the downriver Upper Skagit.

But perhaps the population difference between the small and large dwellings was somewhat greater than in the above approximation. If the estimate for the small house occupant number is increased to 25, that of the large dwelling to 125 persons, and that of those few houses of unreported size to 75 people (Table 1-3, formula B), then the summer/winter ratio for the eastern Skagit segment becomes approximately 21.5 percent (350 divided by 1,625 persons) and that for the downriver segment drops to about 19.8 percent (500 divided by 2,525). From these rough computations it is plain that only about 20 percent of the people of the winter dwellings are accounted for in the warmer seasons.

There is certainly nothing in these data at either formula level to argue for a meaningful movement of upstream people to summer locations downriver. It is likewise apparent that, assuming Collins' findings are substantially accurate, the Upper Skagit did not occupy as many fishing camps along the Skagit as might be expected in view of the reported abundant salmon supply in the river and the importance of anadromous fish as a food resource (Collins 1974a:45-52).

The answer to this problem is not entirely obvious. Having reported that winter villages were generally at fishing sites (Collins 1974a:20), did Collins think it unnecessary to mention the summer settlements at these same localities? But if so, why distinguish between winter and summer communities in her village listing and why call specific attention to one winter village site that also supported summer dwellings and one case of the opposite sort? Indeed, why were nearly all cold season communities at fisheries as she avers, when the fishing season was during the summer months?

And still more questions are raised by the winter/summer population problem just discussed. Were many Upper Skagit scattered too widely along the Skagit during the fishing season and in too small groups to be considered as comprising "villages"? Did many of the people occupy structures too insubstantial and impermanent to be classified as "houses" by Collins -- and presumably also by the Upper Skagit themselves -- while at summer fishing centers? Evidently it was not that their fishing season was too brief, in view of both the Fraser River salmon data presented in the following chapter and Collins' own information to be reviewed presently, for settlements of some permanence to form. But perhaps the people were typically drifters in the warmer months, moving back and forth between river fishing localities and food quest grounds up in the hills and mountains back from the Skagit River. At a later point these questions, bearing directly or at least peripherally on the Upper Skagit utilization of the Park Complex area, will be addressed more fully.

In addition to these maps with their village plots, several village cluster or "extended village" lists exist. Collins herself has produced three different versions. Two appear on maps reproduced here as Figures 1-4 and 1-7 and have been already discussed; the third is a text roster. The listing on Figure 1-7 is a mixture of extended villages, individual villages, and geographical designations and is, in addition, too indistinct in its published form for me to read with confidence. Consequently, these data are not considered further here. The third appears in the Indian Claims Commission (1974:316-317) case document. This catalog is a particularly informative contribution because the Skagit terms are reduced to their more popular equivalents. These terms, as well as the ten-item list in Figure 1-4, are given in Table 1-4.

Table 1-4. Upper Skagit Village Clusters, Ordered East to West.

Collins (1974a:16-19, 17 Map 2)	Collins in ICC (1974:316-317)	Smith (1941:210)	Gibbs (1877:180a)	Eels (1985:19)
Skagit River Groups
1. k'wabacáb	Kwa-bats-absh
2. sk'wixw	Shay-wih	bsqéqwiuqu (J)
3. sk'axaxúcid	Shay-ay-ot-sid
4. s?ilayucid	Say-lay-ot-sid	bsélelotsd (I)
5. s.báliuqw	Sabaliuhk	sbáleuqu bsxádsadsiuqu	Smali-hu	sab-li-hu sba-le-hu
6. basla?alux	Baosloh-a-loh	bslo''xlo'x
7. ub?ab	Cho-bah-ah-bish	tcubábc
8. sk'ikwigwilc	Seek-weeg-gweelts Mee-see-qua-quilch	sikwigwilts	Miseekwigweelis
9. dqwab?s	Nook-wa-cha-mish Nookachamps Dook-wa-shabs	dvkacábs dqwtcabs		do-kwa-tcabsh
Suiattle-Sauk Groups
10. sákwbixw	Sah-ku-mehu	sábkiuqu (K) suyátl'biuqu (L)	Sakumehu	sak-wi-be-hu

Note: One Collins listing, that on her map reproduced here as Figure 1-7, is to difficult to read with confidence. It is omitted in this table.
^aA number of Gibbs' groups cannot be matched with Collins' extended villages.

Another recent catalog of Skagit extended villages is that of Smith (1941:209-210). No differentiation is made between Lower and Upper Skagit village complexes, but the separation between the two divisions is clear from Collins' data. In some respects Smith's list differs from that of Collins on Figure 1-4, as shown in Table 1-4. Specifically, it includes two extended villages not reported by Collins: i.e., besxadsedsiuqu, on the south bank of the Skagit from Hamilton to Birdsview, and suyatlbiuqu, in the Suiattle River region. On the other hand it omits two in Collins' series (see Figure 1-4). Since we are concerned only with the uprivermost extended villages of the Upper Skagit, the complete data from Smith's settlement catalog are not reproduced here. Her four upstream communities are as follows, lettered as in her roster: (I) baselelotsed on the Skagit River from Van Horn to roughly three miles above Rockport and on the Sauk River almost to the mouth of Suiattle River, with the village of tcagwalq' at the mouth of Sauk River; (J) basqeqwiuqu on the Skagit River above Rockport with a settlement of the same name at Marblemount at the mouth of the Cascade River; (K)sabkiuqu, the Sauk River above the confluence of the Suiattle, with a village of the same name on Sauk Prairie below Darrington; and (L) suyatlbiuqu, Suiattle River with a village of the same name not far above the Suiattle River mouth (Smith 1941:210). Of special interest is the fact that the differences between the Collins and Smith rosters are almost entirely concentrated in the Upper Skagit back country.

Two nineteenth century attempts to itemize the Skagit village groups are known to me. The earliest is that of Gibbs (1877:180) who states, surely based on his 1854 field findings, that the "Skagits" include the "Kikiallu," Nukwatsamish, Tow-ah-ha, Smali-hu, Sakumehu, Miskaiwhu, Miseekwigweelis, Swinamish, and Skwonamish" (cf. Gibbs 1877:241). The Samish were a separate non-Skagit people. Several of the remaining Gibbs' groups may be identified more or less certainly. The Kikiallu were on one of the northern Skagit River mouths and so are too far from the Park to be of interest in this study (Spier 1936:36). Perhaps the Nukwatsamish were a subunit of the Samish and the Tow-ah-ha may be equated with the Huwhaha (Upper Samish); if so, these groups likewise fall beyond our concerns. The Smali-hu, however, appear to be Collins' s.baliuqw around the mouth of Baker River, an Upper Skagit "extended village." The Sakumehu are surely Collins' sakwbixw village cluster of the Sauk River country (Spier 1936:36). Gibbs (1855b:472) terms the Sauk River the Sah-kee-me-hu. The Miseekwigweelis are undoubtedly Collins' skikwigwilc. The remaining groups itemized by Gibbs defeat me. The population units identified appear in Table 1-4.

Eells (1985:19), who was aware of the Gibbs listing, learned of four Skagit "bands," excluding the Swinomish as with Gibbs above: the do-kwe-tcabsh at the Skagit mouth; the sba-li-hu on the middle Skagit River; the sba-le-hu whose country was on the Baker River; and the sak-wi-be-hu on the southern branch of the Skagit. The first must be Collins' deqwcab?s close to the river mouth; the second and third Collins' s.baliuqw on the Skagit River in the Baker River area and up that stream (Spier [1936:36] suggests that these two groups must be the same); and the fourth, Collins' sakwbixw, the Sauk River people. These equivalences appear in synoptic form in Table 1-4.

Travel and Transportation

The Upper Skagit lived predominately along rivers and to a lesser degree on lakes. They were therefore preeminently a canoe people, carrying out as much of their travel and transportation by watercraft as possible.

Three types of cedar dugouts were manufactured. Of these the shovel-nosed, river form "was by far the most important." Typically it was about four feet wide, two feet deep, and sometimes long enough to provide transportation for ten people. A tree was cut, trimmed, and floated to the maker's home. There, on uprights, it was hollowed out by careful adzing, burning, and steaming. The two ends were identical, both blunt. It was fitted with thwarts to insure that the sides maintained their shape. The second type was a small, one-man canoe with bow and stern sharply cut away. This was used exclusively on swamps in duck hunting. The third canoe was a three-piece saltwater craft, made by some Upper Skagit for sale to shore people. It was not employed on the Skagit River and is therefore of little interest to this study.

Paddles of vine maple with a T-shaped handle were used less frequently than poles in moving canoes. The latter were always employed in propelling canoes upstream. Paddlers knelt on the canoe bottom, sometimes padded with mats. Young people sang, keeping time with their paddle strokes (Fornsby in Collins 1949:310; cf. Mayne 1962:61 for the middle Fraser Indians).

Occasionally when a ceremonial guest had many gifts to distribute, two canoes were "united with a platform of planks placed over them to hold them together" (Collins 1974a:66). The objects were thrown from this platform to the shore where people scrambled for them. (Collins 1974a:64-66)

According to Upper Skagit tradition, Collins (1974a:66) reports, people who once lived in far upriver villages (e.g., on the Sauk) were unacquainted with the making and use of canoes. When the river level was below their heads, they crossed "by carrying a heavy stone on the shoulder. They had trails in the Sauk River for fording in this manner." One must give such "traditions" little credence: they seem preposterous on the surface. It is of special interest that tales of mythical "foolish folk" were widespread in the nearby Plateau, where informants reported that they were commonly believed -- or half-believed -- in the old days (cf. Smith 1936-1938).

The Skagit are described by Gibbs (1877:180), drawing upon data probably collected in 1854, as having no horses in the 1850s, as being "altogether canoe Indians." This is undoubtedly wholly or essentially correct for the Lower Skagit. However, the Upper Skagit groups, or at least those closest to the mountains, almost surely had a few equines in the 1825-1850 period, used primarily on over-the-mountains journeys to the Plateau. But for this view I have not yet found historical or ethnographic evidence beyond Gibbs' (1855a:432) general statement already noted that the peoples near the western end of the Cascades passes owned a few horses.

Snowshoes were used by upriver Upper Skagit men and women, though not by the downriver and saltwater villagers. Their form and use are not described except that these headwater people wore them in winter in getting about in the mountains and when crossing the Cascade divide. (Collins 1974a:66, 74)

Trails

In regard to the travel patterns and preferences of the Upper Skagit, Collins (1974a:6) provides a brief but very informative statement. She writes:

[The forests of Upper Skagit country were] so dense that travel through them was extremely difficult. Land travel was confined for the most part to places near timberline where the vegetation was relatively sparse. The Upper Skagit making the long trip across the Cascade Mountains to visit relatives in the Plateau preferred to travel on the snow itself.

In light of the above, Gibbs' (1877:169-170; also 180) observations of the mid-1850s are especially informative. Concerning the cross-Cascades trails in general in the Upper Skagit neighborhood and on south, he writes:

In former times, before the diminution of the tribes and the diversion of trade to the posts, there were numerous trails across the Cascades by which the Indians of the interior obtained access to the western district. Of late many of these have fallen into disuse, becoming obstructed with timber and underbrush which they have not industry enough to clear out. In fact all their trails through the forest, though originally well selected, have become excessively tortuous, an Indian riding around the fallen trunks of tree after tree sooner than clear out a road which he seldom uses. . . . The northern trails from the Winatshapam [Wenatchi] and Tselann [Chelan] Lake to the Sto-luch-wha-mish [Stillaguamish] and Skagit seem to be altogether so [abandoned].

These comments suggest at least a three time-level trail-condition sequence: (1) narrow, prehorse, winding trails suitable for foot travel in which obstacles -- like fallen trees -- were generally detoured around, since the labor of clearing them away with stone tools would have been altogether too great; (2) wider trails required by burdened horses, running over terrain suitable for mounted horses and cleared to some extent with metal axes; and (3) the very difficult trails described by Gibbs, associated with population reductions through epidemics and other adverse conditions and the diversion of trade to the posts.

The preceding statement by Gibbs speaks of Plateau people visiting the Skagit. The extent of Skagit movement on visits for trade and other purposes eastward over the Cascades is not specified. Referring evidently to the 1850 time frame, Gibbs (1877:180) remarks only that the Skagit "formerly had some communication with the Indians beyond the mountains; but it is supposed to have been discontinued in consequence of obstructions to their trails." I find this observation ambiguous as to the identity of the journeying parties.

While some Upper Skagit certainly traveled as the data of this section clearly demonstrates, some persons, if Smith (1941:199) is correct, were non-journeyers. For the inland tribes as a whole, she reports the "territory visited by one individual in his life-time was normally limited and old men and women of inland villages who have never visited the Sound are still living." In fact, this may well have been correct for many Upper Skagit. John Fornsby, himself an Upper Skagit by birth, reported that the people of the middle and upper Skagit River areas "never wanted to go down to salt water. [These folk] ... were scared that somebody would take them for slaves or kill them. Oh, sometimes they went down and dug clams and took some up the river. They were just like wild fellows and stayed up there [on the upper river] all the time." (Collins 1949:302) Still, considering the extent to which the mountain uplands of Upper Skagit territory had to be scoured for game and other resources, it would surely be inaccurate to view these inland Skagit as stay-at-homes. But they evidently moved mainly about within the limits of their own country, not much as traders or subsistence seekers into the homeland of other groups, even of the Lower Skagit.

Certain scraps of information exist in regard to specific overland routes, though these are not always well located. They certainly provide a far from complete listing of the total network of Upper Skagit trails to the territories of the other groups of interest to this study.

One overland trail, Collins (1974a:11) reports, led from a village on the upper Skagit River to a Nooksack community, thus avoiding a long descent of the Skagit River to the salt water, moving along the shore to the Nooksack mouth, and then moving up that stream. Unfortunately neither of these settlements is identified nor are any of the details of the route between the two furnished. Perhaps this trail is that described by Smith (1950a:339) as having run from the uppermost Nooksack village on the South Fork of the Nooksack River above the present town of Acme "over to the upper Skagit River country." This trail is said to have been six miles long. This distance figure, assuming its approximate accuracy, together with the "upper Skagit River country" observation suggest that the route led up the Nooksack fork to about its closest point to the Skagit River and then southward, perhaps to the neighborhood of Lyman or Hamilton, though this is hardly more than guess work. At any rate it could not have proceeded south from the Acme area to the lower Skagit River sector. Unquestionably there were other cross-country Upper Skagit-Nooksack trails as well (Smith 1950a:331).

Some Upper Skagit went to salt water by ascending the Sauk River, making the short portage over to the Stillaguamish River, and then following it downstream. For some Upper Skagit villages this was a shorter route then straight down the Skagit. Furthermore, the Skagit was blocked in the 1800s by a log jam upriver from Mount Vernon, which required a portage at that point (Collins 1974a:10, 11,38). Evidently this Sauk-to-Stillaguamish trail is the one referred to by Gibbs (1855b:472): "The Indians state that they have a short portage for their canoes . . . from the north fork of the . . . [Stoluckwamish] (Stillaguamish) to the Sah-kee-me-hu [Sauk] branch of the Skagit."

A land trail evidently ran up the Skagit River to its headwaters, although most Upper Skagit travel was by canoe. How difficult it was to negotiate I do not know, but the upper reaches of the river plainly presented real problems for canoe navigation. On this matter Collins (1974a:38) is worth quoting:

. . . in 1852, . . . [an improved] trail was cut [apparently by or partly by Whites] along the Skagit River. Where it began is not certain but it extended to Ruby Creek [at the southeastern corner of Ross Lake] and also beyond into British Columbia. Here at the head of the main channel of the Skagit, the trail crossed the Niccolum Trail. Even after the trail had been made along the Skagit River, miners used the Nooksack River rather than the Skagit route in order to reach the mines because of the difficulties of navigation in the latter.

Further information concerning the Niccolum trail is not available.

The Upper Skagit are known to have made occasional trips over the Cascades into the Plateau and parties from the eastern slopes of the mountains journeyed westward to the Upper Skagit country. A number of trails obviously united these two regions. Several can be enumerated with reasonable certainty, although hard data describing their early use, by the Skagit or their westside neighbors heading to the Plateau or by Plateau peoples moving westward, cannot yet be cited. Nor can the incentives that led traveling parties to undertake this difficult cross-mountain journey be specified beyond the obvious desire to exchange goods (see following Trade section). (Collins 1974a:13; Suttles 1957:177)

The trip over "the passes which was made in winter on snowshoes took about six weeks. A man or woman (and women took this route less frequently than men) might make this trip only once in a lifetime. (Collins 1974a:13)

The following trails are certainly among those that must have been used to some extent by both Upper Skagit and Plateau parties:

(a) People living in the villages farthest up the Skagit traveled through at least two passes into eastern Washington (Collins 1974a:13). One must have been that which ascended the Skagit River and then Granite Creek to cross over Rainy Pass to the Methow country. A second trail went up the Cascade River and the North Fork of that stream and over Cascade Pass to the Stehekin headwaters in Chelan country. The Upper Skagit are known to have "kept a canoe hidden on Lake Chelan for their use" when they arrived at its shores (Collins 1974a:13).

The second of these passes (Cascade) is the one crossed in late August of 1882 by Lt. Pierce (1973:22-24) and his party. As he reports, he made his way over a wide snow bank at the crest of the divide and, heading westward, began a very rapid, winding, and precarious descent through loose rock and a tangle of high alders. Reaching a creek, a headwater of the Cascade River, he wound downward in and out of the stream for "a considerable distance." He came then to giant cedars and a "powerful waterfall." His night's camp was in a magnificent forest of white pines, red firs, and cottonwoods. Over the next four days he followed the Cascade River toward the southwest and then northwest along a serpentine trail through heavy forest, by cliffs, up bluffs, and down canyons to the Cascade River mouth where he found a salmon fishing camp of an old Skagit Indian and his son and their families.

(b) A trail evidently went up the Suiattle River and over Ross or Suiattle Pass or one or more other passes in the area to the headwaters of streams that flowed northeastward to the Stehekin River or eastward to the west side of Chelan Lake. "Harry Moses, whose father was from Entiat . . . made the trip overland by foot three times during his life" (Collins 1974a:13). Access to the upper Wenatchee drainage from the Suiattle may well have been via Buck Creek or High Pass or some other mountain break in their vicinity.

(c) A trail ascended the Sauk and over White Pass to the upper Wenatchee area. Presumably it is to this route that Collins (1974a:80) alludes when writing of William Moses' itinerary. In 1852 Moses, "who lived at Sauk or one of the other upriver villages, together with other men took a load of furs to Portland, Oregon, to trade, . . . They went by canoe up the Sauk River, crossed the Cascade divide and went down the Wenatchee River and the Columbia River to Portland."

Similarly Plateau peoples, with different motivations, occasionally went west to Skagit country. Some small groups, attracted by the salmon available in the Skagit River headwaters, crossed with the idea of settling in the area. Collins (1974a:13-14, 66) sees an early settlement of parts of Skagit Valley by Plateau people in search of salmon as the explanation for Upper Skagit traditions reporting that certain toponyms in their country were Interior Salishan in origin. Unfortunately, no examples of these place names are provided by Collins, making impossible a linguistic test of this hypothesis.

The Upper Skagit report a case of a Nespelem man, fearing a Nespelem shaman, who came west over the Cascades in the early 1800s to live among the Snohomish south of Upper Skagit territory. Some time before 1855 his son married an Upper Skagit woman from a downriver village and settled in her country. "Like some other Indians on the Skagit River he occasionally went to see his relatives among the Interior Salish." (Collins 1974a:32-33)

Sometimes Plateau parties traveled the cross-Cascades routes for brief visits, doubtless partly to trade.

What is known of travel by Thompson and Chelan parties over the mountain crest to the Skagit Valley is to be found in those sections of this report dealing with these individual Plateau tribes.

Trade

A general statement regarding cross-Cascades goods exchange is made by Gibbs (1877:170). The trade between the western and eastern sides

. . . was once considerable. The western Indians sold slaves, haikwa [shells], kamas, dried clams, &c., and received in return mountain-sheep's wool, porcupine's quills, and embroidery, the grass from which they manufacture thread, and even dried salmon, the product of the Yakama fisheries being preferred to that of the sound.

Writing of trade over the Cascades between the middle Columbia people and their nearest coastal neighbors, Teit (1928:121) states:

A great impetus was given to trading by the introduction of the horse. Root-cakes, dried berries, buffalo robes, and many other heavy or bulky packs, which in former days it did not pay to carry, were now transported across the mountains. Before the introduction of the horse, the trading with Coast tribes was chiefly in light and valuable articles. Pipes, tobacco, ornaments of certain kinds, Indian-hemp, dressed skins, bows, and some other things, were sold to the Coast tribes, the chief articles received in return being shells of various kinds. Some horses were also sold to the Coast people.

Without question, the Upper Skagit were active participants in this trade network.

A very early report of the regular exchange of valuable coastal shells for wild hemp is that of Alexander Ross, on information secured from the Sinkaietk (South Okanagan) in 1814. In the years preceding his arrival, parties from this tribe were accustomed to cross the Cascades to carry on this trade. The shells were secured from the Skagit, or at least the journeying parties passed through the Upper Skagit country en route to the salt water. Ross (1956:37) puts the matter this way:

The Red Fox [to whom the "Red Fox road" over the Cascades refers] ... was the head chief of the Oakinacken nation and had formerly been in the habit of going to the Pacific on trading excursions, carrying with him a species of wild hemp which the Indians along the Pacific make fishing nets of; and in exchange, the Oakinackens bring back any marine shells and other trinkets, articles of value among the Indians.

The Skagit were also visited by Chelan trading parties (Walters in Spier 1938:77). Undoubtedly these, as well as visitors from other neighboring Plateau tribes, brought to the Upper Skagit the bow string -- most probably of the Indian hemp referred to by Ross (cf. Ray 1932:89) -- to which John Fornsby (in Collins 1949:295) alluded: the Upper Skagit "used to have good string from east of the mountains for . . . [their bows]." They also, it may be supposed, carried to the Upper Skagit the cream-colored grass -- it grew "near a lake east of the divide" (Chelan Lake?) -- for use in imbricating designs on their baskets, the basket hats worn by some women, and the plant perfumes used to enhance appearance (Collins 1974a:68, 72, 74).

In regard to downriver trade there appears to be little specific information. It is reported, however, that the Upper Skagit took dried salmon to the Lower Skagit country to exchange for blankets (Fornsby in Collins 1949:297).

Intertribal Relations

We are concerned here with Upper Skagit relations at a rather general level with other tribes, particularly those who held or at least used on a frequent basis the territory within the Park Complex and its neighboring areas. Relations involving intertribal marriage and hostilities are dealt with in the following two sections.

On the north the Upper Skagit were on a particularly friendly basis with the Nooksack and even the salt-water Swinomish. Similarly with the Stillaguamish immediately to their south. In fact, with this latter group the Upper Skagit shared their root grounds at Sauk Prairie between the Sauk and the North Fork of the Stillaguamish Rivers. (Collins 1974a:9, 11) In the neighborhood of Darrington the two streams, the former turning to the south as one ascends it and the latter bending sharply to the north, are within no more than 2.5 miles of each other. And they are separated here by a lowland area devoid of travel obstacles (USGS Concrete quadrangle 1955).

Occasionally crossing the Cascades to the Plateau, the Upper Skagit visited in a friendly fashion both interior Salishan-speaking groups (e.g., Okanagan, Methow, and Wenatchi) and Sahaptian-speaking groups (e.g., Kittitas and Yakima) (Collins 1974a:9, 13). Some of the Northwest Coast cultural influences that reached the Southern Okanagan came over the mountains from the Skagit (Spier 1936:3). However, the extent to which the Skagit were the travelers and carriers of these coastal objects and ideas and the degree to which it was the Southern Okanagan -- or Chelan or Wenatchi, for example -- who crossed the Cascades to Skagit territory is not stated.

To some extent Plateau tribes traveled the passes westward into the territory claimed by the Upper Skagit.

Upper Skagit made chance contact with these peoples sometimes in Skagit homeland and at times on the cross-mountain routes followed by these foreign groups in traveling to the Skagit area; typically the relationship was an amicable one. To the degree that data are available, these contacts are discussed in the Chelan chapter of this study. Upper Skagit contacts with the Thompson, on the other hand, are described in the following Hostilities section, to the extent that the data are derived from Upper Skagit informants and are described from their point of view.

Intertribal Marriage

Intermarriage not only between certain Skagit villages but also between Skagit and certain villages of other tribes occurred in a strongly patterned way, decade after decade. Parents preferred their children to live after marriage in some village where they already had relatives. Sometimes the residence-changing spouse was the man; not infrequently it was the wife. The Skagit had such regular marriage arrangements with the Nooksack, the Musqueam and Chilliwack and other Halkomelem peoples, the Stillaguamish, Snohomish, Upper Snoqualmie, and occasionally even more distant Coastal Salishan groups. Upper Skagit also entered into marriage alliances with nearby Swinomish and Upper Samish (Nuwhaha) persons to the northwest, but whether these were under the patterned structure or less formal in character is not known. (Collins 1974a:10, 11, 98-99) In addition, Smith (1940:11, 1950a:337) knew of one Skagit man who, in the late prehistoric or very early postcontact period, lived by choice in Puyallup country and married a woman of that tribe.

Some intermarriage occurred with nearby Plateau tribes on the eastern side of the Cascades, at least by the mid-1800s. These included the Okanagan, Methow, and Wenatchi as well as other nearby Salishan-speaking groups and Sahaptian-speaking peoples like the Kittitas and Yakima. (Collins 1949:291, 1974a:13)

Hostilities

The Upper Skagit were basically peaceable, fear of retaliation keeping hostilities at a minimum, even though "warrior" guardian spirits appear to have been quite important. They are specifically reported to have been at peace, for example, with the Snohomish (Haeberlin and Gunther 1930:12). And also with members of the Upper Stalo Tait tribe of the middle Fraser Valley just above Seabird Island. Tait hunting parties were accustomed to go back into the mountains several days' travel east of the Fraser. There they frequently met Skagit hunting groups without conflict (Duff 1952:21). These encounters must have taken place in the upper Skagit drainage which suggests that Skagit hunters may well have gone into the country some distance above Ross Lake, surely north of the present international boundary, and that Tait as well as Lower Thompson and Skagit made subsistence use of this mountainous back-country on occasion. It may be assumed that the Upper Skagit also maintained peaceful relations with the various peoples with whom they frequently intermarried (see preceding Intertribal Marriage section) .

Raids, however, were occasionally carried out by Upper Skagit parties against other groups in response to raids by members of those tribes. They were mounted primarily by, or at least led by, men with guardian spirits relating to hostile activities. Such a "warrior" gathered interested men and a few strong women (wives of the men?) to comprise his party. The role of the women was to cook and perform other noncombatant services. No dances were held prior to the raid. The party usually went by canoe to a point some distance from the village to be attacked, hid during the night, and, leaving the women to guard the canoes, attacked just before dawn with bows and arrows, spears, and clubs. They killed all except those they wanted to enslave. Sometimes they burned the dwellings. Heads were taken to prove the success of the venture; usually they were not publicly displayed for any length of time. There was no victory dance. (Collins 1974a:115, 118)

Since raids were typically retaliatory in nature, hostilities tended to continue, back and forth, though at a very low level. The downriver Upper Skagit were most exposed to raids by other people, since most fighting occurred with coastal groups. Some of these lower Skagit Valley villages were palisaded. (Collins 1974a:114-115,118) The Lower Skagit are described in the 1850s as having been "long at enmity with the Clallams, who have attempted to encroach upon their lands" (Gibbs 1855a:433). Presumably it was against raiding parties from this tribe -- perhaps among others -- that these village protections were constructed.

What little fighting was carried out by the upriver Upper Skagit was, according to tradition, mainly against the Thompson (Collins 1974a:115, 118-119). In fact, in early days their relations with the Thompson were apparently almost uniformly hostile. They considered the Thompson to be "different in language . . ., physical type, and culture from other visitors to Upper Skagit territory" (Collins 1974a:15), an appraisal which is essentially accurate on all scores (see, e.g., Thompson 1979:693; Boas 1895:547-549; Teit 1900).

There is nothing in the literature to suggest Upper Skagit raids into Thompson territory above the international border and no information, so far as I am aware, to indicate frequent early Thompson raids down into that part of the Skagit Valley where even the uppermost of the Upper Skagit settlements were situated. Rather the hostile encounters appear to have occurred in the upper Skagit River region -- i.e., in the Ross Lake area and the mountainous country within the Park limits -- territory that was utilized by both Upper Skagit and Thompson in their subsistence quest. The Thompson came into this area, country that the Upper Skagit considered to be theirs, in winter and sometimes remained through the spring and summer to hunt and fish. It was during these periods that the Thompson, encountering Upper Skagit in the forest, killed the men and women, and captured the babies for slaves. (Collins 1974a:15)

Owing to our special interest in this group as rival claimants to the upper Skagit Valley (see Thompson chapter) and rival hunters in that region, the following summary by Collins (1974a:118-119) deserves quoting:

The major enemy with whom the Upper Skagit were concerned was not . . . saltwater peoples, but the Thompson, . . . The northern headwaters of the Skagit River were in British Columbia not far from the Fraser River and its junction with the Thompson. The Thompson came south along the river, traveling on land.

The Upper Skagit . . . feared the Thompson and regarded them as strange in a very different way from other Plateau peoples like the Methow, the Okanogan, or the Wenatchee. The Thompson earned this reputation by camping in the woods in a secretive way and not openly approaching the nearest Upper Skagit village and identifying themselves. They captured a few Upper Skagit as slaves and killed some others by torturing them, that is, running the living person through with a pointed stick and then placing him upright as though he was sitting on the ground for his relatives to find.

The Upper Skagit tell of many warlike encounters with the Thompson. Finally they made friends with them. There were some intermarriages with them, and some Upper Skagit at the time of writing [date unclear but between 1942 and 1974] are part Thompson.

Again the role of the Upper Skagit in this relationship was that of defenders of their own villages. They retaliated when forced to do so for their own protection. Some Upper Skagit still fear the Thompson and feel bitterness for children captured by them.

Among the Plateau tribes known to them, the Upper Skagit fought only with the Thompson (Collins 1974a:13). On the other hand, up in this same back country, Skagit hunting parties sometimes came upon Chilliwack hunting groups of the middle Fraser country, with hostilities as a consequence (Duff 1952:21). One may assume that these chance encounters took place on the northwestern side of the upper Skagit Valley or in the uppermost Nooksack drainage, in either case in territory assigned by Teit (1900:166) to the Lower Thompson (Figure 3-1).

The Upper Skagit had no memory of hostilities among their own settlements (Collins 1974a:13). Relations within the tribe are considered in the sociopolitical sections that follow.

Kinship Bonds

According to Collins (1974a:83), each large house was occupied by related families and, I gather, comprised a single household. Whether this was invariably true -- after all some houses measured as much as 120 by 40 feet (Collins 1974a:19) -- and precisely how the families interlocked in daily operational terms to form an integrated "household" is not made clear.

Residence following marriage was determined by the husband, who generally chose to continue to live with his father's household, thus continuing to hunt and fish with his own kinsmen. But when he considered it to his own advantage, he agreed, sometimes under pressure and persuasion of his wife's family, to join her household. Residence, however, tended to be impermanent: e.g., it changed with deaths and remarriages and sometimes with persons visiting relatives for long periods as parents visiting children and vice versa. (Collins 1974a:83-84)

Kinship was at the very core of the social system. Relatives, no matter how distant and whether by birth or marriage, were basically friends while non-kinsmen were ipso facto outside the group -- suspected if not actual enemies with whom one competed in potlatching and of whom one thought the worst. As Collins (1974a:86) states: "The principal basic features of the [kinship] terminological system are that it is descriptive and bilateral and distinguishes relatives, according to generation, sex, age, and whether the intervening relative is dead or alive." (Collins 1974a:84-90)

In functional terms the central kin unit was that consisting of adult siblings: it was they who formed the core of the household. It was typically a male group of this kind that jointly erected and owned a dwelling, built weirs and set snares, sponsored potlatches, and cooperated in winter dances. These were joined by their sisters' husbands if resident with them and by their own male siblings who might be living in other households or even villages. The female side cooperated in a similar fashion, as in berry-picking, though because of predominant patrilocality sisters after marriage were less likely to be members of the same household and so in a position allowing the perpetuation of a close relationship. However, because sisters often married into the same household or village, close cooperation of married sisters was by no means unknown. Close cooperation between brothers and sisters likewise continued following marriage to the extent permitted by the situation. (Collins 1974a:90-91)

In Upper Skagit society relations between close kinsmen were fairly well formalized.

(a) Parents were responsible for providing their children with the essentials of life and for disciplining and training them in the ordinary patterns of their society and in such special skills as they themselves possessed. Children were expected to respect their parents, to avoid joking with them, and to solicit and consider seriously their advice even when they had attained adulthood and had begun families of their own. Bonds between parents and children were very close and affectionate, particularly between those of the same sex. (Collins 1974a:92-94)

(b) Throughout life older siblings exercised authority over and assumed certain responsibilities of care and instruction for younger siblings of the same sex. And male siblings to some degree over female siblings. (Collins 1974a:90-92)

(c) The brother and sisters and the cousins of a man and his wife were expected to take great interest in the latters' children -- their "nephews" and "nieces" according to the Upper Skagit kinship structure: "to train them, to discipline them (although the parents might not like this [for they were inclined to be especially strict]), and to assist them at ceremonies of their life cycle." These responsibilities increased if a parent of the children died. Nephews and nieces visited their uncles and aunts, were respectful to them, obeyed them, and eschewed joking with them as children. (Collins 1974a:94-96)

(d) Grandparents and grandchildren maintained a very warm and close relationship. The latter could speak more familiarly with them than with their own parents, even joke with them. Should a parent die, the children were most often brought up by the grandparents, frequently in the same house with the remaining parent but as a member of different family group. (Collins 1974a:96-97)

(e) The relationship between husband and wife, unless they were tied by earlier affinal links, was an uneasy one, at least until a union had endured for some years. Having come from unrelated families and generally different villages, they felt themselves as strangers and even feared the guardian spirits of their new partner. Their primary loyalties remained with their own individual kin groups. Often the wife sorely missed her relatives and earlier acquaintances. One's rights and obligations to one's natal family frequently conflicted with those to one's spouse. It is not surprising that a married couple for a time acted toward one another with something other than trust and real attachment.

Husband and wife shared the burdens of maintaining the new family, but the products of one's labors belonged to that person alone. Neither spouse inherited from the other. (Collins 1974a:106-107)

As Collins (1974a:107) reports, "parents-in-law were regarded as substitutes for the parents," especially in the case of women, who lived with their in-laws more frequently than did men. A woman was expected to show them the same respect and obedience as her own parents: she sometimes invited them to eat with her own family; she exchanged gifts with her mother-in-law. Some friction occasionally developed between her and her mother-in-law, but less often then with sisters-in-law. (Collins 1974a:107-108)

It was much the same for a man living with his wife in her parental home. He gave part of his game to his in-laws, and cooperated in the food quest with his father-in-law. The father-in-law, in turn, was expected to have the economic interests of his daughter's husband in mind: to teach him techniques, provide canoes and tools, even help him develop his skills as a shaman. (Collins 1974a:108)

Siblings-in-law were expected to treat one another like brothers and sisters. A man was in many ways dependent on the assistance of his wife's brothers: he often visited their village to secure food as their special seasonal resources became available or, if he lived in their village, relied on them to guide him in local hunting. A woman worked closely, as in berrying, with her brothers' wives before her marriage and, because of the usual patriolocality, with her husband's unmarried sisters after marriage. (Collins 1974a:108-109)

Siblings-in-law of the same sex were supposed to maintain a friendly, cooperative relationship in which joking was allowed. Often, however, tensions of rivalry and jealousy developed between them. As noted above, siblings-in-law of the opposite sex were potential mates as, when the connecting relative died, through the levirate or sororate. Theoretically they treated one another like brother and sister, but, as possible spouses, their behavior toward one another was inclined to be more constrained. (Collins 1974a:109)

The parents of a married couple showed respect to each other, exchanged gifts, and worked together to help their married children. (Collins 1974a:109-110)

Property and Inheritance

Individual ownership of land seems to have been limited to that by certain women of the root plots at Sauk Prairie and in the lower Skagit Valley to which reference has already been made. With this exception the country in the vicinity of each village cluster -- the fisheries, mountains, berry-grounds, and meadows -- was common property of the villagers. These rights extended from the Skagit River up to the Skagit-Nooksack watershed on the north, to the Skagit-Stillaguamish divide in that direction, and to the Cascades on the east. But they became progressively weaker as these divides were approached.

Nevertheless, Upper Skagit did not expect to find unrelated Nooksack, Stillaguamish, Thompson, Methow, or Okanogan on their sides of the divides. Persons who wished to hunt and fish in the locales of neighboring villages were expected to visit their relatives there first to announce their presence, their intentions, and to be given tacit permission. This was never withheld. Part of the objection to the Thompson incursions into Skagit territory was that they did not behave in these accepted routines. (Collins 1974a:80-81)

The key to the character of the Upper Skagit subsistence round is hinted at in the above quotation. Different parts of the tribal territory had their special food resources or varied in the abundance of the same resources. These were "owned" by the village clusters nearby. However localized in their occurrence, the foods came in fact to be widely distributed through the tribe by individuals and families visiting relatives as the latter's food resources became seasonally available and exploiting these sources with their kinmen owners. Under these conditions, the worth of large extended families and of village cluster outmarriage becomes at once apparent. Reciprocity was the key element in this arrangement.

Just as a person had the right through inheritance or by marriage to secure his food and necessary natural resources in the territory considered "owned" by his village, so he had the right on these same grounds to live in a particular dwelling. This right could not be sold or given away.

Individuals and families varied in their wealth, but, save for whatever differences in resource availability might come as the result of linkages with more or fewer kinfolk and belonging to different village clusters as mentioned above, each person and family had approximately the same opportunities to become wealthy. Inheritance of property was of minimal importance. Industry was highly esteemed and in general wealth was related to hard work and, in part, to the development of specialist skills. Abstemiousness in food consumption, care in using material objects, and downright hoarding were all culturally encouraged, partly because these traits led to good marriages and promoted the accumulation of property that could be manipulated in gift exchanges to augment family prestige. Certain guardian spirits, however, were thought to convey power to acquire material possessions and others to grant power -- as in fishing or craft skills -- that more indirectly led to wealth.

Slaves were very few and their effect on wealth and the economy was negligible. Multiple wives, on the other hand, were common and an advantage to a man because they produced many more baskets, blankets, and mats and so allowed the giving of potlatches which raised the status of the family and its relatives. (Collins 1974a:79-81)

Property passed from family to family in several ways. Visitors brought gifts for their kinsmen and in return received presents of equal value when the former became hosts to the latter. These formalized presents tended to be the specialities of the giver's village:

People from upriver villages brought prepared skins, dried venison, mountain-goat wool blankets, and coiled baskets as gifts. In return those on salt water might give smoked clams, oysters, saltwater fish, and cattail matting. (Collins 1974a:77-78)

Formal gift exchanges occurred between parents-in-law and between them and the newly married couple. Potlatches distributed manufactured articles, raw materials, and prepared food. In the midwinter guardian spirit ceremony, sponsors and guests exchanged gifts. At death most of the property of the deceased -- heirloom-type objects were held back -- was taken by more distant relatives who hastened to the house and appropriated what they especially wanted; they were expected, however, to give the bereaved family something in return.

Through the operation of these exchange patterns, property was kept "circulating from family to family. Portable, durable items like baskets circulated widely, reaching villages many miles from their origin." There were no trading centers in Upper Skagit territory. On the other hand, people sometimes journeyed to the mouth of the Fraser River where there were "frequent opportunities for [property] exchange." (Collins 1974a:77-80)

Social Classes

Among the Upper Skagit there were two classes: a wealthy upper class and a poor lower class. In addition there were a few slaves, who were considered to be outside the class structure. (Collins 1974a:123)

The upper class comprised people whose parents had been upper class and most of whose relatives were of the upper class group. When they lacked the expected wealth, this was attributed to their lack of the proper guardian spirits. The inheritance of property was an insignificant element since virtually all valuables of the deceased were given away; certain rights to natural resources, however, were inherited, as already noted. Many members of the class, especially if they were the oldest children in their family, had "honored names," acquired from an ancestor, particularly a grandparent, because of some physical or aptitude similarity. Its acquisition had to be publically celebrated either by or in a gift-exchange potlatch and witnessed and certified to by those in attendance. Upper class persons had to act as a potlatch sponsor at least once in their life. A few held slaves, a right limited to the upper class. (Collins 1974a:123-126; Suttles 1957:174)

Upper class people also strictly observed the recognized code of etiquette, as in eating in small bites, in being quiet, dignified, and unaggressive in public, in being truthful and even-tempered, in not engaging in loud quarrels or fighting with other family members, and so on. They were expected without fail to return a gift with one of greater value. (Collins 1974a:123-126)

Lower class people, it is said, starved in the winter through lack of stored food and "begged for dried fish from people who had them" (Collins 1974a:123). Even when, as a result of industry and daring -- the consequence of possessing the proper guardian spirits -- persons of this class accumulated wealth, they never fully mastered the behavioral patterns, the good breeding, required of upper class people. (Collins 1974a:123-126)

By and large, the Upper Skagit disapproved of slavery. "Everyone was the same," according to Fornsby (in Collins 1949:303), and so it must have seemed to one who, though Upper Skagit born, had lived his life with the Lower Skagit and other nearby coastal groups, where class distinctions were stronger than with the Upper people and slaves far more numerous. The few slaves of the Upper folk were obtained, Collins reports, either by purchase or capture; according to Fornsby, however, the Upper Skagit never went raiding for slaves. Surely Fornsby comes close to being correct: Upper Skagit parties did not often set out for slaves. In any event, they were supposed to come from distant groups to reduce the possibility of escape to return home.

Slaves were not supposed to marry free persons, though it occasionally occurred much against the will of the free person's relatives. Children of such marriages, though not slaves, had to live with the disgrace of their slave ancestry and could not take "honored names" or sponsor potlatches. Slaves were allowed to marry other slaves, their children also being slaves. They were sometimes able to accumulate property, but their behavior, it was felt, remained that of slaves.

Many village clusters (Collins' "extended villages") had one or two dwellings that "were regarded as 'low class,'" the others being 'high class.' Their origin is in doubt: perhaps they were established by men of partial slave descent or possibly they represented village expansions into locations less desirable for securing food. This information regarding their limited assessibility to subsistence resources makes it reasonably clear that Collins means that one or two villages within the larger village cluster were low class villages, not that one or two houses of a larger number in a single village were low class in their ranking. (Collins 1974a:129)

Concerning these class distinctions and possible Plateau relationships, Collins (1974a:129) states:

This basic equality of the members of the society and the ordering of social relations by kinship and household groupings is characteristic of their nearest Plateau neighbors. It may reflect an old type of society found among Salish speakers, which at one time may have been characteristic of much of the inland and coastal area. The place names along the Skagit River in Interior Salish rather than Coast Salish . . . also suggest this possibility.

In addition to these two class and the slave distinctions, the Upper Skagit also recognized status differences between individuals within the same class. Certain persons, because of their higher level of achievement as, for example, having hosted a potlatch, were seated at ceremonies in the more prestigious places, received gifts first in potlatches, and were served food first at meals. But these were not firm, locked-in positions of prestige. It was up to the hosts on these occasions to establish their own order of precedence. These might differ somewhat according to the host and through time. (Collins 1974a:130)

Potlatches

While serving several less direct functions, the potlatch had as its principal objectives the validation of membership in the upper class and improvement of the social status of their sponsors and their families. They were also given to mark special events, such as the bestowing of an "honored name" and the passing of significant points in a person's life cycle: e.g., the termination of a girl's seclusion following her first menses to announce her marriageability and increase her eligibility, a wedding, a woman's first visit home after marriage, and a funeral. The most common of such occasions, however, was as a memorial to a dead ancestor, particularly on the reburial of the bones, the final tribute to the person's memory. They were not often given when one was named, the naming ceremony being commonly incorporated into a potlatch given for some other purpose.

Potlatches were held in the late-summer or early fall when the chum salmon were running. They possessed a strong competitive aspect, especially overt in those in which guests vied physically for blankets and other valuables thrown to them. But they were also happy, gay affairs, for they included gambling games and athletic contests like canoe and foot racing and wrestling. During them all persons were supposed to put aside the enmities which were certain to exist since guests came from various tribes. To give vent to these inappropriate feelings at the outset and promote amity during the rest of the occasion, the host customarily arranged a contest between guests already at the place as one team and each newly arriving group in turn as the other team. Persons who were pulled over a plank marker were expected to give gifts to the winners. Though this game sometimes became rough, all were supposed to put aside their differences during the potlatch ceremonies that followed. (Collins 1974a:131-137, 141)

Both sexes sponsored potlatches, but both a man and his wife as well as their relatives were intimately involved in a cooperative fashion in the event. The planning of such an occasion required as long as two years, during which time all gathered materials and made objects to be distributed and accumulated food for the feasts. When the ceremony was held, each participating family group came dancing into the potlatch structure, bringing their contributions. As the master of ceremonies distributed these materials, he named the contributor as he called the name of the person to receive the object. In this distribution upper class people could not be slighted; lower class people sometimes were. Receiving a gift at a potlatch erased any hurt feeling that a person might have acquired against the donor on some earlier occasion, but it also entailed a return gift of equal or greater value at some future time. (Collins 1974a:137-139)

People felt proud to participate in these ceremonies, even if only distant relatives of the sponsor, and to give objects away in them. In actual fact the Upper Skagit had little to give in traditional times, principally baskets, a few mountain goat and dog wool blankets, possibly a canoe or two, cedar bark, and unworked cedar and spruce roots for basketry. Furthermore, there was very little wasting of things and no intentional property destruction as farther north on the coast; the Upper Skagit simply had too little of everything. (Collins 1974a:137-139,141)

Additional costs were involved in payments to the master of ceremonies, "who had a special guardian spirit in order to speak publicly," the male and female cooks, and the men who drummed with sticks in accompaniment to the singing and dancing.

Another sort of property distribution accompanied both the potlatch and the guardian spirit ceremonial. In this instance only men were participants as donors and recipients. The gifts were thrown from a house roof or from a plank platform across two floating canoes onto the shore or bank or into the shallow water. The property was scrambled for.

Still another form of gift giving occurred when a young man wished to improve his social status and enter the potlatch system without the expense of a full potlatch. Selecting a wealthy old man at the time of a potlatch, he woke him in the middle of the night and piled gifts on and around him, while the man taunted him to make him even more generous. Under pain of lowering his own status, the man was obligated to return to the young man property of still greater value at a later time. Awakened, the potlatch visitors enjoyed the scene and devoted the remainder of the night to singing. Women had a similar gift exchange, which, however, was apparently not part of the potlatch ceremony. (Collins 1974a:140-141)

Among the Upper Skagit hilarious "joking" potlatches were also held. In these the boastful speeches of the serious potlatch were mimicked and broken articles and spoiled foods were given away. Sometimes in these the women potlatched the men. (Collins 1974a:141-142)

The social and economic importance of potlatching was considerable, especially since, as noted above, property was not willfully destroyed. It resulted in a continual flow of articles and food through the population. While more and better goods went to the wealthy, the "scramble" and other distribution occurrences brought things and foods to people of lower status and class. As Collins (1974a:142) observes: "People who did not hunt in the mountains still could get blankets of mountain goat wool; women who did not know how to make the coiled baskets still received them." At the same time it developed in the people a lack of attachment to a particular object, for it was realized that sometime it might have to be given away. Similarly potlatch feasts were greatly anticipated: they brought an abundance of food and various kinds of food to people who ordinarily could not enjoy these luxuries.

Potlatches encouraged industry on the part of persons and families and stimulated the manufacture of articles of quality. They brought together people from distant villages and different tribes, even from enemy groups under a kind of safe conduct understanding, thus broadening the cultural and social horizon of all of the participating groups. Owing to their geographical isolation, this must have been of particular importance to the Upper Skagit. Through sponsoring potlatches, persons and families became well-known by name far beyond their own village and tribe. And potlatches provided a welcome period of social interaction: at the close of the event a guardian spirit dance was commonly held in which the host and his family members sang their wealth and warrior songs, these spirits, unlike others, being summonable during this warm season. (Collins 1974a:134, 142-143)

Intragroup Relations

Life within the community was "relatively peaceful" in traditional times. A pattern of consideration for and cooperation with others and of mild and inoffensive behavior was inculcated in the young. Most offenses that occurred were handled by the individual or immediate family, if retaliation in any form occurred, without involving the larger community. In the case of serious injury or death to another, payment of an acceptable amount of "blood compensation" was generally agreed to by the heads of the two households involved and the perpetrator was compelled to undergo a period of seclusion in the forest, fasting, and bathing. Failing this peaceable resolution, a blood feud erupted. Similarly, smaller property payments erased guilt in lesser quarrels and offenses.

Every effort was made to avoid the inception of a blood feud. Because of widespread and complex interlocking consanguinous and affinal relationships, all people had divided loyalties. Feuds also severely interfered with the holding of religious ceremonies and potlatches and the pursuit of the subsistence quest, this latter requiring much seasonal travel along the river to fisheries and to root and berry grounds. Even after a conflict had been technically settled by proper payments, people feared to travel widely because of residual ill-will. (Collins 1974a:119-122)

One and perhaps two offenses, however, were regarded as so serious as not to be resolvable at the individual or single family level. This was sometimes so in the very rare case where a shaman was thought to have caused the death of several people through his supernatural powers. Members of the aggrieved families collaborated in ambushing him and striking him fatally at the same time to distribute the onus. Those involved in exacting this revenge paid no "blood money." Since ill-will lasting for generations was created by killing a shaman in this way, a more usual method of securing justice was by a family member or hired shaman bringing about the death of the offending shaman by supernatural means. (Collins 1974a:119-120)

Incest -- sexual relations between persons with any known consanguineal bonds -- was likewise considered an especially reprehensible crime, since such a union was thought to be barren or to produce children who would die young. Relatives of both the offending man and woman joined in killing the man. In any event, incest was a rare occurrence. (Collins 1974a:119-120)

Economic Structure: Division of Labor and Specialization

Tasks among the Upper Skagit were divided along two dimensions: between the sexes and between specialists, both male and female. Men hunted and fished, worked wood, stone, and bone, made the dwellings and canoes, and were the warriors and most shamans. Women dug and gathered the plant foods, "assisted in fishing and the drying or smoking of foods, prepared skins, made baskets, wove, and carried wood and water. They cared for children, cooked food, and kept house." (Collins 1974a:75) Rather informally, women also cured by supernatural power and recovered lost souls. This labor division, was, however, not entirely rigid. Men sometimes cooked, as on ceremonial occasions and when out hunting; helped the women of their family in gathering berries in the mountains, in carrying wood and water when women were busy, and even to some extent tended children. Similarly, women occasionally aided in canoe fishing, even to manning a canoe in a netting operation. Collins was told of one women who had been a famous hunter and another who had made canoes.

No specialists applied their skills on a full-time basis, but the amount of time devoted to them was proportional to the importance placed on the tasks by their fellow villagers, others having to discharge their normal daily tasks -- like acquiring food -- while they were otherwise occupied. Among the male specialists were shamans who cured and led the ceremonials; carpenters who made canoes and supervised the building of dwellings; men who directed the construction of the larger fish weirs; and warriors (a small group). Women specialists included those with particular aptitudes in weaving, coiling baskets, serving as midwives, and recapturing the souls of infants.

Specialists were hired for a fee by those requiring their particular services. Persons who performed occasional duties like those of cooking, making canoes and dwellings, drumming, leading songs at ceremonies, diagnosing and curing illnesses, and performing certain tasks at funerals were also paid. Payment was usually in goods -- especially baskets, mats, and blankets -- since the Upper Skagit had very little shell "money."

"Within the joint household, a head woman organized the labor of all the women." One might tend the children and another make baskets or prepare the food while most were out securing food. (Collins 1974a:75-77, 80)

Pregnancy and Childbirth

Families looked forward to many children, possibly partly because of the high mortality among them. Nevertheless, they attempted to space births about two years apart: more frequent births were considered harmful to both the mother and the new child. Barren women were pitied, and were believed to have somehow displeased a supernatural being. Evidently abortions were occasionally brought about, for women herbalists were said to be able to produce them. (Collins 1974a:215)

Women with illegitimate children married, but less advantageously than otherwise. Their children were tended as carefully as other children, but were disadvantaged as they grew up by lacking the social benefits and status of other children whose fathers had given substantial property to their wife's family on marriage, a large wedding potlatch, and could receive an "honored name." (Collins 1974a:215-216)

There was no general preference for children of one sex over the other. Where such existed in a particular case, an attempt was made to control the sex of a birth by placing sex-associated objects (e.g., a bow and arrow or basketry) under the sleeping platform. (Collins 1974a:216)

A pregnant woman observed certain restrictions, some magical and others credible from our point of view. Among the former were taboos against eating fresh salmon and the practice of undressing by removing clothing downward. The latter included watching her weight, walking a good bit, and remaining active. She was also expected to bathe twice daily in cold river or lake water. (Collins 1974a:216-217)

Birth was in the winter house or summer camp, apparently generally not in a specially constructed shelter. If no temporary partitions were erected in the dwelling to insure privacy during the delivery, children were sent to a nearby house. The woman was attended by a midwife, either her mother or another of her relatives who was often a specialist in aiding births. The father gathered large quantities of firewood to insure that she was kept warm. The birth was accomplished in an erect position over a mat, the woman holding onto a house post. In cases of prolonged or difficult births, a shaman with special guardian spirit power was summoned. (Collins 1974a:217)

The birth of twins was regarded as a disaster, a sign of supernatural disapproval. The mother was thought to become "like a wild animal" and the father lost his guardian spirits. The twins were exposed to die, sometimes placed in trees near the house. The parents had to purify themselves by bathing twice each day and scrape their bodies with boughs. (Collins 1974a:217-218)

Postpartum practices were numerous. The infant was rubbed with grease by the midwife. The afterbirth was safely buried and the umbilical cord preserved, often secreted in a tree hollow. For a few days the infant was sometimes fed juice from salmon or cooked grouse. For about two weeks the mother rested and bathed in cold water, partly to aid in healing her birth injuries and partly to remove possible supernatural dangers to her husband and children. The husband swam morning and night for the next six months. (Collins 1974a:218)

The infant was kept in a board cradle -- less frequently a basketry cradle -- for two years, with cedar bark padding between it and the board. Its legs and arms were fastened to the board to keep them straight; its head was protected by a cloth or mat thrown over a half hoop attached to the top of the board; its forehead -- perhaps only in the case of upper-class children -- was gradually flattened as a mark of social status and a feature of beauty by a padded board, buckskin strip, or bag of sand pressing against it. It was diapered with soft, shredded cedar bark and wrapped in mink skins, soft fur inside, before being fastened on the board. The cradle was suspended by two straps from a tree branch when the mother worked outside. In the house it was hung from a bough brought inside and inserted in the ground upright and swung back and forth with a cord to rock the infant. In traveling, the board was carried on the mother's back, suspended from a forehead tumpline. (Collins 1974a:218-219)

Babies were nursed at their own demand. The mother increased her milk supply by squatting, wrapped in a skin or mat, over hot rocks upon which water had been poured to steam her breasts. Infants were removed from the board for cleansing twice daily and as it grew older allowed to be free of it when awake for increasingly long periods. When they began to toddle, the board was abandoned, but the child retained great affection for it, sometimes wanting to be tied to it when sleepy and dragging it around with him when awake. (Collins 1974a:219-220)

Naming

Naming ceremonies could be held in infancy, the time depending "on the type of name and the wealth and social ambitions of the parents." Children without special names were called by nicknames, many alluding to physical appearances. "In daily speech, people most commonly referred to each other and addressed each other by kinship terms." (Collins 1974a:220)

Special names were of two types, guardian spirit names and honored names. Both were respected, the former relating to the religious life and the latter to the social class and rank system. A parent sometimes gave the name of one of his guardian spirits to his child in the hope that the spirit would protect the child; this was a highly secret name never used in daily speech. As an adult, a person might even take as his name the name of one of his own guardian spirits. These names were not sex-specific. In Collins' view, this spirit-based naming system was related to that found in the Plateau. (Collins 1974a:220, 222)

The "honored name" was an upper class cognomen of very high prestige customarily given by a grand parent to a child or even an adult after consultation with the kin of his generation and the parents. If it was his own name rather than that of a deceased relative, he changed his name since it had become the property of its new owner. The particular name chosen in such cases reflected the person's similarities in "personality, talents, and intelligence" to the ancestor whose name he inherited. A potlatch for this purpose or a special smaller ceremony in a potlatch given for some other purpose was held to announce and validate the naming. This was a "happy, gay affair." Honored names, most untranslatable, were either male or female names unlike those relating to the guardian spirits. This naming system was, Collins reports, related to coastal culture. (Collins 1974a:220-222)

Childhood

Until four or five years of age, children were usually treated with relative leniency, especially by grand parents, and fed when hungry. They were toilet-trained once they could walk. Within an extended family in a single dwelling one woman, especially fond of small children, often took care of all the family's nonnursing offspring to free their mothers for other tasks. However, children were frequently taken by their mothers when they went out to pick berries or dig roots. (Collins 1974a:222)

From the age of four or five years to puberty children followed a different regimen, relatively strict in some ways. Each morning they were awakened by an adult "trainer," whipped four times as they lay on their mats, and sent to the river or lake to bathe before they could warm themselves before the fire. This was required even when on the upper reaches of the Skagit River in the mountains where there was considerable snow and ice in winter. To avoid their children's hostility, parents preferred to have a "trainer" perform this unpleasant task for them. This practice had its associations with the guardian spirit world but was also one aspect of the procedure for developing physical hardiness. (Collins 1974a:222-223)

Both boys and girls were encouraged to fast, again partly because this was thought to attract a guardian spirit and partly because it trained boys to be active (e.g., in hunting) later in life even when food was scarce and girls to be careful of food so a surplus could be accumulated to make her family after marriage wealthy. (Collins 1974a:223)

Children from about four years on were trained to acquire guardian spirits, as noted below. At night they were told the myths and traditions of the tribe. (Collins 1974a:212-213, 223, 224)

When approximately eight years of age children began to receive regular formal advice concerning how to treat old people and other appropriate adult behavior from parents and other older kinsmen. Girls spent their time with their younger siblings and their female family elders. A girl "took care of younger brothers or sisters, dug roots, carried water, and learned to do household tasks." Boys were trained by their fathers and older male relations, the father's brothers more often than the mother's because of the tendency toward patrilocal residence. As soon as they could understand, they were taken fishing: taught how "to handle a canoe, [to] make nets, fishhooks, and lines; and [to] do all the other tasks necessary to the varied fishing operations. Both boys and girls were taught to respect fish, especially salmon, not to make fun of them or wantonly to destroy them when they were running." (Collins 1974a:224)

When five or six, a boy was given a small bow and arrows and his first small kill was made the occasion for rejoicing and much public praise. When it was dangerous or otherwise inappropriate to accompany fathers away from the village, small boys stayed home with the women, older people, and girls, often playing in gangs sometimes to the point of being mischievous and even destructive. Small girls enjoyed playing together with imitation houses, house furnishings, and people. (Collins 1974a:224-225)

Puberty

Puberty marked a great change in a girl's life. During her ritual she lived for six weeks alone in a special, small, well-built house constructed exclusively for her use. She observed many regulations, supervised by her mother or some other older female relative. She was considered dangerous, especially during the first five days of her menses and then again after the next four days had passed -- her "dark days." During these "dark" periods only her mother was supposed to see her: she wore a hood over her face and had her face painted red. During the other days of her seclusion -- her "light days" -- she was visited by her female relatives on both her mother's and father's side of the family, who made a point of giving her behavioral advice and sometimes presents. (Collins 1974a:225-226)

Once daily the girl was brought food by her mother. Both in the morning and in the evening she was led by her mother to water where she dived four times and bathed, rubbing her arms and legs with a long willow wand so she would have a long life. Back in her special house, she had her hair combed by her mother, since she was forbidden to comb it herself. Nor was she allowed to scratch her head except with a scratching stick. She expectorated only in baskets that she had made, containers that were thrown away. And she was given tasks like making basketry. All these duties and prohibitions were designed to insure that in this condition she could not harm herself and anyone else and that in later adult days she would lead a safe, long, and productive life. (Collins 1974a:226-227)

At the end of this period of seclusion, the girl returned home, though she had to go back to her special house during her next three menses to fast and bathe. This sequence of periods of isolation behind her, she was given a dinner by her family and was regarded as eligible for marriage, although this was normally delayed for some years. During this interval she learned under her mother's tutelage to cook and perform the other tasks of the Upper Skagit woman. She was kept under the watchful eyes of her mother and other female relatives; she no longer, for example, went out on spirit quests. She was not, however, secluded in a compartment within the house as was the practice among coastal groups to the north. (Collins 1974a:227-228)

Boys underwent no puberty ceremonial. Their guardian spirit searches intensified and they took over a greater share of the adult work of males. (Collins 1974a:228)

Some irresponsible behavior was anticipated of adolescent boys and girls, which explains in part why a newly married couple remained under the authority of their parents until they had their first child. (Collins 1974a:228-229)

Marriage

Most men and all women married in the normal course of events. Marriage between persons of any "blood" relationship was prohibited and punished: i.e., in Upper Skagit terms between people out to and including three degrees of consanguinity. This regulation, as well as the economic and social advantages of developing cooperative kinship links with persons of different village clusters, was responsible for what amounted to a village outmarriage pattern. On the other hand, intermarriage into villages in which relatives already lived was considered most desirable, for the in-marrying woman -- it was generally the wife who left home on marriage -- thus had kinsmen to guide and take her part. It was to guarantee the presence of a close relative, affinal in this instance, that the best marriage arrangement was thought to involve a woman's union with the brother of her brother's wife: i.e., a brother-sister exchange arrangement. Parents also preferred to have children marry into not too distant villages so that they could see them at least yearly, and into villages with cultural patterns, including foods, similar to their own. (Collins 1974a:97-99, 106)

The Upper Skagit maintained marriage links with the Nooksack as well as with other nearby coastal groups of no direct relevance to this study. Given the considerations outlined in the preceding paragraph, it is understandable why fewer marriage unions took place with Plateau people than with coastal folk. When, however, Plateau marriages occurred, the man was apparently normally from the east, though Collins fails to make this point. According to her, however, the Plateau man ordinarily moved to Upper Skagit country rather than the wife crossing the Cascades. (Collins 1974a:98)

First marriages of the more structured and honored variety were arranged by older relatives, the wishes of the young person being rarely considered. The father played the formal role or, if he was dead, an older brother or an uncle. It was especially important for the partner to be of the same class or higher if possible, wealthy, diligent, strong, and healthy owing to the rigors of travel and the food quest, intelligent, and of good nature. Relative ages were a minor consideration.

Prior to the marriage no courtship occurred. In fact, the couple might never have seen one another. Most marriages involved persons who had been previously married, in part the consequence of great pressures on a woman to marry a deceased sister's widowed husband and on a man to do the same with a deceased brother's wife. The rationale for these "sororate" and levirate marriages was that they guaranteed that children would have stepparents that were related to them and who consequently might be expected to treat them better than might otherwise be the case.

The marriage discussions could be quite relaxed or could have their formal elements, sometimes pursued only in a ceremonial fashion after the marriage had already been agreed upon. As an example of a formal pattern, a boy brought firewood bark to the girl's house on four occasions, to have it thrown out by her father the first three times. If he allowed it to remain inside on the fourth occasion, the boy's suit was accepted. Or the young man brought property to the entrance of the girl's dwelling, but was barred access by her father until he had returned one or more times with additional gifts. In any event to legalize the union, a payment was made to the bride's father or another of her male relatives, the first of a long series of property exchanges. (Collins 1974a:229-230)

A wedding ceremony was usually held: sometimes a small wedding dinner, sometimes a ceremony made part of a potlatch scheduled for some other purpose, and rarely a full-fledged potlatch with an expensive feast. "These events were financed by the families of both the bride and the groom," but were celebrated at the bride's home to which the groom's group came in canoes. On this occasion the couple received lectures on the responsibilities of marriage and all enjoyed the good and abundant food. Marriages were legalized by gifts of property by the husband and his relatives to the bride's family. The value of the marriage gifts influenced the social position of the children. (Collins 1974a:104, 230)

Marriages that were not arranged, at least not formally, occasionally occurred. Marriage dances attended by young unmarried men and women with their relations were held: in these a man sang and danced around a room and as he passed the girl he favored put a feather in her hair. Leaving it in place on the fifth occasion indicated that she had accepted his marriage offer. Collins (1974a:102) believes that this dance custom, which was poorly integrated into Upper Skagit marriage patterns, "may have been an introduction during the . . . [1880s] from the Plateau." Also as a very rare occurrence a woman wishing to marry an already married man went with some female relatives to the man's village and fought with his wife. If she was victorious, the man "was expected to marry the victor" (Collins 1974a:102-103). Such forthright action was considered, however, most unfeminine, for women were supposed to be courteous and gentle. Whether following such marriage arrangements a formal wedding of any kind took place is not stated. (Collins 1974a:99-103, 229)

Post-marriage residence was the choice of the husband, but was generally virilocal: i.e., with his family. This meant that usually the wife had to adjust to a new life with new companions, who in general might be expected to favor her husband. However, because of the prevailing pattern of marriage exchange, she could normally anticipate finding some "blood" kin in her new community. Furthermore, she could properly expect to make frequent and sometimes long visits back to her own family. Until the birth of their first child -- or until they had given up hope of offspring -- the couple were under the authority of the parents with whom they were residing. (Collins 1974a:230-231)

At the end of the first year, if all went well, the couple, with their first-born and some of the husband's relatives, visited the wife's parents and received gifts equivalent to the earlier bride price with interest. This visit was an altogether happy occasion. Still people continued through life to feel most comfortable with and to have the greatest attachment to their own "blood" kinsmen and to the degree possible continued to carry on many activities with them.

Most marriages were monogamous. Multiple wives, however, were permitted. Where the man was a great hunter or entered such unions to care as stepfather for a deceased brother's children or to provide his own children with a closely related stepmother, as in the sororate, more than one wife was regarded as especially honorable. Cases of as many as ten simultaneous wives, where husbands were upper class and wealthy, were remembered by Collins' informants. (Collins 1974a:103-104)

Divorce was rare in traditional times, partly because people wished to avoid leaving children with a stepparent, partly because no one wanted to return property acquired during the marriage, and partly in the case of the wife because she was afraid to travel alone back to the village of her family. It was certainly easier for a man to leave his wife than the reverse. "Acceptable grounds for divorce were barrenness of the wife, physical cruelty, or adultery." (Collins 1974a:104-106)

Neither male nor female transvestites were known among the Upper Skagit. (Collins 1974a:106)

The middle years of adult life were vigorous and productive and ideally brought respect and social status. It was the time of the guardian spirit return and of giving potlatches, if a family was fortunate in accumulating with the aid of relatives the necessary food and wealth. (Collins 1974a:231)

Old Age and Death

By their own choice the elderly -- even the blind and otherwise physically handicapped -- continued to work up to their physical abilities. If, however, they were ill or feeble, they were carefully tended.

As death approached, a person's guardian spirits left. His soul also departed, came to a river or lake, and was welcomed by the ghosts already there. In a rotten log the ghosts paddled across and ferried the newly arrived soul to their village in the land of the dead, which was like the living world except not as pleasant. The ghosts of all former relatives lived in the same dwelling, as their kinsmen did on the Skagit. They spoke a language -- understood in real life only by shamans -- in which the sounds of words were reversed. They strongly resisted all efforts by earthly shamans to recover a newly-arrived soul-ghost. (Collins 1974a:232)

Bereaved husbands and wives cut their hair to their shoulders and hid the clippings to prevent others from exercising hostile "contagious magic" with the cuttings. To make it difficult for the newly generated ghost to touch a living relative and so capture his soul and bring about his death, a surviving spouse, child, or parent usually left the house for a time. The closer the relative and the stronger the affection for the dead person, the longer the period away from home. The dwelling was rarely burned following the death of one of its occupants. The rapid distribution of the property of the dead person among the relatives, for which return gifts were later made to replenish the family's necessities, reduced the attraction of the ghost to living close relatives. (Collins 1974a:232-233)

The body was kept for four days, burial being on the fifth day. Men or women were paid to sit with the corpse at night through the period. A man with special "power" qualifications was hired to supervise the burial. He dressed the body and bound it tightly in cedarbark mats. (Collins 1974a:77, 233)

Several methods were followed in disposing of the corpse: it was buried in a shallow grave, the least prestigious method; it was placed on a plank or in a canoe (or canoe half) and tied in and to the branches of a tree; or it was put in a small wooden house constructed on stilts, the least common method. The most expensive and prestigious burial was in the canoe -- or half a canoe or even two canoes, one upside down on the other -- since these objects were very valuable. Some of the person's most valued objects were placed with the body. (Collins 1974a:233-234)

The man in charge, with his helpers, carried the body to the graveyard. Occasionally he removed it from the dwelling through an opening created temporarily in the side by the removal of a plank. All the guests followed the body to the burial locality, many of the women wailing. After the interment, all retired to the house for a feast and perhaps a potlatch. (Collins 1974a:233-235)

After the death of a spouse, the survivor was expected to fast and to bathe and scrape his body with boughs to remove all lingering contamination from the death. About ten years following a death, the person was honored with a potlatch if the family finances could support the ceremony, and the bones, cleaned and rebundled, were reburied, placed in a tree, or buried in a box in the house floor. Actually bones "of a number of dead persons were usually reburied on the same occasion." (Collins 1974a:235)

Leadership

Concerning what might be loosely considered the political structure of the Upper Skagit, Collins makes two summary statements: (a) each village cluster (Collins' "extended village") was politically autonomous; and (b) the only authority extending beyond the village cluster existed in the kinship system according to which "the oldest active man or woman in each generation of brothers and sisters and cousins [in the Upper Skagit extended sense] was the head of this generation" irrespective of where this individual resided. This leaves much untold. For example, whether the individual generations were somehow linked in some sort of older-younger hierarchical system, though one might guess from other evidence that this was probably the case. Further, how affinal kin and nonrelatives resident in the village cluster -- the first were obviously present and at least a few of the latter presumably so -- were placed within this consanguineal kin-directed "political" system. Whether there was invariably only one set of these generational tiers in a village cluster, all residents being descendents of a single actual (or hypothetical?) ancestor, save for the affines and occasional residents of known outside origin. And finally, if there was more than one set in a village cluster, descendents of different known (or presumed) ancestors, how were these different parallel sets interlocked into a single, multi-family system. These and other serious data gaps aside, the village cluster authority structure possessed, according to Collins, the following characteristics.

Generation leaders normally gave no orders but led younger members of their groups by persuasion. Their advice was routinely sought and their approval secured by these persons in important matters, especially those requiring the cooperation of others like holding a potlatch or constructing a dwelling.

When weighty issues arose, these leaders called together their kin group within the village cluster for discussion and, if a unanimous decision was attained, for joint action. When truly important questions developed, a rare occurrence as whether a hostile shaman should be killed or a retaliatory raid against the Thompson should be mounted, siblings and "cousins" living in other village clusters within traveling distance were summoned. These assembly sessions could, in fact, be attended by nonrelatives in the village cluster or even by such persons in other clusters. In such assemblies anyone, man or woman, was privileged to speak, though generally only those with practiced oratorial skills did so. Others made their views known by private conversations with those who were sure to speak. If necessary, interpreters were provided. Action was contingent upon a unanimous decision. (Collins 1974a:111-113)

Apart from the kin leadership structure, individuals attained positions of prestige and prominence by becoming experts in house building, fish weir construction, driving game over cliffs or into ravines, presiding at group ceremonies, curing illnesses, and leading raiding parties. Others followed their leadership because -- and only because -- they individually gained from this participation as in sharing in the weir catch, and considered themselves fortunate in having persons with these abilities in their village cluster. These leaders exercised their leadership qualities only in the areas of their particular competence; indeed, as in the case of shamanas and raid leaders, they were sometimes distrusted and even feared when out of their vocational province. (Collins 1974a:113)

At the other end of the prestige scale, sluggards benefitted from the work of others, but were sneered at, insulted, and publicly demeaned for their lack of industry and cooperation in group tasks. (Collins 1974a:113)

Guardian Spirits and Power

In traditional times the spirit world "permeated nearly all other aspects . . . [of Upper Skagit culture] and was . . . the central theme of their lives" (Collins 1974a:144). This world manifested itself above all in the guardian spirits and their relations with living people. These spirits were so numerous and so varied in their characteristics, powers, method of acquiring a human "owner," and so on that the description that follows can only suggest their general features, present a few examples of specific spirits, and intimate the complexities of this supernatural world and its significance to the people.

Guardian spirits were of two basic types: those that bestowed on persons lay powers, such as those for fishing, hunting, woodworking, fighting, or acquiring wealth; and those that gave shamanistic powers useful for curing or killing by supernatural means. They were acquired as explained below. Individual spirits of the lay class became personal helpers of virtually every man and woman; those of the shamanistic variety appeared only to the few who were to develop shamanistic knowledge and skills and who, it may be observed, possessed lay spirits as well. (Collins 1974a:144-145)

Much of the year the lay spirits lived in their spirit world, with the exception of the warrior spirits and a few others that were believed to travel most of the year. After the persons with whom they had developed their personal relationships (their "owners") reached adulthood, they came to them from time to time during the year when they needed their supernatural assistance. A person needed only to speak the spirit's name to summon it, a fact that explains in part why people were very reticent to speak of their spirits in casual conversation. Each winter the spirits also came to their "owners," who, at some considerable expense, sometimes sponsored a "dance" for them. Shamanistic spirits, on the other hand, remained close to their shaman, one reason why people feared these persons, available on short notice to be called to aid in curing. (Collins 1974a:145)

The nature of the spirits is summarized well by Collins (1974a:146):

Spirits were associated with many but not all animal species, the elements, and natural phenomena like lakes and strangely shaped boulders in the Skagit River. Some spirits had only human characteristics. (All spirits, those linked with animals included, had also a human or semihuman form which they could assume at will.) Certain spirits could be obtained either as lay or shamanistic spirits but not all spirits had this dual potential.

The spirits were very numerous. Those that gave both lay and shamanistic power were mostly those of mammals, though they included both the eagle and hummingbird as well as a spider variety found east of the Cascades, a point of particular interest to this study. Spirits that were exclusively shamanistic appear to have comprised no mammals but a number of birds and likewise minnows, a lizard, and a snake. Other spirits were associated with trees (e.g., the cedar), with meterological phenomena (e.g., rain, rainbow, thunder), with topographic features (e.g., Mt. Baker), and with the winds of the four directions; whether these were lay or shamanistic supernaturals or both is not reported by Collins. Spirits with human or semihuman form were all physically peculiar or deformed, or dwelled in unlikely or impossible places (as underwater), or possessed unusual supernatural skills like the ability to cure smallpox. In the spirit domain there were even beings that granted the ability to prophesy, eat to excess, lead singing, pound sticks in rhythm, and make berries and roots grow. The spirits were, in short, essentially numberless in terms of both their identity and the specific powers they endowed.

Each spirit category (e.g., blackbear or eagle) had its unique native name and each individual spirit in that category had its particular name that was revealed only to the one person with whom it developed the protector relationship. (Collins 1974a:146-147, 150)

These guardian spirit beliefs led to a very special attitude toward animals in general. The Upper Skagit "saw animals as humanlike in their motivations and desires. They believed that man had to show respect toward them." Since many animals were associated with guardian spirits, the appearance of an animal, especially under somewhat unusual circumstances, caused people to wonder whether its associated supernatural was attempting to communicate to them. Children were taught not to kill animals or insects that might be related to the spirit helpers of their relatives and that, therefore, might later become their own tutelaries. Nevertheless, people were permitted to kill food animals -- like bears -- even if they "owned" the guardian spirits of these animals. (Collins 1974a:213-214)

Lay Spirits

Among the many lay spirits were the wolf, blackbear, and grizzly, all three of which granted hunting power. It was believed that bears could become human and that people with bear spirit helpers could change into bears. A special ceremony was held for the first bear killed in the spring. The raccoon and wildcat helped their "owners" in various ways. Hummingbird was an important curing spirit. Among the several wealth spirits were one which, in human form, "lived [in a house] at the bottom of a deep place with swirling water in the Skagit River," a two-faced woman, one who dwelled in a house beneath the waves in salt water, and a little bent-over man spirit who assisted men in hunting and fishing and women in making basketry and other crafts and so both sexes in accumulating wealth. (Collins 1974a:147-153)

A particular female spirit, with long head and body hair, protected the dwelling, insisted on a neat house, guarded the family when traveling, and made people lucky in catching salmon. A giant spirit which lived in swamps and wore moss on its head walked through the forest knocking over trees, making trees appear to burn, and creating a great noise as it moved. A spirit associated with Mt. Baker caused berries to grow on the slopes of that mountain. The cloud spirit, like the rain spirit itself, made driving rain. Several different spirits were associated with prophecy. There were spirits, too, that animated "guiding objects": vine maple boughs; flat, square cedar boards with four holes, painted white and red; objects made of mountain goat wool mixed with cedar bark and twisted into an "e" shaped loop; painted cedar ducks with handles on each side; and still other objects. The duck power was obtained by diving to the bottom of Big Lake in the lower Skagit Valley. A person with these spirits could make these objects come to life and compel people holding them to move at their will. The cedar boards, for example, gave power to fish successfully and led persons holding them to hidden objects and people who were lost. The ducks, in turn, provided their "owners" with the power to bring ducks down out of the air where they could be killed and also protected dwellings. Lay spirits likewise protected the people residing in houses and, coming in human guise, gave curing instructions. In use, the "guiding objects" were grasped by two men. While the owner sang, the object was motivated by the spirit, which pulled the men about. (Collins 1949:293 fn. 6, 1974a:153-156; Smith in Collins 1949:339-341)

A skeleton-like spirit, although of the lay variety, was atypical in giving one who possessed its power the ability to cure illness; but it likewise caused sickness by entering another person and brought about his death unless removed shamanistically. Two spirits who spent the entire year "flying around the world" were acquired by being close to a person who already had them. One was distinguished by being human in form, wearing an elaborate costume with eagle feathers, and granting war prowess. A person who fell sick with this spirit ran wildly off into the woods, heedless of possible injury, and had to be captured, restrained, and kept from biting pieces of flesh from onlookers while he sang the spirit song and danced to satisfy the spirit and return to normal. (Collins 1974a:160-167)

A few Upper Skagit possessed dog spirits that caused them, when they first acquired them, to be "killed" by others who already had these spirits and who comprised a kind of secret society. Taken to the woods for five days by society members, they were brought "back to life" and then returned to the village. There, prevented from biting spectators by being held back with cedar bark ropes, they passed through a ceremony that included devouring a dog and were returned to a normal human state. Since the Upper Skagit lacked this society, persons who secured this spirit were compelled to undergo this initiation in one of the nearby tribes who had the society. (Collins 1974a:166-168)

The Upper Skagit had and used no masks, nor did they have curing or true secret societies. (Collins 1974a:169)

Shamanistic Spirits

One of the most highly regarded shamanistic spirits was the two-headed snake, associated with the rainbow and thunder. Others were the lizard, which aided in hastening infant deliveries and on the evil side could be used to kill; an owl-like spirit that helped in recovering lost souls from the land of the dead; the loon spirit that had power to cure any illness; and the spirits of the whale and spider, both of which assisted shamans in curing. (Collins 1974a:169-171)

Criteria of Spirit Ownership

The "ownership" of one or more guardian spirits was attested to, in the Upper Skagit view, by success in life: specifically wealth, fighting prowess, physical health, and general well-being. Accordingly, no one feared a poor man for supernatural reasons, for his poverty was proof enough of his lack of power to harm through supernatural helpers. (Collins 1974a:172-175)

The possession of guardian spirit power was sometimes demonstrated publicly in a contest context. A man challenged others with the same spirit as his who were present at the time to best him by performing more spectacular feats of supernatural power. The loser was deeply embarrassed and might even die by having his guardian spirit purloined by the victor. (Collins 1974a:174-175)

Acquisition of Guardian Spirits

It appears that virtually every Upper Skagit man and woman, possibly in fact every one, secured at least one guardian spirit at some point in life. These essential helpers could be acquired at any age. From about four years of age to puberty, boys and girls -- both needed supernatural power equally badly -- were sent out in search of them but occasionally might even obtain one around the house. Later in life, too, they could be secured if one wished. It was critical especially for a child first to become clean and pure through a period of fasting and bathing in isolation. In general, the longer one fasted and bathed, the more powerful the spirit that one secured. Children were told what spirit to seek (e.g., a hunting, fighting, gambling, or curing spirit) and for some spirits was given an object associated with the spirit to take along on the quest: e.g., a bow and arrow for a hunting or fighting spirit. Fearing to go out alone especially at night, children were forced to go on the search, through sometimes prepared for the experience by being sent alone on errands to other houses at night. To be certain that they had gone on their search where told to go, children were sometimes given an object to leave at the quest site; a parent checked the locality the following day. They were sent, for example, to a lake where they dreamed of a spirit house under the water, woke and descended into the lake, became unconscious and met the spirit, and woke dry on the shore with a spirit-given song clearly in mind. Breaking one's fast while on a quest (even eating berries) was most serious, resulting in an early death. Few adults, however, sweat-bathed before setting out on a quest, a practice common among Plateau tribes. (Collins 1974a:171, 172, 175-180)

If an animal, the spirit invariably appeared to a person as a human and in this form gave him his specific powers, sometimes even a choice between lay and shamanistic power. It apprised him of how to use the power, taught him his spirit song and dance, and instructed him concerning the clothing he must wear and the fasting he must observe during his winter initiation dance. This information conveyed, the spirit turned around and changed into its animal guise as it disappeared. (Collins 1974a:180)

If a child, on returning home from a quest, avoided his parents and went to sleep at once, his parents knew that he had acquired a spirit and the following morning sent him to bathe and fed him a little food. From that time on until, as an adult, the spirit returned to him, the child was supposed to forget his experience. Some winter years later he fell ill and his sickness was diagnosed as the consequence of his spirit coming back to him. Still, soon after his spirit experience, he was expected to show its effects by exhibiting special aptitudes of some sort. (Collins 1974a:179)

In general, a person acquired spirits that had belonged to an ancestor, either on the father's or on the mother's side. Because spirits were so numerous and everyone, or nearly all people, had them, those available from the ancestral group were considerable in number. When a person was about to die, his spirits left him -- sometimes dramatically as with thunder and lightning in the case of a thunder spirit -- and remained in the vicinity, each attempting to become a supernatural helper of the dying person's siblings, children, or even more distant relatives. If these spirits failed to develop this relationship, they returned "to wherever they came from." Later in theory they became the powers of unrelated people. A person could also send one of his spirits to protect a descendant in difficulty; whether he himself lost it in the process is unclear. Guardian supernaturals could also be stolen by shamans and hidden in the mountains, causing illness in the person losing it and even his death. (Collins 1974a:180-183)

The Shaman

The shaman was a specialist in diagnosing and curing (as well as causing) illnesses with the aid of his supernatural powers and in conducting shamanistic ceremonies. The amount of time that he devoted to this practice was dictated by his reputation and the size of his clientele. It was a rewarding specialty since he received payment both for the diagnosis and, if he attempted a cure, for it as well. When not engaged in "shamanizing," he hunted, fished, and performed daily secular chores like any other Upper Skagit. (Collins 1974a:190)

A person became known as a shaman only when he (or she) began to practice publicly, often only after reaching middle age and after having some quiet successes within his immediate family. Success in attending ill persons was considered to demonstrate the "ownership" of the necessary supernatural power and the guardian spirits from which this power was derived. Among the Upper Skagit there was no shamans' society to certify his skills. Though several shamans might work together on a difficult case, there were no bonds among them otherwise; generally, in fact, they were suspicious and hostile toward one another, an attitude to which explicit expression was given, for instance, in their serious public power contests. (Collins 1974a:190-191)

A shaman tended to have many and especially powerful supernatural helpers. Most of these were lay spirits, but one or more had to be shamanistic in nature, since it was from them that he obtained his power over illnesses. Shamanistic spirits were secured by fasting in the forest in the usual way. They were even thought to be easier to acquire than lay ones, since as ancestral tutelaries they were more insistent on being taken over by descendants. They were, however, more dangerous and hence more feared than the lay variety. (Collins 1974a:191)

Persons who were to become shamans were considered to be somewhat different at any early age owing to their apparent interest in and ability to gain supernatural power. Moreover, they and their relatives sought to promote these differences. They also possessed certain behavioral quirks: e.g., they were quick to take affront and slow to forget, traits that, retained into adult life, led to taking supernatural revenge. When their spirit returned to them later in life, they did not fall ill as persons with lay spirits did but simply began to cure. They held a winter dance, however, which, I take it, served to call general attention to the fact that they were beginning to practice. Shamans could summon their spirits at any time during the year, unlikemost lay spirits which returned only during the winter season. (Collins 1974a:190-194)

The same shamanistic spirits could be used -- indeed, sometimes insisted on being used -- to kill as well as to cure. A shaman harmed another person either by sending one of his own lay or shamanistic spirits into that person or by taking away one of the latter's own spirits. The former could be accomplished by certain supernaturally threatening gestures in the presence of the person. He could even send his spirit to enter a victim a long distance away. When the person became ill, he assumed some of the idiosyncratic behavioral patterns of the hostile shaman -- like a peculiar cough -- which aided a curing shaman in identifying the person responsible. (Collins 1974a:194-196)

Alternatively, a shaman could harm a person by taking his spirit and hiding it in the mountains, burying it, or throwing it into a river. He might even accomplish his purpose by merely looking intently into the other's eyes; this was one of the reasons why children were taught to fear and avoid strangers. Sometimes the curing shaman could see the lost spirit but, owing to his lesser powers, found it impossible to recover it. (Collins 1974a:196)

The typical curing ritual is of some interest. Shamans who were "blood" relatives of sick persons were preferred, since there was always the possibility that a non-relative would have a lesser interest in the patient's welfare and in fact, if he were a local person, might even have caused the illness for the sake of the fee. An alternative to the summoning of a kinsman was to call a shaman from a distant village, but this was costly. Shamans were even known to sing a single curing song and then demand a fee (e.g., a deer skin) before singing a second one. They actually received a fee, though a smaller one -- even if the person died. (Collins 1974a:196-197)

Illness caused by an intrusive spirit was cured by the shaman's sucking the foreign spirit out of the ill person, generally without leaving a mark, and plunging it into water in a cedar-root basket beside him. There his spirit and the intrusive one sometimes fought one another until drops of blood appeared in the liquid. Finally, the enemy spirit, in the form of a worm or insect, was retrieved from the water by the shaman. When the sickness appeared to be the result of spirit capture, the shaman dispatched his own spirit to hunt the missing power down, capture it, and bring it back. (Collins 1974:197-198, 208-209)

When a harmful, intrusive spirit was removed from an ill person and captured, it could be disposed of in several different ways. It could be returned to the offending shaman, the usual procedure unless he had sent this same spirit to perform its odious task as many as three times previously. In cases of repeated offenses, it could be sent to the mountains, buried, or placed in a river to prevent its return to its shaman owner, causing him to become sick or suffer some physical disability (e.g., deafness or blindness) unless he possessed other powerful spirits as well. But if an evil shaman who had caused endless trouble proved too powerful to be injured by supernatural means in punishment for his string of excesses, his neighbors, banding together as a last resort, ambushed and killed him. Because the provocation for this deed was widely recognized, no interfamily feud resulted but the act generated ill-will that persisted for generations. (Collins 1974a:198-199)

Shamans likewise cured two types of illnesses not caused by supernatural means under the control of hostile persons: (a) those brought about by an animal, a particular white worm, or a tiny red insect that entered the body and fed on it; and (b) those resulting from the loss of the soul, as distinct from the guardian spirit. (Collins 1974a:199-200)

It is of considerable interest that the dramatic shamanistic curing ceremony involving the use of "power boards" that was so characteristic of the tribes just south of the Upper Skagit -- the Snohomish, Skykomish, and Snoqualmie among others -- was not part of the Upper Skagit curing complex. These boards were fashioned of planks carved with a vaguely anthropomorphic or bird-resemblant "head-like" top and were sharpened at the lower end for insertion into the ground. The flat surfaces were painted with simple geometric and semirealistic animal designs. Boards were always used in pairs. These and the accompanying elaborate rituals were considered essential elements in recovering the soul of an ill person among the groups immediately south of the Upper Skagit. (Smith 1941:205, 1946:310-311; Wingert 1949b:78-80)

Shamans as Ritual Leaders

An entirely positive function of the shaman was to act as the ritual leader of the community. One example: as a part of his central role in curing a person when his guardian spirit first returned to him, the shaman helped him sing his spirit song, saw that he obeyed his tutelary's food and other behavioral restrictions, and protected him from the possible dangers of supernatural harm by hostile guests at the ceremony. His various ritual services were so highly regarded by his community that he received from it great prestige and authority. A shaman was especially well thought-of by his own relatives, who customarily expressed their approval by regarding him "with great respect and affection." (Collins 1974a:205)

Winter Spirit Dance

As already mentioned, a spirit first returned to a person when he had become an adult. Typically the "owner" fell ill and his sickness was determined by a shaman to be the consequence of his spirit's return and desire for acknowledgement in the form of a winter dance and property distribution. If he were not to die, the ill man was obliged to sponsor such a ceremonial. An announcement of the coming event was sent to relatives in other villages; other persons with the same spirit were aware without this information of the spirit's return. These kinsmen were expected to gather all the food and gifts they could spare and hasten to the village of the ill person, lest he die before the rite could be performed. Arriving at the "power" structure, they danced into it single file, placing their food and property in the center so all could see how much they brought. A master of ceremonies, a well-respected relative hired for the occasion, announced the arrival of each group. The ill person lay in one corner. The shaman who had diagnosed his illness was likewise present. Then a "singer," guessing the probable identity of the initiate's spirit from his earlier behavior and particular accomplishments, tried various spirit songs. When he hit upon the correct one, the ill person rose and, supported by two kinsmen, danced four circuits of the house. The entire assembled group then sang his song, accompanied by a skillful drummer who had to get the rhythm precisely correct. Spirit singing could not be carried out properly by the individual: it required the help of others to be satisfactory to the supernatural. During this ceremonial period, the initiate was compelled to observe the restrictions given him at the time of his spirit experience or of its return lest the spirit leave him: e.g., fasting, drinking no water, singing the proper song and dancing the appropriate steps, dancing a specific number of days, wearing certain paraphernalia like deer-hoof rattles and red facial paint. When he tired, another person took his place, singing his song until the initiate himself was rested. At the close of his singing period, others with the same spirit power were permitted to sing their guardian spirit songs, though few except kinsmen of the initiate availed themselves of this opportunity out of fear that their spirits might be stolen. (Collins 1974a:172-175, 183-186, 288)

Also part of the ceremony were spectacular demonstrations of supernatural power, feasts of abundant food, and formal addresses beginning with an expression of gratitude to all relatives for attending. Distributions of the property gathered followed, kinsmen being especially favored, though in early days the Upper Skagit had little to give away: perhaps a mat, a basket, a part of a blanket, or raw materials (e.g., basketry fibers or wool). Return gifts were made at some future time, but this property exchange was not a principal element in this winter guardian spirit dance as it certainly was in the potlatch. (Collins 1974a:187)

Guardian spirits were supposed by the Upper Skagit to bring success, including wealth. Substantial property was necessary to give the winter ceremonial validity: to demonstrate that the ritual sponsor actually possessed the guardian helpers he claimed and to permit the ceremony to be accepted as a proper response to the spirits' demands. When a man who acquired a spirit was too impoverished to subsidize this ceremonial with his relatives' aid and died as a consequence of this failure to reply to the spirit's requirements, the Upper Skagit system was faced with a contradiction that appears never to have been resolved. (Collins 1974a:188)

The Soul

The soul was considered to be an entirely different entity from the guardian spirit. It was "a nonmaterial, life-giving essence" possessed by each person. It was generally invisible, though it could be seen by persons who had been given this power by a guardian spirit. When out of the body, it was thought to be identical to the body in form but smaller. It might leave a person in a dream, returning before he woke. On a trip, the soul might precede a person to his destination without harm to the person. It could also be taken from a person with untoward consequences unless it was soon recovered. (Collins 1974a:200)

Soul loss was at least sometimes diagnosed on the basis of the actions of the person. At times the soul of a person was taken by a ghost of someone who had died, if the living person mourned the death too deeply. For ghosts tended to remain in the vicinity of relatives in the hope of recruiting companions on their journey to the hereafter. In this after-land life was quite like that on earth but somewhat less pleasant. A new arrival was visited by the deceased kin of people still alive to inquire about the well-being of their relatives. In addition, living people were thought to visit the land of the dead several times for brief stays before their final death. (Collins 1974a:200-201, 204-205, 207)

Shamans who specialized in recovering souls were believed to go to the land of the dead to bring the soul back. As they made the "journey," they described their progress in detail to spectators at the ceremony: their search in the land of the ghosts for the soul, frequently hidden by the ghosts to prevent its recovery, their fight with the ghosts for the soul if such occurred, and their return journey with the missing soul. With his soul restored by the shaman, the ill person recovered. It was possible for shamans in the ghost country also to steal the souls of persons dead for some time and to bring them back to enable them to be reincarnated as reborn infants. (Collins 1974a:201-204, 221)

Souls could also be stolen by supernatural beings associated with unusual natural phenomena along much of the Skagit River. For example, near Rockport stood a large cedar tree with long hanging moss. If a person walked near the tree, it took his soul, threw it across the river to another tree that hurled it back, each time with the sound of thunder. And so it went back and forth, causing the person's death unless retrieved in time by a shaman. "Near the junction of the Suiattle with the Sauk lived a dangerous supernatural being ... who is described as an 'animal.' People who encountered this being in the woods might die." (Collins 1974a:214)

Religious Sanctions

Upper Skagit religion had its price in time, property, and anxiety as well as its benefits in explaining the unknown and unknowable in life, in bringing the ill back to health, in tying the group of relatives together for their general benefit, and in supporting through the guardian spirit system the right of people to be different. How was this complex religious system maintained? Largely through fear of spirit retaliation through illness and death as a consequence of failure to conform to the beliefs and practices of the system. And illness -- the prelude in the Upper Skagit mind to death -- was a source of substantial difficulties to one's family. Other kinsmen had to assume his or her normal duties that were essential to the functioning of the family: the subsistence quest and the manufacture of the objects required by the ceremonial life. Others had to contribute to the costs of the curing ceremony. And the entire family might well lose social status through the family's inability to do its part in subsidizing large family and public functions. (Collins 1974a:206)

On the other hand the positive rewards of proper and full participation in the supernatural system were also considerable. It kept alive and flourishing the beliefs and practices regarding the guardian spirit world, the supernaturals that gave people their individual powers that resulted in success in the practical and religious world, in good health and a long life, in children, and in wealth, high status, and respect. (Collins 1974a:207)

Magic

Magic among the Upper Skagit belonged to a different realm from the guardian spirit world of religion. It involved no supernatural tutelaries but rather charms for success in love and hunting, for improving the appetite, for curing sickness, and for preventing fatigue. These charms worked with a certainty if one knew the correct actions to follow and formulae to utter. When one wanted what could be gained by magic, one hired a man or woman who possessed the necessary knowledge to work the necessary spell. (Collins 1974a:207-208)

Malevolent magic occurred also but was of even less importance than good, sanctioned magic. While aware of the practice of making a small image of a person and damaging it to produce the same injury in the person represented, the Upper Skagit claim that they themselves never practiced this custom.

Both men and women, though few shamans, were the owners of magical lore. But in comparison with the guardian spirit system, magic was relatively unimportant. (Collins 1974a:208)

Concerning the traditional diversions of the Upper Skagit -- art in the narrow sense, dance, and music as well as games and smoking -- virtually no data appear to exist.

Art

Prehistoric stone carvings, anthropomorphic in form and some as much as 2 or 3 feet high, have been found along the upper reaches of the Skagit River (Smith 1950b:27). In light of this fact it is curious that no mention of these Skagit figures seems to be made by Wingert (1949b) in his detailed study of native Northwest Coast sculpture.

In ethnographic times coiled baskets were ornamented with designs worked upon the basket surface by the imbrication process: popular patterns were frogs, butterflies, salmon back-fins, fish gills, fish nets, and a small fly. Upper Skagit women took great pride in their coiled baskets. Presumably the ornamentation of skin clothing and facial painting provided opportunities for simple esthetic expression, but the ethnographic literature is largely silent on the subject. Dwelling posts, which were sacred, and at times the house front around the entrance were carved and painted with the guardian spirit designs of the house owner. The external, self-standing totem poles of the coast were unknown. (Collins 1974a:63-64, 68-69)

Music

At least as late as the 1940s when Herzog (1949:93-109) reviewed what was then known of the music of the Salishan groups of coastal Washington and British Columbia and the interior of that province, the musical traditions of the Skagit were still very poorly described. Nevertheless, enough was understood of the principal musical characteristics of this large group of tribes to suggest that those of the Skagit could hardly have been significantly divergent. With this premise accepted, Skagit music must have been very like that of the Chilliwack in sharing both the general North American Indian characteristics and the particular traits of Salishan music and in being performed with the same inventory of musical instruments. These data are summarized in the Chilliwack chapter that follows.

Mythology and Folklore

As Hess (in Hilbert 1985:xviii-xxiii) has written in a fine summary of the characteristics of the old myths from which the following is borrowed and synthesized, the Lushootseed-speaking Salishan tribes of Puget Sound (including the Skagit and others south to the Puyallup and Nisqually) separated their tales into two categories: "myth-age tales" and "history." "Personal accounts" of the news and individual experiences constituted a third, non-traditional genre. The myths, in turn, may be divided loosely into two varieties, a distinction apparently not made by the Indians themselves: (a) rather light, humorous myths, which, while "reflecting the foibles of human nature," taught proper behavior by detailing the unfortunate consequences of tabu-breaking; and (b) longer myths that "examine[d] the world order and deal[t] with points of tension within the social fabric [and in which their] . . . major characters were heroes who acted wisely rather than impetuously or foolishly as in the lighter stories."

Myths deal with the early time when all beings shared the traits of animals and people and possessed supernatural power and human emotions. They had, however, many of the physical and behavioral characteristics that animals possess today. As Hess illustrates:

These characters, each with human and animal traits, provide[d] the reconteur with great flexibility . . . to draw upon on one or the other of these attributes as plot, imagination, and sense of humor require[d]. . . . [T]he sisters Boil and Hammer [in their myth] . . . looked and acted just like people, except they ate only berries. At the same time, however, they were like a real boil and a real stone maul. Therefore, Boil burst when a fir needle fell on her head and Hammer drowned when she fell into the river because, after all, she was stone.

With the aid of their supernatural power, they were able to perform superhuman and superanimal feats. During this primitive time-period there were several worlds -- e.g., a sky world and one of the Salmon People -- more or less similar to each other and with travel possible between them.

Particularly prominent in the Lushootseed myths were three unprincipled, impulsive tricksters: Mink (a rake), Raven (a glutton), and Coyote, borrowed in tales from the nearby Plateau east of the Cascades. Their episodes are immoral, dishonest, farfetched, and humorous. While tricking others, they themselves were often tricked. They were also capable, however, of performing occasional heroic deeds, like procuring daylight.

These myths are of special ethnographic interest in that they reflect facets of traditional culture. For example, their beings are organized into rather loose high and low classes with slaves at the bottom. High class beings are wealthy in goods, natural resources, and rights to songs, dances, and ceremonies, possess strong supernatural powers, and are influential in their villages. Low class people, in addition to lacking the above, are often uncouth. Movement between the two classes is possible by gaining or losing wealth and developing competence and exhibiting correct behavior, and by acquiring guardian spirits. Sibling rivalry is prominent, with the youngest of the series invariably getting the upper hand.

It is, of course, impossible to reflect in these traditions as written the teller's informative but non-lexical expressions, emotions, rhythmic and pitch variations, vocal mannerisms, and so on, as well as the fragments of song often incorporated into the tales.

As time passed, a transformer of great supernatural power walked through the Lushootseed world, changing as he went these mythic beings into humans, animals, and plants and the landscape into its modern form.

So much for the generalized picture of the mythology of this group of linguistically and culturally closely related people, which included the Lower and Upper Skagit. Fortunately, we have from Collins (1974a:211-212) a parallel generalized description for old Skagit mythology alone. It is of interest to compare her data with the preceding geographically much wider summary of Hess. In some ways it conforms well; in other ways it differs somewhat. But the Collins material is important in its own right, for it offers a fair amount of new information.

According to Collins, the past was divided by the Upper Skagit into two periods. The first was a remote, mythical time when animals were much larger, had the ability to speak, and behaved like humans. Then the world was different from now: there were no death and no mountains and the rivers followed different courses from the present. This old world was transformed into the one of today by several mythological figures, including Raven and Mink (coastal myth personages) and Coyote, who, as the Upper Skagit stated, came to them from the Plateau. (Collins 1974a:211-212)

But their most important culture hero was an entity in human form that was associated with the sun. He "was entirely a benevolent being, noble and upright," not a trickster. A fine supernatural, he reduced the "animals" of the mythological period to their present size and their present wholly animal, generally non-threatening status, altered the world's topography to its current configuration, and taught humans arts and crafts. But he also occasionally changed people into striking stone formations still to be seen today. Such was the odd-shaped stone, once a woman, near Portage on the Skagit River. Addressed as "grandmother" by the Upper Skagit, she was believed to supervise the fishing on the Skagit above the mouth of Cascade River. People from villages upriver from Marblemount asked her respectfully for fish when they went for them and thanked her for her favors as they passed her. (Collins 1974a:211-212)

Raven, Mink, and Coyote were all mythological trickster transformers, supernaturals in the form of these animals but as large as humans, capable of speech, living in houses, and to a certain extent behaving like humans. "They were greedy, mischievous, and lacked the nobility and disinterestedness" of the highminded transformer described in the preceding paragraph. Of the three, the Raven was the most important, a being who never hurt anyone though he was responsible for death owing to a bit of short-sightedness. (Collins 1974a:212)

Myths were told in evenings for the children, who were expected to sit just so and be attentive. To increase interest and give the tale a vitality of its own, a good raconteur changed his voice for different characters and added details of his own choosing. (Collins 1974a:212-213)

Folklore, in contrast to myths, told of events that happened after the world became as it is today, including happenings to ancestors of living people. All were localized either on the Skagit River or its tributaries or in the country of neighboring tribes. Many dealt with encounters of people with guardian spirits and many possessed a moral content, exemplifying the importance of caring for the aged, the handicapped, and stepchildren, of observing the levirate and sororate, and so on. (Collins 1974a:213)

Several collections of myths and legends in whole or in part from the Skagit have been published. These include (a) Hilbert's Haboo (1985) with 17 tales told by Upper Skagit, 10 by part-Skagit, and two by Sauk-Suiattle reconteurs; (b) Haeberlin's (1924) "Mythology of Puget Sound" with 13 Skagit myths; (c) Langen's (1984) "Four Upper Skagit Versions of 'Starchild'"; and (d) Elmendorf's (1961a) "Skykomish and Other Coast Salish Tales" with one myth reciting an occurrence at the Skagit River mouth.

Conception of the World

The world was believed to be an "immense round island ... in an ocean" surrounded by a high wall. Behind this barrier was the pitch-black world of the guardian spirits, when they were not with humans, and of the Salmon People. Annually these salmon folk passed through an opening in this wall in their canoes and once beyond this obstacle tumbled from their canoes to become salmon and ascend the rivers of the Indian world to spawn. The sky, some believed, was a roof above which the stars lived as people. "The sun, the moon, certain individual stars, and constellations of stars were regarded as associated with supernatural beings." (Collins 1974a:209-210)

Calendrical Concepts

Traditional Upper Skagit months were lunar months. At least some had names that were meaningful in terms of the procession of the animal behavior through the year. Thus the third moon was the "moon when the frogs come out and begin to sing" and the eighth was the "month when the spring salmon . . . [i.e., chinook] spawn." (Collins 1974a:210)

Secular Medicine

The Upper Skagit possessed an extensive knowledge of curative herbs and other substances, a third means for curing ailments quite apart from both the guardian spirit and the magical systems. Most herbalists were women. If shamans administered plant medicines, this was entirely distinct from and secondary to their guardian spirit powers.

With herbs or bark steeped in water, sore eyes, headaches, constipation, and urine retention were treated. Broken limbs were set in splints. Ill persons were encouraged to remain in bed as little as possible. Contagion, though explained in guardian spirit terms, was understood and prevented by isolating in the forest the person affected. (Collins 1974a:209)

Introduction

Such fragmentary data as have been located in the ethnographic literature are assembled -- and extended by a certain amount of reasonable speculation -- into four sections: the subsistence use of the Park Complex, its use as sites for villages and task camps, its use in travel routes, and such other secondary uses as may be identified. Source citations are omitted in this section where the data can be found in the preceding topical discussion.

The Upper Skagit were described by Eells in the late 1800s as a group that "roamed in the Cascades mountains." Subsequent ethnographic information confirms this statement. The tribe claimed as a part of their traditional country the mountainous middle or upper Skagit Valley and nearby high country north to the international boundary, territory wholly within the Park borders. This northern line is, however, obviously unrealistic from a geoecological perspective, an artifact of today's political world. Evidently Upper Skagit parties at times went a bit farther north up the valley and its side mountains, far enough to encounter hunting parties from the Tait tribe, whose particular homeland was on the Fraser River between the Agassiz bend and the mouth of the Fraser River canyon. Realistically, this northward roaming could hardly have been a frequent occurrence, for the Upper Skagit, if Collins is correct, laid no claim to this north-of-the-present border country and, moreover, would surely have made use first of the natural resources of the Skagit Valley and high lands closer to their winter villages. In addition, the Lower Thompson regularly hunted south of today's Canadian line within the Skagit drainage, sometimes as far downriver as the Beaver Creek region and, in fact, considered this area to be their proper hunting and gathering territory. Coming from the west, the Chilliwack, in their turn, entered the Skagit River country within the confines of the present Park. Plainly we have in the Ross Lake region an area claimed and routinely utilized as food-source territory by the Upper Skagit and Lower Thompson and at least used to some extent by the Chilliwack and perhaps even by the Tait, though not claimed by these last two groups. Encounters in the area between Upper Skagit and Lower Thompson often, possibly even generally, generated moments of tension, with hostilities the consequence. Contacts between Upper Skagit and Chilliwack were sometimes inimical. Those between the Upper Skagit and Tait were, however, generally friendly, perhaps because they must have occurred in the northernmost periphery of the Upper Skagit tribal range and because they were, it may be conjectured, relatively rare events.

Consistent with these data, Upper Skagit culture, while having its coastal-resemblant traits, likewise possessed others that were clear adaptations to an inland, mountain environment. These adaptations included, for example, an extensive use of large and small game, of roots, and of berries as food, the wearing of moccasins and other skin clothing when the weather was cold in the mountains, and the possession of gear to assist in cross-country travel -- snowshoes in winter and, in protohistoric or at least early postcontact times, a few horses for open season use. It is cultural baggage of these sorts, about which we, unfortunately, know so little, that so sharply points toward a broad utilization of the mountains and high forests and their animal and other resources throughout the year, and that, accordingly, indicates a consistent and thorough exploitation of that part of the Park area to which the tribe laid claim. Indeed, there would have been no reason for the Upper Skagit to have claimed territory which they failed to make use of in some measure: no neighboring groups would have considered legitimate any such empty claims.

Further, the fact that a significant part of the Upper Skagit winter population cannot be accounted for in the known summer villages and camps along the upper Skagit River seems best explained by the presumption that during the warmer months a considerable portion of this missing population was off in the mountains -- many within the Park region -- pursing their subsistence activities in small task groups. On the other hand, as noted below, there are a few fragments of ethnographic information that suggest some reluctance on the part of Upper Skagit folk to spend more time than necessary in the high country, in particular the stated dislike of the women to dig roots in the mountains.

Subsistence Use

The region of the Park and Recreation Area in and above the Skagit River gorge between Newhalem and Ross Lake was utilized by the Upper Skagit primarily as a subsistence source area. Among the nearby tribes these Skagit people, in fact, held a reputation as hunting specialists. Parties journeyed into the Ross Lake region throughout the warmer seasons in search of game, like deer and elk, that could be found to some extent also at lower, downriver elevations, but more particularly of those animals that either spent these months typically in the higher country or were limited by their habitat preferences to these loftier elevations. In this latter category, according to the ethnographic data, were the grizzly bear and especially the mountain goat that furnished robes and wool. This white wool was woven into cloaks and blankets that were something of a specialty of the Upper Skagit and comprised an important item of trade.

It is obvious that the ethnographic sources provide an incomplete catalog of those animals native to the higher mountain terrain of the Park Complex that were taken by the Upper Skagit for food or their other products (like peltries). A more exhaustive list of indigenous fauna can be compiled from the zoological distribution maps of Dalquest (1948), a task simplified by the fact that the entire Park lies within the Cascade Mountains Mammalian Area. Animal use information culled from ethnographic accounts of other tribes in the vicinity of the Upper Skagit territory reveals that these forms, in some cases in different taxonomic varieties, were hunted, trapped, or killed as chance provided by one or more of these groups. Hence it may be presumed that they were likewise secured to some degree by the Upper Skagit within the Park boundaries. The roster includes, in addition to the grizzlies and mountain goat, the Canadian lynx, fisher, red fox, and western marten whose range did not extend much if at all beyond the western border of the Park. Found widely in the high country of the North Cascades Complex and also westward outside the Park were beaver, black bear, bobcat, cougar, coyote, deer, elk, ermine, marmot, mountain beaver, pika, raccoon, snowshoe hare, timber wolf, and weasel.

Anadromous fish certainly occurred up the Skagit River within the Park as far as Newhalem (presumably as far as the gorge) and were taken by the Upper Skagit in this sector. Whether they ascended the river through the gorge and into the Ross Lake region I do not yet know. In any event, fishing within the lake sector seems to have been a rather incidental pursuit.

While the men were engaged in their hunting tasks in the mountains, women were occupied in processing the meat and other game products. To some degree they were also root gathering, berrying, and collecting and preparing for later consumption other plant substances according to the season. For some unexplained reason, as previously noted, Upper Skagit women rather disliked root digging in the high meadows. Perhaps the soil tended to be conspicuously more rocky than on the flood plains farther down the Skagit and on the lower Sauk and Suiattle. Possibly the task of transporting the processed roots back to the winter village was too laborious, for the tribe traveled overland to a considerable extent as well as by canoe along the stream courses. It may be that women were uneasy in the elevated country out of fear of resident supernaturals. And in the Park region the possibility of meeting hostile Lower Thompson hunters and gatherers was always present. Whatever the explanation, they preferred to avoid securing their root supply in the high country.

Berries were extensively utilized as food by the Upper Skagit. When not preparing meat, women in the mountains devoted considerable attention to berry collecting and processing. Some of these berries, like the mountain blueberry, were typically high country fruit, while others were gathered, in their season, in both lower and higher elevations. But how many berry varieties were harvested within the Park borders and how great were the quantities of each kind are not reported.

Off in the northwesternmost corner of the Park, in the upper Baker River sector, the Upper Skagit of that stream country must likewise have spent at least some of their open season time in similar hunting, fishing, and collecting activities high up on the mountain slopes.

Related to this subsistence exploitation pattern of the mid-year months was a shift from the large, concentrated riverine winter and salmon-fishing villages to dispersed gathering camps with their temporary mat, lean-to shelters. During the summer hunting, root-digging, and berrying periods, the Upper Skagit of the permanent villages scattered to the various grounds to secure their food resources. In the process task groups were formed from members of various winter communities and these groups, while they existed, were as firm as the membership of the winter settlements. Thus the social units that prevailed in the Park region were not come-and-go, loose coalescences but closely-knit groups, though precisely how these links were expressed if not overtly formalized is not made clear. Nor are we informed of how the membership in a small summer grouping was determined. Owing to the importance of kinship bonds in Upper Skagit social life, it seems probable that each was largely a cluster of families, from various winter villages, that were related closely or distantly through kinship. Such an arrangement would have the advantage of bringing together for a time kin units that saw far less of one another during the remainder of the year than the families resident in a winter settlement and so of reaffirming the common relationship links of the larger kin network.

Village/Camp Use

Three traditional winter villages of the Upper Skagit are known to have been located within the confines of the Park Complex. All three were on the Skagit River: a large winter dwelling at Newhalem, a small winter house of three families east of Thornton Creek, and a large winter dwelling "at Portage, west of Damnation Creek," all within a stream stretch of some 5 miles. Without question, the residents of these communities must have been among the tribal members who made the greatest use of the resources of the Park. Other winter villages as well as summer settlements began immediately below the point where the Park boundary today crosses the Skagit River. Surely the people of these slightly more downstream settlements must also have visited the Park on their subsistence and other resource forays, though possibly to a somewhat lesser extent than the occupants of the three villages of the uppermost village cluster.

While there were, at least in ethnographic memory, no established settlements, winter or summer, inside the Park Complex above the Newhalem community, task-oriented camps must have been occupied for brief periods while the neighboring meadows, streams, and mountains were scoured for their resources. Unfortunately none of these camps can be located from the extremely meager ethnographic data, the principal interests of which are focused on the main salmon river areas of the tribe to the west and southwest.

Special note must again be taken of the northwestern corner of the Park where an Upper Skagit village was located on Baker River within a very few miles of the Park boundary. From this community hunting and foraging parties must have regularly gone year-round into the upper Baker River country and the mountain flanks bordering the valley, and hence well inside the Park limits.

Trail Use

Only four old Indian trail routes into or through the Park and Ross Lake Recreation Area can be defined with some confidence. One led up the Skagit Valley to enter the Park Complex above Marblemount, to follow the river northward through the Ross Lake sector, and to leave the Area at the present Canadian border. This was the water and land route by which the Upper Skagit moved into the Park country to pursue their resource quest activities, whether largely by canoe or on foot the data do not reveal. It was likewise the route followed by the Upper Thompson in journeying southward from their Fraser Canyon village centers to attend to their food and other needs in the high mountain terrain of the Park. Unless they kept canoes at or near the point where, having come overland, they reached the upper Skagit River (as the Upper Skagit folk did at the head of Lake Chelan), they must have used a land trail down the river, but whether on the east or west side of the stream or both in unknown.

One or more trails led into the Park from the upper Chilliwack River region. Parties of the Chilliwack tribe are reported to have been encountered by the Upper Skagit in the Park river country, the area considered by the latter to be their tribal domain. The most obvious route, it would appear, would have been via Little Beaver Creek.

A trail to the east broke away from the Skagit River at the lower end of the present Ross Lake, ascended Ruby Creek and Granite Creek, and continued eastward over Rainy Pass into the upper Methow Valley. It soon left the Park region, actually about 4.5 miles above Ross Lake. Another trail moved up the Cascade River, proceeded along its North Fork, and came into the Park about 1 mile upstream from the mouth of the Fork. Still within the Park limits, it threaded the mountains through Cascade Pass and descended the Stehekin River to Lake Chelan. While other paths also passed eastward over the Cascades from traditional Upper Skagit territory, these apparently all crossed the mountain crests somewhat south of the Park perimeters.

A considerable amount of trail travel must have taken place, even in old, prehorse times, by small parties of Upper Skagit and Plateau Indians, moving east and west, with trading, visiting, and other objectives in mind. Apparently the Rainy Pass route mentioned above was not the east-west trail most followed; evidently it was a path that led through one of the more southern mountain breaks. According to the Upper Skagit, cross-Cascades movement was particularly undertaken in deep winter on snowshoes. The reason, though not reported by the ethnographer, seems rather plain: while the weather was obviously less hospitable and subsistence game on the trail was less plentiful, the trails were more easily traveled when deep snow covered the rough path -- the rocks, snags, windfalls, low and tangled bushes, and the like. Temporary traveling camps were positioned, of course, along all of these trails, but their locations are not defined to my knowledge in the scanty ethnographic and ethnohistorical records.

Other Uses

It is axiomatic that Indian groups living in high altitude areas or journeying to them with task-specific objectives made as extensive use of the natural resources of these regions as possible within the needs defined by their culture, their knowledge of the terrain and its products, and the time they had available. These uses included not only those that the group considered primary, like, for the Upper Skagit, the hunting of mountain goats or the collecting of high altitude berries. They embraced as well the obtaining of other materials of secondary use, often widely varied in their nature, of limited availability, and desired only in small amounts.

Non-local travelers hastening along the mountain trails toward some particular locality with a few specific, sharply focused objectives in mind were obviously less well acquainted with the resources in the neighborhood of their routes -- particularly resources of a secondary character -- and, in general, surely had less time to exploit those they were aware of than were travelers and especially task parties of the group that "owned" the country. This is to say that the Upper Skagit must have been notably more familiar with the Park/Lake region and its resources than were Chelan, Methow, or Wenatchi journeying parties from the Plateau and than were coastal groups near the Upper Skagit when they traveled eastward over the mountainous watershed to the Plateau.

But what were the nonfood resources of the small streams and mountain slopes of the Upper Skagit River drainage that were secured by the Upper Skagit task-oriented groups? We know of mountain goat horn needed for fishing points, the hides of game killed for food, and mountain goat wool for weaving into cloaks and blankets, removed from skins brought back to camp, all materials that were excellent trade and potlatch items. But what more? Lithic supplies for the making of tools and other artifacts? Plants used technologically -- as in fashioning nets and basketry -- or as medicines and other curatives? Colored earths from which important paints could be made? Were there localities -- high lakes, craggy mountains, peculiarly formed rocks, or strangely shaped trees -- where especially powerful guardian spirits might be obtained? And so on almost without end. Unfortunately, the ethnographic data to demonstrate such Park and Recreation Area utilization by the Upper Skagit are simply not in the literature.