« Previous Region | Index | Next Region »
Umatilla 35UM1 (35UM35)
Hells Canyon Creek
McGraw Creek 35WA1
Ferry Creek 35WS261
"The Columbia Plateau
Here again we find the winter-summer dwelling pattern: the substantial house for winter and the light mobile, easily constructed mat shelter for summer when people moved to the fishing grounds. Permanent houses were of two clearly distinguishable kinds with perhaps a third kind that was certainly in use farther to the north in Washington. The three types are (1) the long, rectangular inverted V of the Nez Perce, (2) the rectangular floor plan similar to the first except it had a rounded end, and (3) the circular house set in an excavated pit" (Cressman 1981:37-38).
"The Nez Perce house type, though built in the shape of an inverted V, did not come to a point at a single ridgepole. In each corner of a rectangular pit, poles were set up and the end ones on each side leaned in toward each other at the top where they were tied to a ridgepole. Thus there were two ridgepoles separated from each other by a few inches. The space between the ridgepoles was left open to allow light to enter and smoke to drift out. Rafters were placed at intervals along the sides from ridgepole to ground and covered with mats overlapped, shingle fashion. Dry grass protected the bottoms of the mats and heaped dirt from the excavation held them down and provided warmth. One or more entrances, depending on the size of the house, were made in each side and covered with mats or hides" (Cressman 1981:38).
"Each house was occupied by a number of families, and divided into two long sets of rooms on each side of the central line under the open space between the two ridgepoles. Each fire, in a shallow pit in the center, was shared by two families, one on either side. Lewis and Clark described one of these houses as 150 feet long and contained twenty-four fires. Probably, therefore, it house forty-eight families. Such a house comprised an entire village. If we estimate five individuals to a family, this would mean a population of about 250 for this dwelling" (Cressman 1981:38).
Chatters 1984:title page
"The circular house was like that described for the Klamath but not so substantial. It was probably a very widespread type, with local variations in the region between the Rocky and cascade mountains and, as I pointed out, probably the earliest. The houses in this area, built in excavations from one to three feet deep, were supported by posts set either on the excavated surface or, rarely, in holes dug in it. Mats and brush probably covered the framework of posts and rafters, for very little timber appears in the archaeological record" (Cressman 1981:38-39).
Summer houses in the area were usually dome-shaped or conical (tipi) structures of a few poles covered with mats, cool, adequate against sun and occasional rain, and very movable" (Cressman 1981:39).
Limestone Timber Sale
In a timber sale report for the Limestone Commercial Thinning Sale on the Wallowa-Whitman National Forest, a single site with seven small shallow saucer shaped depressions was recorded. They were located at the edge of a meadow on both sides of a small creek. They were noted to resemble Plateau house pits in outline and cross section, but with a caution that they may be tree blow downs (Womack 1979:2-3).
Lower McNary Reservoir
Osborne, reporting work for the River Basin Surveys on the Lower McNary reservoir, tested a series of sites in 1948-1949. Site 35UM7 contained housepits. "Map 2, of site UM-7 itself, indicates the lack of settlement pattern, or at least of planned pattern. Only one possibility of such a thing emerges and that is the complete division of the 16 house pits (of which three: 1, 10, and 16, are of questionable validity) into two concentrations. The upstream group is, of course, less of a concentration than is the downstream. It may be recalled that the report on site 45-BN-53 suggested that rather vague groupings of house pits might exist at that site. Site 35-UM-7 obviously does not help in clearing up the question raised by the earlier excavation. There is certainly a division here but we do not know whether or not the two groups were contemporary and it is doubtful if complete excavation would settle the point. As far as the archaeological picture goes, they do appear to belong to the same period. As far as terrain is concerned there is no reason for the grouping--nor none which would preclude a more or less continuous line of houses along the terrace. As matters now stand, therefore, there may well have been a social reason for the settlement plan. The houses are all grouped close to the edge of the terrace and to the rather abrupt drop from the surface of the eolian-fluvial fill to the basalt slope to the river. This area includes the higher sections of the terrace itself. These are presumably remnants of a natural levee or accumulations of wind-blown soil held by the somewhat heavier vegetation of that part of the terrace surface. That this portion is desirable is evident in its better drainage and proximity to the river itself. Surface changes have obscured the outlines of many of the house pits but the picture was essentially as sketched above" (Osborne 1949:8-9).
The house pit were shallow circular depressions. They were eroded. House pit 3 was about two feet deep. No floor could be found during excavations or in profiles. House pit 4 was a NE-SW oval two feet deep. No housepit floors were again found, but there was an indistinct change in fill suggesting a saucer shaped house. An earth oven was found in the house. The oven was 11.5' by 5'. "It is relatively certain that, had the pit dug for house pit 4 been very deep. it would have left its mark in the sandy soil even if this trace were obscured by the midden. As it was not possible to find an outline we may feel rather sure that the pit was shallow and, as there were no superstructural remains found in this pit or in house pit 3 we may conclude that the walls and roofing were of a light and possibly moveable or temporary nature. This sound very much like a typical mat house, and is the same situation as was encountered at 45-BN-53 in 1948. After abandonment of the house pit as a dwelling place it was used as a convenient depression for an earth oven" (Osborne 1949:28).
House pit 12 was also tested and described as "about 30 feet in diameter, a typical house pit of 35-UM-7" (Osborne 1949:29). A probable house floor was encountered a foot below the sand fill in the pit. The 'floor' was only seen in the profile. A probably fireplace was found as well as a cache of hammerstones as well as a possible rush mat. Other possible floors were also recorded, the deepest suggesting a true pithouse. Two post holes were associated with this deep floor.
"(1) the average north-south diameter (28.03') is smaller than the average east-west (30.24'). (2) The average east-west diameter of the upstream group (25.7') is much smaller than the same for the downstream houses (32.27'). (3) the north-south for the same groups are 22.84' and 29.85'. Thus the upstream pits are smaller but in both groups the east-west measurement is the larger" (Osborne 1949:34).
Shiner reported on excavations in the village site of 35UM17 in 1951. This site contained 19 house pits in two groupings as well. The houses were just behind a natural river levee. The downstream group consisted of 14 house pits and the upstream had 5 pits. House pit 14 was 40 by 30 feet. The house when excavated was 2.5 feet deep at the center of its saucer-shaped depression. The house floor contained a charcoal streak that was discontinuous floor as well as a second floor. House 15 had a floor that was seen as a horizontal plane with two occupation levels. The house was 14 feet in diameter and was excavated to a depth of three feet. House pit 18 had no discernible floor. The house pit dimensions were 30 by 26 feet and an unknown depth.
"The Indians that lived on Techumtas Island had roughly circular houses, slightly larger on a northeast-southwest line, averaging about thirty feet in diameter and about two or three feet in depth. The houses were erected over saucer shaped pits probably by lashing mats to a framework of poles. The poles were not individually self-supporting but the structure was given stability by heaping dirt against the mats outside of the house. A small fireplace was usually placed in the center of the floor, serving both for heating the house and for some of the cooking. Cooking on a large scale, such as for feasts, was done outside of the house on an earth oven. Each dwelling was usually large enough for one or more families, but there were one or two very small houses that may have served for the isolation of menstruating women. There was no settlement pattern as far as alignment of the houses along a street is concerned, but the houses were more or less staggered along the river bank. The village was divided into two parts, presuming of course, that all or nearly all of its houses were occupied at the same time" (Shiner 1951:16-17).
He also worked at 45BN6 on the Washington side of the river. This village contained about 60 houses. It also consisted of houses along the river terrace. Houses were of different sizes. House pits 5 and 6 were about 13 feet in diameter and 2.5 feet deep. Both houses had several floors showing up as brown streaks in the profiles. No evidence for post holes or structure were recovered. House 7 was 17 by 15 feet and 2.5 feet deep. It had three floors. House 59 was 65 by 16 feet and thought to be a large mat house. The floor was only about 1 foot deep and fits descriptions of Umatilla houses. The houses contained trade goods.
Upper John Day Reservoir
Dave Cole did research on the Upper John Day reservoir in the 1960's for the National Park Service. Site 45KL5 on the Washington side of the river was excavated in 1964-65. Fourteen pit house floors were recorded during the work. Of the 14 house floors, only four had not been impacted by looting. Five house pits were partially excavated (Cole 1966:5).
"Several house types occurred at Site 45KL5, the most common of which was a circular house that was basin shaped in profile. The largest house excavated, feature 38, was this type. It was 5.6 meters in diameter. Approximately 25% of this house had been removed by pothunters and by clearing the site, in addition the floor had been disturbed by six intrusive pits and the rim had been truncated by erosion and other disturbances, hence very little of the house remained" (Cole 1966 5-7).
"One group of house floors, features, 31, 31B, 31C, 31D, and 31E appear to be a succession of floors within a single house rim. Much of this group of house floors was either disturbed by pothunters or was unexcavated" (Cole 1966:7).
"Perhaps the most significant characteristic of the circular basin-shaped house is that all the houses, which lack fire lenses and abundant floor debris, are of this type" (Cole 1966:7).
"Two house, features 10A and 33A, can be characterized as being rectangular and basin shaped in cross section. Feature 10A had a suggestion of a double rim or a slight bench on one side. This house had been disturbed by pothunters along one side and another side was indefinite, but it measured at least 4 meters in each direction and was at least 42 cm deep" (Cole 1966:7).
Feature 33A was one of the larger houses excavated, measuring 5.11 meters long, 4.36 meters wide and 40 cm. deep. This house was only slightly disturbed by pothunters and intrusions. It was the lowest pit in the site, excavated so far, and possibly the earliest. There was a fire pit and a scattering of charcoal on the floor, but the carbon sample from it was contaminated in the testing laboratory and could not be used for dating" (Cole 1966:7-8).
Cole 1966:no page numbers
The house floor contained a hopper mortar, a bone flesher, a maul, a projectile point, a notched sinker and several flakes. It had a radiocarbon date of 1720±120 (GaK 916).
The illustration shows this house as recorded by Cole.
Umatilla Village 35UM1 (35UM35)
The Umatilla site (35UM35) was excavated on the Oregon side of the river near the mouth of the Umatilla River. While the report and artifacts were given a separate site number, this site was later incorporated into 35UM1, a huge housepit village site containing hundreds of houses and burials. A single house was dug in this project. The house was described as the earliest evidence found in the testing project.
"The floor was approximately 4.3 meters long and 3.6 meters wide and the pit was at least 40 cm. deep. The fill immediately above the house floor contained charcoal material related to the house floor, other than this the floor was isolated by sterile deposits both above and below, hence a single cultural unit is represented here... The house had shallow post holes in evidence along two walls. There was a depression in the center of the floor which contained ash and charcoal and charcoal was scattered over the floor around the pit. The fire pit was closer to the southern end of the house but along the center line. A radiocarbon date for this charcoal was 2420±120 (GaK 914)" (Cole 1966:16). Artifacts found inside the house included projectile points, shell beads, bone tools, four choppers, eight utilized flakes, five knives and three scrapers. The following illustration shows the rectangular shape of the house.
Schalk (1979:28-34) summarized the prehistory of the region around 35UM1 (aka 35UM35) for what is termed the "winter village pattern". The presence of any pithouse structures has been implied to define this pattern based on ethnographic parallels. Village sites were noted as ranging in size form a single house to around 180-200 houses. Such sites are found near the moths of tributary streams, floodplains in major canyons, and near rapids where fish were taken. Some feel that this pattern represents an economic dependence on fishing. Researchers have noted what appears to be an increasing number of houses and an increase in the number of houses in villages over time, and the very large villages only appeared around the time of the arrival of the horse. Schalk noted that "features which have been referred to as pithouses are almost certainly variable in their size, shape, depth, wall construction, kind of superstructure, associated features inside and out, and their special relationship to other houses within sites. Such variability appears to have both temporal and geographic patterning that is presently known in only the vaguest way" (Schalk 1979:29).
"Generalizations about pithouse size have usually drawn from partial excavation of one or only a few sites and there has been no systematic and comparative study of regional variations in house size or form. In their work on the McNary Reservoir, both Osborne (1957) and Shiner (1961) concluded that house size tended to decrease through time. Others have suggested virtually the reverse sequence (Strong et. al. 1930:37). Nelson (1969) has proposed a linear sequence of house types with vertical-walled, flat-floored, and benched variations occurring early followed by vertical-walled with no benches, and culminating near the end of the prehistoric sequence with a shallow, saucer-shaped cross-section variant. Others believe that contemporary occupations include such different house types, and, therefore, that the linear sequence is not valid (c.f. Holmes 1966: 110)" (Schalk 1979:30).
"House shape is generally characterized as circular. Where house perimeters have been adequately outlined, however, it turns out that many tend toward a rectilinear form. In the John Day reservoir, several houses have been excavated which are nearly square or rectangular, or even pentagonal in form (Cole 1967, 1968; Ken Beste 1974:8)" (Schalk 1979:30).
Schalk also notes that superstructures tend to be light and simple. Houses show re-occupation, often over previous floors without cleaning them out... just reuse of an existing pit. He notes that some sites suggest that houses nearest the river contain the greatest accumulations of trash deposits. Storage pits also seem to increase in number in houses nearer the river. Sometimes larger houses appear to be associated with one or more smaller structures. The smaller have been suggested to be for dependents, sweat houses, or for storage.
In the 1970's the Mid-Columbia Archaeological Society excavated about 2,832 cubic yards of 35UM1. Many houses were dug (20 mentioned) and perhaps 150 features. Schalk estimated that perhaps 176,335 cubic meters of site remain intact in this huge village. He recorded eleven housepits in profiles from looter pits or slumps. Houses appeared to be essentially continuous for over 720 meters and well back from the river edge. In their 1973 field notes, Kenneth & Lois Beste note: "We uncovered parts of six pit houses in this area... House pit No. 1 appeared to be the oldest dwelling in this trench. Its lower side wall had a sixty-five degree slope, which extended half way up, then was a slope of forty degrees. Its floor was one hundred two inches from the surface. There were twenty-six inches of yellow clay silt below this floor to river gravel.... House pit No. 2 was used at a later date than No. 1, as it had been superimposed over house No. 1. It also had the same type of sloping wall. From the few artifacts found in these two houses, it would appear they were from the same general time period. The floor of this second house was eighty-seven inches from the surface and covered with ash even on the lower side wall. This indicated that the house may have burned down and fallen into the pit. We were unable to get an accurate measurement, but theses houses appeared to be about twenty feet in diameter. We believe houses 3, 4, 5, and 6 were all from the same general time period and much later than houses 1 and 2" (Beste 1973: 8).
"House floor No. 3, which was sixty-seven inches from the surface, contained a rectangular storage pit thirty by forty-eight by twenty-nine inches deep. This was in the west third of the house. No artifacts or ashes were found in this storage area. This house was circular and about thirteen feet in diameter. The floor of house No. 4 was also sixty-seven inches from the surface and was twelve feet ten inches in diameter. It appeared to be octagonal in shape, as we uncovered five straight sided walls in the part excavated. The wall contained a fifteen inch ledge from ten to twelve inches above the floor. The upper walls ranged from fifteen to seventeen inches in height... The fifth possible house intrudes into house No. 4, but due to lack of time, we were unable to gather sufficient information to evaluate either Nos. 5 or 6" (Beste 1973:8).
"Trench F-G-H, which was incomplete, contained signs of other pit houses. The floor of one of these was one hundred twenty inches from the surface. This house was very likely under part of house No. 1" (Beste 1973:8).
David Rice also wrote up a short report on the MCAS excavations. He noted: "On the basis of form and stratigraphy the housepits seen to fall onto two distinct time periods. The earliest house structures are deep, circular, benched pits with central hearths and/or storage pits. The later house structures are shallow, circular and without benches. they are often associated with hearth areas, but rarely storage pits" (Rice 1971: page 6 but no page numbers).
In a 1979 survey for restoration projects on Elk Creek, Tom Churchill recorded a site with three housepits and three possible sweat house depressions. The possible housepits were arranged in a row above Elk Creek, and the three sweat house depressions were closer to the creek and adjacent to a tributary stream. No measurement were taken of the depressions (Churchill, 1979).
Hells Canyon Creek
Pavesic excavated several houses on Hells Canyon Creek in the 1980's. In area A, the West house was 4 by 3.75 meters and had been heavily vandalized. The depression was 90 cm deep. The rim was indistinct but burned beams and grass suggested a pole and grass or sod covering. South house was 8 by 7 meters. Excavation of the rim revealed a bench with large rocks placed there. North house was 6.5 by 7 meters. Again, large angular rocks were found around the perimeter of the house. The floor was 77-90 cm deep. A pole fragment and post mold were also found as well as a cache pit. In a different area, the North house was 7.7 by 6.5 meters. Another rock concentration along the rim was found as well as a well defined hearth area 118 cm in diameter. The house also had a cache pit. The area B South house had the same rocks around the rim area and a hearth.
"The architectural features of the prehistoric houses represent the most important finding of the filed excavation. The massive rock rims that appear repeatedly in South and North houses, Area A, and the North house, Area B, cannot be considered accidental or a natural sedimentary unit. The size of the individual rocks and the consistent placement beyond the edge of the house fill supports the interpretation of a cultural feature. The rock rim of North house, Area A, also revealed a burnt, vertical 'stake' within the unit and a possible post mold noted along the occupational edge of the fill in pit C12. Of additional importance was the recovery of support poles and beam fragments in west house, Area A, associated with burnt grass blades suggesting a possible sod covering.... Cache pits were commonly identified throughout as having a large, flat stone covering and small rocky or pebbly bottoms. It was also revealed that processing stones, mortar bases or anvils, were purposely set in the house floors" (Pavesic 1986: 42-43).
35WA814: Magden/Wallane Land Exchange
Dan Klug with the Vale BLM recorded site 35WA814 during the Magden/Wallane Land Exchange project. The depressions were described as "possibly four semi-subterranean house pit depressions approximately two meters in diameter and one meter deep" (1992: site form). They were in a scatter of flakes with depth. Site 35WA815 contained two depressions again 2 meters across and 1 meter deep! Artifacts were associated with the depressions.
Keith (1993) recorded site 35MW103 with 44 depressions, because of the number of features, they were not measured nor depths taken. The site form notes they ranged from 1.5 to 4 meters across and 20-30 cm deep. FCR was noted in some depressions. She also recorded (Keith 1994) site 35MW135 with 32 surface features in four clusters (cluster #, E-W, N-S, and depth):
|1||2.4 m||2.3 m||.24 cm||2||2.8 m||2.8 m||.13 cm||3||3.7 m||4.5 m||.37 cm|
|1||2.3 m||2.0 m||.20 cm||2||2.4 m||2.8 m||.08 cm||3||3.4 m||3.4 m||.28 cm|
|1||1.6 m||1.4 m||.31 cm||2||2.7 m||3.8 m||.25 cm||3||3.6 m||3.4 m||.08 cm|
|1||3.2 m||3.6 m||.10 cm||2||3.3 m||2.6 m||.46 cm||3||3.4 m||3.5 m||.24 cm|
|1||2.5 m||2.6 m||.21 cm||2||1.5 m||2.4 m||.27 cm||3||3.7 m||3.5 m||.21 cm|
|1||1.8 m||1.7 m||.10 cm||2||2.3 m||2.6 m||.28 cm||3||2.8 m||3.0 m||.23 cm|
|1||2.0 m||2.5 m||.17 cm||2||1.8 m||1.9 m||.13 cm||3||2.8 m||3.0 m||.23 cm|
|1||2.3 m||3.1 m||.11 cm||2||---||---||---||3||4.7 m||3.1 m||.21 cm|
|1||2.1 m||0.9m||---||2||---||---||---||3||2.8 m||3.2 m||.18 cm|
|1||1.8 m||2.1 m||.23 cm||3||3.7 m||3.5 m||.33 cm||4||3.0 m||2.5 m||.33 cm|
|1||1.5 m||1.8 m||.22 cm||3||4.1 m||4.7 m||.31 cm||4||1.5 m||1.2 m||.18 cm|
|1||2.4 m||2.3 m||.10 cm||2||---||---||---||3||3.7 m||3.7 m||.13 cm|
|1||2.0 m||2.3 m||.09 cm||2||---||---||---||3||4.2 m||4.4 m||.40 cm|
McGraw Creek Site 35WA1
Claude Warren, Robert Yohe and Max Pavesic published in 1999 the excavations at the McGraw Creek Site (35WA1) that had been done in 1966-1967 by Idaho State University by Robert Blankenship for his unfinished master's thesis. Three houses were partially excavated in 1966 and house clusters sampled in 1967. The north house was dated to 1570±60, the center house to 920±50 and 780±60, and the south house to 770±70. The north house contained a cluster of stacked rock and mussel shell and an inverted hopper mortar base. The house also contained three possible hearths and hopper mortars. The central house contained a hearth, a burial, three inverted hopper mortars, a storage pit and a cairn-covered storage pit. The south house a cluster of FCR, and photos of probably inverted hopper mortars.
The projectile points from the houses were dominated by barbed-shoulder and expanding stem points but Desert side-notched were found in all three. An anthropomorphic bone pendant was also recovered as well as bear canine pendants, as well as disk, tubular and tube beads.
Ferry Canyon Site 35WS261
Aimee Di Scipio (1997) described testing of the Ferry Canyon Site (35WS261). Five carbon dates were taken: 730±50, 780±90, 850±80, 610±50 and 660±80. Three houses were bisected with block trenches. The site was interpreted as a Tenino winter village placed for its proximity to hunting, water, wood, fishing and river mussel resources. There were no measurements of house size in the thesis, but the grid suggests the house depressions were roughly 5 meters across. The houses had been heavily looted and no internal features were reported.
« Previous Region | Index | Next Region »