Western Aleutian Archaeology and Paleoecology Project

In 1991 two archaeologists, and a biologist began archaeological fieldwork to study the historic distributions of seabird species in the western Aleutians. The project begun that season has since grown into an international, interdisciplinary research program aimed at understanding the human and environmental history of the Western Aleutians.

The work grew out of biologically oriented research on archaeological collections. In the mid-1980's Dr. Douglas Siegel-Causey, then at the University of Kansas, had a contract to re-analyze large collections of bird remains from the Aleutian Island and Bering Sea regions made by Ales Hrdlicka and Henry Collins in the 1930's and 40's. The remains had been reported in several articles by Friedmann (reference) but were badly misidentified. In 1987 Christine Lefevre, a zooarchaeologist at the Museum national d’Histoire naturelle, in Paris, and Siegel-Causey collaborated on an analysis of cormorant remains from the southern hemisphere using collections housed at the University of Kansas Museum of Natural History. Siegel-Causey began planning a similar project for northern hemisphere sea birds. Lefevre and Siegel-Causey then analyzed bird remains from archaeological excavations made on Amchitka Island in the late 1960's and early 1970's, as a result of underground testing of nuclear weapons. The collections had been scattered around the country and were fraught with problems derived from poor excavation technique and curation. Eventually about a third of the faunal collection reported from the excavations was located and analyzed.

Vernon Byrd, the chief biologist for the U.S. Fish and Wildlife Services, Alaska Maritime National Wildlife Refuge, which encompasses the Aleutian Islands, suggested new excavations and mentioned an eroding site, rich in bird remains, on Buldir Island. Siegel-Causey contacted the University of

Alaska Fairbanks to find an archaeologist interested in collaborating on a biological project. Debra Corbett had just finished a Masters degree on Western Aleutian Island settlement patterns. As part of the research for the thesis she had excavated a single site on Shemya Island and sent the bird remains from that excavation as an additional sample for the biological study.

From 1991 through 1994 the work was funded by grants from the National Geographic Society, with support from the US Fish and Wildlife Service, the Museum national d'Histoire naturelle in Paris, and the Natural History museum at the University of Kansas. Project fieldwork began on Buildir and continued on that island in 1993. In 1992 we exscavated a single site on Little Kiska in the Rat Island group, east of the Near Islands. In 1994 the group returned to Shemya to expand the work begun by Corbett in 1989. The teams archaeologists and biologist were joined by geologist Thomas Corbett in 1993 and by Russian paleoenvironmental researchers Arkady Savinetsky, Nina Kisselova and Bulat Khassanov from the Academy of Sciences in Moscow, in 1994.

The project gained a name and entered a new phase when Dr. West received an NSF grant to study human ecology, culture history and paleoenvironmental problems in the Western Aleutians. The Western Aleutians Archaeological and Paleobiological Project, WAAPP, returned to Buldir in 1997 to continue excavations and begin paleoenvironmental research on that island. The team now consists of archaeologists from the Fish and Wildlife Service, National Museum of France, and the University of Kansas, biologists from Russia and Nebraska, and geologists from Russia and Alaska.

The scope and complexity of this project have expanded annually since the simple beginning in 1991 on Buldir. Research that year focussed on

establishing a site chronology and evaluating the sites research potential. Preservation of organic material in the site is remarkable even by the generally good standards of the area and we recovered an enormous number of well preserved plant and faunal remains, mainly birds, which formed the basis of the subsequent environmental research. Research also focussed on establishing a site chronology. In 1993 we again approached Buldir with very general research questions. In addition to refining our knowledge of the culture history, we began to look at economic strategies to understand how Aleuts adapted to the local environment. We wanted to know: 1) why Aleuts used such a remote landfall, 2) how regularly the island was used, and 3) whether the occupations were seasonal. We are still trying to learn who used the island, and why.

By the 1997 season project goals had been more closely defined: The three primary goals are 1) to document Holocene environmental change in Beringia and determine to what extent observed changes can be ascribed to natural factors or to anthropogenic factors, 2) to study Aleut economic strategies to understand the effects of regional environmental change on the culture, and finally, 3) to examine evidence for contact between the Rat and Near Islands to determine the nature and extent of cultural exchange between politically and culturally independent Aleut subgroups. Buldir Island, the only landfall in a long stretch of empty ocean between Kiska in the Rat Islands and Shemya in the Near Islands is uniquely placed to address questions about interisland contacts. Along with these goals we are continuing work on the chronology of occupation in the Near Islands, including the timing of the initial colonization, and are examining the complexity of social and economic adaptations

Research on Buldir Island

The Western Aleutian Archaeology and Paleobiology Project is focussed on

the Near Islands and Buldir, the westernmost pieces of land in the Aleutian Chain. Archaeological research has a long history in the area, beginning with W.H. Dall in 1880 and carried on through Waldemar Jochelson in the early 1900's, Ales Hrdlicka in the 1930's and Albert Spaulding in the late 1950's. Still, little is known of the culture history and development of the Near Island Aleut culture and its place in the broader Aleutian context. Recognised by Aleuts and scholars alike as the most divergent group of Aleuts their real history is virtually unknown.

Because Buldir had never suffered from fox predation it is considered a "relatively pristine remnant of the Aleutian ecosystem" (Byrd and Day 1986) and its extant faunal population was felt to be an excellent analogue for past biodiversity in the Aleutians. The island hosts enormous numbers, over four million, of breeding seabirds. This made Buldir an excellent place to compare modern fauna to past seabird numbers from the midden. The modern fauna includes 32 breeding species, of which 65% are seabirds. Over 99% of the total population are storm petrels (Oceanodroma furcata and O. leucorhoa). Twelve species of Alcids make up 88% of the remaining bird population with gulls and kittiwakes the next largest group.

The only known site on Buldir is a small to medium sized midden situated on two parallel beach ridges (Figure 3). The main occupation area is between two small streams but the site extends across the mouth of the valley. Nearly a meter of windblown sand covers the site obscuring surface features but six possible housepits were defined along the rear beach ridge and one was visible in an eroding beach profile.

Siegel Causey, Lefevre and Corbett spent 10 days on Buldir in August of 1991 and excavated two small test pits there. The first season we cleared 11 profiles along the beach face. After examining the profiles two, labelled Pits 1 and 2, were selected for small, one meter square, test excavations.

These pits revealed up to seven alternating cultural and sterile sand layers. We returned to Buldir in 1993, joined by taphonomist and zooarchaeologist Dr. Dixie West of the University of Kansas and geologist Thomas Corbett of the consulting firm WGM, Inc., in Anchorage. Three additional pits were excavated, two, Pits 3 and 5, trenched the forward beach ridge and Pit 4 exposed a structure of whalebone. In 1997 we enlarged the excavation of Pit 4, and, to sample the older levelsdeposits found originally in Pit 2, opened Pit 6 and 7 near the east end of the midden.

Unusual depositional and preservation factors enhanced the value of the site for study. First, in all pits the cultural levels are separated from each other by layers of sterile sand. This eliminated the problem of mixed deposits associated with most Aleut middens. The cultural levels are distinguished by a black, dense, greasy feeling soil full of pieces of wood, worked wood, unmodified bone and some stone tools. Second the deposits contain unusually well preserved plant remains, hair and eggshell, and egg membranes in addition to large quantities of cultural wood. This unusual preservation is due to constant percolation of water through tye beach ridges from the marsh inland, combined with anaerobic conditions caused by the decay of the masses of organic materials which used all the soil oxygen and stopped decomposition.

Dating the SiteAs the site consists of a series of discrete occupation levels separated by sterile sands it is easy todefine discrete episodes of occupation for the island, making chronological and typological work exceptionally easy for an Aleutian site. We are still refining Buldir’s chronology but so far have two well dated late prehistoric occupations. The radiocarbon dates for Pits 1, 3, 5, the top level of Pit 2, and the whale bone house range between 530 and 220 years before present with the corrected dates clustering in the mid 17th century, around 1640 AD.

Pit 6, and the upper part of Pit 7 date to the mid 14th century. These Pits have a different suite of wooden and stone tools from the 17th century levels. Whether these differences are simply temporal, or whether they represent use of the Island by the peoples of the Rat vs the Near Islands remains to be determined. The 17th century occupation appears to show joint use by both groups, while the 14th century level may be dominated by Rat Island material culture. The nature of the material used for lithic tools on Buldir, friable, platey phyllite, makes direct comparisons with neighboring islands difficult.

We have one late 12th century date from Pit 2. The middle levels of Pit 7 are as yet undated. The lowest level has returned a date of . We know that human hunters and fishermen occupied the Rat Islands to the east by 4800 years ago. The earliest dates to the west are 3500 years old and appear to correlate with a decrease in storm activity which perhaps permitted sea voyages to the Near Islands (Corbett et al:1997). We hoped the lower levels of Pit 7 would confirm this evidence of early colonization voyages. Interestingly in this earliest occupation level we find that people were exploiting the lithic material sources on the south side of Buldir almost immediately after landing.

We have not found any more recent occupations. Russian records in the 1700's record Aleuts living on Buldir. In the 19th century hunting parties of Aleuts were left on Buldir for years at a time. In the 1960's the refuge manager, Robert Jones, found and collected some copper scraps which are the only indication we have of historic use of the island.

Buldir Kis-008 1/1 wood 460+-50 Beta-54253Buldir Kis-008 1/6 wood 280+-50 Beta-54254

Buldir Kis-008 2/3 wood 530+-60 Beta-54255Buldir Kis-008 2/7 wood 1160+-50 Beta 54256

Buldir Kis-008 3/4 cut2 wood 530+-60 BetaBuldir Kis-008 3/4 cut8 wood 240+-60 Beta

Buldir Kis-008 4 charcoal 220+-70 Beta 108969Buldir Kis-008 4 charcoal 250+-70 Beta 108970Buldir Kis-008 4 bone 390+-70 Beta 108972

Buldir Kis-008 5/2 cut1 wood 350+-80 Beta-71566Buldir Kis-008 5/2 wood 320+-60 Beta 71567Buldir Kis-008 6/10 wood 630+-60 Beta 108965Buldir Kis-008 6/10 wood 760+-60 Beta 108966

Buldir Kis-008 7 charcoal 2347+-84 IEMAE-1177

Descriptions of the ExcavationsThe 17th century levels of Pits 1, 3 and 5 are characterized by a black, dense, greasy feeling soil full of pieces of wood, worked wood, unmodified bone and some stone tools. The deposits also hold plant remains, hair and eggshell, and egg membranes. Wet, anoxic soils are the reason for this preservation. In thin deposits, as in Pit 2 and in the house, the preservation is not nearly so good because they lacked the same anoxic conditions. This level is about 60 cm thick, thinning to the east and extends nearly 70 meters eatward from the western creek. It covers the first beach ridge and is only about six to eight meters wide.

These levels yielded a wealth of wooden artifacts. From the western Pits, 1, 3 and 5, the most common were a variety of slats, shafts and rods as well as fire drill shafts, and six different kinds of projectile point. Two pieces

recovered in 1993 were presumed to be models of kayak bow pieces. The quantity of carving debris, coupled with knives, wedges and scraprs suggest at least the western portion of this level was a wood working area. The kayak parts and several bone awls, for punching holes in hides, raises the intriguing possibility that kayaks were being built or repaired here also.

The range of materials recovered from Pits 6 and 7 greatly expanded the number of types of artifacts known. Among the new finds are several projectile point types, possible mask fragments, and pieces from two popular gambling games. The shaft and slats from the previously excavated pits were almost completely unknown in these excavations. We also recovered yet another bow piece and several deck beams of model sized kayaks, making at least three kayak models found so far. These materials are providing the only information available anywhere on western Aleut kayak construction and design.

Pit 4 was excavated into the only structure identified so far on the site. The meter of sand over the entire area has concealed most surface indications of pit houses. The structure is located on the south side of the midden overlooking the creek and marsh. The surface depression indicates the Buldir house measures about 10 meters long by 5 wide. It is oriented northeast to southwest and probably had the traditional roof entry.

In 1993 a 2x2m test pit exposed a cluster of whalebones and it became obvious the depression was a housepit. Stratigraphy in the house is complex but the floor is marked by a dense black organic rich layer. Dominating the east half of the 1993 excavation was a pit feature 1.5 m in diameter and at least 129 cm deep, lined with a thick mat of grass and hair and sea lion scapulae, held a medium sized whale skull placed snout downward in the center. Two smaller pits excavated into the sand floor held dense concentrations of bird and fish bone, with a few bone needles.

In 1997 the house excavation was expanded to a 3 x 8 meter excavation, exposing cobbles, griddle stones and bones from whales and other sea mammals. Most of the whalebones are embedded in a black, sticky clay layer, the presumed floor of the structure. Concentrations of bird and fish bones were observed under and between cobbles and larger whale and pinniped bones.. To the north of the whale skull found in 1991 was a large amorphous hearth. The skull would have been an a unavoidable focal point for all activities in the structure. Three stone lamps were found within the structure, one was decorated with pecked lines and grooves.

A double burial was found in the southeast corner of the house excavation. One burial, represented by two femurs, a humerus and phalanges, appeared to be a young adult. The position of the bones, with the head of the humerus near the distal end of the femur, suggests the body was tightly flexed. A smaller rib cage and spinal column is from a child. We had gotten permission from the Aleut people to conduct field analyses of human remains but our osteologist was injured getting to the field and returned home. Excavation in the unit ceased when the burials were recognized and the remains were reburied in situ. The presence of bodies in an Aleut house is not unusual although these two do not appear to have been buried, there was no evidence of a pit. Their bundled bodies may have been hung from the rafters or propped in the corner, a practice reported for the eastern islands.

Whale skulls are often associated with burials though there is no obvious connection to these burials. Whalebone is a common structural element in house remains all along the Aleutian chain but none of the houses excavated and reported have had a substantial whalebone component. A house with a whalebone superstructure was excavated on the lower Alaska Peninsula but is associated with an intrusive Thule eskimo occupation in an otherwise

Aleut area. An early historic house from Reese Bay on Unalaska Island, excavatd by McCartney and Veltre, incorporated two whale skulls and other skeletal elements in non-structural contexts. The Buldir house represents the first excavation of an Aleut whalebone house, and one of very few house excavations for the western Aleutians.

Pit 6 was located on the forward beach ridge. It measured 2 x 3 meters with the long axis running north/south. Each one meter unit within the pit was given a letter designation, A-F. Units were excavated in natural levels. Levels 1, 3, 5, 7 and 9 were sterile sands. Level 1 was the thickest with the aeolian sands streaked with buried soil levels with grass roots . These represent different episodes of dune construction then stabilization. The first three cultural levels were thin, and lacked good bone and wood preservation.

Level 2 held a feature consisting of a cluster of bone, griddle stones and lithics. An articlated sea lion flipper was found in Level 4. Wood preservation improved in Levels 6 and 8 and faunal materials also increase in abundance. Level 10 appeared to be the edge of a large hearth. These alternating levels appear to represent outdoor activity areas. They are not substantilly seperated temporally. The interveneing sand layers appear torepr esent single storms rather than breaks in the site occupation.

The Russian paleoenvironmental team dug four pits in non-cultural contexts then began a profile in the midden 20 meters east of Pit 6. This excavation was designated Pit 7. The profile measured 2 meters long along the bluff and was 1 meter wide. The Pit was excavated to a depth of 2 meters when time constraints led the team to reduce the profile to a 1 meter square. The total depth reached 4 meters and terminated in the same volcanic gravel found in the non-cultural excavations. Cultural levels were well defined and clearly separated by sterile sands.

The upper cultural levels, 2-4, yielded few artifacts and have not been dated. The next cultural unit, levels 6-11, are presumed on the basis of similarity in artifacts, to be the same age as Pit 6. Levels 13 and 14 are also cultural but are not as rich as the uper level. The lowest cultural level is dominated by a hearth and several griddle stones Interestingly, all of the lithic materials in the lowest levels are the locally available phyllites, indicating even the earliest occupants of the island were familiar with local resources and relied on them for tools.

Two fragments of Sea cow bone were recovered from the presumed 14th century occupation levels. We felt that one of the attractions for the earliest settlers to the western Aleutians would have been populations of sea mammals not yet accustomed to human hunters. These animals would include not only the usual sea lions and hair seals but also possibly fur seals, walrus and sea cows (Hydromalis gigas). We hoped to find some of these animals in our earliest levels to give us information on the paleoenvironment and confirm expectations about man the voracious hunter spreading havoc in virgin territory. At the time of Russian penetrationof the North Pacific in 1741 Sea Cows were confined to the Commander Islands.. The huge inoffensive animals were exterminated by Russian fur hunters victualling their ships within 26 years of discovery. The presence of living sea cows has been one of the strongest arguments for the Commanders being unknown to humans prior to the Russian arrival.

Finding sea cow bones in more recent levels would imply the animals were able to hang on in human occupied areas until quite late in prehistory and confirms Aleut stories of sightings and even of hunting the giants. Lucien Turner, a weather observer on Attu Island in the 1880s, reported stories that Sea Cows were present in the Near Islands and that the beasts, so inoffensive and helpless, were scorned by the men but hunted and

butchered by women, who gathered them from the intertidal zone like large shellfish.

ShemyaIn 1994 the Project moved to the Shemya Island in the Near Islands to expand the scope of the environmental and cultural historical research. Prior to the beginning of our project some limited survey and testing had been done on the island by the US Fish and Wildlife Service and the Bureau of Indian Affairs. In 1988 a BIA crew had worked on Shemya attempting to locate and record four sites applied for by the Aleut Corporation as Historic and Cememtery sites under Section 14(h)1 of the Alaska Native Claims Settlement Act. this section Alal\ska's regional Native Corporations could claim significant historic properties for transfer to that Corporation. The Aleut Corporation (TAC) applied for four known sites on Shemya. One site was found, SH-6, all others were presumed destroyed by military construction. In 1989 , Corbett, travelled to Shemya with a small team to evaluate that last remaining site. Three 2x1 meter pits were placed in the western end of the midden. While on the island contract construction workers alerted the team to the presence of at least two other sites, SH-7 and SH-8. Later while working with aerial photographs Corbett identified the locations of two other sites, SH-3, and SH-4, and, while working for BIA, again travelled to Shemya to record them for the Aleut Corporation. As part of this project, oral history on Near Island land use and subsistence was collected to supplement historical records.

Shemya is no ones idea of an ideal research target but it is accessible, and relatively inexpensive, to reach. However logistics in the Aleutians are chancy and expensive and compromises must be made. To continue fieldwork in 1994 on a low budget we had to select an island we could get to; Shemya. Dr. Doug Siegel-Causey, Christine Lefevre, Debra Corbett, and Dixie West arrived on the island in August and about a week later were

joined by Stephen Loring and a crew from the Smithsonian. Russian paleoenvironmental scientist, Arkady Savinetsky of the Russian Academy of Sciences in Moscow joined the team that year. The goal was to sample as many of the island's sites as possible to create a regional chronological and resource use picture. Large excavations were made in two sites on Shemya Island, SH-3 and SH-7. Another pit was excavated in the site previously tested in 1989 by Corbett. A small remnant of a previously unknown site was discovered and tested. By the time we left the island we had collected material to date six of the nine reported sites, and had excavated large faunal and artifact samples from three sites. We had also confirmed the destruction of three sites. Savinetsky collected samples of buried soils which have provided a detailed climatological record extending back to the end of glaciation on the island around 9000 years ago.

There are nine known archaeological sites and a single pre-World War II historic site on Shemya Island; all have suffered from modern disturbance. The historic site and three of the archaeological sites appear to have been totally destroyed. We know of their existence from reports by construction workers, and through the use of aerial photography. All of the known sites are described here.

The first survey was made by T.P Bank in the 1950's. Banks work was never reported but A.P. McCartney, compiling a list of all known sites in the Aleutian chain, used Bank's notes and reported four sites on Shemya (McCartney 1972). In 1965 Mike Aamodt, an anthropology student at the University of Alaska Fairbanks, working construction on Shemya made a report on collections he salvaged during the summer. Because the sites were subjected to continuous damage from construction and looting, he salvaged over 750 artifacts and described five sites, two of which no longer exist. Finally, in 1990 Debra Corbett, a graduate student at the University of Alaska Fairbanks, spent 3 weeks excavating at ATU-006 (Corbett 1990).

This work, reported only preliminarily, supplemented the 1994 excavations.

Sites 1, 2 and 5 have not been located by archaeologists in spite of repeated searches and are believed to be totally destroyed.

Site SH-4 is located on the eastern coast of Shemya Island, midway along a crescent beach. The site sits on a wave cut platform at the base of steep bluffs bisected by a small intermittent stream. From aerial photographs we know that the site stretched for over 250 meters along the central portion of the crescent beach. A smaller midden occupied a point about 250 meters to the east of the main site. Together these two middens made up the largest site on Shemya. Today the wave cut platform has been partially destroyed by earth moving and road and building construction. The smaller midden has been completely destroyed. The larger midden is visible only in test pits, road cuts, and vandalized holes, no surface signs of a site are apparent.

The site was reported as a small midden by T.P. Bank (McCartney 1972). Aamodt (1965) called it the North-northeast site, and estimated its size as over 100 meters long and up to 5 meters in depth. He collected 487 (53%) of his artifacts from this site. The Aleut Corporation claimed the site under section 14(h)(1) of ANCSA. It was relocated in 1990 by Corbett (1990), using aerial photographs dating from 1943. Though heavily damaged by construction activity and past vandalism, in 1990 the site appeared to be unknown to island personnel. During the BIA investigation also in 1990 three small test pits were excavated to assess the integrity of the remaining midden deposits (U.S. BIA 199?).

In 1994 the site was considered for excavation but rejected in favor of SH-3, which appeared to have more integrity. Reconnaissance of the site revealed recent vandalism and looting. U.S. Air Force Security Police began patrolling the site and posted signs making the area off limits. Follow up

investigation in 1996 suggests the vandalism has stopped. One collection of 236 artifacts was recovered from a looter by U.S. Fish and Wildlife Law Enforcement officers in 1994. The looter forfeited the artifacts to the government and paid restitution of $500.

Three small test pits and some brief notes on looters' pits provide the only information on the statigraphy and age of the site. The test pits, all 50 cm square trowel tests, were excavated by BIA archaeologists in 1990 to help determine the areal extent of the site and, if possible, provide material for a date. Lithics and faunal material were collected from test pit 2. The faunal meterial was submitted for dating and produced an uncorrected date of 2680+-70 BP (U.S. BIA 199?).Virtually nothing can be said about the history of this site beyond concluding that undisturbed deposits do exist. The evidence suggests the midden recorded a complex depositional history. The remaining deposits probably represent the earliest occupation of the site' upper, more recent deposits having been destroyed by construction.

A larger excavation in 1994 produced three dates from the bottom f a 3 meter sequence.

SH-6 (ATU-003) is located on the southwest coast of Shemya Island, at the west end of the main east/west runway. McCartney (1972) first reported this as a large site based on unpublished field maps by Bank (1948). Aamodt (1965) called it the South Site and collected 109 bone and 86 stone artifacts from the site. The Aleut Corporation claim was investigated by the BIA in 1988.

In 1990 Corbett returned to Shemya to conduct excavations to determine the research potential of the heavily damaged sites on Shemya. At that time SH-6 was thought to be the only site remnant on the island. The work,

sponsored by the U.S. Fish and Wildlife Service, was also part of Corbett's graduate research. Three 2 x 1 meter pits were excavated in the western end of the site (Corbett 1990). In 1994 we returned to SH-6 and reopened the pothole tested by BIA archaeologists in 1988 to obtain additional materials for dating the site and to increase the artifact sample size.

Corbett (1991) estimated the site measured 160 meters long and 70 wide, covering about 7,000 square meters. Like SH-4, SH-6 is one of the largest sites on Shemya. The creek on the western edge of the site, constituted the largest and longest stream on the island, draining a series of lakes. This system may have supported salmon runs in the past. Today the streams are rerouted or dry. Approximately 80% of the midden is gone. The largest remnant, extending along the coast road, is 120 meters long and up to 11 meters wide. The smaller section measures 18 x 6 meters.

Three profiles, four test pits and a small trench have been excavated in SH-6. These investigations are concentrated in the western half of the midden with a single profile in the smaller midden remnant. The three 1990 test pits (Pits 1-3) were also located at the west end of the midden. Extensive samples of faunal remains were collected for each level of these three pits. The 1994 test pit (Pit 4) was excavated into a looters pit on the top of the midden, near the western end of the site. This pit expanded the 1990 BIA. The pit was excavated in arbitrary 20 cm levels. Faunal remains were collected to provide samples for dating.

Site history is difficult to interpret because an unknown portion of the upper levels of the midden has been removed by construction and vandalism. The site may have begun as a camp about 2500 years ago, represented by hearths and thin lenses of sea urchin shells in Pit 4. A period of disuse is suggested by a layer of sterile sand. A more extensive, and intensive, occupation is evident between about 2200 and 1500 years ago with another

layer of sterile sand suggesting another interrupted occupation beginning about that time. The site was occupied after this hiatus but the disturbance to the midden does not allow any estimate of the timing of this reuse.

SH-7 (ATU-061) is located on the southwest coast of Shemya Island, west of Laundry Lake. The site is not visible in aerial photographs nor was it reported by Bank or McCartney. Aamodt (1965) called it the South Southwest Site and noted that it sits about 100 meters back from the coast. He collected 40 bone and 34 stone artifacts and salvaged a flexed burial exposed by road construction. The site was relocated by Corbett et al in 1990 with the help of contract construction workers. The site was one of two known artifact collecting localities on the island (Corbett 1990). BIA excavated two small test pits in order to assess site condition and collect material for dating (U.S. BIA 199?). The tests revealed undisturbed stratigraphy and yielded a date on bone of 3540+-60 BP . This date is nearly 1000 years older than Spauldings date of 600 BC for Krugloi Point, Agattu.

The site presently measures 70 meters long and 45 wide and was probably one of the smallest on Shemya. The unusual site placement has been noted, it sits well back from the present shoreline. The site overlooks a cove defined by Skoot Island to the west and a rocky headland to the east. To the east, 2-300 meters, a creek draining Laundry Lake enters another small cove. Closer to the site, on the southeast edge, a spring is evident. Virtually the entire surface of the site had been removed by construction before 1947. In 1944 the area was used for ammunition storage (Map 1944). The earliest aerial photographs show the site area striped with trenches from earthmoving equipment. In addition the site has been subjected to illegal relict hunting, including attempts with heavy equipment, for years. Efforts by the U.S. Air Force and the U.S. Fish and Wildlife Service in 1994 have evidently stopped the looting

In 1994 archaeologists returned to SH-7 and opened one large excavation pit, two smaller pits and cleared a profile in a pothole (Siegel-Causey et al. 1994). The main excavation consisted of two stratigraphic units. The first is an intrusive feature, probably a house pit dating to about 3000 ybp. Contents included a single lamp and a few stone tool fragments. Few structural details were noted. The house was excavated into midden and along the east wall of the Excavation unit we noted post holes. The south wall profile showed a shallow trench along the edge of the structure wall, as is described in most discussions of Aleut houses. The second stratigraphic unit is the surrounding midden consisting of altermating bands of soil, sea urchin shell, and fish bone. At the bottom, overlying sterile redsand was a dense, black, greasy organic rich soil with abundant fish bone. On top of the sand was a lens of urchin with the remains of several adult and subadult fur seals.

Two dates run on bone from the structure are , 2570+-140 (IEMAE-1206) and 3080+-110 BP (IEMAE-1205). One date on bone collagen from below the intrusive feature is 3096+-155 (IEMAE-1175) Two additional dates from a column sample adjacent to the structure are 2630+-60 BP (Beta 108967), and 3120+-80 (Beta 108968). All of these dates are roughly contemporaneous., and confirm this as the oldest known site in the Near Islands. The entire excavated portion of the site is considered a single cultural unit spanning a few hundred years. It may have begun as a camp 3500-3000 years ago.

SH-8 (ATU-062) is located on the southwest coast of Shemya Island, near the end of the abandoned east west runway was not reported by Bank , McCartney (1972) or Aamodt (1965). The site is not visible on World War II aerial photographs. Corbett (1990), with the assistance of contractors, relocated the site in 1990. The site was considered the best local source for

artifact collecting. According to local legend the site had been bulldozed off the bluffs and totally destroyed. This may be the site Aamodt reported as bulldozed into a bay with at least 100 skeletons destroyed. A single small test pit was excavated to evaluate site condition and collect materials for dating.

From the meager surface remains we estimate the site measures about 100 meters long and 25 wide overlooking a small, reef-choked cove defined by Skoot Island to the south. Construction of the causeway to Skoot Island may have precipitated large scale damage to the site. At some point, possibly to use as fill for the causeway, the entire surface of the site was bulldozed. In addition the site has been subjected to illegal relict hunting for years. Today all that remains is a shallow remnant of the former midden on the edge of the bluff.

A single test pit was excavated in 1990 to assess site condition and recover materials for dating. The pit was 1 meter square and reached a depth of 75 cm before terminating in sterile red sand and gravel. The first 50 cm are disturbed, homogenous brown soil with sea urchin shell, other shell and fish and sea mammal bone. Between 50 and 55 cm the soil became a dark brown sandy horizon. This level, less than 15 cm thick, may be the undisturbed base of the midden. A single charcoal sample from this level returned a date of 2060+-90 BP (Beta-40424).

SH-9 (ATU-066) on the south coast of Shemya Island, just west of the sewage lagoon, was located in 1994 by Arkady Savinetsky during a survey for paleoenvironmental sample sites (Siegel-Causey 1994). The site sits atop 50 foot bluffs overlooking the valley of the largest stream system on the island. The current site measures 20 meters along the bluff and 5 meters wide. It is a maximum of 50 cm deep. Quarrying of the bluff exposed a series of buried soil horizons that proved invaluable for the

planned paleoenvironemental work. A single test pit was excavated in 1994 to recover materials for dating. A sample of faunal material returned an uncorrected date of 2244+-182 BP (IEMAE-1176). Existing remains suggest the deposits either sit at the edge of a onec much large site or that the entire site was a short term seasonal camp.

Research in the Aleutians is heavily dependant on available transportation and sometimes research goals must be modified to reflect the logistics. In 1992 we could not return to Buldir but could reach Little Kiska. The team, joined by Stephen Loring from the Smithsonian Institution, Arctic Studies Center, spent two weeks on Little Kiska excavating a site first explored in the late 1930's by Dr. Ales Hrdlicka and his boys. We hoped to be able to find materials to put this earlier work into a more useful context and to provide well dated material for comparison with Buldir.

Environmental reconstruction

A major thrust of our work is environmental and economic. During the Holocene the Bering Sea region experienced rapid climatic change combined with geological restructuring of the landscape to produce environmental changes on a scale almost unsurpassed anywhere else on the planet (Hopkins 1972; 1979). Most marine organisms have narrow limits of tolerance for changes in temperature and salinity (Odum 1959). Even minor changes in ocean conditions therefore, would have a clear measurable effect on sea life, including sea birds. Paleobiological investigation of faunal remains, especially birds, connected with anthropological study of the human context, can give us a richly detailed picture of Holocene environmental changes in Beringia.

Environmental reconstruction involves two distinct efforts. The Russian team examines buried soil horizons from a variety of non-cultural depositional environments for information on changes in rainfall and temperature. Soil profiles from Shemya, Adak and Buldir are being analyzed to create a regional paleoenvironmental framework. The second effort involves analysis of faunal remains, especially sea birds, from archaeological sites to correlate changes in species composition and relative abundance with environmental factors. Using ethnographic information and zooarchaeological analysis we can factor out the human selective behaviors which affect the composition of the archaeological faunal assemblage.

Fieldwork on Shemya resulted in an environmental record spanning 9,000 years. Arkady S. located and sampled two long soil profiles, one at site SH-9 and the other in a stream valley on the west coast. Rainfall estimates were derived from analysis of soil grain size and rate of deposition. Temperature information from the ash content of vegetation within the soil samples indicates an oscillating cycle of warming and cooling periods. These profiles show a steady decline in the amount of rainfall and an increase in average annual temperature beginning about 4500 years ago and leveling off to present day levels shortly after 3000 tears ago.

In the earliest stage of the project, faunal material from previous archaeological work at Amchitka in the late 1960's (Desautels 1970) and Shemya (Corbett 1990), was analyzed. The Amchitka materials were excavated from sites 49 Rat 31 and 49 Rat 36. To increase sample sizes for analysis several excavation units in Rat 31 were combined; dates are therefore broad estimates. All strata in Rat 36 were combined into a single analytical unit with a single date. All species are reported as Minimum Number of Individuals (MNI) (Table 1). The Amchitka samples had several problems which limited their reliability. The collections had been divided

into at least three samples and sent to different Universities. Provenience information was lacking or confused on many of the bags. Keeping these limitations in mind the Amchitka samples analyzed still provide invaluable preliminary data.

Amchitkan avian faunal remains clearly demonstrated changes in the distribution and abundance of breeding seabirds over the last 2,650 ybp (Siegel-Causey et al. 1991). Relative abundances of three cormorant species, Pelagic Shag (Stictocarbo pelagicus), Red Faced Shag (S. urile) and Double-crested Cormorants (Hypoleucus auritus) have remained similar throughout this period and their numbers remained relatively constant for the last 1,800 years. In contrast Kenyon's Shag (S. kenyoni) appear to increase, the greatest jump occurring in the last 200 years. A single wing bone of Pallas' Cormorant, previously known only from the Commander Islands, suggest this species was more widespread in the past.

Combined analysis of all bird species recovered from Amchitka middens revealed other patterns (Table 1). Aleutian Canada Goose (Branta canadensis minima) numbers fluctuated widely during the 2,650 year occupation of Rat 31. Peak abundances corresponded with climatic maxima; all but one of the population lows corresponded with climatic minima. The single exception to the pattern was for goose bones recovered from the upper strata, laid down in historic times (Desautels et al. 1970). Although global climate was warming during this period, goose numbers were extremely low. Of all species present goose and Short-tailed Albatross (Diomedea albatrus) numbers were significantly lower than expected. Apparently the introduction of Arctic Fox by Russian and American trappers and traders had profound effects on some bird populations, in particular, those that nest on flat ground, such as geese and albatross (Siegel-Causey et al. 1991).

Other species contrast sharply with the pattern noted for the Aleutian Canada Goose. A goose population high seen roughly 1,000 years ago is associated with population lows in Slender-billed Shearwaters (Procellaria tenuirostris) and auklets (Aethia spp.). Similarly a low population of geese 1,800 ybp is associated with peak abundances for Northern Fulmar (Fulmarus glacialis) and auklets. This leads us to suspect these changes in relative bird abundance reflect environmental, not human, perturbations.

Preliminary analysis of remains from Little Kiska, Shemya and Buldir, suggest patterns observed on Amchitka hold true elsewhere in the western Aleutians. Abundant remains of Aleutian Canada Geese and Short-tailed Albatross were found in middens on these islands. On Little Kiska and Shemya numbers decline drastically in historic levels. We found no such drop in abundance on Buldir, which never had Arctic foxes introduced. We can only conclude population declines in these and other species are due entirely to the actions of Europeans in introducing foxes (Siegel-Causey pers. comm.).

We recovered 6,000 bird bones from the two 1991 test pits, nearly 75% from Pit 2. The subfossil assemblage represented in the midden includes 23 species, of which 21 are seabirds (Table 2). In looking at the total assemblage there are some interesting differences between modern and past abundances. Kittiwakes, well represented in the modern fauna, are not found in Pit 2, the older part of the site. Least auklets (Aethia pusilla), the most abundant modern Alcid, make up only 2-3% of archaeological specimens (Lefevre and Siegel-Causey 1993).

Murrelets (Brachyramphus sp.) and Glaucus gulls (Larus hyperboreus)poorly represented or absent in the modern fauna. Murrelets make up nearly 7% of the remains in Pit 2 while Glaucus gulls are a small but consistent component of the past fauna.

Whiskered auklets (A. pygmaea) are the most common species in Pit 1 with 28% of MNI. In Pit 2 the most abundant remains belonged to crested auklets (A. cristatella) and Cassin's auklet (Ptychoramphus aleutiacus) with 25% each, and rhinoceros auklets (Cerorhinca monocerata) with 10%. Whiskered and Crested auklets each make up less than 1% of the modern fauna. Cassin's auklet is represented by 200 breeding pairs while only 12 pairs of Rhinoceros auklets breed on Buldir (Lefevre and Siegel-Causey 1993).

In order to compensate for human hunting choice as a bias in the sample we compared size and weight of the different species (Lefevre and Siegel-Causey 1993; White 1953). For example, the small size of Least auklets may have made them less economically attractive than the less abundant larger auklets. On the other hand the high incidence of Crested and Whiskered auklets in the midden, in spite of their small size, could indicate an interest in the ornamental feather crests for decorating clothing (Lefevre et al. In press).

Accessibility of colonies was also analyzed in order to factor out human choice from environmental process. The different ratios between murres (Uria sp), about 3% of remains and puffins (Fraturcula sp), 5-9% of remains, may be due to easier access to puffin burrows. The clearest evidence of a change in species numbers is suggested by Rhinoceros auklets. The large size of these auklets, over 500 gm, coupled with low modern numbers, may suggest these birds were more abundant in the past.

The environmental work ties in to our second goal, understanding how changes in the regional environment affected the development of Aleut

culture. Faunal material is being examined in order to establish functional and/or seasonal differences between sites in the Near Islands and to provide information on regional settlement patterns, economics and social life. Establishing a regional chronology is critical to achieving these goals. We have recovered material for dating from as many sites as possible and are beginning to place settlements within a temporal framework. At the same time recovering artifacts to establish a regional typology has grown in importance.

WAAPP is documenting environmental changes in an effort to define environmental influences on Aleut culture. Mason and Gerlach (1995) correlate changes in social complexity and technological innovation in Nort Alaska to changes in regional climate paterns during the late Holocene.

Given the above information on climate change we suggest that the shift to a drier warmer climate regime, resulting in fewer and less violent storms, allowed Aleut colonists to safely make the boat journey from the Rat Islands to the Near Islands. The gap between Kiska and Shemya is a critical problem for people using small boats. Until Kiska Island is reached the islands in the Aleutian archipelago are visible at sea level from neighboring islands and even on cloudy, rainy days mariners could see the next landfall. This is not the case from Kiska to Buldir or Buldir to Shemya or Agattu. Though both Kiska and Agattu are visible from Buldir on a clear day, one must be on top of the highest peak on the island to see them. Prior to the climate changes the long journey, out of site of land in open skin boats, may have seemed too hazardous for families to attempt.

Another of our goals is to study the effects these early colonists had on the environment of the Aleutian Islands. The faunal remains in SH-7 are revealing an interesting history. At the bottom of the pit, lying on the sterile sands of the beach ridge, we found the remains of 3-4 fur seal pups,

and 2-3 adult female fur seals. The pups ranged in age from possibly several months old to about 1 year. Until pups are 3 months old they are unweaned and unable to swim. The youngest pups, still reliant on their mothers, would indicate the pups were born in rookeries in the Near Islands, and probably in the Semichis. The evidence so far is inconclusive but if unweaned pups are found in middens on Shemya it would indicate the earliest colonists found rookeries. Exploitation by the early human colonists may have caused the fur seals to abandon the Near Islands, and retreat to their modern rookeries on the Commanders and the Pribilofs.

Another surprise from both SH-6 and SH-7 is found in the fish remains. Two analysts, Megan Partlow, of the University of Wisconsin, and Susan Crockford of Pacific Identifications are conducting detailed analyses of all fish remains from the sites on Shemya. Our samples show that about 60% of our fish were gadiformes (cod) and scorpaeniformes (rockfish). A smaller percentage are flatfish and we have one salmon and a few miscellaneous other species. Cod are the most abundant species of fish used and some of the fish from these sites are of enormous size, well over the maximum recorded for Pacific Cod, about 150 pounds. Crockford estimates the larger cod reached 200 pounds. I will leave it to others to ponder the implications this has for the effects of man on the ecosystem.

Short-tailed albatrosses currently nest on only one small island belonging to Japan. The birds were driven to the brink of extinction early this century and have not recovered to anywhere near their former numbers. The bones of these birds are extremely common in Aleut sites, and biologists have speculated that they nested in the Aleutians. Remains from sites on Shemya and Agattu have been analyzed with an eye to addressing the question of the former range of these birds.

One of the attractions for colonists to the western islands may have been

populations of sea mammals not yet accustomed to human hunters. These animals would include not only the usual sea lions and hair seals but also possibly fur seals, walrus and sea cows (Hydromalis gigas). At the time of Russian penetrationof the North Pacific in 1741 Sea Cows were confined to the Commander Islands. The huge inoffensive animals were exterminated by Russian fur hunter victualling their ships within 26 years of discovery. The presence of living sea cows has been one of the strongest arguments for the Commanders being unknown to humans prior to the Russian arrival. We hoped to find some of these animals in our earliest levels to give us information on the paleoenvironment and confirm expectations about man the voracious hunter spreading havoc in virgin territory.

Our Russian colleagues, excavating Pit 7 on Buldir Island in 1997, recovered two fragments of Sea cow bones. Interestingly the remains are not from the lowest, colonization level of the site. They came from the upper level of Pit 7 which while not yet directly dated has an artifact component identical to Pit 6 which dates to the 14th century. Comparisons with other bones in Russia have confirmed the identification as sea cow. Currently one of the fragments is undergoing DNA study using comparative material from the Smithsonian Institution.

This would imply the sea cows hung on in human occupied areas until quite late in prehistory and confirms Aleut stories of sightings and even of hunting the giant animals. Lucien Turner, a weather observer on Attu Island in the 1880s, reported stories that Sea Cows were present in the Near Islands and that the beasts, so inoffensive and helpless, were scorned by the men but hunted and butchered by women, who gathered them from the intertidal zone like large shellfish.

Culture History and ColonizationThe earliest evidence of human occupation for the Near Islands is from the

bottom of Site SH-7. These dates suggest humans arrived in the Near Islands between 3500 and 3000 years ago. As previously stated we suggest that the shift to a drier warmer climate regime between 4000 and 3500 years ago allowed Aleut colonists to safely make the boat journey from the Rat Islands to the Near Islands.Six of 9 known sites on Shemya have been dated. Due to the damage the sites have sustained over the years, all have yielded dates from near the bottom of the deposits. We are confident of the initial dates for occupation of all these sites. There is one known site on Shemya dating to 3500-3000 years BP. There is a small cluster of dates about 500 years later at Sh-7. About 1000-1300 years after the colonization several sites, SH-3, SH-4, SH-6, SH-8 and SH-9, are being used. We do not know how long most continued in use because all the later levels have been destroyed. The latest date from Shemya is 1720 BP from SH-6 Undocumented collections from sites SH-3 and SH-4 have yielded ground stone artifacts of the Late Aleutian Trait complex which appeared on Umnak and on Amchitka about 1000 AD. We do not know when this complex arrived in the Near Islands. Bone harpoon points made with metal tools may also belong to this complex, but they may also represent early historic occupation. No clearly historic artifacts have been reported from sites on Shemya although Russian records indicate the Semichis were occupied at the time of contact.

The work on Shemya aimed at obtaining comparative collections of tools and faunal remains, and good dates from as many sites as possible to examine Aleut land use and subsistence strategies. We were able to retrieve datable material from undisturbed contexts in 6 sites. If there is an advantage to working on Shemya it is that we were able to reach the bottoms of all the sites and therefore have an excellent record of the initial occupation for most of the sites on the island. The obvious drawback is that we have no idea how long any of the sites continued to be occupied. The later occupation of the island has been obliterated.

This slide shows the locations of all known sites on the island and lists all dates for each site. The earliest evidence of human occupation for the Near Islands is from the bottom of Site SH-7. These dates suggest humans arrived in the Near Islands between 3500 and 3000 years ago. About 1000-1300 years after the colonization several sites, SH-3, SH-4, SH-6, SH-8 and SH-9, are being used. We do not know how long most continued in use because all the later levels have been destroyed. The latest date from Shemya is 1720 BP from SH-6 Undocumented collections from sites SH-3 and SH-4 have yielded ground stone artifacts of the Late Aleutian Trait complex which appeared on Umnak and on Amchitka about 1000 AD. We do not know when this complex arrived in the Near Islands. Bone harpoon points made with metal tools may also belong to this complex, but they may also represent early historic occupation. No clearly historic artifacts have been reported from sites on Shemya although Russian records indicate the Semichis were occupied at the time of contact.

We are still in the early stages of analysis of the faunal remains and the artifacts from the sites on Shemya. Intersite comparisons are beginning to yield information on changes in stone tool technology and use of raw materials. The environmental and economic potential is just beginning to be realized. A monograph report is currently being compiled. Shemya is no ones idea of an ideal research target but these poor raveaged sites have amply repaid our efforts to understand the prehistory of the Near Islands.

Table 1. Western Aleutian Radiocarbon Dates (uncorrected)Island Site Pit & Level material C-14 date BP Lab no.Shemya3 ATU-022 1/cut 8 charcoal 2020+-110

Beta-110114Shemya3 ATU-022 1/9 bone 1856+-69

IEMAE1231

Shemya3 ATU-022 1/10 bone 1589+-60IEMAE1229

Shemya3 ATU-022 2/9 bone 2047+-82IEMAE1226

Shemya4 ATU-023 TP 2 bone 2680+-70 Beta

Shemya6 ATU-003 1/4 charcoal 1790+-160 Beta-39090Shemya6 ATU-003 1/6 charcoal 1860+-90 Beta-39091Shemya6 ATU-003 2/1 charcoal 1770+-120 Beta-39092Shemya6 ATU-003 3/1 charcaol 1720+-70 Beta-40420Shemya6 ATU-003 3/3 charcoal 2030+-70 Beta-40421Shemya6 ATU-003 3/3 charcoal 1810+-60 Beta-40422Shemya6 ATU-003 4/2 bone 2148+-70 IEMAE-1172Shemya6 ATU-003 4/7 bone 2555+-126 IEMAE-1177

Shemya7 ATU-061 TP2/bottombone 3540+-60 Beta-39104Shemya7 ATU-061 1/110-120cm bone 3096+-155

IEMAE-1175Shemya7 ATU-061 feature bone 2570+-140 IEMAE-1173Shemya7 ATU-061 feature bone 3080+-110 IEMAE-

1174Shemya7 ATU-061 Column bone 2630+-60 BetaShemya7 ATU-061 Column bone 3120+-80 Beta

Shemya8 ATU-062 1/bottom bone 2060+-90 Beta-40424

Shemya9 ATU-066 TP 1 bone 2244+-182 IEMAE-1176

Agattu ATU-001 Unit 4, Bottom charcoal 2500+-300 M-12Agattu ATU-001 Unit 4, bottom charcoal 2630+-300 M-12Agattu ATU-001 Upper levelcharcoal 1300+-150 M-13

B3 Interisland ContactsCultural isolation has long been a paradigm of Aleutian archaeology. The Near Islands, the westernmost group in the chain, are the most isolated and the people the most culturally divergent. This paper examines the distribution of several typically Aleutian cultural traits, burial and artifact types, and social and ideological features to outline the extent and nature of western Aleutian isolation.

By 3000 BC the Aleutian island chain was occupied by sea mammal hunters and fishers with a bifacial stone tool industry and an elaborate boone and wood working technology. About 1000 AD widespread cultural change occurred throughout the North (Dumond 1986). Thule Eskimo culture traits spread throughout Alaska and across Canada to Greenland. Thule artifacts, included ground slate tools, new harpoon styles, clay lamps and thick walled pottery, and ivory animal figurines (Anderson 1984). Iron became available in larger quantities, sparking an artistic florescence with new

decorative elements, on tools and new forms of symbolic expression. The mechanism for change is not well understood but in North Alaska it is linked to whale hunting and a possible influx of people from Asia.

Population increase and sweeping changes in social institutions are apparent on the Alaska Peninsula, Kodiak and in the eastern Aleutians. In the eastern Aleutians single family dwellings give way to long houses up to 50 meters long, sheltering whole villages under one roof. The largest long houses are found in the eastern islands, but they extend west to the Central Aleutians and east to Ugashik Lakes, more than 320 miles from the islands (U.S. BIA 1990; 1991). All such communal houses in the RatIslands are in sites with historic remains (U.S. BIA 1985). No large long houses are known from the Near Islands.

New forms of religious and symbolic expression are reflected in the development of mask ceremonialism and elaborate bentwood hats. The eastern Aleutians formed one of the earliest Alaskan mask making centers, nearly 2000 years old (Ray 1967). By late prehistoric times a variety of regional styles had proliferated. Masks have been found through the central Aleutians but have not been found in the Rat or Near Islands (Black 1982). Bentwood hunting hats are found throughout southern and western Alaska from Prince William Sound to Bering Strait (Black 1991). They come in a variety of styles from open crowned visors to elaborate conical affairs decorated with carvings and beads. Only the simple open crowned visor known from the Rat Islands and no carved wooden hats are reported for the Near Islands (Black 1983).

Burial patterns also indicate widespread social and ideological change and complexity in the late prehistoric period of much of Alaska. Aleut burial practices were correspondingly complex but a near total lack chronological information inhibits understanding of their significance. Bundled burials in burial houses, and isolated extended burials with scattered evidence of dismemberment and skull curation (Jochelson 1925; Hrdlicka 1945), are found throughout the chain. Possibly later developments, with more restricted ranges, include cremations, bundled and extended burials in caves and above ground log sarcophagi (Bank 1948; Hrdlicka 1945; Weyer 1931), and stone and whalebone crypts covered by mounds of earth, called umqan (Aigner and Veltre 1976). By AD 1200 Aleut burial customs included intentional mummification of important people (L. Johnson, pers. comm.). The most elaborate and diverse burials are concentrated in the east with an irregular spread to the central islands. The only types of burials reported from the Near and Rat Islands are the bundled burials in houses and extended burials.

There appears to be a strong cultural boundary separating the Rat and Near Islanders from the central and eastern Aleuts. These examples indicate the western Aleuts did not participate fully in what is considered classic Aleut culture. Distance certainly had a role in the diffusion of traits from east to west, but political and cultural factors must have played a part in filtering ideas.

In spite of these obvious differences the western Aleuts were clearly part of a larger Aleutian oecumen. Aleutian matral technology is remarkably similar all along the chain suggesting regular and frequent contacts between the different island groups. McCartney (1971;1977) compared artifact assemblages throughout the chain and found that, while every island group possessed distinctive artifacts, with the Near Islands the most divergent, overall similarities were more important. He concluded the

different island groups had been in more or less continuous contact for at least 4000 years.

The evidence also suggests that transmission of material traits and information could proceed rapidly. A Late Aleutian Trait Complex including ground slate and iron knives, has obvious similarities to Thule. This complex spread west through the Aleutians around 1000 AD. Some, but not all, of the elements of this complex, including ground stone tools and nipple ended needles have been found in the Near Islands. As with so much evidence from the Near Islands there are no dates for the arrival of this complex. Even later, about 1600 AD another cluster of artifacts, including rod-like unilaterally barbed harpoon points, long socketed foreshafts, small bilaterally barbed harpoon points and bear figurines, appeared in the eastern islands and spread west. Again some, but not all, elements of this complex, including bear figurines, spread to the Near Islands by 1750.

The question of western Aleutian contact with Asia is far from clear. Asian influences on Near Island culture may be manifested in non-artifactual spheres. Laipunova lists Asian traits widely distributed in south central Alaska. Some, like roof entry houses and the associated notched log ladder, and the use of Aconite poison for whaling are restricted only to the Aleutians. Some Asian traits found in the eastern Aleutians are not reported for the west, most notably the use of poison on darts. The Near Islanders shared a number of Eskimoan traits with the people of southwest Alaska and Siberia. These include Raven mythology featuring Raven as trickster/creator (Bergsland and Dirks 1990) and large chiefs houses, used for ceremonies, analogous to Yup'ik Eskimo qasgiqs. Raven mythology and ceremonial houses are not found in the central or eastern islands (Black 1984). Also some Near Island art, particularly ivory figurines, show strong resemblances to carvings produced by peoples living on the Sea of Okhotsk littoral (Black 1982). Any attempt to decipher the prehistory of the western

Aleutians must make an attempt to explain these seemingly anaomalous cultural patterns.

Buldir Island as a Cultural Crossroad

Buldir Island is uniquely placed to address questions about interisland contact. All activity on the island was restricted to a single locus, and the island, between Kiska in the Rat Islands and Shemya in the Near Islands, is a convenient and necessary stopping point between the island groups. Our research suggests both Rat and Near Islanders used Buldir concurrently, though we do not know if the contacts were friendly or hostile. Good comparative material exists for both the Rat and Near Island material culture, and the Near Islands especially possess several unique artifactual traits.

We looked at 194 stone, and 100 bone or ivory artifacts and over 13,000 flakes of raw material, to learn both the functions of the site on Buldir and to try to determine the identity of the people using the island. All of these materials came from levels dated to the mid to late 1600's. Russian traders in the 1780's reported two mutually antagonistic groups on the island (Black 1984). By examining lithic material types we hoped to identify materials imported from either the Rat or Near Islands. Aleuts overwhelmingly used local materials for tool making. In the Near Islands fine silicified argillites and propyllitized andesite were favored. In the Rat Islands people relied on basalts and andesites. Low grade, friable phyllites, and siliceous phyllites dominate the collection from Buldir (Corbett et al. in press). Of all the thousands of flakes analyzed, only 71, of three varieties of andesite, are probable imports. Fourteen flakes of propyllitized andesite, the most common tool material on Shemya, suggest a possible connection to the Near Islands.

Bone tools, especially harpoon heads, are the most sensitive to minor stylistic change and are most often used in comparisons. Our bone assemblage is small and contained only 5 widely distributed types of harpoon points. The rest of the bone assemblage consisted of awls and wedges, are common all along the chain, or are unique. Three dense bone or ivory blunts are the only bone tools to suggest a connection with the islands to the west (McCartney 1971).

All stone tools recovered were made on the locally available phyllites. The material is easily shaped but brittle and unsuitable for the decorative flaking, incising, and regular edge serrations that distinguish Near Island collections from other Aleutian materials. The one Near Island trait apparent in the Buldir materials was an emphasis on linear forms with parallel sides. Of the 194 complete or nearly complete stone tools, 35% (68) have parallel sides. Tools in the Rat Island tend to be triangular; of 41 classes described by Desautels (1970), 30 were triangular points and knives. We recovered 27 (14%) triangular points displaying stylistic similarities to Rat Island points. While neither linear nor triangular forms are exclusive to the Rat or Near Islands the styles may indicate trends.

Though far from definitive, evidence from Buldir suggests the occupants during the late prehistoric period were from the Near Islands. Although not exclusive to the Rat Islands, the presence of a sizeable number of triangular points suggest Rat Islanders may have also used the island. This period in late prehistory corresponds to the spread of a cluster of artifact styles from the eastern islands to the west. Contact between the Rat and Near Islanders on Buldir may have allowed transmittal of the new cultural traits to the west.

B4 Cultural complexity

B5 Buldir as Refuge/BoundaryResearch on Buldir has posed additional questions and brings in the third goal. Excavations on Buldir have provided an unending string of surprises. The site, once thought to be simply a seasonally or sporadically used hunting camp, is obviously much more complex. In our first season we found well preserved fragile organics and clearly separated occupational layers which gave great promise for future research. In the second season we expanded the artifact sample from the 17th century occupation and began work on a whalebone structure, the first to be excavated in the Aleutians. The third season added deeper cultural levels and more definition of the house. We now believe Buldir represents a frontier/boundary. The frontier is an edge, away from the populous cores of the island groups to the east and west. Buldir, on this edge, may have been a refuge during periods of resource or social stress in the main island clusters. This may explain the sporadic, but relatively intense occupations. It is clear from previous work the site was not a seasonally used hunting camp, nor does the island have the resource base to support a permanent resident human population.

The most recent excavated levels of Buldir, dating from the mid seventeenth century, consist of a large outdoor work area rich in fragile wooden artifacts, and of a house with whalebone structural supports.

The house is the only structure identified so far on the site. A meter of windblown sand covers the entire site concealing most surface indications of pit houses. The structure is on the south side of the midden overlooking the creek and marsh. The surface depression indicates the house measures about 10 meters long by 5 wide. It is oriented northeast to southwest and probably had the traditional roof entry. In 1993 a 2x2m test pit in the depression exposed a cluster of whalebones and it became obvious the

depression was a housepit. In 1997 the excavation was expanded to a 3 x 8 meter excavation, exposing cobbles, griddle stones and bones from whales and other sea mammals. Stratigraphy in the house is complex but the floor is marked by a dense black organic rich layer. Most of the whalebones are embedded in this level. In the approximate center of the 1997 excavation and dominating the 1993 excavation was a pit feature 1.5 m in diameter and at least 129 cm deep, lined with a thick mat of grass and hair and sea lion scapulae, with a medium sized whale skull placed snout downward in the center. The skull would have been an a unavoidable focal point for all activities in the structure. To the north of the skull was an amorphous hearth. Concentrations of bird and fish bones were observed under and between cobbles and larger whale and pinniped bones.

A double burial was found in the southeast corner of the house excavation. One burial, represented by two femurs, a humerus and phalanges, appeared to be a young adult. The position of the bones, with the head of the humerus near the distal end of the femur, suggests the body was tightly flexed. A smaller rib cage and spinal column is from a child. We had gotten permission from the Aleut people to conduct field analyses of human remains but our osteologist was injured getting to the field and returned home. Excavation in the unit ceased when the burials were recognized and the remains were reburied in situ. The presence of bodies in an Aleut house is not unusual although these two do not appear to have been buried, there was no evidence of a pit. Their bundled bodies may have been hung from the rafters or propped in the corner, a practice well attested in the eastern Aleutians.

Whalebone is a common structural element in house remains all along the Aleutian chain but none of the houses excavated and reported have had a substantial whalebone component. A house with a whalebone superstructure was excavated on the lower Alaska Peninsula but is

associated with an intrusive Thule eskimo occupation in an otherwise Aleut area. An early historic house from Reese Bay on Unalaska Island, excavatd by McCartney and Veltre, incorporated two whale skulls and other skeletal elements in non-structural contexts. The Buldir house represents the first excavation of an Aleut whalebone house, and one of very few house excavations for the western Aleutians.

Analysis of the artifact assemblage is just beginning but appears overwhelmingly weighted toward projectile points of various types, knives and hammerstones. There are very few items for domestic functions such as cooking, sewing, skin working or tool manufacture. Three stone lamps were found, one decorated with pecked lines and grooves. We could be dealing with a structure built and used by a predominantly male work party or we may have a special use structure analogous to the Eskimo gasgiq or mens house where men lived and worked apart from their wives. Mens houses are not characteristic of the Aleutians in general but early Russians reported that Near Island chiefs houses were larger than the rest and that ceremonies took place in these houses. Further analysis will undoubtedly help to define the function of the structure, for now all we can say is that this feature is not like anything described from elsewhere in the chain.

The seventeenth century work area on the front of the midden has yeilded a wealth of information on Aleut manufacturing techniques as well as a large sample of wooden artifacts, primarily tool handles, harpoons, fire drills etc. Few wooden artifacts are known from the Aleutians and these are an invaluable adition to our knowledge of Aleut technology. With the woodworking tools, awls used for preparing the skin covers of the boats and a series of kayak model pieces we also have evidence the site was being used for kayak construction or repair.. We do not know if the models were toys or if they were actual used as plans for full sized boats. So far at least three models are represented in our collections.

DThe team plans to work on Attu Island for the next several years. Attu is the largest island in the Near Islands and the fifth largest in the Aleutians. It is also the closest to Asia. This proximity suggests that, prior to the introduction of foxes in the 18th century, the island was one of the most biologically interesting in the chain. We will spend the first season surveying as much of the coastline as possible for sites and in testing these sites for material for dates. We will be trying to establish occupational histories for the sites. We will also be sampling whale bone in structures to establish temporal patterns of whale use. DNA analysis will allow us to identify the whales used and together this information may allow us to make inferences about whale hunting in the Near Islands. Another question we hope to address is the timing of the appearance of the so-called chiefs houses. Jochelson suggests the larger structures found in some sites were later than the more typical small houses.

While part of the team is conducting this survey, a crew will begin large scale excavations at a site selected during the early reconnaissance. Planned excavations include testing of a chiefs house and ordinary house to determine any differences in arrangement, artifact assemblages etc which could indicate functional or status differentiation.

Meanwhile we are conducting research to round out the ethnographic and historic picture of the Near Island Aleuts since contact. The goal is to connect the archaeological materials with the historic community to demonstrate the continuity of tradition. As part of this documentation we are working with Aleut historian Alice Petrivelli to interview and record the lives of the few surviving Attuan Aleuts as well as Atkans who at times lived, worked or visited Attu before the war.

The oral history research focusses on the history and culture of the Near Island Aleuts. We are collecting genealogies, life histories, and general social information. We are also gleaning as much economic and subsistence information as possible. This approach involves presenting the informants with pictures of the animals recovered from archaeological excavations and soliciting any information on seasons, capture techniques and preferences.

Oral history will be supplemented by available historical documentation-mainly Coast Guard (Revenue Cutter) records, Alaska Commercial Company records, health service records, etc. The goal is to construct as complete a cultural history and historical ethnography as possible for the Near Island Aleuts, a unique, little studied and almost forgotten people at the end of the world.