Founded 8000 B.C.

The Native Village of Afognak

Ancient Afognak

Precontact History and Archaeology of Afognak Bay

Donald W. Clark
Oct 1, 2004 - Feb 2005

          Late Kachemak
     Survival of the Archaeological Record
     The place of Afognak in Ancient Times.

     A dozen archaeology "digs" at Afognak Bay have contributed to the results reported here. Projects at other sites on adjacent Kodiak Island during the past 70 years have added to our knowledge of the peoples who lived on Afognak Island. The Afognak work is listed below and the excavations are described further in Appendix III.

     1951 Observations by D. Clark while he was a fisheries stream guard at Litnik.
     1964 Three days survey by Clark and W. Workman, observed damage to sites.
     1971 Joint Alaska Methodist University-National Museums of Canada project by
             Workman and Clark at Ocean Bay culture sites at mouth of the Afognak River.
     1994-1997 Dig Afognak excavations by Patrick Saltonstall at Settlement Point site.
     1994-1998 Dig Afognak research by K. Woodhouse-Beyer at historic Katanee at Settlement Point.
     1996 Excavation at Afg-012, inner Afognak Bay, by Megan Partlow
     1998 Alutiiq Museum survey of part of Afognak Island by A. Steffian and Saltonstall.
     1999-2000 Dig Afognak excavations for BIA by Clark at Afognak’s Aleut Town.
     2001 Dig Afognak excavation by Clark at Kachemak culture “Tsunami site” at Litnik.
     2002 Dig Afognak project by Clark at Kachemak culture “Salmon Bend” site, Litnik
     2003 Further work on the Tsunami site.”
     2004 Dig Afognak excavation at Early Kachemak culture site Afg-088 at Litnik.

     In addition, there have been Alutiiq Museum-ANC (Afognak Native Corporation) excavations at Malina Creek, and two projects on Shuyak Island by the State of Alaska. They are outside the geographic scope of the present report. There also have been minor tests at other sites.
Where the artifacts are: Artifacts from the 1964 survey are at the University of Wisconsin (Madison) Department of Anthropology. The 1971 excavations were done under a U.S. Forest Service permit. The University of Alaska-Anchorage, Department of Anthropology, has been designated a collections repository. A few items obtained by Clark in 1961 are in the Baranov Museum, Kodiak. With the exception of the State collections, all the remaining collections are at the Alutiiq Museum and Archaeological Repository, Kodiak, where many items may be seen on exhibit.
This work does not give in-text reference citations (except for quotations), but there is a bibliography at the end. The relevance of bibliographic entries is obvious in most cases; otherwise, bracketed guides appear in the bibliography.

     For 7000 years, Afognak Island has been the home of numerous maritime hunters, ancestors of the historic Alutiiq people. Their villages were located in every bay and at the mouth of every important salmon stream. Afognak and Kodiak both have several distinctive ecological features that were important in the development of Native cultures. A well-stocked littoral (intertidal shore) zone and surrounding waters provided for a large population that numbered 8,000 persons on Kodiak and Afognak at the time of European conquest in 1784. Possibly there were more than that at an earlier date.

     Before recent introductions, land mammals were very limited, notably giant brown bears, river otters, and foxes. They were not of major economic importance compared with sea mammals. Instead, the land was essentially a platform for living around the ocean, for maritime hunting, fishing, and gathering. Technically, this lifestyle is referred to as foraging.

     The cool, stormy, wet climate – in a word, the Aleutian low pressure system – constrains nature and human activity. Spring comes slowly, and before the recent warm trend, little green vegetation was to be seen until the middle of June. The heads of inlets where salt water is diluted with fresh water freeze, for instance at the mouth of the Afognak River as far out as Lipsett Point. But with their nearly temperate climate, the coasts remain ice-free during winter. Thus, hunting and fishing techniques were not oriented to hunting from sea ice, unlike the case along the Arctic coast of Alaska and Canada. Until less than a millennium ago, possibly just 700 years ago at Afognak Bay, there was no spruce forest to shelter settlements and buffer the impact of storms. Older residents remark on the presence of hillside clearings in the forest within their own time. But there is an upside to rough weather. High winds and wind-enhanced currents actually increase the primary oceanic productivity (of phytoplankton) of the western Gulf of Alaska and thus make a positive contribution towards supporting the abundance of fish, and consequently of sea mammals and of seabirds.

     The rich oceanic setting includes fishing banks located east of Afognak Island, sea mammal rookeries and haul outs, bird rookeries, migratory populations of whales and fur seals and birds, and a bountiful littoral zone exposed by tides that range by up to 13 feet between high and low extremes. The complexly embayed and channelled shoreline thus supports ecological diversity and an abundance of virtually everything from whales to periwinkles. And there is a lot of shore habitat in relation to land area (3600 km for the Kodiak Archipelago by one measure), again in part due to coastline complexity.
The wet climate (about 80 inches annual precipitation) and complex topography combine to endow Afognak Island with more than a score of streams, perhaps two score, that support salmon runs. These formed the basis for a major summer fishery at the mouths of larger streams, as at the Afognak River (Litnik), Selezneva (Little Afognak), Malina Creek (Malinovski Litnik) and Portage River (at Discoverer Bay). Pauls Bay and Kitoi Bay are recent fisheries created by habitat enhancement and propagation. Thanks to the gouging action of Ice Age glaciers, the islands were given not only their intricately embayed and fiorded coastline, but also lakes to which red salmon return to spawn.

     Though there was a certain degree of cultural continuity throughout 7,000 years, this duration was punctuated by numerous developments in technology and houses, and possibly also in organized salmon fishing. The most outstanding and easiest documented development was from flaking tools like knives and spear heads from hard stone to grinding them from sheets of slate. This and other developments lead to the convenient division of prehistory into a series of periods – Ocean Bay, Kachemak, and Koniag – here called traditions. Deglaciation of the islands and adjacent Alaska Peninsula at the end of the Ice Age about 12,000 years ago sets a limit to the possible onset of settlement. As yet, though, no known archaeological remains on either Kodiak Island or Afognak Island come even close to dating back to the time of the glaciers and ice caps.

     We have some information on how the Russian fur seekers first met the inhabitants of Afognak Bay. Possibly some Afognak hunters had gone to the Russian outpost at Chiniak) which had been established near the site of the town of Kodiak a year or two after the Russians had landed at Three Saints Bay. At that time, Shelikhov also had sent two men to Shuyak Island, who were killed by the Shuyak chief, as reported by Afognak Islanders, who bore the news to Chiniak. To establish his firm rule, and possibly because he feared a Native conspiracy to destroy the Russian foothold on Kodiak, Shelikhov eradicated the native settlement(s) on Shuyak and possibly also on northern Afognak Island. Soon the interests of Shelikhov's company also turned to Afognak Bay. A village or house is shown at Afognak Bay on Izmailov's 1786 map of Kodiak (published in 1787 as part of a larger map). A Russian party that had just over-wintered and had built an artel (a fur hunting and provisioning outstation) at Karluk explored Afognak Bay in 1787 and established an artel at Little Afognak. The map noted above, and other maps of about the same date, show a storehouse or depot, sometimes referred to as a fort, on the north side of Afognak Island. The so-called fort appears to have been short-lived, if it ever was staffed, as it is absent from later maps and is not mentioned in Russian records other than Shelikhov’s account. Dealings with Afognak Islanders from unspecified villages are mentioned in historic records from before the end of the 1700s . After 1802, there are numerous references to the two artels: Igvak (the Igvetsk artel), noted above, apparently associated with the Alutiiq village Igwik, which later is identified as Selezneva or Little Afognak, and Malinovskii Litnik at Malina Creek on Shelikof Strait. The latter probably was downgraded to an odinochka or one person outstation before long. There was another letnik or summer fishing station at Litnik at the Afognak River and an odinochka (a small or one-man post) at Afognak Village. According to one account, the odinochka managed the fishing at Litnik in season. A mid-19th century report also links the Litnik fishery with Kattak. Kattak was located near Settlement Point across from the innermost end of Afognak Village. It is at the site of the field base for the Dig Afognak Program. At Kattak (Quatat, spelled variously) there was a Russian-Creole-Alutiiq settlement that persisted to about 1940, by which time it was associated with the Nekrassoff family. Possibly some references to “the Afognak artel” refer to this settlement, but the records are a riddle.

     What kind of people did Shelikhov's men meet? The answer to this question is, in a sense, the top layer of salt fish in the deep barrel of prehistory with which this essay deals. These Kodiaks, Native Americans, Koniags or "Aleuts," now called Alutiiq, lived by, with, and from the sea. Whalers were important specialists in Alutiiq society, though somewhat feared because of their medicine or shamanistic power. The whales floated to the surface and sometimes drifted ashore after they were struck with poison-tipped spears. The deep bays around Afognak, like Danger Bay and Kizhuyak Bay, were especially suited for the Alutiiq method of whaling. Also, the whales were plentiful in Marmot Bay, especially in the tide rips outside Whale Passage, Afognak Strait, and Shuyak Strait. Hrdlicka reported that a ritual whale trap was drawn across outer Kizhuyak Bay by traversing the bay towing a pouch of fat from a corpse. Ordinary hunters focused on harbor seals. Supplemented by many porpoise and sea lions, they probably supplied most of the red meat and also oil and hides. The seals were harpooned from kayaks or from the shore after being attracted within range by a decoy. The decoys consisted of seal head-shaped helmets and inflated seal skins. Seals also were entangled in large nets and were clubbed at haulouts, as sea lions probably were, also. Additionally, hunters managed to take many fleet porpoises with harpoons or nets. Bears were hunted with bows and arrows, although not many were killed. Occasionally there were trips across Shelikof Strait to the Alaska Peninsula for caribou. Some meat and antler probably were obtained by trade with Peninsular Alutiiqs. Sea birds were taken with multi-pronged spears and in nets. Rookeries were raided for eggs.

     In the sea, many fish obligingly attached themselves to hooked lines, especially cod, rockfish, and halibut. Of equal importance was the summertime focus on the salmon fishery at the mouths of streams, especially the Afognak River. Weirs located close to the summer settlements held back and penned the salmon so they could be speared. The food quest followed the natural migration cycles of animals and fish. This cycle formed a calendar for other aspects of Alutiiq life. In the fall, after the salmon runs had peaked, people returned to their main settlements and hunted sea mammals as the time for winter festivities approached. The main, or winter, villages were located close to the outer coasts, so as not to be left isolated and frozen-in at the heads of bays.

     Extended families or cooperating households of about 18-20 persons lived in each dwelling. Thus, a settlement of not many houses could hold 100 to 200 persons, though some were larger or smaller. The main or central common room served as a workshop and kitchen, and also for storage, while nuclear or individual families occupied appended anterooms. One of the side chambers was used for a wet "sweat bath" similar in some aspects to the Russian banya (but the rocks were heated in the main room and then taken into the bath chamber). The structures were set into a rectangular pit, dug two feet or deeper into the ground for protection from the weather, banked with turf, and covered with thatch. The superstructure was of driftwood posts and beams; the roof probably was cribbed and had a central smoke hole-skylight. Historical ethnographers and surviving traditions fail to describe the composition of a household. It very likely was an extended family, possibly sisters and their husbands and offspring, plus surviving parents. There also could have been unmarried brothers, foster children, attached persons of low status, and actual slaves. Persons who died were interred within the village area, sometimes even in one of the anterooms of their house, or were taken to an old abandoned village for burial (at their ancestral home?). In a sense, they continued to be members of the community.

     Important persons might be mummified and placed in secluded rock shelters where there was a risk of being discovered by whalers and used in secret rituals.

     Other rituals and ceremonies were more public. These are incompletely recorded but included dances, masked theatrical performances, and feasts during the winter season. At that time, the recently deceased were honored with a memorial feast. Certain ceremonies had the objective of pleasing and propagating game, while some were invitational feasts for trading and socializing with neighboring villages.

     Important persons included chiefs. They often were "rich men." Their office appears to have been inherited to some degree, from father to son or uncle to nephew, but also was attained or maintained on a personal basis. Shamanism was prominent and was practiced by both men and women, some of whom were transvestites. In addition, there were herbalist curers and other persons with medical expertise. The office of wise-man or kasek, which organized religious ceremonies, was held by yet another specialist later equated to an Orthodox priest. Whalers had a special, but somewhat feared and "unclean," status. Finally, the Alutiiq are widely acknowledged to have owned slaves, but few details are recorded and slavery does not appear to have been essential to Alutiiq economics and society. It certainly would have been useful for chiefs to have slaves to procure and split wood for the sweat bath if it was like the banya of later times.

     Some aspects of the environment mentioned earlier merit further discussion. Shallow sea banks located off the southeast side of the island, today the site of highly productive fisheries, supported fish stocks, which in turn supported sea mammals and birds dependent upon fish and other resources of the banks for food. Aided by the long complex coastline, these resources then enabled the islands to support a dense human population. But there were natural and practical constraints that limited a person’s ability to make a living. Prolonged stormy periods made it difficult to hunt and fish, damp weather could spoil drying fish, and cycles and unexplained shifts of long duration occurred in the natural abundance and distribution of fish and other animals. Changes in ocean currents and temperature and climatic cycles may be factors. Limnologists studying lake sediments for diatoms and nitrogen isotopes are discovering that at certain times there may have been a marked decrease in the abundance of salmon. (Salmon have the nitrogen 15 isotope of marine origin. When red salmon die after spawning in lakes, traces of this form of nitrogen appear in lake sediments proportionate to the size of the salmon run, it is thought.). In addition, disputes with neighbors, or fear that raiders were going to appear suddenly from around the headlands, might have cut off access to certain territories. Thus, even in a land of abundance there were times of "belt tightening."

     The relatively high precipitation (around 80 inches or 200 cm) supplies a large number of streams, which, because of topography, tend to be short. This results in almost innumerable streams and "lagoons" where salmon, Dolly Varden char, and steelhead spawn. Most numerous are the pink salmon and red salmon. The latter ascend the streams until they reach a lake, but “pinks" and the other species (silver, chum) do not require lakes to spawn, though some also go there. In late summer, silver (coho) salmon go into Afognak Lake. Greater numbers of red salmon arrived as early as the end of April in former years when runs were large. There also were pinks during the summer. Sliver salmon linger in a necrotic state at the spawning beds well into winter, but salmon are available primarily during the late spring and summer months.

     Most of the interior of Afognak is hilly and even mountainous, except the east side, and offers few subsistence resources compared with the sea. Settlements and subsistence activities thus were oriented towards the sea, though salmon fishing camps sometimes were located above the mouths of major streams. Afognak River had the largest salmon runs, though Portage River (Perenosa Bay) and Malina Creek also bear noting. For the Afognak River, spawning escapement at the weir from 1991 through 1999 has varied between 66,767 and 131,374 red salmon per year. (The latter figure is an atypically high aberration prompted by artificial nutrient enhancement or "seeding," followed after 1999 by a drastic crash in numbers.) Between 193 and 8604 coho salmon, and between 8294 and 64,546 pink salmon, also pass through the weir yearly. Many additional pinks spawned below the weir. If it were possible to add in the commercial catch of fish headed for this stream, the totals would be higher.
Forests are limited essentially to spruce trees, and they were absent until the second millennium A.D. Older trees from the northern and southern edges of Afognak Island are about 800 years and 600 years old, respectively, but the first appearance of spruce trees on Afognak has not been closely determined. It is thought that these examples may be among the first spruce to take root on the island. Stands of balsam poplar and cottonwood (trees usually not differentiated by Kodiak residents) and Kenai birch are rarely found on Afognak, which contrasts with the case on Kodiak, where these deciduous trees not only form gallery forests along valleys but also carpet hillsides. However, poplar groves may have occurred on Afognak before the rise of the spruce forest. The most prominent vegetation for many thousand years prior to the arrival of the spruce forest was shrub birch, willows, and alders. Today, after decades of climatic amelioration, the alders and willows grow to tree size at favorable sites. Now, spruce has crowded the alders into patches in their former habitat. Prior to European contact, the primary source of timber for construction was driftwood, brought in by ocean currents and conveniently barked and trimmed through battering in the surf. In early times, after people had combed their home beaches for a generation, driftwood may have become scarce. Accordingly, good timbers were salvaged and reused when houses were abandoned or rebuilt. Even stacks of rock sometimes were used in place of corner posts. Today, most driftwood in the area originates through West Coast logging or other human activity.

     Major environmental changes affected Afognak in the past. These included partial melting of the Ice Age (Pleistocene) glaciers that had buried the land and subsequent, brief readvance of the ice about 11,000 years ago, and then, within a few centuries, final emergence from the Ice Age. This was followed by colonization by plants and animals culminating in the arrival of people. There is a time gap, years when the islands simply lay there before human colonization. Probably the bears got to the island before people, and most certainly sea mammals did, but little is known about the immediate postglacial paleontology. This occurred during a time of encroaching seas, before the relationship between the edge of the land and ocean stabilized roughly 6,000 years ago. (Dates proposed for this event vary widely.) Land-sea relationships posed a complex scenario of waters rising as global ice caps melted (eustatic rise), the land rising in turn as it became free of the weight of the ice (isostacy), as well as further changes in level, up or down, as the earth's plates shifted and earthquakes rippled through the crust (tectonics). On the whole, except for the short-term effects of earthquakes, and gradual loss from coastal erosion, the shore probably has been near its present position for approximately the last 6,000 years. Earlier edges of the land now are under the ocean because of the eustatic rise in sea level, but it is not safe to generalize inasmuch as rising sea level was partially offset by the land mass rebounding after it was relieved of the weight of the local ice cap. Additionally, the ends of the Kodiak archipelago have been affected differently by tectonic uplift and depression. The actual shoreline is constantly being cut back by powerful storm driven seas, as is readily documented by observations that extend back decades, by Alutiiq settlements, and by structures, including WWII fortifications, that are tumbling over the cliffs. All this is in addition to the immediate effects of earthquakes and subsidence in 1964.

     There were climatic changes of a magnitude that would have seriously impacted farmers but may have had little impact on the maritime hunters of Afognak. Nevertheless, there could have been significant changes in oceanic currents and temperatures affecting the abundance and distribution of fish stocks and of sea mammals and birds dependent on fish for food. These remain to be adequately documented. Analysis of food refuse from archaeological middens, especially fish bones and the isotopic composition of clam shells, should provide pertinent, though indirect, information. For instance, we noted that analysis of lake-bottom deposits by Bruce Finny at the University of Alaska also is yielding information about fluctuations in red salmon runs during the last 2,000 years. The size of runs, i.e., the number of salmon that die in the lakes, affects the composition of gradually accumulating lake deposits. These deposits can be sampled, analyzed quantitatively for chemical content, and dated in increments representing the passage of time, in order to construct a time trend that indirectly represents the abundance of salmon, notably red salmon.

     Our rough understanding of climate is as follows: Soon after the end of the Ice Age, climates warmed rapidly in the Northern Hemisphere and, for a period, it was even warmer than today. The first settlers thus may have found Afognak to be a pleasant place. This “Hypsithermal” warm interval ended about 5,000 years ago with steep, fast cooling, followed by rapid warming, not quite to the level of the Hypsithermal. Then 3,500 years ago, the so-called "Neoglacial" began. There were oscillations and warm spikes in temperature, especially during the Medieval Warm Period about 1,100 years ago. (Note that some researchers deny existence of the Medieval Warm period outside Europe.) During the Neoglacial period, climate became cooler and probably wetter, according to the interpretation of pollen recovered from bogs on Kodiak and Afognak. Perhaps significantly, that is the time when the Ocean Bay tradition ended, Early Kachemak began, and a Bering Sea people, called the Arctic Small Tool tradition, appeared on the north side of the Alaska Peninsula, and even at Kachemak Bay, and possibly visited Afognak Island. Changes in the distribution of fish stocks could have occurred again, and some mammals peculiar to the Arctic and Bering Sea regions might have wandered south to Kodiak, but if that was the case, they did not stay. The Medieval Warm Period is coeval with the last centuries of the Kachemak tradition, but we do not link it to the development of the Koniag tradition a few centuries later. Later, early in the second millennium A.D., spruce trees arrived at the northeast end of the islands and began their halting march towards the southwest to reach only recently halfway to the end of Kodiak. The spruce now is filling in meadows and fields of fireweed left behind in its advance. More recent climatic cooling during the second half of the second millennium A.D., called the "Little Ice Age," coincides with the final blooming of precontact Alutiiq culture and may have been a cause of culture change, according to archaeologist Richard Knecht. There actually were a number of cooler and warmer periods within the last millennium, so without specification of particular dates, invoking the term "Little Ice Age" mainly shows a lack of precise information. Generally, the size of salmon runs increased during the second millennium A.D., although there actually were marked decreases during cold periods, such as was the case in one of the Little Ice Age episodes late in the millennium. Climate change undoubtedly affected sea conditions, including possibly the abundance and distribution of sea mammals upon which the inhabitants of Afognak depended, but presently little is known of the actual affects of these variables. (Temperature diagrams can be found in Kopperl 2003 and in various discussions and critiques of the "hockey stick" climate reconstruction for the past 600 years widely published in scientific reports and newspaper articles.)

     There also were short term adverse effects from earthquakes, tidal waves, and resultant episodes of increased shoreline erosion. Some villages had to relocate or move back from the edge of the shore. The net affect of this as far as research into prehistory is involved is that, along the coast, mostly late-dating sites have survived, although even most of those were washed away after subsidence in 1964. Littoral zone invertebrates that had been a source of food were killed off temporarily, whether the land sank or rose, but they came back after a brief period.

     It is interesting to speculate on whether any of Kodiak's limited indigenous land fauna (brown bear, red fox, vole, river otter, weasel, little brown bat, and ground squirrel) was introduced by humans.
Ground squirrels, especially the race found on Chirikof Island, were prominent early during the contact period as a source of pelts for parkas. Zoologists think that the squirrels are an introduction because of the sporadic distribution of colonies on Kodiak and offshore islands. However, it now appears that most ground squirrel colonies are ephemeral and readily appear and disappear. The squirrels are known to swim well, but it is unlikely that they could manage the distance from the Barren Islands to Shuyak, which is very close to Afognak, and the even greater expanse of sea from the mainland to Chirikof Island. The absence of this common species from middens (the food refuse deposits associated with old habitation sites), except on Chirikof where they are prehistoric, can be interpreted as supporting the opinion of some biologists that they were introduced at the beginning of the Russian period to replace more desirable furs that were diverted to the European trade. However ground squirrels actually were present and well established on Marmot Island at least by 1792 A.D., according to Russian narratives of exploration. In the 1760s garments of ground squirrel were seen on Kodiak. This date long predates fur farming and the introduction of blue foxes and comes soon after the establishment of Russians on Kodiak. They also were reported on Chirikof by the 1790s. Thus, the little animals appear to be an indigenous species. Their ephemeral presence at some localities (Pillar Mountain and the Coast Guard Station, Woody Island, Marmot Island, Tugidak, possibly Sitkalidak and Raspberry Island) may relate to human intervention, predation by dogs, or attempts to colonize marginally suitable habitat. If sites with preserved animal bones older than those presently known of the Ocean Bay tradition are discovered, they should go far to help establish the antiquity of some animals in the Kodiak Archipelago.
Bears and river otters could have swam to the islands, but foxes would have found the passage more difficult and weasels, voles, and ground squirrels even more so. These are animals that young Alutiiqs might have had as pets. But why were there no hare ("rabbits"), wolves, lynx, porcupines, marmots, beaver, muskrats, and ungulates (such as deer) before modern introductions? Possibly in another ten thousand years they would have reached the islands, one way or another.


     Three sequential archaeological traditions or cultures are recognized on Kodiak and Afognak. Kodiak and Afognak prehistory differ very little. There was some degree of continuity or transition between traditions.

     The first tradition, Ocean Bay, which appeared about 5,500 B.C., is punctuated by technological developments that stand out both within and beyond the Kodiak region, especially the development of ground slate implements. Then about 1,900 B.C., the Kachemak tradition developed. This is a basic old North Pacific culture with strong ties to ancient Eskimo cultures of the Bering Sea region, and with the Aleutian Islanders, and even to the Northwest Coast region southward to Vancouver Island. By 1,200 A.D., Kachemak had become basically the ancestral culture of the Alutiiq people encountered by the Russians in 1763. This latter has been called the Koniag tradition.

     The chronology used here is based on corrected radiocarbon dates, listed in Appendix III. Radiocarbon dating is one of archaeology's most powerful tools, but it errs slightly, usually by understating true age. Appropriate corrections can be applied on the basis of information derived from comparing tree ring (true) dates with raw radiocarbon dates. Now, calibrated (corrected) dates are supplied by the laboratory that does the radiocarbon analysis. This was not the case originally and variations of dates obtained many years ago often are cited because of differing data used for correction. Generally, dates of the last 2,000 years fluctuate by a few decades on either side of real age, but before then they gradually fall short. By 4,000 years ago, the shortfall has increased to about 350 years; 5,000 years ago it is about 600 years (i.e., a radiocarbon date of 4,400 years actually means the age is 5,000 years). Radiocarbon dates should be interpreted with a degree of latitude because laboratory methods entail statistical imprecision (the + "error"). Additionally, the date may be indirect. For instance, it is assumed that the age of driftwood burned in a hearth is close the date when the fire was lit, but that might not be the case because the driftwood may have been drifting and lying around some time.

     Both the artifacts and the nonmaterial culture of these traditions can be thought of in terms of functional groups, particularly (a) hunting and fishing, including boats and weapons; (b) tools including tools to make things or process materials; (c) household items such as lamps and pottery; (d) clothing and personal adornment; (e) ceremonial, ritual, and games; and also (f) architecture, meaning houses and other structures. In some cases, they can also be indicators of gender, occupational specialization, and of social distinction or ranking. In numbers, the tool category predominates, but the most common tools like hammer stones, abraders, and cobble spall (boulder flake) scrapers and knives are not very distinguished. They are essentially natural utilized stones.

     Artifacts usually were made of local materials, although some items were imported from the mainland, including certain furs, tusk-shaped dentalium shells, ivory, marble, amber, and jet (coal) used for ornaments and carvings, caribou antler, beaver and porcupine teeth for implement bits, possibly red ochre for rituals, and some food, especially caribou venison. Most of these were low-bulk items that traders could easily transport. Metal, mainly copper, was imported through middlemen at Cook Inlet. Its ultimate source was the Copper River, and possibly also Prince William Sound. Copper usually is late in time, less than 1,000 years old, possibly because the tribes living in the source areas did not realize how useful the metal could be and they had not developed trade in this metal. The extent to which iron was used is problematical. Small pieces of iron, used for tool bits, had been imported into northwest Alaska from Siberia fully a millennium before European contact. On Kodiak, evidence for iron prior to European contact is limited to apparent rust stains on artifacts that might have held iron blades or bits. When these materials are found, they show that there was trade by wealthy persons who could afford prestige items. The island was self-sufficient in resources, and it was primarily luxury goods that were imported.

     Locally-produced pottery partially replaced bentwood tubs or other containers for preparing food on southern Kodiak during the last 1,000 years, but it was not adopted on Afognak Island. Undoubtedly, though, Afognak Islanders knew about pottery, as it was briefly used a short distance across Marmot Bay at Anton Larsen Bay about 1,100 A.D., and several years ago a potsherd (fragment) was found at Malina Beach on the Shelikof Strait side of the island. Afognak Islanders would have seen pottery on the Alaska Peninsula as well when they went across Shelikof Strait to hunt caribou and trade. Finally, they might have married girls from southwestern Kodiak whose mothers made pottery. Area differences as seen, for instance, in the distribution of ceramics, show that the Kodiak archipelago was not inhabited by a single homogeneous tribe. Probably, some of the customs, games, songs, stories, rituals, and style fashions (ideational culture) that complemented technology first appeared among neighboring groups and then were adopted by ancestral Alutiit. Numerous region-wide features were held in common, and a shared language helped maintain a distinct Pacific Eskimo or Alutiiq culture area that persisted for several millennia, although differences at the sub-dialect level reminded people that theirs was a diverse area.

     The distinctive character of each of the three traditions of the past depends very much on styles of hunting implements and of a certain other artifacts. There also were differences in house construction and, evidently, in level of salmon utilization. Examples are the distinctively styled boat-shaped early Ocean Bay lamps, Kachemak stone lamps with animal sculptures, heavy grooved Koniag splitting adzes, and, at the micro level, attributes of Kachemak and Koniag labrets.

     Bentwood containers, basketry, kayak and umiak parts, weapon shafts, labrets, figurines and dolls, masks and other ceremonial gear, and game pieces are prominent at waterlogged sites where wood, cordage, and baskets have been preserved. The only waterlogged site intensively investigated thus far is a Koniag site at Karluk, and we can only surmise that similar items were present in earlier cultures before Koniag times, but have disappeared. There also has been very limited recovery of wooden artifacts from Late Kachemak sites. Waterlogged sites occur when ground water to rises into the site deposits, as happens at the base of a hill: the water excludes air, and this slows the decay of wood.
Some customs are inferred from indirect evidence. For example, massive accumulation of fire-cracked rock strongly indicates use of the so-called sweat bath, which appears to have been a steam bath or wet sauna similar to the Russian banya. This bath actually is recorded ethnographically and may have been an important adaptation to damp, cold conditions and a cure for the incipient hypothermia that a person would face after a long boat ride. Accounts state that sometimes kayakers arrived home so cold and stiff that they had to be carried ashore. The Koniags were exceedingly fond of the bath, and at their sites, accumulations of discarded fire-cracked rock that was no longer usable or had broken up are up to six feet thick. In the historically recorded Native bath, rocks were heated in the main room of a house and then taken to the appended sweatbath compartment or side-room (thus differing from the Russian banya).

     The terms used for arms and weapons–spears, lances, bayonets, darts, harpoons, leisters and arrows–can be confusing. A spear freely leaves the hand, i.e., it is thrown, and it does not have a line attached to it. The hand-held fish-spear or trident is the exception. If the spear has a feathered (fletched) shaft, it may be called a dart. Some darts are only slightly larger than arrows, but whales were killed with six-foot-long darts! An implement that is held and jabbed is a lance. Lances tend to be larger and heavier than spears. With a harpoon, a connection is maintained with the quarry by means of a line attached at one end to the harpoon head and, at the other end, to a float or to the projectile shaft, or held in the hunter's hand. Harpoons can be cast like spears, as in the case of harpoon-darts, thrust like a lance, or, shot from a bow, as in the case of harpoon-arrows used for hunting sea otters. Leisters are spears, darts and arrows with multiple prongs or tips arranged around the end or at a midway location on the shaft, used on birds and fish. Bayonet simply is a descriptive term for a long lance blade or spear head.

     In the description of each tradition, attention will be given to housing. Houses, together with clothing, protected a person from the environment. The house also placed a person in a family context. Household organization helped a person articulate with the rest of the community and may have organized the activities of junior members, the families of daughters for instance. (Newly married Alutiiq men usually went to live with their wives.) Houses are the supreme artifact. Today, realizing the importance of houses and their attributes for historical reconstruction and social analysis, field archaeologists investigating Kodiak prehistory are focusing on houses. This was not always the case as excavation techniques, such as digging a trench through a site, were not oriented towards uncovering whole houses. And some archaeologists, like Ales Hrdlicka who dug at Larsen Bay, were in too much of a hurry to take the pain to properly uncover and record houses. The size and layout of houses reflect the organization of households. Construction of large houses, like those found in Koniag tradition villages, implies planning, a household leader, and probably also wealth in order to assemble the necessary materials and secure labor to help build a dwelling. Earlier houses were smaller, easily constructed, and probably were the abode of nuclear families. It is instructive, too, to compare the features of housing in permanent villages with those of summer villages or fishing camps. Data for this comparison, suggest that houses were similar at both types of settlements, but sometimes were smaller and were more lightly built at the summer fishing camps. Houses also contain features that give clues about cooking techniques and storage. Knowledge of precontact Alutiiq housing on Afognak is derived primarily from the Koniag tradition village site at Settlement Point, supplemented by observations of surface outlines that have survived from old house pits. For the Kachemak tradition, three houses have been excavated, or partially excavated on Afognak, but more Kachemak houses have been investigated on Kodiak Island. Because of their age most early Kachemak house pits are much muted, or are obscured by overlaying Koniag refuse layers. The same observation applies to the Ocean Bay tradition. For the Ocean Bay tradition, only parts of houses have been excavated at Litnik, but others have been excavated by the Alutiiq Museum elsewhere on Kodiak Island.

     In order to live on Kodiak and Afognak, the first people to move to the archipelago would already have mastered the techniques and lifeways demanded of a North Pacific maritime hunter and fisherman, including sea hunting and fishing, housing adapted to stormy cyclonic weather, craft to get around on rough water, and waterproof clothing. They likely learned these things while living on the adjacent Alaska Peninsula before they ventured across Shelikof Strait for the Kodiak Islands. These people are referred to as the Ocean Bay culture or tradition, named after a locality on Sitkalidak Island. At 7,500 years ago, villages of the Ocean Bay tradition appeared almost simultaneously on the Shelikof Strait side of the Peninsula and on Kodiak Island. On the Bering Sea side of the Alaska Peninsula, there is an earlier “Paleo-Arctic” culture that shows links to the late-stone age cultures of eastern Siberia and central Alaska. Possibly Ocean Bay developed as an open-coast offshoot of the Paleo-Arctic tradition. Indications are that such a development was taking place in the Aleutian Islands and western Alaska Peninsula 9,000 years ago. Sometime later, their descendants found a new pristine land, the Kodiak Archipelago, and prospered there.

     The highly successful Ocean Bay hunters managed for several thousand years with a single type of non-toggling harpoon head, made in a range of sizes—– “if it works, don't fiddle with it” seems to have been their axiom. There also were spears in several varieties, including ones for birds and fish, some with flaked stone tips, and some with bone heads. It is not known if they had the bow and arrow, which is believed to be absent in North America at this time, but early Ocean Bay used very tiny flaked stone points, which might have been used to tip arrows.

     Early Ocean Bay had microblades set in slots along the sides of slender unbarbed bone spear points, a technological trait handed down from much earlier antecedents. A microblade is a thin, very regular, narrow (quarter inch or less) elongate flake with straight razor-sharp edges produced by controlled kapping techniques from specially prepared small blocks or cores of stone. Microblades formed ready-to-use cutting edges for small tools, and points. Those found on Kodiak and Afognak are 5,500 years old and older; then they were replaced with flaked stone points. It is not known what kind of game these points were used on, but they likely were used on sea mammals. Otherwise, they would have been for defense and aggression against other people, and possibly bears. Archaeologists like to point out that with microblades it is possible to produce a considerable length of tool cutting edge from a single block of stone. This is advantageous where it is difficult to obtain desired raw material, especially during the frozen winter, and a long term supply must be carried from place to place. But this consideration definitely does not apply to southern coastal Alaska. Traditional usage, rather than material economy, may be the reason why Ocean Bay people used microblades. There is evidence for the early use of spear throwers (atlatls) from throwing board pegs recovered by P. Hausler from the Rice Ridge site located near Chiniak. Later, in Kachemak times, toggle harpoon heads and additional styles of non-toggling heads appeared.

     Part of the necessary equipment for hunting at sea was a waterproof parka (kamleika). No garments have been recovered from archaeological digs in the Pacific Coastal area, but numerous delicate eyed needles that would have served for sewing water repellent gear were recovered at the Rice Ridge. Other Ocean Bay implements included stone lamps; adze blades; fish hooks; cobble spalls which were used as knives, probably to split salmon; scrapers for hides and also for sawing slate to make tools; and the occasional large grooved cobble, probably used for sinking a fishing line to the floor of the ocean. At the “Slate” site at the Afognak River mouth, 44% of boulder flakes had a smoothed edge from use in sawing slate, 21% had a natural more or less sharp edge because they were knives or rejects, and the remainder had been shaped into choppers or toothed objects. The earliest lamps were long, narrow, and pointed at one end and squared-off at the other end, and had deep bowls. They are so distinctive and unique to that culture that even incomplete lamps can be attributed to Ocean Bay at a glance. Very early in Ocean Bay, large parallel-sided flakes, almost shaped like a prisms, termed “blades,” were used as blanks for the production of stone tools. Blade technology was a carryover from the earlier Paleo-Arctic era on mainland Alaska and Siberia, as also were the slotted bone points armed with inset microblades. The adze blades also are distinctive to Ocean Bay. Usually, they were flaked to shape over both surfaces—top and bottom—and at the working end are shaped slightly like a gouge: slightly convex at the top and slightly concave on the underside. Grinding to sharpen the bit is limited to the convex (top) surface.

     On Afognak, Early Ocean Bay remains were recovered from the so-called Chert site, originally designated Afo-109 (now Afg-011). This site is an eroded remnant on the tidal reach of the Afognak River. Little of substance remained there when the site was excavated in 1971; only thin traces away from the shore remain today. Late Ocean Bay, described below, was found in the upper layers of the same site, farther up the river estuary in the basal part of site Afg-088 excavated in 2004, and across the river at the so-called Slate site, originally designated Afo-106 (now Afg-008), also excavated in 1971. None of these sites has any preserved bone or other organic artifacts.

    About 6,000 or 6,500 years ago, these people adapted the bone working techniques of sawing, scraping, and grinding to fashioning slate. Thenceforth, ground slate tools partially replaced flaked flint (chert) tools and later, in Ocean Bay II (Late Ocean Bay) times, slate grinding became the main mode of working stone. This handy technique became popular and spread down the coast to the vicinity of Sitka and beyond into British Columbia. Mainly pointed implements, probably weapons, were produced in slate. Most of them followed the elongate tapered outline of sawn slate strips or blanks, but some were copies of smaller flaked chert tools. Some of the long stemmed points and blades are truly like bayonets. They have no known prototypes, but it is suspected that the large slate weapons and tools incorporate the attributes of what formerly had been a flaked chert blade or tip, plus its bone or wooden haft. Many ground slate blades bear fine edge serrations or tiny barbs along the blade and stem and cut-line decorations on the faces. It is not clear that the cut lines were purely non-functional. They could they have held poison or have facilitated bleeding. A change of this magnitude often would be heralded a new culture tradition, but in the present case the adoption of slate technology was clearly generated within the Ocean Bay culture, it was an added feature. It did signal a kind of watershed in prehistory inasmuch as thereafter the characteristic technogy of the area was slate grinding.

     The broad, sometimes crescentic knife, commonly known by the Eskimo term "ulu," was absent until the end of Ocean Bay times. Cutting and butchering was done with double-edged blades shaped somewhat like large lance blades or daggers. Some experimental archaeologists have found that it is very difficult to cut through salmon skin with slate ulu-type knives. The same would apply to tough animal hides and whale skin. They did find, though, that sharp cobble spalls could be used to split salmon. We noted that there was an appreciable number of natural cobble spalls in the Ocean Bay II site at the mouth of the Afognak River. Such spalls or boulder flakes continued to be abundant in succeeding Kachemak tradition sites at the mouth of the river. The earliest ground slate tools were mainly pointed knives and spears. Perhaps pointing rendered these implements better for cutting into hides and salmon skins than the broad ulu-type knife. It was necessary to split the fish if they were to be dried, but not necessary if they were to be eaten fresh or cured in a pit as “stink fish.” The fact remains, however, that ground slate ulu blades are especially abundant at apparent salmon fishing camps and the premise that salmon skin cannot be cut with slate knives needs further examination.
Development of the ground slate industry provides an interesting case of innovation in which the technology for fabricating one material—bone and antler—was transferred to another material, slate. This development can be seen as being related to hunting on the water where there is less danger of breaking brittle points like those made of slate, and also to the local availability of slate at the Afognak River and elsewhere on Kodiak. Concomitantly, it probably took less skill to master slate grinding than chert knapping. Significantly, the first use for ground slate on Afognak was for projectiles and pointed implements. Settlements at the mouth of the Afognak River contributed significantly to the invention of ground slate tools. This development then spread eastward from Kodiak to Southeastern Alaska.
Not all ground slate tools were made from sawn-out blanks or strips. Some slate blanks were chipped or flaked to form. And some slate tools were left in their flaked state without any further work or finishing by scraping and grinding. Nevertheless, the appearance of slate sawing and scraping is almost coeval with the adoption of the ground slate tools. Interestingly, though, use of slate for tools started somewhat differently in Early Ocean Bay times with the modification or rough pointing by grinding of natural rod-shaped stones culled from beaches. The beach rod tools generally are 6 to 10 inches long, the pointing more often blunt than sharp. In the production of sawn strip blanks noted above, sometimes a slate sheet was simultaneously segmented into a number of strips. The sharp edge of a cobble spall formed a suitable saw. Cobble spalls also could be used to scrape slate. Usually, the slate sheet was sawn part way through from one side, turned over and sawn further from the other site, and then the remaining septum of stone was snapped. It is very easy to recognize saw kerfs which remain along the edge of tool until the piece is completely finished by grinding. Even then, perfect straightness of converging sides is a clue to how a piece was made.

     One of the best expressions of Late Ocean Bay, though without any preserved bone, is found at the mouth of the Afognak River at the so-called “Slate” site. People must have stayed at his location to catch and process salmon, but the site also was a tool factory. With free time between salmon runs and a good local source of slate, people took the opportunity the make large numbers of sawn slate tool blanks, as well as finished ground knives and lance or spear points. The number of tools and blanks produced was far in excess of immediate needs. The surplus undoubtedly was intended for use elsewhere during the year and for trade to other communities who did not have access to good slate. Chert flaking had almost died out as a manufacturing technique at this locality, but was revived to some extent later, and it continued thus during Kachemak times. There are two radiocarbon dates for this site. Similar finished and partially finished slate artefacts are found across the river, but there is much less early stage manufacturing debris.

     For the best definition of Ocean Bay, it is necessary to look to a site on Kodiak, Rice Ridge located near Cape Chiniak, where there has survived bone artifacts and bones from the game and fish caught. At the Afognak River sites, only stone has survived. Robert Kopperel has identified the faunal remains from the Rice Ridge site. Most remains probably are food refuse, though some may be from animals taken for their fur and hides (but if the flesh was not used, the bones should not be in the site deposit along with food refuse). All the animals found in later sites (see below) are found there. In addition, there are uncommon marmot remains, including mandibles on which the incisor tooth is sharpened for use as a carving bit. Sea otter bones are unusually common, much more so than later, and often in deteriorating condition. They may have been cooked. Brown bear are well represented, though not abundant. There was a large suite of bird bones, also fish bones and shellfish remains.
Ocean Bay people were fond of red ochre, with which they liberally dusted the floors of their houses. The largest artifacts found in their houses are large flat base stones, resembling metates, used for grinding red ochre. Some archaeologists have proposed that red ochre was used for tanning hides, but there are cases in which a thick (ca 5mm) layer covers a broad expanse or nearly an entire floor. To explain that by spills from the tanning process or decayed hide roofing seems feeble.

     Excavations have partially uncovered Ocean Bay houses. Rice Ridge seems to have had rectangular floors up to 12 by 18 feet in extent. There are numerous floors, but they are superimposed in a small topographically delimited area, which suggests that houses were rebuilt more or less in place atop one another. Family continuity is suggested. There was a succession of rectangular stone slab hearths, except near the bottom where the second-oldest hearth was enclosed in a ring of cobbles, and, below that, the oldest hearth was simply a place where there had been fires. Late Ocean Bay deposits there also revealed a dug-in or semisubterranean circular house 12 feet in diameter. A relatively small subrectangular depression within the historic Litnik fish camp appears to be an Ocean Bay house judging from the presence of a deep red ochre streak, probably the house floor. This feature shows that dug-in houses, often thought to have been for the snug winter mode of living, may even have been built for summer use.

     Evidence for substantial housing also comes from the Slate site, AFG 011, at the Afognak River, though the details are fragmentary. Here, boulders had been used in construction. Numerous irregular stone blocks of various sizes occurred in swath 18 feet wide. These probably are from a tumbled house or two, the front edge(s) of which was lost to bank erosion. Tumbled and disturbed house ruins would be difficult to interpret even when entire, and are almost impossible to understand when part of the structure is missing. Two stone lamps were localized next to three post holes in one area of this house feature. Some of the stones lay more or less atop others but not as a solid, coursed wall. An orange-brown volcanic ash that underlay the occupation had been removed down to glacial till from about half of this feature, probably for a house pit or to level the floor of a house situated on sloping ground.

     Boulders also were used to construct an Ocean Bay II shelter at the Blisky site on Near Island. There, disarranged stones roughly describe two ovals, each about 8 feet wide and 13 feet long. Within each oval, there is a hearth surrounded by rocks. Some of the rocks were piled four high, which suggests walls. The top of the shelter could have been made of skins supported by arched poles.

     Many volcanic ashes fell after the glaciers melted and the islands came out of the Ice Age. Intense volcanic activity on the adjacent mainland lasted (no volcanoes on Kodiak) until about six-thousand years ago, forming the thick brown, orange, and yellow bands sometimes referred to as "butter-clay." In time, the ash has been altered to clay. When it is wet, the butter-clay is very plastic and slippery. Early Ocean Bay settlements often lay directly atop, and even within the butter-clay deposit, which frequently provides an imprint of post holes and ancient house pits. The butter-clay is a composite of several volcanic ash falls. More ash or "tephras" also fell later, but with centuries between major ash falls. Thus, volcanic ash layers found higher up in the deposits can be used to help date and correlate sites. Finally, the highest layer, the Katmai-Novarupta ash of 1912 that smothers all Afognak sites, can be credited with protecting site surfaces from disturbances. On low-lying sites, there also is a reminder of the 1964 tidal waves: a secondary deposit of Katmai ash and sand stirred up from the head of Afognak Bay by the tsunami. The mouth of the Afognak River was a major settling basin. There the secondary deposits that settled out of the water are almost as thick as the primary tephra of 1912, from which they are separated by a very thin band of soil that formed during the 1912-1964 interval.

     Continuing exploration has resulted in the discovery of a substantial number of Ocean Bay settlements in the Kodiak archipelago, many of them overlain by Kachemak tradition occupations in a manner that suggests the one is derived from the other. After making reasonable allowance for the progressive loss of sites through erosion, it thus is likely that there were about as many settlements during Ocean Bay times as there were during succeeding Early Kachemak times.

KACHEMAK TRADITION, 1,900 B.C. to 1,200 A.D.

Early Kachemak

     Kachemak arose from Ocean Bay culture through many technological additions and modifications that occurred within the span of a very few centuries. This is relatively fast for a culture to change without conquest and assimilation or replacement. This tradition began during a period of cooler climate, called the Neo-Glacial. With the inception of the Neo-glacial, a long era of climate as warm as, and sometimes warmer than, that of today came to its end. Now the bays would freeze over during the winter and snow would lie on the ground until the end of May. Early Kachemak appears to have been a spartan, utilitarian, basic old North Pacific culture with crossties to the earliest Eskimo cultures found farther north. In fact, it could have been ancestral Eskimo. Our understanding of Early Kachemak is biased because, with the exception of a small collection from Cook Inlet called Kachemak I, artifact recovery is largely limited to stone tools. At Cook Inlet, Kachemak I has an archaic style of toggle harpoon head and labrets and greater use of flaked chert tools than had been the case in Late Ocean Bay. There also may be some Early Kachemak bone artifacts in an Afognak collection from Malina Creek. Most early Kachemak deposits excavated on Kodiak lack organic preservation and it has been surmised that, should an early bone and shell midden ever be found on Kodiak, it would contain bone tools and provide us with a more balanced view of Early Kachemak culture. That may not be the case. In 2004, the Alutiiq Museum excavated an Early Kachemak site on Uganik Island that had a thick band of well preserved fish bones, with a few mammal bones. There were almost no artifacts in the midden, and none that was outstanding. Above the midden, there was a major concentration of cobble spall tools. This site thus reinforces the impression of Early Kachemak as being dominated by crude tools.    

     Late Kachemak was more elaborate, with much attention having been given to ritual treatment of the dead with many aspects; to personal ornamentation, with a profusion of labrets, beads and pendants; and an oil lamp art that probably expressed male and female deities. Compared with Ocean Bay, there were many changes in fishing and hunting artifacts. This does not necessarily mean that there was less fishing earlier: there are alternative ways to catch fish, especially salmon.

      Kachemak origins are poorly documented, but test excavations made at site AFG-088 located at the mouth of the Afognak River have revealed a late Ocean Bay occupation followed by sparse occupation, where some tools show technological and stylistic continuity with late Ocean Bay, and followed, in turn, by a 3,000 to 4,000-year old early Kachemak occupation. The succession in place, in directly superimposed layers without any break or parting soil layer, from Late Ocean Bay to Early Kachemak, suggests that in fact there is a cultural (technological) transition at AFG-088. This interpretation is reinforced by the fact that on Kodiak such succession or transition also is seen at a number of sites excavated by the Alutiiq Museum. Across on the Alaska Peninsula at Takli Island, a 3,000-year-old occupation, Late Takli Birch, has been found. It seems to be thoroughly transitional from Ocean Bay to Kachemak technology.

     Some tools in the Kachemak kit continued to be flaked from chert (flint) which had been the dominate tool making mode of early Ocean Bay times. Little-modified stone slabs, bars, cobbles, and small boulders also were used extensively for whetstones, fish line and net weights, for very large hide scrapers, as material for stone lamps, cobble spall tools, mauls, ochre grinders, and for weights grooved and attached to a line. Cobble spalls (boulder flakes) often were shaped to make scrapers and choppers and also were used for saws. Sometimes slate broke with a very sharp highly angled (acute) edge that did not need further modification to make a slate slab ready for use as a knife. In a sense, the stony shore was the Early Kachemak hardware store.

     Early Kachemak made much greater use of grooved cobble and notched pebble weights than did Ocean Bay, probably indicating changes in fishing techniques or emphasis. Although cobbles and pebbles could be attached to lines and nets without having to be notched or grooved, Kachemak people and their Koniag successors did make great numbers of them, nonetheless. So we see some significance in their absence, wherever and whenever that occurs. One peculiarity still to be explained is that notched pebble weights, probably net sinkers, became smaller through time. At Early Kachemak sites as at AFG-088 at the mouth of the Afognak River, the notched shingles were very large, 10 cm long and longer, thus weighing several times as much as ones half as large found at Late Kachemak fishing sites. We do not think this was an adaptation to the river current, as smaller weights were used nearby at Late Kachemak localities.

     One style of grooved stone, found at coastal sites but rarely on the rivers, was grooved near one end, leaving a knob. It is difficult to see how a line could be attached to one of these plummets. This type became popular to the westward, on the Aleutian Islands, and at Kachemak Bay, too. Everywhere it belongs to the Early Kachemak and the Paleo-Aleut cultures of 3,500 years ago. Then it went out of style.

     Labrets, lip and cheek plugs, were now adopted, though hardly any have been recovered from sites on Afognak predating Late Kachemak. (They probably await discovery there; they were worn early not far away at Kachemak Bay and on the Shelikof Strait side of the Alaska Peninsula from Early Kachemak times onward.) The custom of wearing labrets may have reached Kodiak from Southeast Alaska and British Columbia, a region commonly referred to as the Northwest Coast, or from Siberia, where they had appeared earlier. Labrets have attracted some attention because they occur in many styles or variations, which may have indicated the wearer’s social or kin-group affiliation. Additionally, the use sometimes of exotic materials for labrets shows that they could have sent a signal of the wearer’s wealth and social status.

     The broad, straight to highly curved, back-hafted, single-edged knife or ulu, a hallmark of Eskimos, became common, though hardly so in the earliest years of the tradition. This finding from the Old Kiavak site on Kodiak in the 1960s was confirmed by work at Afg-088 in 2004. On Afognak, all recognized ulus are sheets of ground slate; such blades are not present in the flaked chert industry. Almost all those from Afg-088 are from the top of the site deposit and probably date to about 2,800 years ago or less, about 1,000 years after the inception of the Early Kachemak phase. But quite a number found washed out onto the shore of the river show traces of the Ocean Bay mode of shaping through scraping, which suggests that they may also have been made in earlier transitional Ocean Bay-Kachemak times. Another time-sensitive trait is the use of drilled holes to assist hafting ulu blades and double-edged flensing knives. None was found at Afg-088, though the sample is reasonably large when fragments and the beach collection are counted. Drilled holes appear at a moderate frequency during Late Kachemak times and during the Koniag tradition. By the latter date, after 1,200 A.D., there also were pecked and sawn holes which provides a further guide for dating undocumented collections.

      Toggle harpoon heads appeared on Kodiak and also in other areas of Alaska and British Columbia at this time. If they have any one point of origin, it has not been determined. This technological addition is of especial interest, considering that, for its entire 3,500 year duration, Ocean Bay sea mammal hunters had flourished without toggling harpoons. (Toggling harpoons completely penetrate through the hide of an animal, and then, due to the “toggle” design of the device, turn sideways like a button securely in its hole.) As was noted, we did not recover this artifact in Early Kachemak sites at Afognak, where early bone is not preserved, but very archaic-looking toggle heads have been found in Early Kachemak sites at Kachemak Bay. However, numerous toggle heads were found at the terminal Kachemak Aleut town site located at Afognak Village.

Late Kachemak

      Kachemak rapidly lost its Ocean Bay features, but it continued in a somewhat impoverished mode until about 200 B.C., when elements of art, ceremony, and decoration became commonplace. Art is a defining feature of Late Kachemak. Better preservation of bone is also a factor in providing a less impoverished developing picture of Late Kachemak. Data come from several excavated sites on Kodiak, from the 1999 and 2000 text excavation at the Afognak Aleut Town site, excavation of the Tsunami house in 2001, and excavation at the Salmon Bend site at Litnik in 2002. A fine Late Kachemak collection also has come from the Crag Point site located across Marmot Bay at the entrance to Anton Larsen Bay. Stone specimens also have been found along the eroded edges of Late Kachemak sites at the Afognak River.

     Tools now include an array of bone implements: awls, delicate eyed needles, whale bone wedges, harpoon heads and sockets placed at the end of the harpoon shaft, arrow heads and spear prongs, pins for fish gorges and composite implements, fish hook shank and barbed parts (the two were lashed together to form the complete hook), fish effigy lures, and sockets for stone adze bits. Most of these items had been known 2,000 years earlier in Ocean Bay times, so possibly their absence from Early Kachemak is due only to poor preservation and inadequate exploration. Barbed harpoon heads were very numerous at Aleut Town and reflect an emphasis on sea mammal hunting.

     Stone adze bits usually are small and not numerous compared with their abundance in Koniag times, possibly indicating that there was little heavy woodworking. Whetstones and abraders in various shapes, sizes, and materials served for smoothing wood, for finishing and edging slate knives and projectile points, and for edging adze bits. Tablets of stone with smooth flat edges may have been burnishers for embossing and smoothing wood. Complementing adzes for working wood were bone wedges. They were recovered in great numbers at the Aleut Town site, the 135 specimens being noteworthy for the very small scale of the excavation. Clamshell-shaped spalls from cobbles made expedient tools for scraping and other uses, such as sawing bone and stone. Tablets of gritty stone also were used as saws.

     Ornaments, mostly for human adornment became commonplace. These included both stylized and naturalistic human and animal figurines, doll parts, miniature harpoons, large hair or clothing pins, combs, "buckles," bear tooth trophies, cylindrical and sometimes globular and pendant-shaped beads in jet (a form of coal), amber, a red stone from Kachemak Bay, shell, ivory and other material, rings for the nasal septum or ear lobe, nose pins, and especially labrets. Labrets were made in several styles, sizes and materials but jet was desired. Wood labrets were most common in the succeeding Koniag tradition, and the case for Kachemak, for which surviving wood is uncommon, may have been the same. Wood labrets likely were for everyday wearing to avoid the weight of stone, to be replaced by ivory and jet when the occasion warranted. Jet also is lightweight, which may be one of the reasons for its popularity.

           Stone projectile points were made in great variety. Some flaked points are like those of the Norton culture of northwestern Alaska. These include small stemmed points, lanceolate and leafshaped points, and delicate long arrow heads or “war points.” The last were found in a cache of nearly 20 specimens at the Salmon Bend site and might have been acquired in trade from a “partner” whose home was on the shores of the Bering Sea, probably Bristol Bay. The “war points” could have been used on caribou. It appears that occasionally hunters went over to the Alaska Peninsula to hunt caribou, although to do so they might have had to make arrangements with the tribe in whose territory they were going to hunt. Visiting and hunting may have been joint activities. One type of barbed ground slate points had carefully-cut bases and barbs. Sometimes they bore a decorative design or so-called identification mark. Others of ground slate were small lanceolate tips. There also were rod-shaped slate points, some of which were plain, others that were copies of slender barbed bone arrow tips. The flaked stone industry, mainly red chert, is moderately strong at the Tsunami and Salmon Bend site. This stands in marked contrast to most other Late Kachemak sites, especially Aleut Town. At Aleut town, there were only two chert implements, three flakes, and three water-rolled pebbles. There may be a sharp temporal cline in chert use as Aleut Town is a few centuries younger than the two Litnik house sites.

     A small number of slate pebbles (four) with incised designs appear to be somewhat like incised pebbles of comparable age found at Prince William Sound, and likely are precursors of the incised figurines that were to be made in great numbers during the Koniag tradition.     

     From the circumstance of their location, it is likely that the sites at the mouth of the Afognak River were fishing stations. The locality later was called Litnik, the Russian term for “summer place,” i.e. fishing camp. At the Early Kachemak site, Afg-088, there were mainly notched net sinkers, also many ulu knife blades. This fits identification of the site as a fishing station, but the case for the two Late Kachemak houses is not that clear—their artifacts are not greatly different than those from coastal sea mammal hunting communities, except that notched pebble net sinkers are common. At Aleut Town, there were no net sinkers.

     The evidence from Kodiak and Kachemak Bay shows that the dead usually were interred within occupied settlements, keeping the deceased as members of the community. Kachemak burials have not been reported at Afognak, although scattered human bones were common at the Aleut Town site. Incomplete burials and scattered human bones are startlingly abundant in both Early and Late Kachemak refuse deposits (but not at Afognak River, where bone has not survived). One chopped femoral fragment from Aleut Town looks like a butcher job. The human bones often show cut marks and breaks, or have been made into artifacts. Some from Crag Point, across Marmot Bay from Afognak, and the Uyak site were drilled for attachment or suspension in some kind of marionette.

     At Aleut Town, sections of jaw with two and three teeth had been cut out of the skull and ground to an even surface across the front of the teeth, perhaps for inlaying some macabre artifact. They and all the scattered bones from that site were reburied there. These occurrences have generated considerable discussion of a Kachemak mortuary complex in the region comprising Kodiak Island and Kachemak Bay, and also of possible cannibalism. The various investigations on the Kodiak Archipelago and at Kachemak Bay have revealed a group of practices that would fit into such a mortuary complex. There are both flexed single burials, which are the norm, and mass graves. There are also secondary burials (reburials) of a bagged or boxed skeleton. Extended burial was rare. Mass burials were interpreted by Hrdlicka as evidence of massacres, but that is not necessarily the case. One grave at the Uyak site (Larsen Bay) contained parts of 18-20 individuals described by Robert Heizer as “skeletons piled indiscriminately; skeletons incomplete. Probably secondary reburial. Long bones split (for marrow?)” (Heizer 1956 "Archaeology of the Uyak Site"). A smaller, more or less similar human bone assemblage was found across from Afognak at Crag Point ( ). Current thought is that the mass graves are mortuary crypts that were reopened from time to time to add more bodies (sometimes disturbing earlier burials in the process), to add the stray human bones encountered during house construction, and possibly "rehabilitated" war trophies and remains utilized for rituals that no longer were wanted. Other mortuarial elements include burial with foxes (now skeletons) and an eagle, artificial eyes placed in the skull, extra skulls or trophies in the grave, and, at Kachemak Bay, a clay mask modeled to the skull as well as bones with drilled holes for suspension or assembly. Cut marks on some bones may be from ritual dismemberment, perhaps to render an enemy harmless, (Such “joint cutting,” a very widespread practice, in Alaska is documented in Athapaskan Indian lore.) Cremations have been found, but they are rare and may even be unintentional. Altogether, Kachemak treatment of the dead and use of human remains was varied, extensive, and decidedly bizarre.

     Later, Koniag burial customs, like Kachemak, also emphasized interment in occupied and abandoned sites or house pits, although cairn burial and disposition in rock crevices also was practiced. (Ocean Bay burial practices are unknown.) Parts of corpses, and even whole bodies, were stolen by whalers and used variously, but this probably involved only a small number of deceased.

     Most early Kachemak lamps are small, plain oval, and little modified from the large beach cobbles from which they were made by pecking out a shallow basin. They would have been used mainly for light, not as cooking stoves, as was the case in the Arctic and sometimes in the antecedent Ocean Bay culture. In Late Kachemak times, stone lamps became very elaborate (decorated) and large—up to 90 pounds in the case of a “breasted” lamp reportedly found Malina Creek about 60 years ago, later sold by Paul Herring to the National Museum of Denmark. There also were miniature lamps as little as two inches long. Some lamps obviously incorporate elements for participation in rituals, judging from carvings of whales, seals, bears, humans, faces with hands, human breasts, and presumed magic circles in the bowl or on the lamp exterior. Some, as found at Litnik, had a prow (as on a boat) at the exterior fore part.

     The area of lamp art includes the Alaska Peninsula adjacent to Kodiak, Kodiak Archipelago, Cook Inlet, and Prince William Sound. This is the distribution of Kachemak culture. The lamps or lamp deities may have been the focus of a ritual at Kachemak Bay, also on Afognak. In addition to female lamps, with breasts, there appear to be male lamps.

     Male lamps are identified at Kachemak Bay by Sphinx-like figures with outstretched hands in the bowl, but on Kodiak there are only human faces or heads in low relief and figures of undetermined gender in the bowl and on the exterior. From the Uyak site, there is a lamp with a platform in the bowl. We speculate that a carving of a deity was placed on the platform facing the flickering flame. A version of the lamp and household deity in which the deity figure was suspended on a cord above and slightly behind the lamp was found in northwest Alaska. At Kachemak Bay, large lamps of both genders were intentionally defaced, battered, and broken apart. These large lamps required a major input of labor and artistic skill for their execution. Their destruction thus must indicate no small level of distress and dissatisfaction with the lamp deities. This facet of Late Kachemak culture appears to have been absent from the Kodiak Archipelago.

     Exotic goods were obtained through trade. Two Late Kachemak houses excavated by Amy Steffian at the Uyak site, Larsen Bay, appear to have been home workshops for producing jet (coal) ornaments, particularly labrets. Such concrete evidence for craft specialization is rare in Alaskan archaeology, though it is documented in studies of historic cultures. The raw material probably came from the mainland. Many elements of Kachemak technology, as well as tool and projectile point styles and ornaments, are identical to those of the 2,000-year-old ancestral Eskimo Norton culture of the Bering Sea region, including the Naknek drainage of the Alaska Peninsula. Norton and Kachemak people probably visited across the Alaska Peninsula for trade and held “invitational feasts.” They also may have intermarried.

     Ancient Afognak was part of the broader community of ancestral Yupik Eskimo peoples. Ivory was among the material traded. Large ivory objects have been found on Afognak (in private collections) and at Karluk (in Alutiiq Museum). We have noted the cache of 20 Norton-style arrow tips found at the Salmon Bend site. They were made off the island and brought to Afognak. A distinctive red stone used for beads and nose-rings was popular in villages around Marmot Bay. It is thought to have come from Kachemak Bay. Beaver incisors which also came from the mainland were favored by both Kachemak and Koniag people for carving tool bits. Antler also was imported. There may have been caribou hunting forays to the Alaska Peninsula, but caribou bones are very rare among the faunal refuse in the middens. Nevertheless, tools made of caribou antler are common at some sites, such as the Uyak site at Larsen Bay.

     There is little information on Early Kachemak houses that is both complete and reliable. Most of them probably did not differ greatly from single-room Late Kachemak houses. At the Old Kiavak site, located southwest of Old Harbor, a trench intersected a 23-foot-long house pit; and a 28-foot-long cluster of hearths, post holes, and floor layers that may have been part of one structure. These dimensions are considerably greater than the size of better-defined Kachemak tradition houses. Three possible Early Kachemak houses uncovered during Hrdlicka's Uyak Site excavations, as reported by Heizer, were single rectangular rooms, slightly dug into the ground, measuring from about 8 feet square up to 13 by 16.5 feet. In each, there was a stone slab fireplace. Excavations at the Bliski site on Near Island by the Alutiiq Museum uncovered part of a floor dated to about 3,000 years ago. The complete structure possibly was subrectangular and may have measured about 18 feet across. It is identified as a semisubterranean house, dug into the ground almost a foot and a half, and probably covered with sod or thatch. Early Kachemak houses have not been investigated at Afognak Bay.

     The best data for Late Kachemak houses are from the Uyak site near Larsen Bay. Heizer reported small single-room rectangular houses, plus a circular one about 28 feet in diameter. Steffian's houses, from new excavations at the Uyak site (published in Anthropological Papers of the University of Alaska), show entry passages half to fully the length of the rest of the house, which was rectangular with a central stone slab hearth. The single rooms tend to be 13 feet square and have floor-level entrance passages. Most of a very late house of similar format was uncovered by the excavations at the Aleut Town site.

     At the Afognak River Tsunami site there was a single-room subrectangular house shown in Figures and . It was at the maximum 6 meters long and 4 meters (13 feet) wide. Unlike the Uyak site houses, it may have been without an entryway passage. However, an entry could have been present inasmuch as part of one end of the house was clipped off by another dug-in house. The house site was complicated by reoccupation after the tidal wave and the construction of an additional house off to one side. Four piles of stone blocks were found at the four corners of the Tsunami House. Two corners also had posts, two among the 14 posts that held up the superstructure. Roof beams probably extended from stone pile to stone pile or from corner post to corner post. These were not very high, and it probably was not possible for a person to stand up close to the wall. The roof then would have been built up to a greater height by cribbing successive tiers of beams. The top of any vertical wall planks would have been leaned against the lower, outer beams like in the construction of a barabara (Siberian Native term for semisubterranean structure widely used in southwestern Alaska; the Alutiiq term is ciqlluaq.). With a cribbed roof, unwanted projections and roots, such as would be on driftwood trees, could be left to stick up and to overhang the outer walls of the structure. This greatly simplified preparation of timber which would have been a laborious task using stone adzes.

     There were at least ten clay-lined pits in the floor of the Tsunami House. There had been more, but probably not all pits were in use simultaneously. The pits took up so much floor space that it would have been difficult to move about within the house unless there were plank covers. Excavation also revealed at least two earlier filled-in pits. Only at one was there a void air space covered with a large slate slab; others were loosely filled with soil similar to the site matrix. The pit lining was tan and yellow-brown butter clay derived from volcanic ash. This is unlike the case at the nearby Settlement Point and Aleut Town sites, where large pits were lined with blue-gray glacial till clay. The pits probably were used for storage, for food in particular, but no telling traces of their contents were recognized. One of the better-defined pits was 46 cm in diameter inside and 60 cm at the top (outside) after the clay lining was removed. It was 43 cm deep. One other pit was larger; none was deeper. For comparison, the clay-lined pits at the Uyak site excavated by Hrdlicka were up to 72 cm in diameter, but only 20 to 30 cm deep. Their height may have been reduced through damage after abandonment. This is considerably smaller than the large Koniag clay-lined pits found at Settlement Point. There was no charcoal, gravel, or fire- cracked rock fill to suggest that the Late Kachemak pits had been used for cooking. There was also a clay apron around one of the hearths at the very late Kachemak Aleut Town site. Such aprons also have been found elsewhere. Though the uses to which these pits were put it not clear, they obviously relate to food preparation and storage. The pits evidently relied on the ability of clay to keep water in or air out. Curing of fish and meat under anaerobic (airless) conditions comes to mind as a possible use, otherwise wooden boxes and baskets would have sufficed.

     There were numerous stone slabs on the floor. One covered a pit; others formed a shallow crypt that had slabs for both its floor and cover, as well as for the low sides; and one sloping group of slabs formed a localized pavement. Most slabs were on, or immediately above, the actual yellow-orange clay floor of the house, thus they were part of the initial construction or date soon thereafter. The reason for the slabs was not apparent. However, one day after much of the clay tephra (yellow-brown) floor layer had been exposed, we attempted to work while it was raining. The spectacle of four people slipping and sliding in a pit was riotously akin to some modern TV stunts. Obviously, the floor had to be kept powder dry or covered with some kind of flooring—grass and slabs.

     Afognak data also include a Kachemak dugout house at the Salmon Bend site at Litnik, described next, and two houses farther out on the bay at Aleut Town. They show that Kachemak people built relatively permanent houses, even at their probable summer settlements.

     The Salmon Bend house, located across the river from the Tsunami house, is unlike any other Late Kachemak house excavated thus far in the Kodiak archipelago. Its uniqueness lies in the fact that it consists of two rooms, not just one. Because of a lack of trained help, it could not be completely excavated. Work was focused on an apparently attached structure, on a stone slab pavement at the front of the house, and on running a one meter-wide trench through the house from front to back. This house was targeted for investigation because artifacts found on the beach suggested that the site would be Late Kachemak in age and because surface depressions showed that the two parts of the house might be joined. Excavation showed that the two structures actually were joined, not simply by a narrow tunnel, but by a broad one-meter-wide passage. The “annex,” which seems to have been the smaller part of the compound structure, was not completely uncovered, but the area where the two structures joined was fully excavated. The interior was found to have hearths and clay-lined pits; thus, unlike the side rooms of Koniag houses which were mainly sleeping rooms, it could have functioned as a house. Numerous stone slabs were found part way down in the fill. Apparently the annex was partially in-filled, either from in-place accumulation of debris or from trash thrown in from the adjacent house. Then it was floored by slate slabs measuring up to one meter long, and occupation continued. Some of the slabs were highly inclined, thus they might also have been on the roof.

     The stone slab pavement at the front of the house was overlain by a thin layer of artifact-bearing dirt from the site occupation. The slabs then overlay a much thicker deposit of occupational debris. This debris filled the end of a trench that extended into the main structure. And the trench itself had been dug into an earlier site deposit. It was on the floor of the trench, at its very end, that the cache of “war” arrowheads was found.

     A complex history of house occupation and rebuilding seems to be indicated. First, there was a dugout house with sunken entry passage. Then the house was abandoned or rebuilt and at that time the entry was filled with soil and refuse. Then the stone patio was laid down. When it was uncovered, the stone surface was very uneven but that might be due to slumping over the past 1,400 years.
Inside the house, there was an accumulation of soil one meter thick, but less deep in the center, so part of this accumulation may be material sloughed in from the walls. Near the center there were two hearths, now somewhat disarranged, and shallow clay lined basins, one round, one rectangular. There also were distinct post holes. The area of floor uncovered was too small for any pattern of post holes to be uncovered, but at the least a substantial superstructure and roof is indicated. There were large chunks of fire-cracked rock over one hearth, which suggests that it had been used to heat sweat bath rock that was not carried to a bath chamber, and it had not been cleaned out or tidied up when the house was abandoned. (In the Koniag bath, rock is heated in a hearth in the central room and then carried to a bath chamber; in the Russian banya it is heated and steamed in place.) There were many stone slabs and large blocks of stone in the outer part of the trench which might have been placed there as fill to eliminate the entry trench depression if the house pit was reoccupied. This is not clear, though. A projectile point like those in the “war arrow” cluster at the end of the entry was found deep inside the “annex” structure, and a fragment of one was found at the base of deposits of the main room in the filled-in trench. Thus these three points of reference appear to be coeval. It is not altogether clear when the “patio” appeared, but it appears to have been very late.     

     At Aleut Town (Afg-004), parts of two very late Kachemak semisubterranean houses were uncovered. Most of the north house was excavated, but one end had to be left so as not to disturb a government survey monument. If there was any entry, it was at the unexcavated end or at a corner that had been removed by historic construction. The area uncovered measured nearly four meters (13 feet) square and we surmise that it could have extended at least one meter farther in one direction. This is the same size order as the Uyak and Tsunami site houses. Floor features included three pits without any clay lining. One pit was covered by two large slate slabs, another pit was partially covered by slabs. The pits had no noteworthy contents, just voids and loose soil. Between two pits there was a slate slab hearth surrounded by a clay apron. Thinner slate slabs lay over the top of the hearth. They might have been a cooking surface. Six post holes were positioned along the east wall and near one corner. The west wall and part of the south wall were not exposed. There were two distinct dark floor streaks at the base of the housepit, but there were almost no artifacts on the floors. Evidently, the occupants had been tidy house keepers. There was later occupation within the same structure pit: a large stone lamp was found upside down on a partial sandy floor at 97 cm depth, while the main floor was at 144-154 cm and deeper. It is not unusual for lamps to be turned over, possibly in accord with some ancient belief.

     The two housepits at Aleut Town were positioned close together, about one meter apart, from which it is suspected that they might have been joined.


     The Koniags were the ancestors of the Alutiiqs prior to the historic changes effected by Russian contact and other outside impacts since the conquest of Kodiak in 1784. The term “Alutiiq,” (plural Alutiit, often anglicized to Alutiiqs), which first entered the area as “Aleut” through Russian usage, is useful for referring to the post-contact phase of the Koniag tradition. Some persons use "Alutiiq" also to refer to the inhabitants of the Kodiak archipelago back to the time of the initial peopling of Kodiak. This usage has a degree of legitimacy as the same people appear to have lived in the area from that time onward.

     The Native people of the Kodiak Archipelago are among the better-documented North American tribes. That is so because of early written accounts, referred to as historical ethnographies; artifact collections that date back as far as the last decades of the 1700s; and because of the remarkable recovery of ceremonial objects and other wooden artifacts from late-dating “wet” (waterlogged) sites. Koniag remains of the 600 years preceding the historic period are closely linked to the people whom the Russians subjugated. This is the starting point for backstreaming, or extending the historic identification of peoples back in time. It becomes more difficult to identify the carriers of Kachemak culture as Koniags or ancestral Alutiit, although that seems to be the case, at least in part. There is a considerable degree of cultural continuity between the Koniag and Kachemak traditions. It is much more tenuous to identify the Ocean Bay people of 7,500 to 4,000 years ago as ancestral Koniags (hence ancestral Alutiiqs), but again there is some cultural continuity between Ocean Bay and Early Kachemak, although the evidence is not strong. Throughout time, there may have been additions to the original population, but there is no firm evidence for complete population replacement on Afognak or Kodiak at any time by conquest and migration. The Alutiit are an Eskimo people. Their ethnicity, culture, and language undoubtedly have evolved or changed in situ. Otherwise we would be faced with the logical conclusion that if Ocean Bay was ancestral Alutiiq, it also is ancestral Eskimo. To say that ancestral Eskimos lived on Kodiak and Afognak 7,500 years ago strongly favors placing the very origin of Eskimos in this region. Most prehistorians and linguists are not enthusiastic to endorse such a proposition, though many will accede that Eskimos may have originated, or that Eskimo culture developed, in the area extending southward from Bering Strait to Shelikof Strait.

     The Koniag or Qikertarmiut (Island) Eskimos, among whom are counted the Alutiiq ancestors of the Afognak islanders, numbered about 8,000 at time of Russian conquest in 1784. However, the existence of many large abandoned second millennium A.D. Koniag archaeological sites on all the islands and inlets of the archipelago suggests that there may have been even more people earlier.

     Alutiiq hunting techniques were so well-adapted that they continued in use after European contact in order to meet the requirements of the fur trade and colonial subsistence. Much of Koniag technology was similar to that of the preceding Kachemak tradition, although there also was ongoing development. Some of the tools now used on Kodiak are stylistically similar to ones used along the Bering Sea coast, as well. The meaning of this has been the subject of scholarly discussion. It is debated whether there was a major migration from the mainland or southern Bering Sea to Kodiak. Nevertheless, Koniag culture came to resemble in many specific ways that of Bering Sea Eskimos, while maintaining links to the Aleutian Islanders and Northwest Coast peoples. Some Koniag artifacts are highly distinctive to the Pacific Coast, and in this manner set the Koniags apart from more northerly speakers of Eskimo languages. Examples are petroglyphs, long slender ground slate whaling dart tips employed in a method of whaling very different from Eskimo harpooning, use of aconite poison, probable mummification, and spruce root basketry. The last is found even where there are no spruce trees, as at Karluk, necessitating that the roots, or the baskets, be imported. In addition, there was a local type of kayak (baidarka) different from even that of the adjacent Aleutian Islands and Bristol Bay.

     The Koniag tradition, beginning about 1,200 AD, is confined to the brief period of 600 years prior to European contact (plus a post-contact phase). Some archaeologists propose a Kachemak-Koniag transitional stage of a century or more in duration, although this is not well- defined. Most archaeologists also propose an Early Koniag phase 1,200-1,400 A.D.. This is followed by a Late Koniag phase, interpreted by some as “Developed Koniag,” but identification of certain artifacts as “Transitional,” “Early Koniag,” or “Developed Koniag” remains highly uncertain. The sequence proposed for diagnostic period or phase artifacts at Karluk may apply only imperfectly to other parts of the Kodiak archipelago. Thus the basis for this scheme of periodization is weakened, as are the concomitant assumptions regarding the development of cultural complexity in Koniag society and lifeways. For instance, on Afognak splitting adzes, incised pebble figurines (both said to be “Developed Koniag”, boat or arch-shaped projectile end blades and long ground slate spear points with diamond cross-section (said to be “Early” or “Transitional”) are contemporary, unlike the sequencing of artifact styles at Karluk. Nevertheless, not all change was simultaneous, nor was it instantaneous. Kachemak and Koniag styles coexisted for a while, as in the case of the shape of compound fish hooks. The actual style differences are too arcane to be discussed here, as is also the case for many other Koniag-Kachemak distinctions. Some features of Koniag and Kachemak stone lamp decoration occur together for a while. Change from Kachemak to Koniag, though not necessarily sudden, was more than a subtle transition. Almost everything that had been Late Kachemak changed in the succeeding Koniag tradition during the centuries 1,000 to 1,400 A.D, but the differences often are very slight, in the position of the line hole in a harpoon dart head, for instance. Nevertheless, some changes were not so minor and have major implications for social institutions and regional interactions. Pottery, for example, appeared locally, but never became adopted on all parts of Kodiak, and is missing from Afognak. This distribution demonstrates local differences that existed among the Alutiit, who probably were not a single tribe. Late in time, the Koniags also obtained a few tools pounded from native copper, a material traded from the mainland to the northeast.

     An important addition to the tool inventory was the heavy grooved splitting adze. This implement is especially common on Afognak, and also at Prince William Sound, and might be correlated with the need for split wood to heat rocks for the now very popular sweat bath. In the Koniag tradition, there also is increased evidence for heavy woodworking. Small planing adze bits became one of the most abundant tools. Often, they were converted into hammer stones after being damaged, but sometimes they are found in perfect condition. Whalebone wedges for splitting wood are very common in both Koniag sites and in the Kachemak tradition, as at the Aleut Town site, but at the latter, adze bits are uncommon, whereas they are very common in Koniag sites. One site on inner Afognak Bay had so many greenstone adze bits that we gave it the name “Adze.” Splitting adzes also were common there.

     The most fascinating artifacts are beach pebbles and fragments of split slate sheets upon which stylized human faces have been scratched. Often the figurines are clothed. A few examples hold a hoop rattle, and thus provide suggestive clues that the figurines portray dancers at festivals. But they are found at both main, or winter, sites and summer salmon fishing camps, so their use was not limited to seasonal festivals. Some details of clothing and personal adornment, such as labrets, can be seen in the figurines, but there is no portrayal of weapons, boats, interacting persons, groupings of artifacts, drums, etc. They appear mainly in Koniag times, used in a ritual for which thousands of these figures were produced and discarded at such sites as Settlement Point on Afognak, Adze (near Litnik), Marka Bay, Monashka Bay, Gull Light (Uganik Bay), and Karluk. Some incised tablets are so rudimentary, showing a few facial features only, that they could have been scratched out and discarded in two seconds, but at one site on Spruce Island, people sometimes took the trouble to grind or rub the figurine blank smooth. Others are very detailed and elaborate, portraying strings of beads by tiny circles, labrets, clothing fringes, feathers, possibly hair, sewn clothing panels, and elaborate parka collars. These pieces were not made hastily. While some are “clothed,” some look uncovered and appear to portray female breasts. For most, there is no clue to gender. The figurine “cult” appears to have lapsed at least a century before the Russians arrived. But by the time it went out on Kodiak, the practice, and possibly some of the underlying belief, had been copied by ancestral Tlingit Indians located from Yakutat to Sitka. This provides a clue that relationships around the Gulf of Alaska were not always hostile and that Alutiiqs and Tlingit ancestors interacted socially before the Russians arrived.

      A variant style of incised figurine produced in small numbers often portrayed a stylized backbone and ribs, and has been found in Late Kachemak sites at Aleut Town and Monashka Bay, as well as at Prince William Sound. Use of a skeletal motif could have played into a belief system that emphasized the importance of bones, as was found among many Native peoples. In this respect, shamanism comes to mind. The early examples might belong to a system of belief and ritual that later fluoresced with the production of more numerous human figurines. Their graphic expression was in the form of humans, but were they really humans, animals, or spirits? There is one other case in Alaska in which large numbers of incised pebbles were produced. This is in the Ipiutak culture of about 500 A.D., possibly slightly earlier than the earliest incised pebbles from Kodiak and Afognak. But the areas where the two occur are separated by several hundred miles. Kodiak-style incised tablets are found as far to the northwest as the Naknek River drainage of the Alaska Peninsula. Ipiutak examples, which appear to portray caribou heads in a very stylized manner, are found mostly north of Bering Strait.

     A large Afognak Bay collection, and much information on houses and diet, were recovered through controlled excavations led by Patrick Saltonstall at Settlement Point for the Afognak Native Corporation. Eight houses are recognized from surviving surface outlines. Every house was tested archaeologically. Radiocarbon dating shows that one house was occupied about 1,350 A.D., with all the rest about 1,500 A.D. and 1,600 A.D., give or take a few decades. It was possible, though, through geological and archaeological data, to further refine the sequence of house construction and abandonment. For example, artifact styles and radiocarbon dating suggest that occupation was during the early part of the Koniag tradition, although it continued until somewhat later. Additionally, when House 1 was abandoned, the pit was filled with midden refuse generated by neighboring households, where continuing occupation was more recent. One of these was House 2, which itself was abandoned before the circa 1,600 A.D. tsunami, as tsunami sand is found in the old slumped house depression. The site is not likely to have been occupied much earlier than House 1, as there had been a large earthquake about 1,200 A.D. The site terrain subsided below the waves, which built a new storm berm that can be seen today behind the houses. Then the land rebounded and people built houses on the beach in front of the berm. In another major earthquake that occurred about 1,600 A.D., the land subsided at least 70 cm, some houses were flooded, and the site was abandoned. This time, the broad gravel berm behind the contemporary beach was built. Many dwellings may have been occupied concurrently, thus a village population of between 80 and 120 persons can be proposed. It is clear from the radiocarbon dates that there were at least two generations of houses; Saltonstall feels that three generations are likely.

      Houses were heated from the earliest times (Ocean Bay) onward with interior hearths, which may have been used primarily for cooking. With warm parkas to wear, people likely did not often need a heated room. But a little heat would have been useful to drive out dampness and to keep food and equipment from mildewing. Stone lamps were for light, though they sometimes reached an impressive size of 100 pounds. There is a persistent story of a lamp somewhere on Afognak that is “too large to move.” Lamps of this size would have served ritual and community needs. Though the lamps often were well-crafted in a distinctive subrectangular form with trimmed vertical sides, others were simply small boulders with a hollowed surface. Many conform exactly to a single distinct style with broad flat rim, border hanging over the sides, sharply formed small wick shelf, and subrectangular shape. The shape could be rounded out to circular. Decorated lamps hardly exist at all. Stylistic uniformity as seen in these lamps is an interesting aspect of Koniag culture. Such lamps are found eastward to Prince William Sound, and as far north as the Nushagak River, but the extremes of distribution of some may have been achieved through trade or gifting.

     Larger, multi-roomed Koniag houses, compared with Kachemak houses, are interpreted as evidence of population increase, as well as evidence for a change in family social structure and community organization. The overwhelming abundance of complex Koniag-style housepits at salmon streams, such Portage River (Perenosa Bay), the Karluk River, Ayakulik River, and elsewhere further suggests population increase during Koniag times, though that remains unproven. A short period of occupancy could account for the large number of houses. There also are many small Late Kachemak houses on the streams, each one hardly large enough to hold a nuclear family.

     Implements found at Settlement Point include bone arrows used to hunt birds, harpoon heads of both toggling and barbed non-toggling barbed formats, fish hook shanks and barbs, notched cobble line weights, hones and whetstones, and numerous greenstone planing adze bits, which is a characteristic of the Koniag tradition. Unlike at most Koniag sites, there were almost no splitting adzes. This supports the view of some archaeologists that splitting adzes were rare before 1,600 A.D. Other artifacts include elongate ground slate dart or spear heads with diamond cross section; ground slate ulus; labrets, including some fashioned of ivory or of jet; a rare imported tusk-shaped dentalium shell bead; stone lamps; and several hundred incised pebble figurines. The spearheads with diamond cross section are a distinctive regional “horizon” style that spread quickly throughout the region from Prince William Sound to the western Alaska Peninsula early in the second millennium A.D. They show that the peoples along the coast were “connected.”

     Another distinctive style of Koniag arrow or spear tip that appeared somewhat later, and was used into historic times, is the arch-shaped or boat-shaped blade that was inserted into the end of arrow and toggle harpoon heads. Initially, the blades were plain flat boat-shaped pieces. Later, they were grooved or slotted along the faces near the base. Finally, they were provided with a carved bed at the base, for insertion into a narrow slot at the end of the spear or arrow. This style (actually two sub-styles) is well represented also from sites that border Bristol Bay and the Nushagak-Kuskokwim River region. Again, widespread contact between Eskimo villages is indicated. Shelikof Strait and the Alaska Peninsula were not significant barriers. Recovery of dentalium confirms that trade goods from British Columbia reached Kodiak in precontact times. Previously on Kodiak, dentalium was limited to ethnographic paraphernalia of the historic period.

     The Koniags developed a complex rectangular house with a large central common room, nine yards square in one early Settlement Point example (House 1), and with several appended side chambers. Each attached chamber was occupied by a family which cooked in the common room. One room was for the steam bath. In the early 19th century, a house could have had 18 to 20 occupants. Houses were dug into the ground (semisubterranean). Dwellings built in this manner may have been easier to heat and provided better protection against storms and cold than ones built fully above the surface. Koniag construction, and probably that of its predecessors, was of the post and beam mode. Posts were erected—low ones near the walls, higher posts near the center—then beams were strung from post to post. Split slabs of wood were leaned against, or placed over, the beams. These in turn were banked and covered with turf for insulation and thatched to shed water. Portions of the walls may have been built of stacked sod alone, but they could have been faced with matting. Driftwood was the source for timber, and we can surmise that it was in short supply after local accumulations were used up during the initial settlement of a locality. Accordingly, floors were covered only with grass, and sometimes gravel, which was renewed from time to time. Dried food was stored inside the house. The little-cabin-on-posts type of cache appears to be a Russian introduction, though earlier there probably were racks or stages on posts for keeping kayaks and other paraphernalia off the ground. At Settlement Point, there were large clay-lined storage or food preparation pits within the houses. Kachemak houses also had numerous clay-lined pits, but of smaller size.

     These large multifamily Koniag houses were very different from the small Late Kachemak houses. As Saltonstall expresses it, “Four hundred years prior to the occupation of Settlement Point, Alutiiq peoples lived in small (c. 20 square meters) single family dwellings. Yet when the Russians arrived in the late 1700’s they reported that the Alutiiq had chiefs, kept slaves and that several related families lived in one large (c. 60 square meters), multiple roomed house.” Pits from abandoned houses are large rectangles with one or two smaller subrectangular pits on the outside along each wall, joined to the main room by a short passage.

     This type of Koniag house apparently was developed soon after or concomitant with the beginning of the Koniag tradition, which was sometime between 1,200 and 1,250 A.D. As yet, the details of its history are not fully understood, but research by the Alutiiq Museum is working out the chronology of such features as sunken or cold-trap entries, small appended corner rooms, the proliferation of appended rooms, and enlargement of the main room. A large fully-developed example with many distinctive architectural features is Settlement Point House 1, dated to between 1,300 and 1,400 A.D. Elsewhere, in 2004 Saltonstall found smaller houses with relatively small main rooms, fewer or no distinctive architectural elements such as stone slab boxes and clay-line pits, and only one or two attached chambers that date to 1,100-1,200 A.D. These are some of the earliest Koniag houses on Kodiak. At this time, we do not know if their characteristics precede developed Koniag or are a local peculiarity of the site, which is on Uganik Island. We noted from Afognak that Late Kachemak houses sometimes had two rooms, each reflected by a separate structure depression and its own hearth.

      The nine houses that comprised the Settlement Point village date to the three centuries immediately preceding about 1,600 A.D., when the site was abandoned because of subsidence following a major earthquake. Some houses had multiple floors, indicating refurbishment. The floors consist of alternating layers of clean beach gravel and hard-packed house floor. When the floor got too dirty with seal oil, food remains, and charcoal, the inhabitants recovered it with clean beach gravel, and possibly also grass. Grass was the common floor covering elsewhere in the Kodiak region,and it has survived at one site on Afognak Bay, the one called “Adze.” It appears that a high degree of wetness is needed for grass floors to survive long enough to become part of the archaeological record. Houses sometimes were rebuilt in place, establishing several centuries’ continuous occupation, at Karluk, for instance. Stacked blocks of turf formed the walls, though the walls of side-rooms, and perhaps some other parts of the house were lined with driftwood planks. Turf walls now have become thick deposits of brown soil. Excavations revealed holes left from the many large posts required to support the sod covered and thatched superstructure. (Most posts had been pulled for reuse.) Each house had a hearth in the main room, contained within upright slate slabs. In some Settlement Point houses, there are pits under the hearths full of charcoal and containing also fire-cracked rock. Saltonstall suggests that these may have been barbecue pits. To use them, meat would have been placed upon a bed of coals and heated rocks, and then covered with soil (and probably grass or sea weed) or gravel to cook slowly.

     Each house pit at Settlement Point had between one and eight side rooms. The appended rooms were slightly higher than the main room, and were entered through a hatch, it seems, and a tunnel dug down about 16 inches lower than the floor. This type of entry formed a cold trap to prevent cold air from flowing out of the central room into the side chambers. These rooms were relatively small, about 10 feet square.

     In addition to the hearths and charcoal-rock cooking pits, each Settlement Point house has several storage or food preparation features built into the floor. These consist of clay lined pits, slate boxes and, evidently, wooden boxes. The largest house has eight clay-lined pits, three slate boxes, and traces of one wooden box. The clay-lined pits are much larger than comparable features found in Kachemak sites, being 3 to 6 feet in diameter and 1.5 feet or more deep. There is tentative evidence that the clay pits contained salmon. The salmon would have kept cool in the subfloor environment, and the pits, being made of clay, could have been sealed at the top to produce an air-free chamber in which the fish would ferment, but would not spoil. Slate boxes also were set into the floor. The slate lids had a slot for a hand grip. In one house, the lid and sides of the box were sealed together with clay and a hole through the lid also was capped and sealed. But the contents had disappeared without a trace: the interior of the box was an utter void even though the seal had not been broken for five centuries. The Settlement Point investigation places food storage and cooking at a more prominent level than previously had been thought to be the case for the Kodiak Archipelago. The evidence includes recognition of charcoal and rock-filled “barbeque” pits.



      Ocean Bay Kachemak Koniag

      5500 BC 1800 BC 1200 AD

      Microblades x x x x x

      Major red ochre use x x x x x x x x x

      Chert Industry x x x x x x x x x x x x less x x x x x x x rare

      Delicate needles x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

      Ground slate tools x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x

      Slate ulu knives ? x x x x x x x x x x x x x x x x x x x x x x x x x

      Giant-size ulus ? ? ? ? x x

      Sawn & scraped slate x x x x x x x x x x x x

      Labrets ? ? ? x x x x x x x x x x x x x x x x x x x x x x x x x x

      Pebble net weights uncommon x x x x x x x x x x x x x x x x x x x x x

      Cobble line sinkers uncommon x x x x x x x x x x x x x x x x x x x x x x x x x x x x xx

      Toggle harpoons ? x x x x x x x x x x x x x x x x x x x x x x x

      Harpoons with

      line holes ? ? ? x x x x x x x x x x x x x x x x x x x x x x x x x

      " " with shoulders x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x less common

      Lamp deity ? ? x x x x x x x x x x x ?

      Ring & Pin game ? x x x

      Clay-lined pits ? ? x x x x x x x x x x x x x x large x x x x

      Steam bath xxxx? x x x xx

      Planing adze uncommon uncommon abundant

      Splitting adze x x x x x x x

      Boat-shaped end blade (projectile tip) ? x x x x x

      Long hollow-ground

      ridged projectile tip x x x x x x ?

      Large house ? ? ? ? x x x x x x x x small x x x x x x x x x x x x x xx x x x x x x x x x x x x x x x x x x x

      Pottery x x x x x x x x

      Slate figurines ? x x x x x x x x x

      Head flattening ? ? x x

     Some clay pits and a few slate boxes were filled with fire-cracked cobbles that differ from the usual slate “banya rock” also found at the site. These appear to be cooking features. Pits, boxes, and other features in the house floors are so numerous that it almost would have been impossible to move about inside.

     Some pits probably had been abandoned and had been filled in while others were in use. Careful investigation of the entry to one house shows that it probably was a cold trap. In this type, the roof of the entry is lower than the floor of the interior room, thus keeping cold air from flowing in. (It would be better to call this “a warm trap.”) Although only the elevation of the entry floor was observable, the roof could be adjusted through the use of baffles to lower it below the floor of the interior room. Where well-preserved house depressions are present, as is the case from the 2003-2004 Alutiiq Museum surveys of the Ayakulik River, depressions can be seen in house and side room entries that apparently indicate cold-trap construction. The initial entryway where the cache of war points was found at the Salmon Bend site described earlier may have been a cold trap, but after it was filled in, that no longer would have been the case. This mode of construction probably was imported from the north. Its presence on Kodiak and Afognak needs to be better dated.

     Additional Koniag houses were partially uncovered in the 1951 and 1996 test excavations at Adze, located across from Rivermouth Point (AFG-012). One feature found there, but not at Settlement Point, is a stone-slab-covered sub-floor drain. A house pit had been dug down to impervious glacial till, hence the need for the drain. The Settlement Point houses built on gravel did not need drains. Hrdlicka found well-constructed drains with slate slab walls and covers at the Uyak site, but he misinterpreted them as walks. None of the Late Kachemak houses described earlier had floor drains, probably because the local topography was slightly raised and drainage was not required. Multiple microstratified floor layers, consisting mainly of compressed grass, and boulders piled as many as five high and which probably were part of a wall, were seen at the eroded face at Adze. The wall is of especial interest inasmuch as otherwise “piled” stone construction has not been demonstrated for Koniag times.

The Historic Period

     Historic period deposits and houses have been excavated at Katanee (Quataat), the small short-term Kazakovskii site, and at the Aleut Town. Also, historic Alutiiq-type houses and historic activity areas have been identified near the mouth of the Afognak River at Litnik, at Malina Creek, and at Little Afognak.

Kataaq in the Russian Period

     This occupied place near Settlement Point, on the east side of Afognak Bay, also is called Katanie. There are extremely few references to Kataaq, probably because in most instances the inhabitants are merged with nearby Aleut Town and Afognak odinochka. There is a report that when Saint Herman died in 1836, a column of light, joining heaven and earth, appeared in the sky over Spruce Island, and was seen by the inhabitants of Kataaq, including Anna Nytsmyshkinak and her Creole husband Gerasim Vologdin (Korsun 2002: 97-98). This is the earliest documentary date we have found that we have to associate with the settlement by name, but indications of a settlement in this vicinity appear on Russian charts as early as 1786-1787. A zilishche [habitation or village] is indicated on the engraved chart compiled by Vil’brekht in 1786 (based on the surveys by Izmailov; published in Black 1992:Fig. 6). This unnamed zhilishche apppears to be either on an islet or promontory in the vicinity of the mouth of Kazakof Bay (Danger Bay), but shoreline details are rudimentary and imprecise. Islets in this area are not suitable for permanent settlements. In 1787, the Russians set up an artel a few miles away at Igwik or Little Afognak.   

     Kataaq is not in the 1795 and 1804 censuses of Alutiiq villages, probably because it was established later. Moreover, Katharine Woodhouse-Beyer’s excavations (2001) at Kataaq have not brought up evidence of a precontact occupation at the site, which helps set it off as a postcontact settlement. She recovered the foundations or floors of six Russian period structures and miscellaneous feature pits, together with an artifact assemblage which convincingly demonstrates a settlement possibly occupied by Creoles (or Russians) together with Native persons. She suggests that this is the site of the Afognak artel. There is firm evidence, however, that the Afognak artel was located at the site later called Little Afognak. This evidence consists of a label “Company settlement” at the appropriate location on Lisianskii’s 1805 chart (published in “atlas” 1812), Davydov’s 1803 identification of the main Afognak artel as Igvetskaia, and a Spanish map showing discoveries up to 1792 that has a Russian post in Izhut Bay. The last location is a short distance east of Igwik or Igvet, but presumably was misplaced on the very crude outline of Afognak Island. Igwik is the locality of Little Afognak, so stated by Afognak elders. Woodhouse-Beyer obtained tree ring dates for foundation posts from the only house at the site that had a brick (Russian) stove. They indicate construction in the 1820s (Woodhouse Beyer 2001).

      An 1849 Ivan Kashevarov chart based on an 1839 hydrographic survey and the 1840 Murashev survey labeled this Kataaq settlement "selen. Afognakskoe." This is somewhat surprising, as one would expect such a label to appear at the large amalgamated Afognak odinochka, Alutiiq barrio and RAC retiree village, later called Afognak. Murashev portrays it the same year with seven structure marks arranged in two rows of three, with the seventh structure by itself. The 1849 Russian-American Company map has 8 rectangles at the "Aleut" village Kattagmiut. We do not know if this was the actual number of houses or buildings. The settlement apparently was not relocated to Afognak at the beginning of the 1840s after the smallpox epidemic, as the RAK map consistently leaves off the settlements that had been closed out or consolidated. Also, it was only Alutiiq settlements, not Russian and Creole habitations, that were consolidated, which leaves us puzzled by the designation “Aleut Village.”

      Teben’kov’s Atlas (1852:44), which uses data earlier than the 1852 publication date, states that because of the prolific salmon run at the head of Afognak Bay, two settlements were established here [Kattak and Rubets the Afognak odinochka]. He does not name the former, but writes that "…on the opposite side of which [Afognak Bay], on the outer cape, is another Russian settlement three miles NE by N of the settlement of Rubets." That accurately pinpoints Kataaq and indicates that people living there may have been RAC retirees or Creoles, hence the designation “Russian.” The 1849 RAC map shows here the village Kattagmiut (probably copied from Murashev, who shows the same from a survey done 10 years earlier). The “miut” suffix on the RAK map suggests a Native Alutiiq settlement. Holmberg refers to this settlement as "Eingeborenen No. 7" on an 1856 [1852—CHECK DATE] map. As the German description indicates, Holmberg considered it a Native village.  

     After 1849, there is a gap in the records of about 20 years. Then in 1871-1872, Alphonse Pinart collected masks here. Seventeen years later, a notation on a map dated March, 1888, signed by Davidson (copy seen by Clark at University of Alaska-Fairbanks) identified this settlement as Little Afognak. This undoubtedly was an error, even though it would have been appropriate. We assume that that notation was meant for Selezneva, located a considerable distance to the east. The 1891 census states that "Catanee" is included in the count for Afognak, as it was most other years, except in 1910, when the U.S. Census enumerated a Nekrasov family of five persons living there. Occupation by this family continued at least until 1935, as it is listed in a BIA census. The Russian and Tlingit Nekrasoff family came to Kodiak soon before, or in the beginning of the 1850s, according to Church records. It is tempting to speculate that they lived continuously at Kataaq, bridging the Russian and American periods. Woodhouse-Beyer states that one of the numerous Naumov families also had a “homestead” at Kataaq close to the time that it was depopulated. Church records and family tradition tell that the two families were linked by marriage.

     Woodhouse-Beyer does not report on the American period remains she recovered. Apparently, some American houses occupied the same sites or foundations as Russian period houses, and artifacts from the two occupations were not always [easy] to separate. She proposed that the location was vacated after the smallpox epidemic of 1838, but was reoccupied later. It is not clear, though, that it actually had been vacated. Since it appears on maps that postdate the epidemic but precede the American period, we believe that it continued to be occupied.

     No two structures were alike. One large house had plank floors, a brick stove flanked by slate flagstones, and yielded abundant glass beads and many fragments of glass window panes. It apparently was constructed without the use of any nails. The excavator proposed that the walls were formed by planks, but does not demonstrate how this conclusion was reached. This structure would have been occupied by Creoles, and by Russians, if any of the latter were living in the settlement. Unless it served also as a store or storehouse, the space could have accommodated two families. An undated saw pit is located near the houses. Other houses are slightly dug into the ground, but none is a typical Alutiiq barabara.

     The artifact assemblage is varied and includes some art carvings in bone and ivory as well as a large number of glass beads but, strangely, no nails, although nails are common at other Russian period sites. An appreciable number of small harpoon heads was found, some of them made so precisely in a historic Aleutian Island style that one can conclude that they came from the Aleutians or that their maker was an Aleut who came to Kodiak. We know that in 1784, or soon after, Shelikhov brought a number of Fox Island Aleuts to Kodiak, some of whom were sent to Afognak.

     In addition to window glass, mica window panes were common. We have noted that the settlement is not likely to have been the one referred to as the Afognak artel, in part because it was only three miles across the bay to the odinochka, which appears to have been especially substantial for an odinochka, and in many ways was like an artel. Nevertheless, the artifact assemblage is much like what might be expected from an artel frequently visited and staffed by Native people and Creoles. It has perplexing aspects, too: lack of fishing gear, scant presence of fox trap prongs, and lack of nails, for instance.

     The locality was not yet completely abandoned after 1935. In the 1940s, one Nekrasov family lived in a house and had a small barn, both built on the gravel bar between the stream and the bay. In 1951, there were still two small single room modified barabaras in the woods between the stream and the bay, huts where trappers or gillnetters stayed. The stream there carries a run of silver salmon that is not large, but appreciable considering the tiny size of the creek which has been dammed, but not completely blocked by beavers. The beaver are a relatively recent introduction.

Kazakovskii Zaliv

           This was the local Russian place name for a deep narrow arm of Marmot Bay on Afognak Island, called today Danger Bay. The name may refer to an artel’s manager Kazakov who served in the 1820s at unspecified outpost. Mys Kozakovskii, shown on the 1849 RAC map, is the name of a headland at the western entrance to the bay.

        A small Creole settlement existed there in 1860s-1870s. This settlement is not indicated on any map. The Alutiiq name for the settlement may be Mautaq, recorded by the Native Village of Afognak place names project in 2001. Kazakovskii Bay settlement appears in Church confessional records for 1867 only. The 16 members of the community were comprised of the Iakushev, Mikhail Seleznev, and widow N. Kostyleva families. All of these families had close kin at Little Afognak (Seleznevo). The settlement apparently existed for only a few years before and after 1867. It appears that the church records usually included its inhabitants with the nearby parent village of Seleznevo (Little Afognak). By 1871, Kostylev’s daughter had married and moved to Afognak.

      In 1872, Pinart traveled by baidarka around Afognak and Shuyak islands. As he moved on to Kazakof Bay from Little Afognak, he noted that on the right side of the bay was the small settlement of Kosakoffski, close to a lake surrounded by woods. In his account (ms., data courtesy Desson), Pinart states that settlers of Kosakoffski came to fish at Marka Bay.

      There is a local tradition that two persons from the hamlet disappeared under uncertain circumstances. One resident was suspected of responsibility for their deaths, and the remaining people moved away. The identity and fate of the alleged culprit is now unknown.

      There is a compound housepit site on elevated sandy beach ridges on the east side of Danger Bay, near the entrance, at a locality known locally (in 1998) as Big Sandy. It is near the site of the ephemeral Silver Logging Camp. Structure outlines are of both simple rectangular format and compound Alutiiq format. Sponge-stamped and painted ceramics have been found in two areas of the site, as well as traces of corroded iron artifacts (Steffian 1998 "Archaeological Survey on Central Afognak Island, Alaska"). The ceramics indicate occupation late in the Russian period and/or early in the American period.

      According to Woodhouse-Beyer’s interpretation, the Russian-period Katanee occupation should be dated within a narrow time frame from the 1820s to about 1841. However, it may have lasted much longer, to the end of the period. Occupation of Kazakovskii followed (but not by the same persons) at the beginning of the American period, and lasted for a few years only.


      The Aleut Town midden is located at the far inner end of Afognak Village. Aleut Town also included the locality of the former odinochka near the last Afognak school building, and at times may have extended southward almost as far as Graveyard Point. The area is thought to have been occupied more or less continuously from two millennia ago up to the tidal wave event of 1964. There were three midden deposits in this area, including the Aleut Town midden investigated in 1999 and 2000. Kachemak tradition artifacts can be sorted out stylistically, and usually are associated with deep occurrence in the midden, but it is difficult to sort out Koniag, Russian period, and American period evidence from among the 3,000 historic artifacts recovered. That is in part attributable to the lack of a recognizable break between these occupations. It is also attributable to the loss of much of the midden and overlying historic refuse and structures to coastal erosion and from disturbance or “landscaping” of the surface. Sometime after 1912, low areas or structure pits were filled in. It is likely that part of the mound deposit was shoved over the bank onto the shore.

      The area in front of the school building is indicated on one 160-year-old map as the site of the odinochka, but careful inspection of the ground after it had been cleared and tilled for a garden revealed only traces of very recent occupation. However, glass beads have been reported in garden plots, now long abandoned, at the south end of the Aleut Town area.

      Most European “historic” items from the Aleut Town midden are recent in age, often from above the 1912 “Katmai” ash, but a few artifacts can be proposed as likely remains of the Russian and early American periods. “Late pre-Katmai American” would encompass 17 years from 1795 (an approximation; approximately applies to inception of wire nails) to 1812, “Early American” covers 27 years from 1868 to 1895, and “Russian” refers to 66 years from the first published date of the odincochka to 1868, or 80 years from the date of Russian settlement on Kodiak.

      For Late pre-Katmai American occupation, one would expect a modern array including 30-30 shells (local people advised that 30.06 was not common). Early American and Russian diagnostics would be missing. Glass beads would be out of style. Wire nails are present, though there would be some cut nails from old structures and boats. Shoe (boot) parts were found to be common, also parts of cast iron stoves. Bottle neck attachment techniques (cessation of separate necks attached before the glass cooled and hardened) and can soldering and sealing (with the advent of machine crimping) underwent changes during this period, but some of the old technique persisted to beyond 1912. The time boundary thus is somewhat fuzzy.

      The following listing includes pre-1912 American, Early American period artifacts, and possibly a few Russian period items. They are lumped together because of the absence of stratigraphic separation: Some white ceramic sherds were found below Katmai (most came from above), a single porcelain sherd, clock parts, a brass samovar or lamp, and eyeless fish hook have been found. There is a rim-fire pistol cartridge; two old .44 cartridges, one with depressed H headstamp; 30-30 WRA, WCF, and USC WRA cartridge cases; FA 36, 45-70, and 45-90 cartridge cases; brass shotgun No. 10; gun butt plate of brass; forged iron lance; thimble; a hinge; a latch hasp; buttons; a comb; a watch part; a suspender strap buckle clip; a fox trap prong; a later fox trap spring; a powder can top; bricks; flat window glass; many cut nails; a small number of wire nails (most wire nails were above Katmai); a clay pipe fragment; and glass of various colors, but most glass came from above Katmai.

      Early American Period material was not common, yet it is moderately abundant, given the short duration of that period. Included are the two Henry rimfire cartridges, a pinfire cartridge, a brass butt plate; possibly the very small number of glass beads, the top of a one-pound powder flask, the spring torsion trap prong (could be Russian), and a few painted ceramic sherds like the ware traded by the Alaska Commercial Company. The butt plate and painted ceramics were found outside the excavation in a recent soil borrow disturbance. Many of the cut nails may belong to this period. It, too, is the best fit for clay pipes. The arms are easiest to date by initial appearance, although a conservative fondness and owner attachment resulted in the persistence of many models of firearms. One Henry cartridge with raised H would have been manufactured between 1866 and 1898, its 0.815 case length (exclusive of Head) indicating that it was produced relatively early within that time range, and a swollen head may show that it was fired from a Model 1866 improved Henry. The second .44 cartridge has a recessed H. It was manufactured between 1873 and the early decades of the 20th century, although its greater length of about 0.845 mm indicates a very late production date. The 45-70 cartridge could go back as early as 1873, but people continued to use this ammunition well into the 20th century. All-metal Winchester shotgun cartridge cases go back to about 1880 but were used for a long time thereafter. Evidence of use of muzzleloading guns, including shotguns, includes the powder can top, two musket balls, a stone ball mould, a lockplate for a flint, and a possible small size pliers-shaped bullet mould. Some of this muzzleloader material could be derived from the Russian period, but may have remained in use into the 20th century. The pinfire became a specialty type of arm that continued in use into the 20th century.

      Early trade goods often are seen in limited amounts in Native sites. This may be the case at Aleut Town, but there is little trace of Koniag tradition or late prehistoric Alutiiq occupation there. Possibly this occupation was better represented in parts of the site that have been lost to erosion or removed through later construction and “landscaping.” Otherwise, glass trade beads should have been abundant. They were not, but about 75 years ago glass beads were found in gardens elsewhere in the area near the south end of Aleut Town (Nadia Mullen pc August 2000).

      Early Alutiiq residents at Aleut Town likely also would have acquired ceramics for tea, metal pots, and metal tools. But this cannot be documented at Afognak. It would be expected to find brass scrap from recycled and cut-up burned out pots, and a small number was found below the Katmai layer.

      There appears to have been no Russian Period living floor in the area tested. This is perplexing, as Russian trade goods were available at the nearby odinochka from some time before 1802, and in the years continuing thereafter. For 1795 and 1804, Russian records name the Alutiiq village that Afognak elders identify as Aleut Town, thus we are confident that the site actually was occupied in the Russian Period. Baranov’s census listed Naschkuchilik, a village of 105 persons in 1795 and 99 persons in 1804. The census is for Native villages and thus does not list the Russian establishments. Naschkuchilik (variously spelled) continued into the American period when Army officer Eli Huggins (1881) visited Afognak, and later, about 1891, when the village is described in the 11th U.S. Census. Three miles across the bay at Quataat, there was abundant evidence of the trader’s presence, as described above.

      A critical problem is to link the former inhabitants of the Koniag sites at Afognak Bay to the people first encountered by the Russians, and to the historic Alutiiq inhabitants of Afognak Village, a topic that returns us to the beginning of this chapter. Thus far, archaeological investigation has failed to identify any “contact” sites of the late 1700s on the basis of the types of artifacts recovered. However, inferential evidence strongly suggests that Selezneva (Mali or Little Afognak), Malinovskie Litnik or Nunalik (Malina Creek), and a site on Afognak Bay were occupied at the time of Russian contact.

      There are Russian period traditional Alutiiq “octopus” houses on the Afognak River. Very likely, there was continuity across the precontact/historic temporal boundary here. These are summer fishing villages and we can expect to find main villages in the vicinity, closer to the outer coast, as at Aleut Town.

      At the beginning of the 1800s, more records of settlements become available. One at Afognak Bay was part of the creole village later called “Afognak.” The other, also a Creole village, later called Selezneva or “Little (Mali) Afognak,” was at Duck Bay. There also was an Alutiiq village and small artel at Malina Creek. When Lt. Davydov visited Afognak in 1803, there was a small Russian station there of the type named Rubtsov odinochka. AFG-004, the Aleut Town site or Naschkuchlik, is a strong candidate for early historic or Russian period contact, and even a point of initial contact, as it appears to have been located within a few hundred yards of the odinochka. We noted, however that the Russian period portion of Naschkuchlik has largely disappeared. There has been serious consideration given to the possibility that the Igvet or Igvak artel, which has been glossed as “a step across” (from the odinochka?), was the Creole-Native settlement at Qat’at or Katanee, uncovered in Katharine Woodhouse-Beyer’s excavations. But evidence that the artel actually was at Selezneva (Mali Afognak) is found in Pinart’s 1872 unpublished observation, conveyed to the writer by Dominique Desson, that “whereas Afognak Village had had only an odinochka, Selezneva had been a (more substantial) artel, together with a large Native population. Too, the Igwik locality was the first place where Russians settled on Afognak Island.” Pinart had visited the Seleznevs at Little Afognak, sons of one of the earlier inhabitants.


      The salmon counting weir is located almost exactly on the site of an 1870s-1880s fishing barricade or “zapor,” published in 1902 in Moser’s Fish Commission report. A short distance below the weir, there are several large Alutiiq-Koniag style multi-roomed “octopus” houses. Together with appended rooms, these houses measure 25 to 35 feet across. A test pit excavated in a side-room of one house by the Dig Afognak program in 1997 recovered a small number of Russian glass beads. This site thus was part of the historic Alutiiq fishing encampment, “Litnik,” where there were forty barabaras late in the 19th century. A shovel probe made in front of one house in 1971 found shells from clams brought up the river from the bay at least two miles away.

      On the opposite (northwestern) side of the river, there are additional deep house pits, not of typical format, i.e., they lack multiple appended rooms. Two test pits placed outside the houses did not encounter any artifacts or refuse deposits, but one test found several feet of mottled brown soil, possibly a "melted" sod wall. It remains very likely that these depressions are from later Alutiiq houses, perhaps those of the people who operated the fish “zapor.”


      A cluster or row of houses, possibly a community hall (called kashim by the Russians), and racks or stages for storage and equipment formed a village. There also were fish drying racks at summer settlements. Koniag villages sometimes were 200 yards long. Many were larger than Early Kachemak and Ocean Bay villages, although data for earlier times is imprecise because most early village sites have been destroyed or reduced by erosion. Settlement was based primarily on large main villages, occupied during late autumn, winter, and early spring, and on summer salmon fishing villages. Houses built at both types of villages were similar in format, at least in their floor plans, that survive as surface depressions. There also were minor settlements, which for lack of details can be called hunting camps. Most settlements were laid out in a single row of houses along the shore. But some villages on Kodiak were so large that not all houses had frontage. Refuse, mostly shells and bones, and also sweat bath (steam bath) rock was disposed of at the site. At minor sites, there are such “middens” in front of and to some extent alongside the houses. As occupation continued, trash also was dumped in pits from abandoned houses, and construction of new dwellings often shifted to old dumps, sanitized through the passage of time. The shell matrix provided good drainage and easy digging for house pits. In this way, thick midden sheets and mounds accumulated at the settlements with the longest occupation.

      Refuge islets and natural forts are especially interesting as evidence for increased population and warfare. The rocky headlands and cliff-bound islets of Kodiak provided many sites for such natural refuges, a settlement type initially identified in historical accounts. Houses on the refuge rocks were similar in plan and in interior layout to those of main villages, judging from Richard Knecht's excavation of Shelikhov's battle site at Sitkalidak Island. Most appear to be late-prehistoric Koniag, which accords well with hypotheses of the recent rise of warfare in the North Pacific region. However, very few refuge rocks have been investigated, most of them superficially. None has been found in Afognak Bay, but, farther out on Marmot Bay, a small islet at the entrance to Danger Bay and a somewhat larger island off the site of Little Afognak would bear examination. Thus far, there has not been adequate examination of potential “forts.”

      In addition to the refuge islets which can be "defended" passively because they are almost inaccessible, there are defensive sites. These have access by way of steep slopes or along a limited approach such as the neck of a protruding headland, and hence would be relatively easy to defend. Again, no Afognak sites of this type have been recognized. Rather than concluding that peace reigned in the area, it appears that more examination of appropriate locations is needed..

      There are about 29 settlement and camp sites around Afognak Bay from Danger Bay westward (Table 2). Some are minor or now have been completely washed away, as some of the observations date back to 1951. Additional sites must have existed earlier, but they now are totally lost from the record. Finally, additional small sites likely remain to be discovered, especially on Whale Island, which has not been examined.

      We have divided the bay into inner and outer zones. The inner zone is, essentially the mouth of the Afognak River, but includes also a shallow water site at the head of nearby Back Bay. People would have moved to the inner settlements in the late spring and summer to catch and process salmon. They could have remained there until December to catch silver salmon if other priorities did not draw them elsewhere. Between the peaks of the runs, they would have had time to make equipment, hunt out on the bay, and to do other things. During Koniag times, a lot of effort must have been put into making wooden equipment, as adze bits are one of the most common artifacts found along the eroded shores at Litnik on the lagoon of the Afognak River. The inventory from the beach at the two Koniag sites there consists of 28.7 percent adzes (of which 8.9 % are splitting adzes). This exceeds the frequency of ulu blades (23.8%) which were one of the most important implements at a fishing camp. But farther down stream (below the old road crossing) in the Kachemak area of Litnik (exclusive of the Ocean Bay II-Kachemak site AFO-088), the frequency of adzes (none of them splitting adzes) drops to 1.6 percent, though ulu blades also drop to a modest 10.5 percent.












2 middens seem in 1964



30 yards L

Historic &


2 ft thick midden plus 20 yd 


+ 20 yards


thin band at base of Katmai 



82 yards L.

Alutiiq, Kon-

Front greatly

5 ft thick.  Tested in 1999

Aleut Town

25 yards W

iag, Kachemak 

 cut back




Was 70 yd L.



Faces petroglyphs,  much FCR






1 house atop bank, slumped



140 yards L

Ocean Bay I


Tested, description, collections 

Chert site



Ocean Bay II


and dates are published




Hist. Alutiiq

Edges eroded

House pits and fire-cracked




& Koniag






Kachemak &

Edges badly

Variable determiners but no









43  yards L.

Ocean Bay II

Edge badly

Tested, description, collections,

Slate site




and 14-C dates are published



40 yards L. 



Much L & W    Tested in 1951 and 1995



+ much lost


thick deposits, once important





Some slump

Annex of AFG-012 across rill



36 yards L.



2 house pits






Excavated by Dig Afognak, 

Settlement Pt.


7 houses?



reports and dates by Saltonstall



240 yards


Badly cut 

 Sandy deposits, house pits, fire-






cracked rock



90 yards L.    



Once 2 house pits, much FCR



105 yds L.


Edges badly

In segments, creek runs through

Marka Bay

up to 35 W






90 yards L.


Badly cut 

 Test pits, collections and  14-C






dates are published, no midden

AFG-211, 212



Historic & pre-


Koniag-style house pits, next to




contact Alutiiq


1800s weir




Historic & pre-

Edges eroded

House pits and fire-cracked 




contact Alutiiq


rock deposit

AFG-214, 215




Some edges

Variable determiners, some 




other early eroded

house pit, no middens




Ocean Bay I


Stone artifacts present in thin

Estuary Islet




layer, little of site remains






Low elevation, flooded?



   51 yards



2 segments, multi-room house



   4 x 9 yards



14 in. thick midden



   13 x 9 yards


  Edge eroded

Thin, early Kachemak?  



    A few yards


  Edge eroded 

 Shell midden, eroding graves 



recorded in 1984

FCR = Fire-Cracked Rock


Cobble spalls seem also to have been used to process fish, to cut through tough fish skin. The combined cobble spall (boulder flake) and ulu blade frequency is nearly equal between the two occupations, at 26.8 percent for the Koniag area and 23.4 percent for the Kachemak area. It seems, then, that on a per-capita basis processing salmon was of about equal importance during each period, although more people may have been at the fishery in the later period.

      When winter came, the river and tidal estuary froze hard, and some years snow accumulated deeply. The inner area was not a place to live during winter and early spring, so people spent most of the year at the outer sites. There are about thirteen inner sites, more or less, depending upon how one subdivides the line-up of camps around the mouth of the Afognak River, but two are historic. There are eighteen known outer sites though the original number must have been greater. Most outer sites belong to the late Koniag tradition, while the inner sites represent the entire span of prehistory. This may be laid to the fact that erosion has been more severe in the outer zone, and the early sites once present there had been subject to attrition for a much longer period than the Koniag tradition sites. The latter represent no more than 600 years’ exposure, the earlier sites more than 6000 years’.

      The Koniags occupied twelve sites, exclusive of three inner sites which are thought to be seasonal aspects of outer settlements. The twelve range in size from a single house to nine houses, at Settlement Point. The 200-yard long site facing the Collector Beach now has seven or more houses, somewhat confused from secondary use for garden plots, but prior to extensive erosion there could have been many more houses there. The Aleut Town site at Afognak Village may also have been a substantial village, but house pits no longer are clear on the surface because of late historic occupation. Among the ten Koniag tradition sites of the outer area, some villages probably had been abandoned by the time others were occupied. Nevertheless, it appears that more than one main settlement and also some of the small sites, were inhabited concurrently. The site “Adze” has radiocarbon dates that place it within the time range of the nearby Settlement Point village. This poses questions as to how these villages were allied or cooperated to share game, fisheries, bird rookeries, and driftwood in the relatively localized area of Afognak Bay, and whether they ever came together under a single leader for ceremonies, trading trips, and war and defense.


      A broad spectrum of food resources was harvested, but to date, few data have been published other than for mammals at Late Kachemak and Koniag sites. An analysis of the fauna, e.g. food refuse, from Settlement Point and Afg-012 by Megan Partlow appeared in 2000, and there are identifications from the first (1964) excavations at Crag Point, at the entrance to Anton Larsen Bay. Faunal material from the Aleut Town dig has not been processed. Faunal analyses show that uniformity in subsistence through time was not the case. While one can assume that there was a tendency to utilize all available resources, availability may have varied from time to time and place to place.

      Sea otters may have been utilized as food mainly during the Ocean Bay tradition. Fox bones are abundant in Kachemak tradition sites, where this animal evidently was consumed for food. They are variably present later in Koniag tradition refuse. Dogs are moderately common in both Kachemak and Koniag deposits, and there also is some evidence that they were eaten (cut marks on bones, opened brain cases), though the abundant dog remains recovered from the Aleut Town midden at Afognak do not bear these indications. During much of prehistory, the most commonly hunted sea mammals were porpoises (notably harbor porpoise), except at Aleut Town where porpoises are rare, and especially harbor seals. A small number of fur seals also were taken. Northern sea lions were hunted but, like whales, it is difficult to assess their importance because the meat and other usable parts may have been stripped from the bones at the kill site. The same applies to brown bears, which were harvested to a limited extent, a limiting factor being the size of the bear population of no more than a few thousand animals recently for the entire archipelago, and 100 years ago far fewer than that. Some caribou meat was imported, especially by Koniag villages located on Shelikof Strait across from the Alaska Peninsula.

      Whale bones are found in sites of all periods and the Koniags are known to have been avid whalers. A whale bone recovered from an earlier context at Crag Point in 1986, in deposits that date to about 2,000 years ago, reported by Kopperel, has the imbedded fragment of a slate point [check for L. Yarborough published article reference identifying the bone through DNA]. This animal is accorded great importance in ethnographic accounts and the contribution of whale oil and meat to the diet reportedly was significant. Whale bones are common and were modified to make implements from the earliest time onward. At contact and in early historic times, whales were struck with long slate-tipped darts smeared with aconite poison. It is believed, according to Bisset, that the poison destabilized the whale by crippling the flipper muscle, so the whale would drown. It rose in a few days and started drifting. The long slate tips used ethnographically on whaling darts have not been recovered from precontact deposits, but Koniag whaling ritual and ceremonialism are sufficiently imbedded in lore, art, and customs that whaling appears to have been indigenous, not a Russian introduction.

      The surround method of hunting sea otters involves several hunters in kayaks surrounding and harassing individual otters, which seldom escape from being struck by a harpoon-dart or harpoon arrow. The case for this technique being indigenous requires further examination, but that cannot be done here.

      A number of wild plants and berries were collected for food and medicines, but the quantitative measure of their contribution to nutrition before commercial foodstuffs became available has not been established. The kamchatka or “chocolate” lily (Fritallaria, sarana in Russian), which has a bulb composed of rice-like granules, was harvested and stored under Russian direction. Earlier, Alutiiqs very likely did the same, but they no longer do so although the bulb is easily harvested and is very palatable. Wild celery (cow parsnip Heracleum and Angelica) also was eaten, and sometimes confused with deadly water hemlock. Sour dock (Rumex) also bears mentioning. Two forms of cranberries, high bush and lingenberry, are abundant and store well. The salmonberry, for which Kodiak is renowned, keeps poorly and, like most berries, is subject to poor crop years. It was eaten mainly in season. Red elderberries were harvested and eaten along the coast of British Columbia. They are available in substantial quantities farther north, but we have not seen records of their use on Afognak (except by bears).

      Historic accounts, together with the range of archaeological fishing implements and layers of fish bones in ancient archaeological sites, show that fishing in the sea was a major occupation. Detailed analyses of the remains are available only for Settlement Point, reported by Partlow. Mainly salmon and cod, far less often sculpin, Irish lords and right eye flounders, were eaten at Settlement Point. This also is the case at nearby Afg-012, although here cod are more than twice as abundant as salmon, even though Afg-012 is closer to the two salmon streams at the head of Afognak Bay. Halibut bones are surprisingly rare at these sites, although today the fish are plentiful just a short distance offshore. Large halibut which may weigh several hundred pounds likely would have been butchered on the shore and the bones left there if the early Alutiit successfully landed any such beasts, but there is little reason to doubt that many smaller halibut would have been caught. At some sites, at Larsen Bay, for instance (reported by Yesner), salmon also is tied with Pacific cod, rockfish trailing behind. Fishing pressure on Pacific cod was so great that, over time, fish size decreased, according to research by Robert Kopperl.

      Hypothetically, with adequate storage and preservation, the salmon fishery could have fed an immense population on the Kodiak Archipelago. In 1995, a peak year, the catch of pink salmon in the Kodiak district was more than 30 million fish, to which several million salmon of other species can be added. But there are several reasons why it was neither practical nor possible for salmon alone to support a very large population. These include the occurrence of alternating good and poor years for salmon, long cycles of scarcity and abundance respectively, and bad weather years for drying and preserving fish. As an alternative to drying, pits in the ground were filled with salmon in many parts of Alaska to produce “stink fish.” This practice has been described for Karluk late in the nineteenth century. There is also the danger of overfishing small streams and killing off the runs if harvesting was too intense. That would leave people dependent on access to a small number of larger salmon streams. Moreover, it was critically important to balance the diet with foods other than fish, and to provide proper nutrition throughout the year. Inasmuch as the land fauna and plant foods that would do this were inadequate on Kodiak, the need remained to round out the diet with sea mammals, especially their fats, to meet the high energy demands of North Pacific sea hunters. Heavy fish eaters risked being perpetually hungry.

      Fishing was done with hook, or gorge, and line in the sea. At salmon streams, spears, nets, and weirs—in conjunction with spears or gaffs and probably traps— were used. Direct evidence for the use of nets consists primarily of numerous notched pebble sinkers found at some Kachemak sites, at the mouth of the Afognak River especially, and of net floats at Koniag wet sites. (There also are mesh gauges, but they could have been used to make bird nets as well as for fishing nets.) Heavier notched and grooved cobbles were for weighting lines to fish in the sea. It also was possible to use unmodified cobbles for this purpose.

      Birds and fowl of almost every kind were utilized, primarily for food, but also for skins for clothing (parkas), and for feathers and bills for ornamentation. A list of 49 species has been drawn up from the bird bones that Hrdlicka collected at the Uyak Site. The list for the smaller-scale excavation at Settlement Point, given by Partlow in her dissertation, is shorter: 11 genera of one to several species each. They are:

      Loons – uncommon

      Grebes – uncommon

      Albatross – common

      Shearwaters - common

      Cormorants – very abundant

      Geese & ducks – very abundant

      Ptarmigan – abundant

      Eagles – uncommon

      Gulls – common

      Murres and puffins – abundant

      Ravens and crows – uncommon

      Cormorants also were most common at Afg-012.

      Large amounts of shellfish were eaten, judging from the volume of shell dumps. Shellfish provided dietary variety, important minerals, and trace elements. They reportedly served as a stopgap when fish and other meat were unavailable, but they are very low in calories. Nearly everything found in the littoral zone, between high and low tide, was collected and eaten. Shell reflects a subsistence diet more directly than other types of faunal refuse because virtually nothing is stripped away and left at the "kill site" on the tidal strand. Shell refuse also reflects immediacy of use. Because of the danger of spoilage, consumption dared not be delayed long, and usually was local. However, whole clams were taken inland to fishing camps, on Karluk Lake for instance. The main components of shell middens are blue mussels; clams, especially butter clams (Saxidomus), cockles (Clinocardium), a few "horse" clams (Spisula), and littleneck clams (Prototheca), May;. Tellina, sea urchins; the tiny periwinkle in immense numbers and smaller amounts of "dog winkles" (Thais;, large whelks (sea snails Neptunea); chitons; barnacles; and limpets. Razor clams are found at only a few beaches and are rare in the middens. Octopus were caught but their remains do not survive. Proportions vary from site to site. Crab, of which there are several edible species in shallow and offshore waters, is absent in the middens, though the red or Dungeness crab would have been easy to harvest. It is possible that in spite of all the un-nice things Europeans and Americans said about Alutiiq food, people were squeamish about eating an animal that feasted on the corpses of drowned persons—at least this modern explanation is cited today. Crab also was little used in the 19th century by Russians and Creoles, even though people were aware of their existence. Massive harvests of blue mussels, clams, periwinkles, and green sea urchins often resulted in the formation of relatively pure blue, white, and green layers in the shell middens. Usually, though, there is a mix of assorted shellfish, fish bones, mammal bones, soil, and burned rock and the occasional artifact. The waters of Kodiak are highly susceptible to paralytic seafood poisoning or "red tide," which can be fatal when infested clams and mussels are eaten. Little is known about the past incidence of PSP. Colonial records, including countries outside Alaska, mention little about it, but there was a serious case north of Sitka a little more than 200 years ago, in which several persons died.

      For further details, we can note the mammal remains collected from the Late Kachemak occupation at Crag Point site in 1964, and from the Koniag settlement at Settlement Point during the 1990s, reported by Clark and Partlow, respectively, and from the 1986 excavation at Crag Point, reported by Robert Kopperl (Table 2). Crag Point is located on Kodiak across Marmot Bay within sight of Afognak.

  Crag Point (Late Kachemak) Settlement Point (Koniag)
  1964 1986  
harbor seal      45.10% 37.20% 76.00%
fur seal       5.40% 0.50% 0.00%
sea lion           1.10% 2.80% 0.70%
sea otter          0.30% 0.50% 1 bone
river otter       3.00% 0.00% 1.00%
red fox             29.70% 14.30%    ---
domestic dog 2.40%  16.7% "canid"  
dog plus fox      32.50% 31.00% 4.50%
Kodiak bear    3.00% 0.00% 0.00%
porpoise 9.50%  23.3%* 10.30%
caribou?         0.50%    0.15% cervid 0.00%
whale                              present 2.20% 7.00%
ermine 0.00%    0.15% (1 bone) 0.00%
* 2 species: Harbor porpoise  and Dall's porpoise.
Preliminary inspection of the Aleut Town collection indicates that dog is common there but that porpoise was scarce.

      All of the uncommon bear and sea otter bones at Crag Point (and also at Three Saints, another Kachemak site) are modified. They have been used for games (sea otter) and evidently for material for tools (bear). A small dog is represented among the Crag Point remains. Wm. Haag reported two sizes of dogs at the Uyak site, and Afognak Islanders probably also kept these two types of dogs. Dog skulls found in 1999 at Aleut Town are not small, nor are they as large as Eskimo malamutes.


      Petroglyphs are designs carved into rock outcrops and on boulders. Usually they are made by pecking with a hammerstone, though sometimes glyphs are chiseled or sawn. They are one of the most common types of “rock art,” another common type being rock painting. Rock paintings are found in outer Cook Inlet and Prince William Sound but are not confirmed for Afognak.


      The largest petroglyph site in the Kodiak region is at Cape Alitak, where hundreds of figures are scattered over acres of smooth granitic rock surfaces. Several additional petroglyph localities have been found on Kodiak and Afognak. They usually consist of small clusters of figures and even isolated carvings. The sites are inconspicuous and easily overlooked, so additional localities may exist. The following ones are on Afognak:

      Discoverer Bay (Perenosa Bay) lagoon
      Peril Cape (single large figure, reported, not recorded)
      Marka Bay
      Afognak River
      Lipsett Point
      Inner end of Afognak Village, also...
      Port Lions (solitary figure on boulder. According to hearsaythere
      are two other petroglyph localities but they have not been described).

Description of Afognak Sites

      Marka Bay. The figures at Marka Bay are located on seaward facing surfaces of a dike or seam igneous rock that, when it was molten, was forced into a crack in rocks of the slate-greywacke group. The dike is exposed on a boulder strewn shore that, even before 1964, was submerged at high tide. It rises about two yards above beach level for about 20 yards, but petroglyphs are found along only part of it. They are very faint and only eight figures could be followed with clarity when the locality was recorded by Clark in 1964. Judging from additional unclear traces, at least 30 figures may have been present. The figures range in size from about 4 to 12 inches. They represent faces, whales, a dancing or drumming figure, and a geometric sign.

      In 1995, the locality was reexamined by Dig Afognak staff. After subsidence in 1964, the figures become densely covered by blue mussels and barnacles. A few were scraped clear, reexposed and photographed. Patrick Saltonstall then revisited the site in 1996, and found that figures exposed the preceding year again were obscured by barnacles. One can surmise from the condition of this site that there may be additional totally obscured petroglyphs within the intertidal zone around Afognak.

      Afognak River. Two petroglyph panels were found in 1971 on an inclined slate outcrop facing upriver along the tidal reach of the Afognak River. The panels are separated by a fissure little more than a foot wide and may belong together as one unit. The larger panel is three feet wide and seven feet long. The petroglyphs consist of small weathered pecked cup-shaped depressions and sawn grooves. Many of the grooves appear to be random but they occur in swarms with a common orientation. There also are many four-, six- and eight-rayed stars formed by the intersection of short lines. One starred hexagon and a few rectangles complete the group. The pits or cups are not necessarily of the same age as the sawn lines. Other slate outcrops and boulders at this locality would have been suitable for petroglyphs, so it can be assumed that the person who made the cups was aware of the sawn glyps and intentionally combined the two. The cups show much greater weathering than the sawn lines, but they were made by percussion or fracturing and pulverizing the stone instead of sawing.

      Boulders in the lagoon at the salmon stream at Discoverer Bay (in Perenosa Bay) also have cup-shaped depressions, and a boulder at the shore end of a stone fish weir there has sawn cuts one to five inches long.

      Lipsett Point. A dike of igneous rock, similar to the one found at Marka Bay, but less prominent, outcrops on the beach immediately inside Lipsett Point, well out from the upper edge of the shore. However, the shore has eroded and has shifted inland many yards since subsidence in 1964, with the consequent loss of a small Koniag site and historic garden plots. Part of a small badly-eroded panel of petroglyphs has survived on an outward-facing surface. The figures appear to be largely above tide water, but are in the splash zone, as they host a few barnacles. Part of this surface had fallen away in the past. Presently, three or four figures can be counted. The clearest ones are a face and a small dancing or drumming figure (an object held high by one arm could be either a drum or a ceremonial object). They were recorded by the Dig Afognak project in 1998. The stone is highly eroded and the glyphs had to be felt by hand, though they are recognizable in the latex mold that we made. Outcrops of similar rock found along the western shore of Afognak Bay were examined but no additional petroglyphs were found.

      Afognak Village. Just beyond the northeast end of Afognak village is a small panel of faces on a vertical seaward-facing surface of a greywacke ledge. Adjacent surfaces, equally suited for rock art, appear to have been ignored by the petroglyph mason. This published site is well-known, and the figures have been used on Afognak Native Corporation letterhead. The faces are wearing labrets. The site remains generally above the level of high tide, although during one visit a few periwinkles were seen lodged on it.

Analysis, Comparisons, Distribution

      Three types of petroglyphs are found on Afognak Island: (a) sawn lines and sawn geometric figures, (b) small cup-shaped depressions, and (c) pecked figures portraying mainly faces, full figures, and whales or other sea mammals. All, to some degree, are compositions, or parts of compositions, rather than random occurrences. The limited occurrence at each site of a cluster of a few figures seems to have been done on purpose, and further suggests that each locality or panel was conceptually unified.

      Kodiak and Afognak Island are at the northwestern geographic limit of a rock art distribution that extends along the coast of Alaska from British Columbia and other areas of North America. Afognak and Kodiak are of especial interest inasmuch as, with the exception of adjacent Cook Inlet and PrinceWilliam Sound, where red-colored rock paintings occur, and other rare exceptions, these islands host the only examples of rock art found in the region inhabited by peoples of the Eskimo family. There appears to be a degree of relatedness between the Kodiak petroglyphs and those of the coast of southeast Alaska and British Columbia, especially as seen in the treatment of faces. Afognak petroglyph making, and the reasons for it, may have been stimulated by contact with peoples located farther east around the Gulf of Alaska. But similarities between the petroglyphs of interior Siberia and the Kodiak archipelago also can be noted, although there is an immense region devoid of petroglyphs between Afognak and Siberia. Resemblances may be coincidental because similar things, such as human faces, are portrayed. The same line of reasoning also weakens the proposed relationship between Afognak and southeastern Alaska petroglyphs. The petroglyphs of the Kodiak Islands thus remain mysterious and isolated.

      Cup-shaped depressions, often confined to boulders, have a very widespread distribution in western North America. Often they are found at streams that carry runs of salmon. For the most part the diverse people who made these “cups” or pitted stones had no direct contacts between one another. Yet, we may wonder if they were not produced in response to some unidentified widely diffused pan-tribal belief.


      The motives of the anonymous ancient stone carvers are obscured by time and may be in the realm of mystery (in the religious sense). But the circumstantial evidence of site location, directional orientation of the figures, and topics portrayed provide clues.

      With the exception of the sawn glyphs and cup marks found at the mouth of Afognak River, all glyphs face towards the sea or towards bay entrances, not inland. They are situated close to tide level, one site being submerged at high tide even prior to subsidence in 1964. Suitable smooth rock surfaces are used.

      The geometric and simple linear shapes of the sawn glyphs differ from others, in part due to the limitation of using the sawing technique. The location of cup and sawn petroglyphs at the mouths of salmon streams suggests that they pertained to beliefs for welcoming, controlling or propagating salmon but this is not apparent from the petroglyphs.

      Other glyphs evidently were made to be seen from the sea, either by human travelers, sea animals, or supernatural beings. Possibly they simply were markers that advertised ownership of fishing and hunting rights in up-bay areas, especially at the salmon streams. But alerting people from other areas to this fact could have been accomplished by more visible means. Instead, the messages of the figures may have been directed to beings in of the sea. Along with this possible function, their obscurity may have reinforced the establishment of an in-group, of people who shared the esoteric knowledge of where to find the glyphs, traveled and saw and touched them, and returned home empowered, like members of a secret fraternity of shamans or wise men.

      The mix of motifs consisting of symbols, of faces (which in many cases show human attributes), the presence of dancing persons or drummers (common to Marka Bay, Lipsett Point, and Cape Alitak), and also the presence of whales, shows that the story or message communicated through these sets of carvings was complex and detailed. Whale figures most certainly pertained to whales or whaling, and dancers and drummers may signify ceremonies and welcoming, but the meanings of symbols are more obtuse. Desson in her dissertation, for instance, points out that a circle painted on the prow of a whale hunter’s kayak was in fact a “trap” that worked like the trap formed by dragging a pouch of fat extracted from a human corpse across the mouth of a bay within which a whale had been struck.


      The durability of rock considered in the light of the obviously weathered and eroded condition of many of the petroglyphs suggests a considerable age for the figures. However, great antiquity may not actually be the case. A small painted box panel from the Karluk wet site has a face similar to the Afognak Village faces, with lines extending downward from the eyes. The Karluk specimen dates to the centuries just before Russian contact. Such eye motifs also are found on historic Alutiiq masks and on small squatting-man figurines for hunting hats of the second millennium AD. Labrets are less telling, as they have been worn in the region for more 2,000 years, and possibly more than 3,000 years.

      The Marka Bay site may have been submerged at high tide even before the land sank in 1964; at least the figures were in the splash zone long enough for a few barnacles to grow on them. Thus, that locality likely was subject to relatively rapid erosion, which is not favorable for petroglyph survival. Following a major earthquake about 1,150 A.D., all the Afognak petroglyphs could have been submerged, at least at high tide. Later, the land rebounded, but during the interval of subsidence it is possible that low level petroglyphs were severely eroded. Accordingly, those seen now may postdate 1,150 A.D., according to Gary Carver.


      In the beginning, Afognak was settled by people who found there “a place in the sun” [Hrdlicka’s expression] following an earlier existence on the Alaskan mainland and along the Bering Sea. The island was a platform in the sea—a strand along which these hypothetical first settlers placed their villages, from which they gathered littoral products and driftwood, and from which they fished and hunted at sea.

      As a consequence of this mode of living at the edge of the sea, the unstoppable process of coastal erosion has washed away most of the archaeological record. The short-term effect of tidal waves in 1964 was to tear out chunks of some sites and to redeposit sediments from shallow bays atop sites, especially at the heads of inlets. But the effect of the land dropping six feet was much more drastic. At high tide waves ate away at sites, opening them to erosion and cutting back their margins. Some sites were reduced by a yard or two, others were eroded more extensively and some were completely obliterated.

      Next, with the beaches and littoral zone being lower, larger waves ran up onto the shore, shoving sand, gravel and cobbles to the upper beach limit, sometimes overriding the old edge of the shore and creating new storm berms. During the years after subsidence, the land has gradually rebounded, thus far regaining about half the elevation lost in 1964. The shores are being restored to a seemingly pristine condition and the only clues to what had transpired are, temporarily, the dead trees that rise from the tidal zone and from ponds near sea level, where drainage was blocked or dammed because of subsidence, and the new storm berms. In protected areas, there now has is an apron or ramp of rebounded terrain—a slight terrace, overgrown with rye grass, between the wave-cut banks of a few years ago and present-day gravel beaches. There also is the permanent record of deposits from tidal waves that can be “read” by the prying eyes of geologists. But there is one difference: the remains of once mighty village sites and midden mounds and rows of old house pits are gone, except sometimes for their inner fringes. A newcomer sees the land as if they never had existed, and he may conclude that they never existed. This is what happened to many sites after 1964. Similar events probably occurred many times in the past, leaving, after rebound, reconstituted shores and beaches, minus older deposits and sites, and we do not see the land as it was once. Geologists have read clues, though, telling that the last major pre-1964 event of this kind was about 1,150 A.D. As part of the past experiences of Afognak Islanders, earthquakes and tidal waves also posed a threat to living communities. But it is not known that they had any disastrous effects on the inhabitants. These events probably damaged and destroyed their houses and made it necessary to relocate some villages.

      Older persons from Afognak recall that areas along the shore, now reefs, once were grassy areas or hay pastures. The priest A. Prisadskii noted (in Nov. 1938) that according to Afognak residents the sea had taken about 200 feet of ground in front of the village during the preceding two decades, separating it from its tidal sand flat. He observed that during storms the sea would throw driftwood into the churchyard (The Russian American Orthodox Messenger, 1939, Vol. 35, No. 7, p. 110). It unlikely that many older traces of the settlement of 200 years ago have been washed away. The extent of damage is difficult to determine because of the lack of surveys and clear photographs of the coastline.

Lab No.
Raw Date++
(Calibrated Age)
Applies to
Tsunami Site      
Beta 165141  880+40 BP (AD 1030-1250)   Reoccupation above tsunami deposit
Beta 165139  1320+80 BP (AD 650-780 1 sigma)   Just before tsunami event
Beta 165140  1750+60 BP (AD 230-380 1 sigma)   Orange clay house floor, start of occupation
Salmon Bend Site      
Beta-170060   1400+80 BP (AD 530-780; 1420-1170 BP)   Annex room, charcoal under slabs
Beta-170061   1330+60 BP (AD 620-790; 1320-1160 BP)   Charcoal high in fill of main room
The “Chert” Site (AFO-106) x    
Gak-3801 5750+240 3800 BC (4470 BC) Very base of site
Gak-3802 4150+200 2200 BC (2630-2670 BC) Top of middle stratum
The “Slate” Site (AFO-109)tx      
S1418t4480+160 BP 2530 BC (3150 BC) tSite base
S1419t4475+125 BP 2525 BC (3110-3145 BC) Near site base
Gak-3804 4200+140 BP 2250 BC (2700-2820 BC) Near site base
Gak-3803 3890+110 BP 1940 BC (2190 BVC) tUpper levels
Beta 88720 2780+110 BPt  (785 BC-1250 BC) TP 3 Structure floor?
Beta 77807 3530+80 BPt (1660 BC-2035 BC) TP 1 53 cm below Katmai
Beta 88719 3490+90 BPt  (1540-2025 BC) TP 2 Base of Kachemak?
      45 cm below Katmai
Beta 195048 4540+70 BP 4020-3500 (5450-4970 BP) Charcoal high in butter clay
“Adze” Site AFG-012t y    
Beta 101916 450+60 BP (AD 1430-1480)t Koniag house
Beta 101915 420+60 BP (AD 1440-1620)t Koniag
Beta 101914 310+40 BP (AD 1515-1651)t Koniag house hearth
Beta 101917 280+60BP (AD 1520- 1670)t Koniag midden
Settlement Point  AFG-015t y    
Beta 101551 620+50 (AD 1305-1400) t House 1 hearth
Beta 118300 570+60 (AD 1310-1430)t House 1 floor
Beta 114202 440+60 (AD 1430-1480)t House 5 hearth
Beta 114204 450+50 (AD 1430-1470)t House 7 hearth
Beta 101912 440+50 (AD 1430-1480)t Refuse midden
Beta 114096 370+80 (AD 1440-1640)t Refuse midden
Beta 101913 390+50 (AD 1450-1630)t Refuse midden
Beta 114097 350+70 (AD 1450-1650)t House 3 hearth
Beta 114098 340+60 (AD 1470-1650)t Midden
Beta 114203 330+60 (AD 1470-1650)t House 4 hearth
Beta 101552 300+50 (AD 1520-1660)t House 2 hearth
Beta 114205 300+50 (AD 1520-1660)t House 6 hearth
Aleut Town      
Beta 150810 950+50 BP (AD 1010-1230 2 sigma) tBase of site?
Beta 150811 1090+80 BP (AD 770-1140 2 sigma) tHearth in midden

x Calibration is from MASCA tables published in 1973, 1 sigma probability (which is 67% probability that the age actually is within the stated range)
y  Calibration by Partlow after CALIB REV 3.0, 1 sigma probabilityOther calibrations are supplied by the dating laboratory.
++ In some cases not the “measured age” but the “conventional age”, as reported by the analysing lab.  The distinction is very technical.

The Dig Afognak Excavation Projects

Aleut Town 1999 and 2000

      In 1999, the Native Village of Afognak (NVA) engaged the Afognak Native Corporation (ANC) to test an archaeological midden site located at the Old Afognak Village. The project was resumed for a second season in 2000. The work was done with U.S. Bureau of Indian Affairs support to mitigate the proposed sale of land upon which the site is situated. In 1999, the excavation consisted of a 4 by 5 meter block (20 square meters) dug to the base of cultural deposits. In 2000, an additional 20 square meters were excavated. Dr. D.W. Clark was the archaeologist in charge for both seasons. Logistical support, staff, and equipment were provided through ANC’s and NVA’s Dig Afognak program. A dedicated assistant, Rose Ward, was hired to assist Clark, and various members of the staff of the Dig Afognak camp, as well as camp visitors, also assisted at all times. Each year, persons from the camp, including youth there for programs, undertook the task of backfilling. Wet and stormy weather impeded the work, less so the second season, when a large weather port was erected over the work area. Each season, 28 days were available for work at the site out of a 53-day and a 44-day field season, inclusive of breaks, respectively. Less time was lost because of bad weather the second season.

      Approximately 500 historic artefacts (400 catalogue lots) were collected from above the 1912 Katmai ash and from the upper part of the underlying soil in 1999. In 2000, an additional 2,500 historic items were found. These are mainly ceramic sherds, nails, bottle glass, window (flat) glass, and parts of cast iron stoves. The site had become abandoned progressively during the 20th century, and the last one was vacated about 1961. Remaining traces of the houses were swept away by tidal waves in 1964. Relatively few items, among them two .44 Henry cartridge cases and a few painted ceramic sherds, represent the early American period before about 1900. The Russian period, before 1868, possibly is not represented at all. There are approximately 800 indigenous artifacts from precontact levels predating the end of the 18th century. Most artifacts are of Late Kachemak styles, and two radiocarbon dates indicate that the site was occupied at the very end of the Kachemak tradition. Remains from the ancestral Alutiiq Koniag tradition, dating after about 1,200 A.D., are very sparse. The Alutiiq village was recorded historically, in 1796, by its name which local persons identify as the location where we excavated. Apparently, therefore, to explain the missing evidence, the Koniag and early Russian period evidence has been lost through erosion that had removed a significant portion of the site and by “landscaping” of the surface during the past century. Almost the entire precontact occupation was by people of the Late Kachemak tradition, which predated 1,200 A.D. Bone preservation was good, resulting in the recovery of numerous harpoon heads and other bone artifacts.

      Two Kachemak tradition houses, or one house with two rooms (a certain area that was not excavated could have held a connecting passage) had been dug to the base of the site through an underlying volcanic ash loam to beach gravel. On the floors, there were almost no implements and trash other than ulu blades. This shows a high standard of housekeeping and neatness in ancient times. Almost all stone tools were made of slate; only two implements were flaked from chert or flinty stone.

Tsunami Site 2001, 2003
      The name of this site is derived from the presence of a tidal wave-deposited layer that may have triggered abandonment, though the site was reoccupied later. It is located at the bend in the Afognak River estuary. The site was identified as likely Kachemak in age, and as consisting of a house structure, from a text pit dug in 1997. It was considered to be of possible Early Kachemak age because a very-early Kachemak and Ocean Bay transitional site was situated nearby. Again, the project was fielded by the Native Village of Afognak and the Afognak Native Corporation. It operated out of, and was supported by, their Dig Afognak camp at Katenee. Three full-time helpers, seasonal employee Lisa Peterson from Kodiak, and student assistants Analisa Gregoire and Kent Cardwell, did the excavation. From time to time, Dig Afognak visitors and youth participating in the Teacher’s Academy and Science camp, particularly Justin and Tristen Kewan, Arthur May, and Brian Knagin assisted. Visiting geologist Gary Carver identified a tsunami (tidal wave) layer within the site and 4 to 8 additional tsunami layers from a core taken of a boggy area next to the site. Senior camp staff and Olga Pestrikoff’s family backfilled the excavation. The second season, 2003, was planned to excavate further at one end of the structure uncovered in 2001. It was briefer and uncovered a smaller area. That year, Clark was assisted by Jimmy Charliaga. Excavation was carried out for 22 days during a 53 day period in 2001. Work was undertaken in the open, but the crew was forced to stay away only 2 ½ days because of rain.

      The excavation consisted of an area five by seven meters (35 square meters). A subrectangular housepit (semisubterranean house) was uncovered. It was approximately 4 meters wide and 7 meters long. One part of the house appeared to extend beyond the area excavated. When this area was investigated further in 2003, it was found that, in fact, the deposits thought to be unexcavated fill at the end of the Tsunami House were from another house of about the same age and that in its construction lay immediately adjacent to, and had clipped off the very end of, the Tsunami House.

      Nearly 300 artifacts were recovered from the house fill and floor. Artifact styles and radiocarbon dating indicate that the house is Late Kachemak, but not as late as the Aleut Town site. Thus, it nicely complements the latter. Artifacts recovered from deposits of another house excavated in 2003 also are of Late Kachemak styles. Implements are mainly slate ulu knife blades, which would have been used for cutting salmon; abraders and hones used to form and sharpen the ulus; flaked and ground slate projectile points; and notched pebble net weights. More flaked stone was found than had been expected on the basis of the excavations at Aleut Town. This suggests that lithic technology may have changed very rapidly during the few centuries intervening between the two occupations. There also are stone lamps and labrets. The complex floor that was uncovered, with clay pits and stone slab features, is described and illustrated in the discussion of Kachemak houses.

Salmon Bend 2002
      This site is located at the bend in the Afognak River estuary directly across from the Tsunami Site. Artifacts collected from the eroded shore-edge of the site indicated that the compound house pit seen several yards in from the bank would be Late Kachemak in age. This was verified by the 2002 excavation and radiocarbon dating. Interest in the site was perked among Afognak Corporation members by the discovery of a female breasted lamp exposed on the shore the previous year, while from an investigative position, the site was considered to be important because the house was made up by two pit structures. Usually there is a single room or pit in Kachemak tradition houses.

      This was an NVA project. D.W. Clark was the archaeologist in charge. Megan Holmes was provided by the NVA to assist. A number of adults at the Dig Afognak Science Camp and other camps participated, usually for one day, and an attempt was made to expose camp youth to archaeology. Young children were of limited value as assistants, as it turned out that their interest could not be held any longer than three hours, and it turned out to be very disruptive to keep them at the site for a whole day.

AFG-088 2004
      This ancient fishcamp site, located along the banks of the river estuary at Litnik, was recognized as a possible transitional Ocean Bay/Kachemak site on the basis of artifacts collected from the shore in 1971. D.W. Clark excavated three test pits here in the 1990s and found an apparently transitional mix of Kachemak and Ocean bay artifact styles, leading him to believe that the site actually documented a transition from the one culture to the other. Radiocarbon dating showed that the deposit was early Kachemak in age, 2,800 to 3,400 years old. (With calibration or correction, the older dates are two or three centuries earlier. See earlier table).

      More extensive minor excavation was undertaken in 2004. It was then determined that there were remains of a Late Ocean Bay or Ocean Bay II occupation in the volcanic ash layers at the base of the site, expressed mainly in ground slate tools. It was about 6,000 years old. This was followed by a zone that yielded very few implements. The few recovered were of forms characteristic of both Late Ocean Bay and Early Kachemak. The upper zone of the site (as defined by artifact styles) yielded abundant large notched pebble sinkers, a characteristic Early Kachemak artifact. There also were a moderate number of ground slate ulu fragments. There were not very many other items. A few show Late Ocean Bay characteristics, but the data are too weak to prove conclusively a technological transition from the Ocean Bay to the Kachemak culture.

      This site was collected and tested by Clark in 1951, and also visited by Clark and Workman in 1964 and 1971. Adjacent site Afg-013 may be included, though today a small creek separates the two. An evidently large portion of the site formerly lay northeast of the present midden remnant and was observed by Clark in 1951. There were many artifacts on the beach the first year, few at the time of the second and third visits. Megan Partlow and three assistants made a test excavation in 1996, for the purpose of obtaining a sample of faunal remains. At that time, traces of two multi-room houses were visible on the surface. Artifacts recovered by Partlow included incised pebbles, adzes, ground slate ulus, and ground slate projectile points. A slate slab box lid with chipped hand-hold was found on the erosion face at Afg-012, and an identical lid was found on the face of Afg-013. These are similar to specimens found at Settlement Point. Earlier, Clark also had found stone lamps and splitting adzes (on the beach), a cobble with three notches, a jet labret, and a “boat-shaped” slate end-blade, plus numerous greenstone adzes.


      Characteristics of the three regional archaeological traditions were discussed earlier. These traditions were distributed across the North Pacific area that extends from Prince William Sound, outer Cook Inlet and the Kenai Peninsula, the Kodiak Archipelago, and the adjacent Alaska Peninsula mainland. The boundaries of this culture area remained largely unchanged for 7,500 years. This region was not wholly uniform throughout time. Local and temporal variations did occur. Viewed in a general manner, the ancient communities of Afognak Bay fit very well within the broader framework of prehistory. Some differences can be discerned with close scrutiny, and we are concerned to know whether these are random variations or part of a broader regional pattern that featured progressive change as one moves northeastward from the south end of Kodiak Island to Cook Inlet.

      Technologically viewed, the late people of the Kodiak archipelago were slate tool makers. In fact, before the Ocean Bay culture was discovered, only slate grinders were known from Kodiak. Elsewhere in Alaska, all cultures of the past, with some late exceptions, had chipped tools flaked from flinty materials. At one time it was thought that technology moved rapidly from flaked stone to ground slate on Kodiak—the Ocean Bay I/II boundary—but it was soon found that a form of Ocean Bay I continued to coexist with Ocean Bay II. It then appeared that ground slate gained final ascendancy in Early Kachemak times on Kodiak, but it was recognized that at Cook Inlet and on the Alaska Peninsula there was a lot of stone flaking in Early Kachemak times. Data from a number of site excavations do show that, for the most part, Late Kachemak of the second millennium A.D. and the early Alutiit after 1,200 A.D. had completely ceased to make flaked stone tools. This is documented by Hrdlicka’s excavations at the Uyak site, the author’s work at Late Kachemak Crag Point (Anton Larsen Bay), Saltonstall’s excavations of Koniag houses at Settlement Point on Afognak, Clark’s work at Late Kachemak Three Saints, and his excavations at the late Koniag Rolling Bay and Kiavak villages. We state “for the most part” because the occasional flaked tool still was found at every site. Additionally, blanks for adz bits were produced by flaking hard stone.

      Six seasons of recent excavations at Afognak Bay now require that this scenario be amended, at least for that area. Specifically, after it had nearly ceased in late Ocean Bay times, flaked stone regained a modest importance at the first millennium A.D. Salmon Bend and Tsunami sites. There were 25 ground slate points and 22 flaked points, plus an additional 17 flaked points found “cached” together. The total flaked chert industry numbered 56 items, not including waste flakes and the 17 points. As expected, though, this number is greatly surpassed by ulu blades and fragments, showing that slate was greatly favored for ulus. In the case of projectile points, though, the early people at Litnik appeared to be at ease with two technological vocabularies. Later, at the end of the first millennium or beginning of the second millennium, there occurred a strong shift at the Aleut Town site. Only two chert artifacts were recovered. This fit the Late Kachemak and Koniag model established elsewhere on Kodiak Island. Aleut Town results are confirmed by collections from the eroded fringes of the lagoon at Litnik, where almost no chert artifacts were found among numerous items washed out of Koniag culture camps.

      In sum, at Afognak Bay, at Litnik specifically, there was a greater, later emphasis on flaked stone artifacts than was the case elsewhere in the archipelago. The case appears to be the same for Kachemak Bay. This may reflect an area trend towards retention of flaking technology towards the northeast. It suggests that Afognak may have had a stronger alignment with the outer Cook Inlet area than with southwestern Kodiak Island. Significantly, though, no stone lamps with sphinx-like human figures carved in the bowl have been found on Afognak (and Kodiak), though they are rather common in the vicinity of Kachemak Bay, and other Kachemak-Cook Inlet lamp styles are common on Afognak.   

     The example of knife and spear hafts appears to fit this scenario. In Ocean Bay II times, the stems of ground slate knife blades and spear heads often were finely notched or serrated to assist fixing to a haft. This practice was carried over into Early Kachemak times. Stem serration also was popular in Late Kachemak times, but mainly on Kodiak Island itself (and Chirikof Island). By then, this practice had gone out of style on Afognak. There were many other regional differences, but altogether an extensive repertoire of Kachemak technology and cultural practice pervaded the greater region.

A product of the Afognak Data Recovery Project

To embrace, protect, develop, and enhance Alutiiq culture, protect our traditional use areas and encourage unity among the Alutiiq of the Kodiak Archipelago