Fishing For Birds

Technological Innovation and Marine Adaptations on the Bering Coast of Alaska


Nunivak Islanders Andrew Noatak and Peter Smith recounted a story about the evolution of walrus skin ropes on Nunivak Island (Hoffman, 1990). According to Noatak and Smith, in ancient times men hung from cliffs on sealskin ropes for the purpose of collecting seabirds and their eggs. But sealskin ropes tend to rot in sunlight, and many hunters plummeted to their deaths. A poor man, unable to afford seal skins for rope-making, made his rope from walrus skin. The stodgy town elders warned him that his cheap walrus skin rope would break, and he would be killed. The rope held up beautifully, however, and within a short period all Nunivarrmiut cliff-hangers used walrus-skin ropes. Cliff-hanging had become a safer and more productive subsistence activity.

Anthropologists often describe the evolution of a subsistence adaptation in ecological or behavioral terms. The tacit assumption is that anthropological peoples “react to” or “interact with” their environments, and as environments and populations change, so do technologies, almost as a matter of reflex. Noatak and Smith’s story suggests that innovation and personal decision-making affect technology in anthropological societies, in much the same way that innovation and choice affect technology in industrial societies. To understand the changes in marine adaptations among Eskimo and proto-Eskimo peoples, we must understand the nature of the technologies that influence these adaptations, as well as the mechanisms by which such technologies evolve.

Ethnographers have represented Eskimo culture as primarily maritime (R. K. Nelson 1969), or as broadly-based on a variety of terrestrial and marine resources (Lantis 1946, 1986). The patchy but seasonally-abundant foodstuffs of the Bering Sea region demand a flexible lifestyle with shifting emphases on the most available resources. An overdependence on a few resources can have disastrous consequences when those few items become scarce (Shaw and Holmes 1982).

I propose that an increased reliance on marine resources apparent in the Norton and Thule traditions (ca. 2500 years b.p. to 1000 years b. p.) represents a broadening of the overall resource base, coupled with more intensive use of all resources. The advent of sophisticated sealing and netting technologies permitted prehistoric Bering Sea Alaskans to exploit a variety of marine and “para-marine” resources including anadramous and pelagic fish, sea birds, and sea mammals. The shift from lifeways focused on only two or three primary resources to lifeways with no particular focus allowed late Norton and Thule peoples to expand throughout the Bering Sea region and across the North American Arctic.


In general, two factors influence the productivity of marine and terrestrial resources in the Yukon-Kuskokwim delta and on Nunivak Island: upwellings of organic materials from deep in the Pacific and the outer Bering Sea, and the stability of climatic regimes that affect the extent of forests and the density of terrestrial biomass.

The nutrient-laden waters of the Bering Sea fuel the rich biome of the western Alaska coast. Relatively warm, saline waters from the North Pacific and outer Bering flow along the edge of the Bering shelf, mixing with the colder and fresher water there. During the summer, the large influx of cold fresh water from the melting polar ice mass, and the silty freshwater runoff of the Yukon and Kuskokwim rivers form a salicline limiting the infusion of nutrients into the waters of the Bering shelf. In the winter, as the ice mass grows to cover the shallow Bering Sea, vertical mixing of the deep saline and shallow brackish water along the edge of the shelf boosts shallow water productivity. The high productivity of the marine biome in the Bering Sea results from this complex process, making the waters there among the richest in the world (Ackerman 1988: 53–58).

Ackerman (1988) summarized sea-level fluctuations for the Bering Sea region. During the late Pleistocene and early Holocene (22,000 b.p. to 10,000 b.p.), sea levels in the region dropped, exposing the Bering Shelf to below the -120m isobath, leaving the coastline as much as 400 km from present-day Nunivak Island. After about 15,000 b.p., sea levels rose gradually, reaching their present level around 5,000 b.p. For the purposes of this paper (which deals with archaeological sequences after this time), I will assume that sea levels were at or near their present level for the last 5,000 years or so.

Shaw (1982: 9–33) described the geography of the Yukon-Kuskokwim delta region. The massive Yukon and Kuskokwim river deltas dominate the geography of their physiographic region (figure 1). The entire delta region lies, on average, below 30m in elevation, sloping gradually from the extensive coastal marshlands up to more stable alluvial/eolian tundra inland. Tidal fluctuations affect the meandering streams of the delta as much as 60km inland. The Bering sea maritime regime softens the harsh subarctic climate, but conditions become progressively more severe toward the interior, with colder winters and warmer summers.

Nunivak Island lies about 30 miles offshore from Nelson island and the mainland
Figure 1, Yukon-Kuskokwim Delta and Nunivak Island. (After Nowak 1982)

Tundra environments predominate on the Yukon-Kuskokwim delta. Grasses, woody shrubs, and dwarf willow support a relatively sparse mammalian fauna of caribou, moose, brown bear, and a variety of small mammals including foxes, voles, rats and mice. The Bering Sea supports a rich ecosystem, however, with large fish, sea mammal, and bird populations. All five major salmon species run in both the Yukon and Kuskokwim rivers, and the estuaries support herring, whitefish, grayling, trout, turbot, and tomcod. Despite the low offshore gradient, sea mammals abound, especially near rocky headland areas. Four species of seal, and beluga whales feed on the fish here; walrus live on the margins of the ice shelf as it retreats past the delta in the spring. In the summer, the delta supports an enormous population of migratory birds. Ducks, geese, swans and cranes breed along the seasonal thaw lakes inland, while shore birds live on extensive rookeries along the coast.

Nunivak Island lies about 30 miles offshore from the mouths of the Yukon and Kuskokwim rivers, at about 60 north latitude. Lantis (1986:209) summarized the geography and biogeography of Nunivak. Winter weather on Nunivak is similar to that of the mainland: fiercely cold and wet, with temperatures falling to as low as -31 C. Summers are cooler and wetter than on the mainland, and the vegetation on Nunivak is denser than the mainland tundra. Like the delta, the terrestrial biome is less productive in general than the marine biome. The island sits on a long, shallow shelf that receives the seasonal Pacific and deep Bering upwellings, as well as silty currents from the delta. Salmon, Dolly Varden, cod, halibut, flounder, needlefish, tomcod and herring feed on the dense plankton of the near shore currents. Harbor, spotted, and bearded seals and beluga whales feed in turn on the fish. In late winter through spring, walruses abound in Nunivak waters, feeding on deepwater clams living on the Nunivak shelf. Many of the terrestrial fauna depend directly on marine resources; Nunivak boasts a dense population of migratory sea-birds, especially on the steep cliffs forming the southwestern corner of the island. A lack of predators other than foxes and minks coupled with the large near-shore fish populations make Nunivak a prime rookery for at least 89 bird species—mostly ducks, murres, cormorants, puffins, and gulls (Hoffman 1990). Caribou are the only large mammals found archaeologically on Nunivak. Euroamericans introduced reindeer and musk oxen to the island this century (Dennis Griffin, pers. comm.). Rats, voles, foxes and minks are the only small mammals.

Based on Ackerman’s (1988) assessment of local resources, I will split the biogeography of the Yukon-Kuskokwim delta and Nunivak Island into two biotic regimes. Productivity in the terrestrial/riverine regime is directly dependent on climatic variability, although some resources (e.g. elk) may be less seasonally variable than marine resources. Good years may be very good in the terrestrial regime, although a single poor season can affect productivity for several years. For example, a recent winter storm on Nunivak Island drove several thousand reindeer out onto the sea ice, where they froze to death or drowned. The reindeer herd will require several years to recover (Dennis Griffin, pers. comm.).

Productivity in the marine regime depends on ocean currents that may be affected by ice melt or climatic patterns further south in the Pacific. The marine biota relies on upwellings from the deeper ocean, and marine resources may be more stable than terrestrial resources from year to year. The winter freezing of the Bering Sea makes the extraction of marine resources seasonally-variable, however, meaning that people utilizing marine resources must alter their procurement strategies during certain seasons to keep food on the table.

The Archaeological Record

The earliest evidence for broad marine adaptations on the Bering Sea of Alaska appears in the Small Tool Tradition, around 4500 years ago (figure 2). Dumond (1986a) sees little evidence for marine resource use at the most northern coastal Small Tool Tradition site, Cape Denbigh, between 4500 and 3400 years B.P. At Cape Denbigh, the relative abundance of small blade tools over triangular bifaces usually identified as harpoon heads suggests that the people living there focused primarily on caribou or other large land game (Dumond 1986a), although the presence of seal bones (in unknown quantities) implies that the early inhabitants there hunted seals (Giddings 1964). The paucity of published faunal data from Denbigh, coupled with its early excavation, make the interpretation of the site frustrating. Dumond (1981) also describes an Arctic Small Tool assemblage during the Gravels Phase (3800–3300 years B.P.) at the Brooks River sites, south of my primary study area and somewhat inland. Dumond’s crew recovered seven salmonid bones from this phase (Dumond 1981: 121), leading Dumond to surmise that the Gravels Phase peoples engaged in riverine fishing, although he also reports the presence of “unidentifiable” mammal bones, the implications of which are uncertain.

Thule assemblages appear after 1000 b.p. throughout the study area
Figure 2, Generalized chronology of Alaskan Bering Sea Prehistory. (After Dumond 1987, Nowak 1982, and Shaw 1982)

In their analysis of Small Tool Tradition adaptations, Dumond (1986b) and Dumond and Bland (1995) outline a broad-based interior-oriented economy focused on riverine and large terrestrial mammal resources, with occasional forays to the “sea-edge,” apparently for the purpose of sealing. Lutz (1982) interprets the shift to marine resources at the end of the Small Tool Tradition as a response to worsening winter conditions in the interior of Alaska. Anderson (1986, 1988) believes that Denbigh and other Small Tool Tradition sites represent seasonally-variant emphases: summer sealing camps found near the coast alternating with caribou hunting settlements found in the interior.

Marine adaptations became widespread during the Norton Tradition, after about 2500 years ago. Overall, Norton peoples displayed an deepened interest in the sea, with the establishment of larger, more permanent coastal villages, and an increase in fishing and sea mammal hunting generally (Dumond 1982). An increase in netsinkers in Norton assemblages and an apparent increase in the intensity of riverine occupations suggests to Lutz (1982) greater exploitation of anadromous fish. Dumond (1987: 127) characterizes late Norton peoples as riverine fishers, catching salmon with traps or spears.

Shaw, in his work on the Yukon-Kuskokwim delta, noted new technologies in assemblages associated with typologically late Norton check-stamped pottery. At MAR-007 on the Manokinak River, Shaw (1983: 118–130, 245–252, 264–265) recovered a single net weight and the remains of a net, in addition to bone fishhooks, dart points, and wooden fishing implements including a fish trap and fish clubs. Carved wooden artifacts representing cranes or ducks and a decorative puffin bill reflect an interest at some symbolic or ideological level in birds (Shaw 1983: 156–161, 252, 360). Norton settlements on the delta were clustered at coastal localities with protected embayments, on headlands, or along rivers in the interior uplands (Shaw 1982: 70).

Shaw (1982: 71; 1983: 359–360) characterizes Norton hunting and fishing as “labor-intensive,” relying mostly on large but infrequent takes of each major resource: fish (especially salmon), seals, and caribou. If Norton people could rely on a small take of at least two of the resources, they could remain well-fed and healthy. But if more than one of the resources collapsed, the people faced hard times. Shaw’s work has implications for the breadth of the Norton resource base (see Understanding Marine Adaptations, below).

On Nunivak, the first evidence for human occupation appears with Norton Tradition materials. At the Mekoryuk Village site XNI-028 (figure 3), Nowak (1982, 1988) identified two components he termed the Early Duchikmiut Phase (150 B.C. to A. D. 250) and the Late Duchikmiut Phase (A. D. 250 to 650), containing typologically Norton check-stamped pottery. Nowak (1982: 78–87) describes a technological variation between the two phases. Small-gauge, thin walled check-stamped pottery and small ground and chipped lithics predominated in the Early phase; thick walled large-gauge check-stamped pottery, bone tools, and large ground slate tools (including knives) formed the bulk of the Late phase assemblage. Nowak apparently recovered a large faunal assemblage from Mekoryuk Village, but this assemblage is poorly-documented. He reports finding hair seal, bearded seal, sea lion, and beluga whale in large quantities at XNI-028. I find the evidence for net fishing technologies at XNI-028 more compelling; Nowak and his crew found more than 1400 notched netsinkers of uncertain size and function. More than 1000 of these netsinkers appear in the Late Phase, although Nowak believes concentrations of netsinkers in the Late Phase represent caches rather than net remains, with the implication that the greater number of netsinkers in this phase may not represent the increased use of nets. At any rate, the large netsinker assemblage suggests an increased reliance on mass fish-harvesting, seal-netting, or bird-netting (see Understanding Marine Adaptations, below). The lack of change in the faunal assemblage from Norton to late historic times indicates to Nowak that historic Nunivarrmiut have been living in much the same way for more than 2000 years. A paucity of quantified faunal data makes fine-tuned assessments of prehistoric Nunivarrmiut subsistence difficult.

Most locations discussed in this paper are near Mekoryuk on the northeast coast of the island
Figure 3, Nunivak Island. (After Hoffman 1990)

Finally, the appearance of human settlements on Nunivak during Norton times (and not earlier) indicates marine lifeways, or at least boat use. Since sea levels had been stable for some 2000 or more years before the initial habitation of Mekoryuk, the only way to reach Nunivak during the summer was by boat. Winter crossings to the island over the ice may have been possible, but I could find no ethnographic evidence indicating that modern Nunivarrmiut ventured more than a few miles out onto sea ice, and then largely for the purpose of hunting or fishing (Lantis, 1986; VanStone, 1989). This line of reasoning follows from negative evidence (the lack of pre-Norton sites on Nunivak); subsequent excavations on Nunivak may push the date of colonization back to an earlier date.

The people of the Thule Tradition (after A. D. 1000) seem to have established the classic Eskimo lifeway that continued until contact times. Evidence for early Thule occupations appears infrequent on the delta, although this is probably an artifact of the lack of archaeological investigation in the region. Shaw (1983) uncovered a late Norton component he terms the Yukon-Kuskokwim Phase (A. D. 400–1000) from MAR-007, then describes a probable hiatus in occupation from A. D. 1300 to the ethnohistoric present. The late appearance of net technology and bird fauna at the site led Shaw to hypothesize that the people at the Manokinak River had adapted net technology for hunting waterfowl and small terrestrial mammals (Shaw 1982: 71–72; see Understanding Marine Adaptations, below).

Researchers on Nunivak Island have published very little data on Thule occupations there. Nowak mentions a potential Thule-era occupation (A. D. 1200) at the Mekoryuk shell midden site XNI-030, and notes that he identified neither walrus nor beluga at the site, but that sea lions "appear in small numbers" (Nowak 1988: 46). Nowak (1982) seems to have dismissed the possibility of substantial changes after about 2000 B.P. in Nunivarrmiut subsistence activities, based on his summary analysis of the faunal assemblage from the Mekoryuk Village site, XNI-028.

Understanding Marine Adaptations

The economy of Norton peoples depended upon the reliable extraction of two of three major resources. Thule peoples extended this economy with the introduction of pelagic sea mammal hunting and small terrestrial game hunting (Shaw 1982, 1983; Shaw and Holmes 1982; Dumond 1982, 1987). Understanding the intensification of marine resource use at the Norton-Thule transition requires a fusion of archaeological and ethnographic data about this “new” economy.

Norton peoples exploited near-shore sea mammals with a certain (largely unknown) level of intensity. Ackerman (1988: 68) implies that the sudden appearance of technologies associated with pelagic sea-mammal hunting in late Norton times represents a sudden realization of the importance of marine resources, followed by the quick adoption of brand new technologies for their extraction. Ackerman couples this “revolution” in open-water hunting with a climatic warming trend around a thousand years ago, which made marine mammals more locally abundant. Ackerman overlooks ethnographic evidence suggesting that sea mammal hunting technologies involve bone and wood artifacts that preserve poorly. Van Stone (1989) describes a large array of tools associated with marine hunting, from a variety of harpoon heads to eyeshades and pronged sea-bird darts. Only a few darts or harpoons have stone components, usually a triangular ground slate tip. The remainder of the sea-mammal hunting toolkit involves organic artifacts that would preserve poorly in an archaeological context, making arguments based on the lack of marine mammal-hunting tools at archaeological sites questionable.

An examination of the ethnographic record for Nunivak Island indicates that sea-mammal hunting usually involved very small hunting parties or solitary hunters. VanStone (1989) notes that seal hunters usually worked from inside a kayak, often alone or in groups of two or three. He observed that lone hunters have the most success in sealing, since a single kayak has the best chance of approaching a seal. Hunters employed a limited toolkit consisting of light harpoons and darts, thrown with an atlatl-like board (figure 4). Harpoon darts were tethered to sealskin floats, enabling the hunters to hit a seal and then follow it at their leisure until it drowned and floated to the surface. Women and infirm hunters may have also hunted seals at air holes in the winter ice (Lantis 1946), but VanStone (1989:8) suggests that the Nunivarrmiut rarely engaged in such a practice.

Figure 4, Nunivarrmiut harpoon heads. (After VanStone 1989)

In addition to actively pursuing seals, the Nunivarrmiut maintained large seal nets, which they set through holes in the winter ice or parallel to shore during the spring and summer (VanStone 1989: 8–9). They weighted their seal nets with large notched pebble sinkers (VanStone 1989:9) which should preserve well archaeologically. Nets may also have provided some of the beluga found on Nunivak, which are too scarce around the island to warrant active pursuit (Lantis 1986: 213). Nelson (1899:130) describes the active setting of nets around rocky reefs on the mainland for the purpose of hunting “white whales” (beluga). Unlike seal nets, which the mainland Yupik placed near to shore, beluga nets were set away from sight of land, requiring almost daily visits in a kayak.

The Nunivarrmiut hunted walrus in the spring, when the edge of the ice mass retreated past the island (VanStone 1989:9). Walrus hunters often worked in small groups, since in the water a walrus is capable of tipping a kayak. Hunters favored taking walruses on the ice while they slept, preferably killing them before they could roll into the water, where their heavy mass would drag down any floats. VanStone (1989:9) also notes that the Nunivarrmiut hunted sea lions “in open water,” presumably with techniques similar to sealing or walrus-hunting.

I side with Shaw (1982) in representing early sea-mammal hunting technologies as simple but labor-intensive. They involved the active corral and clubbing of seals, sea lions, and walruses on land or on the ice. While such activity would have yielded a large amount of meat, sinew, bone, ivory and blubber, it would have required the participation of several hunters, and concerted techniques of hauling and processing to make efficient use of the meat and blubber. Such technology would have tied prehistoric peoples to areas plentiful in these animals, meaning that settlement sizes near these areas would grow. Indeed, Shaw (1982), Dumond (1982), and Shaw and Holmes (1982) report a growth in settlement sizes during the Norton period. Further, Shaw (1982) hypothesizes that during times when the population of one animal resource (e.g. sea mammals) dropped, Norton peoples would have to intensify their harvest of the other resources (riverine fish and land mammals) and hope that these populations remained stable.

Netting technologies permitting the easy capture of seal and whales also provided a means for hunting birds. Hoffman (1990) recorded oral historic accounts of cliffhanging bird hunting practices on Nunivak Island. The important birding season was late spring through summer, when migratory seabirds came to nest on the sea cliffs on the western edge of the island (figure 3). Nunivarrmiut bird hunters caught cliff-dwelling birds with long-handled scoop nets or snares, or with unweighted throw nets (figure 5a), (figure 5b).The hunters would lower these nets from atop the cliffs, or suspend themselves from the cliff face by walrus-skin ropes. Cliff-hangers usually worked in small teams of two or three, with one hunter hanging over the cliff edge and the others steadying his ropes. Nelson (1899:131–135) describes a variety of other birding techniques in practice on the delta. The Yukon-Kuskokwim Yupik employed a variety of snares and loops probably similar to those used on Nunivak, as well as simple ptarmigan and duck decoys, throw nets and bolos weighted with decorated wood, ivory, bone or stone. Notably, most of the bird-hunting toolkit utilized organic tools, which may not survive well archaeologically, or which archaeologists may misinterpret as part of a fishing toolkit.

Figure 5a, Nunivarrmiut bird snares and "dip" nets. (From Hoffman 1990)

Figure 5B, Nunivarrmiut birding throw net. (From Hoffman 1990)

Bird hunting also provided important non-food resources, such as skins, feathers, and colorful beaks. Pratt (1990) summarized the economic value of bird skins by comparing the number of skins required of a certain type of animal necessary to sew a single parka, and Nunivarrmiut estimates of the trade worth of each skin. Pratt concluded that cliffhanging provided important surplus foods in the form of dried meat and eggs (both of which could be kept through the winter), and that bird skins were the third most valuable trade item for Nunivarrmiut (behind whale sinew and bearded seal skins). Pratt (1990: 76) also suggests that cliff-hangers collected wild lettuce and other greens clinging to cliff faces, a reminder that not all resources (even in the Subarctic) come in animal packages.

Advances in net technologies (probably coupled with improved boats and kayaks) allowed Norton fisherpeople to exploit a very profitable resource: pelagic fish feeding on the nutrients upwelling from the Pacific and deep Bering Sea. On Nunivak Island, Lantis (1986) describes ethnographic fishing techniques that do not emphasize netting, principally riverine fishing with stone weirs and traps, and hook-and-line fishing from canoes (in the summer) or through holes in the ice (in the winter). VanStone (1989) describes Nunivarrmiut fishing for tomcod with small three-pronged spears, and the taking of large codfish and halibut with hooks. He also reports that the Nunivarrmiut fished for tomcod and smelt in near shore waters with fine-gauge weighted nets, and for flounder and herring with seine nets. Dipnets and a variety of traps rounded out the ethnographic toolkit, but appeared to have been less important than spears and nets. Nelson (1899: 188–189) illustrated gill nets with wooden net floats and antler or walrus-rib sinkers. All of these fishing techniques required only a few people; VanStone (1989: 13) describes a seine-netting method where a woman remains on shore with one end of the net line, while a man paddles from shore in a kayak holding the other end. Considering the large number of netsinkers Nowak (1982) found at the Mekoryuk Village sites (see The Archaeological Record, above), ancient Nunivarrmiut may have engaged in more large-scale, intensive net fishing in the past. Overall, recent ethnographic fishing practices (like other hunting strategies) emphasize the ability of a lone fisherperson or small group of fisherpeople to catch a large number of fish rapidly. The importance of solo hunter/fishers in modern and late prehistoric Yupik culture may have evolved from earlier strategies that focused more on mass-harvesting.

The gradual evolution of large kayaks, composite toggling harpoons, nets, and floats allowed small groups of hunters to engage in more “casual” or occasional hunting, with lower processing overhead. The addition of birdnets, snares, cliff-hanging techniques and bird decoys allowed ancient Bering Coast Alaskans to expand their resources into new areas altogether. Advanced gill nets probably became the most important technological innovation during late Norton times. By allowing the rapid extraction of large numbers of fish by a few fisherpeople, on virtually any day of the spring or summer, Norton peoples had access to a rich resource with relatively little fluctuation from year to year.

Norton peoples probably didn't arrive at these technologies overnight; rather, they developed them with the accumulation of small technological advances over a few hundred years. In fact, the magnitude of resource expansion from early Norton times (ca. 2000–2500 years ago) to Thule-transition times (ca. 1500–1000 years ago) impresses me as quite rapid (only 1000 years or so). Prehistoric Alaskans didn't devise these tools and techniques from scratch; they molded existing technologies (single-piece harpoons, barbed darts, riverine fishnets and traps) into novel applications (composite toggling harpoons, birdnets, sealnets, gill nets).

I propose that the seemingly “sudden” appearance of complex net, dart, and harpoon technologies around 1000 years ago represents an artifact of differential preservation and limited archaeological exploration, rather than a sudden realization by ancient Alaskans of the productivity of marine resources. Dumond (1987: 133) also discounts the possibility that the Norton-Thule transition represents a migration or replacement of Norton populations by Thule peoples, arguing that similarities between house styles and tool types imply an in situ cultural evolution.

While the archaeological record does suggest an increase in marine resource use around 2000 to 1000 years ago, this increase may be a result of accumulated technological advances in the Norton toolkit. The resulting highly-specialized and multiple-use toolkit emphasizing the solo hunter persisted until the ethnographic present.

Conclusions & Caveats

A variety of shortcomings in the archaeology of Nunivak Island and the Yukon-Kuskokwim delta complicate the interpretation of marine adaptations here. The tool-seriation focus of earlier researchers led to an underreporting of faunal materials and "uninteresting" artifact groups. With a lack of published paleobotanical or faunal studies, the paleoecology of the region remains uncertain, as do the subsistence behaviors of early inhabitants. A refocus on distinctive technologies that may indicate marine adaptations (e.g. netsinkers) helps to address this issue, with the caveat that unevenly-employed recovery techniques may skew artifact distributions as well. The lack of published primary data also leads to an over-reliance on derivative secondary literature, suggesting the possibility that netsinkers or other artifacts may have been recovered in greater abundance than reported.

With the above caveats, I propose that sufficient archaeological data exist to warrant a reconsideration of subsistence activities in proto-Eskimo lifeways. Following Shaw (1982) and Shaw and Holmes (1982) I characterize the late prehistory of Bering Coast Alaska as a series of expansions in the resource-exploitation base, coupled with the development of advanced sealing, fishing, and birding technologies. Before the Norton period (2500 years ago and earlier), the inhabitants of the Yukon-Kuskokwim delta probably focused most heavily on large marine and terrestrial mammals. In the absence of direct data suggesting pelagic fishing or hunting, I see the overall lifeway as directed toward large-package resources, with intensive focus on only a few species. Lacking highly-specialized and efficient hunting and fishing tools, such a strategy would yield a reliable amount of food throughout the year, but with high labor costs.

Lutz (1982) suggests that by the Norton period, climatic changes had stressed inland forest resources, prompting a shift towards marine resources. Whatever the cause, after about 2500 years ago, a broadening of marine resource exploitations seems to have occurred, with harpoons, netsinkers, and wood-working adzes (sug-gesting increased boat-building) becoming increasingly more frequent (Bockstoce 1973). Instead of a shift in the focus of resource use, I see this expansion into new maritime niches as a deepening of lifeways already oriented in some degree toward the ocean. The increase in marine adaptations expanded the number of animal species upon which Norton preyed. Previous peoples relied on large terrestrial game, then riverine or anadromous fish and sea mammals (proabably in that order). Norton peoples increased their take of sea mammals and fish of all kinds, without apparently abandoning terrestrial hunting to any great degree. The Norton subsistence pattern forms Bockstoce’s (1973) “tripartite” strategy.

Some 1500 to 1000 years ago, Thule peoples established the stereotypical highly-flexible Eskimo lifeway. With the introduction or invention of specialized multiple-utility nets, snares, and composite toggling harpoons, Thule peoples could add seabirds, small mammals, and pelagic fish and sea mammals to their previous “tripartite” economy of large land mammals, inshore sea mammals, and riverine fish (Shaw 1982). The high variability of many of these resources dictated that dependence on only a few could have disastrous consequences. Thule peoples responded by exploiting in a very efficient (non labor-intensive) manner whichever resource happened to be abundant locally. Nowak (1982, 1988) posits that Nunivarrmiut lifeways have not altered substantially since early Norton times. I suggest that a re-examination of netsinker and bone tool assemblages, as well as the excavation of other sites elsewhere on Nunivak, may cause archaeologists to reconsider technological ingenuity as a motivator of changes in subsistence activities. Since nets and bone darts increase the number of animals killed per hunter, and since bird hunting in particular utilizes a toolkit that largely disappears archaeologically, assessment of subsistence activities on Nunivak will require more finely-tuned faunal assemblage data.

Archaeologists, like most anthropologists, explain changes in subsistence behaviors with mechanisms of “culture:” perception, attitude, or understanding. Researchers investigating increased marine adaptations among proto-Eskimo peoples characterize this increase as precipitated by either an increased understanding of marine resource use, or as a response to an environmental or population pressure. Such explanations sell short human ingenuity. Alaskans have eaten fish and sea mammals for 4000 years or more; I suspect that prehistoric Alaskans understood intimately the potential of marine resources. However, they lacked sophisticated localized technologies that enabled them to exploit these resources with a high degree of efficiency. The archaeological record contains clues suggesting that, with time, prehistoric Alaskans developed technologies that enabled them to expand (not merely shift) their subsistence activities.

Understanding changes in proto-Eskimo lifestyles may require questioning the anthropological concept of “technology.” Technology, often described with language like “shifting,” “changing,” and “revolutionary,” is an accretional process, where improvements in technique and materials accumulate (sometimes very rapidly) through time. Armed with this perception, archaeologists may find parsimonious explanations for increased marine resource use in prehistoric Alaska through the cumulative nature of technological processes.

References Cited

Ackerman, Robert E.
1988 Settlements and Sea Mammal Hunting in the Bering-Chukchi Sea Region. Arctic Anthropology 24(1):55–79
Anderson, D. D.
1986 “Prehistory of North Alaska” in The Handbook of the North American Indians, vol. 5: The Arctic, David Damas, ed. pp. 80–93. Washington: Smithsonian.
1988 Onion Portage: the Archaeology of a Stratified Site from the Kobuk River, Northwest Alaska. Archaeological Papers of the University of Alaska, vol. 22 nos. 1 & 2. Fairbanks: University of Alaska.
Bockstoce, John
1973 A Prehistoric Population Change in the Bering Strait Region. Polar Record 16:793–803.
Dumond, Don
1981 Archaeology on the Alaska Peninsula: The Naknek Region, 1960 1975. University of Oregon Anthropological Papers no. 21.
1982 Trends and Traditions in Alaskan Prehistory: The Place of Norton Culture. Arctic Anthropology 19(2):39–51
1986a “Prehistory of the Bering Sea Region,” in The Handbook of the North American Indian, vol. 5: The Arctic, David Damas, ed. pp. 94–105. Washington: Smithsonian
1986b “Prehistory: Summary,” in The Handbook of the North American Indian, vol. 5: The Arctic, David Damas, ed. pp. 72–79. Washington: Smithsonian
1987 The Eskimos and Aleuts. London: Thames and Hudson, Ltd.
Dumond, Don E., and Richard L. Bland
1995 Holocene Prehistory of the Northernmost North Pacific, Journal of World Prehistory 9(4):401–451
Giddings, J. L.
1964 The Archaeology of Cape Denbigh. Providence.
Hoffman, Brian W.
1990 Bird Netting, Cliff-Hanging, and Egg Gathering: Traditional Procurement Strategies on Nunivak Island. Arctic Anthropology 27(1): 66–74.
Lantis, Margaret
1946 The Social Culture of the Nunivak Eskimo. Philadelphia: Transactions of the American Philosophical Society, new series, vol.35, pt.3.
1986 “Nunivak Eskimo” in The Handbook of the North American Indian, vol. 5: The Arctic, David Damas, ed. pp. 209–223, 105. Washington: Smithsonian.
Lutz, Bruce J.
1982 Population Pressure and Climate as Dynamics within the Arctic Small Tool Tradition of Alaska. Arctic Anthropology 19(2):143–149
Nelson, Richard K.
1969 Hunters of the Northern Ice. University of Chicago.
Nelson, Edward William
1899 The Eskimo About Bering Strait. Bureau of American Ethnology, vol. 18 pt. 1
Nowak, Michael
1982 The Norton Period of Nunivak Island: Internal Change and External Influence. Arctic Anthropology 19(2):75–91
1988 Sea Mammals in a Mixed Economy: A Southwestern Alaskan Case. Arctic Anthropology 25(1): 44–51.
Pratt, Kenneth L.
1990 Economic and Social Aspects of Nunivak Eskimo “Cliff-Hanging.” Arctic Anthropology 27(1):75–86
Shaw, Robert D.
1982 The Expansion and Survival of the Norton Tradition on the Yukon-Kuskokwim Delta. Arctic Anthropology 19(2):59–73
1983 The Archaeology of the Manokinak Site: A Study of the Cultural Transition Between Late Norton Tradition and Historic Eskimo. PhD dissertation, Dept. of Anthropology, Washington State University.
Shaw, Robert D. and Charles E. Holmes
1982 The Norton Interaction Sphere: An Orientation. Arctic Anthropology 19(2):3–10
VanStone, James W.
1989 Nunivak Island Eskimo (Yuit) Technology and Material Culture. Fieldiana/Anthropology n.s. 12. Chicago: Field Museum of Natural History