C. Melvin Aikens
Hunting-gathering societies which flourished during post-glacial times in Japan, eastern and western North America, and northern, Europe display important similarities in economic and social patterns, but somewhat different growth trajectories. The present paper compares these developments to elucidate some of the simple basic factors that structured them. The growth of archaeological data around the world offers increasing opportunity for comparative studies of prehistoric cultural continua that can expand our understanding of the basic factors shaping sociocultural evolution over long periods of time.
In this account, we restrict our attention to the Northwest and New England coasts of North America, the Baltic Sea region of northern Europe, and Japan from the Kanto (Greater Tokyo) area northeastward into the Tohoku region (Fig. 1). These are all north temperate lands with much-indented coastlines backed by wooded hinterlands. Because of the vast distances separating each of the areas under discussion, as well as disparities in the timing of cultural developments, we feel safe in postulating that the parallels noted are primarily the result of convergent evolution, rather than common heritage or cultural contact (Aikens and Dumond, 1986). The prehistoric folk of these four regions converged on certain generic organizational features, we believe, because of the nature of hunting gathering economics and production in general, and the biotic similarity of the environments they exploited in particular.
We begin from the thesis that these hunter-gatherers all fall into what Sahlins (1972) has termed the domestic mode of production, here referred to in a hunter-gatherer guise as the collector mode of production. The domestic mode has these important features: 1) It is characterized by continual under-production. "The main run of [these societies] (do) not … realize their own economic capacities. Labor power is under used, technological means are not fully engaged, natural resources are left untapped" (Sahlins, 1972:41); 2) The domestic mode of production is "an economy for use, for the livelihood of the producers" (Sahlins, 1972: 69). They produce what they need at a level to meet only those needs; 3) Increasing production or increasing per capita productivity, that is intensi fication, results immediately from social and/or political action, rather than directly from technological change alone.
We suggest that the socioeconomic similarities between the widely separated cultures that are the focus of attention in this paper arise firstly and fundamentally from the need for hunter-gatherers in all of these regions to depend upon storage to some degree. This need is enforced simply by the natural seasonality in biotic productivity that dominates middle and higher latitudes. The natural rhythm requires that at least some resources must be intensively exploited in order to produce stores, and it controls not only the timing of group activity, but also the stability of social aggregations. Groups operating in these regions must be "collectors" sensu Binford (1980). Those discussed here all were characterized-at least in their mature stages-by central residential camps from which specialized task groups exploited particular resources. As Binford points out, this is a subsistence strategy for dealing with resources which are disjunct in space and/or in time. It is therefore a strategy common among hunter-gatherers in middle and northern latitudes. Conceivably, other factors besides global climatic patterns may produce disjunct resource distributions, but it is climate and seasonality-combined with the specific similarities of the regional environments involved-that are basic in the cases to be discussed here.
The second fundamental factor giving rise to sociocultural similarity between the regions compared is simply that all four of them share highly similar kinds of exploitable resources, which are obtainable in highly similar settings. Those resources include medium-sized to large terrestrial herbivores, marine mammals, resident and anadromous fish, a variety of terrestrial plant foods, and the resources of intertidal zones. As will be shown, the general similarity in mode of production shared by the different regions, when combined with specific similarities in the biotic resource base, determined certain fundamental similarities in the life of all of them.
The region is geographically complex, consisting of the continental shelf and coastal strip, the outer mountains, the coastal trough, and the Cascades/Coast Range mountain system (Fig. 2). The two largest and most productive river systems are the Fraser in the north, and the Columbia in the south. While the outer mountains and western slopes of the coast ranges are drained by thousands of short rivers and streams, only these two and a handful of lesser rivers drain significant interior areas. Common large land mammals include deer, elk, bighorn sheep, black bear, and grizzly bear, with moose and caribou to the north. The coastal waters are rich in whales, dolphins, killer whales, and sea lions. They are also home to a great array of fish, among them the five apecies of Pacific salmon, steelhead, and olachen, all anadromoua in habit. Other economically important fish include herring, cod, halibut, toad fish, and dog fish. The intertidal zones are rich in bivalves, clams, cockles, whelks, barnacles, and limpets. Sea urchins and other intertidal creatures were also exploited aboriginally.
Available data suggest differences in economy, technology, mode of production, and social organization along the Northwest Coast before and after about 5000 years ago. These differences will be briefly summarized from the archaeological record; overall, the evidence indicates a trend toward increasing logistic organization, with the logistically structured, collector mode of production developed to a high level by 2500-2000 B.P. Increasing ritual elaboration and expansion of the web of social control were integral to this development.
Subsistence, Technology, Settlement, and Mode of Production
Before 5000 B.P.
Much of the coast was deglaciated by 10,000 years ago, and the earliest well-dated coastal sites are consistently about 9,000 to 10,000 years old (Fladmark, 1975, 1982). Among these are Ground Hog Bay (Ackerman, 1973) and Hidden Falls (Davis, 1980) in southeastern Alaska, and Namu, on the central British Columbia coast (Carlson, 1983). The Dalles locality (Cressman et al., 1960) at the upper reach of tidewater on the Columbia River, has an early occupation dating between 9800 and 9100 years ago. After about 9000 B.P., and particularly after 8000 B.P., the number of dated sites on the Northwest Coast increases dramatically.
The distribution of early sites along the coast is good presumptive evidence for the use of watercraft, but there is, as yet, no direct evidence that the early people had a strongly marine-oriented economy. Faunal assemblages postdating 8000 B.P. show that the coast's inhabitants relied as much, if not more, upon terrestrial resources as upon littoral or pelagic species. Microblade tools comparable to those of the Anangula site in the Aleutians have led some to envision an Aleut-like adaptation to littoral and marine environments.
Farther south, the cultures of the Vancouver Island/Puget Sound vicinity seem to have shared a generalist, or fine-grained, economic adaptation. Deer and elk are the most common modern wild herbivores throughout this area, and their remains are the most common in archaeological sites. Seals occur at The Dalles (Cressman et al., 1960) and Glenrose Cannery (Matson, 1976), along with elk and deer. At the critical Glenrose Cannery site, salmon were recovered from a component dating 6000 to 8000 B.P. ; at The Dalles, the early component contains an enormous number of salmon vertebrae, indicating an apparent early intensification of salmon fishing there. The Milliken site in the Fraser River Canyon has an early major occupation which Borden (1975) interprets as evidence for massive exploitation of salmon, though no actual bones were recovered. Mussels were collected along the coast, but not in quantities great enough to produce shell middens. Evidence for plant collecting is very sparse, and no reported artifacts from the coast are obvious plant processing tools.
No direct evidence for food storage exists at any of the sites so far discovered. There are no known sites analogous to the winter villages of later times, which comprised popu lation aggregates sufficiently large to require extensive stores for survival. This implies that logistical organization-the development of a network of task-specific supportive work camps (Binford, 1980)-though perhaps indicated at sites such as The Dalles and Glenrose Cannery, was much less developed than it later became. Thompson (1978), in her study of prehistoric settlement patterns in the Gulf of Georgia, was able to document only one kind of settlement for this early period, a generalized salt water/fresh water type. Before 5000 B.P., population seems to have been everywhere sparse, though gradually in creasing with time.
It is thus difficult to characterize the early mode of production for this region concretely, given the paucity of archaeological information. The following speculative statements, however, seem plausible:
1) The basic resources which supported the ethnographic economy appear to have been exploited even in remote prehistory in this region.
2) It is reasonable to believe that these people depended on storage at least to some extent, given the seasonal rhythm of biotic productivity natural to the area. However, it is conceivable that small groups of hunter-gatherers operating along the Northwest Coast prior to 5000 B.P. may not have depended as heavily upon storage as did their descendants. Early populations were no doubt small, and this is the period of the postglacial climatic optimum. Conditions may not have been as difficult as they became later in the Holocene and small, highly mobile groups of hunter-gatherers may have been able to operate without a heavy dependence on stored foods during the winter. Schalk (1982) has suggested that the important changes along the coast after 5000 B.P. resulted from technological inno vations which made storage of large amounts of fish more feasible. Simple population growth and societal elaboration may also have made storage more necessary.
3) Settlement pattern data are as yet too meager to demonstrate how logistically organized these peoples might have been. The Anangula Aleut model certainly suggests a very strong logistic collector mode of production, although Thompson's (1978) study is tanta lizing evidence pointing in the other direction. She finds no evidence for specialized sites in Puget Sound before 1500 B.P., but since hers is a single study in a limited portion of a vast stretch of coast, little can at present be built upon it. It is fair to say that for the entire Northwest Coast there is no definitive evidence for specialized collecting sites during this period, although one might claim that the few known microblade sites indicate the existence of specialized task groups. In this case, one must also argue that microblades represent some kind of specialized tool kit. Recent work by Loy (1983) may ultimately resolve the issue of microblade function, but at present, it seems likely that microblades armed a wide variety of compound tools (cf. Clarke, 1976).
After 5000 B.P.
After about 5000 B.P., shell midden sites appear all along the Northwest Coast, reflecting intensified gathering of intertidal molluscan species. An increasing diversity of species generally appears in the faunal records, and there is artifactual as well as faunal evidence for the intensification of fishing. Fishing tackle continues to increase in abundance and diversity as time passes. At Glenrose Cannery, there is evidence of an increase in the rela tive importance of salmon after 5000 B.P. and at sites in Prince Rupert Harbor, salmon constitute almost half of the identifiable fish remains. These sites are located adjacent to major salmon rivers; in other areas, nonanadromous fiah were important, such as herring, rock fish, and toad fish (Calvert, 1980). Reports of prehistoric tuna fishing on Vancouver Island indicate the range of piscine species and marine habitats now being exploited (McMillan, 1979).
Evidence for storage is indirect, but suggestive. In the Prince Rupert middens, great numbers of small post holes may represent supports for fish-drying racks. Thin, well-made ground slate knives appear in southwestern British Columbia sites shortly after 3000 B.P. These are taken by Mitchell (1971) to indicate the mass processing of salmon for storage, although Burley (1980) offers alternative explanations. More broadly, evidence for popu lation growth and sedentism may reasonably be taken to suggest extensive storage as a precondition for such developments (cf. Schalk, 1977).
A settlement pattern study for the Puget Sound/Gulf of Georgia region (Thompson, 1978) demonstrates an increasing frequency of specialized sites there after 5000 B.P. Prior to that time, one generalized site type was recognized; after that time, the number increased, reaching a total of seven types after about 1200 B.P. By 2500 to 2000 B.P., it becomes possible to distinguish village sites from non-village sites, and in the village sites to infer the presence of the large multifamily plank house type known from the historic period. In Prince Rupert Harbor, the contrast between village sites and special purpose sites is extreme in terms of material culture, site features, volume of midden, and so forth (MacDonald and Inglis, 1980-81; Ames, 1976, 1981). All this evidence suggests the increasingly partitioned envi ronment and increasingly specialized material culture that one expects with a well-developed logistical strategy.
Thus, while clear and direct evidence for storage is proving hard to find, the large middens and other indicators of relatively dense populations are good indirect evidence for storage. There is good though fragmentary evidence for logistical organization all along the coast, and the logistically organized collector mode of production becomes in creasingly pronounced over the last 5000 years. In the final analysis it is simply incon ceivable that the cultural systems existing along the Northwest Coast during this period could have developed without heavy reliance upon stored foods.
Midway through the period, large domestic structures of ethnographic type appear, accompanied by indications of social ranking. Thus by that time the basic economic unit was probably the extended family or lineage organized along hierarchical lines. Evidence for logistical organization is stronger after these developments than before them; the logistically organized collector mode of production, unmistakably present soon after 5000 B.P., was established in its mature, highly productive, ethnographic form by about 2500-2000 B.P., and in this form was clearly a function of a highly structured society.
Before 5000 B.P.
Archaeological evidence currently available is insufficient to reconstruct early Northwest Coast social organization. The nature of the domestic group and of larger groups remains to be documented. No structures predating 5000 B.P. are known anywhere along this stretch of the North American coastline, although Anangula (Aigner, 1976) in the Aleutian Islands does contain large pithouses that arc earlier than this. Anangula is usually interpreted as representing a cultural system like that of the historic Pacific Eskimos and Aleuts. The presence of microblades at sites along the coasts of Alaska and British Columbia similar to the microblades at Anangula has sometimes been taken as evidence for similarly orga nized domestic groups along these coasts as well. This, however, is conjecture.
After 5000 B.P.
The intensification of logistic collector strategies on the Northwest Coast was accom panied by obvious developments in social organization, as measured by elaboration in material culture and burial practices. In Prince Rupert Harbor, for the period 5000-3500 B.P., the recovered material culture documents mainly subsistence-oriented tasks. Between 3500 and 1500 B.P., subsistence-related artifacts increased in variety, and decorated tools appeared. Items of personal adornment including labrets (an ethnographic status marker), pendants, beads of shell, amber, and copper, and bracelets became common (Borden, 1983; MacDonald, 1983; Stryd, 1983). Trade items also appeared. This increasing richness of material culture accompanied an elaboration of burial practices. It was during this period that the Rupert middens began to accumulate rapidly, and evidence for large multifamily plank houses first appeared (MacDonald and Inglis, 1980-81).
By 1500 B.P., the classic Northwest Coast pattern was fully developed in the Prince Rupert region:
Massive and elaborate pecked and ground stone artifacts occur. Zoomorphic art flour ishes, being applied to a wide range of bone and stone tools and personal objects such as combs and pins. Features from different areas of the site (GbTo: 313), such as house size and associated materials, reflect ranked village structure (MacDonald and Inglis, 1980-81: 52).
The complex alliance among art, mythology, and socialization ethnographically described by Boas (1890) and others was fully established by at least 1500 years ago, and there has been essential stability and continuity in material culture and society during the last millen nium and a half.
Farther south, evidence of social elaboration appeared somewhat later than in Prince Rupert Harbor, during the Marpole phase of ca. 2800-1200 B.P. The trends were similar however; material culture became more elaborate, a distinctive zoomorphic art appeared, burials provide evidence of social differentiation, and large multifamily plank houses appeared (Mitchell, 1971; Burley, 1980). After about 1200 B.P., with social ranking already clearly achieved, there was an increase in the complexity of settlement patterns in the Gulf of Georgia region. Seven different settlement types were in use after that date, con trasting with only two types for the preceding period. This contrast apparently reflects a late intensification of logistic resource procurement, perhaps fostered by a more complex and centralized social organization (cf. Ames, 1981, 1983).
New England and Baltic
Maritime adaptations of hunter-gatherers are common in the North Atlantic. Two North Atlantic areas, the littoral of the Gulf of Maine in North America, and a southern portion of the Baltic Sea in Europe, witnessed quite comparable adaptations to marine environments. Specifically, we compare the central coast of the State of Maine, USA, with Denmark (Figs. 3 and 4). The cultures are the Late Archaic and Ceramic periods of Maine, dating between 4400 B.P. and 400 B.P., and the Ertebølle culture of the island of Zealand, dating between 6600 B.P. and 5300 B.P. In both areas, the marine zones were backed up by a variety of nearby terrestrial environments, capable of providing seasonal or even year-round support when necessary. Both areas produced cultures with high population densities and relatively permanent settlements. In many parts of the world, these characteristics appear as forerunners, some would say prerequisites, to food production in the form of agriculture. That they did in one of our North Atlantic examples and not in the other is of some interest.
The postglacial ecological history of Denmark is well known and traditionally divided into a series of time units based on pollen analytic data. The Ertebølle culture falls in the Late Atlantic chrono-unit with the Littorina Sea transgressions (Brinch Petersen, 1973). It was the time of the thermal maximum, a period when oak, elm, and especially lime dominated the deciduous forests. Major game animals included the red deer, roe deer, and pig. Plant foods such as hazel nuts were abundant. The littoral of Zealand was con voluted and full of estuaries and other shallow, highly productive environments for fish, shellfish, sea mammals, and aquatic birds. There were off-shore islands useful for specific resources, and a rich fishery, with many anadromous and catadromous species (Paludan-Müller, 1979).
The ecological history of the Gulf of Maine is much less well documented (Davis and others, 1975). The central Maine coast during late Ceramic times experienced the increased cold of the Neoglaciation. This is reflected in the spruce-fir forest of the coastal strip, which reduced the vegetable food potential of the terrestrial zone and disadvantaged the white tail deer that depend on deciduous forests. The marine ecosystem was highly productive, however. Many attractive environments occurred along the highly convoluted coast of estuaries, bays, and islands, all within easy reach of Indians with birch bark canoes. Anadromous and catadromous fish utilized the many rivers and creeks in their spawning cycles.
A brief comparison between the two areas suggests that in Zealand the woodland vegeta tive resources were definitely superior during Ertebølle times. It is more difficult to esti mate the relative productivity of the marine resources; however, marine transgressions in the Baltic probably resulted in an enhanced littoral environment (Paludan-Müller, 1979: 125). The history of the Gulf of Maine indicates a potential for maximum productivity during the late Ceramic (Sanger, 1975, 1983), when high tidal ranges produced extensive water column mixing, nutrient recycling, and great plankton growth (Apollonio, 1979).
Subsistence, Technology, Settlement, and Mode of Production
The earliest documented maritime-adapted cultures in Maine are only about 5200 years old (Bourque, 1976). Although earlier cultures may have existed, a subsiding coastline has drowned any older sites. How many older sites may have existed is unknown; however, simulation models of marine biological activity predict that lowered sea-levels would result in reduced carrying capacity due to lesser tidal mixing of nutrients (Sanger, 1975, 1983), so it may be that littoral exploitation was rather limited in earlier times.
The Late Archaic Moorehead phase of 4400 B.P. to 3800 B.P., especially as documented at the Turner Farm site, features a maritime adaptation associated with a heavy emphasis on large terrestrial animals (Spiess and others, 1983). Sites of this phase were located on the mainland and on large islands with mainland-like habitats. There is some evidence for the use of shellfish at this time, and utilization increases steadily over the next few millennia. At the same time, inland areas show signs of being less intensively exploited. Employing varied seasonality indicators a case can be made for an intensification on marine resources, through an increasingly long period of time being spent each year in the littoral zone.
As the Ceramic period began in Maine about 2500 B.P., there were numerous large clam-shell midden sites located in a variety of habitat areas (Sanger, 1979, 1982). Analysis of seasonality points to some year-round occupation in the coastal environment. The outermost islands, some involving a 16 km paddle across open water, were probably used on a more restricted basis, mostly in the summer. Observable site density is high despite the erosion affecting the archaeological record of the Maine coast. For example, on thecentral Maine coast, in one recently surveyed area with a radius of 20 km, some 400 shell midden sites-most with Ceramic period components-have been documented. Site size is variable from a few square meters to 10,000 square meters, while still larger are the Damariscotta River shell heaps (Snow, 1972), a special group of American oyster middens. Land mammals occur in all sites; however, the sites themselves are situated to accom modate the requirements and needs of living on the seashore (Kellogg, 1982; Sanger, 1982).
The variety of species represented in the shell middens attests to the generalist nature of the exploitation pattern. Fish, shellfish, sea mammals, land mammals, and birds occur in varying percentages depending on age, location, and season of the sites (Sanger, 1982). Little has been preserved in the way of technology for exploiting the marine environment. Barbed, non-toggling harpoons and fixed barbed points occur, as do small pointed bone pieces that might have barbed fish hooks. Chipped stone projectile points are common. Much of the exploitation of marine habitats can involve simple technology that may not be preserved. There is no evidence for food preservation and storage, such as well defined storage pits, despite the natural low productivity in the spring which implies the need for a logistic collecting strategy. There is some evidence, based on the study of shell growth, to suggest that shellfish gathering was greatest in the spring of the year when other food resources are scarce or are of lesser quality.
Prehistoric cultural systems of the central Maine coast remained at the hunter-gatherer level until contact by Europeans in the early 17th century A.D. Effective maize agriculture came late to much of northeastern North America. In fact, in those regions to the south and west of Maine, there is scant evidence for agriculture before 1000 B.P. Perhaps agriculture could never have spread to the central Maine coast due to the limited number of frost-free days, especially with the onset of the Neoglaciation. Alternately, the diffuse subsistence system based on the rich marine and terrestrial environments may have dampened enthu siasm for a maize growing pattern that would have been marginally successful at best.
The Ertebølle culture of Denmark is the last in a developmental series of Mesolithic cultures (Maglemose, Kongemose, and Ertebølle), and in many ways is the climax of the culture type (for a detailed listing of artifacts, subsistence, and sites associated with these cultures see the descriptions by Brinch Petersen, 1973 and Clark, 1975). The Ertebølle occurs in the Atlantic chrono-unit, a time of great marine and terrestrial productivity, and immediately precedes the Neolithic which begins around 5300 B.P. According to Brinch Petersen (1973: 98), there is a marked increase in the number of sites in the later phases of the culture.
Analysis of site locations indicates both coastal and interior occupation, although as Price (1981) reminds us, no place in Denmark is more than 50km from the sea. In Zealand, early Ertebølle sites tend to concentrate in the estuaries where some year-round settle ment is suggested. In the latter phases of the culture (Dryholm II stage) there is more of a tendency for resource areas of lesser carrying capacity, such as the outer islands and the riverine habitats, to be used with increasing frequency. This trend, according to Paludan-Müller (1979: 153, 4), is a reflection of population pressure, a direct result of the favorable High Atlantic ecological situation.
In these conditions, the last major resource area to be systematically exploited was the climax deciduous forest. Increased familiarity with and dependence upon the inland forested areas set the stage for the later acceptance of animal husbandly and agriculture. Unlike Maine, there are no obvious environmental restrictions. Although hypotheses based on demographic models are notoriously difficult to prove in prehistoric archaeology, it does seem that densely populated hunting and gathering societies, accustomed to exploiting vegetable resources, are especially receptive to innovation in the area of food production.
The marked increase in the number of sites noted by Brinch Petersen (1973) in the Dryholm II stage of late Ertebølle may well be functionally related to the increasing popu lation pressure hypothesized by Paludan-Müller (1979) for Zealand. The apparent increase in population and the greater use of less productive resource areas may have resulted in (or reflected) a kind of resource partitioning that imposed a greater degree of logistic or organizational structure on the Ertebølle culture.
Relatively little is known of the social organization of the prehistoric societies that lived on the central Maine coast. Drastic depopulation during the first decades of the 17th century undoubtedly altered the pre-contact social organization. Although ethnohistorians have argued the case for a tribal level of complexity above the band level of organization, the data are equivocal. Burials, a potential source of social information, are few during the Ceramic period, and so far uninformative with regard to social organization. The most elaborate burial ceremonialism actually occurs in the Late Archaic and red ochre inhuma tions of the Moorehead burial tradition (3800 B.P.-4400 B.P.) (Sanger, 1973), but the societal implications are unclear.
Mouses may provide some clues as to domestic group size. Those described on the Maine coast are small, usually less than 5 m by 6 m, and would seem appropriate to nuclear families or moderately extended families (Sanger, 1979).
Price (1981) has commented on the dearth of information on social aspects of the Ertebølle culture despite the recent influx of data on technology, subsistence, and settle ment pattern. The settlement data do not, unfortunately, include dwellings. The same author does, however, conclude that the evidence "argues for increasing social and ritual complexity" (Price, 1981: 7).
The evidence referred to is in the form of human burials recovered from a disturbed area at Vedbaek, 30 km north of Copenhagen (Albrethsen and Brinch Petersen, 1977). Only 17 graves, representing 22 individuals, were recovered, although the number was greater prior to disturbance. Red ochre-covered, extended, supine inhumations, most accompanied by mortuary offerings in single graves, are the rule. The site dates to the early stages of the Ertebølle culture (Dryholm I stage), or around 6000 B.P. Analysis of the C-13 content of the bones suggests a heavily marine diet. As the authors point out, the small number of individuals and the disturbed nature of the site caution against placing too much weight on the grave lot associations. There are very few other burials of this period. Despite the small numbers, there does seem to be a pattern of male burials with items normally thought of as used in male economic activities. That entire tool kits are represented would appear to be ruled out by the scarcity of implements with each individual. At this time, there would not appear to be adequate data from these few burials to really support the claim, for an increase in social complexity, if for no other reason than that we have so little by way of comparison.
On the other hand, simply from the overall growth displayed in the Maglemose-Konge-mose-Ertebølle continuum, it seems reasonable to allow that Ertebølle represents some sort of social climax for the Mesolithic culture type in the Baltic region. Rowley-Conwy (1983) argues, on the basis of biomass data, that the Ertebølle people were most probably sedentary complex hunters of the type represented by the ethnographic Ainu, Northwest Coast, and Eskimo peoples. His argument stresses ecological potentials more than archae-ologically demonstrated facts. Nevertheless, though much work remains to be done on the problem, his analysis is highly suggestive.
Any more detailed attempt to relate social organization, demography, and maritime adaptations in either Maine or Zealand must face the fact that the data are still inadequate. Confronted with the lack of the kind of ethnographic base available for the Northwest Coast of North America, archaeologists dealing with the above areas will have to depend on the development of other data to reconstruct the social organizational principles.
Central and Northern Japan
Central and northern Japan, with its greatly indented coasts and mountainous hinter lands, is biotically rich and diverse (Fig. 5). Many marine fishes, seals, and small whales were to be found in the bays and adjacent open seas. Vast tidal flats produced clams, oysters, and mussels of various species, and provided a habitat for water birds. Salmon ran up most of the rivers of northern Japan, and many species of estuarine and fresh-water fish were available as well (Akazawa, 1980, n.d.). The woodlands harbored bear, deer, boar, rabbit, badger, ermine, fox, and other mammalian species. The Japanese flora is rich in edibles suited for human consumption, including acorns, walnuts, chestnuts, and seeds, fruits, roots, and greens of many kinds. Even in late historical times, the gathering of wild foods in quantity was of major economic importance in the mountains of central Japan (Koyama, 1981; Matsuyama, 1981; Akimichi, 1981).
The Japanese evidence shows clearly that a logistically organized collector mode of production was achieved there early in postglacial times. Based on this economic adapta tion, a high point of cultural development was reached in the Jomon tradition by shortly after 5000 B.P. Increasing elaboration of ritual paraphernalia suggests the concomitant growth of mechanisms for social and economic management (Aikens, 1981; Aikens and Higuchi, 1982).
Subsistence, Technology, Settlement, and Mode of Production
Before 5000 B.P.
At the shell mound of Natsushima on a small island in Tokyo Bay, the Japanese Jomon culture was present in a basically mature form by the latter part of the Initial Jomon period, dated between 9500 and 6000 B.P. Littoral and marine exploitation are indicated by the remains of 34 species of shellfish, 17 species of marine fishes, and seven species of birds. The dolphin was also exploited. Land hunting on the adjacent shore was indicated by the bones of 10 species of mammals, including boar, hare, dog, and the Japanese raccoon-dog. Flaked stone arrowpoints and bone points and fishhooks give evidence of the means by which these species were taken. At other sites, in both the Tokyo area and the region around Sendai to the northeast, such fishing gear as net floats of bark or pumice, girdled and notched stone net weights, and bone or antler harpoon points and leister prongs provide additional evidence (Akazawa, 1980, 1982). Grinding slabs and mortars from Natsushima indicate the processing of seeds and roots, and many roughly flaked tools resembling axe or adze-heads may have served to tip hoes or digging sticks. Pottery, mainly the sherds of cord-marked, pointed-bottom cooking vessels, was abundant and characteristic of the latter part of the Initial Jomon period generally (Sugihara and Serizawa, 1957).
The situation of Jomon shell mound sites along the shore and on small offshore islands demonstrates the early use of watercraft. In fact the occurrence, in Paleolithic sites of the Tokyo region, of obsidian from Kozu-jima Island in the Pacific south of Tokyo Bay shows that watercraft capable of quite formidable journeying were in use long before Jomon times (Suzuki, 1974). At Kamo, a Jomon site across the bay from Natsushima, a large fragment of a dugout canoe is dated at 5100 B.P.; at Torihama, on the Japan Sea coast, a nearly complete canoe and paddles similarly predate 5000 B.P. (Matsumoto et al., 1952; Morikawa, 1976).
Fairly large and substantial semisubterranean houses have been identified at the Hanawa-dai site, east of Tokyo Bay, at a period roughly contemporary with the major occupa tion at Natsushima. Another very important site containing Initial Jomon houses is Hakeue, in western Tokyo. Here, numerous probable storage pits around and between the houses contained pottery sherds of the same types as found in the dwellings, clearly establishing the contemporaneity of houses and pits. House structures of the Initial Jomon period have been found in other parts of Japan as well, as far south as Kyushu, indicating that stable residential patterns were established early in the postglacial period (Tsuboi, 1971; Kidder, 1983; Yawata and Kagawa, 1955).
Food storage is strongly suggested by the presence of numerous associated pits in these early sites, and food remains of many kinds, both plant and animal, are increasingly being directly recovered through the use of new analytical approaches at sites throughout Japan (Kato, Sahara, and Koike, 1984; Watanabe, 1984). Indeed there seems little doubt that food storage was widely practiced in Japan from early Jomon times onward.
Thus, as the above review shows, the initial establishment of a logistically organized collector mode of production is clearly indicated by the evidence of Natsushima and other Initial Jomon sites. The wide range of animal species, the presence of shell middens, and the occurrence of houses with associated storage pits, are all indicative. The typical distri bution of landscape types in central and northeastern Japan, with bays, inlets, tidal flats, estuaries, alluvial plains, and wooded hills all in close juxtaposition-especially in the Tokyo and Sendai regions-undoubtedly fostered a logistic form of social organization as the most rational and economic approach to subsistence and settlement. A strategically located residential site had all the necessaries sufficiently nearby as to obviate incentives for any wider-ranging form of hunting-gathering involving residential rather than logistic mobility.
After 5000 B.P.
Subsistence and technology manifested strong continuity throughout the Jomon age. The same species were exploited, and the same technology utilized in this exploitation, both before and after 5000 B.P. During Middle and Late Jomon times, shell middens con tinued as the major site type along the coasts, and where conditions of preservation are favorable, such sites continue to yield the bones of boar, deer, small mammals, birds, sea mammals, and fish of many kinds. Where exceptional conditions of preservation obtain, nut shells, acorn hulls, and other plant remains give evidence of the products which were processed on the grinding tools found everywhere in sites dating both before and after 5000 B.P. Some very large, waterlogged storage pits filled with acorns are known from southwestern Japan, and elsewhere the variety of storage pit contents is becoming better known through plant macrofossil analysis (Makabe, 1979; Otomasu, 1984).
The same continuity is seen in functional artifact assemblages. There was certainly an expansion in the quantity and variety of ceramic artifacts during and after Middle Jomon, but this was related to changes in the social realm, as will be seen. The production tools of Middle Jomon-arrowpoints, knives, scrapers, fish-spears, hooks, sinker stones, leisters, harpoons, canoes, mortars, pestles, milling stones, nutstones-were little if at all 'different from their earlier counterparts; no new tool classes were added to the cultural inventory which would change the character of the long-established subsistence technology.
Apparently fostered by this economic stability and its concomitant sedentism, there was over time a gradual expansion of the human population, which reached a peak during the Middle Jomon, about 4000-5000 B.P. Figures compiled by Koyama (1978: Table 3) show that the number of archaeological sites known for Initial Jomon throughout Japan is almost doubled for the succeeding Early Jomon, and that figure is doubled again for the Middle Jomon period, reaching a total of nearly 11,000 locations. Thereafter, population remained relatively stable in northeastern Japan until the end of the Jomon age, although elsewhere Jomon sites became progressively fewer, particularly toward the south. This last was no doubt significantly related to the northward spread from Kyushu of the rice-producing Yayoi culture, and relates to a set of questions not at issue in the present discussion. At this point, the fact of interest is the persistent growth of population, which culminated shortly after 5000 B.P. in quite dense swarms of Jomon sites clustered in particularly favorable settings on bays and inlets and in the mountains of central and northeastern Japan.
Patterns of human settlement are notably more differentiated after 5000 B.P. Hayashi (1983) recognizes three broad categories of sites for the Kanto-Tohoku region: 1) large settlements with manifold activity loci, including residential pithouses, workshop areas, storage pits, ritual centers, cemeteries, and refuse disposal areas; 2) smaller sites with one or two pithouses, lacking evidence for a wide range of activities; and 3) very small or sparse sites which lack pithouses and yield only light scatterings of pottery sherds and flakes. Site distribution, and the fact that large settlements are few while smaller sites are many, suggests the existence of a unified "village system," or settlement unit, in which a few core settlements are surrounded by more numerous satellite or daughter hamlets. The smallest (and most numerous) sites, lacking evidence of habitations, are seen as exploitative stations linked to the residential occupations.
As noted earlier, well before 5000 B.P. the collector mode of production was established in Japan. The review just concluded shows that the species which formed the basis of sub sistence, and the technology which served in obtaining and processing these species, did not change after 5000 B.P. Nevertheless, there was a significant intensification of economic production in Middle Jomon times; not only were much larger populations being supported then, but they were being supported at a higher level. Collection intensified in tandem with the increasing complexity of Jomon society.
Before 5000 B.P.
The Initial Jomon settlements where house remains have been found are all small, and the number of Initial Jomon sites is relatively few as well. At Hanawadai, the remains of two houses were identified, and at Hakeue there were 11, composing two groups separated by about 60 meters. At Sozudai, in Kyushu, only one well-defined house outline could be traced, but a welter of pits and postholes at the site suggest the repeated rebuilding of structures there. The evidence thus suggests that the earlier Jomon settlements were small, limited to a few families, but stable. Logistic organization is indicated, as noted above, although the extent to which ephemeral activity stations may have been associated with the residential sites remains unclear.
After 5000 B.P.
An increasingly complex form of domestic organization is indicated by Middle and later Jomon settlement patterns. According to Hayashi, settlement units of the kinds described for the Kanto-Tohoku region occupied a fairly discrete territory, such as a river valley or a large inlet, and reflected an extended form of household organization:
Analysis of the distribution of settlement facilities, and constitution of the cemetery, allows us to infer that a settlement is neither an aggregate nor a complex of private or personal facilities, but rather a complex of such facilities prorated to each household, Thus a "territorial unit" in conventional definition is ascribed to a "household community" in Marxian terms, in which the individual household has secured its position as a subject of usufructury, although its claim over both property and production is still to be submitted to the community as a whole (Hayashi, 1983: 8).
The growing maturity and complexity of Jomon society is reflected in both population growth trends and in a striking elaboration of certain classes of artifacts related to display and social status. Specialized ceramic items and jewelry became abundant in Middle and Late Jomon times, giving evidence of the increasingly differentiated social structure that managed and manned the productive organizations.
By 5000 B.P. or shortly after, elaborate sculptured pottery was being made in quantity. Some of these vessels have such ornate-even fantastic-sculptured rims that the pots are impractically top-heavy, and obviously something more than expressions of a mere urge to beautify objects of daily use. Incense burners made for suspension, pottery drums, hollow pottery figurines both anthropomorphic and zoomorphic, engraved ceramic plaques, car-spools, and fired clay magatama beads are some of the notable ceramic elements which indicate the growing importance of display and the emergence of increasingly marked social distinctions in later Jomon Japan. Other elements such as polished stone beads and carved bone hairpins enhance this impression, while a proliferation of ground and polished phallic stones and elaborately detailed female figurines suggests the nature of cult observ ances that probably served as the theater in which unequal social relationships were dis played and reinforced.
In short, there is unmistakable evidence for the emergence of social elites having impor tant roles in the organization of group activity. It is important to note that artifacts indicating social elaboration have early beginnings in the Jomon tradition, but seem to have reached a florescent stage during the Middle Jomon period, at the same time that the archaeological record suggests the establishment of a more numerous and more differen tiated population. The fact that the very peak of Jomon population in the Tokyo area (in late Jomon times) followed directly upon the Middle Jomon strengthening of status divisions within society suggests that more effective management of economic activities, enhancing the general prosperity, was one important outcome of the increasing social stratification.
Comparisons and Interpretations
In this paper, we set out to describe some basic parallels in the background and develop ment of affluent hunter-gatherer societies in Japan and on the northern coasts of North America and Europe. The concept of a collector mode of production was introduced at the outset to focus attention on our thesis that these societies operated on the basis of funda mentally household economies, where increases in productivity required social or political action. We also argued that many of the economic similarities and therefore developmental similarities among these regions arose from common structural features in their adaptations and environments. We specifically listed a common dependence upon stored foods, a common collector strategy, and finally, very similar sets of resources.
In our narrative we observed that:
1) All the societies mentioned existed in north temperate coastal/woodland environ ments of similar structure, with high species diversity and high biotic productivity. As a matter of fact, despite the great distances separating them, much of their flora and fauna is shared at the generic level; all four regions fall within the biogeographer's Holarctic province. In all of them the economic security provided by high biotic diversity, and the prosperity underpinned by general biotic richness, fostered continuous long-term popu lation growth.
2) All four cultural traditions exhibited long, gradual developments culminating in large sedentary or nearly sedentary populations of collectors of varying degrees of social elaboration, who lived clustered in substantial communities surrounded by smaller satel lites of differentiated function.
Within the bounds of these broadly parallel developments, there are also some contrasts to which we attach importance:
1) The Japanese and north European societies reached plateaus of social complexity significantly earlier than did their North American counterparts. The Japanese Jomon tradition was most precocious, reaching an early plateau between 9500 and 6000 B.P., but by about 5000 B.P. both it and the European Ertebølle culture had achieved levels of richness and complexity probably not equalled on the Northwest Coast until at least 2500 B.P., and on the New England Coast even later.
2) In the sequel to the periods here discussed, the Japanese and north European societies went on to even higher levels of societal complexity, ultimately to develop feudal states and embrace fully agricultural, economies. By contrast, Northwest Coast and New England coastal societies grew little more and never adopted cultivation.
How may these similarities in pattern and differences in trajectory be accounted for? We offer four suggestions founded on the simple factors of environment and time, with population growth an important dependent variable.
1) The development of affluent hunter-gatherer societies of similar form in these four widely separated parts of the world is simply a reflection of similar ecological structure, including environment and mode of production, which shaped the most basic forms of social life (see also Rowley-Conwy, 1983).
2) Differences in the apparent tempo or trajectory of societal elaboration in the four areas are probably a function of place, time, and attendant population growth. Clearly, elaborated social structures can only exist where there are sufficiently large and dense populations to actualize and benefit from them. The relative precocity of Japan and Baltic Europe in the present frame of reference, as compared to New England and the Northwest Coast, probably began from the simple fact that in the Old World cases, long-established Paleolithic societies provided an initial population base certainly much larger than the initial base available in the New World, which humans only entered at the very end of the Pleistocene. Population growth and the related developments of post-glacial times thereby got off to a much better start in the Japanese and European cases.
The fact that Japanese development seems to have run a bit ahead of that in northern Europe is accountable in the same terms. Northern Europe had to be re-occupied after the end of the last glaciation from populations resident south of the ice, but Japan was never significantly glaciated and Paleolithic populations there were already well established by the time the Jomon tradition emerged at the beginning of the Holocene.
3) It also seems appropriate to invoke environmental circumscription (Carniero, 1970) and argue that the Kanto region of Japan, especially, was a "hot house" for social elabo-ration among its hunter-gatherer populations. Its large population was (at least relatively) densely packed within the limited and topographically circumscribed space of the Kanto plain. There, populations grew in a restricted environment; mobility was not favored, nor was fissioning and pioneering new areas a useful strategy for long. Thus the Jomon people inevitably came to rely more heavily on logistical strategies than would otherwise have been necessary. Under these conditions of high density, the environment was subdivided into bounded territories containing competing groups. Hence we see early evidence for logistic collector strategies, for formation of social hierarchies, and so on. Because the Kanto is topographically circumscribed and human mobility limited, population densities increased more rapidly there than on the British Columbia coast or elsewhere among the regions compared, "forcing" the development of complexity among these affluent collectors.
4) The fact that two of our subject areas ultimately took up agriculture, while the other two did not, may be at least in part related to the above-noted differences in time trajectory. As argued elsewhere by one of the present authors (Aikens, 1981), that con-centration of human effort on propagating and managing plant and animal foods which we term agriculture may be best seen as stimulated by social elites for (their own) social ends. Thus, the thoroughgoing adoption of agriculture demands the pre-existence of substantial populations and a strong managerial class. In Middle and Late Jomon Japan, and in the Baltic region of Ertebølle times, such were undoubtedly present. In the case of New England, a sufficient level of population and social complexity may simply not have had time to develop, given the late beginning there of a productive coastal culture pattern. For the Northwest Coast, which had the societal prerequisites, it could be argued that environ-mental limitations there precluded a focus on species that we readily recognize as agricultural-but that nevertheless the observed intensification of its coastal economy had many of the characteristics of agriculture; in short that the Northwest Coast subsistence economy was in important respects a quasi-agriculture based on the most productive and manageable native resources of the region.
We cheerfully admit that the foregoing exercise is higly speculative, and certainly too brief and simple to provide a really satisfying account of the course of events in the four world regions discussed. We have essayed it nevertheless because we believe in the potential of comparative study for identifying factors that have been significant in cultural evolution, and because we think it important to begin with basics.