reached 94 percent.87 On Molokai, soils on the leeward slopes with the correct rainfall were too old—most had formed from volcanic eruptions up to 1.4 million years earlier—and thus too mineral-poor for extensive field systems. But at Kalaupapa Peninsula, the “stone leaf” formed at the base of the windward sea cliffs by an eruption 330,000 years ago, ideal soil conditions combined with a rainfall band between nine hundred and thirteen hundred millimeters to allow a very productive ‘uala field system, begun about 1400 and encompassing at its zenith in the nineteenth century a grid of up to four thousand walls covering a combined 230 kilometers, according to partial surveys.88
Studies of the dryland field systems indicate that they, like the irrigated valleys but in an accelerated amount of time, followed the path of intensification. Likely beginning as long fallow swidden areas carved out of the mesic forest to help support small numbers of people, they gradually became more permanent and more intensive, with labor needs and population growth rising reciprocally, leading to the rapid expansion of the area under cultivation but also to increased subdivisions within the fields.89 Productivity also trended downward as nutrients were depleted, spurring farmers to shorten fallow periods and increase mulching.90 Robert Hommon summarized the probable net effect on the dryland farmers: “Increased labor requirements, the addition of field work to women’s traditional tasks, diminished productivity, rapid population growth, reduced soil fertility resulting from shortened fallow, government levies, and finally infrastructure development that reached the limits of cultivable land—probably led to an increased frequency and degree of food stress among the commoners who were dependent on rain-fed systems.”91
Even so, the inherent instability of the dryland field systems slowed neither their expansion nor the increase in overall average production as well as population. This is in contrast to the situation in the older, western islands of Kaua‘i, O‘ahu, and probably Molokai, where the population, after expanding enormously from 1400 to 1650, reached a peak in 1650 and then went into a slow but steady decline.92 Yet on Maui and Hawai‘i, population continued to grow from 1650 to contact.93 Had population in the western islands reached the limits of their environment’s carrying capacity, or were other processes at work?
Patrick Kirch outlined a theory of why population growth in Polynesia slowed, oscillated, or even reversed in many places at or about the same time: the “full land” scenario, where population pressure was felt first in a lack of access to land, not in constrained food supply, “which could often be offset through” intensification. Population densities of one hundred people per square kilometer or higher were not uncommon on many Polynesian islands, without causing famine. Instead, “cultural controls on population growth began to be implemented,” including “means of reducing fertility (celibacy, contraception, abortion), as well as of increasing mortality (infanticide, suicide voyaging, war, expulsion of certain groups, ritual sacrifice, and even cannibalism).94 The most successful of these societies self-regulated to adapt to their resource base before their diminishing resource base did the regulating for them.
The fact that agricultural intensification is, if not entirely independent of demography, at least not deterministically produced by population pressure is one of the most interesting discoveries of Pacific anthropology. Instead, it is dependent on social variables. There is a primary relationship between production and power. In anthropological theory the salient feature of chiefdoms as distinct from other simpler modes of social organization (usually described as the domestic model of production and typically organized around many equivalent-status household units producing only enough for their subsistence) is that chiefdoms are ordered by rank and impose “social production” on the households that make it up. That is to say, a small, nonproducing class of chiefs extracts a surplus of food and goods from the majority of producers and then deploys it in the service of continuing and expanding the hierarchical relationship. In a sense, the surplus is a kind a capital accumulated and spent to sustain or increase itself, the extraction of surplus. Kirch, echoing his colleague and frequent collaborator University of Chicago anthropologist Marshall Sahlins, characterized the political economy of chiefdoms as “the ceaseless extraction of surplus from individual households that otherwise might be thought of as intrinsically antisurplus.”95
In Hawai‘i, status was conferred not solely by genealogy but by wealth, which took the form of food and labor extracted from the people and of ritual, status, and durable goods, such as feather cloaks, produced by commoners and specialists for the ali‘i—goods that could in turn be deployed to obtain more status and power—a system of political economy referred to as staple finance.96 Kurashima and Kirch wrote: “The political economy of precontact Hawai‘i was based fundamentally on food surpluses.”97
There have been two classic analyses of Polynesian chiefdoms, both from the mid-twentieth century (predating the postwar expansion of field archaeology in the Pacific), attempting to explain both the broad similarities and diversity of types by dividing them into three classes. The first is Sahlins’s 1954 study, “Social Stratification in Polynesia.” In it, Sahlins identifies Type III societies as simple; small (not more than two thousand people); generally based on atolls like Tokelau; and ruled by hereditary chiefs, little removed in lifestyle from commoners, who are responsible for both religious and secular guidance. Type II societies are midsized both in terms of population and island size (like Mangaia), with at least two strata of chiefs. Type I societies are large (ten thousand or more people), with distinct and often complex social hierarchies (Tonga or Hawai‘i) counting as many as seven to eight distinct strata, with separate priestly and warrior classes. In this framework, environmental limits had a direct social manifestation: greater aggregate productivity of an island or system equaled greater social differentiation between producers and distributors—which is to say that social stratification is ecologically adaptive. Accordingly, the physical size of the island or island system, insofar as it equates with productive potential, correlates in Sahlins’s conception directly with social typology.98 Big islands have big societies; small islands have small ones.
The second classic analysis is Irving Goldman’s 1970 Ancient Polynesian Society, in which his three categories are sorted by degrees of “status rivalry” between chiefs and are called traditional, open, and stratified. A traditional society is governed by the same hereditary chief as Sahlins’s Type III; it is “conservative” and close in structure to the ancestral Polynesian model described by Kirch of the ‘qariki (hereditary leader of a common descent group).99 Open societies are typified by competition between politicomilitary claimants and are less rigid, and less religiously ordered, than the traditional type. They tend to be located on midsized islands, such as Mangaia, the Marquesas, and Easter. Stratified chiefdoms, including Hawai‘i, Tahiti, and Tonga, are rigidly divided by caste and rank and supported by elaborate religious structures yet retain some of the competitive, warlike qualities of open societies as a sort of alternate method of social advancement.
Clearly, there is overlap between the two analytical frameworks; both describe the dynamic social and productive evolution of Polynesian societies, though from different optics: one resources, the other status. These optics derive from and illustrate two fundamental outlooks in anthropology: Malthus’s focus on environment and Marx’s on political economy. Both share a basic correlation with size, though only Sahlins’s ecological approach interprets it explicitly. In Sahlins’s model, greater size equals greater productive potential and therefore greater potential or tendency toward sociopolitical stratification and, when conditions are right, toward greater intensification. Simple atoll environments sustain only simple, relatively stable social types, as the opportunities for agricultural intensification are radically limited. In midsized environments—which correspond with the list of environmentally marginal islands—progress toward stratification is derailed by the shortage of good land. When the “full land” state is reached in the valley cores, intensification extends onto more marginal lands, pushing them into the cycle of degradation—deforestation, erosion, and so on—at first increasing and then markedly decreasing the carrying capacity of the land and creating conditions of competition and strife too unstable and
