Institute for Research on World-Systems University of California, Riverside Working Paper #4 The Ancient Hawaiian World-System: Research Questions Christopher Chase-Dunn and Elena Ermolaeva Abstract: This paper is part of a project that compares world-systems in order to understand the emergence of intersocietal inequalities and their role in historical social change. We examine the small world- system of the ancient Hawaiian archipelago before that system was transformed and incorporated in to the modern world- system in the early nineteenth century. The objective of our research is to construct a social geography of local, district, island-wide and interisland interaction networks in the Hawaiian archipelago in order to understand the extent and nature of intergroup inequalities and the ways in which these were developing in the ancient Hawaiian world- system. We focus on two locales in order to compare dimensions in the Hawaiian system: the Puna District of Kauai, and intradistrict relations on the big island of Hawai'i. We seek to determine the insertion of these locales in the larger interaction networks that composed the precontact Hawaiian world-system. The Hawaiian Islands are easily understood as a whole system because they were relatively isolated from interactions with other areas in the Pacific before the arrival of Captain James Cook in 1778. This paper was presented at a session on "Earlier, smaller world-systems" held at the annual meeting of the International Studies Association, Washington, DC March 29, 1994. The comparative world-systems perspective analyzes whole networks of intersocietal interactions. Exchange (trade, etc.), intermarriage, warfare, communications, settlement systems, and similarities and differences in social structural and cultural characteristics are studied on a regional and interregional basis. Because the nature and spatial extent of different kinds of interaction networks vary, it is necessary to pick focal groups from which to begin one's study. This should be done in a way that maximizes the opportunities for understanding the systemic nature and transformational processes of the system under study. Our Hawaiian studies focus on two locales: the island of Kauai because it may have been the old core of the system -- the area where complex chiefdoms and class-stratification first emerged; and the big island of Hawaii because it was the most recently developed core region in the system and produced the largest polities that were capable of conquering other islands. Whether or not our hypotheses about Kauai and Hawai'i are correct, focussing on these locales provides two windows on the whole Hawaiian system and allows us be fairly complete in our utilization of all the available evidence. You can not do in-depth research on even a small world-system if you try to study the entire thing in great detail. The "whole thing" is not merely the largest networks of interaction, but the ways in which larger networks are woven in with more local networks. It is most feasible to study a whole system by focussing on particular locales. Of course this approach risks the possibility that one will end up with a view of the larger system that is distorted because of the choice of focal locality. Ideally one could focus intense attention on more than one strategically-chosen location. We guard against the sin of distortion by working with other researchers who are focussing on local networks on Maui and Oahu. The main objective of our research project is to carry out a case study of a chiefdom world-system in order to examine hypotheses suggested by the comparative world-systems perspective. A secondary goal is to understand the unique features of the case under study. The main focus will be on local, district and regional interaction networks and the ways that spatial inequalities operated in the Hawaiian world-system. The Comparative World-Systems Perspective The comparative world-systems perspective is a structural approach to systemic intersocietal interactions (Chase-Dunn and Hall, 1993). We employ a multicriteria notion of spatial interaction that analyzes local, regional and interregional interaction networks of all kinds -- communications, trade (both bulk goods and prestige goods), intermarriage, and political-military interaction. This structural approach allows us to abstract from spatial scale and to compare very large with very small intersocietal networks. The emphasis is on the structural nature of whole systems be they large or small. The modern world- system is global because the interaction networks of which it is composed are dense, global in extent, regularized, and the events on one continent have consequences for people on other continents fairly quickly. Earlier systems were smaller, not because there were major discontinuities in interaction networks (because all societies interact with their neighbors), but rather because the fall-off of the consequences of events in one region was more rapid and so systemic interaction occurred in smaller regional systems rather than in a single global system. One of the main concerns of the comparative world- systems approach is the structural and processual nature of intersocietal inequalities. In the modern world-system national societies are parts of a larger stratification system which is called the core/periphery hierarchy. For purposes of comparative research we have found it useful to distinguish between core/periphery differentiation and core/periphery hierarchy (Chase-Dunn and Hall,1993). Core/periphery differentiation is defined as a situation in which two different societies with different degrees of complexity, population density and/or internal stratification are interacting with one another in a systemic way. A core/periphery hierarchy exists when one society is exploiting or dominating another society. These two kinds of core/periphery relations are often found together (e.g. in the modern world- system) but one can exist without the other. The point of the comparative approach to world-systems is to investigate these structures rather than assuming them. The main purpose of the comparative world-systems approach is to understand how core/periphery relations emerged and how they were connected with major changes in the organization of power and the logic of accumulation. We do not assume that all world-systems have core/periphery hierarchies. On the contrary, the existence and nature of core/periphery relations is the main object of research. We include egalitarian systems in the scope of comparison precisely because we hypothesize that there are important causal connections between the internal stratification of societies and intersocietal stratification. A broad comparative framework that includes egalitarian systems allows us to examine how egalitarian intersocietal relations were reproduced and how socially structured intersocietal inequalities emerged. The general hypotheses suggested by the comparative world-systems perspective are set within the framework of an heuristic typology of world-systems (Chase-Dunn and Hall, 1993:866-71). The typology includes the modern global political economy and earlier regional intersocietal interaction networks. Many of the larger systems are ones in which states competed with one another in a context of regional trade networks. The study of state-based world-systems other than the modern world-system has experienced a florescence in archaeology (reviewed in Hall and Chase- Dunn, 1993; see also Algaze, 1993) and also among historians, sociologists and political scientists (e.g. Mann, 1986; Abu-Lughod, 1989; Tilly, 1990; and the articles in Hall and Chase-Dunn, 1991; and Chase-Dunn, 1992) The comparative world-systems approach has also been applied to stateless intersocietal systems. Early work by Friedman and Rowlands (1977) formulated a world-system theory of chiefdom formation and early state formation and Friedman (1981) has applied this approach to Oceania. The inclusion of stateless systems in the comparative framework of world-systems research has been undertaken by Randall Collins (1992) and several researchers have carried out studies that are directly relevant to this task (e.g. Kristiansen, 1987,1991; Baugh,1991). The study of stateless world-systems is itself a huge topic. Small scale intersocietal systems varied from those composed exclusively of sedentary hunter- gatherers to those much larger systems in which complex chiefdoms interact with one another. Sedentary foragers -- stone age hunter-gatherers living in fairly permanent villages -- first emerged about ten thousand years ago in the Near East (Henry,1985). A system composed completely of sedentary foragers continued to exist in California until the middle of the nineteenth century CE. The most general hypothesis about this sort of egalitarian system is that core/periphery differentiation and hierarchy will be mild and unstable in such systems. An earlier study utilized the comparative world-systems perspective to study the nature of interaction networks and potential core/periphery relations between valley dwellers and hill people in late prehistoric Northern California (Chase-Dunn, Clewett and Sundahl, 1992). The present study jumps over a number of potentially interesting types of world-systems (e.g. big man, simple chiefdoms) to examine one composed of complex chiefdoms in which rather extreme class stratification has emerged. The questions remain generally the same: Do core/periphery differentiation and hierarchy exist? If so how are they organized and what role do they play in the development of the system? These case studies are undertaken in order to explore basic problems in conceptualizing small-scale world-systems for the purpose of comparing them with larger world-systems. The long term goal is to enable us to test propositions about variation in world-system structures and processes of change by comparing a large number of world-systems to each other. This kind of cross-world-systems research needs to be preceded by careful case studies that analyze the nature of key structures and processes in particular systems. It will only be possible to know how representative the findings of the case studies are after larger numbers of cases have been compared. Chiefdom world-systems Chiefdom world-systems were those in which the largest and most complex polities were chiefdoms. These may also have contained less complex societies that were in interaction with the chiefdoms. But once hierarchy exists within a system there are often pressures for all the interacting societies to develop their own hierarchies in order to provide protection against hierarchical neighbors. Exceptions are people who live in mountainous terrain or other inaccessible and easily defended regions. It is costly to dominate and exploit such people and so they may retain rather egalitarian forms of organization even in a context where they are surrounded by large and powerful neighbors. Hierarchy formation undoubtedly occurred in several different ways. Friedman's (1982) world-systems analysis of Oceania distinguishes between chiefdoms based on prestige goods systems, in which the power of chiefs is based on their ability to monopolize imported status goods, and caste-like "feudal" systems like those found in Eastern Polynesia, including Hawaii. This difference is similar to Earle's (1991) distinction between hierarchies based on wealth finance and those based on staple finance. Friedman stresses that hierarchy-formation is reversible. He points out that, if hierarchies automatically emerged with the passage of time, Melanesian societies should be the most hierarchical because Melanesia has been occupied for much longer than has Polynesia. But Eastern Polynesia, the most recently settled region, is also the region with the most stratified societies. Friedman contends that earlier Melanesian and Western Polynesian hierarchies based on the control of prestige goods declined in a context of increasing trade density (because chiefs lost their monopoly of prestige goods imports). Chiefdoms of the Eastern Polynesian sort, which are based on staple finance rather than prestige goods monopolies, sometimes collapse because of overexploitation of the environment (e.g Easter Island). There are also cases in which hierarchy formation and political centralization becomes stalled. Kirch's (1991) analysis of the Marquesas Islands societies argues that relatively poor ecological conditions prevented the emergence of island-wide polities and perpetuated an endemic warfare and cannibalism that performed the function of demographic regulator. The model of social evolution underlying this approach is multilinear, path dependent and allows the possibility of devolution. The broad typology of world-systems proposed by Chase-Dunn and Hall (1993) utilizes the structural distinctions proposed by Johnson and Earle (1987). They distinguish between big-man systems, simple chiefdoms and complex chiefdoms. These types are mainly based on the degree of internal hierarchy, the nature of the institutionalized power of leaders, and the size of independent polities. Complex chiefdoms have rather large polities and have class-stratified social structures based on a hierarchy which separates the class of chiefs from the commoners and strongly institutionalizes the control of the chiefs over resources that are necessary to the daily lives of commoners. Johnson and Earle (1987) make the cut between complex chiefdoms and early states in terms of the emergence, in true states, of specialized institutions of regional control such as standing armies or permanent bureaucracies. Chiefdoms rely exclusively on kinship-based marriage alliances and a metaphor of lineage seniority for controlling distant regions, while states employ non-kin-based institutions of regional control. Johnson and Earle also utilize the polity size criterion in this distinction. States are usually larger than complex chiefdoms. Though these distinctions may be difficult to apply to particular cases, we agree with Johnson and Earle that the underlying dimensions of organizational differentiation that they specify are important for distinguishing between different types of systems. There has been a lot of attention given to chiefdoms by anthropologists and archaeologists recently (see Earle, 1991; Upham, 1990) Much of this work is relevant for the project of comparing world- systems despite the fact that the main theoretical questions are usually somewhat different than those that are posed by the world-systems perspective. One study of a complex chiefdom that explicitly employs world-systems concepts is Peter Peregrine's (1992) analysis of the Mississippian civilization centered at Cahokia. Patricia O'Brien (1992) has also analyzed this same case from a worlds-systems perspective. While this is an important case of the rise and fall of a complex chiefdom (or early state), our knowledge of it is limited entirely to archaeological evidence. One reason to study Hawaii is that we have ethnographic and documentary, as well as archaeological, evidence. Another reason is that the cultural region of which Hawaii is a part, Oceania, has been the subject of comparative studies from both diffusionist and evolutionary perspectives for over a century. The island societies of Melanesia, Micronesia and Polynesia ostensibly represent a set of cases that are isolated as if in a laboratory. Thus similarities and differences can be studied under conditions in which outside impacts are controlled, or at least, these influences can be more easily known and taken in to account. The theoretical analyses of processes of development in Oceania present a useful backdrop for the comparative world-systems perspective. Precontact Hawaii was a proto-typical complex chiefdom with large polities and a radical separation between social classes. It is important to study a chiefdom world-system because this kind of system is intermediate between the kind of egalitarian relations we found in California and the much more familiar state-based systems with their obvious core/periphery structures. All state-based systems exhibited certain analytically similar features. They went through a sequence of political centralization and decentralization in which a single state conquered a large chunk of the contiguous core region and then this conquest empire fell apart to once again take the form of a system of competing states (an interstate system). New core regions and dominant states emerged as these systems developed, with formerly semiperipheral regions attaining core status at later points in time. Chiefdom world-systems are much more similar to the modern system than California was. Chiefdom systems also experience a sequence of political centralization and decentralization in which larger conquest chiefdoms rise and then fall apart. This was described and analyzed for the Hawaiian case by Sahlins (1972:141- 148). And, like state-based systems, chiefdom systems occasionally go through radical expansions in which the whole scale of the system gets larger. And, in both state-based and complex chiefdom systems, conquest and the political/military cycles are connected to the dynamics of class exploitation within polities. There are other important differences between chiefdom systems and state-based systems in addition to size and the degree of specialization of control institutions. The strategies of chiefdom formation and expansion differ in important ways from the strategies of state-formation and expansion. And there are different strategies of chiefdom formation, as discussed by Friedman (1981). These differences are discussed in Chase-Dunn and Hall (1993). We hypothesize that the unspecialized nature of internal political hierarchies in chiefdom systems will correspond with an intermediate degree of core/periphery hierarchy because kin-based control institutions are relatively poor vehicles for exploiting and dominating distant regions. In the case we want to study, precontact Hawaii, did core/periphery differentiation and/or hierarchy exist? If so, how were they organized and what role did they play in the processes of political rise and fall and the growth of polity size? These are the main questions that our research on Hawaii is trying to answer. Why Hawaii? We have already mentioned that ancient Hawaii is a rather well-known example of a complex chiefdom. We have also mentioned that there is more evidence and different kinds of evidence about ancient Hawaii that there is for most complex chiefdoms. Archaeology in Hawaii has made great strides in recent years (e.g. Kirch,1985, 1990b) and ethnographic, traditional and documentary evidence are extensive relative to what we have for other complex chiefdoms. In addition, this evidence is more available because we can communicate easily with social scientists working in and on Hawaii. Hawaii is atypical in the degree to which it was isolated from outside influences, but this is primarily an advantage for world-systems research. One difficulty with world-systems research is that it is hard in practice to distinguish between exogenous factors which occasionally have important impacts on a system and endogenous factors which have systemic consequences. This is less a problem with the global modern world- system than it is for earlier regional world-systems. The diffusion of a technique or idea may have big consequences for a regional system even though the origin of that factor was outside the regularized interactional networks of the system. The isolation of the Hawaiian archipelago before the arrival of the Europeans is an advantage because it reduced the frequency of such exogenous impacts, though it did not eliminate them. The fact that the sweet potato somehow got from Peru to the Marquesas and hence to Hawaii made possible the development of dry field agriculture, which raised the human carrying capacity of the Hawaiian environment and provided an important alternative form of intensification of production. It is likely that institutional diffusions from Tahiti had some impact on Hawaiian religion and political organization in the twelfth or thirteenth centuries CE (but see Spriggs,1988:68). These examples show that even such an isolated region as the Hawaiian archipelago, thousands of miles across the open ocean from other inhabited islands, was not completely free from exogenous impacts. Yet they also suggest that these kinds of impacts were fewer than for systems that were less isolated. One might suppose that another possible advantage to studying Hawaii flows from its isolation. The restriction of exit options due to the long distances to other inhabitable lands might have intensified the circumscription factor and thereby intensified the processes of hierarchy formation. But this was apparently not the case in Hawaii. The Handys argue that the development of agricultural terraces and irrigation systems had not yet progressed to the point of taking up all the lands that were appropriate for such development (Handy and Handy, 1972:280-4). Apparently complex chiefdoms emerged in Hawaii despite the fact that the land was not completely "full." This does not despoil theories of population pressure and circumscription, however, because these do not require that absolute constraints on carrying capacity be reached. Kirch (1988:422-3) hypothesizes that the reason why political evolution was more intense in the newer islands of Hawai'i and Maui was because these had less arable land and thus experienced population pressure more intensely. The Hawaiian World-System Hawaii is culturally a subregion of Polynesia. Ocean-voyaging Polynesians, probably from the Marquesas, first occupied the Hawaiian archipelago as early as the third century C.E.(Kirch,1985:87). The first settlers bore the cultural heritage of their Polynesian ancestors: knowledge of agriculture, sophisticated fishing techniques, boat-building, navigation, the principle of hierarchical kinship based on successive primogeniture, and the ideas of mana and kapu (taboo). When Captain Cook first sighted Oahu and Kauai in 1778 the Hawaiian world-system was an inter-chiefdom system composed of four independent chiefdoms: Kauai and Niihau; Oahu and probably Molokai; most of Maui, Lanai and Kahoolawe; and Hawai'i and part of East Maui (Spriggs, 1988:63-4). Each island was divided in to major districts and these were further divided into ahupua'a, typically pie-shaped areas including seaside, flat land and a section of mountainous area. In theory each ahupua'a was self-sufficient because it contained all the necessary ecological zones from which food and raw materials were extracted. Each ahupua'a had a local chief, a konohiki, who was responsible for collecting and delivering food and other products of the commoners for the provisioning of the higher chiefs. The konohiki also acted to maintain the population of his ahupua'a. He encouraged marriage within the boundaries of his lands and sought to discourage emigration. The settlement system of the ancient Hawaiians was rather dispersed. Housing was located near fields or fishing places. People did not build houses close together in villages except under rare conditions in which a more dispersed pattern was made difficult by topography. There were big differences in the population densities of different areas, but these were due mainly to the uneven distribution across space of prime sites for agriculture and fishing (Weisler and Kirch,1985). The interisland density of interaction must have been fairly high at the time of Captain Cook's arrival. Some authors have argued that the wide and dangerous channel between Kauai and Oahu limited interaction between Kauai and the other islands, and we know of historical incidents that support this contention. King Kamehameha was prevented from mounting an expedition of conquest against Kauai because a storm at sea disrupted his war-canoe navy . And yet we know that there were many intermarriages of chiefs from Kauai with chiefs of the other islands. Captain Cook discovered that it took venereal disease less than ten months to travel from Kauai to Maui, a distance of over two hundred nautical miles (Sahlins,1985:1-3). Interisland interaction density must have been fairly thick. The first Polynesian settlers in the Hawaiian archipelago found a land full of food. For the first few hundred years they derived a large proportion of their protein from the meat of birds, including a species of flightless geese that had no natural predators until the arrival of humans. Hierarchy was probably minimal in these early settler societies as people spread out across the archipelago occupying the prime sites for agriculture, fishing and birding. As population density increased and all the islands became populated, the ancestral Polynesian cultural script (contained in the principle of successive primogeniture, and the ideas of mana and kapu) were used to fashion a more hierarchical society. Polity sizes grew and class stratification became more extreme. A simple model of this process of population growth and lineage stratification is presented by Kirch (1984:199-202). The principle of successive primogeniture is organized in Hawaii and Polynesia on the basis of conical clans. Genealogical lines are ranked in terms of closeness to the original ancestor and each person theoretically has a unique and intransitive position in this system of ranked seniority. In practice this system operated as a theoretical ideal. Competing claims were settled by force and genealogies were reconstructed according to the outcomes of struggles among contending chiefs. Nevertheless, the principle of genealogical seniority was an important one in legitimating rule, and this principle evolved, in Hawaii, to include a radical separation between chiefs and commoners, and the practice of brother-sister marriages among sacred chiefs and chiefesses to produce the highest possible seniority for offspring. Kirch's model for colonization and population growth works as follows. A colonizing party landed on a hypothetical island and occupied the most favorable location on the windward side of the island. The windward side receives the most rainfall and is generally the best for agriculture. Other considerations were offshore reefs that make for good fishing locations. When the best location became fully occupied this community hived off to occupy the second most favorable location on the windward side. The second community was headed by a chief who had less seniority than the line that ruled the original founding community. This process continued until the windward side was occupied, at which point locations on the leeward side of the island were occupied in the order from best to least desirable. This produced an island in which the ecological desirability of locations mapped perfectly with the kinship hierarchy. Kirch goes on to propose that, after the island was filled up, continued population growth brought about pressures for either further migration to other islands or resulted in intensification of conflict and, eventually, pressures for island-wide hierarchy formation. He suggests that successful conquest efforts are likely to have come from regions that were ecologically less favored such as the leeward side of the island of Hawai'i, where large-scale irrigation was not feasible (Kirch,1984:204). It is thought these would have been the first to experience population pressure and they would have the most to gain from a policy of conquest. This model may be interpreted in world-system terms as follows. An intra-island core/periphery structure emerged in which ecological factors and kinship hierarchy corresponded. Continuing population growth created pressures for intensification of production, intensification of conflict and hierarchy formation. These pressures eventually resulted in the formation of island-wide chiefdoms by means of conquest, and the successful conqueror chief was likely to come from one of the junior lines on the less ecologically favored side of the island. This is last proposal is fascinating because of its analytic similarity with a process known to occur in state-based systems -- the formation of empires by semiperipheral marcher states (Chase-Dunn, 1988; Mann, 1986:130 f.f.). This model is an ideal-type that was probably not ever perfectly realized for particular islands. The model assumes that a single island was populated until it was full and only then did people migrate to the next island. It also assumes that there was a clear order of ecologically desirable locations and that these were determined by one or two major factors. It is reasonable to guess that the Hawaiian islands were populated by groups that both moved to new locations within their original island and to other islands. Interisland migrations within the archipelago, and the very occasional arrival of migrants from the distant Polynesian south, may have complicated the relationship between seniority and ecology on any single island. Also, the ordering of desirable locations is relative to the kind of subsistence system that was being pursued. The best locations for birding were probably not also the best locations for fishing or for agriculture. As the nature of economic activity and technology changed, the ranking of desirable locations would also have changed. And we also know that chiefs preferred certain locations because they were beautiful or because the surfing was good. Thus the determination of spatial desirability was multifactorial and changed over time with the development of population, investments in terracing, irrigation, fishponds and religious architecture (heiaus). Nevertheless, with all these complications in mind, Kirch's model is a good beginning for studying core/periphery relations, the emergence of hierarchy, and changes in the size of polities. We will extend Kirch's model to the interisland level in order to construct a set of hypotheses about the development of the Hawaiian world-system. This exercise is the beginning of research, not its end. The research proposed below will evaluate this model and consider others. In exploring interisland interactions and possible cultural and structural differences among the islands we are taking a course that has been recommended by Kirch (1990) but which has not yet been much pursued. The dominant tendency in research on Hawaii has been to assume homogeneity across the whole system, so a finding from one locale is generalized to the whole archipelago. The world-systems approach encourages us to analyze within-system differences as well as similarities. The Hawaiian archipelago is laid out on a northwest to southeast axis. The geologically oldest islands are those to the northwest (Kauai, Oahu) and the youngest island is the big island of Hawai'i, with still-active volcanos. Weathering, valley-formation, soil-formation and off-shore reefs are much further developed on the older islands, while the younger islands are steeper, rockier and have fewer and smaller permanent streams and little shallow water off-shore. The original settlers probably did not care much about soils and flat, irrigable land. They were birding and fishing. But once agriculture became a more important proportion of subsistence, people would tend to move to islands with flat land, better soils and more water. These would be the first areas to develop extensive agricultural terraces and irrigation systems. Population densities would likely be greater earlier on the geologically older islands and these would be the most likely to first develop class structures and large chiefdoms. Our study focusses on and compares two locales -- the Wailua River and Puna district of Kauai and the Kona and Hilo districts and the Pololu valley on the big island. Why Kauai? It is our hypothesis that Kauai was the first island to develop complex chiefdoms in the Hawaiian archipelago and that the seniority of chiefly ranks of the islands of the archipelago reflects this in a way similar to that for Kirch's model of a single island. Oahu and Kauai were the homes of the most senior chiefly lineages (Valeri 1972:59; Fornander 1880:21- 28). Traditional sources claim that the first settlers landed at Wailua on Kauai. A later party from Tahiti, bringing a shark-skin drum and a more rigid system for preventing marriages between chiefs and commoners, is supposed to have also landed at Wailua. We hypothesize that Kauai was the old core. It was the first to develop the institutions that became the cultural and organizational nexus of the Hawaiian world-system. Kauai's developmental primacy gave it advantages over the peoples of the other islands. But the institutional developments on Kauai spread to the younger islands and eventually to the big island, Hawai'i. Hawai'i remained the frontier. Class relations were less hierarchical there. Commoners still made claims to the land through a kinship group called the 'ohana (Handy and Pukui, 1972), a feature which is not found on the older islands (Sahlins, 1958). And it was from the leeward side of the youngest and most junior island (Hawai'i) from whence came the conqueror of the whole archipelago, Kamehameha. Kauai was the Florence or Venice of the system, the repository of high culture, good taste, a former center of real power that had declined but still had high symbolic standing. Hawai'i and Maui were the frontiers with class relations that were somewhat less extremely stratified and with more recently developed chiefdoms. They were the last to develop island-wide polities. But these factors operated as the "advantages of backwardness" and the big-island kingdom, once formed, was able to mobilize its soldiers to conquer the other islands. Of course the guns and ships that the British provided to Kamehameha also played a role. Some have argued that the process of inter-island polity formation would have eventually taken a similar course even without the exogenous factor of new ships and guns. The big island was like the United States during the first half of the twentieth century -- less stratified, larger, able to adopt new institutions and technologies more easily because of the recency of political expansion and polity-formation, tasteless but bold and naively self-possessed. This analogy with the modern world-system is rough, but is meant to suggest a scenario that can be investigated in the Hawaiian case. Most of the research on Hawaii has tended to assume a single homogenous culture across all the islands. Kirch (1990) has argued that new studies should investigate regional variation within the archipelago in cultural and material features and he suggests some interesting examples and explanations. The world-systems approach requires that we examine questions of this sort. What was the real interisland ordering of the development of intensified production, classes and chiefdoms? Which islands were first unified under a single chief and did this occur in the same way on all the islands? How was the sequence of the development of monumental architecture (Kolb,1992) similar or different on the different islands? In looking for core/periphery relations we are encouraged to investigate these questions. Within Kauai we will focus on the Puna district and the drainage system of the Wailua River. Wailua was the ancient summer capital of the island-wide kingdom of Kauai, and before that it was the chiefly residence of the smaller kingdom of Puna. The Wailua River is the largest river in the Hawaiian archipelago. It is navigable by large canoes for quite a distance upstream. The river valley cuts between two mountains just before the river enters the sea. This whole area was kapu to commoners and was the home of the ruling chiefs and their retainers and families. This ancient capital is widely known in the traditions of Native Hawaiians. All chiefs of Kauai were ideally supposed to be borne at a particular sacred spot in Great Sacred Wailua. A commoner-woman who expected to have a boy could receive permission from the priests to give birth at the sacred place, and the boy child would be raised in Wailua as a chief. The focus on Kauai provides a needed balance to the general knowledge of Hawaiian political history. A great deal of attention has been paid to the political history of the big island because that is the island that Kamehameha came from. This is a familiar consequence of power transition. The newly powerful are the most likely to receive the attention of chroniclers, and the authors of chronicles are likely to be those who are attached to the retinue of the now- powerful. Thus it is not surprising that we know most about the political history of Hawai'i and very little about the political history of Kauai. One reason to focus on what we hypothesize to have been the old core is that we will be able to help balance this state of affairs to some extent. Since the amount and quality of information is probably biased in favor of the more recent core (a common methodological problem in world- systems research) it may be necessary to read between the lines to some extent. The danger here is that we may impose our interpretation on the evidence. We will make every effort to guard against that error, including working with colleagues whose focal regions are on Maui. Why the Big Island? It is our hypothesis that the Big Island of Hawai'i was becoming the core of the ancient Hawaiian world-system at the time of Captain Cook's arrival. Kirch's model of island-level colonization and population growth must be applied to the Big Island with the attention to spatial desirability as multifactorial. According to the Handys, on the Big island a factor of prime importance affecting the development of plantation areas was propinquity to good fishing grounds. The land areas that were intensively developed were always in localities where good fishing grounds were easily accessible. As a general principle Hawaiians developed their land resources only where they lay not too far distant from good fishing grounds which would give them their needed protein food. Such localities on the island of Hawaii were Kona and Ka'u. The legendary evidence also indicates the presence of ancient chiefdoms in certain localities. Waipi'o, the great canyon-like valley toward the north end of Hamakua District, was the seat of renowned ancient ali'i, notably Kiha, Liloa, and Umi. Another important factor in spatial desirability worth mentioning was the habits and values of the Polynesian peoples who came to the Hawaiian islands and the environments and opportunities that they had before (Holly McEldowney, personal communication 1993). For example, it may have been "too wet" to settle on the windward side of the Big Island for people who came from Tahiti or the Marquesas. On Hawai'i the intensive dryland field systems were predominant in contrast with the westerly islands (Kaua'i, O'ahu, Moloka'i), where taro irrigation prevailed. Intensive dryland systems have much higher labor requirements. Kirch cites Tuggle and Tomonari- Tuggle who, on the basis of their North Kohala studies, proposed that "if all this is translated into simple terms of food production and population, it suggests that population growth on Hawaii had reached a plateau while on the other islands it was still increasing... If so, it can be argued that demand for agricultural land, particularly irrigation land, increased competition among polities, thus acting as a variable in the process of political elaboration" (Kirch, 1985:235). Confronted with diminishing returns, the chiefs of Maui and Hawaii imposed taxes on both men and women and got the name of being oppressive to the people, while the chiefs on Oahu and Kauai demanded taxes of the men alone. On Maui and Hawaii women worked outside as hard as the men (Kamakau 1961). We want to investigate both interisland and intraisland core/periphery relations. At the interisland level this will mean comparing the dimensions of core/periphery differentiation and hierarchy across the islands. We have picked Kauai and the Big Island because they are hypothesized to be polar cases. At the intraisland level it is our hypothesis that Kona was the core district of the Big Island. We will also investigate this hypothesis by comparing Kona with other districts, especially Hilo and the Pololu valley. The intriging question concerning the political history of the Big Island is what prevented any of the regions of the island of Hawaii or any regional fraction of the ruling class from achieving island-wide political centralisation? For several centuries there existed two autonomous political units on the Big Island. One under the control of "Kona" chiefs - Kohala, Kona, and Ka'u. And the Hilo chiefs controlled Hamakua, Hilo, and Puna. Having no quarrel with the existing explanations of the long-term political and economic equilibrium, we intend to examine the role of the Hawaiian religion in social transformation. Did it set limits on growth or not? We hypothesize that the Hawaiian religious system set limits on development in the sense that it prevented the accumulation of finance goods. Religion in pre-contact Hawaii was highly integrated across several islands. There were important differences in types of gods worshiped by people of different occupations and islands (an important point which will be discussed later), but religious manifestations were island-wide and concerned with problems that transcended island boundaries. The main gods were Kane, Lono and Ku. The Handys argued (1972:15), that Kane was introduced by the first settlers, and that it was these colonizers who established systematic agriculture in those areas that were capable of systematic development by means of irrigation; in other words, primarily the windward coasts and the valley areas on leeward sides of Kauai, Oahu, and West Maui. On Hawai'i, people retained the amalgam of archaic religious forms, however they combined them with more recent forms. The sacred organization was divided into two cults, each of which was administered by a separate hereditary order of the priesthood (Kamakau 1964:7; Handy and Handy 1972:322). The cult of Ku, of which Ku-ka'ili-milu was an important manifestation, was the most powerful of the two. The other cult was that of Lono. Hawaiian ritual annually alternated between the gods Lono and Ku. Lono was preeminently the god of growth and presided over the life of the people in general. As such he was a nourishing god. In warfare he was responsible for defence. Ku, as a war and sorcery god, enable his worshippers to conquer lands. He is the god who seized governments. Although all the major gods contributed to the strength of a chiefdom, the forces of Ku were seen as somewhat antithetical to those of the other gods. The prudent paramount chief, therefore, was expected to maintain a reasonable balance between the two (Davenport 1969:9). Thus, every temple dedicated to Ku had to be counterbalanced by temples for the others (Malo 1951:189). Davenport's (1969) analysis points out that, in response to factional pressures, some paramount chiefs divided the chiefdom into several smaller ones and designated a successor to rule each. Another stratagem was to divide the chieftainship and designate two successors as co-rulers. The legendary prototype for this was the action of the great king Liloa who designated his son Hakau to be successor to the chieftainship but made his other son, 'Umi, custodian of the state god Ku-ka'ili-moku. This meant that although Hakau was the senior chief, 'Umi had administrative control over the religious branch of the government. In these situations, however, the legendary evidence also dramatically points out that the factional quarrels were only temporarily dampened by such divisions of power and that the inevitable struggles to recombine and recentralize the rule soon erupted in violence. The point can be made here that it is possible to associate senior and junior chiefs with certain gods: Ku - with junior nobles, and Lono - with senior. By this Ku and Lono are becoming symbols of strata within the chiefly class. The god and the particular manifestation worshipped depended on the identity of the worshiper and the circumstances. So, the point can also be made that in their struggle for political power chiefs' aspirations found more "support" in either Ku or Lono. Lono was the god of "clouds, winds, the sea, agriculture and fertility" (Pukui and Elbert 1973:392). Handy and Handy (1972:320, 333) state that Lono was particularly associated with sweet potato and the southerly or kona storms of the rainy season. The rains that these storms brought were of utmost importance to agriculture, particularly cultivation of the non- irrigated crops, dry taro and sweet potatoes, on the drier leeward coasts. The cult of Lono was important in those areas, particularly in Kona on Hawaii (1972:14). Historical evidence shows the kings who worshiped Ku - Umi, Keawe-nui-Umi, Lono-i-ka-makahiki, Kalani- 'opu'u (e.g. Kaeppler 1982:99). All represented the junior line, and all initially resided not in the core /leeward/ districts of their time. According to Hommon (1976:153) in traditional sources the explicit reasons given for war include the accumulation of chiefs' prestige (Kamakau 1961:45-6, 230), revenge (1961:66-7; Fornander 1969:210, 289) and the elimination of oppression (Kamakau 1961:35; Fornander 1969:88). One of the mechanisms of the latter is discussed in Valeri's (1985) political analysis of the Hawaiian legend of 'Umi: the practice when commoners were looking for a "good" king. Using a systemic approach the point can be made that a (semi)peripheral area (Hilo) attacked the core (Kona) based on the incentives born purely of their (Hilo's) locality. People from Hilo compared and saw that the "level of life" was higher in Kona and "pushed" their chief to attack. The unequal distribution of the rewards of trade did not necessarily generate dissatisfaction with the chiefs among their followers. The literature on conspicuous consumption suggests that it is quite likely that the members of a clan saw the conspicuous consumption of their chief as a public symbol of their clan's prestige. "As the rat will not desert the pantry (kumuhaka) where he thinks food is, so the people will not desert the king while they think there is food in his store-house" (Malo 1951:195). Some other evidence shows that people from Kona came to Hilo and asked Hilo's chief to "secure" them from the exploitation and corruption of Kona's chiefs. About 1781 I-makakoloa of the Puna district, Hawai'i was moved to revolt against Kalaniopu'u (H19-3k; AH-23) because of the latter's profligate ways (Kamakau 1961:105-107; Fornander 1969:200). However, the points that we want to add here are, first, the reduction of exploitation as the consequence of inter-district raids or wars resulted in the prevention of accumulation of staple goods in the core areas; second, in their search for a "good" king commoners were appealing to junior chiefs in peripheral areas who worshiped the god Ku since Ku legitimates an opportunity to rebel against high exploitation of the core areas. In the Legend of Popoalaea Ku is called "a god of justice" (Beckwith 1976:381). Finally, it is important to point out that Ku was invented and praised, not by a paramount, but by the people down the social hierarchy. In the events of October 1819 (when the indigenous religion was abolished) Ke-kua-o-ka-lani, as custodian of the state god Ku-ka'ili-moku, was inalterably committed to defend the religious tradition against the actions of the abolitionists. The following of Kekuaokalani consisted largely of the traditional territorial chiefs of the middle rank (Webb 1965:23). That the opposition to the destruction of the old system should so largely consist of lesser, local chiefs reflects the fact that any person outside the government would have no real power or place at all once the sacredness of the old system was ended. We tentative conclude that social development on the Big Island was subject to very contradictory forces. From one side, the stress on the god Ku increased the opportunity of lesser chiefs for upward mobility; but from another, cultural resistance legitimated by the Hawaiian pantheon had a leveling effect on districts within an interchiefdom system. As a result none of the fractions within the ruling class could achieve stable political centralization of an island. We want to consider the social and political consequences of ritual sacrifice. Classically, the leveling effects of redistribution systems are discussed in terms of equalizing possessions of commoners, or of commoners and chiefs. Similarly, in the context of Hawaiian history, redistribution in the form of the "sacrifice to the altar" is discussed as a mechanism of leveling the possessions of chiefs and commoners (Earle 1977). The research questions that we want to pose are: what kind of good is primarily used for sacrificial activity? Is this good important for the social reproduction of the members of the society? What class or fraction within a class does accumulate this good as a result of a redistribution and who is denied? If, the division of a class into fractions is associated with the territorial division of a system, what are the consequences for the development of the regions? Our reading of the Hawaiian history and ethnography suggests that bird feathers were the prestige goods used in gift-exchanges to establish the loyalty of the supporters. There is also evidence that feathers had a tendency to be accumulated by lesser chiefs who resided on the windward side of the Big island and worshiped the god Ku. The main lands that produced the valuable feathers were in the Hilo and Puna districts (Malo 1951:77). Feathered images were often associated with the war god Kuka'ilimoku (Kaeppler 1982:86). Lesser chiefs who were denied the royal power were seeking alternative means to establish their power. From another side, feathers were intensively used in the redistribution system. According to Malo (1951:77), feathers were the most acceptable offering to the Makahiki idol and the lands that produced feathers (Hilo and Puna) were heavily taxed at the Makahiki time. The paramount chief, as the earthly manifestation of the Makahiki god Lono, received the feathers directly. During the next phase of the annual cycle - the Luakini temple ritual - the king and his god, as representatives of Ku, redistributed feathers among his nobles (Valeri 1985:270). We hypothesize that the Hawaiian ritual of sacrifice served as the idiom of factional competition and conveyed manifestly political messages. In the annual religious cycle of the society the sacrifices consecrated to the god Lono (who was mainly associated with the senior nobles) included the items "prestigious" for Ku (associated with junior nobles). As the result, the junior nobles were constantly denied the accumulation of prestige goods as the means for the reproduction of local power structures and hierarchy formation. We also suggest that by equalizing the symbolic manifestations of power of the competing nobles through sacrifice, the annual cycle slowed down the development of "coreness" and "peripherality". Thus, we hypothesize that redistribution had a leveling effect on the amount of valuables possessed by any fraction within the ruling class, and if the evidence for intentionality can be found, then sacrifice served as a mechanism of political struggle among fractions of the ruling class. This hypothesis has an important implication. Jonathan Friedman's (1981) explanation of the absence of a prestige goods system in Hawaii makes the point that there was not any Hawaiian social group who's power depended on the monopoly of prestige goods. The above hypothesis suggests an alternative explanation of the absence of a prestige goods system. There were prestige goods and there was a group that could have potentially benefitted from monopolizing their control, but the religious institutions of ritual sacrifice operated to undercut this possibility by equalizing access to these valuables. We may also wonder if the leveling mechanisms that operated within the island of Hawai'i had any affect on the structural position of the Big island within the Hawaiian world-system? The theory of semiperipheral development (Chase-Dunn, 1988) contends that a relatively low level of stratification within a polity can be an "advantage of backwardness." A kin-based reciprocal society can more easily mobilize collective energies (for either production or warfare) than can a more stratified class society. Increasing stratification between classes makes mobilization of the commoners for warfare more dependent on material incentives and less susceptible to calls for sacrifice in the name of the collectivity. With the coming to power of Kamehameha the semiperipheral island of Hawai'i combined elements of a peripheral kin-based mode of production (as represented, for example in pantheon) with elements of core-like tributary mode based on taxation. Kamehameha succeeded in eliminating many of the vestiges of the kin-based mode that had remained in the old core areas of the western islands and in establishing a more centralized, more exploitative, purer form of the tributary mode of production than had ever existed in the archipelago before. The hypothesis of semiperipheral development also provides a wide role for voluntary action and open- endedness in social development. As discussed by Valeri, Kamehameha introduced changes in the traditional Makahiki festival by making room for his own gods (particularly the sorcery gods and Kuka'ilimoku - cf. Kamakau 1964, 19-20; 1961, 180). He also prescribed a tribute payable during this festival (Kamakau 1964, 19-20). Here perhaps we see a conflict between political and cosmological motivations: the development of royal authority under Kamehameha imposes the preeminence both of the king's tribute over the firstfruits sacrifice and the god of force (Kuka'ilimoku) over the god of the festival (Lonomakua). Another important factor in Kamehameha successful conquest of the other islands was polity- scale. At the moment of conquest, the Big Island was unified. Of course we have not mentioned the crucial matter of the introduction of British ships and guns, and Kamehameha's success in obtaining the lion's share of these new "techniques of power." While these were undoubtedly important factors, we contend that structural processes already at work within the Hawaiian world-system would likely of have eventually led to state formation even without the exogenous impacts that accompanied incorporation in to the modern world-system. Methodological Considerations Our research employs several different kinds of evidence. Virtually all of the investigation involves secondary analysis of data gathered for other purposes. Secondary analysis is fraught with potential for errors of inference because the lack of fit between the intentions of the original researchers and the current uses may introduce systematic biases. In addition, the main kinds of evidence we use -- archaeological, ethnographic and documentary sources -- are unfamiliar to most sociologists. It is necessary to be schooled in the kinds of problems that are specific to these kinds of evidence in order to avoid mistakes of inference, especially when secondary analysis is undertaken. Our work on California has already taught us to understand the methodological conventions and difficulties of archaeology, ethnography and documentary analysis and we have consulted with archaeologists in Hawaii who have local knowledge of the particular areas we are studying. We are not doing additional archaeological excavations, surveys, or ethnographic interviews. Rather we are carrying out secondary analysis of archaeological survey data and improving upon the surveys that have been published by adding more recent materials available in the contract literature. We also are coding "events" or "instances" from documentary and ethnographic materials in order to study local and regional interaction networks. The purpose of this approach is to be able to evaluate and go beyond the general characterizations of interaction networks in the ethnographic literature. For example, most characterizations of Hawaiian exchange patterns contend that ahupua'a were self-sufficient and so there was little exchange across ahupua'a boundaries except for the ritual giving of tribute. Study of early reports by explorers, missionaries, settlers and recorded indigenous histories will enable us to evaluate this claim. The problems with this procedure have to do with bias due to selection effects and problems with the analysis of small samples. For example, it is possible that there were dense trading networks across ahupua'a boundaries but instances of these kinds of exchanges were never recorded because they were not of interest to observers. This is why we are also using archaeological evidence of exchange networks whenever this is possible. Unfortunately the archaeological study of exchange networks in Hawaii is yet in its infancy. But there are several archaeologists and geologists now pursuing this topic (e.g. Marshall Weisler). Mapping Intergroup Interaction Networks The key to understanding any world-system is the social geography of interaction networks. In the comparative world-systems approach world-systems are defined as intersocietal networks in which the interactions (e.g. trade, warfare, intermarriage) are important for the reproduction of the internal structures of the composite units and importantly affect changes that occur in these local structures (Chase-Dunn and Hall, 1993:855). Ideally we would like to know the spatial characteristics of social networks exchanging bulk goods, raw materials, prestige goods, information, marriages, conflict events and alliances, celebrations, religious rituals, and travel for each individual, household, family, ahupua'a, district, island and the whole Hawaiian archipelago. Theoretical reasons for studying these kinds of interaction networks are presented below. We would also like to know how the patterns of these nested interactions changed over time from the arrival of the first settlers to the early years of the nineteenth century. It would also be ideal to have this detailed map of interaction structures for every locale in the system under study, but this is not a feasible possibility. We will focus our events coding on the two focal locales on Kauai and Hawai'i. With our events coding project we are reconstructing the interaction patterns that link the focal locales in to the larger networks at the period of contact with the Europeans, roughly around 1800, and for some types of interaction we are able to examine changes over time prior to 1800. We want to know what the system was like before its incorporation into the expanding modern world-system changed it dramatically. For some of the interaction networks we are only able to obtain a cross-sectional picture for the period around 1800 from our coding project. Better time depth is possible for the intermarriage network for chiefly marriages using genealogical studies. And the coding of conflict events has sufficient time depth to enable us to examine long- run changes in patterns. Our hypotheses about regional variation will be examined by working closely with colleagues who are studying locales that are hypothesized to be located in very different structural positions within the larger Hawaiian world-system. Dr. Michael Kolb of the Hawaii State Historical Preservation Division is serving as an archaeological consultant to this project to help sort out regional variation by facilitating comparisons with his careful study of the Hana District of Maui (Kolb,1991,1992,1993). Coding Interaction Events In order to construct estimates of these network characteristics we will use a methodology developed in our study of Northern California. This involved coding all the specific events that have been recorded in ethnographic, documentary and traditional literature about a region. We are also recording general descriptions of practices, but analysis is restricted to events. We are constructing five "events files" and coding all the literature available on the focal districts, their internal interactions and their interactions with other areas of the archipelago. The five files are: 1. Marriages: all known marriages are coded. Each entry contains information about the approximate date of the marriage and about the statuses, place of birth, current home, and other relevant characteristics of the partners. 2. Conflict events: this file contains information about conflict events including disputes, fights, murders, raids, battles, retaliations, punishments, sacrifices, crimes and the outcomes or settlements of these encounters. Information about the parties to the disputes, their numbers, the number of deaths, the location of their homes, their statuses, the weapons they used, and the approximate time that the event occurred are also included. 3. Exchanges: all exchange events including sharing of food, gifts, collective gathering, hunting or fishing events, trade, collection of offerings, taxation, tribute, and community feasting or celebrations, and the approximate time at which the event occurred. We also record general characterizations of trade and other exchanges. 4. Travel: all journeys short or long that are recorded in the literature are coded, the purposes of the journey, identities and statuses of the travelers, the distance of the journey, the place of beginning and the destination, and the approximate time at which the travel event occurred. The coding of marriages enables us to examine patterns of intermarriage across space and according to status. Friedman and Rowlands (1977) contend that chiefdom formation entails a marriage strategy that shifts from symmetrical intermarriage based on wife- giving to asymmetrical marriage based on wife-taking by the lineage rising to power. We are investigating patterns of wife-giving and wife-taking between different locales in Hawaii among both commoners and chiefs. Part of the notion of ahupua'a self-sufficiency is the idea that the konohiki tried to encourage endogamy within his ahupua'a. We are examining this notion with information about commoner marriages. This is difficult, however, because the main impetus to traditional genealogical accounting was the legitimation of chiefly lineages as senior. Thus commoner marriages with chiefs tended to be edited out of these accounts, though there are some famous exceptions (e.g. the story of Umi). Important sources are Fornander (1880), McKinzie (1983, 1986) and Hommon (1975,1976). In coding conflicts we can see where the lines of conflict were in this system. We will code the events in a way that will be designed to facilitate the study of change over time. The hypothesis stated above about the status of Kauai as an old core in the system can only be evaluated by comparing the political/military traditional history of Kauai with that of Hawai'i and Maui. Coding exchange and trade provides valuable information on patterns of local and regional interaction. We already mentioned our strategy for evaluating the hypothesis of ahupua'a autarchy. We also are examining the notion that interregional exchange was almost wholly monopolized through the mechanism of chiefly extraction and redistribution. The travel events file includes general characterizations of travel as well as instances of specific trips. And we are assembling information about trails, transportation, religious processions and journeys, etc. These allow us to examine the spatial extent of this kind of interaction. We hope to be able to find reports of enough specific events to enable us to estimate the relative frequencies of different kinds of travel. Spatial Patterns in the Built Environment We are studying the patterns of spatial distribution of certain important features of the Hawaiian built environment. The purpose of these spatial studies is to make possible comparisons across islands in order to examine our hypotheses about sequences of development. We are also improving upon the existing knowledge of land section boundaries, including the boundaries between ahupua'a. We employ existing designations of the boundaries between districts. Settlement Systems We employ archaeological surveys, excavations and land records to map the location of residential house locations, irrigation systems and ahupua'a boundaries in our focal locales. Housing locations are used to compare the settlement systems across districts and islands (e.g. Molokai as studied by Weisler and Kirch, 1985). The older published archaeological surveys of Kauai (e.g. Bennett, 1931) will be updated using more recent surveys, excavation reports and contract archaeology reports. The use of land records to improve upon ahupua'a boundaries requires searching the archives of several institutions on Kauai and in Honolulu. The boundaries of ahupua'a are allegedly shown on U.S. Geological Survey Topographical Quad maps as red lines. But these boundaries, which were mapped when the quad maps were first published in the early decades of the twentieth century, are notoriously inaccurate because the surveyors often mistook the boundaries of smaller parcels for ahupua'a boundaries. It is possible to correct the quad maps by using the land records produced in the middle of the nineteenth century at the time of the Great Mahele -- the land division and the granting of alienable titles. The land records of disputes over claims to parcels are a fertile source of information that can be used to determine the actual ahupua'a boundaries. This has already been done for the district of Hanalele on Kauai by Earle (1973;1978:120-8). We will do the land record research for our focal locales with the aid of our consultant, archaeologist Michael Kolb of the State Division of Historic Preservation. Several of the personnel in Michael Kolb's office are employed in this kind of work for other areas of the archipelago, and so the task has become much more efficient than when Earle did his original work on Hanalele. Subsistence and Production Systems We also employ archaeological surveys and excavation reports to study the location and distribution of terraced agriculture, irrigation systems, and fish ponds. The information contained in Bennett (1931) Handy and Handy,1973) on irrigation and terraces has already been coded by Kolb (1988) for the island of Kauai. We are updating this data using recent archaeological studies and adding fishponds. Much of the fishpond information for Kauai is contained in DHM Incorporated (1990). William Kikuchi's (1976) study of fishponds shows that these were an important aspect of the Hawaiian production system and that there were significant differences in the kinds of fish ponds built on different islands. One important local dimension of spatial differentiation is the seashore/upland (mauka/makai) relationship. It has been suggested by Spriggs,1988:65) that inland expansion led to a coastal-inland economic orientation which supplanted the former links between communities along the coast and led to ahupua'a self-sufficiency and endogamy. Spriggs also suggests that ahupua'a formation and investments in irrigation and terracing may have played a role in (and/or been caused by) the development of caste-like relations between chiefs and commoners (1988:70). We are examining the coastal/upland relationship in our focal locales. We also are considering the position of rather more isolated communities such as those far up Waimea canyon (Handy and Handy,1972:397) and in the isolated valleys of the Napali coast on Kauai in order to develop an understanding of local and regional core/periphery relations. We hypothesize that the fishponds are an important indicator of class stratification within a region. We will compare the numbers of Royal fishponds and the fishponds used by commoners, the kind of fish grown in fishponds (ex., mullet vs. milkfish, where the former was the restricted or "konohiki" fish,) and the physical maintenance of the ponds). Monumental Architecture: Heiaus To the spatial data sets for our focal locales we are adding the location, dimensions and architectural characteristics of heiaus, the temple platforms and enclosures that Hawaiians constructed for purposes of worshipping their gods. As with the other archaeologically known features of the built environment, the earlier surveys will be corrected and expanded based on the entire corpus of recent evidence, including contract archaeology. Help with this mapping project will be gained from our consultant, William Kikuchi, an archaeologist who has worked on Kauai for several decades. For the big island we will consult Ross Cordy. It is worth mentioning that there is no agreement among archeologists concerning the correlation between architectorial features and functions of temples. According to some, functionally specialized heiaus had well identifiable features; luakini temples could not be built just anywhere, but only upon sites formerly built on by the people of old. If so, then the larger heiau should be architecturally stratified, with the platforms and terraces of older heiau burried under later additions and elaborations. An opposite oppinion is represented by Stokes (1991:22): the different types of foundations described above seem to have had no connection with the classes of worship to which the heiau belonged. An intermediate solution is proposed by Valeri (1985): the Hawaiians had two heiau classification systems, which have not been clearly distinguished. The first was based on function - fertitily, production, or war. The second was based on architectural typology. The functional classification could be combined with the architectural to produce a wide array of subtypes. Sourcing of Basalt and Volcanic Glass In California we learned about archaeological techniques for studying trade patterns based on the sourcing and hydration dating of obsidian artifacts. In Hawaii the techniques of sourcing are still in their infancy. It has, however, been shown to be feasible to determine the quarrying sources of basalt used in the production of adzes (e.g. Cleghorn, et al, 1985) and the sources of volcanic glass used in the form of small flakes for cutting. The volcanic glass can also be dated by measuring the depth of hydration because volcanic glass slowly absorbs moisture when a newly broken surface is exposed to the elements. Hawaiian studies of basalt and volcanic glass have not yet produced much definitive knowledge about trade patterns and changes in these patterns over time (but see Weisler, 1990). It will eventually be possible to use this kind of evidence to shed light on local and regional patterns of exchange. Though this kind of research is beyond our competence we are endeavoring to motivate archaeologists to undertake further sourcing studies. This paper reports intentions, hypotheses and research plans. We are not far enough along in our research to report even preliminary results. At this point we hope to stimulate other scholars to consider our hypotheses and our approach in order to improve our Hawaiian research and to make comparisons with other cheifdom-based world-systems in Oceania and elsewhere. Notes References Abu-Lughod, Janet 1989 Before European Hegemony: The World System A.D. 1250-1350 New York: Oxford University Press. Algaze, Guillermo 1993 The Uruk World System. Chicago: University of Chicago Press. Apple, Russell, and William Kikuchi. 1975. Ancient Hawaii Shore Zone Fishponds; an Evaluation of Survivors for Historical Preservation. Honolulu: U.S. National Park Service. Barrera W. M., Jr., Anaeho'omalu: A Hawaiian Oasis (Preliminary Report of Salvage Research in South Kohala, Hawaii PAR 15, Bishop Museum, Honolulu, 1971. Barrera, W. M. and Hommon R. J. Salvage Archaeology at Wailau, Ka'u, Island of Hawaii. Report 72-1. Bishop Museum, Honolulu, 1972. Baugh, Timothy 1991 "Ecology and exchange: the dynamics of Plains-Pueblo interaction." PP107-27 in Katherine A. Spielmann (ed.) Farmers, Hunters and Colonists: Interaction between the Southwest and the Southern Plains. Tucson: University of Arizona Press. Beckwith, M. W. 1976. Hawaiian mythology. New Haven: Yale University Press. Bennett, Wendell Clark 1931 Archaeology of Kauai Bernice P. Bishop Museum, Bulletin 80, Honolulu: Bishop Museum. Brumfiel, E. M. 1988 "Factional competition in complex society" Pp. 127-137. In: Domination and Resistance. Ed. Miller, Rowlands M., Tilly C. London: Unwin Hyman. Chase-Dunn, Christopher 1988 "Comparing world-systems: toward a theory of semiperipheral development," Comparative Civilizations Review 19:29-66 (Fall). ________ (ed.) 1992 Review 15,3 (Summer) Special Issue on Comparing World- Systems. _________ 1993 "Intersocietal inequalities in small world-systems: Northern California." Final project report to the National Science Foundation. Grant # SES- 9110853 Chase-Dunn, Christopher and Thomas D. Hall (eds.) 1991 Core/Periphery Relations in Precapitalist Worlds. Boulder, CO.: Westview. __________________ 1993 "Comparing world-systems: concepts and working hypotheses." Social Forces 71,4:851-886 (June, 1993). __________________ 1994 "The historical evolution of world-systems" Forthcoming in Sociological Inquiry. __________________ Forthcoming, Rise and Demise: Comparing World-Systems, Boulder: Westview. Chase-Dunn, Christopher, S. Edward Clewett and Elaine Sundahl 1992 "A very small world-system in Northern California: the Wintu and their neighbors." Presented at the 57th annual meeting of the Society for American Archaeology, Pittsburgh, April 8-12. Working Papers Series #7, Program in Comparative and International Development, Department of Sociology, Johns Hopkins University. Chase-Dunn, Christopher and Mahua Sarkar, "Place Names and Intersocietal Interaction: Wintu Expansion into Hokan Territory in Late Prehistoric Northern California," A paper presented at the annual meetings of the Society for Economic Anthropology, Durham, N.H. April 23, 1993. Forthcoming in Peter Peregrine (ed.) Economic Analysis Beyond the Local System. University Press of America. Ching, Francis Jr. 1968 "Archaeological surface survey: Wailua State Park, Kauai." Department of Land and Natural Resources, State of Hawaii. Clark, Jeffrey T. 1988 "Paleodemography in leeward Hawaii." Archaeology in Oceania 23:22-30. Cleghorn, P.L., T. Dye, M. Weisler and J. Sinton 1985 "A prelinimary petrographic study of Hawaiian stone adze quarries." Journal of the Polynesian Society 94:235-51. Collins, Randall 1992 "The geopolitical and economic world-systems of kinship-based and agrarian-collective societies." Review 15,3:373-89 (Summer). Cordy, Ross 1981 A Study of Prehistoric Social Change: The Development of Complex Socieites in the Hawaiian Islands. New York: Academic Press. Davenport W. 1969. The "Hawaiian Cultural Revolution": Some Political and Economic Considerations. American Anthropologist 71:1-20. DHM Incorporated, 1990 Hawaiian Fishpond Study Honolulu: DHM Inc. Dickey, Judge Lyle A. 1917 "Stories of Wailua, Kauai." Annual Report of the Hawaiian Historical Society, Honolulu: Hawaiian Historical Society. Earle, Timothy 1973 "Control hierarchies in the traditional irrigation economy of Halelea district, Kauai, Hawaii." Doctoral dissertation. Ann Arbor: University Microfilms. Earle, Timothy 1977 "A Reappraisal of Redistribution: Complex Hawaiian Chiefdoms." Pp. 213-232 in T. Earle and J. Ericson (eds.) Exchange Systems in Prehistory. ______ 1978 Economic and Social Orgainzation of a Complex Chiefdom: the Halelea District, Kauai, Hawaii. Anthopological Papers #63, Ann Arbor: Museum of Anthropology, University of Michigan __________1980 "Prehistoric irrigation in the Hawaiian islands: an evaluation of evolutionary significance." Archaeology and Physical Anthropology in Oceania 15,1:1-28 (April). _________ "The evolution of chiefdoms," Pp. 1-15 in T. Earle (ed.) 1991 Chiefdoms: Power, Economy and Ideology. Cambridge: Cambridge University Press. Ellis, William 1963 [1827] Journal of William Ellis. Honolulu: Advertiser Publishing. Emory, Kenneth. Religion in Ancient Hawaii. Aspects of Hawaiian Life and Environment. Honolulu: Kamehameha Schools Press, 1965. Ermolaeva, Elena 1992 "Hawaii, 1678-1778" Unpublished ms. _______________ 1993 "Symbolic networks in the Hawaiian world-system." A paper presented at the annual meeting of the World History Association, Honolulu, June 27. _______________ 1994 "Chiefdom to state: cultural resistance in the Hawaiian world-system." Ph.D. dissertation proposal, Department of Sociology, Johns Hopkins University. _______________ 1994 "In the support of the concept "culture fatigue". A paper to be presented at the Annual meeting of American Sociological Association, August, Los-Angeles. Fornander, Abraham 1880 An Account of the Polynesian Race. Volume 2. London: Trubner. Freidman, Jonathan 1981 "Notes on structure and history in Oceania." Folk 23:275-95. ________1988 "No history is an island: review of Sahlins Islands of History."Critique of Anthropology 8,3:7-39. Friedman, Jonathan and Michael Rowlands 1977 "Toward an epigenetic model of the evolution of 'civilization.'" Pp. 201-78 in J. Friedman and M. Rowlands (eds.) The Evolution of Social Systems. London:Duckworth. Ii, John Papa 1959 Fragments of Hawaiian History. Honolulu: Bishop Museum Press. Hall, Thomas D. 1991 "The role of nomads in core/periphery relations," Chapter 7 in Chase-Dunn and Hall, 1991. Hall, Thomas D. and Christopher Chase-Dunn 1993 "The world-systems perspective and archaeology: back to the future" Journal of Archaeological Research 1,2:121-143. Hammatt, Hallett H. 1988 "Archaeological reconnaissance of the upper Wailua hydroelectric project, Wailua, Kauai". Prepared for Bonneville Pacific Corp., Honolulu: Cultural Surveys Hawaii. Handy, E.S. Craighill and Elizabeth Green Handy 1972 Native Planters in Old Hawaii: Their Life, Lore and Environment. Bernice P. Bishop Museum Bulletin #233, Honolulu: Bishop Museum Press. Handy, E. S. Craighill and Mary Kawena Pukui 1972 The Polynesian Family System in Ka'-u, Hawai'i. Rutland, VT.: Charles Tuttle. Henry, Donald O. "Preagricultural sedentism: the Natufian example." Pp. 365-384 in T. Douglas Price and James A. Brown (eds.) Prehistoric Hunter -Gatherers: The Emergence of Cultural Complexity. New York: Academic Press. Hommon, Robert 1976 "The formation of primitive states in pre-contact Hawaii." Unpublished Ph.D. thesis, University of Arizona Hommon, Robert J. 1975. Use and Control of Hawaiian Inter-Island Channels. Polynesian Hawaii: A.D. 1400 - 1794. Office of the Governor, Joesting, Edward 1984 Kauai: A Separate Kingdom. Honolulu: University of Hawaii Press. Johnson, Allen W. and Timothy Earle 1987 The Evolution of Human Societies: From Foraging Group to Agrarian State. Stanford: Stanford University Press. Kaeppler, Adrienne 1982 "Geneology and disrespect: a study of symbolism in Hawaiian images." RES 3:82-107 (Spring. Kamakau, Samuel.M. 1961. Ruling Chiefs of Hawaii Honolulu: Kamehameha Schools Kamakau, Samuel M. 1964. Ka po'e kahiko: the people of old. Honolulu: Bishop Museum Special Publication No. 51. Kelly, Marion. 1980. Land Tenure in Hawai'i. Amerasia Journal 7 (Fall/Winter): 57-73. Kikuchi, William K. 1976 "Prehistoric Hawaiian fishponds." Science 193:295-9. _______1984 "The petroglyphs of Wailua" Archaeology on Kauai 11,2 Kikuchi W. K. and J. C. Belshe. Examination and Evaluation of Fishponds on the Leeward Coast of the Island of Hawaii. Report for Hawaii County Planning Commission, Hilo, 1971. Kirch, Patrick Vinton 1984 The Evolution of Polynesian Chiefdoms. Cambridge: Cambridge University Press. __________1985 Feathered Gods and Fishhooks: An Introduction to Hawaiian Archaeology and Prehistory. Honolulu: University of Hawaii Press. 1988"Circumscription theory and sociopolitical evolution in Polynesia." American Behavioural Scientist 31,4:416-427. __________1990a "Regional variation and local style: a neglected dimension in Hawaiian prehistory." Pacific Studies 13,2:41-54 (March). __________1990b "The evolution of sociopolitical complexity in prehistoric Hawaii: an assessment of the archaeological evidence." Journal of World Prehistory 4,3:311-45. __________1991 "Chiefship and competitive involution: the Marquesas Islands of eastern Polynesia," Pp. 119- 145 in T. Earle (ed.) Chiefdoms: Power, Economy and Ideology Cambridge: Cambridge University Press. __________ 1992 Anahulu: The Archaeology of History. Chicago: University of Chicago Press Kolb, Michael 1988 "Heuristic concepts of space and distance in archaeology: a case study of prehistoric labor investment from Kauai, Hawaii." unpublished seminar paper, UCLA _________1991 Social Power, Chiefly Authority and Ceremonial Architecture in an Island Polity, Maui, Hawaii. Ph.D. Dissertation, University of California, Los Angeles. ann Arbor: University Microfilms. _____________1992 "Diachronic design changes in heiau temple architecture from the island of Maui, Hawaii" Asian Perspectives 31,1:9-37. ______________1993 "Ritual activity and chiefly economy at an upland religious site on Maui, Hawaii." Journal of Field Archaeology. Kristiansen, Kristian 1987 "Centre and periphery in Bronze Age Scandinavia." Pp. 74-85 in Michael Rowlands, Mogens Trolle Larsen and Kristian Kristiansen (eds.) Centre and Periphery in the Ancient World. Cambridge: Cambridge University Press. ________1991 "Chiefdoms, states and systems of social evolution," Pp. 16-43 in Timothy Earle (ed.) Chiefdoms: Power, Economy and Ideology. Cambridge: Cambridged University Press. Lenski, Gerhard 1966 Power and Priviledge. New York: McGraw-Hill Linnekin J., 1990. Sacred queens and women of consequence. Rank, gender and colonialism in the Hawaiian Islands. Ann Arbor, The University of Michigan Press. Malo, David 1951 [1898] Hawaiian Antiquities (Moolelo Hawaii). Translated by Nathaniel B. Emerson. Honolulu: Bishop Museum. Mann, Michael 1986 The Sources of Social Power. Cambridge: Cambridge University Press McKinzie, Edith Kawelohea 1983 Hawaiian Genealogies Volume 1, Edited by Ishmael W. Stagner II, Honolulu: University of Hawaii Press. __________1986 Hawaiian Genealogies Volume 2. Newman, T. Stell. 1970. "Makai-Mauka; Fishing and Farming on the Island of Hawai'i in A. D. 1778." Ph.D. Dissertation (Anthropology), University of Hawaii. O'Brien, Patricia "The world-system of Cahokia within the middle Mississippi tradition," REVEIW 15,3:389-418 (Summer). Peregrine, Peter 1992 Mississippian Evolution: A World- Systems Perspective. Madison, WI.: Prehistory Press Pukui, M. K. and Elbert S. H. 1973. Hawaiian Dictionary. Univ. of Hawaii Press. Honolulu. Rosendahl 1972. "Aboriginal Agriculture and Residence Patterns in Upland Lapakahi, Island of Hawaii." Ph. D. Dissertation (Anthroplogy), University of Hawai'i. Sahlins, Marshall 1958 Social Stratification in Polynesia. Seatlle: American Ethnological Society. _________1972 Stone Age Economics. Chicago: Aldine. _________1981 Historical Metaphors and Mythical Realities: Structure in the Early History of the Sandwich Islands Kingdom. Ann Arbor:University of Michigan Press. ________1985 Islands of History. Chicago: University of Chicago Press. ________1992 Anahulu: Historical Ethnography. Chicago: University of Chicago Press. Sahlins, Marshall D. and Dorothy Barrere. 1973. William Richards on Hawaiian Culture and Political Conditions of the Island in 1841. The Hawaiian Journal of History, Vol. 7, pp. 18-40. Honolulu. Schortman, Edward and Patricia Urban 1992 Resources, Power and Interregional Interaction. New York: Plenum. Smith, Walter J.Legends of Wailua Lihue, Kauai: Garden Island Publishing Co. Spriggs, Matthew 1988 "The Hawaiian transformation of Ancestral Polynesian society: conceptualizing chiefly states." Pp. 57-73 in John Gledhill, Barbara Bender and Mogens Trolle Larsen (eds.) State and Society , London: Unwin Hyman. Stauffer, Robert 1992 "Draft notes for an annotated bibliography for Wailua materials" Hawaii Division of State Parks. Stokes, John F.G. Heiau of the Island Hawaii. A History Survey of Native Hawaiian Temple Sites. Bishop Muse um Bulletin in Anthropology 2, Bishop Museum Press, Honolulu, 1991. Thrum, Thomas G. 1907 "Tales from the temples," Pp. 49- 69 in All About Hawaii from Thrum's Hawaiian Annual , Honolulu: Star-Bulletin. Thrum, Thomas G. 1910. "Heiaus: Their Kinds, Construction, Ceremonies." Hawaiian Annual for 1910. Honolulu: Thrum's Hawaiian Annual, pp.53-71. Tilly, Charles 1990 Coercion, Capital and European States, AD 990-1990. Cambridge, MA. Blackwell. Upham, Stedman 1990 The Evolution of Political Systems: Sociopolitics in Small-Scale Sedentary Societies. Cambridge: Cambridge University Press.. Valeri, Valerio 1972 "Le fonctionnement du systeme des rangs a Hawaii." L'Homme 12:29-66. Valeri, Valerio. 1984. "The conqueror becomes king: A political analysis of the Hawaiian legend of 'Umi". In: Transformations of Polynesian society, ed. Antony Hooper and Judith Huntsman. Wellington: Polynesian Society. ________1985 Kingship and Sacrifice: Ritual and Society in Ancient Hawaii. Chicago: University of Chicago Press. Wallerstein, Immanuel 1984 "The quality of life in different social systems: the model and the reality." Pp. 147-58 in I. Wallerstein The Politics of the World- Economy Cambridge: Cambridge University Press. Webb, M.C. 1965. The abolition of the taboo system in Hawaii. Journal of the Polynesian Society 74:21-39. Weisler, Marshall 1990 "Sources and sourcing of volcanic glass in Hawaii:implications for exchange studies." Archaeology of Oceania 25:16-23. Weisler, Marshall and Patrick Kirch 1985 "The structure of settlement space in a Polynesian chiefdom: Kawela,Molokai, Hawaiian Islands." New Zealand Journal of Archaeology 7:129-158. Wichman, Frederick B. Kauai Tales Honolulu: Bamboo Ridge Press. Wilkinson, David 1991 "Cores, peripheries and civilizations." Pp. 113-166 in Chase-Dunn and Hall (eds.). Yent, Martha 1989 "Mapping and testing of Hikinaakala heiau and Hauola, Lydgate Area, Wailua River State Park, Wailua, Kauai" Hawaii Division of State Parks.