Note: This page is part of the archives of the Phrontistery, but is no longer being updated.
This is a revised version of a paper originally written in 1995 as part of an honours degree project. It is somewhat academic in nature, so be warned. While my opinions on some of the issues discussed herein have changed somewhat, my theoretical interests are substantially similar to those in this paper.
I. IntroductionNo thing great is created suddenly, any more than a bunch of grapes or a fig. If you tell me that you desire a fig, I answer you that there must be time. Let it first blossom, then bear fruit, then ripen. Epictetus, The Discourses, I, 15 The study of state formation, and social evolution in general, has fallen out of fashion in recent years, particularly among ethnologists. However, there is no methodological justification for this collapse, although some researchers have been convinced to abandon this topic. Relying on evidence from archaeology, ethnohistory and ethnology, I argue that despite the serious errors committed by some scholars of political evolution, the general evolutionary framework is essentially valid. I feel that an approach to state formation generally informed by systems theory is essential and workable; however, consideration must be given to the specific historical conditions extant in any specific state society. We must come to recognize that societies are neither infinitely particular nor widely generalizeable, and that while many causal factors may lead to state formation, trends of increasing political complexity are evident in widely divergent societies. I offer the proposal that many aspects of societies which are invisible to archaeologists can be fruitfully studied by ethnographic and ethnohistorical research, and that interdisciplinary co- operation is essential if the evolutionary framework is to be refined. Upon opening a discussion of a topic which is no longer popular in the anthropological literature, one must consider whether it has failed due to the irrelevance of the paradigm, or due to other circumstances. What has gone wrong with the study of the state among ethnologists? Certainly, one might argue validly that overly broad synchronic comparisons have been made between societies separated widely in their history and geography. The method adopted by Claessen and Skalník in their Early State Project which seeks to define the state by a number of structural characteristics, while useful for purposes of description, has little explanatory value in and of itself, although it makes no such claim (Claessen 1978). Other scholars, arguing purely from (often dogmatic) theoretical positions, which may require that the data be made to fit the theory, seem always to fail the test of generalizability. Does it make sense, however, to reject the subject altogether, or should we seek rather to refine our theories? It is the second approach which seems most fruitful, and which I adopt throughout this paper. It is apparent, upon examining recent scholarly works on the study of the state, that archaeology has come to have a greater effect on the topic within anthropology. Indeed, Karl Hutterer has argued that sociocultural anthropologists no longer have any role to play in this subject. I dispute strongly the claim that "it will be necessary to shift the responsibility from ethnology to archeology" in order to construct evolutionary sequences (Hutterer 1991: 237). It is essential, however, that the ahistorical, synchronic ethnographic models which are used primarily for taxonomic purposes be re-examined in light of the specific trajectories developed by ethnohistorians and archaeologists. I should also make clear to what I am referring as "a state". As a preliminary definition, I suggest the following: a form of government in which there exists centralized authority, hierarchical differences in access to power and basic resources between individuals, reinforced by institutionalized coercion, usually accompanied by the decline of kin structures. As such, there is no reason to see these characteristics of "the state" as more important than any other agglomeration of characteristics. However, the recurrence of particular sets of traits in diverse areas of the globe (Mesoamerica, South America, Africa, India, China, etc.) indicates that there may still be some use in talking about "state formation" as opposed to the more general "political evolution". Political evolution, then, is the process by which societies change their structures through time, generally in the direction of greater complexity. I do not intend to discuss issues of social collapse or "devolution" in this paper; certainly, the value of such terms within an evolutionary framework can be debated. There appear to be four fundamental illusions which some state theorists have fallen under. Firstly, too much emphasis has been placed on the dichotomy between coercive and integrative theories of the state, leading to overly polarized positions which obscure the complex reality of societies. Next, false notions of hierarchy and progress have led some to see state formation as an end point of social evolution rather than part of an ongoing process of increasing social size and complexity. Third, the systemic forces which cause social evolution to occur have been misidentified and misinterpreted by scholars under the influence of paradigms from elsewhere in the life and social sciences. Finally, many have existed under the illusion that state formation is a process which imposes centralization upon non-state societies, rather than being an outgrowth of their existing institutions and historical practices. I intend not only to dispel these four illusions but to replace them with constructive means by which anthropologists should revive state formation as a worthwhile area of study.
II. Illusory Dichotomy: Coercion and integration in early statesThe chief foundations of all states, new as well as old or composite, are good laws and good arms; and as there cannot be good laws where the state is not well armed, it follows that where they are well armed they have good laws. Niccolo Machiavelli, The Prince, ch. 12 Until very recently, theorists on the topic of state formation have largely been polarized into two schools: conflict theorists, who emphasize the coercive role of the state in increasing and legitimating inequality, and integration theorists, who point to the state's greater redistributive ability as its primary causal feature. Neither of these two extreme positions is wholly satisfactory in every state, and I would hazard that neither is sufficient explanation in any state. As Machiavelli astutely noted almost four centuries ago, both laws and arms make a good prince, and it is impossible to have one without the other.(1) The history of the study of states could theoretically be extended back to the ancient philosophers, and thenceforth to the Enlightenment philosophes such as Hobbes, Locke, and Rousseau. It is evident that the tension between social contract, or integration models, and social conflict, or coercive models, extends throughout this period, as argued effectively by Haas (1982). In the anthropological literature on states, one would expect no objection to the claim that most Marxists and materialists adopt coercive models, while those emphasizing managerial functions would propound integrative theories of state formation. However, the considerable evidence on both sides of the debate (which I do not intend to review in detail) immediately leads one to suspect that neither theory is wholly satisfactory. Scholars who argue for the primacy of warfare as causal agent in state formation have generally combined the two orientations of integration and conflict due to the nature of their "prime mover". Webster notes that warfare facilitated "the emergence and survival of privileged managerial groups" (1975: 469) in order to co-ordinate military efforts and marshal resources. However, this integrative role is contrasted with his assertion that "Concentration of power is integrally related to concentration of wealth, and derives from it" (1975: 468). This apparent contradiction suggests that while initially, the need for integration enabled meritorious or cunning individuals to acquire control over these resources, coercion and differential access to resources reinforced this position due to the resulting power differential. Warfare is surely a necessary but insufficient condition for state formation. However, this position, which has been advocated by Cohen (1978: 46) as part of complex events which synthesize various factors leading to the centralization of authority, in no way argues that warfare is a prime mover in social evolution. Neither do I accept the conclusion of Naroll and Divale that because cross-cultural studies do not show a correlation between warfare and territorial growth that "peaceful diffusion is more likely than warfare to be the primary mechanism of selection in cultural evolution" (1976: 121). The causes and consequences of warfare are both varied and closely intertwined with other social events in societies. Warfare is neither indicative of "coercion" or "integration", but requires that both factors operate simultaneously. Surely we can accept that war has serious influences on political leadership and social structure without arguing that its presence somehow determines the future of the society. The greatest methodological problem which scholars have faced in interpreting archaeological data on archaic states is the establishment of causal links which would support or reject the conflict or integration positions. How, for instance, can data whose date can only be placed accurately within a century be used to establish whether warfare preceded the state, or the other way around, and so on? This problem is faced most clearly by Haas (1982) in attempting to re-evaluate the archaeological data on the "primary civilization", and is not completely resolved. However, we do have a number of instances of historical state formation in Polynesia, Africa and elsewhere, for which such details are known. Might it not be productive, then, to examine the better known ethnohistorical examples in order to aid in constructing analogies in order to understand the lesser known archaeological states? The best evidence we have, then, suggests that both coercion and integration have their place in state formation. I offer the complementary proposition that the dichotomy between the two is both unnecessary and unimportant. Whether a mechanism is coercive or integrative is not the primary determinant of its importance for the formation of states. The excessive focus on the polar opposition of these two principles may stem from the philosophes and their proposals of "social contract" and "social coercion", neither of which were based on any real evidence, anthropological or otherwise. Thus, while we may maintain that organizing irrigation systems is "integrative" while enslaving one's conquered enemies is "coercive", I offer as a testable hypothesis the notion that there may be more important guiding principles of state formation as yet undiscovered.
III. Illusory Pyramid: Egalitarianism and the Growth of ComplexityThere is no steady unretracing progress in this life; we do not advance through fixed gradations, and at the last one pause ... Herman Melville, Moby Dick, ch. 111 A great deal of ink has been spilt over the question of typology of societies with regards to their political evolution. The classic theories of Service (1962) and Fried (1967) are seminal in this matter. Each is generally concerned with how egalitarian (or band) societies evolve through intermediate stages into modern states. I agree with Upham that Service's band-tribe-chiefdom-state and Fried's egalitarian-ranked-stratified-state models "do not exist in their entirety in the real world; they are categorical representations of reality that condense meaningful variability" (Upham 1990: 89). The fact that these typologies are constructions does not, however, eliminate their value as analytical tools, as long as we understand that they have limited explanatory value and do not reflect the fullness of social reality. The term "state" certainly should not be used as a predictor of the existence or absence of specific traits associated with state societies. It should be emphasized that "egalitarian" societies are by no means characterized by complete equality but of course are stratified in terms of authority by age, sex and personal ability (Service 1975: 49). Trigger reports that in many cases, even economic equality is not achieved within these societies due to differences in merit and the perceived generosity (or lack thereof) of individuals (1985: 50). The evolution of political inequality should therefore not be treated as the origin of this inequality, which has its basis in nature, but rather, in the true meaning of evolution, its "unfolding". Likewise, "complex" societies often maintain considerable vestiges of their earlier egalitarian condition, as evinced by the retention at least somewhat of kinship and reciprocity as important principles in states. I am, then, referring only to ideal types in the use of "egalitarian", "complex" and other such terms, while acknowledging that real-world societies are always found somewhere on a continuum of complexity. Regardless of the existence of broad stage-like categories, the way in which these categories are related to one another has been fundamentally misinterpreted. It is customary to imagine a "hierarchy" of political evolution, with states at the pinnacle, chiefdoms and tribes in the middle and band societies at the bottom. This pyramidal analogy can be found in many structural Marxist arguments, such as that of Friedman (1975) and Kristiansen (1991) which place the state at the centre of a spatial hierarchy with chiefdoms and tribes on the periphery. It is essential, however, that we examine the consequences of this geometric-spatial model of social evolution for the errors of cause and effect which it engenders. Kristiansen asserts that chiefdoms "are in many cases, perhaps most cases, a secondary development" and that simpler societies are merely devolved polities separated by geography or other circumstances from the states upon which they are dependent" (1991: 27). This argument is wholly teleological and obscures the processes by which such hierarchies of societies (which certainly exist in colonial and post-colonial regions) are formed. How, then, do we explain the existence of chiefdoms in places such as pre-contact Polynesia, where states were unknown and ties to outside state structures non-existent? It is extremely arrogant of those in state societies to assume that merely because they can enforce dependence on smaller-scale societies that such societies are somehow created by the spatial relations between core and periphery.(2) Such attitudes also reveal a hidden bias: that states are somehow the culmination or end-point of processes of social evolution, and that modern nation-states have somehow transcended the evolutionary paradigm. It is entirely possible for change to be directional without being teleological. I wish to make my point emphatically clear: political systems are not arranged hierarchically, excepting specific instances where there is a relationship between a central (core) state and various less complex polities in the periphery. Gall and Saxe offer the following caution regarding taxonomic classifications: Since the taxonomy is concerned with trends of increasing complexity, we should not assume that this is the only trend in nature or that it is "progressive" in nature.... Nor does this taxonomy substitute for an analysis of cultural processes. Rather it provides a frame for assessing what is there and what changes have taken place in the relationships between the variables as it developed and why (1977: 257). A hierarchy is a real structure, expressing relations of information exchange and power between real entities. Hierarchical constructions of evolutionary taxa do not express such relations, but rather describe a fictitious evolutionary pattern which inevitably leads towards the state. There always exists a danger that this model will become a straw man by which some may seek to discredit the notion of the state. We should seek to express the concept of complexity in quantifiable (if not perfectly so) terms, before we may continue with the study of social evolution. How, then, should we describe the directional process of increasing complexity which characterizes social evolution? The most profitable course, I think, is to adopt linear models which have descriptive value based on real-world data such as settlement size, subsistence form and the presence or absence of political traits. While researcher bias certainly may influence a measurement for a single society, if trends are visible throughout many societies, the cross-cultural research will have described something with at least some basis in social reality. This sort of project is undertaken by Carneiro (1967) using forty-six randomly selected societies, in which population size is plotted against the number of organizational traits present. Indeed, a fairly reliable logarithmic linear correlation between the two variables exists. What the sorts of studies that Carneiro and others have done accomplishes is to establish variation between societies. What it does not do is establish the variation of this change over time. However, as Jerome Rousseau notes, "we do know something of archaic relations of production", and may reasonably assume that hunter-gatherers of today, while certainly different from archaic ones, possess the same basic subsistence forms and political traits (Rousseau 1985: 38). We know further that "all primitive nonegalitarian societies were not primordial, in the sense that they all were preceded by egalitarian societies" (Khazanov 1985: 94); that is, the original state of humanity was (relative) egalitarianism, and all non- egalitarian societies are transformations of earlier egalitarian ones. These factors, then, form the basis for the claim that societies generally increase in complexity through time. (3) However, once we have established the linear model as one based in factual data rather than speculation and teleology, what do with it? It is important to realize that we cannot simply read the course of human history by travelling up our new line of state formation. There are inevitable dips and gullies in the process of state formation, as cities collapse, chiefdoms divide, and polities everywhere experience periods of stability with little structural transformation. We must be cautious, recognizing that a descriptive model can do nothing for our understanding of process and causation until we examine the specific histories of state societies. However, this caution has not been heeded by all scholars. One of the most egregious conclusions reached by those scholars who place descriptive characteristics of states above the factors causing them is that of Carneiro. Carneiro notes that the number of autonomous political entities extant in the world has declined steadily from the beginning of humanity to the present, and that as the number of states grows smaller, their size grows larger - fair enough, so far. He then asserts that the curve of declining polities leads inevitably to one conclusion: the emergence of a world state (Carneiro 1978: 222). To do so, he invokes the principle of competitive exclusion, an idea borrowed from biological evolution which states that two species cannot coexist using the same niche indefinitely, and one will eventually prevail, forcing the other to evolve or die out (Carneiro 1978: 221). Carneiro's contention that there is an inevitable decline in the number of political groups requires that one accept that independent aggregates existing in pre- and proto-human periods were also "political groups". Otherwise, one would have to acknowledge that there was a time in human history when there were no political groupings at all, followed by an increase, which produces a curve rather unlike Carneiro's. Taken to its extreme, all biological populations are in fact autonomous political groups in Carneiro's sense of the term. We must also consider whether the conditions of harsh competition and warfare required for this model actually existed. Furthermore, it is by no means a foregone conclusion that the world state is an inevitable result. Carneiro here confuses directionality with teleology, a similar error to the pyramidal fallacy of the state described above, when it is entirely unknown where the curve of autonomous polities might be leading. We cannot predict "where" social evolution may be going, simply because it is not "going" anywhere at all, but rather processes acting at the present act to shape the future without any inherent goal or end point. The conclusion which we must draw from this analysis is clear. The pyramid must be toppled, not only because it presents a misguided geometry of political systems, but also because this analogy has resulted in the rejection of social evolution entirely as a legitimate process. Yet the limitations of the linear model of evolution must also be recognized. There is surely a trend towards increasing complexity among societies over time; however, due to the vagaries of societal collapse and stagnation, we cannot clearly place the descriptive template of settlement size and political differentiation atop the overall causal scheme of political evolution. There is the ever-present danger that, as Carneiro and others have done, we will extend the general evolutionary curves into predictive and explanatory models which ignore or devalue process. What, then, can be done to establish the process by which this evolution occurs?
IV. Illusory Force: Processes and Causes of State FormationEvolution is not a force but a process, not a cause but a law. John, Viscount Morley of Blackburn (1838-1923) The linear model of state formation which I have presented suffers from a further defect of great consequence, hopefully not a fatal one. While the linear model has descriptive value, it has little or no explanatory value. Merely establishing that the Chinese state has greater differentiation among its political officials than do !Kung San bands tells us nothing as to what causal factor or factors led to Chinese society being as complex as it is. It also cannot explain why the !Kung San did not develop inequality to the same degree. Although many scholars from different fields have tried to explore the mechanisms of state formation, no synthesis has been reached, as many of these models are purely speculative and have no basis in ethnographic data. This flaw cannot be easily addressed using synchronic ethnographic studies, but must be examined in light of historical and archaeological data. The question of evolutionary mechanism is central to social evolutionary theory. On the one hand, biological models of adaptation have been introduced in many forms to explain how each stage of political evolution is "selected for" by virtue of their adaptive advantages. On the other, we have materialist arguments from both cultural ecologists and some Marxists which see the alteration of political structures as a response to environmental conditions or to the mode of production of the society. Both of these arguments have a certain amount of merit; selection of some sort obviously occurs among societies, while material conditions undoubtedly influence the emergence of states due to their centrality for human existence. While certainly these arguments have lost much of their explanatory force as universal predictors, it would be erroneous to assume that they are no longer advocated. Indeed, due to the relative parsimony of biological and "crude" materialist arguments, they retain a degree of respectability as theories of state formation that I do not feel is warranted. Attempts by evolutionary biologists to explain the emergence of the state have used the term "evolution" in a purely reductionist sense of an increase in adaptive fitness in a particular environment. Jack Hill, a parasitologist, suggests that it is a given that human beings will inevitably seek greater prestige for the reproductive advantage it will give them. States, then, are the culmination of a never-ending quest for greater and greater prestige in order to have more offspring (Hill 1994: 224). Hill's analysis, however, is wholly teleological as it assumes that the end result (the state) is "adaptive" and then determines why it might be adaptive for people to seek out this adaptive state. Why, for instance, do states not evolve among chimpanzees, since it is also true that the great apes have dominance hierarchies that give them greater reproductive success? Have hunter-gatherers such as the !Kung San somehow missed out on the opportunity to gain reproductive success by not evolving to a "state" level of organization? While Hill is certainly right that a desire for power and prestige is an essential part of any leader assuming command of a group, any increase in reproductive fitness in the leader would necessarily be offset by the loss in reproductive success of the underlings. In fact, why would close-knit bands of kin have any competition for prestige at all, since one would be reducing one's inclusive fitness by competing with close kin? Hill does not account for the acceptance of the non-prestigious of the laws and ordinances of those who gain prestige -why would one not simply leave, or eliminate one's leader, or use other means to reduce the resultant inequalities? He must assume that all states evolved in highly circumscribed environments with strong competition for prestige, a situation for which he provides no evidence. The final blow to Hill's argument derives from the fact that in hunter-gatherer bands stratified by age and sex, and thus young men know that they have only to wait for a few years until their status rises, along with their reproductive success. There is no reason to overthrow the leadership in such a stable situation, and there is no evidence that this is how social evolution actually occurred. While there are certain correlations between the acquisition of power and the continuation of one's line, Hill has reduced a complex situation to such a ridiculous degree that his argument has no merit whatsoever. Roger Masters equates political evolution with game theory in evolutionary biology, specifically the classic Prisoner's Dilemma. Masters contends that the inevitable result of individual rational choices would be some sort of "social contract" in which it is "natural" that the benefits of law and government outweigh its costs (Masters 1982: 446-447). His sources on state origins are those of Hobbes, Rousseau, Kant and Hegel, without consideration of the anthropological evidence which suggests that the "social contract" is an illusion of Western philosophers rather than a real phenomenon. Masters' contention that the state is caused when equals gather together to determine codes of proper or "altruistic" conduct supposes that this "natural state of equality" existed (Masters 1982: 442), while I have suggested above that it did not. This analysis falls prey to the same sort of fault as Hill's, in that it may explain something about human sociality but that it fails to consider the non-emergence of states as an equally relevant fact as their actual formation. It is clear that while there may be correlations between state formation and biological adaptive advantages, evolutionary biologists have failed completely to produce adequate causal agents for political evolution. Having explored the ridiculous arguments from the biologists, let us turn now to the materialists, whose arguments regarding state formation can be separated roughly into two distinct groups. The first, Marxist and neo- Marxist scholars, argue that material conditions and the mode of production of a society are essential in transforming its political structures and ideological behaviour. The other, called by some "vulgar Marxists", are cultural ecologists who argue that a society's environmental conditions are the primary determinants of its social structure and evolution. While both of these arguments fail, they fail for different reasons, as I will detail below. Marxist arguments on state formation no longer rely extensively on the writings of Marx himself; certainly, few would be found who still argue that all societies must pass through a capitalist phase before reaching "the end of history" with the advent of communism. Charles Maisels, a neo-Marxist, rejects the notion that there is one universal line of evolution. Instead, he argues that there are at least two divergent lines: one, that of east Asia and the Andes, leading to the so-called "Asiatic" mode of production, and the other, found in Mesopotamia, leading to a "city-state" system of urban centers (1987: 331). The Marxist line of thought regarding state formation generally requires the conversion of human beings and their labour-products into use by wealthy elites (see Tosi 1976). As such, it is a "coercion" theory which has as its primary mechanism the interaction of material and human factors. Marxists have certainly come farther than many in their emphasis on historical processes and structural and ideological change in societies. However, many Marxist scholars still have a tendency to see state formation according to an orthodox Marxist scheme for which there is little evidence. (4) For instance, Patterson's description of the "pre-capitalist" Inca state requires that one accept that "capitalism" is the inevitable successor to whatever sort of "pre-capitalism" existed among the Inca (Patterson 1991). I believe this to be entirely unsupported. Furthermore, the direct translation of material conditions into ideology seems overly simplified; while certainly one's subsistence method and production relations influence one's way of thought, the notion that they are thereby determined seems a bit far-fetched. Many cultural ecologists have been centrally concerned with the evolution of states, arguing that the primary mechanisms of cultural evolution and the creation of states derive from the physical environment. Heirs to the legacy of White and Steward, ecological models are not necessarily deterministic, and are often quite refined with regards to the interplay of cultural and material factors. Carneiro's argument that state formation can only occur in highly circumscribed physical environments is of this sort, as it recognizes that there is always the possibility that "social circumscription" by other human populations may have the same effect (Carneiro 1970: 737). Few scholars would now argue for a direct translation of ecological circumstances into political expansion or contraction (however, see Harris 1977). Terry Rambo, a cultural ecologist, presents his case eloquently: To a considerable extent, we have moved beyond use of gross environmental variables (e.g. rain forests, deserts) into making of quite fine-grained descriptions of the specific local habitats within which specific cultural groups are functioning (Rambo 1991: 92). Environmental and ecological factors are such persuasive candidates for "prime mover" status simply because they are so clearly related to means of subsistence. The development of agriculture is surely the key to the production of surplus, which is a necessary condition if a society is to have non-producers among its leaders. However, most ecologists know very well that agriculture is not adopted in many places where a surplus might be produced, simply because of the intensive labour required (see Boserup 1965). We also need to explain, then, why states did not form in every place where the environment was suitable; to do so, we must regard environmental factors as necessary but insufficient conditions, as with so many other factors. While ecologists have gone beyond the Marxists in recognizing that societies are not simply material conditions writ large, Marxist thought is much more amenable to the sort of historical thinking which I feel is essential in order to understand the interaction of variables in social systems. We have seen that the evolution of the state has very different features from evolution in nature. It should also be evident that states do not represent crude transformations of material and ecological conditions, as advocated by some cultural ecologists and Marxists. Nevertheless, it is clear that societies represent systems of some sort, and are not merely agglomerations of individuals even in the most undifferentiated of peoples. By "system" I mean only any organized assembly of inter-related components, and as such many types of systems (mechanical, organic, astrophysical, social, etc.) can be identified. A key insight which is raised by Walter Buckley is that a system "is not to be confused with the structure or organization its components may take on at any particular time" (1967: 5, italics in original). It may thus be said that while there are evolutionary systems or social systems, there are only state structures which are the physical manifestations of a social system at any given time. Lest the reader think I am advocating some sort of metaphysical force behind such systems, I must emphasize that systems are entirely the result of human actions, the natural and constructed environment and the interaction of the two. Nevertheless, the consequences of systems and the structures they produce are often indifferent to or even contradictory to the desires of their human components. Anyone who has ever invested in the stock market is well aware that it is a system in which each actor desires financial gain, and yet often the market can crash, causing financial harm to all individuals concerned. Most social systems are, of course, far more complex and cannot be so easily described. It is the job of the systems theorist to explain how human actors and actions and the physical environment are the sole causal agents which result in the creation of state structures. An important (and much misused) concept in systems theory is that of feedback, of which positive (deviation-enhancing) and negative (deviation- reducing) are the two forms. Buckley notes that feedback should not be "equated simply with any reciprocal interaction between variables" (1967: 52). The classic example of negative feedback is a thermostat; when a change to the initial condition of the system occurs (increase or decrease in temperature), the system is sensitive to the change and responds by reducing the deviation (either by sending a signal to introduce more heat or by sending another signal to turn the heat off). To witness positive feedback in action, the stock market example above is instructive. As soon as any slight decrease occurs in a stock or a negative circumstance arises, the system's sensors (human actors) react by attempting to sell the stock before a more drastic decrease occurs. Since the market then has less demand than supply, the overall effect (feedback response) is an increased rate of decline. In the case of state formation and social evolution, it is positive feedback with which we are most concerned, although negative feedback and homeostasis are important and ongoing processes which maintain stability in all societies, whether or not they are states. Perhaps the most influential systems theorist working within a cultural evolutionary framework has been Kent Flannery (Flannery 1972). Flannery suggests that cultural evolution is most adequately explained "multivariant causality" in which various factors are interrelated in a complex of feedback loops, causal chains and other relationships (Flannery 1972: 423-424). By rejecting "prime movers" such as population pressure or warfare, without denying the importance of such factors, multivariant causality allows that these factors may operate in different directions, and that each social system is unique. One may debate the value or existence of Flannery's "generative rules" of the state as too simplistic or incomplete scenarios. However, this objection does not require the rejection of the theory, but rather careful research to test its validity. Systems theory has been used with moderate success to analyze instances of state formation among archaic civilizations which are known archaeologically. Examining evidence in southwest Iran from Mesopotamian sites, Wright and Johnson conclude that "only multiple variable changes" could produce vast shifts in the political economy and exchange networks present (1975: 284). The shift towards greater political centralization and the loss of autonomy among smaller centres may have overloaded the existing system's capability to process information, leading to the creation of hierarchies to govern the new system. While Wright and Johnson admit that their research is unsuccessful in that it is unable to answer what specific causal factors led to the creation of centralized redistributive centres, they offer a model which can be tested in order to explain specific instances of state formation. J. Stephen Athens attempts to criticize the systems approach of Flannery, Wright and Johnson, arguing instead from an ecological perspective that "an understanding of the evolution of complex social systems can only proceed through an examination of factors relating to the stability of agricultural production and the use of relevant variable dimensions of analysis" (1977: 377). He claims that system theories "are unsuccessful because they cannot predict state formation" (Athens 1977: 360). Certainly, no one has ever claimed the usefulness of any model, including that of systems theory, to predict when state formation will occur. Athens, however, assumes that a theory which successfully addresses the question of causation will be fully equipped for prediction. By arguing that any successful theory of state formation should be universally applicable, Athens forces himself into a position where he must reject any theory such as systems theory which postulates that many possible evolutionary trajectories can lead to the formation of similar structures. His argument that multivariant causation is unfounded is therefore a highly tautologous one. Systems theorists have a response to Athens' qualms regarding causation in the concept of equifinality. This is defined by Buckley as the principle that a final condition of a system may be achieved by any number of developmental routes (Buckley 1967: 60). To attempt to separate out distinct causes, or even to emphasize one as primary, may not be useful because the result is unknown until it emerges. The "final" result of the interplay of factors is not merely the cumulation of causes to produce an effect, but a process of interaction and amplification. It is thus not possible to predict the emergence of the state with any regularity, or to separate causes from their end result. Given that the evidence regarding state formation suggests that there is no single factor that is sufficient, and that even the necessity of many factors can be debated, it seems reasonable to see whether different equifinal models of state formation can be created. A word of caution is necessary before continuing. Systems theory as a model creates testable hypotheses, not complete accounts of incidences of state formation. It should be acknowledged that these are not mechanical, electronic or organic systems, but human creations, subject to the vagaries of human existence and random chance. We must recognize that something as simple as the untimely death of one's leader can have profound consequences on history (as it did on the Mongol hordes), and there is nothing like a hurricane to crush an incipient rebellion and unite individuals in a rebuilding effort. Charles Spencer discusses the effect of the eruption of the volcano Xitle as a precipitating event for the conquest of Cuicuilco by the early Mesoamerican state of Teotihuacan (Spencer 1990: 22). As noted above, the consequence of this randomizing factor is that we never know exactly what parts of the system have priority in a given society, rendering prediction impossible. And in no way do I mean to imply that because a system is stable (has achieved equilibrium or homeostasis) that it is "adaptive", to use a much-misused term. As Rappaport notes, "Social systems ... can make mistakes of which biological systems may be incapable" (Rappaport 1977: 51). Still, recognizing this fact, we should proceed with the knowledge that even an imperfect system is not a random aggregate, but a structured grouping.
V. Illusory Universality: Systems and HistoryHistory abhors determinism but cannot tolerate chance. Bernard de Voto (1897-1955) I have placed myself thus far in the "systems theory" camp of evolutionary explanation. However, I feel that this is a necessary but insufficient means of depicting social evolution. After all, the day after the stock market crashes, life goes on, and the rules of the game are the same. Yet when societies evolve, they often undergo fundamental structural change. I now intend to come to some sort of analysis regarding the changes of such structures over time: in short, to indicate and explore the fundamentally historical nature of social systems. I have, so far, rejected any of the claims of evolutionary biology as to its applicability to processes of state formation. However, there is one respect in which the two types of systems, social and biological, are clearly similar and in which a useful analogy can be drawn. It is self- evident that all evolutionary processes are essentially historical processes which do not operate in some sort of vacuum. There has been a recent movement among some evolutionists to modify the Darwinian paradigm of evolutionary biologists, in recognition of the historical constraints which limit the variation which can exist within populations. At the forefront of this movement is Harvard biologist Stephen Jay Gould, whose works The Panda's Thumb and Bully for Brontosaurus suggest that neither biological nor cultural evolution can be understood in ahistorical terms (Gould 1980). This principle of historical constraint aids in explaining why humans have four limbs rather than some other number. Since humans' evolutionary predecessors - fish, reptiles and other mammals - all had four limbs, it is not possible for that trait to have been selected against, since no better variation existed within those species (and even if one evolved as a mutation, it would almost certainly be non-adaptive). Indeed, just as the panda cannot simply evolve a thumb out of nowhere, but must rely on existing morphological traits in order to produce an adaptive advantage, Gould asserts that cultural evolution may operate according to similar historical principles. He cites the example of the retention of the suboptimal QWERTY keyboard as a historical quirk which is now retained because of the incredible inertia that would have to be overcome in order to change to a better system (Gould 1992: 71-72). While, by his own admission, Gould's perspective on cultural evolution is that of an interested amateur, there has been little consideration given to these sorts of constraints among most social evolutionists, other than vague statements to the effect that "of course, there are some historical particularities in such-and-such society". Consequently, these particularities are usually dismissed as being of little importance. However, I see no reason to make such a broad assumption, and, indeed, many scholars have begun to consider issues of historical constraints when examining the evolution of particular societies. A recent paper by William Keegan and Morgan Maclachlan, whose ethnohistorical and archaeological analysis of the Bahamas traces the evolution of Taino avunculocal chiefdoms, suggests ways in which political evolution might be given its proper historical context. They assert that the traditional (initial condition) of Taino matrilineality, bolstered by the emergence of avunculocal residence due to internal warfare, led to structural changes and a shift towards greater political differentiation (Keegan and Maclachlan 1989: 620). One consequence of this change was a shift in settlement pattern towards settlement clusters and larger towns, "resulting from a male's desire to participate in his own lineage while resident with his wife's" (Keegan and Maclachlan 1989: 624). With larger population units came the inevitable need for co-ordination and conflict resolution through the power of a chief. Environmental factors must also have played a major role in Taino evolution. Keegan and Maclachlan remark that avunculocal chiefdoms are always found in highly circumscribed environments such as the Bahamas because they entail a great deal of internal conflict that can only be resolved upon the emergence of one dominant clan in order to produce stability (1989: 622). Finally, technological changes such as the advent of large ocean-going canoes for trade further bolstered the need for leadership and co-ordination among the Taino in order to marshal resources (Keegan and Maclachlan 1989: 626). We see in the emergence of Taino chiefdoms that (at least) four distinct elements of the society - residence pattern, kinship structure, physical environment and technology - act in combination to create a system by which political evolution is the favoured or "natural" outcome. It would be na‹ve to regard any one of these as a "prime mover". The removal or alteration of any one of those factors would have altered the evolution of Taino political centralization, perhaps even halting it. It is quite probable that if there had not been a great deal of initial conflict due to the circumscribed environment, that disaffected Taino clans could depart the region. It is equally likely that without the initial condition of matrilineality, the men of the clans would have no reason to desire clustered settlement patterns. We must also consider that, as this is a historical system: what happens once political evolution occurs? The new "initial conditions" are not only matriliny and warfare, but the presence of canoes, long-distance exchange and chiefs. If the new technology is used to migrate, thus removing the factor of environmental circumscription, will the chiefs lose their power? My point is not to describe "what if" (for we certainly cannot use a systems model for prediction) but to explain "what for" and "from what" certain institutions developed, and to understand the processes by which this might occur. I will turn now to an examination of the consequences of a historical perspective on social evolution to one of the key problems in the study of state formation: the distinction between primary (or pristine) and secondary states. Most authors make a strong distinction between the two types of state. Jonathan Haas claims that in every known ethnographic case of state formation, the situation has been "contaminated" by the presence of outside complex societies (Haas 1982: 51). Anatolii Khazanov notes that "the states founded as a consequence of the expansion or colonial policies of more developed states - which hence represented a kind of secondary developed states - were definitely not early ones, even at the time of their origin" (Khazanov 1978: 79). Jerome Rousseau, in discussing the prerequisites of inequality, asserts that "societies transformed by more developed modes of production (and this applies eminently to the colonial situation) cannot be expected to follow this sequence" (1985: 45). One would certainly have to concede that the distinction between primary and secondary states is a meaningful one at least with regards to its basic definition. There is no apparent reason, however, to sharply divide episodes of state formation on these lines any more than along the axis of circumscribed vs. non-circumscribed states, or distinguishing the presence or absence of long-distance trade. Two reasons come to mind immediately as to why the primary-secondary distinction might come to assume a logically prior position in the description of states. Firstly, many early researchers were seeking "prime movers" rather than considering a multivariant scheme. An inevitable consequence of this quest would be that secondary states had to be relegated to "impure" status in order to study the single cause which resulted in primary state formation. Inevitably, cultural diffusionism came to be looked down upon as a secondary process, unworthy of study. A second reason might be cultural arrogance; we do not, for instance, often think of Anglo-Saxon state formation as a secondary process, or somehow "contaminated". Just because the processes of state formation among the Iroquois or the Zulus were relatively recent does not deny their validity. An important issue which most scholars have not raised is why, given contact between large centralized states and smaller, non-state societies, certain societies developed into states and others did not. Why did the Zulus and Polynesians respond as they did, by evolving centralized political institutions of their own? Why did other societies such as the Iroquois, who were at the same "level" of evolution as the first two, not do so until much later? And why did other areas, such as the Pacific Northwest, not evolve states at all? It seems ridiculous to claim that all instances where diffusion occurred should fall under the rubric of "secondary states". The number of truly "pristine" formations is very small, perhaps as low as two (see Wright and Johnson 1975). Why, then, do we insist on making much ado about nothing by continuing to ignore the processes of state formation right on our historical doorstep? It is already very well-known among anthropologists that various institutions and structures may take different forms when imposed on or borrowed by societies to which those institutions are not native. Alan Macfarlane discusses capitalist economics as a cultural creation of the West, specifically England, in The Culture of Capitalism (Macfarlane 1992). The creation of pseudo-Christian cargo cults in New Guinea is well- documented, as are various forms of syncretic religion which derive from contact between organized religions and the religious beliefs and traditions of conquered and colonized peoples. There is no reason, then, other than blind insistence on the existence of "pure" political forms, to argue that the diffusion and/or imposition of state structures occurred in a pure, non- contested form. The position taken by Barbara Price regarding secondary state formation is quite close to my own. Price argues that secondary state formation depends on a dialectic approach in which the relationship is not simply one of the more complex society "imposing" the state upon the simpler, but upon careful consideration of the conditions of both the state and the non-state society, and the degree of contact between the two (Price 1978). However, where Price and I diverge is in her materialist assertion that it is the amount of energy use of each society alone which determines the resultant secondary state structure. For reasons I will detail below, I believe that the situation in many secondary states is far more complex than the expression of material factors. I will do so by considering two cases: the Iroquois of the Great Lakes region of North America, and the Zulu of southeastern Africa. The evolution (or, more accurately, the homeostasis) of the Iroquois has been well-documented by Bruce Trigger in a manner generally consistent with a systems-information theory perspective. He notes that immediately prior to and at the time of contact with Europeans, Iroquois society possessed many of the necessary conditions for statehood: a great potential for agricultural surplus production, the presence of warfare, long-distance trade with and borrowing from more complex Mississippian peoples, and concentrated villages under the control of chiefs (Trigger 1990: 124-127). However, despite material conditions and stresses similar to those creating states elsewhere, Iroquois society did not become significantly differentiated during this period. Trigger attributes the non-emergence of state structures to a concerted effort to maintain social equality through censure of those seeking political gain, and through rewarding generosity (1990: 136). The fact that "[t]he material benefits of chiefly office were more than counter-balanced by its expenses" which included dispensing goods ensured that those seeking political gain were discouraged from doing so (Trigger 1990: 133). Only much later, when colonial rule was at its height and the Iroquois found themselves under more circumscribed conditions and engaged in large-scale wars, did state formation occur.(5) From a systems perspective, it is evident that many factors ensured homeostasis in the resultant political structure by means of negative feedback. Trigger thus concludes that egalitarianism is not merely a "default" option for societies, but must be explained, just as the creation of states which result from the uncoupling or dissolution of such factors require analysis (1990: 145). The example of the Iroquois is also instructive in that ideological and cultural factors, presumably retained from an earlier period, take precedence over material and ecological factors which provide impetus towards social inequality into the system. It is evident, then, that human social systems are not merely slaves to their environment but have the potential to organize themselves along a variety of ideological principles which help in determining the course of social change. In the case of Zulu state formation, a "successful" instance (meaning only that a state was actually formed), the situation is quite different, and yet the systemic principles involved are very much the same. Service discusses the Zulu case briefly, noting that the personal traits of its first ruler, Shaka Zulu, were of secondary importance to the more gradual transformations in the politico-legal system (1975: 116). Stuart Romm, a political scientist, re-examines the Zulu data in a systems-oriented light, tracking the responses of various actors to changing military and political conditions, culminating in the end of homeostasis and the formation of the Zulu state. He argues that the eventual stabilization of Zulu rule under Mpande, its third king, was a result of his willingness to compromise with traditional chiefs, whereas his predecessors required brutal force and coercion to maintain their positions (Romm 1986: 637). The fundamental difference between the Iroquois and Zulu situation can be equated directly with the question of positive and negative feedback. Romm argues forcefully that pre-state polities "will be most subject to these disequilibrating tendencies when economic, social, and military conditions are such that actors of this nature are able to attract followers and mobilize resources" (1986: 611). Since the Zulu kings Shaka, Dingane and Mpande were able to revise the traditional chiefly ideologies in order to gain power, the constraints on state formation were removed. The two factors of personal military might and ideological considerations, previously holding each other in check, become coupled and positive feedback is the result. The conclusion that Romm reaches is inevitable: "one cannot have 'just a little' sociopolitical change" (1986: 641). Of course, not all instances of state formation are as rapid as the Zulu case, which was hastened by other factors such as the British and Boer colonizers. Although both the Iroquois and Zulu cases of political evolution (or its absence) are "secondary" formations, both show the remarkable properties of systems in the same way as detailed by systems theorists such as Flannery. Furthermore, with the cumbersome requirement of a single prime mover discarded, it is possible to re-integrate diffusion as a legitimate source of cultural evolution. My proposition, therefore, is this: The distinction between primary and secondary states, if it is to be retained, must be reduced in importance and its weight considered in all cases of state formation. In systemic terms, diffusion from states or responses to contact with states provide impetus for structural change to the initial conditions of the system. Whether there is positive feedback (probably leading to a state) or negative feedback (inertial forces preventing state formation) depends on the initial system under consideration, the nature and degree of unbalancing forces, and the interaction of the two. The structural change which occurs may then be considered a new set of "initial conditions" so to speak, upon which the evolutionary process may work.(6)
VI. Conclusion: Implications for Future State ResearchTime is a sort of river of passing events, and strong is the current; no sooner is a thing brought to sight than it is swept by and another takes its place, and this too will be swept away. Marcus Aurelius, Meditations IV, 43 Throughout this paper, I have traced the construction of four fundamental illusions regarding state formation. I have also, I hope, shown that despite these illusions, there is a process of social evolution which exists at the core of them all. If I have succeeded, the illusions should be stripped away, bringing us closer to an understanding of the processes that have constrained and governed the evolution of all societies, and which still continue to act upon them today. It would, of course be overly arrogant to claim that I have not at the same time constructed my own illusions; much work needs to be done before state formation can be resuscitated from its dormant state. I have indicated that the fundamental complexity of all political systems prohibits their reduction to a small number of categories in which one or a few scenarios lead to the development of the state as the "pinnacle" of human political evolution. Instead, we should recognize that the same processes that occurred in the past continue to operate now, both upon small-scale and industrial states, albeit in a greatly modified form. Is it possible that we will someday refer to the state as just one of many stages in the evolution of societies? The question of what such a future "post-state" might look like is one for the astrologer, not the anthropologist. Nevertheless, Western societies should recognize that they too are embedded in these historical processes. I am certainly not the first to suggest that all universal models of state formation are grossly oversimplified; most scholars, I suspect, regard this sort of theoretical simplicity as wrong. Nevertheless, the trend among some scholars to consider only what can be found in the archaeological record as potential causes has led to a sort of "methodological" simplification that I contend is entirely unnecessary. Rather than searching for "pure" or "pristine" states in the earth, I offer that it may be equally profitable (albeit in a different manner) to study the "messy" states of the historical and ethnographic record. While the situation thereby becomes muddled with regards to state formation, we should avoid oversimplifying a topic simply due to an outdated notion of the purity of earlier times. Historical events, "Great Men" (and Great Women), kinship structures, and other elements may all have played important roles in determining the course of any particular society's political evolution. If we are to reject any of these, it should not be because we cannot find them in the earth but because we cannot find room for them in our model of state formation. It may be that ethnographic study of modern states is no longer a viable alternative to archaeology, inasmuch as all human social systems are now under the control of Western-style state structures. However, it is inevitable, insofar as archaeologists recognize prehistoric states as human systems, that some degree of analogy will always be required between living human societies and the archaeological record. In this respect, I am in agreement with Gledhill (1988: 24), who asserts that the problems inherent in creating such comparisons "may not preclude archaeologists from continuing to draw insights about process from ethnographic and ethnohistorical cases, if we handle such analogies with greater care". I would go further and suggest that the invisibility of certain elements in the archaeological record has caused their role in societies to be overlooked, emphasizing instead what has survived the tests of time. If this is the case, then it is essential that we examine modern societies of all degrees of political complexity in an attempt to create meaningful dialogue between the sub-disciplines of archaeology and ethnology on the nature of the state. While there have certainly been many errors committed by early state theorists, many of these do not pose insurmountable barriers to the discernment of patterns of long-term directional social change. We must gain a clearer understanding, not only of the descriptive characteristics of complex polities, but also of the interrelated causal factors - ecology, warfare, trade, ritual, kinship, cultural diffusion, personal achievement, and others - which lead to their creation. Anthropologists should come to recognize the systemic features of societies so that the process of their evolution can be understood more fully. Above all, however, human systems must be recognized as historical systems, in which the initial conditions of the system play a key role in constraining its structural change. We must cease our search for the "pristine" or "pure" state of the past, which probably never existed, and look at the real examples around us. It is essential to examine societies of particular examples of these constraints in action, so that we may not fall off the thin ledge between excessive particularism and radical determinism. Rather than rejecting the evolutionary paradigm altogether, a critical examination of its failures and shortcomings can thus lead to a greater understanding of its relevance for contemporary theory.
Notes(1) Unlike Service (1975: 23), I do not regard Machiavelli solely as an early "coercion theorist", for he does recognize the need for the prince to have the consent of his subjects in many circumstances. (2) Such insistence on the creation of state "templates" which are imposed upon and eliminate earlier settlement patterns has been successfully refuted by Nocete (1994) with regards to locational analysis such as Central Place Theory. (3) Of course, social evolution in no way negates the possibility of no change in complexity - "stagnation", for lack of a better term, or societal collapse or "devolution". As I show later, these instances are useful in the study of the evolutionary process and, in fact, the exceptions prove the rule that societies tend to increase in complexity. (4) I do not include the "structural Marxists" such as Friedman and Godelier in this camp, for many scholars influenced by Marxist thought emphasize material factors along with ideological and cultural factors in state formation. (5) This occurrence would be an instance of negative feedback, since the newly introduced structure would tend to reduce the deviation caused by the initial condition of circumscription. (6) The reader may note certain similarities between this approach and the Hegelian dialectic (thesis + antithesis = synthesis), with each instance of change producing a new set of initial conditions.
Works CitedAthens, J. Stephen 1977 Theory Building and the Study of Evolutionary Process in Complex Societies. In For Theory Building in Archaeology, edited by Lewis R. Binford, 353-384. New York: Academic Press. Boserup, Ester 1965 The Conditions of Agricultural Growth. Chicago: Aldine. Buckley, Walter 1967 Sociology and Modern Systems Theory. Englewood Cliffs, NJ: Prentice-Hall. Carneiro, Robert 1967 "On the Relationship Between Size of Population and Complexity of Social Organization", Southwestern Journal of Anthropology, 23: 234-243. Carneiro, Robert 1970 "A Theory of the Origin of the State", Science 169: 733-738. Carneiro, Robert 1978 Political Expansion as an Expression of the Principle of Competitive Exclusion. In Origins of the State: The Anthropology of Political Evolution, edited by Ronald Cohen and Elman R. Service, 205-233. Philadelphia: Institute for the Study of Human Issues. Claessen, Henri J.M. 1978 The Early State: A Structural Approach. In The Early State, edited by Henri J.M. Claessen and Peter Skalník, 533-596. The Hague: Mouton. Cohen, Ronald 1978 State Origins: A Reappraisal. In The Early State, edited by Henri J.M. Claessen and Peter Skalník, 31-76. The Hague: Mouton. Flannery, Kent 1972 "The Cultural Evolution of Civilizations", Annual Review of Ecology and Systematics 3: 399-426. Fried, Morton 1967 The Evolution of Political Society: An Essay in Political Anthropology. New York: Random House. Friedman, Jonathan 1975 Tribes, States and Transformations. In Marxist Analyses and Social Anthropology, edited by Maurice Bloch, 161- 202. London: Tavistock. Gall, Patricia and Arthur A. Saxe 1977 The Ecological Evolution of Culture. In Exchange Systems in Prehistory, edited by Timothy Earle and Jonathon E. Ericson, 255-268. New York: Academic Press. Gledhill, John 1988 Introduction: the comparative analysis of social and political transitions. In State and Society, edited by John Gledhill, Barbara Bender and Mogens Trolle Larsen, 1-32. London: Unwin Hyman. Gould, Stephen Jay 1980 The Panda's Thumb. New York: W.W. Norton. Gould, Stephen Jay 1992 Bully for Brontosaurus. New York: W.W. Norton. Haas, Jonathan 1982 The Evolution of the Prehistoric State. New York: Columbia University Press. Harris, Marvin 1977 Cannibals and Kings: The Origins of Cultures. New York: Random House. Hill, Jack 1994 "Biology, Prestige and the Origin of States", Journal of Social and Evolutionary Systems 17(2): 213-225. Hutterer, Karl L. 1991 Losing Track of the Tribes: Evolutionary Sequences in Southeast Asia. In Profiles in Cultural Evolution, edited by A. Terry Rambo and Kathleen Gillogly, 219-246. Ann Arbor: University of Michigan Press. Keegan, William F. and Morgan D. Maclachlan 1989 "The Evolution of Avunculocal Chiefdoms: A Reconstruction of Taino Politics and Kinship", American Anthropologist 91: 613-630. Khazanov, Anatolii 1978 Some Theoretical Problems of the Study of the Early State. In The Early State, edited by H.J.M. Claessen and Peter Skalník, 77-92. The Hague: Mouton. Khazanov, Anatolii 1985 Rank Society or Rank Societies: Processes, Stages and Types of Evolution. In Development and Decline, edited by H.J.M. Claessen et al., 82-96. South Hadley, MA: Bergin & Garvey. Kristiansen, Kristian 1991 Chiefdoms, States and Systems of Social Evolution. In Chiefdoms: Power, Economy and Ideology, edited by Timothy K. Earle, 16-43. Cambridge: Cambridge University Press. Macfarlane, Alan 1987 The Culture of Capitalism. Oxford: Basil Blackwell. Maisels, Charles Keith 1987 "Models of Social Evolution: Trajectories from the Neolithic to the State", Man (NS) 22: 331-359. Masters, Roger D. 1982 "Evolutionary biology, political theory and the state", Journal of Social and Biological Structures 5: 439-450. Naroll, Raoul and William T. Divale 1976 "Natural selection in cultural evolution: warfare versus peaceful diffusion", American Ethnologist 3: 97-128. Nocete, Francisco 1994 "Space as Coercion: The Transition to the State in the Social Formations of La Campina, Upper Guadalquivir Valley, Spain, ca. 1900-1600 B.C.", Journal of Anthropological Archaeology 13: 171-200. Patterson, Thomas C. 1991 The Inca Empire: The Formation and Disintegration of a Pre-Capitalist State. New York: Berg. Price, Barbara J. 1978 Secondary State Formation: An Explanatory Model. In Origins of the State: The Anthropology of Political Evolution, edited by Ronald Cohen and Elman R. Service. Philadelphia: Institute for the Study of Human Issues. Rambo, A. Terry 1991 The Study of Cultural Evolution. In Profiles in Cultural Evolution, edited by A. T. Rambo and Kathleen Gillogly, 23-112. Ann Arbor: University of Michigan Press. Rappaport, Roy A. 1977 Maladaptation in Social Systems. In The Evolution of Social Systems, edited by J. Friedman and M.J. Rowlands, 49-72. London: Duckworth. Romm, Stuart 1986 "Political conflict, role differentiation and state formation: a case study of the Zulus", Social Science Information 25(3): 607-645. Rousseau, Jerome 1985 The Ideological Prerequisites of Inequality. In Development and Decline, edited by H.J.M. Claessen et al., 36-45. South Hadley, MA: Bergin & Garvey. Service, Elman R. 1962 Primitive Social Organization: An Evolutionary Perspective. New York: Random House. Service, Elman R. 1975 Origins of the State and Civilization. New York: W.W. Norton. Spencer, Charles 1990 "On the Tempo and Mode of State Formation: Neo- Evolutionism Reconsidered", Journal of Anthropological Archaeology 9(1): 1-30. Tosi, Maurizio 1976 "The Dialectics of State Formation in Mesopotamia, Iran and Central Asia", Dialectical Anthropology 1: 173-185. Trigger, Bruce 1985 Generalized Coercion and Inequality: The Basis of State Power in the Early Civilizations. In Development and Decline, edited by H.J.M. Claessen et al., 46-61. South Hadley, MA: Bergin & Garvey. Trigger, Bruce 1990 Maintaining Economic Equality in Opposition to Complexity: An Iroquoian Case Study. In The Evolution of Political Systems, edited by Steadman Upham, 119-145. Cambridge: Cambridge University Press. Upham, Steadman 1990 Analog Or Digital? In The Evolution of Political Systems, edited by Steadman Upham, 87-115. Cambridge: Cambridge University Press. Webster, David 1975 "Warfare and the Evolution of the State: A Reconsideration", American Antiquity 40(4): 464-470. Wright, Henry T. and Gregory A. Johnson 1975 "Population, Exchange and Early State Formation in Southwestern Iran", American Anthropologist 77: 267-289.