Religion in Human Evolution: From the Paleolithic to the Axial Age

by Robert N. Bellah

  One last typological remark: Hawaii seems to be a good example of what Max Weber called the patrimonial state, and which he defined as a state growing out of the household (court) of the king. Here again there is a matter of degree. It is only when the king’s “household” reaches the size and effectiveness that enable it to function as a genuine secondary formation that it can be called a patrimonial state. I would argue that in ancient Hawaii it had reached that stage.

  In my discussion of tribal religion I chose three examples for close examination: the Kalapalo, the Australian Aborigines (the Walbiri), and the Navajo. From the thousands of tribal peoples, this choice could not be defended as “representative,” even though each was chosen from a different continent. In considering chiefdoms as the form of organization intermediate between the tribal and the archaic, I chose to look mainly at Polynesia because of the clarity of the record there in which archaeology and ethnography combine to give a sense of development over many centuries, starting with Neolithic villages and ending with an early state in Hawaii. Still, given that we have data for hundreds of chiefdoms in many parts of the world, the choice of Polynesia can be defended as strategic but not as representative. With early states or early civilizations, what I have chosen to call archaic societies, we are in a very different situation. Though exactly how many there are can be argued, the number is surely quite small compared to tribes or chiefdoms, and those for which data is adequate are fewer still. Looking ahead to what follows the archaic, namely the axial age, there are only four cases: ancient Israel, ancient Greece, India in the second half of the first millennium BCE, and China in the same period. I have therefore decided to look closely only at those archaic societies that significantly contributed to axial ones: ancient Mesopotamia and Egypt, which influenced both Israel and Greece; and Shang and Western Zhou China, from which there is a smooth transition to the Chinese axial age. Had the data been adequate I would have included the Indus Valley civilization in India as well.

  I have, of course, in Chapter 4, already considered at length one other archaic society, Hawaii. This I have used as an example of the transition to an early state, with the advantage that we know more about it at an early stage than any other case. For none of the archaic societies we will consider in this chapter was there anyone like David Male, around to report on their early stages, nor can we reconstruct the probable developmental sequence for over 2,000 years before their emergence as early states with the clarity we now have about Polynesia. For studying the beginning of an archaic religion, the Hawaiian case is invaluable because of the wealth of information we have about it, not available for any other case.

  Before turning to the cases with which this chapter will be concerned, it will be useful to consider Bruce Trigger’s instructive survey, Understanding Early Civilizations, a compendious comparative analysis of seven cases: Old and Middle Kingdom Egypt, Mesopotamia from Early Dynastic III to Old Babylonian times, China in late Shang and early Western Zhou times, the Aztecs from the late fifteenth to early sixteenth centuries, the Classic Maya, the Inka kingdom during the early sixteenth century, and the Yoruba peoples of West Africa from the mid-eighteenth century to the late nineteenth century.’ Trigger has chosen his cases largely because they are the ones for which there is adequate data; the inability to understand the Indus Valley civilization on the basis of archaeological evidence alone (what little writing there is has not been deciphered) has forced him to exclude this important case, rightly in my view.’ Trigger’s sample is of mature states; because he does not include Hawaii, he has no example of a really early state. Of course we should remember that the “early state” is more a process than an event-it is almost always impossible to “pinpoint the precise moment of the birth of the state.“3 Even in Hawaii the state was clearly forming well before Western contact, though the process is more evident than in any of Trigger’s cases, and for this reason the Hawaiian case remains invaluable.

  One of the defects of my sample of archaic societies is that it excludes all the New World cases, so a summary of Trigger’s findings-three of his seven are from the New World-can go a little way to make up for that deficiency. It will be useful to begin by considering what Trigger means by “early civilizations,” because his definition is very close to what I mean by archaic societies:

  Anthropologists apply the term `early civilization’ to the earliest and simplest forms of societies in which the basic principle governing social relations was not kinship but a hierarchy of social divisions that cut horizontally across societies and were unequal in power, wealth, and social prestige. In these societies a tiny ruling group that used coercive powers to augment its authority was sustained by agricultural surpluses and labour systematically appropriated from a much larger number of agricultural producers. Fulltime specialists (artisans, bureaucrats, soldiers, retainers) also supported and served the ruling group and the government apparatus it controlled. Rulers cultivated a luxurious style of life that distinguished them from the

  If we think of Hawaii, the distinction between the ali’i and the commoners is just such a clear class distinction. Another way of making the same point without focusing quite so centrally on class is to say that the key distinction is between the state as a secondary formation and the rest of society. That this is close to what Trigger means is clear when he writes, “wealth tended to be derived from political power far more frequently than political power was derived from wealth.“5 So it is not class as defined in terms of relation to the means of production that is critical in these societies, but class as defined in relation to political power.

  Also important for Trigger is the point that kinship, although remaining significant in different ways for both the rulers and the ruled, no longer, as in tribal and chiefdom societies, is the “basic principle governing social relations.” He adds one further point of great importance: “Just as class has replaced real and metaphorical kinship as a basis for organizing society, so religious concepts replaced kinship as a medium for social and political discourse.“6 Of course, symbolic action and expression that can be called religious appear at every level of social organization, but something new in the religious realm appears in archaic societies: gods and the worship of gods. My reading of Trigger’s study reinforces my sense that what makes archaic society different from its predecessors is a complex religio-political transformation that gives rise to two ideas that are essentially new in the world: kingship and divinity, in many ways two parts of a single whole.

  Hawaiian society as we described it focused on the king and his relation to, even identity with, the gods, particularly Ku and Lono. Kingship is central in every one of Trigger’s cases, and everywhere the king had a unique relation to the gods, was frequently considered a god himself. Some form of divine kingship can be found in Old Kingdom Egypt, the Aztecs, Mayas, Inkas and Yorubas, and in Zhou China the king was the “Son of Heaven,” though he was not himself considered divine. In Mesopotamia, the earliest period of what was probably priest-kingship is obscure, but there were sporadic claims to divine status by kings in the Akkadian and Ur III dynasties in the third millennium BCE, and perhaps even in the Old Babylonian dynasty in the first half of the second millennium.?

  Human sacrifice associated with royal ritual was present in some form in every case, and was, as in Hawaii, always an indication of the extraordinarily exalted status of kingship, although the extent of it was variable. The commonest form was what is called retainer sacrifice, in which wives and retainers, sometimes in large numbers, were buried with the dead king. In Egypt this practice was found in the First and probably the Second Dynasties; in Mesopotamia only in the Early Dynastic royal burials at Ur-in each case no later examples are known. Although the numbers decreased markedly in China after the Shang, some retainer burial was practiced for centuries. But in most cases human sacrifice in rituals other than funerals was not uncommon: Shang China, the Mayas, Inkas, and Yorubas, and most extensively of all, the Aztecs, where thousands of w
ar captives were sacrificed at the great temple at Tenochtitlan right up until the Spanish conquest.’

  The extraordinary exaltation of the ruler puts Hawaii firmly in the category of (early) archaic society, where such exaltation everywhere went to extremes unknown in earlier or later periods, but there are other features that we normally consider indicative of archaic society that were not present in Hawaii: urbanism and writing, for example. Trigger argues, however, that cities are not an indispensable marker of early civilizations; rather, such civilizations divide into two types, city-states and territorial states. Whereas Mesopotamia, the Yorubas, the Aztecs, and the Mayas were city-states, Egypt, China, and the Inkas were territorial states.9 City-states were large, multipurpose, urban conglomerates, usually located near highly productive agricultural areas, and from which larger states were sometimes formed, usually by subjecting other such cities to tribute status. In territorial states it was the court, not the city, that provided the center, and the court was often peripatetic. There were important ceremonial centers, but the court could visit them only intermittently or move from one to another. Hawaii was clearly in the category of the territorial state, building its empire across the archipelago rather than reaching out from a single city. Of course, established territorial empires eventually gave rise to cities, though cities were not the basis of state structure. Conversely, city-states sometimes became territorial states, though extending city institutions to a large territory usually proved a daunting and often in the long run an impossible task, Rome being the great exception.

  When we use the word “civilization,” as we inevitably must in speaking of archaic societies, we usually think of writing as an essential criterion. But in Trigger’s seven cases, writing was entirely absent among the Inkas and the Yorubas, and rudimentary among the Aztecs, Mayas, and perhaps the Shang Chinese (though there may have been more extensive writings than the oracle bones, on which our knowledge of Shang writing depends, they have not survived). Even in Mesopotamia, where writing was “invented” around 3200 BCE, it was first used mainly for accounting and for lists, and continuous texts cannot be deciphered until about 2500 BCE.

  Another feature of most archaic societies is the presence of monumental architecture, mainly for ritual and/or royal use. The Hawaiian heiau (temple) was a modestly monumental structure, one of the largest of which, on the island of Maui, was over 4,000 square meters in area and required an estimated 26,000 labor-days for construction during ten separate occasions? Such temples do not compare with the Mesopotamian ziggurats, the Aztec, Maya, or Inka temples, or, of course, with the Egyptian pyramids. But neither the Shang Chinese nor the Yorubas appear to have produced monumental architecture much more impressive than the Hawaiian heiau.

  Trigger indicates that when he began his study he expected to find economic practices to be the most constant in his sample and religious beliefs and practices the most variable. In fact he found the opposite: subsistence patterns varied quite widely due to differences in ecological context, whereas religious beliefs and practices were remarkably comparable across his seven cases.” Comparable, but, as we shall see, still significantly different. In Chapter 4 we saw how the relation of religion and power, only incipient in tribal societies, came to a kind of climax in Hawaii. Having used Trigger’s book as an introduction to the field of mature archaic societies, we can now try to understand better the relation of religion and power, of god and king, in such societies by taking a closer look at three of them.

  Ancient Mesopotamia

  On the face of it, Hawaii and Mesopotamia could hardly have had more opposite starting points. Hawaii was located in just about the remotest spot on the planet, out of contact with any other society for centuries before the arrival of Europeans. Mesopotamia (literally, “the land between the rivers”- the Tigris and Euphrates-present-day Iraq contains all of ancient Mesopotamia) was at the center of the vast Eurasian (and North African) land mass and was never out of touch with its many neighbors, near and far. This geographical difference alone helps account for the fact that the Mesopotamian state began about 5,000 years before the Hawaiian state. Not only geographically, but in terms of other variables as well, Hawaii and Mesopotamia are far apart among archaic societies, so that beginning the consideration of mature archaic societies with Mesopotamia allows maximal contrasts to appear.

  Archaeology reveals that, in spite of their many differences, in both cases settlement began on largely virgin territory. After about 4000 BCE, in the alluvial plain of Southern Mesopotamia, only very sparsely settled before, a large number of fairly large settlements appeared rather suddenly, and by about 3200 BCE the first true cities in the world had emerged.12 These cities focused on monumental temple compounds but also had palaces, markets, and extensive residential quarters. The new level of population density that these cities evidenced was made possible by extensive cultivation of the alluvial soil. But the economic basis of these cities was not just local irrigation agriculture, but area-wide economic innovations that Andrew Sherratt has called the secondary products revolution, a transformation that he believes was as significant as the beginnings of plant and animal domestication themselves, at least 4,000 years earlier.13

  Early animal domestication was at first simply for the purpose of having a stable meat supply. With the secondary products revolution, for the first time animal power began to replace human power in agriculture. (It is worth remembering that, due to the absence of cattle and sheep, there was no secondary products revolution in the New World, or, of course, in Hawai’i). Yokes and harnesses were invented so that cattle could pull plows and carts. Sherratt estimates that the plow, because it can go deeper into the soil, is four times more efficient than the hoe in preparing the soil for sowing. 14 And carts make it much easier to bring grain in from outlying fields. These inventions appeared first in northern Mesopotamia by about 4000 BCE, in the old zone of agricultural settlement, but they helped to make possible the rapid urbanization in the south, which followed soon after. The changes involved in the secondary products revolution were not only agricultural; they involved a new kind of pastoralism as well. For the use of milk and milk products (yogurt, cheese) originated at about this time, as did the use of sheep to supply wool for textiles, earlier textiles being of vegetable fiber. Again, Sherratt estimates that the use of herds as a source of milk products is four to five times more efficient, in the amount of protein and energy produced relative to the same amount of feed, as using them only for meat.15 Although southern Mesopotamia had rich alluvial soil that could be very productive when irrigated, and lands beyond the possibility of irrigation that would support pastoralism, it had little else: no wood, no stone, no metal. In spite of great ingenuity in the use of resources indigenous to the area, it is clear that trade, including longdistance trade, was essential from the very beginning. Thus a region-wide economy, involving plow agriculture and intensive pastoralism, together with a considerable amount of trade, had appeared by the end of the fourth millennium BCE.

  Susan Pollack catalogs some of the developments in southern Mesopotamia evident by the end of the Uruk period (4000-3100 BCE):

  The Uruk period witnessed a massive increase in the number of settlements. Although many of them were small villages, others grew rapidly into towns and cities. By the end of the Uruk period, some larger settlements were walled. Temples and other public buildings became larger and more elaborate, and their construction must have employed large workforces for lengthy periods … Mass production was introduced for manufacturing some kinds of pottery using technological innovations such as mold manufacture and wheel-throwing. Systems of accounting … were elaborated and diversified, and writing-the premier accounting and recording technology-was invented toward the end of the period. Representations of men with weapons and bound individuals, presumably prisoners, attest to the use of armed force. The repeated depiction of a bearded individual with long hair, distinctive style of headdress, and skirt engaging in a variety of activities s
uggestive of authority is among the indications [of] the public exercise of power.16

  By 2900 BCE the city of Uruk, perhaps the most important city of Sumer, had become enormous by the standards of ancient cities. Hans Nissen shows that it was larger than Athens in 500 BCE or Jerusalem in 50 CE, and almost as large as Rome in 100 CE.17 It has been estimated that by 2500 BCE the population of Uruk was about 50,000. The main temple of the city was immense, with a stepped tower that had been rebuilt several times, each time with increased height.

  With only archaeological evidence (the script was used almost solely for accounting and contains no decipherable narratives) to go on, we simply cannot say what the structure of authority in Uruk and other comparable cities emerging at the same time was like. Hans Nissen details some of the previous theories: that the early rulers, entitled en, or ensi, were in effect priest kings; that later, temporary military leaders called lugal (meaning “the great man”) were appointed, and these over time became permanent “kings,” rivaling the chief priests for dominance in the city. Nissen feels the whole terminology of en, ensi, and lugal is too inconsistent in the surviving records to support such a theory and that we simply do not know how power was wielded in the earliest period. By the early dynastic period (2900-2350 BCE) it is clear that there were royal dynasties in the major Sumerian cities, but that great temples were the focus of both wealth and power, and indeed their upkeep was a major royal responsibility. Both temple and palace have been referred to as “great households” or “great organizations” because they were major landowners, had large staffs, and in some cases engaged in textile manufacturing, the major Mesopotamian export.18