5 How Were Societies Organized? Some of the most interesting questions we can ask about early societies are social. They are about people and about relations between people, about the exercise of power and about the nature and scale of organization. As is generally the case in archaeology, the data do not speak for themselves: we have to ask the right questions, and devise the means of answering them. There is a contrast here with cultural or social anthropology, where the observer can visit the living society and rapidly form conclusions about its social and power structures before moving on to other matters, such as the details of the kinship system or the minutiae of ritual behavior. The social archaeologist has to work systematically to gain even basic details, but the prize is a rich one: an understanding of the social organization not just of societies in the present or very recent past (like cultural anthropology) but of societies at many different points in time, with all the scope that that offers for studying change. Only the archaeologist can obtain that perspective, and hence seek some understanding of the processes of long-term change. The first question to address is the size or scale of the society. The archaeologist will often be excavating a single site. But was that an independent political unit, like a Maya or Greek city-state, or a simpler unit, like the base camp of a hunter-gatherer group? Or was it, on the other hand, a small cog in a very big wheel, a subordinate settlement in some far-flung empire, like that of the Incas of Peru? Any site we consider will have its own hinterland, its own catchment area for the feeding of its population. But one of our interests is to go beyond that local area, and to understand how that site articulates with others. From the standpoint of the individual site - which is often a convenient perspective to adopt - that raises questions of dominance. Was the site politically independent, autonomous? Or, if it was part of a larger social system, did it take a dominant part (like the capital city of akingdom) or a subordinate one? If the scale of the society is a natural first question, the next is certainly its internal organization. What kind of society was it? Were the people forming it on a more-or-less equal social footing? Or were there instead prominent differences in status, rank, and prestige within the society - perhaps different social classes? And what of the professions: were there craft specialists? And if so, were they controlled within a centralized system, as in some of the palace economies of the Near East and Egypt? Or was this a freer economy, with a flourishing free exchange, where merchants could operate at will in their own interest? These are just some of the questions we may ask. Provided we ask the right questions, the issues touched on above, and many more besides, can find some sort of answer from the archaeological record. Different kinds of society need different kinds of question. For example, if we are dealing with a mobile group of hunter-gatherers, there is unlikely to be a complex centralized organization. And the techniques of investigation will need to vary radically with the nature of the evidence. One cannot tackle an early hunter-gatherer camp in Australia in the same way as the capital city of a province in China during the 5hang Dynasty. Thus, the questions we put, and the methods for answering them, must be tailored to the sort of community we are dealing with. So it is all the more necessary to be clear at the outset about the general nature of that community, which is why the basic social questions are the first ones to ask. Precisely because the scale of a society is crucial in determining the way many aspects of its social organization work in practice, this chapter deals first with smaller, simpler societies, building toward larger, more complex ones. Certain questions, such as settlement archaeology or the study of burials, are therefore discussed in the context of each type of society. This involves some repetition between sections but it allows us to deal more coherently with the different social aspects of societies organized on approximately the same scale. 165 PART II Discovering the Variety of Human Experience ESTABLISHING THE NATURE AND SCALE OF THE SOCIETY The first step in social archaeology is so obvious that it is often overlooked. It is to ask, what was the scale of the largest social unit, and what kind of society, in a very broad sense, was it? The obvious is not always easy, and it is necessary to ask rather carefully what we mean by the "largest social unit," which we shall term the polity. This term does not in itself imply any particular scale or complexity of organization. It can apply as well to a city-state, a hunter-gatherer band, a farming village, or a great empire. A polity is a politically independent or autonomous social unit, which may in the case of a complex society, such as a state society, comprise many lesser components. Thus, in the modern world, the autonomous nation state may be subdivided into districts or counties. each one of which may contain many towns and villages. The state as a whole is thus the polity. At the other end of the scale, a small group of hunter-gatherers may make its own decisions and recognize no higher authority: that group also constitutes a polity. Sometimes communities may join together to form some kind of federation, and we have to ask whether those communities are still autonomous polities, or whether the federation as a whole is now the effective decision-making organization. These points are not yet archaeological ones: however, they illustrate how important it is to be clear about what we wish to know about the past. In terms of research in the field, the question is often best answered from a study of settlement: both in terms of the scale and nature of individual sites and in relationships between them, through the analysis of settlement pattem. But we should not forget that written records, where a society is literate and uses writing, oral tradition, and ethnoarchaeology - the study from an archaeological point of view of present-day societies - can be equally valuable in assessing the nature and scale of the society under review. First, however, we need a frame of reference, a hypothetical classification of societies against which to test our ideas. Classification of Societies The American anthropologist Elman Service developed a four-fold classification of societies that many archaeologists have found useful. Associated with these societies are particular kinds of site and settlement pattern. 166 Bands. These are small-scale societies of hunters and gatherers, generally of fewer than 100 people, who move seasonally to exploit wild (undomesticated) food resources. Most surviving hunter-gatherer groups today are of this kind, such as the Hadza of Tanzania or the San of southern Africa. Band members are generally kinsfolk, related by descent or marriage. Bands lack formal leaders, so that there are no marked economic differences or disparities in status among their members. Because bands are composed of mobile groups of hunter-gatherers, their sites consist mainly of season· ally occupied camps, and other smaller and more specialized sites. Among the latter are kill or butchery sites - locations where large mammals are killed and sometimes butchered - and work sites, where tools are made or other specific activities carried out. The base camp of such a group may give evidence of rather o. ';' .Qrl:.~ b()~ iWorld population : 10 million Hunters:100 percent World population : 3 billion Hunters: 0001 percent (Above) Before the advent of farming, all human societies were hunter-gatherer bands. Today bands scarcely exist. (Right) Classification of societies. TOTAL NUMBERS SOCIAL ORGANIZATION ECONOMIC ORGANIZATION SETTLEMENT PATTERN RELIGIOUS ORGANIZATION ARCHITECTURE ARCHAEOLOGICAL EXAMPLES MODERN EXAMPLES 5 How Were Societies Organized?Social Archaeology BAND SEGMENTARY SOCIETY CHIEFDOM STATE San hunters, South Africa Less than 100 Egalitarian Informal leadership Mobile hunter-gatherers Temporary camps Shamans Temporary shelters Paleofithic skin tents , Siberia Al l Paleolithic societies, including Paleo-Indians Eskimo Kalahari Bushmen Austra lian Aborigines Man pfowing, Va/camonica , N. Italy Up to few 1,000 Segmentary society Pan-tribal associations Raids by small groups Settled farmers Pastoralist herders Permanent villages Religious elders Calendrical rituals Permanent huts Burial mounds Shrines Neoffthic shrine, catat Huyuk, Turkey All early farmers (Neolithic/ Archaic) Pueblos, Southwest USA New GUinea Highlanders Nuer & Dinka in E. Africa Horseman, Gundestrup caldron 5,000-20,000 + Kinship-based ranking under hereditary leader High-ran ki ng warriors Central accumulation and redistribution Some craft specialization Fortified centers Ritual centers Hereditary c hief with religious duties Large-scale monuments Stonehenge. England - fina/form Many early metalworking and Formative societies Mississippian, USA Smaller African kingdoms Northwest Coast Indians, USA 18th-century Polynesian chiefdoms in Tonga, Tah iti, Hawaii t : t:~) : ... .. \ . !<.~ " ," -' T • " :~ . ' 'L l ' " , Terracotta army, tomb of first emperor o/China Generally 20,000 + Class-based hierarchy under king or emperor Armies Centralized bureaucracy Tribute-based Taxation Laws Urban: cities, towns Frontier defences Roads Priestly class Pantheistic or monotheistic religion \ Palaces, temples, and other public buildings Castillo, Chichen Itza, Mexico All ancient civilizations e.g. in Mesoamerica, Peru Near East, India and China; Greece and Rome All modern states 167 PART II Discovering the Variety of Human Experience insubstantial dwellings or temporary shelters, along with the debris of residential occupation. During the Paleolithic period (before 12,000 years ago) most archaeological sites seem to conform to one or other of these categories - camp sites, kill sites, work sites - and archaeologists usually operate on the assumption that most Paleolithic societies were organized into bands. Ethnoarchaeology (see below) has devoted much attention to the study of living bands of hunter-gatherers, yielding many insights relevant to the more remote past. Tribes. These are generally larger than bands, but rarely number more than a few thousand, and their diet or subsistence is based largely on cultivated plants and domesticated animals. Typically, they are settled farmers, but they may be nomad pastoralists with a very different, mobile economy based on the intensive exploitation of livestock. These are generally multicommunity societies, with the individual communities integrated into the larger society through kinship ties. Although some tribes have officials and even a "capital" or seat of government, such officials lack the economic base necessary for effective use of power. The typical settlement pattern for tribes is one of settled agricultural homesteads or villages. Characteristically, no one settlement dominates any of the others in the region. Instead, the archaeologist finds evidence for isolated, permanently occupied houses (a dispersed settlement pattern) or for permanent villages (a nucleated pattern) . Such villages may be made up of a collection of free-standing houses, like those of the first farmers of the Danube valley in Europe, c. 4500 BC. Or they may be clusters of buildings grouped together - so-called agglomerate structures, for example, the pueblos of the American Southwest, and the early farming village or small town of <;atalhdyUk, c. 7000 BC, in what is now Turkey. Chiefdoms. These operate on the principle of ranking - differences in social status between people. Different lineages (a lineage is a group claiming descent from a common ancestor) are graded on a scale of prestige, and the senior lineage, and hence the society as a whole, is governed by a chief. Prestige and rank are determined by how closely related one is to the chief, and there is no true stratification into classes. The role of the chief is crucial. Often, there is local specialization in craft products, and surpluses of these and of foodstuffs are periodically paid as obligations to the chief. He uses these to maintain his retainers, and may use them for redistribution to his subjects. The chiefdom generally has a 168 center of power, often with temples, residences of the chief and his retainers, and craft specialists. Chiefdoms vary greatly in size, but the range is gener· ally between about 5000 and 20,000 persons. One of the characteristic features of the chiefdom is the existence of a permanent ritual and ceremonial center that acts as a central focus for the entire polity. This is not a permanent urban center (such as a city) with an established bureaucracy, as one finds in state societies. But chiefdoms do give indications that some sites were more important than others (site hier· archy), as discussed later in this chapter. Examples are Moundville in Alabama, USA, which flourished c. AD 1000-1500, and the late Neolithic monuments of Wessex in southern Britain, including the famous ceremonial center of Stonehenge (see boxes, below) . The personal ranking characteristic of chiefdom societies is visible in other ways than in settlement patterning: for instance, in the very rich grave-goods that often accompany the burials of deceased chiefs. Early States_ These preserve many of the features of chiefdoms, but the ruler (pe~ps a king or sometimes a queen) has explicit authority to establish laws and to enforce them by the use of a standing army. The society no longer depends totally on kin relationships: it is now stratified into different classes. Agricultural workers or serfs and the poorer urban dwellers form the lowest classes, with the craft specialists above, and the priests and kinsfolk of the ruler higher still. The functions of the ruler are often separated from those of the priest: palace is distinguished from temple. The society is viewed as a territory owned by the ruling lineage and populated by tenants who have the obligation of paying taxes. The central capital houses a bureaucratic administration of officials; one of their principal purposes is to collect revenue (often in the form of taxes and tolls) and distribute it to government, army, and craft specialists. Many early states developed complex redistributive systems to support these essential services. Early state societies generally show a characteristic urban settlement pattern in which cities play a prominent part. The city is typically a large population center (often of more than 5000 inhabitants) with major public buildings, including temples and work places for the administrative bureaucracy. Often, there is a pronounced settlement hierarchy, with the capital city as the major center, and with subsidiary or regional centers as well as local villages. This rather simple social typology, set out by Elrnan Service and elaborated by William Sanders and Joseph Marino, can be criticized, and it should not be used unthinkingly. Some scholars find the concept of the tribe a rather vague one, and prefer to speak of "segmentary societies." The term "tribe," implying a larger grouping of smaller units, carries with it the assumption that these communities share a common ethnic identity and self-awareness, which is now known not generally to be the case. The term "segmentary society" refers to a relatively small and autonomous group, usually of agriculturalists, who regulate their own affairs: in some cases, they may join together with other comparable segmentary societies to form a larger ethnic unit or "tribe"; in other cases, they do not. For the remainder of this chapter, we shall therefore refer to segmentary societies in preference to the term "tribe." Certainly, it would be wrong to overemphasize the importance of the four types of society given above, or to spend too long agonizing as to whether a specific group should be classed in one category rather than another. It would also be wrong to assume that somehow societies inevitably evolve from bands to segmentary societies, or from chiefdoms to states. One of the challenges of archaeology is to attempt to explain why some societies become more complex and others do not, and we shall return to the fundamental issue of explanation in Chapter 12. Nevertheless, if we are seeking to talk about early societies, we must use words and hence concepts to do so. Service's categories provide a good framework to help organize our thoughts. They should not, however, deflect us from focusing on what we are really looking for: changes over time in the different institutions of a society - whether in the social sphere, the organization of the food quest, technology, contact and exchange, or spiritual life. For archaeology has the unique advantage of being able to study processes of change over thousands of years, and it is these processes we are seeking to isolate. Happily there are sufficiently marked differences between simple and more complex societies for us to find ways of doing this. As we saw above in the description of Service's four types of society, complex societies show in particular an increased specialization in , or separation between, different aspects of their culture. In complex societies people no longer combine, say, the tasks of obtaining food, making tools, or performing religious rites but become specialists at one or other of these tasks, either as full-time farmers, craftspeople, or priests. As technology develops, for example, groups of individuals may acquire particular expertise in pottery-making or metallurgy, and will become full-time craft specialists, occupying distinct areas of a town or city and thus 5 How Were Societies Organized? Social Archaeology leaving distinct traces for the archaeologist to discover. Likewise, as farming develops and population grows, more food will be obtained from a given piece of land (food production will intensify) through the introduction of the plow or irrigation. As this specialization and intensification take place, so too does the tendency for some people to become wealthier and wield more authority than others - differences in social status and ranking develop. It is methods for looking at these processes of increasing specialization, intensification, and social ranking that help us identify the presence of more complex'5.Qfieties in the archaeological record - societies here termed for convenience chiefdoms or states. For simpler band or segmentary societies, other methods are needed if we are to identify them archaeologically, as will become apparent in a later section. Scale of the Society With this general background in mind one can develop a strategy for answering the first, basic question: what is the scale of the society? One answer may come from an understanding of the settlement pattern, and this can only come from survey (see below) . For a first approximation, however, an elaborate field project may be unnecessary. If, for instance, we are dealing with archaeological remains dating to before about 12,000 years ago, then we are dealing with a society from the Paleolithic period. On present evidence, nearly all the societies known from that enormously long period of time - spanning hundreds of thousands of years - consisted of mobile huntergatherers, occupying camps on a seasonal and temporary basis. On the other hand, where we find indications of permanent settlement this will suggest a segmentary society of agricultural villages or something more complex. At the other end of the scale, if there are major urban centers the society should probably rank as a state. More modest centers, or ceremonial centers without urban settlement, may be indicative of a chiefdom. To use these c1assificatory terms is a worthwhile first step in social analysis, provided we bear in mind again that these are only very broad categories designed to help us formulate appropriate methods for studying the societies in question. If it is clear that we are dealing with communities with a mobile economy (Le. hunter-gatherers, or possibly nomads), highly intensive techniques of survey will have to be used, because the traces left by mobile communities are generally very scanty. If, on the other hand, these were sedentary communities, a straight- 169 PART II Discovering the Variety of Human Experience forward field survey is now called for. [t will have as its first objective the establishment of settlement hierarchy. The Survey The techniques of field survey were discussed in Chapter 3. Surveys can have different purposes: in this case, our aim is to discover the hierarchy of settlement. We are particularly interested in locating the major centers (because our concern is with organization) and in establishing the nature of the more modest sites. This implies a dual sampling strategy. At the intensive level of survey, systematic surface survey of carefully selected transects should be sufficient, although the ideal would be a total survey of the entire area . A random stratified sampling strategy - as outlined in Chapter 3 - taking into account the different environmental areas within the region, should offer adequate data about the smaller sites. However, random sampling of this kind could, in isolation, be misleading and subject to what Kent Flannery has called "the Teotihuacan effect." Teotihuacan is the huge urban site in the Valley of Mexico that flourished in the 1st millennium AD (see box, pp. 86-87). Random stratified sampling alone could easily miss such a center, and would thus ruin any effective social analysis. The other aim of the strategy must be, therefore, to go for the center. Means must be devised of finding the remains of the largest center in the region, and as many lesser centers as can be located. Fortunately, if it was an urban site, or had monumental public buildings, such a center should become obvious during even a non-intensive survey, so long as a good overview of the area as a whole is obtained. [n most cases the existence of such a prominent site will already be well known to the local population, or indeed recorded in the available archaeological or antiquarian literature. All such sources, including the writings of early travelers in the region, should be scrutinized in order to maximize the chances of finding major centers. The main centers usually have the most impressive monuments, and contain the finest artifacts. So it is imperative to visit all the major monuments of the period, and to follow up the circumstances of any particularly rich finds in the region. Where appropriate, there is plenty of scope too for remote sensing methods such as were described in Chapter 3. Settlement Patterning Any survey will result in a map of the areas intensively surveyed and a catalog of the sites discovered, 170 together with details of each site including size, chronological range (as may be determined from sur· face remains such as pottery) , and architectural fea· tures. The aim is then to reach some classification 01 the sites on the basis of this information. Possible site categories include Regional Center, Local Center, Nucleated Village, Dispersed Village, and Hamlet. The first use we will make of settlement pattern information is to identify the social and political terri· tories around centers, in order to establish the political organization of the landscape. Many archaeological approaches here give prominence to Central Place Theory (see below), which we feel has some limita· tions. It assumes that the sites in a given region will fall neatly into a series of categories according to vari· ations in site size. All the primary centers should be in one size category, all the secondary centers in the next, etc. This technique cannot cope with the true situation which is that secondary centers in one area are sometimes larger than primary centers in another. More recent work has found a way of overcoming thil difficulty (the XTENT technique), but we will deal here with the earlier methods first. Central Place Theory. This theory was developed by the German geographer Waiter Christaller in the 1930s to explain the spacing and functions of citiel and towns in modern-day southern Germany. He argued that in a uniform landscape - without moun· tains or rivers or variations in the distribution of SOi!l and resources - the spatial patterning of settlementl would be perfectly regular. Central places or settle· ments (towns or cities) of the same size and nature would be situated equidistant from each other, sur· rounded by a constellation of secondary centers with their own, smaller satellites. Under these perfect Site Number Ch-LT-41 Natural Setting: 2450 meters, in the Smooth Lower Piedmont. Situated on gently sloping ground. Shallow to medium soil cover. Moderate erosion. Modem Utilization: Rainfall cultivation. Archaeological Remains: Light surface pottery over an area of 0.8 hectare. No structural remains. Mixed with heavier Aztec material (Ch-Az-1 02). Traces of Terminal Formative pottery also occur. Classification: Small hamlet, 5-10 people. An example of a formal site descriplion (rom the catalog of sites produced (or the Basin o( Mexico sllrvey by Jeffrey Parsons llnd his colleaglles. 0 • • 0 • • • • • • • • 0 • • • • • 0 • • • • • • 0 • • 0 • ' CITY o VILLAGE OrOWN • HAMLET (efllra l Place Theory: in a flat landscape, with no rivers or IIlIriatiollS in resources, a central place (town or city) will dominate a hexagoTlalterritory, with secolldary centers (viI/ages or hamlets) spaced at reglllar intervals around it. conditions, the territories "controlled" by each center lVould be hexagonal in shape, and the different levels of center would together give rise to an intricate settlement lattice. Such perfect conditions do not occur in nature, of (OUfSe, but it is still quite possible to detect the workings of Central Place Theory in the distributions of modern or ancient cities and towns. The basic feature is that each major center will be some distance from its neighbors and will be surrounded by a ring of smaller settlements in a hierarchically nested pattern. In political and economic terms the major center will supply certai n goods and services to its surrounding area and wi ll exact certain goods a nd services in return. Even in an area so far from uniform as Mesopotamia (modern Iraq) , Central Place Theory has its uses (see box overleaf). Site Hierarchy. Despite the reservations we have expressed about Central Place Theory, the analysis of site sizes is a useful basic approach. In archaeological studies, the sites are usually listed in rank order by size (i.e. in a site hierarchy), and then displayed as a histogra m. There are normally many more small villages and hamlets in a settlement system than large towns or cities. Histograms allow comparisons to be made between the site hierarchies of different regions, different periods, and different types of society. In band societies, for example, there will usually be only cl narrow range of variation in site size and al! the sites 5 How Were Societies OrgaTlized? Sucial Archaeology will be relatively small. State societies, on the other ha nd , will have both hamlets ,lIld farmsteads ,md large towns and cities. The degree to which a single site is dominant within a settlemcnt system will also be evident from this type of anillysis, and the organization of the settlement system will often be a direct reflection of the organization of the society which created it. In a general way, the more hierarchical the settlement pattern, the more hierarchical the society. Thiessen Polygons. Another relatively simple method that ca n be used in the study of settlement patterns is the constrllction of Thiessen polygons. These are simple geometrical shapes that divide an Mea into a number of separate territories, each focllsed on a single site. The polygons are created by drawing straight lines between each pair of neighboring sites, then at the mid-point along each of these lines a second series of lines, at right angles to the first. Linking lip the second series of lines creates the Thiessen polygons, and in this way the whole of an area can be apportioned among the sites it contains. It should be noted, however, that this procedure takes no account ot" differences in size or importance of sites; a small site wi ll have as big a polygon as a large site. Thus it is important to use only sites of the sallle rank in the settlement hierarchy when this technique is being applied. A further question, more difficult to l'esolve, is contemporaneity, since clearly it would be meaningless to draw Thiessen polygons between sites which were not in occupation at the same time. XTENT Modeling. Gne of the shortcomings of Central Place Theory and other approaches is that sites occupying the same level in a settlement hierarchy might not be of the same size. Thus the capital city of a state on the periphery of a distribution could be smaller than a secondary city in the center. We are now able to cope with this using the technique of XTENT modeling. This has the aim of assigning territories to centers according to their scale. To do this, it assumes that a la rge center will dominate a slllall one if they are close together. [n such a case, of so-called dominance, the territory of the smaller site is simply absorbed in the study into that of the larger one: in political terms the smaller site has no independent or autonomous existence. This approach overcomes the limitation of the Thiessen polygon method , where territories are assigned irrespective of the size of the center, and where there are no dominant or subordinate centers. In XTENT mocleling, the size of each center is assumed to be directly proportional to its area of 171 PART II Discovering the Variety ofHuman Experience influence. The influence of e,xch center is thought of as analogous to a bell or bellftent in shape: the greater the size of the center the higher the tent. Centers are considered to be subordinate if their associated bell tents fall entirely within that of a larger center. If they protrude beyond, they will have their own autonomous existence as centers of political units. Ancienl watercourse Warka C Ceramic wall cones (Top) XTENT model territories, Late Uruk period, Warka area, Mesopotamia. Arrows indicate four centers that emerge as autonomous. Compare Greg Johnson 's hierarchy (above) for the same region. Note how four of the five "large centers" correspond with the autonomous ones in the XTENT model. 172 15 (fJ 10 w I- (fJ U. 0 a: w m ::;; ::J Z 5 o 5 SETTLEMENT PATTERNS IN MESOPOTAMIA Gregory Johnson's work in the Diyala region of Mesopotamia, to the east of Baghdad in modern Iraq, provides a good illustration of the way in which Central Place Theory can be applied 10 archaeological survey results. Thirtynine settlements of the Early Dynastic • • I10 15 20 25 SET TLE MENT SIZE IN HECTARES Site hierarchy for 39 settlements in the Diyala region, expressed as a histogram. AI is usually the case with such hierarchies, there is a decline in the number of sites as site size increases. There are normally ITlBIrf more small vii/ages and hamlets in a settlement system than large towns or citit! Any analysis of this kind has to make certain assumptions - for instance, that evidence for sites in each category has been uniformly preserved, which may not always be the case. period (c.2800 BC) are kno~n from this area.They range in size fro~ 25 ha (60 acres) to just over one-tenth of a hectare (0.25 acre), and on this basis Johnson divided them into five categories: large towns, towns, large villages, small villages, and hamlets. The distribution of sites suggested that there were four lattice cells, each lattice cell being the network of settlements grouped around a first order center or central place. In theory, each cell should have had a large town at the center, towns at each of the four corners, and large villages at the midpoints between the towns and at the N Large town Town Large village Village Hamlet ;.~"~-" Ancient watercourse -s-..-. Modem watercourse 15 miles I I 25 km 5 How Were Societies Organized?Social Archaeology mid-points between the towns and the large towns. Small villages and hamlets completed the pattern to create a model settlement lattice which could be compared with the real pattem as revealed by the Diyala survey. Discrepancies could then be identified and explained. It is precisely the discrepancies from the expected pattern that are of interest. Johnson found that maximization of usable land (which would have been implied had there been even spacing of settlements) was less significant in determining settlement location than were water Early Dynastic settlement pattern in the Diyala region of Iraq, based on survey work originally carried out by Robert Adams. transport networks. Settlements of successively smaller size were located along watercourses - lines of communication - between the larger settlements. Nevertheless, it was only after considerable modification that the lattice model could be made to fit the Diyala evidence. Several of the predicted primary and secondary centers were lacking, while others were smaller than they were expected to be. Thus, though the exercise was certainly valuable, it highlighted the difficulties of applying Central Place Theory to a real archaeological case. 590 Derivation of the proposed settlement lattice for the Diyala region, from the idealized, regular four lattice cells (top) to the final pattern (above) that seemed best to fit the actual settlement locations on the ground. 173 PART II Discovering (fle Variety of Human Experience Although the XTENT model can never offer more than a simple approximation of the political reality, it does allow a hypothetical political map to be constructed from appropriate survey data (see illustration on p. 172). By methods such as these, information derived from settlement surveys can be used to produce what is in effect a political and administrative map, even though such maps will always rely on certain basic assumptions that cannot easily be proved. And while the exa mples given in the box above have been drawn from studies of state societies, it is possible to apply similar techniques to the settlement patterns of less complex societies, such as the Neolithic of southern Britain (see box, pp. 190-91). In the Iron Age of south· ern Britain, more hierarchically organized societies developed, with prominent hillforts dominating the tribal territories. A pioneering analysis by David Clarke interpreted the social position of the Iron Age site of Glastonbury (see box, pp. 38-39) in these terms, as belonging within a territory dominated by such a fortified center. FURTHER SOURCES OF INFORMATION FOR SOCIAL ORGANIZATION If the first approach by archaeologists to the study of social organization must be through the investigation of settlement and settlement pattern , this should not exclude other possible avenues of approach, including the use of written records, oral tradition. and ethno- archaeology. Here it is appropriate to mention the argument of Lewis Binford, that if we are to bridge the gap between the archaeological remains and the societies those remains represent we need to develop a systematic body of what he terms Middle Range Theory. For the moment, however, we believe it is difficult to justify the division of archaeological theory into high, middle, and low. We choose not to use the term Middle Range Theory. Some scholars also lay great emphasis on the concept of analogy. Arguments by analogy are based on the belief that where certain processes or materials resemble each other in some respects, they may resemble each other in other ways also. Thus it may be possible to use details from one body of information to fill the gaps in another body of information from which those details are missing. Some have considered an analogy a fundamental aspect of archaeological reasoning. In our view this emphasis is misplaced. It is true that archaeologists use information from the study of one society (whether living or dead) to help understand other societies they may be interested in, but these are usually in the nature of genera l observations and comparisons, rather than specific detailed analogies. Written Records For literate societies - those that use writing, for instance all the great civilizations in Mesoamerica, China, Egypt, and the Near East - historical records can a nswer many of the social questions set out at the 174 beginning of this chapter. A prime goal of the archaeo· logist dealing with these societies is therefore to find appropriate texts. Many of the early excavations of the great sites of the Near East had the recovery of archives of clay writing tablets as their main goal. Major finds of this kind are still made today - for example, at the ancient city of Ebla (Tell Mardikh) in Syria in the 1970s, where an archive of 15,000 clay tablets yielded evidence of a previously unknown lan· guage and state of the 3rd millennium BC. In each early literate society, writing had its own functions and purposes. For instance, the clay tablets of Mycenaean Greece, dating from c. 1200 BC, are almost without exception records of commercial trans· actions (goods coming in or going out) at the Mycenaean palaces. This gives us an impression of many aspects of the Mycenaea n economy, and a glimpse into craft organization (through the names for the different kinds of craftspeople), as well as intro· ducing the names of the offices of state. But here, as in other cases, accidents of preservation may be impor· tant. It could be that the Mycenaeans wrote on clay only for their commercial records, and used other, per· ishable materials for literary or historical texts now lost to us. It is certainly true that for the Classical Greek and Roman civilizations, it is mainly official decrees inscribed on marble that have survived. Fragile rolls of papyrus - the predecessor of modern paper - with literary texts on them, have usually only remained intact in the dry air of Egypt, or buried in the volcanic ash covering Pompeii (see box, pp. 22-23). An important written source that should not be overlooked is coinage. The findspots of coins give interesting economic evidence about trade (Chapter 9). But the inscriptions themselves are informative about the issuing authority - whether city-state (as in ancient Greece) or sole ruler (as in Imperial Rome, or the kings of medieval Europe). The decipherment of an ancient language transforms our knowledge of the society that used it. The brilliant work of Champollion in the 19th century in crJcking the code of Egyptian hieroglyphs was mentioned in Chapter 1. In recent years, one of the most significant advances in Mesoamerican archaeology has come from the reading of many of the inscribed symbols (glyphs) on the stone stelae at the largest centers. It had been widely assumed that the Maya inscriptions were exclusively of a calendrical nature, or that they dealt with purely religious matters, notably the deeds of deities. But the inscriptions can now in many cases be interpreted as relating to real historical events, mainly the deeds of the Maya kings (see boxes, pp. 124-25 and 388-89). We can also now begin to deduce the likely territories belonging to individual Maya centers (see box, p. 197) . Maya history has thus taken on a new dimension. A more detailed example of the value of written sources for reconstructing social archaeology is Mesopotamia, where a huge number of records of Sumer and Babylon (c. 3000-1600 BC), mainly in the form of clay tablets, have been preserved. The uses of writing in Mesopotamia may be summarized as follows: Recording information for future use Communicating current information Communicating with the gods Administrative purposes Codification of law Formulation of a sacred tradition Annals Scholarly purposes Letters Royal edicts Public announcements Texts for training scribes Sacred texts, amulets, etc. The Sumerian king list provides an excellent example of annals recording information for future use. It is extremely useful to the modern scholar for dating purposes, but it also offers social insights into the way the 3umerians conceived of the exercise of power - for ~xa mple, the terminology of rank that they used. )imilarly, inscriptions on royal statues (such as those )f Gudea, ruler of Lagash) help us to perceive how the )umerians viewed the relationship between their 'ulers and the immortals. This important kind of nformation concerning how societies thought about hemselves and the world - cognitive information - is liscussed in more detail in Chapter 10. Of even greater significance for an understa nding of he structure of Sumerian society are the tablets 5 How Were Societies Organized? Social Archaeology associated with the working or organizing centers, which in Sumerian society were often temples. For instance, the 1600 tablets from the temple of Bau at Tello give a close insight into the dealings of the shrine, listing fields and the crops harvested in them, craftspeople , and receipts or issues of goods such as grain and livestock. Perhaps most evocative of all are the law codes, of which the most impressive example is the law code of Hammurabi of Babylon, written in the Akkadian language (and in cuneiform script) around 1750 HC . The ruler is seen (illus. p.1??) at the top of the stone, standing before Shamash, the god of justice. The laws were promulgated, as Hammurabi states, "so tha t the strong may not oppress the weak, and to protect the rights of the orphan and widow. " These laws cover many aspects of life - agriculture, business transactions, family law, inheritance, terms of employment for different craftspeople, and penalties for crimes such as adultery and homicide. Impressive and informative as it is, Hammurabi's law code is not straightforward to interpret, ilIld emphasizes the need for the archaeologist to reconstruct the full social context thcH led to the drafting of a text. As the British scholar Nicholas Postgate has pointed out, the code is by no means complete, and seems to cover only those areas of the law that had proved troublesome. Moreover, Hammurabi had recently conquered several rival city states, and the law code was therefore probably designed to help integrate the new territories within his empire. Written records thus undoubtedly contribute greatly to our knowledge of the society in question. But one should not accept them uncritically at face value. Nor should one forget the bias introduced by the accident of preservation and the particular uses of literacy in a society. The great risk with historical records is that they can impose their own perspective, so that they begin to supply not only the answers to our questions, but subtly to determine the nature of those questions, and even our concepts and terminology. A good example is the question of kingship in Anglo-Saxon England. Most anthropologists and historians tend to think of a "king" as the leader of a state society. So when the earliest records for Anglo-Saxon England, The Anglo-Saxon Chronicle, which took fina l shape in about AD 1155, refer to kings around At) 500, it is easy for the historian to think of kings and states at that period. But the archaeology strongly suggests that a full state society did not emerge until the time of King Offa of Mercia in around At) 780, or perhaps King Alfred of Wessex in At) 871. It is fairly clear that the earlier "kings" were generally less significant figures 175 PART II Discovering thf Variety of Human Experience 176 The variety of historical evidence. (Left and above) Scribes were accorded high status in ancient civilizations. Among till Maya, a rabbit god (left) is shown as a scribe on an 8thcentury AD painted vase. A scribe from Classical Greek times (above left) is depicted on a 5th-century BC bowl. Egyptian military scribes (above center) record on papyrus rolls the submission of Egypt's New Kingdom foes - a relief carving from Saqqara. The Inca (above right) had no writing system as such, but kept records of accounts and other transactions using knotted ropes called quipu. Clay tablets and coins. (Left) Some of the 15,000 clay tablets discovered in the royal palace at Ebla (Tell Mardikh in modern Syria), dating from the late 3rd millennium BC. The tablets formed part of the state archives, recording over 140 years of Ebla's history. Originally they were stored on wooden shelving, which collapsed when the palace was sacked. (Below) Hoard of Arabic coins found in Gotland, Sweden, from the Viking period (8th/9th centuries AD). Coin inscriptions call be informative about dating (Chapter 4) ant! trade (Chapter 9), and also about the issuing authority. Inscriptions. (Above) The famous law code of the Babylonian king Hammurabi, c. 1750 BC. The laws are carved in 49 vertical columns on a black basalt stela, 2.25 m (7 ft 4 in) high. In this detail the king is seen confronting the seated figure of Shamash, god of justice. See also main text p.17S. Early medieval documents. (Below) An Anglo-Saxon king and his council depicted in an 11th-century AD manuscript. Historical dDqlments require careful interpretation just as much as archaeological evidence. 5 How Were Societies Organized? Social Archaeology Seals and seal impressions. (Above) Rollout impression from a cylinder seal of c. 500 BC which depicts the Persian king Darius in his chariot hunting lions. The inscription is written in the cuneiform script, like Hammurabi 's law code (left) . The scene is intended to convey the authority, strength, and dominant status of the king. Such seals were used to mark ownership or authenticity. Many thousands have been recovered from Mesopotamian sites. Oral tradition. (Below) Scenes from the Hindu epic, the Ramayana, on a late 18th-century AD temple-hanging, Mathum, India. The story describes the exploits of a great ruler (Rama) in his attempt to rescue his consort, carried off to Sri Lanka by a demon king. The legend may have its origins in southward movements of Hindu peoples after 800 BC but - as always with oral tradition - the difficulty comes in disentangling history from myth. 177 PART II Discovering the Variety of Human Experience than some of the rulers in Africa or Polynesia in recent times, whom anthropologists would term "chiefs." Thus, if the archaeologist is to use historical records in conjunction with the material remains, it is essential at the outset that the questions are carefully formulated and the vocabulary is well defined. Oral Tradition In non-literate societies, valuable information about the past, even the remote past, is often enshrined in oral tradition - poems or hymns or sayings handed on from generation to generation by word of mouth. This can be of quite remarkable antiquity. A good example is offered by the hymns of the Rigveda, the earliest Indian religious texts, in an archaic form of the language, which were preserved orally for hundreds of years, before being set down by literate priests in the mid-1st millennium AD. Similarly, the epics about the Trojan War written down by Homer in about the 8th century BC may have been preserved orally for several centuries before that time, and are thought by many scholars to preserve a picture of the Mycenaean world of the 12th or 13th century BC. Epics such as Homer's Iliad and Odyssey certainly offer remarkable insights into social organization. But, as with so much oral tradition, the problem is actually to demonstrate to which period they refer - to judge how much is ancient and how much reflects a much more recent world. Nevertheless, in Polynesia, in Africa, and in other areas that have only recently become literate, the natural first step in investigating the social organization of earlier centuries is to examine the oral traditions. Ethnoarchaeology Another fundamental method of approach for the social archaeologist is ethnoarchaeology. It involves the study of both the present-day use and significance of artifacts, buildings, and structures within the living societies in question, and the way these material things become incorporated into the archaeological record - what happens to them when they are thrown away or (in the case of buildings and structures) torn down or abandoned. It is therefore an indirect approach to the understanding of any past society. There is nothing new in the idea of looking at living societies to help interpret the past. In the 19th and early 20th centuries European archaeologists often turned for inspiration to researches done by ethnographers among societies in Africa or Australia. But the so-called "ethnographic parallels" that resulted - 178 where archaeologists often simply and crudely likened past societies to present ones - tended to stifle ne\\" thought rather than promote it. In the United State; archaeologists were confronted from the beginning with the living reality of complex Native American societies, which taught them to think rather more deeply about how ethnography might be used to aid archaeological interpretation. Nevertheless, fully· fledged ethnoarchaeology is a development really of only the last 20 or 2S years. The key difference is that now it is archaeologists themselves, rather than ethnographers or anthropologists, who carry out the research among living societies. A good example is the work of Lewis Binford among the Nunamiut Eskimo, a hunter-gatherer group of Alaska. In the 1960s Binford was attempting to inter· pret archaeological sites of the Middle Paleolithic of France (the Mousterian period, 180,000-40,000 years ago). He came to realize that only by studying how modem hunter-gatherers used and discarded bones and tools, or moved from site to site, could he begin to understand the mechanisms that had created the Mousterian archaeological record - itself almost cer· tainly the product of a mobile hunter-gatherer econ· omy. Between 1969 and 1973 he lived intermittently among the Nunamiut and observed their behavior. For instance, he studied the way bone debris was pro· duced and discarded by men at a seasonal hunting camp (the Mask site, Anaktuvuk Pass, Alaska). He saw that, when sitting round a hearth and processing bone for marrow, there was a "drop zone" where small fragments of bone fell as they were broken. The larger pieces, which were thrown away by the men, formed a "toss zone," both in front and behind them. Such seemingly trivial observations are the very stuff of ethnoarchaeology. The Nunamiut might nOI provide an exact "ethnographic parallel" for Mousterian societies, but Binford recognized that there are certain actions or functions likely to be corn· mon to all hunter-gatherers because - as in the case of the processing of bone - the actions are dictated by the most convenient procedure when seated round a camp fire. The discarded fragments of bone then leave a characteristic pattern round the hearth for the archae· ologist to find and interpret. From such analysis, it has proved possible to go on to infer roughly how many people were in the group, and over what period of time the camp site was used. These are questions very relevant to our understanding of the social orga niza· tion (including the size) of hunter-gatherer groups. With the benefit of his observations at the Mask site, Binford was able to reinterpret the plan of om habitation at the French Paleolithic site of Pincevent, Ethnoarchaeology: the work ofLewis Binford. (Right) From obseroations among living Nunamiut Eskimo in Alaska, Binford derived this model of bone processing around an outside hearth. Small bone fragments fall in a "drop zone" around the men, while larger pieces are thrown both in front and behind them in two "toss zones." (Below center) At the Paleolithic site of Pincevent, France, dating from about 15,000 years ago, the excavator Leroi-Gourhan interpreted three hearths as being evidence for a complex skin tent (reconstruction, center right). (Below) Binford applied his 'outside hearth model" to the three Pincevent hearths, and deduced from the distribution of bones that his model fitted the evidence better than that of Leroi-Gourhan: i.e. that the hearths lay outside, and not within atent. (Below right) Classic semicircular arrangement around an outside hearth as demonstrated by Nharo Bushmen at Ganzi, Botswana, c. 1969. ..... I ( ;. , " ~\..t Fo~WA~O 7b6S ZONE --~ BON E.. D 16TR.1 BUT ION RE:L.ATIVE TO "ME.N'!>' OUT61 DE t-IEARTH MODEL.. Plt-JCE.VE.NT No. 1 5 How Were Societies Organized? Social Archaeology "MEN'o" OUTOIDE... HEARTH MODEL. 179 PART 11 Discovering the Variety of Human Experience occupied during the last Ice Age about 15,000 years ago. The excavator, Andre Leroi-Gourhan, interpreted the remains as indicating a complex skin tent covering three hearths. Binford at the Mask site had noted how when wind direction had changed, people seated outside next to a hearth would swivel round and make up a new hearth downwind so as to remain out of the smoke. The distribution of debris around the Pincevent hearths suggested to Binford that two of them were the result of just such an event, one succeeding the other as wind direction changed and a seated worker rotated his position. He further argued that this kind of behavior is found only with outside hearths, and that therefore the excavator's reconstruction of a covering tent is unlikely. Recent analysis, however, suggests that these hearths had slightly different functions. Work at Pincevent, and other similar sites in the Paris Basin, is finding useful insights, as well as errors, both in Leroi-Gourhan's very focused interpretations and in Binford's generalized observations from ethnoarchaeology. Ethnoarchaeology is not restricted to observations at the local scale. The British archaeologist Ian Hodder, in his study of the female ear decorations used by different tribes in the Lake Baringo area of Kenya, A· type of ear flap (Njemps) B-type of ear flap (Njemps) C-type of ear flap (Tugen) Metal~coi l ear decoration (Pokot) Ear decoration (Njemps) Tugen Poko! Njemps undertook a regional study to investigate the extent 10' which material culture (in this case personal decora· tion) was being used to express differences belweeD the tribes. Partly as a result of such work, archaeologists no longer assume that it is an easy task 10 take archaeological assemblages and group them inlo regional "cultures," and then to assume that each "cul· ture" so formed represents a social unit (see Chapter 12). Such a procedure might, in fact, work quite wen for the ear decorations Hodder studied, because the people in question chose to use this feature to assert their tribal distinctiveness. But, as Hodder showed, tl we were to take other features of the material cultwe, such as pots or tools, the same pattern would not nec· essarily be followed. His example documents the important lesson that material culture cannot be used by the archaeologist in a simple or unthinking manner in the reconstruction of supposed ethnic groups. At this point it is appropriate to move on to consider how one actually sets about systematically searching for evidence of social organization in archaeological remains, using the techniques and sources of inform· ation just outlined. Here we will find it useful to look first at bands, then segmentary societies, and finallyal chiefdoms and states. 5 miles Bkm Mukutan Ethnoarchaeology: the work of Ian Hodder. In the Lake Baringo area of Kenya, East Africa, Hodder studied the female ear decorations worn by the Tugen (right), Njemps, and Pokot tribes, and showed on a map (left) how these ornaments were used tl assert tribal distinctiveness. Other features of the material culture (e.g. pots or tools) would reveal a different spatial pattern. 180 ANCIENT ETHNICITY AND LANGUAGE Ethnicity (i.e. the existence of ethnic groups, including tribal groups) is difficult to recognize from the archaeological record. For example, the view that Mousterian tool assemblages represented different social groups, as suggested by Francois Bordes, has been criticized (see discussion in Chapter 10); and the notion that such features as pottery decoration are automatically a sign of ethnic affiliation has been questioned. This is a field where ethnoarchaeology is only now beginning to make some headway. One field of information, however, once overused by archaeologists, has in recent years been much neglected: the study of languages. For there is no doubt that ethnic groups often correlate with language areas, and that ethnic and linguistic boundaries are often the same. But it should also be remembered that human societies can exist quite well without tribal or ethnic affiliations: there is no real ne8d to divide the social world up Into named and discrete groups of people. Ethnicity should not be confused with race, which insofar as it exists (Chapter 11) is a physical attribute, not a social one. The ethnos, the ethnic group, may be defined as "a firm aggregate of people, historically established on a given territory, possessing in common relatively stable peculiarities of language and culture, and also recognizing their unity and difference from other similar formations (self-awareness) and expressing this in a self-appointed name (ethnonym)" (Oragadze 1980, 162). 5 How Were Societies Organized? Social Archaeology This definition allows us to note the following factors, all of them relevant to the notion of ethnicity: 1 shared territory or land 2 common descent or "blood" 3 a common language 4 community of customs or culture 5 community of beliefs or religion 6 self-awareness, self-identity 7 a name (ethnonym) to express the identity of the group 8 shared origin story (or myth) describing the origin and history of the group Ethnicity, however, is a much-abused term, and one that is sometimes used to mask directly political motives. Since 1992, for instance, within the former republic of Yugoslavia, there has been serious fighting between Serbs, Croats, and others (mainly Muslims) over territories. The irony is that there are relatively few underlying differences among the communities involved, the principal distinctions being religious (Orthodox Christian, Roman Catholic, and Muslim respectively). It is sad that blind prejudice along ethnic and religious lines which underlay the horrors of the Holocaust during World War II should once again lead to the mindless slaughter in Yugoslavia termed "ethnic cleansing." The perversion of ethnicity is the curse of our century. It seems likely that in some cases the scale of the area in which a language came to be spoken was influential in determining the scale of the ethnic group that later came to be formed. For instance, in Greece in the 7th and 6th centuries BC the political reality was one of small, independent city states (and some larger tribal areas). But in the wider area where Greek was spoken there was already an awareness that the inhabitants were together Hellenes (i.e. Greeks). Only Greeks were allowed to compete in the great Panhellenic Games held every 4 years in honor of Zeus at Olympia. It was not until later, with the expansion of Athens in the 5th century BC and then the conquests of Philip of Macedon and his son Alexander the Great in the next century, that the whole territory occupied by the Greeks became united into a single nation. Language is an important component of ethnicity. In Mesoamerica, Joyce Marcus has drawn on linguistic evidence in analyzing the development of the Zapotec and Mixtec cultures. She notes that their languages belong to the Otomanguean family, and follows the assumption that this relationship implies a common origin. Marcus and Kent Flannery, in their remarkable book The Cloud People (1983), seek to trace through time "the divergent evolution of the Zapotec and Mixtec from a common ancestral culture and their general evolution through successive levels of sociopolitical evolution" (Flannery and Marcus 1983,9). They see in certain shared elements of the two cultures the common ancestry suggested by the linguistic arguments. Using glottochronology (Chapter 4) Marcus suggests a date of 3700 BC for the beginning of the divergence between the Zapotec and Mixtec; she then seeks to correlate this with archaeological findings. It is questionable whether glottochronology can be used in this over-precise way. But this criticism in no way undermines the relevance of her introduction of the Zapotec and Mixtec languages into the discussion of the social evolution of the two cultures. 181 drawn here between segmentary societies and centralized ones. In segmentary societies, craft production is mainly organized at the household level - what the American anthropologist Marshall Sahlins in his book Stone Age Economics (1972) has termed the Domestic Mode of Production. In more centralized societies such as chiefdoms and states, on the other hand, though the household unit may still play an important role, much of the production will often be organized at ahigher, more centralized level. This distinction is useful at the practical level of survey and excavation. Even small villages in segmentary societies will show signs of household craft production in the form of pottery kilns or perhaps slag from metalworking. But only in centralized societies does one find towns and cities with certain quarters 5 How Were Societies Organized? Social Archaeology given over almost entirely to specialized craft production. At the great 1st millennium AD metropolis of Teotihuacan, near modern Mexico City, for instance, the specialized production of tools from the volcanic glass obsidian took place in designated areas of the city. Quarries and mines to extract the raw materials for craft production developed with the crafts themselves, and provide another indicator of economic intensification and the transition to centralized social organization. For example, the flint quarries of the first farmers of Britain, around 4000 BC, required less specialized organization than the later flint mine at Grimes Graves in eastern Britain (c. 2500 BC), with its 350 shafts up to 9 m (30 ft) deep and complicated network of underground galleries. TECHNIQUES OF STUDY FOR CHIEFDOMS AND STATES Most of the techniques of analysis appropriate to segmentary societies remain valid for the study of centralized chiefdoms and states, which incorporate within themselves most of the social forms and patterns of interaction seen in the simpler societies. The investigation of the household and degree of differentiation on the rural village site are just as relevant; so too is the assessment of the degree of intensification of farming. The additional techniques needed arise because of the centralization of society, the hierarchy of sites, and the organizational and communicational devices that characterize chiefdom and state societies. Once again, it is the nature of these devices that interests us, not simply the classification of society into one form or another. Identifying Primary Centers Techniques for the study of settlement patterning were discussed earlier in the chapter. As indicated there, the first step, given the results of the field survey, is to consider the size of the site, either in absolute terms, or in terms of the distances between major centers so as to determine which are dominant and which subordinate. This leads to the creation of a map identifying the principal independent centers and the approximate extent of the territories surrounding them. The reliance on size alone, however, can be misleading, and it is necessary to seek other indications of which are the primary centers. The best way is to try to find out how the society in question viewed itself and its territories. This might seem an impossible task until one remembers that, for most state societies at any rate, written records exist. Their immense value to the archaeologist has already been outlined. Here we need to stress their usefulness not so much in understanding what people thought and believed - that is the subject of Chapter 10 - but in giving us clues as to which were the major centers. Written sources may name various sites, identifying their place within the hierarchy. The archaeological task is then to find those named sites, usually by the discovery of an actual inscription including the name of the relevant site one might for example hope to find such an inscription in any substantial town of the Roman empire. In recent years, the decipherment of Maya hieroglyphs has opened up a whole new source of evidence of this sort (see box, overleaO. In some cases, however, the texts do not give direct and explicit indications of site hierarchy. But placenames within the archive can sometimes be used to construct a hypothetical map by means of multidimensional scaling - a computer technique for developing spatial structure from numerical data. The assumption is made that the names occurring together most frequently in the written record are those of sites closest to each other. The British archaeologist John Cherry has developed such a map for the lands of the early Mycenaean state of Pylos in Greece (c. 1200 BC) (see box, pp.198-99). Even myth and legend can sometimes be used in a systematic way to build up a coherent geographical picture. For instance, the so-called "Catalogue of Ships" in Homer's Iliad, which indicates how many ships each of the centers of Greece sent to the Trojan 195 PART II Discovering the Variety of Human Experience War, was used by Denys Page to draw an approximate political map of the time. It is interesting to compare it with a map drawn using only the hard archaeological data for fortified sites and palace centers in Mycenaean Greece (see illus. below): the archaeological and the historical pictures correlate very well. Usually, however, site hierarchy must be deduced by more directly archaeological means, without placing reliance on the written word. The presence of a "highest-order" center, such as the capital city of an independent state, can best be inferred from direct indications of central organization, on a scale not MACHAO N ... FCiRTIFIED SITES l50miles t 80km ! Late Bronze Age Greece: a map of tenitories derived from Homer's Iliad (top) compares well with a territorial map (above) based solely on archaeological evidence. 196 exceeded elsewhere, and comparable with that 01 other highest-order centers of equivalent states. One indication is the existence of an archive (even without understanding anything of what it says) or 01 other symbolic indications of centralized organiza· tion. For instance, many controlled economies used seals to make impressions in clay as indications 01 ownership, source, or destination. The finding of a quantity of such materials can indicate organizational activity. Indeed, the whole practice of literacy and 01 symbolic expression is so central to organization that such indications are of great relevance. A further indication of central status is the presence of buildings of standardized form known to be associ· ated with central functions of high order. In Minoan Crete, for instance, the "palace" plan around a central court is recognized in this way. Therefore, a relatively small palace site (e.g. Zakros) is accorded a status which a larger settlement lacking such buildings (e.g. Palaikastro) is not. The same observation holds true for buildings of rit· ual function, because in most early societies the con· trol of administration and control of religious practice were closely linked. Thus, a large ziggurat in Mesopotamia in Sumerian times, or a large plaza with temple-pyramids in the Maya lowlands, indicates a site of high status. Failing these conspicuous indicators, the archaeo· logist must turn to artifacts suggestive of the function of a major center. This is particularly necessary for surface surveys, where building plans may not be clear. Thus, on site surveys in Iraq, workers studying the Early Dynastic period, such as Robert Adams and Gregory Johnson, have used terracotta wall cones as indicators of higher-than-expected status for the small· er sites where they are found. The cones, known to form part of the decoration of temples and other public buildings on larger sites in the region, suggest that such smaller sites may have been specialized adminis· trative centers. Among other archaeological criteria often used to indicate status are fortifications, and the existence of a mint in those lands where coinage was in use. Clearly, when settlement hierarchy is under con· sideration, sites cannot be considered in isolation, but only in relation to each other. The exercise is one 01 early political geography. Functions of the Center In a hierarchically organized society, it always makes sense to study closely the functions of the center, con· sidering such possible factors as kingship, bureau· cratic organization, redistribution and storage of goods, organization of ritual, craft specialization, and external trade. All of these offer insights into how the society worked. Here, as before, the appropriate approach is that of the intensive site survey over the terrain occupied by the center and its immediate vicinity, together with excavation on as large a scale as is practicable. Again, this is a sampling problem, where the objective of comprehensiveness must be balanced against limited resources of time and finance. In the case of smaller centers, just a few hectares in extent, an intensive area survey will be perfectly appropriate. But for very large sites, a different approach is needed. Abandoned Sites. Many of the most ambitious urban projects have been carried out at abandoned sites, or at sites where the present occupation is not of an urban character, and does not seriously impede the investigation. (The problems of continuously urban sites, Le. ones that remain major centers today, are considered below.) The first requirement, which may present practical difficulties if the site is forested, is a good topographic map at something like a scale of 1:1000, although this may not be convenient for sites several kilometers in extent. This map will indicate the location of major structures visible on the surface, and some of these will be selected for more careful mapping. On sites where extensive excavations have already been conducted, their results can also be included. Such topographic maps are among the most costeffective undertakings of modern archaeology. One of the most interesting examples is Salvatore Garfie's survey of the site of Tell el Amarna, the capital city of the Egyptian pharaoh Akhenaten, as part of the British project of survey and excavation there. The site was occupied for only 13 years in the 14th century BC, and was then abandoned. The buildings were of mud brick and are not well preserved as surface features, so the map draws heavily on excavations over the course of a century. In the New World, there have been several projects of comparable scale, one of the most notable being the University of Pennsylvania's great mapping project at the Maya city of Tikal, and similar work is now under way at several Maya sites. Perhaps the most ambitious project of all, however, has been the survey at the greatest Mexican urban center, Teotihuacan (see box, pp. 86-87). The preparation of a topographic map is only the first stage. To interpret the evidence in social terms means that the function of any structures revealed has next to be established. This involves the study of the 5 How Were Societies Organized? Social Archaeology MAYA TERRITORIES NORTH The Classic Maya lowlands of AD AMERICA 300-900 were a densely settled area with many large population centers. D The first clues to their political organization came with the discovery of "emblem glyphs," hieroglyphic compounds that seemed to identify individual cities. It is now known that 6 these combinations are the titles of Maya kings and describe each as the "divine lord" of a particular polity. The discoveries showed that the lowlands Copan were at this time divided into a dense a "mosaic" of numerous small states. Today, a lively debate continues as to what degree this arrangement reflects the full political landscape. Tikal Some scholars think such states were i! autonomous and of roughly equivalent strength and influence. Others see evidence for a hierarchical ranking between kingdoms, arranged either in Calakmul a "quadripartite" model of regional states, or a more loosely structured ~ "hegemonic" system, in which dominant powers exercised some control over subject states, without .. interfering directly in their internal Palenque affairs. These reconstructions give ~ greatest prominence to centers 0 known from surveys such as: Copan, Tikal, Calakmul, and Palenque. Caracol IJ GULF OF MEXICO Maya Political Territories in AD 731 L~: , Calakmul @Naranjo o """," :-- 'Yaxchilan EGkal ~ - \) .,'Palenque. •• • •• "Q'.. Piedras Negras "Do Territories Petexbatun • Major cities _ Smaller centers 1100 miles , , . ' " Copan 100 km Emblem glyphs (left) naming 7 of the most important Maya polities. The map shows one suggested arrangement of Maya political territories, c. AD 731. 197 1'1(f) :<: ~ @ u PART II Discovering the Variety of Human Experience MULTI-DIMENSIONAL SCALING (MDSCAL) Multi-dimensional scaling (MDSCAL) is a niultivariate statistical technique, which, like factor analysis and cluster analysis, seeks to simplify complex information. The main aim is to develop spatial structure from numerical data. The starting point is a series of units, and some way of measuring or estimating the distances between them (often in terms of similarity and difference, where a larger difference is treated as much the same as a larger distance). The method allows one to reach the best arrangement (usually in two dimensions) of the various units in terms of similarities and differences. Q)