MORTUARY PRACTICES IN THE ALTO MAGDALENA: THE SOCIAL CONTEXT OF THE “SAN AGUSTIN CULTURE* Robert D. Drennan Dept. of Anthropology University of Pittsburgh In press in Tombs for the Living: Andean Mortuary Practices edited by Tom Dillehay. Washington D.C.: Dumbarton Oaks. The tombs of Colombia’s Alto Magdalena region have always been the easiest aspect of its archeological record to clscu:s. They are its only conspicuous monumental architectural remains, and they have been known and reported as curiosities since long before the beginnings of archeology as a scholarly discipline. These tombs and the sculpture associated with them came to be the defining feature of the “San Agustin Culture," named after the modern town near which the largest number of most spectacular examples occurred. ‘his archeological culture has been defined, redefined, explicated, and compared to other archeological cultures of the New World with primary reference to its tombs and its sculptural style (e.g. Preuss 1931; Pérez de Barradas 1943; Duque Gémez 1964; Reichel-Dolmatoff 1972; Herndndez de Alba 1979; Duque Gémez and Cubillos 1979, 1983, 1988; Cubillos 1980; Gamboa Hinestrosa 1982; Sotomayor and Uribe 1987). My attention here, given the subject of this symposium, will be directed to the burial customs that the tombs represent and to the ways in which consideration of these customs can help us to understand the nature and development of the societies that produced them. I will summarize these burial customs and the archeological remains they have left for us, but the real nature of the contribution to our knowledge that I hope to make involves placing these archeological remains in the broader context of other kinds of archeological information. In this effort, I will rely on results of relatively recent studies of residential zones and regional settlement patterns (e.g. Duque Gomez and Cubillos 1981; Llanos and Duran 1983; Drennan, ed., 1985; Llanos 1988, 1990; Herrera, Drennan, and Uribe, eds., 1989; Drennan et al- 1989, 1991), I will suggest some ways in which changes in burial customs in the Alto Magdalena might fit with other kinds of evidence of social, political, and economic change through time. And finally, I will discuss how such thinking helps to advance our understanding of early complex societies and their development. I would prefer to start my discussion with the earliest known periods in the Alto Magdalena, and proceed chronologically through the sequence. It is the middle segment of the sequence, however, for which information is most abundant and which gives efus our best starting point for relating burial practice to other kinds of archeological evidence. I have thus chosen to begin the story in the middle, and to follow up with the beginning and the end. Available radiocarbon dates indicate that the monumental tombs of the Alto Magdalena were constructed primarily between about 1 and 800 AD. I will refer to this period here as the Regional Classic, adopting terminology most recently codified for the Alto Magdalena by Duque Gémez and Cubillos (1988), although my usage is not identical to theirs. My usage here is intended to make it easier to relate archeological remains from the vicinity of San Agustin to those from the nearby Valle de la Plata (all within the Alto Magdalena, see Figure 1). Unresolved issues of contemporaneity leave certain inconsistencies, particularly in where we p!ac> the beginning date of the Regional Classic, but this is not an appropriate place to address those issues.’ Their ultimate resolution will not, in any event, affect the conclusions I arrive at here, and so for present purposes these inconsistencies do not matter. Along similar lines, the length of the periods in the basic chronology available for the Alto Magdalena has bothered some scholars, in particular one of the discussants at the symposium for which this paper was written. We would, indeed, find greater chronological precision most helpful in studying various aspects of the sequence in the Alto Magdalena. For precisely that reason, I have devoted considerable attention in my own work in the Valle de la Plata to chronological refinement. These efforts have met with some successes (Drennan 1993). This is not an appropriate place to report them, however, because, among other reasons, greater chronological precision is not necessary to the achievement of the goals of this paper. ‘The commonly used scheme of three long periods divides the sequence precisely where it needs to be divided to make the points I wish to make here. The fact is that the period when monumental tombs were commonly made in the Alto Magdalena is some 800 years long, and it is that period I wish to contrast with the preceding and following periods. There is no need here to subdivide it. At the same time, it will be readily recognized that the picture of changing social patterns in the Alto Magdalena that can be painted at present is done in rather broad strokes. Many of the other contributions to this volume are able to discuss the social and symbolic context of mortuary practices with a richness of detail that still eludes us in the Alto Magdalena. To argue that we should not address such issues until our knowledge is much more detailed, however, is a prescription for halting progress in archeology. To begin to paint the picture of society, even if the first strokes are broad, is the only way to know what kind of archeological information is needed to fill in the details. And research conducted with no clear idea of what information is sought and why it is needed is more likely to confuse than to enlighten. ‘This paper, then, attempts to identify some broad -2changes in social organization in the Alto Magdalena sequence; it does not pretend to assign full meanings to all aspects of mortuary practices. Regional Classic Period Mortuary Practices Any discussion of mortuary practices in the Alto Magdalena must recognize some deficiencies in this aspect of the region’s archeological record. In particular, preservation of bone is extremely poor as a consequence of very acid soils. ‘This means that it is rare to actually recover skeletal remains, even from a carefully excavated tomb. What would, in many other regions, be well-preserved skeletons are, in the Alto Magdalena, faint powdery traces or less. It is only by the presence nf grave goods and structures that burials are ordinarily recognized Thus it is usually necessary to work without even the most rudimentary information available from biological remains, like age and sex or even position and number of individuals interred Moreover, burials accompanied by no goods of a preservable character may be difficult to identify as burials at all. If such burials are simple inhumations in pits of modest size, even the pits can be mistaken for pits dug for other purposes (or, if one adopts a less conservative position on identifying graves, there is the risk that other kinds of pits will be counted as burials). It seems almost certain, then, that the burials of the majority of the population of the Alto Magdalena are systematically excluded from the evidence upon which the following discussion is based Another serious defect in the archeological record for the Alto Magdalena is the extent to which its remains have been damaged by treasure hunters. The earliest known written account of the statues and tombs near San Agustin, that of Juan de Santa Gertrudis’ visit in 1758, speaks of the disappointment of the priest in San Agustin (then a hamlet of five houses) that the six experienced looters he had brought from Popayan had already opened 19 tombs, and found only one small ornament of gold (Santa Gertrudis 1970, Tomo 2: 97). The most extensive early campaign of archeological fieldwork, led by Preuss in 1913 and 1914, expended considerable effort in documenting barrows already devastated by previous uncontrolled and undocumented excavations (Preuss 1931). Much of Preuss’s documented work was subsequently undone by further vandalization of the tombs before adequate protection could be provided in such a remote region. (It took Preuss 16 days on muleback to reach San Agustin from the point where he had to abandon slightly more comfortable transportation. ) The extent of such looting, together with the poor preservation of skeletal remains, has led some scholars to apply the term "temple" to several monuments that were clearly tombs. It also means that, for all the most elaborate examples (which were 23looted), we have no information whatever about objects that may have been included with the principal burials. We do not even know for certain just what was the placement of the statues and stone slabs of which the architecture was composed nor even what were the original dimensions of the earthen barrows that covered them. Between 1970 and 1972, Duque Gémez and Cubillos (1979 and 1983) carefully collated the records of the various investigations suffered over 200 years by the elaborate tombs of the Mesita A, Mesita B, and Alto de los Idolos sites near San Agustin (Figure 1). They conducted new excavations and attempted to replace the statues and stone slabs of which the tombs had been made in their original positions as nesrly °s cold be determined. These now represent as good a guess as can be made about the characteristics of the largest and most elaborate tombs of the Alto Magdalena. Figure 2 shows a plan and section of the West Mound at Mesita A as reconstructed by Duque Gémez and Cubillos. The tomb itself was a chamber some 1.5 by 3 mand not much over 1m high. Its floor was of earth and its walls and roof were built of rough stone slabs (Figure 3). Just beyond the northeastern end of this chamber were three large statues (Figure 4). Two of them, together with two other vertically set stones, held up a single large slab which formed a roof over the central statue. This whole construction was at about the natural ground level. It was evidently originally covered over completely by an earthen barrow of irregular ovoid shape, perhaps 20 by 30 m (Figure 5). Duque Gémez and Cubillos guess that it may have originally been about 3 m high at its peak. The barrow had been so thoroughly wrecked by previous excavation that no trace of any goods that might have been included with the burial was identifiable. Figure 6 shows a plan and section of Mound 1 at Alto de los Idolos, also as reconstructed by Duque Gémez and Cubillos. Here the stone slab tomb chamber, similar in size to the one described above, contained a stone sarcophagus 2.5 by 0.6 m (Figure 7). Preuss and the looters had been less thorough than usual in this case, and, in an undisturbed layer of artificially deposited soil around the sarcophagus, Duque Gomez and Cubillos found several small irregularly shaped thin pieces of gold with holes as if for sewing to clothing. There were also fragments of stone beads and sherds of pottery which may have been offerings destroyed by looters. A single statue stood beyond one end of the tomb chamber, also roofed by a large stone slab supported by four vertically set stones (Figure 8). The distance of 4.5 m between the end of the tomb chamber and the statue was occupied by a narrow elongated chamber or passageway about 0.7 m high, walled and roofed with unmodified slabs. This chamber or passageway may not have been disturbed by looters before Duque Gomez and Cubillos uncovered it, but no trace of items that might have been eaeplaced in it remained. All this had been placed on a level surface created by artificial fill and was covered in turn by an earthen barrow perhaps 2.5 m high. More than 2.5 m below these constructions of Mound 1, but following the same axis, was another tomb chamber. This small cist, 1.8 m long by 0.5 m wide by 0.5 m high, had walls, floor, and roof of very small thin stone slabs. Above it, lying on its side, was a small statue. This tomb was undisturbed when Duque Gomez and Cubillos found it and contained two small beads in the form of fish covered with gold, one similar fish-shaped bead not covered with gold, one tubular stone bead, several tiny gold beads about 3/mm across, and a polishing stone. Stone slab tombs and cists have been found at numerous sites throughout the Alto Magdalena. Figure 9 illustrates several such structures excavated by Duque Gémez (1964) at Mesita B. Those at Mesita B, of course, were in an area that contained barrows of the sort described above, but others occur at sites that have produced no statues or barrows. Their construction of stone slabs often approximates that of the main tomb chambers in the barrows; sometimes the scale and elaborateness are similar, and sometimes they are considerably smaller and simpler (Figures 10 and 11). Enough of them have been encountered in undisturbed condition to make it possible to discuss the goods included in them. These are remarkable principally for their modesty. A small number of ceramic vessels (rarely more than three or four in a single tomb), polished stone celts, beads, and sometimes small gold ornaments were included. A surprising number of apparently undisturbed tombs, including some chambers as large as those in the most elaborate barrows, contained no preserved goods at all. Burials in pits of rectangular or oval shape, are also known from several sites of this period in the Alto Magdalena. Sometimes a partial roof over the pit was made of stone slabs, sometimes not. Modest offerings of the kinds mentioned earlier were included in some of them; others had none. Regional Distribution of Tombs and Statues The elaborate material remains of mortuary practices are scattered widely across the landscape in the Alto Magdalena. They cluster together in groups ranging from one or two barrows and a half-dozen or fewer statues to as many as ten barrows on a single hill with several dozen statues. The largest number of such groupings and those with the largest numbers of statues and barrows are in the vicinity of the modern town of San Agustin, but others occur throughout a zone some 120 km long lying on the eastern flank of the Central Andean cordillera at the headwaters of the Rio Magdalena. Throughout the Alto Magdalena, which ranges in elevation from about 600 m to peaks over 4500 m above sea level, these funerary monuments concentrate heavily in thezone between 1500 m and 2000 m (cf. Sotomayor and Uribe 1987: 19). The monuments defined small restricted ceremonial precincts associated with substantial evidence of human settlement. Excavations at Cerro Guacas in the Valle de la Plata produced little evidence of domestic activities immediately adjacent to such monuments. Here, at a modest monumental complex of only a pair of small barrows and probably originally no more than five or six statues, stratigraphic test pits in the immediate vicinity of the barrows revealed very shallow deposits and low artifact densities (Drennan 1993). Less than 100 m away, however, are abundant remains of residential activities, including post-hole patterns of circular houses, ntly excavated by Jeffrey Blick. At Morelia, Llanos (1988: 39-62) documented a similarly small- scale ceremonial complex attached to a residential zone of substantial size. Reichel-Dolmatoff (1975) excavated stratigraphic trenches in deposits of residential refuse in the general vicinity of some of the largest and most elaborate tomb complexes near San Agustin. Information at a larger geographic scale is available for one portion of the Alto Magdalena, the Valle de la Plata (Figure 1), where systematic regional survey has produced settlement pattern data for an area totaling over 500 km? (Drennan, Herrera, and Pifieros 1989; Drennan et al. 1989 and 1991). In the surveyed region, the broad pattern of population distribution during the Regional Classic was similar to the distribution of barrow tombs and statues in that both were heavily concentrated in the zone between about 1500 and 2000 m above sea level. Such a pattern of population distribution makes good environmental sense in the Valle de la Plata, since this elevation band contains the most promising agricultural resources for Pre-Hispanic populations. In lower part of the valley, higher temperatures, less available moisture, and smaller areas of fertile, easily worked soils place limitations on potential production. At the opposite extreme, above 2000 m, low temperatures and waterlogged soils increasingly impede simple agriculture (Botero, Le6n, and Moreno 1989). Between 1500 and 2000 m, though, temperature and moisture conditions are very good, and substantial areas of fertile, easily cultivated soils occur on relatively gentle slopes. This zone was very attractive to the Pre-Hispanic inhabitants, and it was here that the societies of the Regional Classic period were focused. Within this favored zone, further unevenness in settlement distribution provides clues to patterns of sociopolitical organization. Figure 12 illustrates the distribution of Regional Classic period settlement in a portion of the area surveyed in the Valle de la Plata. Elevations in this area range from about 1400 m to over 2400 m. Settlement, as just noted, was concentrated between 1500 and 2000 m, but even there its density - 6 =varied considerably from place to place. Some of this variation corresponds to variability in topography and soils, but there remain clear concentrations of settlement unrelated to such environmental parameters, Between these settlement concentrations were areas where settlement was much less dense despite the fact that the agricultural resources they offered to their inhabitants were quite similar to those of the areas where the settlement concentrations occurred Regional Classic Social Organization As can be seen in Figure 12, each of the Regional Classic ceremonial sites included in the surveyed area falls within one of these general settlement concentrations. From the opposite point of view, each of the settlement concentrations contains one site with barrow tombs and statues. What we seem to be seeing here is the representation in regional settlement patterns of several small separate chiefdoms. (I use the word "chiefdom" here simply as a convenient label for a highly diverse group of societies with many forms of complex hierarchical organization but without the bureaucratic political institutions of the state.) The concentrations of settlement within a few km of each of the ceremonial sites represent the familiar centripetal tendency of complex societies to focus populations on central places. Although I describe these as concentrations of settlement, xesidences within them were still dispersed with considerable open space between them. ‘Trimborn (1949: 129-130) has compiled evidence from several early Conquest period eyewitness accounts of such concentrations of dispersed settlement in southwestern Colombia. Houses in some places were not clustered tightly enough together that early chroniclers were comfortable calling the settlements “pueblos,” but they did form clearly defined communities that the same authors identified without hesitation as "pueblos" in the other sense of that word, emphasizing social rather than spatial relations. Although the zones between the settlement concentrations of the Regional Classic in the Valle de la Plata were by no means deserted, occupation was noticeably less dense than within the concentrations, and this suggests the separateness (and quite possibly from time to time the competitiveness) of these chiefdoms. Each chiefdom would have consisted of a few thousand people in a settlement concentration with a radius of no more than about 10 km. The individuals buried in the valle de la Plata’s elaborate tombs seem most likely to have been the leaders (or "chiefs") of these societies. Presumably they lived in the general vicinities of the ceremonial locations where monuments marked the burial places of their predecessors and where they would themselves would eventually be interred. ‘he monumental nature of these burials made an emphatic and permanent statement about the importance of egiesthe individuals so commemorated. If we adopt the perspective suggested to us by the title of this symposium, "Tombs for the Living," we are led to think why such a statement might have been so important to the survivors of a deceased leader. It was, after all, those who survived who surely shouldered the principal responsibility for this elaborate mortuary treatment. And whatever message may have been communicated by this behavior, it was communicated to those who lived on after the leader’s death. Providing permanent reminders of the importance of specific past leaders is likely to be of particular importance in societies where leadership is a highly personal affair--that is, in societies where the institution of leadership is not highly developed. If leadership posit‘or: are fully institutionalized, the demise of a leader automatically creates a vacancy to be filled. Chiefdoms, however, often lack such a high degree of institutionalized leadership, which would seem logically related to the established bureaucratic systems of states. Each chief may have to do more than claim the position; he may have to create or at least actively maintain the position so that it is C there for him to occupy. In such a situation, not only who is chief, but also how powerful the chief is, or even whether there is a chief at all, depends to some extent on the personal abilities of the individual chief to persuade his people of his legitimacy and to exercise effectively whatever actual power he has--in general to use whatever means are at his disposal to induce cooperation or compliance from the people he purports to lead. one such means is to forge strong links between himself and previous effective chiefs. The commemoration of individual predecessors through permanent public monuments provides a chief (or hopeful chief) with the opportunity simultaneously to glorify those predecessors who were effective leaders (and thus their actions) and to attach himself to their tradition. It is, of course, not a novel observation that such activities are important in maintaining and reinforcing leadership and the social order in general in many societies, and it has long been noticed that deaths of leaders were pivotal events in the societies of the northern Andes. Early Conquest period accounts pay considerable attention to the elaborate treatment accorded deceased leaders in native Colombian societies (cf. Trimborn 1949: 226-232), and the Andes in general have been characterized as having a similar preoccupation. A focus on connections to previous leaders and their power can thus be a powerful political tool in the hands of those leaders’ descendants. Such legitimation is not necessarily incompatible with other means of exercising leadership, and may well combine with them. In the Alto Magdalena, however, the only works of a permanent, monumental character of which we have archeological evidence are funerary monuments created to commemorate specific =§ =individuals. There are no permanent monumental remains in the Alto Magdalena of the temples, plazas, palaces, or other kinds of structures (often of unclear function) that appear in the archeological record of some complex societies elsewhere in the Andes and in other parts of the world. It is reasonable to suggest, then, that the societies of the Alto Magdalena during the Regional Classic were ones in which the basis of leadership was highly personal (as opposed to institutional)--that these were what Renfrew (1974) has labeled "individualizing" chiefdoms, ones heavily focused on status rivalries between individual chiefs or aspiring chiefs, as opposed to "group-oriented” chiefdoms, where leadership seems to have a more collective and less personalized basis. Up to now I have concentrated primarily on the monumental character of the tombs in which important individuals were buried in the Alto Magdalena during the Regional Classic and of the statues associated with these burials. The nature of the offerings included with the burials also presents us with some interesting indications of the basis for leadership in these societies. Perhaps the most salient characteristic of these offerings is their sparseness. Compared with the architectural and sculptural elaboration of the tombs, the inventories of objects included as offerings with the corpses seem quite impoverished. This remains true, even after making allowances for the fact that the largest and most elaborate tombs were looted long ago. Archeologists are ordinarily accustomed to encounter much richer offerings in tombs far less distinguished in terms of their construction, their placement in ceremonial precincts, and their sculptural associations. Cieza de Leén, along with other sixteenth century chroniclers, makes much of the richness of the chiefs he observed in southern Colombia and the quantity of elaborate material goods included with their burials: "They have their mortuaries and tombs, in the custom of their land, with vaulted chambers, very deep, and with the opening to the east. When a lord or principal has died, they place him inside with many tears, leaving with him all the arms and clothing and gold that he has, and food as well" (Cieza de Leén 1553: 113 [Chapter XV]). In the Alto Magdalena, however, during the Regional Classic, a good number of fairly elaborate tombs contained no non-perishable offerings at all, even though the usual range of non-perishable artifacts (including pottery, flaked stone, stone beads and pendants, and gold ornaments) were clearly regarded as appropriate burial offerings and were included in modest quantities in some tombs. Such patterns of burial offerings are consistent with the notion \||that the basis of acquiring and exercising leadership in the Alto Magdalena during the Regional Classic, however personal or attached to individuals it may have been, was not primarily =gs{through the accumulation of personal wealth. While the ||individuals buried in the tombs clearly had great personal ||prestige, there is little to suggest that they possessed much ||wealth. ‘The commemoration of individuals suggested is rather of their public stature and of supernatural characteristics or connections of some kind. One general point on which all interpretations of the Alto Magdalena’s sculpture agree is that the themes represented are in some way supernatural--it would seem hard to argue otherwise for representations of people with long fangs or with two-headed crocodiles clinging to their backs and peering fiercely over their heads. Sculpture with broadly similar characteristics has been argued to be particularly appropriate in sanctifying a social order and the role of leaders in it, for example, in the case of Olmec art (Coe 1972; Drennan 1976). These artistic expressions of the supernatural characteristics or connections of certain individuals spring from and reinforce the beliefs that provide the reasons why ordinary people are willing to follow these individuals’ leadership. Belief (often religious) in the legitimacy of leaders and their directives certainly makes people more likely to heed them, and attainment of such legitimacy is often posed as an alternative to use of coercive force or control over basic resources (which makes coercive force possible) as a means of social control. Processes of sanctification could, of course, operate either in tandem with considerable concentration of resource control in the hands of leaders (i.e. personal wealth) or in lieu of much resource control. The archeological evidence just described leads one to think in the latter terms--of social integration | more heavily dependent on ideological/religious bases than on economic ones. The mortuary practices of the Alto Magdalena, then, form a basis for suggesting that societies of the Regional Classic period were characterized by minimally institutionalized leadership roles rather dependent for their force and even continued existence on justification in personal or individual terms. The principal’ foundation of this largely personal leadership would seem to lie in the realm of belief systems rather than in ability to exercise force or control basic resources, and leaders, despite the individualized nature of their positions, would seem to have had relatively little personal wealth. These | suggestions derive from a number of different characteristics of (the mortuary evidence, including the architectural elaboration of tombs, their associated sculpture, their location in apparent ceremonial precincts at the heart of regional settlement concentrations that seem to represent social units, the paucity of offerings included with burials, and the absence of archeological remains of any other kinds of permanent monuments.Increasing Confidence in Our Social Reconstruct ions There are, of course, other possible interpretations of such mortuary patterns. For instance, one can imagine a society with marked wealth differentiation--indeed, in which wealth differences were a major organizing principle--but with an ideology denying the legitimacy or even the very existence of those wealth differences. In such a society the rule of behavior for the wealthy would be, not conspicuous consumption, but covert consumption. Measures might be taken to prevent any obvious display of wealth, and such measures would probably extend to the treatment wealthy individuals received after death as well. some archeologists are tireless in their efforts not only to imagine such counter-examples to the apparent indications of the archeological record but also to find ethnographic instances to prove that the imagined counter-examples are more than just pipe dreams. Could the Regional Classic societies of the Alto Magdalena be such a counter-example? It certainly must be recognized as a Possibility. Is there some other way of looking at mortuary Practices that would enable us to come to a more certain reconstruction of social patterns? It is surely more productive to look instead at other, independent lines of evidence. No matter how careful we are at sorting out the mortuary evidence, and no matter how clever we are at thinking of ways in which the conclusions we base on it could be in error, we will probably never succeed in completely eliminating the possibility that the Pre-Hispanic inhabitants of the region are telling lies by the way they buried their dead. After all, an important point of departure for this symposium is consideration of the role that public mortuary ritual plays in living societies, that is, the kinds of public statements that mortuary practices make about establishing and reaffirming the social order. When the material remains we study as archeologists are shaped by their role in making public statements about social facts, then we risk being taken in by statements that are self-serving or worse. This realization need not cause paralysis in archeology. After all, ethnographers have always risked being lied to in depending primarily on asking a small number of informants about what people think, say, and do in the societies those informants are taken to represent. The solution to the ethnographer’s problem is not to connect informants to polygraphs (if for no other reason than that informants may really believe in the truth of the lies they may tell). Similarly, the solution to the archeologist’s problem is not a rigorous cross-cultural search for "material correlates" of particular behaviors for which there is never a counter-example. The solution, for ethnographer and archeologist alike, is to pay attention not only to the statements people make but also to observe their behavior in as many different ways as possible--in esshort to seek different, independent lines of evidence bearing on the conclusions we wish to make. To return to the reconstruction of some aspects of social organization I have drawn out of the mortuary evidence from the Alto Magdalena, we must recognize that it is based on several different, but only partially independent lines of evidence, all concerned with burial practices. The route toward being more certain of the reconstruction I have suggested (or alternatively toward rejecting it in favor of something else) lies not in worrying the burial evidence to death. it lies in turning to other completely unrelated kinds of archeological evidence. At least two principal directions immediately suggest themselves. In the first place, because of their generally conspicuous character in the archeological record, we are well aware of the tombs and statues of the Regional Classic. That no obvious, monumental, and permanent constructions for other kinds of public, ritual, or communal activities have been noted for the Regional Classic, however, does not necessarily mean that such activities were not conducted. Temples may have been constructed of perishable materials, as were residences; plazas for public assembly of one kind or another may have been integrated into residential zones; and so on. ‘The fact that such features did not have the permanent monumental character of chiefly burials has implications for the personalized nature of leadership, as argued above, but knowledge of the size, location, and character of spaces created for other public, ritual, or communal activities would help us to arrive at a more complete picture of sociopolitical organization. The lack of evidence for such things may indicate a truly overwhelming focus of public life on burial-related commemoration of specific individuals. On the other hand, this lack may simply be a consequence of the absence of detailed community-level studies applying such standard archeological approaches as intensive surface survey of single communities (with a much higher level of resolution than that possible in regional-scale survey) and excavation of substantial areas in residential zones. A second major direction in which future gathering of evidence should proceed focuses on other ways of using archeological remains to reconstruct aspects of wealth differentiation. Despite the obvious differences in prestige between inhabitants of the Regional Classic societies of the Alto Magdalena, the burial evidence has suggested little or no wealth differentiation. If this reconstruction is accurate, then there should be other evidence of evenness of wealth distribution. We would expect to see little variation in house size or construction; we would expect deposits of household refuse to yield similar ranges and frequencies of possible luxury goods for ‘all households; we would expect remains of likely preferred foods to be fairly evenly distributed; and so on. At least some of ~12-these independent lines of evidence would not be at risk from the kind of lying discussed above, since people are unlikely, for example, to be consciously or unconsciously making any kind of statement by disposing of household garbage. Similarly, the independence of such lines of evidence makes it likely that they will escape other sources of confusion or error to which the interpretation of the burial evidence may be subject. On both these scores, the relevant evidence is not currently available for the Alto Magdalena. It does not come from the excavation of tombs or from regional survey, although both of these have contributed mightily to our knowledge thus far. It can only come from intensive attention to use of space at the community level, from painstaking exposure of residential architecture, from rigorous reporting of artifact inventories organized not simply for purposes of stylistic description but in meaningful units for purposes of social reconstruction, from careful recovery and quantitative analysis of several kinds of food remains, and so on. This will require patient excavation of substantial areas at several different sites of the Regional Classic period. Some such excavation is being carried out in the Valle de la Plata as this paper is being written; more is planned; and still more will surely be required to provide the necessary complement to the information already obtained. There is no need to wait for ethnoarcheological study of households, for the development of new archeological methods, or for the elaboration of more sophisticated conceptual tools. We simply need to practice certain aspects of what has fairly widely been preached in archeology for at least two decades. The Formative Period Taking a broader diachronic view is difficult, because our ability to discuss either patterns of social organization generally or mortuary practices specifically for the Formative period (here taken to be before about 1 AD) is quite limited. Surface collections, shovel tests, and small scale stratigraphic excavations in the Valle de la Plata show a clear association of sites with statues and barrow tombs to the Regional Classic, although it would be premature to assert that the beginnings of these practices do not reach back into the Formative. Available radiocarbon dates associated with the tomb and statue complexes are strongly clustered in the first few centuries AD. A few fall slightly earlier, back to 150 BC. Duque Gémez and Cubillos (1988: 106) report a date of 800 BC + 30 for a slab tomb in a barrow at Alto de las Piedras, but they provide no information about the nature of the carbon sample or its context. A date of 555 BC t 50 (Duque Gémez 1964: 456) has a somewhat dubious history, coming from a wooden sarcophagus obtained in 1937 by Pérez de Barradas (1943: 109) at Alto de Lavapatas. Duque Gomez and Cubillos (1988: 107) attribute its actual excavation to looters. iaIn all it seems likely that the mortuary practices our attention has been focused on up to now began around the time of Christ or possibly slightly before. Since the only direct evidence for complex social organization in the Alto Magdalena has until recently been the tomb and statue complexes that flourished in the Regional Classic, it has been customary to treat the Formative implicitly as a period of egalitarian societies. In the Valle de la Plata survey area, population at the end of the Formative was probably less than one-fifth the size it reached. in the Regional Classic, so the region’s demographic scale was certainly much smaller then. On the other hand, the distinct settlement concentrations that seem to correspond to small chie‘dems in the Regional Classic are clearly present in the settlement pattern maps of the middle and later Formative as well. This suggests that such centralized social organization had its beginnings well back in the Formative. If chiefdoms did come into existence by the middle of the Formative, then they apparently managed successfully without the elaborate mortuary practices of the Regional Classic. Such issues as the extent of prestige accumulated by leaders and the existence of differentiation of wealth or resource control, however, remain entirely undocumented. To begin to deal with these subjects, we will need the kind of information called for above to Complement the burial and regional survey data for the Regional Classic. I will pass on, then, from discussion of the Formative, simply noting that the beginnings of complex societies in the Alto Magdalena may well antedate by a substantial margin the appearance of the tombs and statues that, simply because of their conspicuous nature, have always seemed such a watershed in the region’s archeological record. ‘whe Recent Period Following 800 AD or so, the construction of the permanent funerary monuments that have been the focus of this paper ceased. The suggestions that the people who created these monuments disappeared and that the Alto Magdalena was abandoned have been fueled by a documentable long period of near-total abandonment after the Spanish Conquest and by unclear and incomplete accounts of the nature and distribution of population in the Alto Magdalena in the 1530s when the Conquistadores first arrived. As archeological attention turned from a single-minded focus on tombs and statues to systematic efforts to build chronology, it became clear that there were, indeed, archeological remains of habitation in the region after the end of the Regional Classic (Duque Gémez 1964, Reichel-Dolmatoff 1975), in a period that has come to be called’ simply "Recent." In the Valle de la Plata survey area, not only were these remains present, but it is quite clear that population levels were at least a$ high as they had been during the Regional Classic and = gtprobably somewhat higher. Llanos (1990: 51) also notes an abundance of Recent settlements in the Valle de Laboyos. As Figure 13 shows, population distribution in the Valle de la Plata during the Recent has the same tendency to form settlement concentrations as it did before. Indeed, this tendency is even stronger, in that the settlement concentrations are now smaller and denser, and the intervening areas more sparsely settled than before. This finding is not consistent with notions of "cultural decline" or "social collapse” spawned by the disappearance of the conspicuous evidence of complex organization provided by the monumental burials of the previous period. Instead, it suggests a form of organization even more centralized in some way than its predecessor. Mortuary practices were quite different. A narrow shaft with a small side tomb chamber from the site of Barranquilla is illustrated in Figures 14 and 15. Offerings, consisting of a metate, two manos, and four ceramic vessels, had been placed in the fill of the shaft 60 to 90 cm below its top (Drennan 1985: 127-129). Other, larger, tombs of similar design have been opened by looters at the same site both prior to and following the 1984 excavations (Drennan 1985: 123). Figure 14 also shows Tomb 47, excavated at Alto de Lavapatas. It is a shallower, somewhat larger tomb chamber, also with an entrance shaft or step to one side. The only indication of offerings was a pile of carbonized wood and other plant matter, including maize cobs, some with kernels still attached (Duque Gémez and Cubillos 1988: 174, 187-188). There is no indication that such tombs of the Recent period were marked by any kind of permanent monuments visible on the surface after the tombs had been filled in. These tombs, then, would not have functioned to reinforce the roles of leaders nearly as well as those of the Regional Classic. And yet, the intensification of the pattern of concentrated settlement seen earlier suggests that centralized leadership continued to exist, possibly in even stronger form. ‘The changes in mortuary practices raise the possibility of changes in the base upon which chiefs‘ ability to exercise their leadership rested. Personal claims to legitimacy, previously bolstered through mortuary practices commemorating past leaders, may have become less important. It is easy to think of at least two bases for leadership that may have become more important. The first is more complete institutionalization of positions of leadership, giving them an existence and power apart from their individual occupants. The second is expanded exercise of economic forms of control. Increased development of either of these would not necessarily imply that legitimation through connections to previous leaders disappeared entirely-- only that it may have lost some of the centrality I have argued that it previously had.We might expect greater institutionalization of leadership to be reflected in more standardized symbols of the office occupied-- symbols likely to be more important and conspicuous during the lifetime of a chief than after his death. Things like the nature of the chief's house, special activities conducted there, and special artifacts that he used might set him apart more markedly. Precisely such things characterized some southern Colombian societies at the time of the Conquest: "The lords or chiefs and their captains have very large houses, and at their doors are placed thick canes of the kind they have there, which resemble small beams; on the tops of these they have placed many heads of their enemies” (Cieza de Le6n 1553: 112-113 [Chapter XV]). Trimborn (1949: 217-220) lists fine cotton garments, elaborate feathers, gold implements and orram2nts, and other special symbols of authority used by Conquest period chiefs according to various early accounts. Very similar kinds of evidence could also relate to economic power as reflected in greater wealth accumulated by the chief and passed down to his descendants. Solidification of institutions of leadership might call for increased emphasis on activities of a communal or ritual nature focused on those institutions rather than on deceased individual chiefs. Plazas, temples, or other spaces might be created for the regular occurrence of such activities. Chroniclers like Cieza de Leén often note the absence of temples, but chiefs’ houses are described as having cages for captives taken in war: “They feed them very well and when they are fat they take them out to the plazas, which are next to the houses, and on feast days they kill them with great cruelty and eat them" (Cieza de Le6n 1553: 128 [Chapter XX}). Unfortunately, the Conquest-period societies of the Alto Magdalena are not among those for which highly useful eyewitness accounts are available, so we must rely on multiple lines of archeological evidence to reconstruct them. At least a little archeological evidence aside from mortuary patterns and settlement distribution already referred to is already available. Ordinary residences of the Recent period have been excavated at Potrero de Lavapatas (Duque Gémez 1964: 232-241), La Estacién (Duque Gomez and Cubillos 1981: 19-110), and Quinchana (Llanos and Duran 1983: 35-60). These houses, evidenced by posthole patterns, were circular, ranging in diameter from about 2.5 to 5.5 m. ‘Tombs of the general sort already described for the Recent period were intermingled with residences at all these sites. Although the small stratigraphic tests at Barranquilla were not large enough to reveal the posthole patterns of houses, tombs were spread through an area of dense residential debris (Drennan 1985: 117-129). At La Estacion a single larger circular structure, some 9 m in diameter, was also uncovered (Duque Gémez and Cubillos 1981: 25- 36). Whether this was a ceremonial structure or chief’s house = 16 —(both of which notions have been suggested by its excavators) or something else altogether is not entirely clear, but it could be evidence for the kind of social changes postulated. (We do not, of course, know that such structures were not present during the Regional Classic, owing to the small amount of excavation data available for residential areas.) The general range of offerings in Recent period tombs, like those of the Regional Classic, is limited, however, and this lends little support to the notion of increased unevenness in the distribution of wealth, but complementary data from such sources as refuse associated with different houses are not available. The notions suggested, then, about social changes from the Regional Classic to the Recent require further investigation. Conclusion Consideration of mortuary practices in combination with other kinds of archeological evidence has led me to suggest that the Regional Classic in the Alto Magdalena, and perhaps the Formative before it as well, were periods of vigorous development of small- scale chiefdoms. In some respects these chiefdoms resembled those documented in written sources for the early sixteenth century in the northern Andes. They may have differed, however, particularly in the extent to which accumulation of personal wealth was an important principle of social organization and possibly in the extent to which positions of leadership were institutionalized. Subsequent societies of the Recent period in the Alto Magdalena may have shown greater development of one or both of these features, although even the archeological evidence available at present fails to fit perfectly with what we would expect of societies like those that populate the written accounts of the sixteenth century for the northern Andes. The notion of associating the most spectacular monumental remains with a period of relatively undeveloped institutions of leadership runs counter to the implicit assumptions most often made about the Alto Magdalena, but it strongly parallels Dillehay’s analysis of Mapuche mound building in Chapter ?? of this volume. If the outline of social evolution I have sketched for the Alto Magdalena is confirmed (in some of the ways I have suggested above), it raises questions whose answers have implications for our understanding of the nature and processes of development of early complex societies. The course of development from Formative through Regional Classic times can be seen as fairly consistent and continually in the direction of larger scale and more complex forms of organization. There is, however, a marked change at around the beginning of the Regional Classic, when population levels soared and the construction of monumental burial complexes flourished. This development did not burst upon the scene with no prior warning; its roots are clearly in the Formative. And yet we cannot say to what extent it involved qualitatively new forms of social, 17 -political, or economic organization and to what extent it was simply the intensification of previously existing patterns. We do not understand the nature of the relationship between the surge in population and the conspicuous monumental manifestation of hierarchical organization. Was one the cause of the other? If so, which one? Were both the result of some other process among the several that have been suggested to bring chiefdoms into being? If the Regional Classic was characterized by only very limited accumulation of personal wealth, then resource control and economic advantage generally probably did not play a major role in the emergence of the earliest chiefdoms in this region. Such a view was once argued as a generalization about chiefdoms, but it has recently been vigorously challenged. The end of the Regional Classic may mark an even more dramatic point of change in the sequence. This time, however, the change is not primarily one of demographic scale or degree of complexity, as the beginning of the Regional Classic seems to have been. To judge from the regional settlement pattern evidence, societies of the Recent period were only very slightly larger and more centralized than those of the Regional Classic. Instead, the change I have suggested is a qualitative one in the basis of the social hierarchy: increased institutionalization of positions of leadership and/or increased economic differentiation. Such changes are taken by some to be involved in the shift from chiefdom to state, but this was clearly not such a shift. This change took place within the class of chiefdom. (Or, to put it another way in case misunderstandings or disagreements about how to use the word "chiefdom" may cloud my meaning, societies of both the Regional Classic and Recent periods inthe Alto Magdalena fall toward the simple end of the complex society scale.) The sequence of social change in the Alto Magdalena, then, seems not to be adequately characterized simply as movement along an axis running from small scale simple organization to large scale complex organization. While several kinds of cultural evolutionary approaches have made a valuable contribution in calling our attention to this axis, they have encouraged us to overlook variation along other axes not parallel to this one. This consideration of changing mortuary patterns in the Alto Magdalena, as related to some other categories of archeological evidence presently available, raises issues of varying bases for authority or power in social hierarchies--varying bases not necessarily related to or best studied in the context either of the initial emergence of chiefdoms or of their transformation into states. More subtle variations in organization, like those discussed here, and a recognition of axes of variation that do not simply correspond to level of complexity, offer us an opportunity to raise (and try to answer) somewhat different questions about the fundamental processes of social change. ~ 18 -References Cited Botero, Pedro José, Jonés C. Le6n P., and Julio César Moreno 1989 Soils and Great Landscapes. In Prehispanic Chiefdoms in the Val. Plata, Volume 1: The Environmenta: Context of Human Habitation (Luisa Fernanda Herrera, Robert D. Drennan, and Carlos A. Uribe, eds.): 1-14: University of Pittsburgh Memoirs in Latin American Archaeology, No. 2. Pittsburgh. Cieza de Leén, Pedro de 1553 La_Crénica del Peri. Edicién de Manuel Ballesteros. Historia 16 (Informacién y Revistas, S.A.), Madrid, 1984. Coe, Michael D. 1972 Olmec Jaguars and Olmec Kings. In The Cult of the Feline (Elizabeth P. Benson, ed.): 1-12. Dumbarton Oaks, Washington. Cubillos, Julio César 1980 Arqueologia de San Agustin: El Estrecho, El Parador, y Mesita C. Fundacién de Investigaciones Arqueolégicas Nacionales del Banco de la Repiblica, Bogota. Drennan, Robert D. 1976 Religion and Social Evolution in Formative Mesoamerica. In The Early Mesoamerican Village (Kent V. Flannery, ed.)j: 345-368. Academic Press, New York. 1985 Archeological Survey and Excavation. In Regional Archeology in the Valle de la Plata, Colombia: A Preliminary Report on the 1984 Season of the Proyecto Argueolégico valle de la Plata (Robert D. Drennan, ed.): 117-180. “Museum of Anthropology, University of Michigan, Technical Reports No. 16. Ann Arbor. 1993 Ceramic Classification, Stratigraphy, and Chronology in Prehispanic Chiefdoms in the Valle de la Plata, Volume 2s_Ceramics--Chronoloay and Craft Production (Robert D. Drennan, Mary M. Taft, and Carlos A. Uribe, eds.). University of Pittsburgh Memoirs in Latin American Archaeology. Pittsburgh. Drennan, Robert D., Luisa Fernanda Herrera, and Fernando Pifieros S: 1989 Environment and Human Occupation. In Prehispanic Chiefdoms in the Valle de la Plata, Volume 1: The Environmental Context of Human Habitation (Luisa Fernanda Herrera, Robert D. Drennan, and Carlos A. Uribe, eds.): 228-234. University of Pittsburgh Memoirs in Latin American Archaeology, No. 2. Pittsburgh.Drennan, Robert D., Luis Gonzalo Jaramillo, Elizabeth Ramos, Carlos Augusto Sénchez, Maria Angela Ramirez, and Carlos A. Uribe 1989 Reconocimiento Arqueol6égico en las Alturas Medias del valle de la In V_ Congreso _Nacioi de Antropologt. s_de]_simposio de Arqu: a Antropologia Fisica (Santiago Mora C., Felipe Cardenas A., and Miguel Angel Rold4n, eds.): 119-157. Instituto Colombiano de Antropologia and Universidad de los Andes, Bogota. 1991 Regional Dynamics of Chiefdoms in the Valle de la Plata, Colombia. Journal of Field Archaeology, 18: 297-317. Duque Gémez, Luis 1964 Exploraciones Arqueolégicas en San Agustin. Revista Colombiana de Antropologia, Suplemento No. Imprenta Nacional, Bogota. Duque Gémez, Iuis and Julio César Cubillos i979 colo. de San Agustin: Alto de Idolos Monticulos as. Fundacion de Investigaciones Arqueolégicas Nacionales del Banco de la Republica, Bogota. 1981 Arqueologia de San Agustin: La _Estaci6én. Fundacién de Investigaciones Arqueolégicas Nacionales del Banco de la Repiblica, Bogota. 1983 Arqueologia de San Agustin: Exploraciones y Trabajos de Reconstruccién en las Mesitas Ay B. Fundacién de Investigaciones Arqueolégicas Nacionales del Banco de la Repiblica, Bogota. 1988 Arqueologia de San Agustin: Alto de Lavapatas. Fundaci6n de Investigaciones Arqueolégicas Nacionales del Banco de la Repiblica, Bogota. Gamboa Hinestrosa, Pablo 1982 La Escultura en la Sociedad Agustiniana. Ediciones Centro de Investigacién y Educacién Cooperativas (Universidad Nacional de Colombia), Bogota. Hernéndez de Alba, Gregorio 1979 La Cultura Arqueolégica de San Agustin. Carlos Valencia Editores, Bogota. Herrera, Luisa Fernanda, Robert D. Drennan, and Carlos A. Uribe, eds. 1989 Prehispanic Chiefdoms in the Valle de la Plata, Volume 1: ‘The Environmental Context of Human __Habitation. University of Pittsburgh Memoirs in Latin American Archaeology, No. 2. Pittsburgh. Llanos Vargas, Héctor 1988 Arquedlogia de San Agustin: Pautas de Asentamiento_en el Cafion del Rio Granates--Saladoblanco. — Fundacién de Investigaciones Arqueolégicas Nacionales del Banco de la Repiblica, Bogota. 1990 Proceso Histérico PrehispAnico de San Agustin en el Valle de Laboyos(Pitalito--Huila) . Bogoté:FundaciéndeInvestigaciones Arqueolégicas NaCionales del Banco de la Repiblica. - 20 -Llanos Vargas, Hector and Anabella Durén de G6émez 1983 Asentamientos Prehispanicos de Quinchana, San_Agustin. Pundacién de Investigaciones Arqueolégicas Nacionales del Banco de la Repiblica, Bogota. Pérez de Barradas, José 1943 ologia _ Agustiniana: _Excavaciones _Arqueolégic. Realizadas de Marzo a Diciembre 1937. Imprenta Nacional, Bogota. Preuss, Konrad Theodor 1931 Arte Monumental Prehist6rico: Excavaciones Hechas en el Alto Magdalena y San Agustin (Colombia). Comparacién Arqueolégica con las Manifestaciones Artisticas de las Demis Civilizaciones Americanas. Escuelas Salesianas de Tipografia y Fotograbado, Bogota. Reichel-Dolmatoff, Gerardo 1972 San Agustin: A Culture of Colombia. Praeger, New York. 1975 Contribuciones al __Conocimiento de la __Estratigrafia Ceramica de San Agustin, Colombia. Biblioteca Banco Popular, Bogota. Renfrew, Colin 1974 “Beyond a Subsistence Economy: The Evolution of Social Organization in Prehistoric Europe. In Reconstructing Complex Societie An___Archaeologicai Colloquium (Charlotte B. Moore, ed.): 69-85. Supplement to the Bulletin of thi i chools of Orienta. No. 20 Santa Gertrudis, Fray Juan de 1970 Maravillas de la Naturaleza. Biblioteca Banco Popular, Bogota. Sotomayor, Maria Lucia, and Marfa Victoria Uribe 1987 EStatuaria del Macizo Colombiano. Instituto Colombiano de Antropologia, Bogota. Trimborn, Hermann 1949 Sefiorfo _y Barbarie en el valle del Cauca: Estudio sobre la Antigua Civilizaci6n Quimbaya_y Grupos Afines del Oeste de Colombia. ‘Translated from the German by José Maria Gimeno Capella. Consejo Superior de Investigaciones Cientificas, Instituto Gonzalo Fernandez de Oviedo, Madrid. Be od ey‘tay Ahote Lang ee avr Br tos toow Py a sous Figure 1: Map of the Alto Magdalena showing archeological sites mentioned in the text (solid circles), the Valle de la Plata study area, and zones of regional survey (cross- hatched) within the Valle de la Plata. -—@- ee‘aot oF wacaviTon Figure 2: Plan and section of West Mound at Mesita A (after Duque Gémez and Cubillos 1983:79).Figure 3: Principal tomb chamber (left) and statues (right) in West Mound at Mesita A.Figure 4: Statues in West Mound at Mesita A.Figure 5: General view of West Mound at Mesita A from the east. -s- 26Figure 6: Plan and section of Mound 1 at Alto de los Idolos (after Duque Gémez and Cubillos 1979: 19, 53, and 64).Figure 7: Remains of principal tomb chamber and stone sarcophagus in Mound 1 at Alto de los Idolos.Figure 8: View of excavated area in Mound 1 at Alto de los Idolos from the north.lee Jona: sm cre cones LL ~ ie. | Tou 2A rows 20 un é Figure 9: Plans and sections of a few of the stone slab tombs at Mesita B (after Duque Gémez 1964: 45, 71, 85, and 116). 30Figure 10: Stone slab tomb at Mesita A.Figure 11: Stone slab tomb at VP0051 near the town of La Argentina in the Valle de la Plata.ee -e- STATUE/BARROW COMPLEXES, Figure 12: Map of Regional Classic Period ssttlement distribution in one of the regional survey zones in the Valle de la Plata.RECENT PERIOD Figure 13: Map of Recent Period settlement distribution in one of the regional survey zones in the Valle de la Plata.ere ~ ALTO DE BARRANQUTLLA TAVAPATAS: TOUS 6 Tous «7 Figure 14: Recent Period tombs at Barranquilla (vP0002) in the Valle de la Plata and at Alto de Lavapatas near San Agustin (after Drennan 1985: 127 and Duque Gémez and Cubillos 1988: 174 and 187). i asFigure 15: View of Tomb 6 at Barranquilla.