The Collapses in the West and the Violent Ritual Termination of the Classic Maya Capital Center of Cancuen
The Collapses in the West and the Violent Ritual Termination of the Classic Maya Capital Center of Cancuen
Causes and Consequences
Abstract and Keywords
The cities of the Pasion trade route had the earliest and most violent collapses beginning before AD 740. Dos Pilas, Aguateca, other cities, hilltop fortresses, and even small villages were completely destroyed with minimal or no evidence of ritual. In contrast, the international port center of Cancuen, strategic head of navigation for the entire system, was completely “ritually terminated” with great care and respect including the mass execution of its royalty and nobles. Here we discuss the nature and possible reasons for variability in the violent demise of the western cities, the stages of the seventh to ninth century collapse process in global perspective, and implications for theories of the root causes of the Classic Maya collapse.
At Cancuen about AD 800 a singular event occurred: one of the largest termination rituals recorded in the Maya lowlands. An entire site—a high kaloomte’ royal capital of a rich riverine kingdom and the center of a sprawling economic exchange network—was ritually terminated. The entrances to the royal palace were buried in meters of clay and rubble, the monuments were carefully defaced, and much of the nobility—over fifty men, women (two of whom were pregnant), children, and the king and queen—were assassinated. Most were deposited in two sacred cisterns, one at the entrance to the palace and a second near the main plaza. But the king, Kan Maax, and his probable consort were simply buried under 60–80 centimeters of mud at the second entrance, without tombs or even cists. Strangely, these nobles and royalty were interred with fine offerings and dressed in full regalia. The Cancuen epicenter was then abandoned completely and never reoccupied.
This event also marked the complete disintegration of one of the great exchange routes of the western Petén and a signal event in the violent collapse of the Classic Maya kingdoms of the west. In terms of scale, details, and clarity of data, there are few more explicit examples of the violent ritual end of royal power in the Classic Maya lowlands.
(p.160) Furthermore, the events at Cancuen provide the best-dated chronology for the late-eighth-century period of the crisis of divine authority that is discussed in this volume. In that period, Cancuen’s contexts are very clearly divided into phases of 760–780 (Los Laureles), 780–790 (Early Chaman), and 790–800 (Late Chaman) based on historical texts, typological and modal ceramic microchronology based on midden deposits, interregional cross-dating of ample ceramic imports (sourced by instrumental neutron activation analysis), a six-stage palace construction sequence in which three reconstructions have in situ monuments with dedication dates, and a general architectural style sequence based on the palace and other well-dated architecture (e.g., Barrientos 2014).
But what happened at Cancuen? Who carried out this strange massacre? Why with such elaborate ritual? Why at that time? What does it tell us about the ideology of the eighth-century Maya states and the reasons for their disintegration? What does it tell us about the crisis of legitimacy described throughout this volume (see especially Iannone this volume)?
Seventh- to Eighth-Century Challenges to Centralization and Interregional Spheres of Hegemony
To understand the root causes of the disasters at Cancuen and elsewhere, we must go back at least a century. The crisis of legitimacy at the end of the Late Classic was not a sudden phenomenon; it was a culmination of a long process of fragmentation throughout the eighth century. This process involved challenges at even higher levels than royal power.
The period of maximum reach and power of royal authority might be seen in the epoch of the greatest kaloomte’s, the “Middle Classic” kings of kings of the greatest centers such as Tikal and Calakmul. In the fifth and sixth centuries Tikal dominated much of the Petén (Martin and Grube 2008). While direct control of the administration of other kingdoms is unlikely, the placement of allied or related rulers, tribute in exotics and commodities, and labor are reflected in the archaeological and epigraphic record of Tikal and other lowland sites in this period.
If we look at the western Petén trade corridor of the Pasión River, Vera-paz, and transversal valley land routes (figure 7.1), we see that Central Petén influence, monuments, and Tzakol-style ceramics have been found in both the highlands and Pasión River sites. Thousands of broken vessels of Tzakol-style ceramics that date to about AD 400 to 600 were found in the cave shrines that line the valley route from Cancuen at the southern frontier of (p.161)
(p.162) the Petén lowlands, deep in the highlands (Woodfill 2010; Woodfill and Andrieu 2012). Central Petén–style monuments were erected at sites such as Punta de Chimino and Tres Islas on the Pasión River route (Bachand 2010). A Petén-style stone coffer was recovered even in the piedmont and highlands at Hun Nal Ye (Woodfill et al. 2012).
Evidence of Central Petén influence, therefore, was uniformly present on both the river and land routes of the Pasión River and the highland valleys. The route to highland jade, pyrite, obsidian, quetzal feathers, Pacific conch, and other exotics appeared to have been under Central Petén control (Woodfill and Andrieu 2012), probably through merchant pilgrimage and other peaceful means. This was reflected in the great abundance of jade, obsidian, and exotics in this period at Tikal and its region and in the Tzakol presence at the trade route sites in the piedmont and highlands (figure 7.2a).
The very beginning of the process that eventually led to the decline of centralized royal power in the lowlands can arguably be traced as far back as the rise of the Calakmul hegemony and its challenges to the status quo on the Tikal-dominated western exchange system and elsewhere. In the sixth and seventh centuries, the great hegemonies of Calakmul and Tikal battled for influence in the west. Through conquest or asymmetrical alliance, Calakmul asserted authority over Dos Pilas, El Perú-Waka’, La Corona, Uxul, and other sites on the northern end of this western route and routes further west, using a complex but systematic strategy of expansion of influence. This sequence of lowland centers, which began at Cancuen, formed the major trade artery of the western Petén, including what we have called the great western exchange route of the Pasión and Usumacinta and what Freidel and others have called the “royal road” to Calakmul, the latter half via the Pasión River, but then due north by land (Canuto and Barrientos Q. 2013; Freidel et al. 2007).
The movement of the Upper Pasión kingdom to its Late Classic seat and its founding at Cancuen by Calakmul in AD 656 (Martin and Grube 2008) was a key element in establishing this unified access to or control of the Pasión River valley and the entire route. The accession of the Cancuen ruler was celebrated at Calakmul. Cancuen was positioned at the head of navigation of the Pasión River valley (figure 7.2b), the transfer point between this lowland river route and the “royal road” with the highland land route that led south through the Verapaz to many resources (Barrientos and Demarest 2007; Demarest et al. 2008, 2009). Also note that figure 7.2b shows only half of the Calakmul sphere of influence; its political machinations were (p.163) also being carried out in the east in interactions with Caracol, Naranjo, and other sites (e.g., Martin and Grube 2008).
The beginning of the gradual process of decentralization of power through the lowlands might be traced to AD 695, when Calakmul and its allies suffered a devastating defeat by Tikal (Martin and Grube 2008). While that led to a decline at Calakmul, it did not lead to reunification of the western exchange route under Tikal. Instead, political relations on the land routes became quite complex and were characterized by greater regional autonomy. Neither the Tikal nor the Calakmul superpowers regained control of the system. However, the Middle Pasión Petexbatun kingdom of Dos Pilas began a campaign of conquest and alliance in an attempt to control or at least assure access to the southwestern Petén and the Pasión River half of the route (figure 7.2c). Conquest and control of other Petexbatun sites was followed by further expansion on the river (Demarest 2006). A celebrated royal marriage alliance with Cancuen brought that site into the Dos Pilas sphere of dominance (Martin and Grube 2008, 60).
It is clear that this Petexbatun Pasión hegemony represented a great reduction in the area of hegemonic control as early as the end of the seventh century. Neither as extensive nor as active as the earlier hegemonies, Dos Pilas unified only the Pasión River portion of the trade route, despite ample evidence (discussed below) of exchange south via the gateway center of Cancuen. By the eighth century the highland Verapaz land portion of the route no longer registered Petén influence in material culture (Woodfill 2010). The Dos Pilas hegemony was limited to the Pasión River valley and did not extend, as it had previously, to the north or east on the Petén land routes to El Perú-Waka’, Calakmul, Tikal and other sites. Furthermore, unlike the Early Classic, during the seventh- and eighth-century period of Cancuen’s role as a gateway center, there is little lowland influence in the piedmont and highland Verapaz sites. In contrast, elaborate highland-oriented ritual settings were created within Cancuen itself. We can assume that control of access to those routes was in the hands of independent highland political entities.
The size and centralization of hegemonies shrank even further after Tikal defeated El Perú-Waka’ in AD 743 (and defeated Naranjo in the east in AD 744). After that, the last Dos Pilas ruler, K’awiil Chan K’inich, struggled to maintain the Petexbatun alliance. In AD 761, Dos Pilas was violently attacked, besieged, and destroyed (e.g., Demarest 2006; Demarest et al. 1997). These events were documented in many primary contexts at Dos Pilas and in other texts. The Pasión region balkanized into battling independent (p.164)
(p.165) (p.166) powers such as Tamarindito, Ceibal, Aguateca, and Santa Amelia, each of which had their own emblem glyph or fought to claim that of the Petexbatun (e.g., Demarest 1997, 2004, 2006; Inomata 1997, 2008c; O’Mansky and Dunning 2004).
At this point, AD 760–800 Cancuen was fully independent and was entering its period of greatest florescence (Barrientos and Demarest 2007; Demarest et al. 2008). However, the Cancuen dominion was only over the Upper Pasión portion of the route, as shown in figure 7.2d (Barrientos and Demarest 2007). The Middle Pasión remained in chaos, and the highland valleys were independent (and apparently not very interested in borrowing elements of lowland ritual or material culture). In the late eighth century, Cancuen’s ruler and elites began to look to the far west and the northeast for alternative routes, since the Middle Pasión was so disrupted by endemic warfare (Demarest et al. 2009; Demarest et al. 2014).
Thus, viewed in chronological order and transregional perspective, there was a sequential reduction during the Classic period in the scale of hegemonies and the area of alliance networks under the great kaloomtes (compare figures 7.2a, 7.2b, 7.2c, and 7.2d). First (figure 7.2a), the Tikal hegemony had extensive influence, if not control, of the entire western route, including the highland valleys through the Verapaz region (Woodfill and Andrieu 2012) and a parallel strategic control in Belize to Honduras (Bell et al. 2004). After that, the subsequent Calakmul hegemony (figure 7.2b) covered much of the same territory, but its power ended at Cancuen and did not register influence in the highland Verapaz valleys to the south, despite confirmed large-scale movement of resources from that zone. Its influence also did not extend southeast to the Motagua. Subsequently, the Petexbatun kingdom (fig. 3c) extended its boundaries, but only to control a smaller portion of the route from the middle to the upper Pasión. Finally, the destruction of Dos Pilas and the Petexbatun hegemony led to the disintegration of its alliance and a period of battling city-states (Demarest 2004b; O’Mansky and Dunning 2004), each with a limited range of authority. That of Cancuen (fig. 3d) covered only the upper Pasión River (Demarest et al. 2008).
Thus, the later changes that led to the demise of the kings at a local or subregional level must be seen against the backdrop of a transregional balkanization and a general reduction in the size and scale of influence of larger political units. These changes would have resulted in a more limited range of power, tribute, and patronage of the k’uhul ajaw of the respective capital centers. The rise of many individual polities and smaller zones of (p.167) power was a factor at the highest level of decentralization of power. It also led to an intensification of all forms of status rivalry, from monuments to warfare, as the reduced subregional elites vied for power. Therein lay the seeds of the late-eighth-century crisis of legitimacy (Iannone this volume; Iannone et al. this volume).
Challenges to Royal Authority: The Rise and Proliferation of Subroyal Elites
By the end of the seventh century, at the same time that these changes were under way at the interregional level, each individual polity was beginning to experience a dispersion of power because of internal changes. Particularly in the west there was a simultaneous proliferation of individual emblem glyphs and an increase in and diversification of subroyal titles and offices. Beginning in the late eighth century, the number of such titles increased significantly. From 184.108.40.206.0 (AD 652) to 220.127.116.11.0 (AD 731) in the Maya Long Count, the number of inscribed subroyal titles increased sevenfold (Jackson 2005, Fig. 3.9). Sajal, ajk’uhuun, yajaw-k’ahk’, bach’ok, ti-sak-hu’n were among the many designations for high-ranking but nonroyal lords (Houston and Inomata 2009, 173). By the eighth century, such lords had their own palaces, thrones, rich burials, and in some cases even their own courts as leaders of smaller petty kingdoms (e.g., Iannone 2005; Schwake and Iannone this volume). Bureaucratic involution, royal and high-elite polygamy, and general growth led to a dynamic in which larger courts intensified the local influence of the state, but they did so at the cost of greater internal divisions of power and probable internal status rivalry in addition to the interstate competition already present between the growing number of polities of many sizes.
The proliferation of elites was even more intense in the Usumacinta and Pasión regions of the western exchange and transport route than in other regions (Jackson 2005). There and elsewhere this reflected a balkanization of power. A similar process has been observed in more detail at sites such as Copan, where rich iconographic and epigraphic evidence has survived. Rulers such as Yax Pasaj were pressured to divide their authority with other nobles, some of whom probably controlled subsectors of the Copan Valley (Fash et al. 2004). Leadership was then coordinated and negotiated at popolna, or mat houses (Fash et al. 1992). Smaller palaces with thrones were associated with nobles with specific titles. It has even been suggested that such a division of power was an element in the internal aspects of collapse (p.168) in the Copan Valley, as the central power of the k’uhul ajaw was weakened (Fash and Stuart 1991; Fash et al. 2004).
At Cancuen, this process of the internal rise of nobles is manifested in multiple lines of evidence (e.g., Demarest et al. 2014). This shift occurred even as the site experienced its apogee in the late eighth century. Elite complexes were found at various strategic locations in the site, and each of Cancuen’s three most studied port areas was associated with such range structures (e.g., Demarest et al. 2014). Some of the site’s noble complexes had impressive three-dimensional stucco sculptures similar to ones in the royal palace itself (Barrientos 2014; Demarest et al. 2014; Demarest 2013). Monuments indicate a sharing of power between the ruler and high elites. Panel 3 at Cancuen (figure 7.3a) displays the king as a water lord, but he shares that scene with an ajk’uhuun and a sajal noble. Meanwhile, although altars set as ball court markers celebrate military victories (Fahsen and Barrientos 2006), they were proxy wars, and Altars 2 (figure 7.3b) and 3 show the k’uhul ajaw with another companion ajaw. In one case, the other ballplayer is identified as the captor of a lord from a site of Sac Witz and later as the captor of the lord of the great site of Machaquila (e.g., Fahsen and Barrientos 2006). In many ways the ballgame monuments celebrate the victories of subordinate lords more than the power of the Cancuen k’uhul ajaw. Thus, the roles of k’uhul ajaw as both master of the ports (the economic raison d’être of Cancuen) and war leader seem to have been relegated to nobles, and other roles may have been reassigned as well.
The division of power is also apparent in the central royal palace of Cancuen in its final stages in the late eighth century (figure 7.4). It reflects changes in the nature of governance. Over a decade of detailed study, excavation, and comparative analysis Barrientos and his colleagues (e.g., Barrientos 2014; Barrientos, Arriaza, et al. 2006) have shown that the huge palace served only minimally as a royal residence and was more of an administrative and ritual complex that was also used by the many nobles at the site (Barrientos 2014; Demarest 2013). The architecture consists of many small nonresidential audience chambers, likely receiving rooms with benches for subroyal officials (figure 7.4). These may have served as the offices for the nobles living elsewhere in the other range structures at the site or even some may have been audience chambers of leaders whose residences were not located at Cancuen but in satellite centers. Only two courtyards of eleven in the palace have been shown to have a residential function. These were strikingly different in size, privacy, and form from the many audiencias. This shift in central architecture at Cancuen parallels that of other (p.169)
(p.171) centers with administrative palaces, including those at Nakum, Caracol, Calakmul, Ednza, and other late-eighth-century and Terminal Classic sites (Barrientos 2014). At Cancuen the association of these patterns with shifts in economy and foreign influence is notable (Demarest 2013; Demarest et al. 2014).
All of these changes indicate another level of challenges to centralized authority. Intrapolity division of power would have led to the type of unstable political dynamics observed in most regions in the late eighth century as rivalry between nobles (and potential rulers) intensified, threatening and weakening the k’uhul ajaw.
Challenges to State Regional Authority
There is much evidence that on the level of sites within states there were also significant challenges to the central power of the states dominated by k’uhul ajaw. Increasing autonomy was already present in the rise of more centers. The number of sites with emblem glyphs spiked at the end of the eighth century, but many of these were new small satellite sites that were expressing heightened local status and degrees of authority. Some sites registered a great degree of autonomy from the k’uhul ajaw and were ruled only by sajals. The situation in areas of extreme competition between larger centers may have allowed minor centers to negotiate for greater autonomy. Numerous smaller centers arose in the zone of conflict between Piedras Negras and Yaxchilan, some of which were fortified and were located in strategic positions (Golden and Scherer 2006; Golden et al. 2008). Similarly, in Belize, the rise of petty kingdoms such as Minanha and Xunantunich have been interpreted as reflecting the decentralization of power of larger polities such as Caracol and Naranjo (Iannone 2005). This proliferation of little kingdoms is also reflected in the increase in the number of sites that erected monuments that celebrated the endings of k’atun periods; the number of such monuments peaked in AD 790 (Sharer and Traxler 2006, 504).
Challenges at the Community Level
Even at the level of local non-elite communities, an increase in autonomy can be observed. Eberl (2007, 2014) has shown that in the Petexbatun region in the early eighth century, small communities began to develop internal hierarchies among family groups and adopt symbols of militarism and (p.172) social differentiation in common ceramics (Eberl 2007). Even these more autonomous small communities were often fortified in the militarized landscape of endemic warfare in the Petexbatun (O’Mansky and Dunning 2004). Indeed, recent fine-grained chronologies indicate that some communities simply chose to leave or to move to defensible locations well before the 761 fall of Dos Pilas, leaving the major sites in a less-populated region (Eberl 2007, 2014; Eberl and Monroy 2007). Conversely, at Dos Pilas the elite abandoned the site after military defeat. Some of the non-elite local population lingered for a brief period, but as a dramatically reduced, comparatively simple community without any public architecture or evidence of elites (Palka 1995).
The degree of community autonomy in the zone of Cancuen may have been even more pronounced. The local populations in surrounding communities used local and Verapaz ceramics with minimal Petén influence or polychromes and, with a few exceptions, only earthen and cobble architecture. Many aspects of material culture in most of the surrounding communities are more similar to those of the highlands, and Cancuen can be regarded as truly defining the isolated frontier of lowland Maya civilization. Significantly, the highland trade corridor registered very little lowland influence in material culture in the late eighth century, despite the fact that it was a period of intensified exchange in jade, obsidian, shell, pyrite, and other exotics that came into Cancuen from the Verapaz route (Woodfill 2010). The mechanisms for that exchange do not appear to have involved direct control from Cancuen or even the presence of leaders from there or the use of coercive force in the highlands. As discussed below, the increased autonomy of the piedmont and highland communities may have been a crucial factor in the final demise of Cancuen.
Site-Level Differentiation and Challenges
In addition to the proliferation of subroyal elites, other indications of factionalism were present at Cancuen. A possible weakening of royal authority may be indicated by the distribution of foreign ceramics there (Forné et al. 2009, 2010). The apogee of Cancuen is in part attributable to the establishment of alternative exchange routes going west along the base of the highlands to Tabasco and Veracruz. This “transversal” exchange route is clearly marked at Cancuen by the presence, in significant quantities, of western ceramics, including variants of Chablekal Fine Grey that have been sourced by neutron activation to southern Chiapas or Tabasco (Bishop et al. 2005) (p.173) and even Campamento Fine Orange from Veracruz (Forné et al. 2010). In this same period of AD 760 to 800, Zaragoza obsidian from Puebla appeared at Cancuen in quantities more characteristic of Terminal Classic to Postclassic sites (Andrieu and Quiñonez 2010; Andrieu et al. 2011a). However, all of these influences are registered near the ports and/or in some complexes in the northern zone of the epicenter. Even more striking are distinct zones of highland influence such as that surrounding a highland-style ball court (discussed below). Thus, the epicenter of the site appears to be segmented in terms of influences, and this might have led to less centralized royal control.
Meanwhile, the royal palace and adjacent compounds have little to no Tabasco or Veracruz material culture and few highland materials (Forné et al. 2009). This pattern might indicate further internal differentiation and increasing division within the general Cancuen community at the same time that subroyal elite power and status was growing, as evidenced by the patterns referred to earlier. At all levels, then, the centralized authority of the k’uhul ajaw appears to have faced new challenges.
Another sign of internal differentiation and challenges at Cancuen can be seen in its ball courts. Extensive settlement pattern studies in Chiapas and the Petén (Montmollin 1997) have linked a high number of ball courts in a single site to internal differentiation and the increased presence of elites who were asserting their identity and competing for power within a center. Comparative studies of nearly 100 ball courts in Chiapas found a correlation between the number of ball courts in a polity and the degree political decentralization (Montmollin 1997). It is significant, then, that at Cancuen, which was not a huge center, three contemporaneous ball courts were located within 200 meters of each other in the epicenter at the end of the Classic period. The trend toward more ball courts in the Terminal Classic and Postclassic might also strengthen the indications that Cancuen experienced a “premature Terminal Classic” in the eighth century, as is also signified by its administrative palace, nobles’ palaces, and fine paste ceramics (e.g., Demarest and Martínez 2010). Differences in style and even orientation in Cancuen’s ball courts may also reflect divided identities at Cancuen.
Royal Responses to Waning Power
Divisions of authority are also indicated by attempts by the k’uhul ajaw to hold onto waning power. These changes fit into royal strategies of containment (p.174) seen at other sites in crisis at the end of the eighth century (Iannone this volume). As discussed above, such concessions at Cancuen include the facts that the administrative palace served the needs of local lords and that nobles shared the stage on monuments. These may have been indicators of royal attempts to mollify increasingly threatening elites. Other mechanisms to hold onto authority might have included military coercion. However, as noted above, such actions may not have been directly taken by rulers; they may have been relegated to other nobles.
Efforts to further impress the general populace were manifested in the exterior palace at Cancuen, which had enormous three-dimensional stucco sculptures. These projected royal power without the need for literacy in written texts (Demarest 2013). The ritual water system running through the site reinforced the imagery on monuments such as Panel 3 (fig. 3.3a), which show the ruler’s definition as a water lord (Alvarado 2011; Barrientos, Demarest, et al. 2006; Demarest 2013; Fernández Aguilar 2010). Such associations of Classic Maya rulers with water (and with blood and other liquids) have been noted generally (Lucero 2006), but at Cancuen may have specifically linked the king with his power over the river. Claims of royal control over the river, ports, and the head of navigation may have been the goal of that symbolism. As an aside, note the remarkably parallel association of rivers with the corporal essence of both ancient and modern divine African rulers (Iannone this volume; Taylor 2010). As in Rwanda, such associations also placed much responsibility on the holy king. Nonetheless, at Cancuen in even the most explicit presentations of a water lord, such as Panel 3, the scene is shared with named and titled nobles and the direct presence at ports was that of nobles’ palaces.
The most obvious attempt to consolidate weakening state power was Cancuen’s northern ball court (Torres 2011). There, close relations with piedmont and highland neighbors were clearly being solicited. The northern ball court is unlike any yet found in the Maya lowlands. It is a 24-meter-long sprawling highland-style court (figure 7.5). Between its low mounds the playing alley was defined by sloping unworked stone slabs like those found in ball courts in the highland Verapaz and Quiche regions, which were the source zones of many of Cancuen’s imports, including jade. Surrounding the ball court were massive middens that included pieces of serving vessels in local piedmont styles. It appears to have been a true feasting ball court like those found in the highlands and southern periphery (Fox 1996; Torres 2011). The most probable interpretation is that Cancuen’s late-eighth-century rulers were hosting highland caciques and elites as part of a (p.175)
ritualized exchange relationship, even though in the highlands themselves lowland influence was no longer registered at sites or cave shrines in the Late Classic period.
Thus, Cancuen’s access to highland resources and jade, which is so clearly manifested in its material culture, depended not on coercion or physical presence but on consensual and ritualized mechanisms. These were elaborate steps on the part of the state to maintain access to the piedmont and highland Verapaz Valley trade route.
The Sequence of Collapse in the West
The collapse of Maya kingdoms began earliest in the west, in the Middle Pasión River Valley, between AD 695 and 760. This process appears to have begun prior to major drought, deforestation, or the other factors that are often proposed as causes of the end of Classic period society. In the west, as throughout the southern lowlands, the kingdoms were destabilized by problems created by the very success of Classic Maya civilization and the growth of its ritual/religious/political state structure. That political system (p.176) generated and reinforced power with massive ceremonies in the architectural theaters in which elites performed in beautiful and semiotically loaded costumes and props for these Negara-like spectacles (Geertz 1980). However, the very successes of the system led to tremendous pressures because of the proliferation of elites through polygamy, alliance, and bureaucratic involution, leading to greater demands on all aspects of the western exotics and commodities exchange systems.
Meanwhile, population growth, ecological stress, drought, and all of the various manifestations of proximate (final, immediate) causes of collapse, decline, or transformation developed in other forms in different regions. In any of these cases of regional or subregional crises, counterproductive strategies of containment such as those described above only added to the costs of maintaining states. Furthermore, such strategies of legitimation at Cancuen and elsewhere only increased further the linkage between the ruler and the increasingly difficult state of affairs—economic, ecological, political, and ideological—in the kingdom.
Specifically, the splendor and royal legitimacy of western cities such as Cancuen, Dos Pilas, Waka’, Piedras Negras, Altar de Sacrificios, Yaxchilan, and Ceibal, was due in large part to their strategic positions on the great western routes of exchange and transport of precious exotics (jade, pyrite, Pacific shell, quetzal feathers, etc.) and commodities such obsidian and (surely) cacao, cotton, and salt (e.g., Demarest et al. 2014). As elite populations, centers, palaces, hegemonies, titles, and patronage increased in the Petén, warfare on the exchange route intensified, in part driven by the increasing demands for the goods imported on that route, which were needed for greater legitimation efforts, patronage, and status rivalry.
In the Petexbatun and Middle Pasión, the frequency and intensity of conflicts increased, culminating in the abandonment of some landscapes and crowding into fortified or defensible centers in others. Sixty years of intensive archaeology and ceramic studies that span from Gordon Willey’s work in the 1950s through the Petexbatun project of the 1980s and 1990s to Markus Eberl’s ceramic microchronology in the 2010s (e.g., Sabloff 1975; Foias and Bishop 1997; Eberl and Monroy 2007; Eberl 2014) have provided specific information about the history of this process. In the period AD 700 to 750, some Petexbatun village populations began to move into defensible locations or simply left the region. By the mid-eighth century, the settlement strategy in many western regions had been reduced to a single variable: defensibility (O’Mansky and Dunning 2004). The exchange system broke under the strain of its success and the increasing demand for (p.177) exotics and the value of the route that supplied them. Sites were fortified (even villages) and centers were destroyed. Some centers such as Dos Pilas were abandoned or were left with scattered populations at 5 to 10 percent of previous levels (Palka 1995). Like dominoes falling, other centers in the greater Pasión Valley were destroyed or abandoned. In the end only a few sites remained, such as Seibal, Punta de Chimino, and Altar de Sacrificios, which were Terminal Classic highly defensible enclaves that represented a very different variant of lowland Classic Maya culture (e.g., Demarest 2004b, 2006; Tourtellot and Gonzalez 2004).
There was remarkably little ritual associated with the dramatic collapse of some Middle Pasión states. They score very high on two of the three variables emphasized in this volume—violence and collapse—but not so high on ritual. At Dos Pilas, there was were some resetting of monuments at the very end and a pit with a dozen decapitated skulls of adult males, undoubtedly warriors, adjacent to the breached defensive walls (Johnson et al. 1989). This deposit seems to be captive sacrifice. More significant termination rituals were found at the rapidly abandoned site of Aguateca (Inomata 2008c; Inomata this volume). However, at Cancuen, excavations recovered far more ritual activity associated with its abrupt ending. During the period of the Petexbatun and Middle Pasión collapse, the kingdom of Cancuen to the south (located at the head of navigation of the upper Pasión River) boomed, creating detours around the crisis zones to the north and access to three separate and important western routes (see figures 7.1 and 7.2d): 1) the river trade route on the Pasión to Tres Islas and El Raudal and then via Machaquila to the central Petén and east; 2) access to the Verapaz valley route through the highlands to the jade sources in the Quiche, past the sources of quetzal feathers and pyrite, and then south to the source of obsidian at El Chayal, then on to Kaminaljuyu and the Pacific Coast; and 3) the east-west transversal route that runs along the sharply defined base of the southern highlands into Chiapas and then northwest, keeping along the base of the hill ranges toward Tabasco, Veracruz, and beyond (Demarest et al. 2009, 2014; Forné et al. 2010). From its strategic position at the head of navigation of the Pasión, Cancuen could monitor or even control the movement of exotics such as jade, quetzal feathers, and pyrite and transversal commodities such as salt, cacao, and cotton. In addition to the importance of these routes and the changing economic and political patterns they point to, this exchange system also identifies some possible suspects for the royal mass assassination and the elaborate site termination ritual that ended Cancuen.
(p.178) The Violent Termination at Cancuen
The most dramatic evidence of the Cancuen termination ritual comes from excavation of the southern palace cistern of the site (figure 7.6). A beautiful plaster-lined red-painted cistern of eight meters by ten meters and over two meters deep was located at the entrance of the royal palace. Fed by the ritual water system of the epicenter (e.g., Barrientos, Demarest, et al. 2006) and by a subterranean spring, it would have been a highly sacred location. Stone steps lead into it, and it might have functioned for ablutions prior to ascending the great staircase and passing the audience chambers to enter the royal palace itself. Excavations of the cistern in 2004 to 2005 recovered an enormous skeletal sample, over 600 bones that had been deposited there at the time of the ritual destruction of the site in about AD 800 (a date confirmed by the artifactual microchronology of Cancuen, texts, ceramic cross-dating, the palace architecture sequence, etc.).
The renowned team of the Forensic Anthropology Foundation of Guatemala (FAFG) conducted the field excavations and analysis of the system. The team normally works for the World Court and the Guatemalan Truth
Commission in the difficult study and forensic analysis of war crimes and modern massacres. For a year, it studied this singular termination ritual at Cancuen, in their terms “an ancient war crime.” The forensic team did an osteological study of the sacred cistern at the southern palace and its thirty-one skeletons (identified as the minimal number of individuals); other Cancuen subprojects also did complete artifactual studies and osteo-logical and isotope studies continue (e.g. Quintanilla 2013; Winburn et al. 2014) (figure 7.7).
Some conclusions of the FAFG team were clear, consistent, and specific and provide clear evidence about this singular event (Suasnavar et al. 2007; Quintanilla 2013; Quintanilla and Demarest 2013). The individuals in the pool were in good health; they had strong, well-preserved bones and exhibited little pathology. Their elite status was marked by fine artifacts (most were probably elements of dress and adornment), by the most sacred location of their watery interment, and by aspects of their physical remains. Analyses of many more conventional burials (110 of 118 studied to date [Quintanilla 2013]) in Cancuen have identified the presence of paraphernalia made of shell and conch and finely made obsidian and chert artifacts (p.180) as indicators of elite status (e.g., Andrieu and Quiñonez 2010; Andrieu et al. 2011a, 2011b; Quintanilla 2013). Furthermore, the individuals deposited here had various forms of dental alteration and decoration and cranial deformation consistent with a more elite status (Whittington and Reed 1997; Harrison-Buck et al. 2007; Tiesler and Romano 2008, 24; Quintanilla 2013). At Cancuen, the ceramics, artifacts, and architecture, all indicate a multiregional population, and preliminary results and isotopic analyses (Winburn et al. 2014) of the victims in the cistern confirm that they were members of the politically and economically dominant Central Petén elites, as the associated artifacts suggest.
The range of individuals was that of a living population, not one with the age profile of natural deaths. Context indicated the deposition of all of the bodies at the same time. Individuals included adults of both sexes, adolescents, children, infants, and remains in a good state of preservation of two fetuses. One FAFG study sampled bones from superior and inferior extremities, ribs, and vertebrae from the southern cistern that bore evidence consistent with blunt-force trauma and sharp trauma, such as stabbing, slashing, and puncturing (e.g., Suasnavar et al. 2007; Quintanilla and Demarest 2013). Of course, the cause of death could not be precisely specified for much of the sample studied if osteological evidence was not present (wounds are usually made to soft tissue). It is clear, however, that for much of the studied sample, the cause of death does not appear to be natural or illness and that some bones had clear evidence of perimortem trauma that was apparently inflicted by different types of weapons (e.g., Quintanilla 2013; Suasnavar et al. 2007). Bone trauma and scars on osteological samples were generally consistent with modern evidence the FAFG has recovered worldwide of execution with bayonets, clubs, or machetes (Suasnavar et al. 2007). In the pre-Hispanic period, this would have indicated wounds by spears, clubs, and axes; this is supported by the artifacts present. One individual was decapitated.
Note that the evidence is not that of an execution of adult male warriors, as at Dos Pilas and other contemporary sites (e.g., Wright 1990). Instead, it is a cross-section of the population; eight were identified as male, six as female, eleven as undetermined adults, and six as subadults. Given the presence of the two fetuses, it is probable that two of the women were pregnant.
Where the killing of these individuals took place is uncertain: at the cistern? in the adjacent palace? at various points in the site from which their remains might have been carried? Around the east port of the epicenter’s peninsula, about 300 meters north, was a probable battle scene; two adult (p.181) male skeletons were left unburied there, adjacent to a defensive wall around the east port and the subroyal elite palace group there (e.g., Alvarado et al. 2006; Berryman and Novotny 2004). Deposits of skeletal material and artifacts were discovered in the east port and of over twenty individuals in another ritual cistern that lies between that east port and the royal palace.
In addition to ceramics and context, the date of the event at AD 800 can be estimated by the last monument dates at Cancuen and by other contemporary textual references to the site. One inscription on a shell dated to AD 810 in a private collection refers to the last king, Kan Maax, but in a reference to the historical past. Recovered ceramic offerings can be securely placed in AD 790–800, in the Late Chaman phase.
One remarkable factor in this mass assassination of Cancuen nobility was that their precious elements of costume were interred in the sacred waters with them. Objects recovered in the southern cistern may have been personal adornments rather than deposited offerings, except for a capping sample of broken vessels that included polychromes—a probable termination deposit. The objects in the cistern included carved artifacts of jade, mother of pearl, and (especially) shell. Similarly, the shallow burial of the king and probable queen also held precious grave goods (see below). Strange as it seems to western thinking, it might have been a respectful, highly ceremonial, mass killing: a religious termination of the dynasty. Nonetheless, it was a thorough termination in terms of people, monuments, and architecture. Consistent with the demise of divine kingship discussed throughout this volume, this event was a religious and political act, not merely a war massacre. Again, the nature of the deposits contrasts with evidence of captive sacrifice or massacres elsewhere (e.g., the deposits at Colha [Mock 1998b; Buttles and Valdez this volume] and Dos Pilas [Wright 1990]).
Nearby, in the royal palace, in an entrance buried by massive earth and rubble fill (possibly a termination deposit), the last ruler, Kan Maax, and his consort apparently were also victims of the mass assassination of royals and nobles. They were hastily buried under only sixty centimeters of earth and without a tomb or even a haphazard cist (Barrientos, Arriaza, et al. 2006, Quintanilla 2013). The context and date indicates that the rulers were assassinated, although poor bone preservation in this context (unlike the context of the mud-filled cistern) made clear forensic confirmation impossible. Despite their apparent assassination and rapid burial in this mud fill, the royal pair also were laid to rest with great respect. Both burials were accompanied by fine imported vessels. The ruler had an engraved (p.182) mother-of-pearl necklace that identified him as “Kan Maax, Holy Lord of Cancuen, Holy Lord of Machaquila” (Barrientos, Arriaza, et al. 2006). His elaborate headdress of cloth, feathers, and shell was stained into the mud above and at the sides of the skeleton, running from his head down to his lower legs. In the headdress, mother-of-pearl and carved coral fishes and water lilies matched the aquatic-themed headdresses shown in the carved monuments and stucco sculptures of the Cancuen rulers.
Again, the contradiction to modern sensibilities is striking: a mass assassination, but with respect and awe given the dead. In that paradox we believe lies some of the evidence concerning the perpetrators of the assassination.
The palace evidence could be consistent with the architectural termination rituals described throughout this volume (see Iannone 2005; Schwake and Iannone this volume). At Cancuen, deposits in the two palace entrances were so thick and massive that they buried parts of entire entrance structures. We had previously suggested an alternative hypothesis that it might have been the beginning of a new stage of reconstruction of the palace (e.g., Barrientos 2014). In the light of subsequent comparative study, these are more likely to have been termination ritual deposits sealing the structures most associated with the ruler, as described at other sites in the chapters in this volume. The rubble and dirt covered both major entrances to the palace, the spectacular monumental royal entrance in the southwest, and the more accessible open entrance to the palace in the east. The major royal entrance in the west, Structure L7-9, was covered with many tons of earth, clay, and loose rock, sealing and entombing the entrance under several meters of deposits. Its doorways over three meters high and its entire wide and high monumental stairway entrance were buried. A similar massive deposit of clay and rubble buried the eastern entrance structure, L7-27. It was in the fill over this second entrance that we found the shallow burials of the last king and his probable consort.
At the royal ball court specifically dedicated to the last two kings, Taj Chan Ahk and Kan Maax, there was more evidence of termination similar to those of the crises of legitimation at other sites. There, four monuments, three altars, and one finely carved panel were carefully defaced to remove just the faces of the rulers a yajaw, a sajal, and an ajk’uhuun. Similarly, the site’s stelae were carefully defaced. This treatment is diagnostic of termination ritual treatment of monuments throughout the lowlands (e.g., Harrison-Buck this volume).
Elsewhere at the site, all evidence points to a rapid termination. There (p.183) were defensive walls and unfinished defensive walls in and around the palace and around the east port. These were identical to the many kilometers of such walls that have been intensively investigated for two decades at the sites of the Petexbatun region (e.g., Inomata 1997, 2008c; O’Mansky and Dunning 2004; Demarest 2004a, 2006). The unburied male skeletons near one of these walls, the great cistern deposits, the probable termination rituals at the palace and elsewhere in the site, and, above all, the placement of decaying bodies in water sources all suggest a rapid end and a rapid abandonment. There was no occupation at this site after the Late Chaman phase (AD 790–800), so it was not conquered and absorbed into a new kingdom. It was ritually terminated and abandoned. The only subsequent occupation in the epicenter peninsula of Cancuen consists of one Terminal Classic residence and a few handfuls of scattered Postclassic sherds (not unlike Minanha; see Schwake and Iannone this volume). In the greater site area, no Postclassic material has been found to date.
Discussion and Speculations: Causes and Culprits?
Many questions surround the extraordinary events that took place at Cancuen. Above all, why was it terminated and why was it not reoccupied? Why were the nobles and rulers assassinated and yet treated with such respect and ritual and interred with precious artifacts? And who carried out those actions and why? These are questions that move us into the realm of speculation and alternative hypotheses.
It is not surprising that Cancuen was a target for military conquest. Its strategic location at the interface between highland and lowland transport routes and at the head of navigation of the great Pasión-Usumacinta exchange route through the lowlands and the “royal road” route to Calakmul and the north was highly desirable (Canuto and Barrientos Q. 2013; Freidel et al. 2007). Cancuen would thus have been a target for conquest and acquisition, not for destruction and abandonment. Furthermore, rival lowland kingdoms, such as the suspect Machaquila, would have celebrated the conquest of a center such as Cancuen in their many stelae that have been dated to the following forty years (AD 800–840). Yet none report such a conquest there, or at Seibal, or anywhere. Thus, suspicions turn to the highland and piedmont communities that surrounded Cancuen, lined its trade routes, and constituted its closest trading partners.
While the end of Cancuen remains an enigma for further investigation, the answers to its particular legitimation crisis and mysterious end surely (p.184) relate to its role as an international exchange center. By AD 800, Cancuen had undergone many transformations. These included intense exchange with distant regions to the west from which ceramics, including Chablekal Fine Grey and Campamento Fine Orange, were imported from Tabasco and Veracruz and Zaragoza obsidian was imported from Puebla. Meanwhile, the control of Machaquila opened up routes to the north and east while avoiding the maelstrom of warfare in the Middle Pasión region (Demarest 2004b; O’Mansky and Dunning 2004).
Most significantly, exchange with the highlands to the south was critical to the florescence of Cancuen in the late eighth century. From that region (or through its valleys) came the jade, shell, quetzal plumage, pyrite, and other materials that were critical to the tribute, ritual, and patronage networks of the lowlands (e.g., McAnany 2013) and central to the role of Cancuen as a nexus of their exchange. Furthermore, more recent evidence indicates intensive trade with the adjacent highlands in commodities, including quantities of obsidian that were unprecedented in the lowlands and, in all probability, cacao, cotton, annatto, and vanilla from the transversal ten to fifteen kilometers south, an area identified by recent studies as a major zone of production of cacao, cotton, and condiments in the Contact and Colonial periods (e.g., Caso and Aliphat Fernández 2006, 2012; Akkeren 2012). Cancuen’s connections to the highlands are emphasized by its highland-style ball court and other features in the center of the site. It would appear that Cancuen’s rulers and elites were providing feasting rituals for their highland neighbors and trading partners at this ball court. Additional evidence of these ties is found in highland ceramics near the ball court and even in royal burials. The burial of even the probable consort of the last ruler had highland ceramics, hinting of a possible marriage alliance (Barrientos, Arriaza, et al. 2006).
The nature of Cancuen’s “ideological solicitation” of highlanders through many ritual devices (Demarest 2013) contrasts sharply with the situation in the highlands themselves. Despite the evidence at Cancuen of massive exchange of both exotics and commodities in the Late Classic, the period of that center’s apogee, Petén influence of any kind is virtually absent from the Alta Verapaz highland trade routes in this period. Both the areas that were the sources of precious ritual goods and the valleys along which they were transported lack lowland imports. The material culture of Verapaz, in all respects, from ceramics to architecture, is distinctly highland in nature. Thus, after the seventh century founding of Cancuen, there seems to have been a radical reassertion of identity by the highland and piedmont neighbors, (p.185) who not only controlled the trade routes but were also the population surrounding Cancuen and probably served as its agricultural support and labor.
Thus, the situation of Cancuen might be broadly similar to that posited in Rathje’s (1971, 1973) hypotheses regarding the relations of the highland and periphery with the lowlands, but in this case at a subregional level and with both sacred highland goods (see Blanton and Feinman 1984) and transversal commodities. In the Late Classic, as Cancuen reached the height of its splendor and its international role in exchange, the adjacent piedmont and highlands may have asserted their autonomy. The “periphery” became independent, if not controlling, as in Rathje’s model. The sites and assemblages of the Verapaz reflect this reinforcement of independence and identity. Meanwhile, Cancuen was even more invested in mechanisms of creation and support of alliance with the highlands. For highland exchange relations it seems to have relied heavily, perhaps almost exclusively, on the ideological mechanisms of legitimation found in its architecture, monuments, and artifacts (Demarest 2013).
In the late eighth century, the many factors that weakened the k’uhul ajaw may have led highland populations to lose faith in the Cancuen divine kingship. Even during its apogee, Cancuen confronted such challenges. Together, highland autonomy and the internal division of power would have weakened the position of the Cancuen rulers. As at many of the other sites described in this volume, the stage was set for a legitimation crisis (Iannone this volume). Meanwhile, downriver in the middle Pasión, the route to Cancuen was torn by endemic warfare and collapse, striking numerous centers and smaller communities. Furthermore, Cancuen may have over-extended its network of relations, which included ties to the more vigorous economies of Tabasco and Veracruz. Like later Terminal Classic polities in the northern lowlands, those zones had more well-developed interregional exchange systems that included the long-distance movement of commodities such as cotton, textiles, cacao, and salt. In contrast, the Cancuen region relied entirely on the adjacent highlands and piedmont for such products.
By the end of the eight century, Cancuen’s reliance on ritual mechanisms for power may not have been enough. Indeed, maintaining the expense of such legitimation may have become increasingly difficult. As economic and military crises spread through the lowlands, the wealth from exchange that had funded such monumental display and the feting of highland elites at Cancuen may have been greatly reduced. Perhaps the highland neighbors, once awed by Cancuen’s majesty, began to lose faith in the k’uhul ajaws (p.186) of that revered center. The center may have even come to be regarded as parasitic because of its need for external subsistence support and massive labor for its construction projects. All of these factors might have led to its ritualized destruction.
But the question remains as to why the termination of Cancuen was so respectful. Why were the ruler and elites interred after assassination with rich costumes and grave goods? Several answers are possible. One is that regardless of disappointment or conflicts, the rulers and nobles of Cancuen might still have been regarded as holy lords and the kings perhaps as divine or semi-divine. Thus, in death they served as holy ritual scapegoats (Iannone this volume; Taylor 2010; Taylor this volume). This may be why the aspects of the built environment most associated with the divine king were the targets for ritual termination: palace entrances, audience structures, and sculpted monuments that displayed the royal personage. Yet even in death the holy lords would need to be respected and propitiated.
An additional reason for the elaborate termination might be that the highland caciques and lowland nobles and kings may have been allied via marriage and thus were relatives. The presence of highland residents at Cancuen was notable in many aspects of material culture and burials. Thus, these two worlds of the Classic Maya in the southwest, the highlands and the lowlands, may have had close ritual and kinship ties that were honored even in the destruction of the site.
Despite such connections, however, the end came to Cancuen rather abruptly. A defensive wall was left half-finished. Adult male bodies were left on the causeway entering the site. The deposition of probable nobles in the sacred cisterns and elsewhere, however respectful and religious it would have been, would have left the site polluted and uninhabitable. However, chert points were recovered amid the bodies of the slain nobles made from chert exogenous to the site yet still from the general region (Andrieu and Quiñonez 2010), again casting suspicion on Cancuen’s highland and transversal piedmont neighbors.
Thus, the close allies and neighbors of Cancuen may have brought the center to its end. All activity ceased at its bustling ports, busy palace, and three ball courts. The sweeping processes of the ending of Classic Maya civilization that were beginning or under way elsewhere may have brought this kingdom and its divine dynasty to an early and dramatic end.