The distribution of urban settlement, as far as our sample permits us to judge, is generally concentric with respect to distance to the site center and does not cluster appreciably around elite buildings. Occupations are assigned primarily on the basis of surface collections. We placed from one to four 2x2-meter squares on all structures surveyed since 1992. These squares were then intensively checked for surface artifacts. Although surface collections in general match subsurface deposits fairly well in cases where both are available, surface collection totals may be biased in favor of later phases. Since, however. Late Classic Yumcab types form the vast majority of sherds, that factor can be discounted with respect to the topic under consideration.
The impression of settlement density must be tempered by the fact that many structures yielded few or no sherds, even when as a last resort we conducted grab samples. This suggests a significant number were not residences, but special-purpose structures or outbuildings of one sort or another. Even where ceramics are present, assignment of occupations is problematic, primarily because of erosion and consequent small sample sizes. For the purposes of this paper, we rely upon a subsample of 306 structures beyond the site nucleus that had a total of ten or more identified sherds. Structures were chosen from all of the survey quadrats except for the central GT. We experimented with two definitions of an occupation, one in which at least 5 percent of all identified sherds came from a given phase, and another where a 10 percent cutoff was used.
By either criterion, occupation was overwhelmingly Late Terminal Classic Yumcab, with more than 90 percent of the sample structures being occupied, even at the higher cutoff level. Occupation was also uniformly distributed, with only squares HU and FU reporting lesser Yumcab ceramic percentages because of high Early Classic occupancy. Most of the identifiable domestic architecture is also from this period.
Postclassic occupation estimates vary from 19 to 28 percent of the sample, depending upon the cutoff used, but the majority of these reflect very light artifact deposition. Overall, Postclassic ceramics formed less than 6 percent of the total from Ek Balam, or less than one-tenth of the number of Yumcab-phase sherds. Postclassic (Xtabay-phase) buildings are clustered in the squares adjacent to a sixteenth-century ramada chapel (quadrats HT, HS, IT, and probably IS), although none were found in HU to its south. Limited architectural evidence also supports this clustering. A Postclassic mound type identified by Hanson consists of a single line of stones delimiting a low terrace along the west edge of a mound; all examples of this type cluster in the above quadrats. The only sign of Postclassic architectural activity within the enclosure walls is a small one-room east coast-style “temple” atop the Early Yumcab platform GT-10 and possibly small one-room shrines on earlier pyramids. Surface offerings of Chen Mul incensario fragments were also noted on several of the major Late Classic structures. Because identifying the Postclassic occupation was a specific concern of the project, and we therefore made an effort to weight coverage to the east side of the site where Postclassic settlement was known, the conclusion seems inescapable that there was a major loss of population in the urban zone.*
Rural survey results indicate that forty-three loci (73 percent) possessed Yumcab ceramics, while only twenty-three (39 percent) had Xtabay types, but this again obscures the true disparity between the two. A total of 11,811 Yumcab-phase sherds weighing 90.99 kilograms were recovered from rural sites compared with just 702 (or 7.68 kilograms) from the following Xtabay phase. This ratio, better than 10:1, was present at all of the major second - and third-tier sites with the exception of X-Huyub, which apparently was temporarily abandoned during the Yumcab phase. Sotuta and Hocaba ceramics were almost nonexistent in the rural zone.
Turning to the intersite survey, about 75 percent of structures at Ichmul de Morley had some Sotuta ceramics, but only 15 percent had enough to suggest occupation. Cehpech-sphere ceramics were evenly and heavily distributed across the site, while Sotuta ceramics were widely scattered but at much lower frequencies. In all cases, Yumcab sherds were many times the number of Sotuta sherds. Interestingly, Sotuta sherds were not well represented in our site-center sample but were present in low numbers on a number of larger platforms of clearly elite status in the outlying settlement sample. The minor role of Sotuta ceramics must be tempered by two observations. First, several examples of Cehpech types with Sotuta forms and modes were noted, including a number of molcajete fragments made with Muna Slate ware paste and slip. Second, our limited obsidian collection from Ichmul suggests it enjoyed much greater access to the green obsidian distributed by Chichen than did Ek Balam.
The Postclassic decline at Ichmul was severe: only 122 (0.75 percent) Xtabay-phase sherds were collected from the entire site. These were scattered across the site in numbers too low to indicate occupations, except for one very small sample whose high Postclassic frequency can be attributed to sampling error. Only a single Postclassic sherd was collected from the site center. Like the rural survey sample, there were no Hocaba-related ceramics.
There can be little doubt from our data that the Postclassic decline of the Ek Balam region was severe and involved more than just the cessation of monumental construction or the redistribution of population. The center of Ek Balam was abandoned, with only the occasional visitor stopping to leave modest offerings of effigy censers among the decaying grandeur of the main plaza. Outside the walls, some minor building additions were made, as for instance to Structure GS-12, but most of our occupation evidence comes from the area east of the site center in squares HS, HT, and IS. Rural communities and those between Chichen and Ek Balam alike were abandoned. Whether the little evidence we have for Postclassic occupation was continuous or not remains unclear.
The two centuries between a. d. 1000 and 1200 remain an enigmatic gap in the northern Maya archaeological record, with little supporting evidence of how this transition was effected. As noted above, in areas not under Itza dominance the major Classic-Postclassic ceramic transition was between the Cehpech and
Tases complexes, with no clear-cut Early Postclassic complex intervening (Bey et al. 1998; Ringle, Gallareta, and Bey 1998). This pattern is evident at both Ek Balam and at Coba, although in each case Postclassic settlement was clearly substantially reduced. Beginning around a. d. 1000-1050, sites with a strong Sotuta presence were marked by a gradually increasing admixture of Hocaba types (Brainerd’s “coarse slateware,” especially Black-on-cream) followed by a replacement of Sotuta by Tases types, but still accompanied by residual Hocaba production. The earliest Mayapan deposits, dating to around a. d. 1200, belong to this latter transition, while those of late Chichen, Zipche II at Dzibilchaltun, late Uxmal, and elsewhere, to the first.
At Ichmul de Morley and Yula (Anderson 1998; personal communication 1999), the absence of Hocaba and Tases sherds indicates these towns probably were largely abandoned by a. d. 1000 and that any remaining inhabitants no longer participated in the Itza production-distribution network. Hocaba ceramics are also absent from nearby Yaxuna (Brainerd 1958; Suhler et al. 1998), despite some evidence of intrusive Sotuta deposits. Although Hocaba ceramics continued to be produced in quantity at Chichen and in the Chikinchel region (Kepecs 1998), at the former these were in apparently post-monumental contexts. The absence of Hocaba ceramics at nearby sites such as Ichmul, Yaxuna, and Yula further supports the contention that the retrenchment of the Itza production-distribution sphere occurred at the end of the Terminal Classic, and probably did not appreciably outlast the florescence of its neighbors.
Late Classic Ek Balam
At Ek Balam, the Late Classic demographic peak was accompanied by increased investment in monumental architecture. The buildings around the main plaza are some of the most impressive in northern Yucatan, although our work and the work of I. N.A. H. demonstrate that elite architecture was being erected from at least the Early Classic period onward. Abundant Formative material from test pits further belies the story that Lord Ek Balam was somehow responsible for the founding of the city, although the story may reflect the advent of a new dynastic line during the Late Classic.
The monumental core of Ek Balam consists of three enormous platforms (GT-1, 2, 3) delimiting three sides of the main plaza. The southern end is enclosed by another large platform (GT-10) and two “temple assemblages,” an architectural grouping consisting of a pyramid at right angles to a long, multiroom or hall-like structure (Strs. GT-15, 16 and GT-19, 20).’ A number of lesser structures are found within and without this complex, including a ballcourt (GT-8) and several shrines and altars, for lack of better terms. The whole is surrounded by double enclosure walls, apparently first identified by Eric von Euw in the 1970s. Such walls are common at a number of northern sites, such as Cuca, Chacchob, Muna, Uxmal, and others (Kurjack and Andrews 1976; Webster 1978; Kurjack and Garza T. 1981: 303-304) and are discussed below. A series of long narrow platforms between the buildings around the main plaza create a third, innermost enclosure wall. The “civic” armature of the site is completed by five sacbeob that radiate from the monumental core. Four of these run generally in the cardinal directions, forming a cruciform pattern suggestive of Landa’s description of Postclassic towns and Uayeb rituals (Bey and Ringle 1989; Ringle and Bey 1992). These causeways generally terminate in smaller complexes of elite architecture, frequently consisting of temple assemblages.
Temple assemblages appear to be one key to understanding Late Classic settlement over much of the northern plains, as they are farther afield (Fox 1989). Temple assemblages form the nuclei of smaller sites, where they are the principal monumental architecture, and of elite clusters at larger sites such as Ek Balam and Chichen Itza. We have suggested this may reflect a segmentary form of organization, perhaps of important noble families and their retainers, with the longer building perhaps being identifiable as a popol ua.'°This pattern may have its roots in the Late Formative period, at sites such as Mirador (Andrews IV and Andrews V 1980: 21-41; Ringle 1999), but by far the majority date to the Late Terminal Classic period. At larger sites of this period, several such units were often integrated into a single whole, resembling quite closely the cuchcabal form of political organization described for the Late Postclassic in ethnohistorical sources.
At least three elite architectural styles span the Late Terminal Classic apogee of Ek Balam. The latest is a simplified version of the Pure Florescent style, which clearly had an enormous influence on local architects. Structure GT-2, for instance, is a sixteen-meter-high basal platform supporting a classic Puuc-range structure along its rear (west) edge, reminiscent of the House of the Deer at Chichen or the House of the Governors at Uxmal. Unlike Puuc structures farther west, however, mosaic facade sculpture is much reduced for the site as a whole. Although we have found a few “Chac” noses, corresponding facial sections of such masks are unknown. (The few masks we have found are carved from single blocks of stone.) Upper wall zones were instead plain or decorated with simple geometric motifs, such as mat designs. Typical Puuc ornaments such as banded colonettes, drum altars, and three-part moldings are also not uncommon. Structure GT-8, which we had identified as a ballcourt on the basis of limited trenching, proved to be a Pure Florescent add-on to an earlier structure when further excavated by Vargas. A fragmentary ballcourt ring bearing only “9 baktun” indicates the style arrived relatively early at Ek Balam compared to other dated examples.
Much of the architecture of the site center structures predates the Pure Florescent style. Exposed facades of GT-8, GT-15, and GT-1, among others, show the use of less finely cut and more irregularly coursed stone covered with thick coats of stucco, sometimes modeled. This masonry closely resembles some of the structures from Coba. Also typical is the use of stucco-covered tableros in the upper wall zone, sometimes in association with stone armatures that once supported stucco sculptures (e. g., GT-15). Finally, several platforms, including GT-8 and GT-1, bear an unusual facade of multiple aprons. Vault stones are often well dressed but are quite long in profile, acting as integral parts of the wall rather than as decorative veneer. Recent discovery of elaborately modeled threedimensional stucco sculpture on the upper facades of GT-1 are without precedent in the north and suggest prototypes to the south or possibly southeast.
Structures GT-10, GT-16, and GT-20, in contrast, are even earlier, but still within the Yumcab complex. Megalithic blocks were used for the retaining walls and foundations of GT-10 and GT-16. GT-20 is distinguished by irregularly coursed walls covered with thick coats of modeled stucco, but most notably by vaults of thin, roughly fashioned slabs. Although perhaps several centuries old by the onset of the Pure Florescent period, ceramic deposits and architectural renovations indicate that these buildings continued to be used and remodeled throughout the Late Classic. For instance, excavation of a small shrine atop GT-20 demonstrated it was a Chenes-style addition. Somewhat later several of the doorways of the main hall were walled up. Whether this was due to a change in building function or for defensive purposes is unclear.
This eclecticism is reflected in the layout of the site center. Structures GT-1, GT-2, and GT-3 and the lesser pyramids GT-16 and GT-18 created an enclosed central space surrounded by massive architecture. This is unlike most northern site centers, where the emphasis is on horizontal, rather than vertical, masses, and where central spaces are often more accessible. The net result is a type of main plaza reminiscent of more southerly centers. On the other hand, the enclosure walls are wholly within the northern tradition.
Ek Balam’s distinctive blend of southern and western influences is also reflected in its sculpture. The main example of the former is Ek Balam Stela 1 (Figure 21.5b), a badly fractured monument apparently from the site center. As reconstructed by Graham, the monument depicts a standing figure holding a God K “manikin scepter.” Below the figure is a zone with a bound captive facing right; most probably a companion once faced left. Above the central figure is an “ancestor cartouche” similar to that of Yaxchilan Stela 4.
Several aspects of the stela are noteworthy. First, its Long Count date is a rarity in northern Yucatan. The only other surviving tenth baktun Long Count date in the north comes from the Temple of the Initial Series at Chichen about forty years later, though several “short count” dates from Chichen and other sites are contemporary. Although late, the iconography is quite traditional, harking back to motifs characteristic of Coba stelae 50-200 years earlier, such as the basal zone of captive figures and the long pendant necklace extending almost to the figure’s knees.' ‘ These motifs are ubiquitous at Coba, being found on Stelae 1, 2, 3, 5, 6, 20, and 21, and probably others were they not so eroded (Thompson, Pollock, and Chariot 1932; Graham and von Euw 1997).
2L5 a. Cobd Stela 20 (Graham and von Euw 1997); b. Ek Balam Stela 1 (preliminary drawing courtesy Ian Graham); c. Ek Balam Emblem Glyph from GT-1 balustrade (author drawing).
These similarities, however, are probably retentions from a general pre-Flo-rescent sculptural tradition across the northern plains. Similar necklaces may be seen at Oxkintok on lintels from 3C7 and 3C10 (Pollock 1980: 523a, 536) and on Uxmal Stelae 2 and 3, for instance. The Coba and Oxkintok versions differ in having a single shin-high crosspiece rather than the three present on the Ek Balam and Uxmal stelae. Ek Balam also differs from the Coba stelae in that the emblem of authority is the God K scepter, whereas nearly all figures at Cobd hold serpent bars. God K scepters are known from Edzna, Uxmal (St. 2 and 6), Dzibilchaltun (St. 9 and 19), Oxkintok, and Sayil (St. 5), among others. Finally, Ek Balam’s Stela 1 diverges from Coba in its emphasis on ancestors (the figures of the cartouche above the main figure and probably the two eroded figures on each side of the monument). No ancestor cartouches are present at Coba, nor were flanking figures carved on its stelae. Here more southerly prototypes are suggested.
Another sculpture in the more naturalistic Late Classic style is a reclining bench figure depicting either a captive or a bacab from the GT-20 shrine. Two ballcourt panels from Ichmul de Morley (Proskouriakoff 1950: Figure 82a, b) also show poses typical of ballplayer panels from the southern lowlands, especially those from “Site Q” and Yaxchilan, although the workmanship of the Ichmul panels is inferior. Fragments of modeled stucco from GT-1 and GT-20 are also excellent examples of this tradition, as are painted capstones recently reported by Vargas et al. (1999).
In contrast, other pieces reflect the cruder style of Puuc sculpture. In 1987 we found two seated figures that may originally have been affixed to a building facade (Figure 21.6). Similar patterns of facial tattooing are present on heads from the Puuc zone, including Maxcanu (Pollock 1980: Figure 576e), Kabah (Pollock 1980: Figure 377c, d), and the “Queen of Uxmal.” These may be from the same workshop as two warrior torsos found in the main plaza. One headless warrior carries several darts; the other is a captive with his hands bound behind his back. Less easily classified are several miscellaneous sculptures of composite “monsters” that probably also decorated architectural facades. While it is tempting to see a chronological progression from “naturalistic” to Puuc styles, Stela 1 and the stucco facades of GT-1, as well as the Puuc-like architecture of GT-7, are apparently the work of a single ruler, U Kit Kan Lek.
World History
