Architectural Style Zones

These regional architectural decorative and formal styles were among the first regional aspects of culture to be defined in studies of Maya civilization. In his extensive and extraordinary surveys from 1887 to 1914, Teobert Maler was very much aware of these distinctive building styles (cf. Maler 1895). A. M. Tozzer says in a 1913 report: “From a study of the Maya remains as a whole several distinct areas can be distinguished, in each of which there is a certain kind of unity as regards assemblage, construction, and method of decoration” (1913:149). Tozzer goes on to specify Northern Yucatan, Southern Yucatan (Chenes and Río Bec), the Peten, and the Usumacinta Valley.

Many studies since Tozzer's time lead to the present conclusion that the Puuc, Central Yucatan, and Greater Peten styles are distinct in space and, to some degree, in time. Peten-style material dealt with here is from both the Early and the Late Classic periods, a.d 250–850. The Peten cultural apogees run for about seven hundred years. The central Yucatan style dates from ca. a.d 550–830 and therefore is generally contemporary with the Peten Late Classic, 650–850 (cf. Potter 1977). The Puuc building style evolves perhaps as early as a.d 500, but seems not to reach its fully developed form until about a.d 650, and lasts until about a.d 910 (Dunning 1992:66–71). All of the styles seem to undergo late florescence at about the same time—that is, during the late eighth and ninth centuries—and last about two hundred years. Arthur Demarest (1992:141) drew the same conclusion some time ago in a similar context.

Site Sizes in the Maya Lowlands

Turning to site size in these regions, enough data of roughly equivalent quality have allowed scholars to rank-order sites (Dunning 1992; Graham 1992; Morley 1938–1939; Pollock 1980; Ruppert and Denison 1943). In 1981, the geographer Richard Jones and I published a study of the Puuc, Chenes-Río Bec, and Peten region sites in which we ranked the sites according to size. We used a method called courtyard counting, although strict volumetric estimate is probably slightly more accurate (Adams and Jones 1981; Turner et al. 1981). In reality, both are assessments of the total mass of formal major architecture at Maya sites, as noted in the lengthy discussion of the methodologies in the cited papers. We also developed a map of the clusters of ranked sites in which we used a nearest-neighbor statistic based on the average day's travel time in the lowland, 37 km per day (Figure 1; Adams 1978; Adams and Jones 1981:Map 2). The patterns that result are distributional clusters of sites around three large cities: around Calakmul and Tikal in the greater Peten, and around Uxmal in the Puuc Hills. No discernible clusters appear in the Río Bec and Chenes regions unless these are combined with the sites of the Calakmul region in the latter case and with Uxmal in the former instance. We then applied the geographers' rank–size rule to the clusters of sites, with the result that we were able to suggest that the Maya of these regions at least briefly had regional states. The authors of the present paper have taken these data and much new published information into account (e.g., Dunning 1992) and reanalyzed them, with the following results (Figures 2, 3, 4, and 5).

Map of the Maya Lowlands showing rank-order and rank-size elicited sites and possible regional states of the Classic period. After Adams and Jones 1981:Figure 1; redrawn by Bruce Moses, Center for Archaeological Research, University of Texas, San Antonio.

Tikal region sites including those of the Three Rivers Zone (n = 146), rank-ordered and displayed on a double log chart. The pattern is approximately log-normal.

Río Bec region sites (n = 33), rank-ordered and displayed on a double log chart. The pattern is pluralistic.

Calakmul region sites (n = 41), including the Río Bec sites, rank-ordered and displayed on a double log chart. The form approximates a primate pattern.

Puuc Hills region sites (n = 143), rank-ordered and displayed on a double log chart. The form approximates a primate pattern.

We might have included lists of rank-ordered sites with this study. However, this would involve publication of a table of 146 sites for the Tikal region alone and of large numbers of sites from the other regions. Rather than essentially republish the data detailed in Adams and Jones, we refer the reader to sources for rank-ordered sites in the southern lowlands (Adams 1999:191–207; Adams and Jones 1981:Table 1) and in the Calakmul Region (Adams and Jones 1981:Table 1); for partially rank-ordered sites in the Puuc Hills (Dunning 1992; Pollock 1980); and for rank-ordered sites in Central Yucatan (Río Bec and Chenes) zones (Adams and Jones 1981:Table 1). The Puuc Hills sites were ranked as a total region. Data missing from the cited references are the rank-ordering of some of Pollock's Puuc Hills sites, and of the Tikal-Uaxactun region sites plotted in Dennis Puleston (1983). These evaluations and the reassessments of the already published results certainly need to be made available for the use of the few specialists interested, and we propose to do this in Ixcanrio Research Reports, Nos. 1–2 (Adams and Valdez 2002), now in press. The methodology used to produce these studies is detailed in Adams and Jones (1981) and Turner et al. (1981).

Correlations

Robert Sharer (1994:496) has criticized the “size-hierarchy” method on two grounds. First, size does not always correlate with relative political power; and second, the method is insensitive to political change. Both reproaches are faulty in that they blame the method for not compensating for the incomplete quality of the data. Skewed, incomplete, and doubtful data are an occupational hazard of archaeology, and the remedy is to produce more high-quality information, not to scuttle a promising and proven analytical tool. Some compensation for data defects may be gained by developing independent size-hierarchy data from other sources. These ranked data would include epigraphy, chronological studies, ceramic analysis, which could then be compared and correlated.

Returning to the present study, there are three possibilities with logarithmic analysis of rank-size data. One is a log-normal display, which implies politically and economically well-integrated systems of long duration. The second pattern is a plural distribution, which implies competitive centers on several levels, with none large enough to achieve dominance. The third is the primate pattern, which implies a city exponentially larger than the cities or centers below it. Primate cities are often found in the early development of a centralized state. Calakmul and Uxmal are primate cities in their regions, implying regional states and ones that had achieved their status relatively recently. The Río Bec region displays a plural pattern, which indicates that it had no one dominant center and perhaps had undergone a slow but long development. There is, however, the possibility that Calakmul at one period dominated the Río Bec region (William Folan, personal communication, 1999). Tikal is the one center that displays a log-normal pattern and therefore implies a regional state of long standing. The few data for the Early Classic 2–3 period (ca. a.d 390–530) indicate that Tikal was then a primate center for its region, again indicating a relatively recent development of the regional state.

Once established, regional building forms in all parts of the lowlands varied relatively little from early to late in their evolution. The greater Peten zone is the one for which we have the best data. Temples vary mainly from Early to Late Classic by being larger in the late period. Palace structures are more massive, more labyrinthine, and more numerous in the Late Classic. As Gordon Willey has said (unpublished comment), there is a megalomaniac aspect to Late Classic Maya buildings.

Iconographic Zones

In the greater Peten there is an Early Classic emphasis on the Storm God at Río Azul and Cerros and on the Sun God in the Late Classic (Tikal, Kohunlich). The Greater Peten is the only area in which change is readily noticeable, probably due to inadequate data on the Early Classic in all other zones. It will be noted that the change is in terms of shifts of emphasis rather than replacement of patron deities.

Buildings in the central Yucatan region commonly use Earth Monster of Sky Serpent Doorways framed by Sky Serpent stacks. These motifs occur on both palace and temple facades. The so-called “false” temple towers actually contained tombs. The distinction between palace and temple is thus blurred in the Central Yucatan region. The temple towers occur on the corners and central zones of palaces—for example, in the famous Xpuhil structure and at Becan. In fact, at Becan there is only one free-standing temple of the kind so common to the south in the Peten zone. The Chicanna palace facade is especially ornate in its decoration. The central doorway is in the form of a large Sky Serpent mouth, complete with lower jaw and teeth. This central doorway is flanked by the usual stacks of serpent masks. Two flanking doorways are decorated as if they were the entrances to thatched roof houses. The facade was once ornamented with small, three-dimensional sculptures, perhaps both humans and monkeys. This building is better studied and preserved than other examples of Central Yucatan architecture, but it seems to be typical. It faces a towered palace building across the courtyard. There appears to be no shift in iconographic emphasis during the period, and there is no reflection of the contemporary militarism. Yet nearby Becan possesses one of the largest fortification systems in the Maya Lowlands. Although built in the Late Preclassic, it was refurbished in the Late Classic, and David Webster thinks that it was open in that period (as cited in Potter 1977:7). If so, then the rejuvenation of the defensive ditch was in response to a military threat. Given that the Puuc was undergoing a militaristic expansion at the time, it is possible that the threat to Becan came from the north. Calakmul also may have constituted a threat to these apparently independent city states.

In the Puuc Hills region, the emphasis on the Earth Monster, the Sky Serpent or Itzamna, and the rain gods (Chacs) in the Late Classic is well known. Nicholas Dunning (1992:142, following Kowalski 1987) has summarized the symbolism of the Puuc facades as follows:

Other Maya deities were also represented by long-nosed masks and “include Kukulcan or Quetzalcoatl (Venus as God of the Winds, manifest as a feathered serpent), and Itzamna, manifest as either the Earth Monster or Celestial Serpent” (Dunning 1992:142). Dunning (1992:150) also notes a shift in emphasis in the elite-dictated art of the late tenth century. “The monumental art of the Uxmal-Nohpat-Kabah triad and associated smaller sites … features martial subject matter. The art portrays various individuals in military regalia, scenes of warfare, or captive taking and display. This militaristic emphasis gradually supplants a regional tradition in which rulers were more often portrayed in more benign ceremonial pageantry.”

Ideology, Maya Rulership, and Regional Variation

Early Classic formal religion, worldview, and political ideology will be dealt with briefly. The following reconstruction is based mainly on data from a single site of the Maya Central Lowlands, Río Azul, and from a single period, Early Classic 2–3, dating absolutely at the site from a.d 390–530, or about 140 years. We are supplementing this reconstruction with information from the related sites of Uaxactun and Tikal.

The Importance of World Directions

East–west appear always to be the most important directions, according to Wendy Ashmore's studies (1991). Río Azul clearly reflects this significance in the layout of the total site and the orientation of nearly all of the 739 buildings to cardinal directions, as well as the strict orientation of the twenty-eight Early Classic tombs and burials in east–west directions, with three exceptions. All Early Classic temple structures faced west. The Early Classic tombs deserve more description and discussion.

A distinctive feature of the Early Classic formal burial chambers at Río Azul is that most of them have well-preserved hieroglyphic texts and other symbols painted on their walls (Adams 1999:91–99). Tomb 12 (Figure 6) has already received considerable attention in the literature because it possesses directional glyphs painted on the appropriate walls with associated celestial signs (Adams 1999:Figures 3-15, 3-16). Tomb 1 was decorated by murals, which included a hieroglyphic text that identified the occupant as a member of the ruling family of Tikal—that is, that the deceased was a son of the Tikal ruler “Stormy Sky” and his wife “Bird Claw” and that he was born in a.d 417. His ultimate ancestry is indicated as derived from the Sun God, whose visage appears above the coat of arms of his grandfather, Curl Nose. Other tombs carry the symbol of the Storm God (Chac) and small texts that apparently identify the deceased. As noted, twenty-five of twenty-eight Early Classic tombs at Río Azul were oriented east–west. Elite male burials at Río Azul were probably oriented east–west, and elite female burials were arranged on a north–south axis. It also appears that elite burials (including those looted before 1983) at the site were overwhelmingly of males. This is plausible, given the known Maya patrilineal kinship system. Ancestor veneration, if not worship, is also indicated by these data (Hall 1984, 1986, 1987, 1988).

Mural paintings from Río Azul Tomb 12 in which the cardinal directions are indicated by giant hieroglyphs on the four walls. Glyph drawings by David Stuart and layout by Grant D. Hall.

Uaxactun

A. L. Smith (1950:88–102, Tables 6, 7) lists six tombs in his most elaborate burial category, four of which contained males considered to have been district rulers (Valdes et al. 1999:112–113) and whose tombs were oriented east–west.

Tikal tombs have not been separately reported, but William Coe's masterly report (1990) on the North Acropolis gives complete data on the largely Early Classic tombs in that building complex. Of the twenty-five burials excavated in the North Acropolis (Group 5D-2), ten can be assigned to the Early Classic (Coe 1990:Table 156; times spans 9 to 5 inclusive). Most Tikal Early Classic North Acropolis burials are those of males. Eight of the ten elaborate burials are oriented north–south. Thus, we have a striking difference between the roughly contemporary burials at the capital city of Tikal and its subordinate city of Río Azul. We suggest that differences in direction preference during the Early Classic may have something to do with differences in the patron deities of the deceased. Here also is evidence of diversity of religious practice within a regional state. Even “Stormy Sky,” ruler of Tikal, was buried in a north–south tomb (Burial 48), whereas his son, Governor X of Río Azul, was buried in an east–west tomb (Río Azul Tomb 1).

No significant burial patterns are known from the Puuc Hills sites or from the Central Yucatan sites.

Turning to another aspect, as is well known, the ancient Maya conceived of the universe as layered into thirteen heavens and nine underworlds. Further, it appears that the underworlds were watery. For lack of space, we will deal with only one new piece of archaeological evidence here. The “Cosmos Jar” from Río Azul (Figure 7) is an Early Classic 2–3 polychrome water jar found in a courtyard of the B-56 building complex. This is a Small Palace (Class 2) residential complex (Adams 1990:25–27). The painted scene on the jar is about 95% complete. It dates from about a.d 425. The analysis of the scene has been done by Jane Jackson Adams, with advice from Jacinto Quirarte. The scene appears to represent not only the cardinal directions but also the layered universe. The core of the interpretation by Jackson Adams is that the scene probably represents activity in the afterlife. Both the celestial worlds and the underworlds are shown and are linked through the Cauac monsters. The vessel therefore appears to be a cosmic diagram in which various roles are assigned to sacred creatures. No overtly human figure can be seen, although humanity may be represented by one or more of the birds or by the composite animal. A tentative suggestion is that the birds and animals may represent totemic creatures associated with patri-clans.

The Río Azul “Cosmos” pot. Drawing by Leonel Alvarado.

To sum up, at the largest scale, major sites seem to be invariably oriented toward the cardinal directions throughout the entire Maya Lowlands, an area of 250,000 km². This appears to reflect a uniform belief in the importance of this principle, analogous to that in traditional Chinese culture, which emphasizes geomancy in the placement of buildings. At the regional scale, there appears to be considerable diversity in elite tomb orientation, but it should be pointed out that these are always placed in reference to cardinal directions. The specific orientations—male rulers placed north–south in Tikal's North Acropolis, for example—are perhaps linked to clans' patron deities and to ancestor veneration. Uaxactun and Río Azul male rulers were buried with an east–west orientation. Could this represent differences in dynastic patron deities between capital (Tikal) and second-level administrative cities (Uaxactun and Río Azul)? At the most specific and smallest scale—that of individual depictive artifacts or art—there is again diversity, but it revolves around the interpretation of specific principles. This level of variety is reflected in items such as the Río Azul “Cosmos Jar.”

Political Ideology

The Storm God appears as patron of the earliest ruling group of Río Azul (Adams and Robichaux 1992). Tikal in the Early Classic period has clear references to the Storm God (Stormy Sky) as well as to the Central Mexican rain gods in Early Classic stelae (esp. Stela 31) and ceramics (Burial 10 pottery). On later Tikal stelae, there appears to be an emphasis on the Sun God.

The later Puuc regional rulers, as noted earlier, associated themselves with the rain gods. However, it is also clear that toward the middle of the Late Classic, predatory birds (eagles, vultures, etc.) appear as motifs on some of the buildings at Uxmal. These have been interpreted as signs of central Mexican influence, together with the possible introduction of the idea of “sacred war.” In other words, the late capital of the Puuc regional state appears to have become more militaristic through the Late Classic. This is probably also the period of the construction of the defensive wall around Uxmal and perhaps of other fortification features.

There are two bodies of iconography that reflect war and violence at Río Azul. The first is a group of Early Classic 2–3 monuments (Column Altars 1, 2, and 3), which show male humans stripped nude and evidently dying, but identified by individualized hieroglyphs on their backs. Similar Early Classic “conquered figures” were discovered by LaPorte and his associates at Tikal in the “Lost World” group. Early Classic stelae of the Greater Peten including those at Río Azul more commonly show captive figures at or under the feet of the principal humans (Robichaux 2000).

Earlier Uxmal stelae seem to link rulers to the rain gods, as noted earlier. Late Classic stelae of Uxmal, together with art from other Puuc Hills sites, depict a variety of organized violence—for example, the murals at the small site of Mul Chic and door-jamb sculpture from Kabah (Pollock 1980:Figures 372–373).

The Epigraphic Political Models

Heinrich Berlin's (1958) identification of emblem glyphs in Maya texts led Proskouriakoff to the revolutionary theory of the historicity of Maya texts, as expressed in her papers of 1960, 1961, and later. Proskouriakoff was largely concerned with deciphering the historical texts on stelae and demonstrating that the texts dealt with real temporal rulers. In the 1970s, Joyce Marcus (1973, 1976) proposed that a pattern of spatial distribution of emblem glyphs derived from a text at Copan indicated that a group of four cities or polities were associated with the cardinal directions. Further, Marcus's work established the existence of a hierarchy of sites. However, Peter Mathews, together with other epigraphers, has chosen to regard all sites identified by emblem glyphs as evidence of independent and equivalent polities (cf. Mathews 1991). This interpretation required that they ignore all variation in affixes, the enormous differences in sizes of sites and their architectural mass and paved areas, as well as big differences in the number of carved stelae. Even in 1973 this seemed an unreasonable conclusion. Some time later, Simon Martin and Nikolai Grube (1994) concluded that the epigraphic evidence supported the idea of at least some regional states, a conclusion that had been reached by field archaeologists years before. Therefore, it is now thought by many epigraphers that the textual evidence indicates variation in polity size through time and space in the Maya Lowlands. This is congruent with field archaeology studies such as the rank-order and rank-size analyses cited earlier.

As shown in Table 1, political-unit size was considerably larger and of greater duration in the Greater Peten than in any other region. Clearly, as has been pointed out by Sharer (1991) and Marcus (1994), there were cycles—or, at least, a sequence of rises and dips—in the political and demographic history of the region. Richard Hansen's (2001) interpretation of massive amounts of data from the Mirador Basin indicate that this tradition of large-scale polities begins in the Middle and Late Preclassic periods. Architectural and iconographic styles form a unit across political boundaries in this case, with the contending states of Tikal and Calakmul sharing commonalities in these aspects. Kowalski's and Dunning's work on the Puuc Hills also indicates a series of state-level societies, but through the Classic period and building toward a very late (Terminal Classic) unified state under Uxmal (Dunning 2001).