4
Chunchucmil Chronology and Site Dynamics

Middle Preclassic: Xpim Phase (700–400 bce)

The Middle Preclassic in Yucatán spans from 1000 to 400 bce, though Chunchucmil lacks pre-Nabanche pottery that pertains to the early part of the Middle Preclassic. Middle Preclassic pottery has been found in the lowest levels of excavations in the central portion of the site, yet the Middle Preclassic occupation is limited. Middle Preclassic sherds (n = 1,238) make up less than 1 percent of the total sherds classified to type at Chunchucmil. The ceramics of the Xpim phase consist of the Joventud (n = 649 sherds), Chunhinta (n = 155 sherds), and Dzudzuquil (n = 413 sherds) groups, which for the most part show characteristics similar in form and surface finish (figure 4.1b, c, d) to the Preclassic ceramics of the Early and Late Nabanche horizon (E. W. Andrews V 1989; E. W. Andrews V and Bey 2011; Bey 2006; Gallareta Negrón and Ringle 2004). Of these groups the Joventud Group presents some differences in slip and paste with respect to the Early Nabanche at Komchen (Ceballos Gallareta and Robles Castellanos 2012). Compared to the Nolo variety of Joventud Red (Andrews et al. 1989), Chunchucmil’s Joventud Group ceramics have a more fragile paste and a waxier slip that also adheres better to the vessel. This type of paste and slip also has been identified at Sihó and Oxkintok (Ancona Aragón 2014; Jiménez 2007; Varela Torrecilla 1998). At these two sites Middle Preclassic ceramics were also found in plaza fills that were remodeled as part of later constructions in the monumental part of the site (Jiménez 2007; Varela Torrecilla 1998).

Figure 4.1. Common ceramics types from Chunchucmil I: (a) Saban Unslipped: Unspecified; (b) Joventud Red: Unspecified; (c) Petjal Red on Black and Cream to Buff: Petjal; (d) Sierra Red: Unspecified; (e) Repasto Black on Red: Unspecified; (f) Sierra Red: Flaky; (g) Sierra Red: Unspecified; (h) Elote Striated-Impressed: Unspecified; (i) Oxil Unslipped; (j) Oxil Unslipped. (Photography by Pedro Tec Chim.)

Ceramics from at least two of the three Middle Preclassic ceramic groups come from excavations in 22 operations (figure 4.2). Middle Preclassic sherds make up 9 percent or more of the total classified sherds in seven of these 22 operations (table 4.2). These seven operations include two quadrangles (Ops. 26 and 28), a non-quadrangle monumental group (Op. 30), a medium-sized albarrada group (Op. 32), a large albarrada group (Op. 80), an isolated platform (Op. 3a), and Sacbe 1 (Op 7A). We provide details of some of these operations in the next paragraph. Figure 4.2 shows that these seven operations are clustered at the center of the site in an area of about 0.5 km². Figure 4.2 also shows that 20 of the 22 operations that yielded pottery from at least three Middle Preclassic groups cluster within a bit less than a square kilometer, also at the site center. Small amounts of Middle Preclassic sherds have been found scattered in other parts of the site.

Figure 4.2. Map highlighting groups with significant quantities of Middle Preclassic pottery. For the groups in black, at least 9% of the pottery recovered is Middle Preclassic. Groups in gray have much less Middle Preclassic pottery but have sherds from at least three Middle Preclassic ceramic groups.

Regarding details of the operations that yielded significant amounts of Middle Preclassic pottery, the isolated platform (Op. 3A/N1E1-H) exhibits megalithic construction, a form of architecture that the Maya of the northern lowlands began to build toward the end of the Late Preclassic (Mathews and Maldonado C. 2006). Op. 32/N1W2-E is a type 10 group centered on a large (2,000 m²) basal platform with a subsidiary patio containing smaller structures to the east. A test pit on the north side of the basal platform uncovered a buried structure with a plaster floor and non-megalithic masonry walls still standing 60 cm high. The fill beneath this plaster floor represents one of only two sealed Middle Preclassic contexts excavated at the site. The other sealed Middle Preclassic context comes from the architectural compound immediately to the northeast of Op. 32. This is the Xpim quadrangle (Op. 28/ N1W2-C/Xpim), located 500 m west of the site datum. Here, Middle Preclassic ceramics were recovered in a sealed context below two floors inside the patio. Since no architecture has been excavated at this group it is difficult to reconstruct its construction history. We know that a patio floor with a considerable amount of fill below it (50–100 cm deep) was built in the Middle Preclassic at this group, although we are uncertain of the architecture associated with it. Test pits at this quadrangle revealed that Middle Preclassic ceramics comprised nearly two-thirds of the classifiable ceramics. The rest came from the Late Preclassic and Early Classic periods. The Middle Preclassic sherds in the fill of Sacbe 1/Op. 7A (see also Sacbe 2/Op. 66) were mixed with Late Preclassic sherds and Classic-period sherds. It is important to note that all the groups that have Middle Preclassic ceramics also show later occupation. Perhaps further archaeological excavations could further define the Middle Preclassic occupation and the origins of the site beginnings.

Late Preclassic: Pich Phase (400 bce–250 ce)

The Late Preclassic occupation consists mostly of ceramics from the Saban (n = 1,489), Sierra (n = 499), Polvero (n = 414), Unto (n = 85), and Tipikal (n = 46) groups (figure 4.1a, e–g). The sherds in these ceramic groups total 2,679, which represent 1.8 percent of the total sherds from Chunchucmil classified to type. The groups Sierra, Polvero, Unto, and Tipikal continue into the Early Classic whereas a portion of the Xanaba Group pottery currently attributed to the Early Classic was likely made in the Late Preclassic. Seventy-three of the excavation operations have no Late Preclassic sherds and another 24 have only one or two sherds. Table 4.3 lists the 22 operations in which Late Preclassic ceramics comprise at least 4 percent of the total classifiable sherds. Figure 4.3 shows the location of these 22 operations on the map. Operations in which Late Preclassic sherds make up more than 10 percent of the total classifiable sherds are in dark gray, those with between 10 percent but more than 4 percent are in light gray.

Figure 4.3. Map highlighting groups with significant quantities of Late Preclassic pottery. For the groups in dark gray, over 10% of the pottery identified is Late Preclassic. For the groups in light gray, between 4% and 10% of the pottery identified is Late Preclassic.

Excluding Operation 59, the groups that have at least 4 percent Late Preclassic sherds cluster at the site core in an area of a little less than 1.5 km². There are both continuities and disjunctions between the Middle Preclassic and Late Preclassic occupation. For example, two operations with high percentages of Middle Preclassic pottery, Ops. 28 and 30, also have high percentages of Late Preclassic ceramics. Likewise, the pits in Sacbes 1 and 2 turned up appreciable amounts of both Middle and Late Preclassic pottery, although these are fill contexts that also contain later ceramics, suggesting that the sacbes were in fact built in the Classic period. In Ops. 3A and 26, 10 percent and 11 percent, respectively, of the identifiable pottery was Middle Preclassic but only about 1.5 percent dates to the Late Preclassic. This is interesting for Op. 3A, whose architecture is megalithic. Although megalithic structures in Yucatán are often built toward the end of the Late Preclassic, they can also be built in the Early Classic. Furthermore, some groups with more than 4 percent Late Preclassic pottery, such as Ops. 19, 20, and 24, had little or no Middle Preclassic pottery.

Of the 22 groups with more than 4 percent Late Preclassic ceramics, four are quadrangles. Thus, about a fourth of Chunchucmil’s quadrangles had Late Preclassic settlement, even though it does not appear that the architecture of the quadrangles visible on the surface was Late Preclassic.

In sum, the Late Preclassic occupation of the site is relatively minor, though at least one Preclassic sherd was recovered from almost half of the architectural contexts tested. We refrain from going as far as offering a population estimate for the Late Preclassic because many Late Preclassic ceramic groups continued into the Early Classic (e.g., a Sierra Red sherd does not necessarily mean a Late Preclassic date of manufacture) and because low quantities of Late Preclassic sherds (2 or 3 or 4) at a building may not indicate that the building was occupied during that period. No architecture definitively dated to the Late Preclassic has been excavated so far. Well-planned civic architecture, including sacbes, is known from other Preclassic sites across northern lowlands, such as Komchen (Ringle and Andrews 1990), Xocnaceh (Gallareta Negrón and Ringle 2004), Poxilá (Robles Castellanos in press), Yaxuná (Stanton and Freidel 2005), and Xtobó (Anderson et al. 2012), so it might have been present at Chunchucmil in the Preclassic.

Sierra Group sherds from the Pich phase show two kinds of surface finish: one in which the slip adheres very well to the surface (Unspecified variety) and one with a flaky slip (Flaky variety) that can be differentiated from the Joventud Group on the basis of its slip color, paste, and characteristic forms. It should be mentioned that these two groups (Sierra and Joventud) were defined by comparisons with ceramics from other settlements from the same phase, taking into consideration that many ceramic groups that pertain to the Late Preclassic period continue in use during the following epoch. At Chunchucmil, ceramics assigned to the Late Preclassic are sometimes associated with ceramics from later periods, such as Tixcacal Polychrome or the Xanaba and Shangurro groups, which delimit the beginning of the Early Classic ceramic traditions of the Yucatán Peninsula.

The concrete phenomenon of continuity in ceramic traditions from one epoch to the next is common in various Maya sites. At the same time, one must take into consideration that the fragmentation and erosion of a large portion of the Preclassic materials at Chunchucmil makes it nearly impossible to identify details of form, surface finish, and decoration on sherds that might have helped clarify details on the site’s ceramic typology. Finally we should not discount the possibility of recycling of ceramic materials from different periods that served as construction fill for the diverse buildings and architectural modifications at Chunchucmil.

Early Classic: Early Aak (250–400 ce) and Late Aak (400–630 ce) Phases

Some of the most common pottery groups found at Chunchucmil—Maxcanú (n = 7,872), Hunabchen (n = 11,668), Oxil (n = 62,249)—are part of what Carmen Varela Torrecilla (1998) has identified as the Oxkintok Regional Complex, which she dates to between 500/550 and 600/630 ce. Since Chunchucmil’s most abundant pottery is extremely similar to the pottery of Oxkintok, which is located 27 km east of Chunchucmil, we begin with a close discussion of the ceramic sequence of Oxkintok. This sequence was developed by members of the Spanish Archaeological Mission at Oxkintok, which worked at the site in the late 1980s. Yet members of this project disagree on some aspects of the sequence. These differences arise from discrepancies in how different investigators valued the variety of chronological data available: architectural phases, epigraphy, ceramics, and differences in the patterns of how buildings were used. These disagreements in the ceramic sequence manifest themselves in the epoch of transition between the Early Classic and Late Classic (500/550–630 ce), which Project Director Miguel Rivera Dorado (1992) calls the Noheb I Complex and which project ceramicist Carmen Varela Torrecilla (1998) calls the Middle Classic Oxkintok Regional Complex.

For Varela Torrecilla (1996, 1998), Oxkintok’s ceramic sequence can be correlated with architectural dates. In Varela Torrecilla’s view (1993, 1998), the Ichpá ceramic complex dates to the Early Classic and the Oxkintok Regional Complex pertains to an approximately 100-year-long period (500/550–630 ce) that is intermediate between the Early Classic and the Late Classic. Varela Torrecilla treats this as an independent ceramic complex with significant modifications in pottery making. This stage matches up with the Proto-Puuc A architectural style and a hiatus in the epigraphic record.

Rivera Dorado (1992, 1996) does believe that there is an intermediate period at Oxkintok, but that the Proto-Puuc architectural style associated with this period has strong connections with later periods. He also sees greater similarities between Oxkintok Regional ceramics and later ceramics: for this reason, he places the ceramic groups that form Varela Torrecilla’s Oxkintok Regional Complex into the Noheb I Complex (early Late Classic), which is followed by the Noheb II Complex.

García Campillo and Fernández (1995:136) propose that the Early Classic should be broken into two phases: Ichpá I (300–450 ce) and Ichpá II (450–600 ce). For them, the Oxkintok Regional Complex was a ritual and funerary subcomplex pertaining entirely to the Early Classic. Robles Castellanos and Andrews (2000:206–207) came to the same conclusion, arguing that it is not appropriate to use the term Middle Classic as a time period because of the lack of sufficient arguments and contexts that would corroborate the temporal isolation of the cultural characteristics of the Middle Classic.

Later, Varela Torrecilla and Braswell (2003) proposed that the term Oxkintok Regional Complex be used to refer to a conjuncture of ceramic groups with a manufacturing technology specific to western Yucatán. This ceramic repertoire has foreign characteristics that relate to other areas, such as Teotihuacan, the Petén of Guatemala, and Campeche, and can be dated between approximately 500/550 and 630 ce. Varela Torrecilla and Braswell 2003 proposed that during this period Oxkintok’s elites acquired political control of the surrounding region and, on ritual and funerary ceramics, imposed a style that contained local interpretations of foreign attributes. The characteristics that they considered foreign date to the Tzakol 3 ceramic horizon, which is the final part of the Early Classic, and persist into the following Late Classic period. The period from 500/550 to 630 ce at Oxkintok is now seen as a late moment of the Early Classic and presents local imitations of Teotihuacan ceramic forms and architecture, often filtered through southern Maya lowland versions of such foreign attributes (Ringle 2014:252).

Following these more recent interpretations, we place the Chunchucmil ceramics that fit the description of the Oxkintok Regional ceramics into the latter part of the Early Classic period. Contrary to Varela Torrecilla (1998), we do not call this the Middle Classic because we have not seen strong evidence of a Teotihuacan intervention, though there is certainly evidence that some people at Chunchucmil embraced pottery and architectural styles often associated with Teotihuacan (see chapter 11). We therefore split the Early Classic at Chunchucmil into two phases: Early Aak and Late Aak.

Like the pottery from the Preclassic phases, Early Aak ceramics at Chunchucmil are scarce. The site reached its apogee in the Late Aak phase, which we date to 400–630 ce. Late Aak pottery comprised a majority of the ceramic assemblages from all but one excavation operation (7a, a test pit in Sacbe 1). The excavation operations that are shaded gray in figure 3.3 represent locations with a strong Late Aak occupation. Put differently, the pottery from nearly every architectural group that we dug was dominated by Late Aak ceramics. Such a date range receives a boost from radiocarbon dates as well as architectural data. An animal bone in a sealed context in association with Oxkintok Regional sherds from the Lool residential compound (N2E2-N) yielded a radiocarbon date of 1550 ± 40 bp (calibrated at 2σ to 420–610 ce).

In terms of architecture, residential complexes in which Maxcanú, Hunabchen, and Oxil Group ceramics dominate invariably exhibit the Early Oxkintok/Proto-Puuc A architectural style, which George Andrews (1985) places between 550 and 650 ce. The dominant configuration of architecture during this time period is the plazuela, or patio group: a set of structures that surround one or more central patios. As we discuss in the next chapter, such patio groups often housed extended families. Most patio groups can also be called houselots because they are partially encircled by houselot walls and therefore fall into architectural group types 8 through 14 (chapter 3).

The Early Aak phase is marked by Usil Ware, which is characterized by a brittle paste with calcite temper. Two slipped groups pertaining to this ware, Xanaba (n = 1,878) and Shangurro (n = 244), are native to the northern lowlands. The somewhat frequent presence of thick-walled Xanaba cajetes (plates) and jars and Shangurro bichrome, composite silhouette cajetes with ring bases, suggests that Chunchucmil was well-integrated into the northern lowland interaction sphere. It is possible that in the later part of the Early Classic, Chunchucmil continued to use ceramics that were common across the northern lowlands. Xanaba and Shangurro pottery may have been replaced slowly during the latter part of the Early Classic with a later version of Usil Ware. Chuburna Group pottery, described below, is probably the later version of Usil that replaces the Xanaba Group ceramics.

Timucuy Polychrome (n = 879), which has been considered autochthonous to the northern lowlands (Brady et al. 1998; R. E. Smith 1971; Varela Torrecilla 1998:39), was also present at Chunchucmil during the Early Aak phase. Although Timucuy Polychrome has similarities to southern lowland ceramics, this northern lowland ceramic group is less lustrous and is distinct in terms of paste and slip. This pottery has a fragile paste with many carbonates and is generally very pale brown (10YR 8/3). The slips, which do not adhere well to the surface, range from yellowish red (5YR 8/4) to reddish yellow (7.5YR 7/8). The slip is applied over a calcareous wash of lighter color. The basic decoration of the Timucuy Group is geometric designs painted in red (2.5YR 5/8) and brown (5YR 5/8) over a red/orange background. Two forms of Timucuy cajetes have been recovered from Chunchucmil: those with tetrapod supports (Early Aak phase) and those with ring bases (Late Aak phase).

At Oxkintok, the cessation of Timucuy indicates a key marker of the end of the Ichpá ceramic complex of the Early Classic (300–500/550 ce; Varela Torrecilla 1998), precipitating a shift to monochrome and Oxkintok Regional Complex ceramics. At Chunchucmil we do not have concrete dates that indicate that Timucuy ceased to be used at the beginning or middle of the sixth century, (500/550 ce). Chunchucmil has small quantities of several southern lowland pottery types—Balanza Black (n = 277), Lucha Incised (n = 45), Paradero (n = 8), Pital Incised (n = 8), Dos Arroyos Orange Polychrome (n = 85), and San Clemente Gouged/Incised (n = 3) that date to the Tzakol 3 (400–600 ce) tradition.

Unslipped pottery from the Oxil Group, a Yucatán unslipped ware established by Varela Torrecilla (1998), is abundant at Chunchucmil (n = 62,249; figure 4.1i, j). We find Oxil Group ceramics commonly associated with Proto-Puuc A architecture and other Oxkintok Regional ceramic groups at Chunchucmil, not to mention a late Early Classic/early Late Classic radiocarbon date (previously discussed), suggesting that it dates to the Late Aak phase. Further below we note that some Oxil was likely produced in the early part of the Early Classic, with some being produced at the beginning of the Late Classic period. Oxil ceramics are coarse, characterized by a smoothed exterior and a slightly compact, red (2.5YR 5/6 6/6) paste with abundant fine gray and white calcite particles. Jar fragments have fine striations that occasionally combine with incisions resembling corncob impressions (Elote Striated: variety not specified; figure 4.1h) and, less frequently, incisions in the style of wavy lines (Elote Striated: Mulix variety). Corncob impressions on Oxil at Chunchucmil are less common than at Oxkintok. The typical form of Oxil jars during the Early Classic period is a composite silhouette with elongated neck and short, divergent, thickened rims. Some of the rims found at Chunchucmil have a lip with a shallow groove. In addition to jars, large basins (cazuelas) with ring (annular) bases and bolstered rims are also common. They exhibit carefully smoothed surfaces with fine, oblique striations (Yalchak Striated). Less common in the Oxil Group at Chunchucmil are incensarios with stucco decoration, smoothed and striated with pill-shaped appliqués (Pech Appliqué type).

New studies focusing on form and decoration have established that Oxil ceramics are not merely local to Oxkintok and Chunchucmil. They also have been found at sites in and around modern-day Mérida. Furthermore, at the excavations of the Cathedral of Mérida, Oxil has been found in association with Sierra Red and Chancenote Striated (Jiménez and Ceballos 2000). The rims of these early Oxil pots have a fragile paste, are much shorter and flatter, and often show impressions from the fingers of the potters. A few rim fragments of this kind found at Chunchucmil suggest that some Oxil pottery may date to the Early Aak phase. At Chunchucmil, the paste characteristics and surface treatment of Early Aak Oxil sherds (the ones with the short, flat rims) are basically identical to those of Late Aak Oxil–phase sherds. Therefore, when we assigned Oxil body sherds to the Late Aak phase, we often did so based on the other ceramics that the Oxil sherds were associated with as opposed to the physical characteristics of the pottery. Basically, the ceramic pastes of the Oxil group are homogeneous in color and texture. Nevertheless, a petrographic analysis might help to identify differences between Oxil pastes from different phases or to propose that these differences result from local production versus importation from elsewhere.

During the Late Aak phase at Chunchucmil new vessel forms identical to those of Oxkintok emerge. Examples of these forms include the ring-base Maxcanú cazuela with a “double bevel” rim (a projection/flange immediately below an exterior thickened rim), large pitchers, tripod plates with basal flanges and conical or slab supports, and very shallow cajetes with ring bases. This repertoire of forms manifests itself in the Maxcanú, Hunabchen, Kanachen and Aak Groups (figures 4.4f–i; 4.5b–e; 4.6b–e).

Figure 4.4. Common ceramics types from Chunchucmil II: (a) Dzununcan Striated: Dzununcan; (b) Ch'ikam Composite: Ch’ikam; (c) Tec Composite: Tec; (d) Tiznuk Composite: Tiznuk; (e) Dzununcan Striated: Dzununcan; (f, g) Hunabchen Orange: Unspecified; (h) Hunabchen Orange: Xcamal; (i) Maxcanú Buff: short neck. (Photography by Pedro Tec Chim.)

Maxcanú Group ceramics from the Early Classic have a medium texture, light red (10R 6/6, 7/6) paste covered by a slightly lustrous pink (7.5YR 8/4) or yellow-reddish (7.5YR 6/6) slip. Maxcanú slip colors get darker over time. Hunabchen Group ceramics have a medium texture, light red (2.YR 6/8) paste with gray calcite particles and a red (5YR 5/8) slip. Sherds from the Kanachen Group (n = 2,573) have pastes that are similar to those of the Hunabchen Group but have a slip whose color ranges from very dark grayish brown (2.5Y 3/1) to black (5Y 2.5/1), with abundant pale yellow (5Y 7/4) spots.

Two other common Late Aak–phase pottery groups at Chunchucmil are Chuburna (n = 3,503) and Aak (n = 1,768). Ceramics in the Aak pottery group were initially classified as Batres Red. Aak forms, such as those in the Ch’ikam type (figures 4.4b and 4.5a), are similar to those of the Aguila, Hunabchen, and Maxcanú Groups. Aak has a compact, yellowish red (5YR 5/8) paste similar to that of the Teabo Group and that of the Chaac variety of the Timucuy Group. Chuburna is represented by cajetes with ring bases, cazuelas with double bevels, and jars with short exterior thickened necks (Dzununcan type, figures 4.4a, c, e and 4.5g). Chuburna pastes have carbonates and the slip is smooth to the touch, exhibiting multicolored shades and abundant fire-clouding, which makes it difficult to know if the original color was yellow (2.5Y 8/4), red (2.5YR 5/8), yellowish brown (10YR 5/8), or black (10YR 8/1). Finally, the Late Aak phase at Chunchucmil also contains thin ware ceramics from the Chencoh (n = 2,673), Kochol (n = 2,323), and Acú (n = 325) Groups, which are part of the Oxkintok Regional Complex at the site of Oxkintok (figure 4.5f–i).

Figure 4.5. Common ceramics types from Chunchucmil III: (a) Ch’ikam Composite: Ch’ikam (annular base); (b–c) Kanachen Brown-Black: Unspecified (annular base); (d) Hunabchen Orange: Unspecified; (e) Hunabchen Orange: Unspecified; (f, h) Kochol Black; (g) Chuburna Brown: Chuburna; (i) Chencoh Thin Orange: Chencoh (Structure Kaab 32); (j) Dzibical Black on Orange: Dzibical (Structure Kaab 34); (k) Hunabchen Orange: Unspecified (Structure Kaab 32). (a– g photographed by Pedro Tec Chim; i–k photographed by Aline Magnoni.)

Figure 4.6. Common ceramics types from Chunchucmil IV: (a) Yaxuna Pre-pizarra: Unspecified; (b) Hunabchen Orange: Unspecified; (c) Maxcanú Buff: Gray Slip; (d) Chunchictok Striated: Unspecified; (e) Maxcanú Buff: Gray Slip. (Photography by Pedro Tec Chim.)

In summary, during the Early Classic, the pottery of Chunchucmil is very similar to Oxkintok pottery, as classified by Varela Torrecilla (1998), indicating that residents of both these sites interacted with people and products from the southern Maya lowlands and the Mérida area. Based on our findings, the Aak phase at Chunchucmil can be related in a general manner to the Cochuah horizon of the northern lowlands (Jiménez 2007, Varela Torrecilla 1998).

Chunchucmil reached its apex during the Late Aak phase. Quantifying this in terms of ceramics is difficult for a number of reasons, including that the body sherds from unslipped coarsewares are sometimes difficult to type correctly and the amount of preserved sherds that date to a particular period may say more about the durability of pottery from that period or the pottery economies of that period than the intensity of occupation during that period. Nevertheless, looking at slipped sherds alone, anywhere from about 70 percent to 80 percent (the variation depends on what portion of Maxcanú is placed in the Late Aak phase or the subsequent Kaab’ phase) of the pottery classified to the type level at Chunchucmil dates to the late part of the Early Classic period. The Late and Terminal Classic periods account for anywhere between 10 percent and 20 percent (depending on how the Maxcanú Group is split up) of the slipped pottery from the site.

Albarradas add an important detail to the chronology of the site. In the residential core (see chapter 5), where domestic settlement is most dense, albarrada walls act as relatively clear boundary walls that encircle domestic compounds. The albarradas form an interlinked and intelligible system of boundaries and circulation routes (callejuelas) through the site. The fact that these albarradas are intelligibly interlinked throughout the site suggests that they were once part of a living, contemporaneous system. The fact that many of the encircled groups were dominated by Late Aak–phase ceramics suggests that the system of albarradas dates to the Late Aak phase as well. To the extent that the albarradas are a signature feature of the site’s spatial organization, we again get the sense that Chunchucmil reached its fullest expression in the Late Aak phase. The fact that the albarradas exhibit a pattern of “filling in” (Hutson 2010; Stanton and Hutson 2012) suggests more specifically that the site reached its apex at the end of the Late Aak phase, which is to say the beginning of the seventh century ad.

Many of Chunchucmil’s architectural compounds, such as S2E1-G/Kaab’ and S2E2-C/Aak, also were occupied in the early part of the Late Classic (i.e., the seventh century). In both compounds, however, the major architecture was constructed in the late Early Classic. For instance, at Kaab’, a Chencoh bowl and a Hunabchen plate with basal flange and slab supports were found in the construction fill of structure 34, the domestic shrine (figure 4.5i, k). The fill underneath the floor of Structure 22 in Group S2E2-F/Aak consisted of sherds coming exclusively from the following groups: Maxcanú, Hunabchen, Kanachen, Timucuy, Chuburná, Aak, Tituc, and Sierra. Though Maxcanú continues into the Late Classic, it is found in association with different pottery during that period (Bond-Freeman and Mansell 2006), as the following section describes.

Late Classic: Kaab’ Phase (630–750/800 ce)

The Kaab’ phase of the Late Classic period at Chunchucmil consists of ceramic types first established at Oxkintok. On the basis of ceramic offerings from Tomb 7 (Structure CA5) and ceramics associated stratigraphically with buildings from the Proto-Puuc B architectural subphase and at Structures CA3 and the Satunsat, Varela Torrecilla (1993, 1998) defined two ceramic complexes at Oxkintok that pertain to the Late Classic period: Noheb (630–750 ce) and Ukmul I (750–850 ce). Four wares from the Noheb Complex at Oxkintok are common at Chunchucmil: Katil Unslipped (Motul Striated and Acanceh Appliqué-impressed types), Yucatán Glossware (Maxcanú Buff and Hunabchen Orange types), transitional slate (Sat Preslate type), and thin slate (Sabero Thin Preslate type). We have one radiocarbon date, 1392 ± 76 bp, calibrated at 2σ 436–806 ce, that likely pertains to the Kaab’ phase. The date comes from a burial of an elderly female in Structure 32 of Group S2E1-G/Kaab’. The burial (Burial 1, see also chapter 5) contained two vessels—a Kinich tripod plate and a Dzibical Black on Orange bowl (figure 4.5j)—from ceramic types that are rare at Chunchucmil but date to the Late Classic at other sites. Though the 2σ range extends back into the Late Aak phase, the pottery offered with this burial suggests that the actual date of this building falls in the latter half of the range.

We begin with the coarse ceramics, which show technological changes over the course of their nearly two-century developmental trajectory in the Late Classic. In this time period at Chunchucmil we see fragments of coarse globular jars with elongated, outcurving necks with striated surfaces (Motul Striated type, figure 4.7g, h; Varela Torrecilla 1993) or smoothed surfaces (long-neck variety of the Oxil Unslipped type, figure 4.7e, f).

Figure 4.7. Technological development of unslipped wares of Chunchucmil. (Photography by Pedro Tec Chim. Drawings by Belem Ceballos Casanova.)

A very common ceramic group at Chunchucmil, likely dating to the Late Classic Kaab’ with perhaps some production in the Early Classic Late Aak phase, is Katil (n = 16,304, figure 4.7d, this volume; Magnoni 2008:293). All but one of the Katil Group sherds pertain to the Motul Striated type. The other sherd pertains to Acanceh Appliqué-Impressed. Katil Group jar fragments have deep striations on their surface, giving them a rough texture. Compared to Oxil, Katil sherds are more durable, the calcite temper is more transparent, and the jars have outcurving necks. Katil Group jars typically have shorter necks than Oxil jars.

The Oxil Unslipped type begins in the early Classic and only the long-necked variety of this type (figure 4.7a) continues into the Kaab’ phase/early facet of the Late Classic. The number of Oxil sherds reported previously in the Aak phase is likely artificially high because some of the Oxil long-necked variety sherds were accidentally included. It is likely that a small number of Oxil sherds that pertain to the Kaab phase were counted as Aak-phase Oxil sherds. The long-necked variety of Oxil has a light red (2.5Y R 6/4; 6/8) paste with a slightly coarser texture than that of Elote Striated-Impressed from the Early Classic (figure 4.7c). Late Classic Oxil pottery has larger and more angular calcite grains than Early Classic Oxil, giving the later Oxil pottery a rougher texture. Corncob impressions from the Early Classic Oxil Group disappear in the Late Classic, during which only striated and smoothed surface finishes continue.

Yucatán Glossy Ware, another important Kaab’-phase pottery, has a beige/buff slip (Maxcanú Buff) or an orange slip (Hunabchen). Chunchucmil’s Maxcanú Buff pottery exhibits a variety of finishes and slip tones, from reddish brown (2.5YR 5/4) to light yellowish brown (10YR 6/4). Late Classic Maxcanú pottery from Chunchucmil generally lacks the vitrified slips common in Maxcanú Group ceramics from Oxkintok (Varela Torrecilla 1998). Chunchucmil’s Maxcanú pottery has lustrous, semilustrous, and cracked slips. This variation suggests that glossy wares with beige slips came from various locations or potting centers.

In the Kaab’ phase, Maxcanú cazuelas continue to be made with an exterior flange just below the exterior bolstered rims, although these features are thinner and vessels with this flange (Tiznuk Striated type) no longer have annular bases. Another cazuela form, this time with a very heavily bolstered double rim, appears during this time period as well. These cazuelas are part of the Yaxuná Preslate type, which, similar to earlier types, have roughly striated and smoothed walls (figure 4.6a). Jars from this period (Maxcanú Buff: Conkal variety) no longer feature composite silhouettes, are of globular form, and have thinner, outcurving necks with pinched rims. Tripod plates from the Maxcanú and Hunabchen groups show certain characteristics from the prior period but also display differences. Changes include slightly thinner walls and smaller supports, with Teotihuacan-style supports no longer being used. Also, the basal flange is less prominent.

During the Kaab’ phase, two new forms appear, both in the Maxcanú Group. We see basins with lightly thickened rims, folded to the exterior (Chunchintok Striated: variety unspecified; figure 4.6d). These basins have striations but no slip on the exterior. The second form consists of thin-walled bowls similar to Ticul-Thin Slateware that we call Maxcanú Buff: thin-walled variety. The slip of Maxcanú Buff: thin-walled variety is soapy to the touch and very pale brown (10YR 7/4), so translucent in color that a reddish slip color (5YR 6/6) is visible underneath. The early forms are composite silhouettes from the beginning of the Classic period (Ticul/Sabero) while its late forms, dating to the Late and Terminal Classic period at Chunchucmil, are bowls with beaded rims. Pottery similar to Maxcanú Buff: thin variety at Chunchucmil is found at the site of Oxkintok, which has been considered a production locale (Ancona Aragón 2014:163). Thin-walled bowls similar to Ticul-Thin Slateware have been amply documented in the region; based on paste difference, perhaps some should be reassigned to the Maxcanu Buff: thin variety.

Important Late and Terminal Classic groups such as Muna (n = 2,731) begin during the early part of the Late Classic and continue for at least four centuries. Kaab’-phase slatewares exhibit certain similarities with early slatewares reported by Varela Torrecilla (1993). These similarities include deep tripod plates with thick walls and cazuelas with marked cavities in the interior walls and annular bases with flattened lips. These slates are generally monochrome; when they are bichrome, the trickle painting technique is so subtle that it is nearly unnoticeable.

Late/Terminal Classic: Xnokol Phase (750/800–1,100 ce)

By the eighth century ce, Chunchucmil’s landscape had undergone significant transformations. Although many residential groups and some quadrangles continued to be occupied into the early part of the Late Classic, nearly all of them were abandoned by the late part of the Late Classic. However, this time period, which we combine with the Terminal Classic, experienced the construction of new buildings that differed significantly from previous architecture. It is difficult to separate ceramics from the early part of the Late Classic (seventh century and early eighth century) from the late part of the Late Classic (late eighth century and the ninth century) and the Terminal Classic because certain groups (e.g., Chum, Muna, and Teabo) were produced throughout the Late Classic and often later. If forced to specify the number of sherds that could be placed in the Xnokol phase, we have 18,512 sherds (12.5% of the pottery that has been classified at Chunchucmil). The most common pottery from this phase comes from the Chum (n = 14,315) and Maxcanu (n = 2,559) Groups.

Buildings constructed in the Xnokol phase consist of basal platforms measuring about 40-by-40 m, usually standing between 1 and 2 m high. All of these fall into architectural group type 6 (see Magnoni 2008 for more extensive details). We have excavated 16 type 6 platforms and 13 of these have substantial quantities of Xnokol-phase pottery (figure 4.4). In addition to these 13 confirmed Xnokol-phase platforms, we place six other unexcavated type 6 platforms in the Xnokol phase based on the presence of Puuc stonework (three cases), Xnokol-phase pottery on the surface (one case), and close formal resemblance to other Xnokol-phase platforms (two cases; Magnoni 2008). All of these platforms are located close to the site center (figure 4.8). Xnokol-phase platforms differ from earlier constructions in many ways. First, they are generally higher than earlier basal platforms. Second, they are never encircled by houselot walls. On the contrary, the builders of these platforms often located them in such a way that they disrupted earlier patterns of albarradas, streets, and sacbes. The builders often scavenged stone from these earlier features as well. Third, many of them have Early Puuc veneer stonework that is still standing, tumbled Classic Puuc cut stones found in association, or Puuc-style banner stones. Early Puuc and Classic Puuc architectural styles date from 650 to 1050 ce (G. F. Andrews 1985). Fourth, these platforms have more metates than earlier architectural groups: an average of 10 per platform compared to the site average of two metates per architectural group. Finally, none of these constructions features Plaza Plan 2 arrangements (Becker 1991), a common layout in earlier Chunchucmil houselots in which a shrine is located on the east side of the group. In Xnokol-phase platforms, the amount of Late and Terminal Classic sherds as a proportion of the total classifiable sherds ranges from 1.6 percent to 46.1 percent. The high amounts of earlier ceramics recovered from these platforms suggests that in many cases the Late/Terminal Classic platforms engulfed earlier constructions.

Figure 4.8. Map highlighting groups with Late/Terminal Classic pottery. The 19 groups in black consist of platforms built in the Late/Terminal Classic. The groups in gray were built in the Early Classic but 10% of their classifiable pottery tentatively dates to the Late/Terminal Classic. The dotted ring shows the location of the defensive barricade.

Several Late Aak–phase patio groups also have respectable amounts of sherds from the pottery groups that we place in the Xnokol phase, indicating that they could have been occupied in the Xnokol phase. For example, table 4.5 lists 17 architectural groups that lack the kind of Late/Terminal Classic construction discussed above but that have 10 percent or more of their pottery assigned to the Late/Terminal Classic (figure 4.8). These data suggest that the Late/Terminal Classic occupation of the site was not limited to the 19 broad, high, Xnokol-phase platforms. In these groups it is not clear what architecture would have been associated with this later occupation. Extensive excavations on structures would be necessary to understand if the Early Classic architecture still visible on the surface of these residential groups was still in use during the Late/Terminal Classic.

Beginning in 850/900 ce, Chunchucmil’s ceramics show important changes in manufacturing technology as well as in decorative styles. Maxcanú Group pottery in the Xnokol phase changes in paste texture and surface finish. The paste is hard and compact and shows a nucleus that is pink (7.5YR 8/3), light grayish (10YR 7/2), or red (2.5YR 5/6) while the edges are lighter (pale yellow, 2.5Y/2/8.5). The paste contains abundant small inclusions as well as medium-sized white calcite. The slip is gray (4/N), light gray (10YR 7/1), or light gray/brown (10YR 6/2) with abundant reddish gray (2.5YR 6/1) fire-clouds. The slip is porous and its texture ranges from rough to lightly soapy (Gray/White Slipped variety of Maxcanú Buff type). The changes in color and surface finish in the Maxcanú Group have been documented previously in the ceramics of Sihó (Jiménez Álvarez et al. 2006:487, figure 3) and Santa Bárbara (Stanton and Jiménez 2007).

The Late/Terminal Classic Maxcanú Group retains many of the forms and decorations of the previous Kaab’ phase. Cazuelas with massively bolstered rims of the Yaxuná Preslate type, bowls with lightly thickened rims folded to the exterior (Chunchintok Striated type), and jars with outcurving necks and pinched rims (Maxcanú Buff: Conkal variety) continue to be produced but now with a lighter and less glossy slip color. These forms are not present in Muna Slate pottery. Although differences exist in paste, form, and finish between Muna and Maxcanú, these differences can be subtle and confusion between Maxcanú pottery and Muna Slateware can exist if special attention is not paid when examining the two ceramic groups.

Maxcanú decorations from the Kaab’ phase continue to appear in the Xnokol phase. In this phase, black trickle paint becomes more common and the same impressed motifs are found on jars and drums. Black trickle paint is only found on cazuelas: in the jars and drums it shows up as almost imperceptible dark brown, vertical lines, possibly painted before firing. This pottery has been classified as Sacalum Black on Slate in various ceramic collections from northern Yucatán, such as the Oxkintok collection in the INAH Yucatán Ceramoteca. At Chunchucmil, we refer to it as Tacopate Trickle, Gray/White Slip variety (part of the Maxcanú Group).

The Xnokol-phase pottery from Chunchucmil also consists of rough pottery with long, outflared necks (Yokat Striated: Interior Neck variety, figure 4.7i, j), Dzitas Slateware, and slates with forms that are common to slatewares from elsewhere in northern Yucatán (chultun-like ollas, tripod plates, and ring-base bowls with beveling on the exterior rim). The Gray/White Slipped variety of Maxcanú often comes in these same forms. The paste of slatewares from Chunchucmil is less compact than true Muna Slatewares and has visible gray and white calcite particles. With the exception of the paste, these plates and chultun-like jars are physically similar to Muna Slate pottery and have been preliminarily placed into the Muna Group.

We have found fine orange paste pottery in association with Maxcanú and Muna at Chunchucmil. Fine gray pottery appears at Chunchucmil with characteristics that are typical of the lower Usumacinta River. Since Chunchucmil’s fine gray pottery exhibits similarities with fine gray pastes reported in other parts of the Maya area, we assign it to 750 ce or after (Ancona Aragón and Jiménez Álvarez 2005; Muñoz 2006; Rands et al. 1982).

The majority of Xnokol-phase ceramics come from off mound test pits and are found intermixed with earlier materials. In some of these excavations one can distinguish abundant quantities of Muna or Maxcanú of the Gray/White variety also in association with the Kukulá Group and a bit of Sotuta Complex pottery. This is interesting because such associations suggest a later date within the 750/800 to 1000 ce time range. Only three test pits penetrated the construction fill of the Xnokol platform and these reveal more or less pure Xnokol-phase pottery. In contrast, pottery deposits from construction fill of most of the earlier buildings are mixed, consisting of sherds from multiple periods, Preclassic to Late Classic (Magnoni 2008).

Postclassic Period (1100–1542 ce)

Archaeological evidence for the Postclassic period (1100–1550 ce) is limited to small amounts of sherds found at a few locations across the site. Our excavations have produced a total of 297 Postclassic sherds, classified to the Mama (n = 114), Navulá (n = 107), and Kukulá (n = 76) groups. This pottery comes from 18 different operations at Chunchucmil (see table 4.6, figure 4.9). Although only 13 Postclassic sherds came from Op. 41 (Group S2E1-E), these sherds make up 21.3 percent of the pottery that has been identified to type from this operation. Approximately 75 percent of the pottery from this group dates to the Late Aak and Kaab’ phases and the size and layout of the buildings resemble those of other groups dating to the Classic period. Thus, it appears that the architecture we see today was not built in the Postclassic. Rather, the area could have been used in some small capacity during the Postclassic period. In the other 17 operations that yielded Postclassic pottery, these sherds never make up more than 5 percent of the total.

Figure 4.9. Map highlighting groups with Postclassic pottery. In the groups in black, at least 1% of the pottery recovered dates to the Postclassic. Groups in gray have at least one Postclassic sherd but Postclassic pottery makes up less than 1% of the classifiable sherds.

No clear evidence for permanent occupation and no definite structures in use during the Postclassic have been identified. Excavations as part of the regional survey (see chapter 8) have also not revealed any Postclassic occupation. A Postclassic depopulation was also documented by Costa Maya surveys conducted in the region north of Chunchucmil by A. P. Andrews and Robles Castellanos (2004). Whereas the Costa Maya project found 130 sites with Terminal Classic occupation, only 21 sites had Postclassic sherds. The major contact-period sites in the Costa Maya survey region—Kinchil, Tetíz, Hunucmá, Ucú, and Caucel—are villages and towns that still exist today and likely conceal underlying Postclassic settlement.

Historic-Period Chunchucmil

The village of Chunchucmil is the closest modern settlement to ancient Chunchucmil, lying 2 km to the west of the site core. Archival research by Tony Andrews (2001) provides information about the historic-period occupation of Chunchucmil. The earliest documentation of any settlement where the contemporary village of Chunchucmil is located comes from 1783, when a resident of Maxcanú who owned the ranch at Kochol, located about 4 km to the southeast of the ancient site core (see figure 1.2), established a ranch at what would become the village of Chunchucmil by buying a well and the land surrounding it. We are less familiar with the historic occupation of Kochol, today a modern village, though we know that it dates at least to the early eighteenth century. In a 1798 map of the area, an unnamed town appears in the precise location of the modern village of Chunchucmil. The core of the Classic-period ruins is located about 2 km to the east of the modern-day village. In the early 1800s, Simon Peon acquired Chunchucmil. The powerful Peon family owned several haciendas in what is now northern Campeche and western Yucatán and held on to Chunchucmil until the mid-twentieth century. In 1828, Chunchucmil, with 46 residents, appears in the documents as a parish of Maxcanú. On Simon Peon’s invitation, Stephens and Catherwood visited Halachó, Sihó, Maxcanú, and Oxkintok, all visible in figure 1.2, but not Chunchucmil (Stephens 1843:109–145). The large buildings visible in the center of contemporary Chunchucmil began to be built in 1872, under the direction of Rafael Peon Losa, who established Chunchucmil as a henequen plantation, or hacienda. By 1906, when Mexican president Porfirio Díaz visited Chunchucmil, the hacienda consisted of, among other things, a house (the casa principal) for Peon Losa’s family, buildings for processing and storing henequen fiber, a chapel, a store, small living quarters for the more-or-less enslaved Maya labor force, and a network of narrow-gauge rails to cart agave spears from the fields to the hacienda. In the late 1930s, as part of nationwide land reform, most of the hacienda land was expropriated and became the core of the Chunchucmil ejido, a communal land-holding corporation controlled by former hacienda workers. Today, approximately 1,200 people live in Chunchucmil, which is in the municipality of Maxcanú. In the nineteenth and twentieth centuries, Chunchucmil’s economy was connected to the coast in ways that we believe resemble ancient Chunchucmil’s Classic-period economy. We explore these resemblances in chapter 12.

Defensive Barricade

One of Chunchucmil’s most intriguing settlement features has been only roughly dated thus far: a rubble wall enclosing 35 ha of the site core (see figures 4.8, 5.1). Dahlin (2000) proposed that this was a defensive barricade, explaining that a 340-m gap on its west side exists because the people of Chunchucmil, not anticipating an attack, constructed it in haste, confronting enemies coming from the east. In some cases the wall was built on top of sacbes and structures. The wall was definitely erected after the Kaab’ phase, since it ran over Structure 32 of the group S1E2-E/Kaab’, which was built in the Kaab’ phase. The group as a whole was occupied until the beginning of the Xnokol phase. We do not think the barricade could have been built across the Kaab’ architectural group while it was still occupied. This gives the wall a Late/Terminal Classic terminus post quem, likely being built when most of Chunchucmil was abandoned and the population consisted of the occupants of the 19 type 6 platforms listed in table 4.4 and a small number of houselots. Such a late construction date for the barricade would explain why several large architectural groups dating to the Early Classic, including the quadrangle with the second-largest pyramid (S1E2-B/Pomoche/Op. 19), were left outside this walled enclosure. It is unclear, though, exactly when the wall was built, since it does not clearly articulate with architecture of a specific time period. The wall might have been built after the Late/Terminal Classic, since the builders left more type 6 platforms from the Late/Terminal Classic period outside of the wall than inside.

Conclusion

Though we recovered ceramics from the Middle Preclassic to the Postclassic, Chunchucmil’s major occupation occurred during the late facet of the Early Classic. Population boomed in the fifth century and reached its apogee at the end of sixth century. An appreciable but much-reduced occupation in the seventh century eventually dwindled in the eighth century to less than a tenth of the site’s previous maximum size. In the following chapter we discuss the size, zonation, and demography of Chunchucmil when the city was at its peak and reflect on urban life in its three zones.