The Belize Postclassic project is a long term program of research devoted to understanding processes of social transformation and survival strategies of populations that recreated Maya society in northern Belize after the10th century collapse of Classic period city states in this region. Sites such as Laguna de On Island represent village settings where this process of cultural change took place, as communities preserved some institutions, discarded others, and transformed still other structures of the past to meet new circumstances of the Postclassic world. The Maya ethnographic record suggests that the community is a highly significant unit that defines social and political identity (Wolf 1959, Redfield 1941, Redfield and Villa Rojas 1934, Thompson 1970, Vogt 1976, Farriss 1984). Based on analogy to ethnographic observations, the community scale of analysis is an appropriate one for examining Postclassic cultural patterns further back in time, at village sites like Laguna de On Island.

The community focus of this project is also based on results of previous research in the southern Maya lowlands, which indicate that adaptations to the Classic collapse are variable and village-based. In fact, interpretations of the relative prosperity and ethnic affiliation of Postclassic sites is highly variable in the southern lowlands (Freidel 1985, Ball 1985, D. Chase 1982, A. Chase 1983, P. Rice 1986, D. Chase and A. Chase 1982, A. Chase and D. Chase 1985, D. Chase and A. Chase 1988, A. Chase and P. Rice 1985, Graham 1985, Graham et al.1989, Hester 1985, Masson 1993, Pendergast 1981, 1985, 1986, D. Rice 1986, Walker 1990). These interpretations variously suggest that foreign migrants replaced earlier populations in Belize (D. Chase and A. Chase 1982, Hester 1985), continuities are observed in transformation among in situ populations (Pendergast 1981, 1985, 1986, D. Chase and A. Chase 1988, Graham et al. 1989, Walker 1990), or that local and foreign ethnic expressions can be viewed at individual sites as a result of mutual assimilation (D. Rice 1986). Investigations of Postclassic community patterns were initiated at Laguna de On Island in 1991. Testing of deposits on the island and at a residential plaza at the lagoon’s southeast shore were conducted at this time (Valdez et al. 1992, Masson 1993). The analysis of these assemblages and their contexts provided preliminary indications of Classic to Postclassic cultural adapations, economies, and chronologies at this site which guided the 1996 season.

Although work has begun at Laguna de On Island, little is currently known at the regional scale of patterns of settlement, demography, economy, and political structure of northern Belize Postclassic culture. The long term goals of this project are to compile data through investigations at multiple communities which will ultimately facilitate regional scale interpretations. In the 600 years of Postclassic culture history in Belize, systems of community and regional organization are not assumed to be homogenous, given the variability observed across the lowlands in space and time (Freidel 1985). The long term objectives of the Belize Postclassic Project will document regional systems of political organization through extensive testing of Postclassic village sites from the Middle Postclassic, Late Postclassic, and Colonial periods. Political organization will be examined through the evaluation of intrasite and intersite social hierarchies, forms of religious monopology, as well as local and long distance economies. The analysis of artifact assemblage variablity and patterning from household and ritual contexts will provide the primary data for this examination. Belize is a significant region in the cultural geography of Yucatan and the southern lowlands in the 11th through 17th centuries, situated between two regions that are more thoroughly documented historically, i.e., the Yucatan and the Peten. The archaeological record of Belize can supplement biased historical accounts such as the Books of the Chilam Balam which emphasize the role of foreign invasion in lowland political history. The untold story of village adaptations in locations further removed from the hot seat of political affairs in northern Yucatan is represented in the Belize Postclassic archaeological record. The Belize Postclassic Project has a long term commitment to analyzing the history and process of these late centuries in the Belize region.

Of broader anthropological interest to this project is a process of cultural change that may be termed selective reproduction (Masson 1997a). Analyzing selective reproduction of cultural institutions in a resistance climate like that which may have existed after the Classic period collapse in Belize assumes conscious historical action and the presence of social memory among Postclassic villages and their leaders. In the Postclassic period, certain institutions of Classic period Maya society were retained (like calendric rituals) while others were intentionally disgarded (like divine kingship and investment in monumental architecture). This investigation assumes that social memory was not lost in the Postclassic, and that Postclassic peoples were not "incapable" or insufficiently "complex" to achieve the centralization of the Classic period city states. It is instead considered that social and political transformations toward more fragmented and autonomous secondary regional state systems were accomplished through active choices made in Postclassic cultural reproduction, and these choices were based on the perception that such changes were generally beneficial to the participants involved. Essentially, patterns observed to date suggest that Postclassic Maya society in northern Belize was reproduced, at the village scale, through the selective replication of institutions of the past (Pendergast 1986, Masson 1997a). The mechanism of replication of ritual or other organizational principles was originally observed by Vogt (1969) who noted this phenomena among communities surrounding the center of Zinacantan, Chiapas.

To observe patterns of selective reproduction of Maya culture archaeologically, excavations at Laguna de On Island are investigating a variety of ritual and domestic contexts through detailed horizontal excavations which will facilitate spatial analysis. The project seeks to collect empirical evidence attesting to village political organization, social hierarchies, ritual practice, and local and long distance economic patterns from which broader questions may ultimately be addressed. To date, this approach has provided evidence for the reproduction of certain religious rituals by political authorities at the island (Masson 1997a). Household production and local economy also exhibit significant reproduction of earlier systems (Masson 1993, 1997b). Reproduction of institutions from the past was accompanied by change and adjustments to new external political and economic realities (Masson 1997b).

The objective of the 1996 season was to gather evidence of village social hierarchy and political organization at Laguna de On Island, through expanding the research investigations of the 1991 season (Valdez et al. 1992, Masson 1993, 1997b). Initial testing in 1991 facilitated general predictions for subsurface architectural recovery, while the 1996 project sought to recover detailed spatial information from ritual and domestic contexts. The primary focus of 1996 excavations was ritual architecture, although at the beginning of the season the function of the structures was not known. The recovery of Postclassic ritual behavior is significant to the project’s interest in political organization, as religious institutions preserved in the archaeological record provide important insights on the survival and transformation of beliefs and customs of the past. They are also a playing field for political control and advantage that is closely linked to community political systems and social organization. Two ritual buildings were exposed in 1996. Additional 1996 testing focused on community investment projects in landscape modification, as such efforts provide a means of measuring the degree of persuasive political authority through the mobilization of cooperative works. Public construction projects were documented and assessed through exploring artificially constructed terraces, fortification walls, and canoe docks at the island.

Evidence for social status differentiation at the island was to be examined in the 1996 research plan by comparing the architectural elaboration, possessions, and production activities of households. Family size and kin-based social identity were to be similary assessed through investigating architectural form, artifact assemblages, and mortuary patterns. Although a number of locations were tested at the island with the aim of uncovering domestic architecture, a household structure has yet to be located. Several domestic middens and associated burial areas have been found, however, and analysis of domestic assemblages has begun in advance of the recovery of associated structures.

Excavations in 1996 employed the technique of broad horizontal exposure of architectural features visible from the surface or encountered in the 1991 season. Broad horizontal exposure allows for maximal observation of spatial relationships between artifacts, features, and architecture. Spatial associations provide the richest data for interpreting the meaning and function of archaeological deposits. As Postclassic remains on the island are found at the surface and extend to a depth of around 30 cm, this village site poses an ideal situation for horizontal exposure due to a lack of overlying deposits. In 1996 this methodology was accompanied by a program of smaller test units that explored areas of the island to be expanded into horizontal exposures in future seasons.

For tight spatial control, large buildings were excavated within 2X3m square (or smaller) collection units. Within these units, artifacts associated with different deposits (or lots) were also collected separately. Lots were designated according to soil characteristics, directional relationship to important features such as walls (internal and external space, for example), changes in artifact densities, and arbitrarily. Vertical levels were customarily divided into 5 or10cm intervals. All soil was screened through 1/4" hardware mesh, and stone tools and debris, ceramics, animal bone, and shell were collected. Additional soil samples were taken from features and profiles for later flotation, pollen or phytolith analysis, or soil humate dating. Suboperation 14 underwater units were tested by shovel scraping surface lagoon sand directly into a 1/4" screen and water-screening the deposits in each unit. Water at Suboperation 14 averaged 80cm deep, which facilitated this method.

Excavations at Laguna de On Island during the Summer of 1996 were concentrated in the upper status precinct of this Postclassic community, located at the island’s apex (Figure 1). Investigations also tested selected areas on the outer terraces of the island (Figure 1) to provide valuable comparative information on the variability in social status, religious and economic activities, and other clues to the organization of Postclassic society at this village. In the upper courtyard of the island, two architectural complexes were horizontally cleared, the Suboperation 8 complex (Structure I, Masson et al. Chapter Two) and the Suboperation 12 complex (Structure II, Rosenswig and Becker, Chapter Four). Additional testing in this area explored the boundaries of a large courtyard area along a boulder wall that delineates the west edge of the apex area in Suboperations 7 and 13 (Barnhart and Howard, Chapter Six). The south area of the island, away from the upper courtyard, was explored in test units designated Suboperation 5b and 5c (Barnhart and Howard, Chapter Six). A partially submerged dock structure at the water’s edge to the east of the apex was tested in Suboperation 14 (Masson and Gonzalez, Chapter Five). These Suboperations are described in detail in the chapters of this report, and their location on the island is shown in Figure 1. A summary of findings is described below, in sections on artifact analysis, ritual architecture and test explorations.

As a result of efforts of volunteers, workmen, and staff during the 1996 excavation at Laguna de On Island, 190 square meters of Postclassic structures and activity areas were exposed. Recovered materials include 12,875 ceramic sherds, 9,663 pieces of debitage (stone tool manufacturing debris), 5,300 pieces of animal bone, 307 obsidian blades, 212 stone tools, 172 net weights, 24 modified ceramic sherds, 21 spindle whorls, five human skeletons, eight pieces of worked marine shell, two clay balls, one incised clay object, one serpentine adze, one piece of coral, one flint eccentric and assorted samples of soil, pigment, burned rocks and carbonized wood for specialized analysis (for relative percentages by Suboperation see Table 1). With the exception of the flint eccentric which remained with the Department of Archaeology of Belize, all materials were shipped to the lab at SUNY-Albany, where they will be analyzed through studies conducted by collaborating professionals and graduate students. Currently, preliminary analysis has been conducted on 100% of the debitage from the site, 100% of the lithic tools, 100% of the obsidian blades, 50% of the ceramics, and 50% of the faunal bone. Preliminary analysis has primarily been classificatory, and future specialized studies are planned for raw material sourcing for lithics and ceramics, tool use wear, residue analysis, and radiocarbon dating. Ongoing analyses that are currently underway include completion of chronological and functional ceramic analysis (by Shirley Mock), completion of the faunal analysis for a master’s thesis at SUNY-Albany (by Jennifer Wharton), analysis of skeletons and mortuary patterns (by Rob Rosenswig), lithic analysis (by Masson and SUNY students), and GIS-based spatial analysis of Laguna functional assemblages, distributions, and chronological trends (by Masson and Rosenswig).

Excavations during the 1996 season in a possible ritual courtyard area at Laguna de On Island yielded important information about 11th-14th century religion. A C-shaped public building (Structure I), a large shrine platform (Structure II), and two burial areas provide initial views of

Figure 1. Map of 1996 Excavations at Laguna de On Island (originally drafted by Kathryn Reese-Taylor).

Table 1. Artifact Frequencies from 1996 Excavations at Laguna de On Island

ritual behavior at this site. Artifact concentrations at these buildings identify activities that took place there, such as the sacrifice and consumption of large game animals (Wharton, Chapter

Eight), incense burning, breaking of ceramic vessels, and caching of offerings of exotic chert stone tools (Masson et al., Chapter Three). Both structures are associated with concentrations of broken ceramics that date to the Middle Postclassic (Mock, Chapter Seven). There appear to be some functional differences, based on architectural composition. Structure I has a three-sided wall foundation in a C-shaped configuration. It is associated with upper status burials. Structure II is a rubble platform, which faces a rubble patio that extends to the south of it. No burials were found in the vicinity of Structure II, and numerous concentrations of highly burned rock were recorded within the rubble of this platform. Functional differentiation of these buildings awaits completion of analysis. However, neither resembles a domestic structure due to a lack of household debris, such as edible portions of large game, manos, metates, or contemporary fire pits. However, in the soil zone beneath each structure, there appear to be midden deposits that suggest a domestic occupation existed in this vicinity prior to the construction of these nondomestic buildings. The midden deposits appear to be primary, as pits full of charred soil are present within them. These pits originate well below the soil level associated with the architecture of Structures I and II (about 30 cm below surface), and are overlain by walls or rubble of each building. These results raise intriguing questions to be addressed in ongoing artifact analysis which will examine temporal trends in these assemblages.

The Postclassic ritual buildings and deposits at Laguna de On may reflect a complex of religious beliefs and rituals associated with calendrics, the agricultural cycle, and ancestor veneration that has its origins in the Preclassic period (at least 350 B.C.) and continued through the Colonial era (Taube 1992, Freidel et al. 1993, Masson 1997a). Evidence from Laguna may thus demonstrate the selective reproduction in the Postclassic period of traditional Maya beliefs that remained useful to affluent village agriculturalists of Belize, in particular, rituals pertaining to rain. Through examining such behavior, Postclassic reconciliation of social memory and social choice in the context of recreating Maya society can be documented. This process has been alluded to in investigations of reproduction of Maya culture at Postclassic sites across the lowlands (Pollock et al. 1962, D. Chase 1982, Freidel and Sabloff 1984:183-184, Pendergast 1986) and in the highlands (Blake 1985).

The 1996 excavations reflect considerable distinctions in social status and the existence of a political hierarchy at Laguna de On Island. Burial deposits in the ritual center included skeletons with filed front teeth (Becker, Appendix I), a cosmetic touch suggesting elite status. These upper status burials were located around a ritual C-shaped structure (Structure I). In contrast, burials recovered elsewhere at the island lack filed teeth and are not in proximity to public buildings. The association of elite cemeteries with ritual activity implies a renewed significance for kin-based political power in the Postclassic.

Excavations outside of the possible ritual courtyard in the vicinity of Structure I and Structure II showed intensive and elaborate landscape modifications that improved the island’s contours and made them more suitable for village habitation. In Suboperation 5b, a lens of fill over a meter deep was documented in an area that appeared natural from the surface. The detection of fill at this location supplements information provided by the original Suboperation 5 excavations in 1991, which documented a terrace wall at the water’s edge that retained an additional meter of fill. Suboperation 5c provides important contrastive stratigraphy that verifies the artificial nature of the Suboperation 5 and 5b deposits (Figure 1). Although Suboperation 5c was only 5 meters upslope (and 20cm higher in elevation) than 5b, only 30 cm of topsoil was present over bedrock at 5c. A gradient of 20cm is insufficient to account for a difference in depth of soil formation from 30cm to 1m, especially at the top of a slope. The 30cm deposits in 5c represent the natural soil formation in the humic layer over bedrock found throughout the island. Bedrock rises to a high point in Suboperation 5c, and dips low in 5b. This dramatic undulation of bedrock was levelled out in Postclassic terrace construction. Suboperation 5 and 5b indicate the extent of artificial terrace construction in this location of the island, which covers approximately 1/3 of the island’s width. In bedrock at Suboperation 5c, a 1X2m trench, three burials have been found which may represent a cemetery area. Broad horizontal exposure in this area planned for 1997 will clarify the extent of this burial concentration, and its associations. These burials may be of a lower status than those recovered at Structure I, as they did not exhibit tooth filing. Additional public facilities constructed at the island include a stone dock that projects into the water on the island’s east side (Suboperation 14) and a large boulder wall framing the west edge of the upper courtyard (Suboperations 7a and 13). These efforts imply that other areas of the island might be similarly artificially constructed, a possibility to be explored in future testing.

The terrace, wall, and dock landscape modifications at Laguna suggest the existence of an influential political office during the Postclassic period at this community. At first view, the island appears to be a natural phenomena geomorphological phenomena. Neither architecture or landscape modifications are visible from the surface. As a result of 1996 investigations, it is apparent that much of the island may be artificially modified through engineering feats undertaken by Postclassic villagers. These observations suggest that Postclassic Belize Maya society was politically complex, more so than might have been surmised from the lack of visible construction projects observed on the island’s surface.

One of the major goals of this project is to reconstruct the political organization and social hierarchy of the Laguna de On Postclassic. Last year’s investigations in the upper courtyard recovered Structures I (Suboperation 8) and II (Suboperation 12), which provide clues about important foundations of political and religious behavior. The organization of rituals at Laguna de On Island may have been orchestrated by important village headmen, as observed in Colonial times based on ethnohistoric analogy (Scholes and Roys 1948, Coe 1965:100-102, Fox 1987, D. Chase 1982, 1986, 1988, D. Chase and A. Chase 1988, Masson 1995). Village leaders could have been elected officials or members of powerful founding lineages of the community. Postclassic political transformations throughout Mesoamerica express a renewed importance of kin-based power (Fox 1987) which seems to be exemplified at Laguna de On.

The importance of lineage-based distinctions at Laguna de On might be supported by the proximity of these public buildings to a burial area from which the skeletons with filed teeth have been recovered. It is significant that the male and female skeletons buried at Structure I exhibit the only tooth filing observed from the island. This pattern suggests that important ancestors, such as lineage heads, may have been buried in this prominent location at the top of the island. The activities and religious architecture found in this area may be directly related to the presence of interments of important elders who were to become venerated ancestors. Many Maya groups believe that ancestors possess powers to intervene on behalf of the living by appealing to supernatural entities in the spiritual realm that control natural forces such as rain (McAnany 1995).

God K is closely associated with ancestors in Maya mythology, and an effigy eccentric in the form of this deity was found in close proximity to Structure I and its associated burials. The flint eccentric is a likely heirloom manufactured in the Classic period, deepening its significance as a cached offering at Structure I. The presence of this artifact on the island links this Postclassic community with the recent past in a special way. Whether it was looted from a Classic period tomb by Postclassic people or handed down through generations, it signifies that ties to the past were important to the villagers of Laguna. This is a thought-provoking observation, as processes of Postclassic social transformation are not well understood in this part of Belize. It is not known to what degree populations of the Classic period were displaced or replaced in the Early Postclassic, nor is it known to what degree institutions of the Classic period were rejected by villagers who faced new opportunities in the Early Postclassic to throw off burdens of labor, military service, and taxation by regional lords. Certainly the lack of large scale monumental architecture, hieroglyphic inscribed monuments, and the institution of divine kinship were aspects of Maya culture that were rejected, not forgotten, in the Early Postclassic. Yet the caching of the God K heirloom suggests that some aspects and objects of the past remained significant and were included in politically orchestrated ritual.

Other significant research issues will be addressed from analysis of materials from within and outside of Laguna’s ritual precinct, including chronology, paleoecology, demography, and economic systems. Ceramic concentrations found at Structure I included vessel types that are often assigned to different time periods (ranging from late Early Postclassic, Middle Postclassic, and Late Postclassic), but their recovery together in features at Laguna may ultimately result in a refined Postclassic ceramic chronology for northern Belize. Bones of sacrificed animals associated with these ceramic features will be used for radiocarbon dating.

Identification of the bones of numerous forest dwelling large game animals, such as tapir, reflects a major rejuventation of the 12th-14th century ecology in northern Belize. Improved hunting opportunities and animal proteins in the diet for the Laguna community are implied from the recovery of abundant terrestrial and aquatic animals in ritual and domestic deposits. By contrast, game is scarce or absent from most Classic period domestic contexts in this region (Scott 1980, 1982) and pollen records show considerable deforestation (D. Rice 1986). Skeletal analysis will further provide testimony regarding the health of Middle Postclassic populations at Laguna. The population at Laguna appears robust and exhibits minimal pathologies, perhaps due to a more diverse and protein rich diet facilitated by reforestation. Bone chemistry studies planned for these remains will supplement current understanding of Postclassic human biology. Middens sampled throughout the island will enable analysis of ceramic and lithic assemblages pertinent to understanding local production and exchange relationships of this community.

Tests in Suboperations 5, 5b, and 5c provide evidence of considerable community labor projects which modified the natural landscape of the island. The upper contour (see Figure 1) is artificially extended through the addition of fill to build a terrace of at least 13m in length, retained by large boulder walls. Similar modifications around the site may have given the island its current tiered and fortified appearance. Future testing will be designed to measure the extent of this effort in other zones of the village. Extending the upper surface of the island in the Suboperation 5 area made it more suitable for human occupation as the original configuration of the island shows dramatic undulations in topography. The modifications are significant as they reflect a considerable degree of village-level social complexity measured by the amount of organization and effort required to change the surface of the island through earth-moving and wall construction. Such modifications require leadership capacities found in socially complex societies. The documentation of such efforts is significant for estimating political complexity and tracking changes in mechanisms of community integration and focus in the Early Postclassic period, particularly as traditional monumental architecture is absent from the site. Future analysis will focus on the composition of the terrace fill itself, as currently it origin is not known. Collaborative geomorphological efforts are planned in conjunction with radiocarbon dates and artifact analysis aimed at determining whether the fill is from elsewhere on the island or whether it was brought by canoe from the lagoon shore. This determination is significant as transporting fill from the lagoon shore would exponentially increase the estimates of labor investment in this project.

A Comment on Chronology

It may be noted that much of this introduction has avoided the issue of Postclassic chronology, in particular the assigment of "Early," "Middle," or "Late" Postclassic to the Laguna de On Island site. There are two reasons for this abstinence. First, radiocarbon results from the 1996 season are not yet completed, although samples have been selected and are ready to be sent as soon as affirmation from solicited sources of support is received. The second reason is that confusion exists in the literature regarding the chronological correlation of the Early, Middle, and Late Postclassic. For example, the ceramics from late Early Postclassic/Middle Postclassic at Colha are thought to be contemporary with the ceramics from what is termed the Late Postclassic at Tulum (Valdez and Mock 1985). As Valdez and Mock (1985:74) proclaim, the Colha dates are corroborated by radiocarbon dates, and the Tulum dates are not. Indeed, the ceramics may be contemporary, but the names of the periods at each site are contradictory. Tulum-like redware ceramics comprise the majority of types identified at Laguna de On Island. At Tulum, these redwares have been dated to AD 1250 (Ball 1982:111) and AD 1300 (Smith 1971:30). According to murals and architecture, however, Tulum "flourished" after AD 1400 (Miller 1982:11). All three dates are based on stylistic observations. Tulum-like wares at Santa Rita have been assigned dates of AD 1150-1300, based on initial radiocarbon results (D. Chase and A. Chase 1988:78).

Santa Rita Corozal is generally referred to as a Late Postclassic site (D. Chase and A. Chase 1988:13), although published dates from this site are primarily 13th and 14th century (1988:Table 2) as are the katun endings recorded on the site’s murals (Long 1919). It is hoped that the difficulties with nomenclature described above will eventually be resolved, and for the time being it appears that Laguna de On Island is a primarily Middle Postclassic site with ceramics that resemble Tulum, Santa Rita, and Mayapan. Tulum-like redwares are found at the site in addition to Rita Red and Chen Mul Modelled forms. The duration of occupation at this site is thought to begin in the "late Early" Postclassic and extend into the Late Postclassic as indicated by the continuum of ceramics recovered at the site in the 1996 season. Detailed temporal affiliations will be determined as ceramic analysis progresses with the aid of radiocarbon dating.

A section of one of the walls of Structure I (Suboperation 8) was uncovered during the 1991 season in a 2X3m unit (excavated by Kam Managhan), which revealed a burial, offering, and fire pit to the west of the wall and a concentration of animal bone (crania of large game) and ceramic censers to the east of the wall (Valdez et al. 1992, Masson 1993). This initial test suggested that this structure was the focus of ritual activity (Masson 1997a). However, more extensive excavations were needed to document the configuration of this building and the associated artifacts that would explain the nature of its use. The position and dimension of Suboperations 8a through 8o are shown on the composite map of the island (Figure 1). These units exposed the walls as well as the internal and external space associated with Structure I. Together these units exposed 67 square meters. Excavations at Suboperation 8 were supervised by Joy Shumake and Evon Moan. The photograph in Figure 2 shows Structure I after most of these Suboperations were completed.

Structure I is a C-shaped building, as it has stone wall foundations on three sides (north, west, and south) and is open to the east (Figure 2, 3). The upper part of the walls would have had perishable pole and thatch extensions to the roof. The front may also have been partially walled with perishable materials, or it may have been open. The walls of Structure I were made of limestone, including a hard nonporous variety and a softer "solidified marl" variety that showed greater signs of erosion, both types available in the local bedrock substrate. The walls of the building were partially outcropping on the surface, which guided the placement of excavation units. Some stones were misplaced due to tree root disturbance. Others were burned, suggesting the structure may have been destroyed by fire.

One very large stone was located in the center of the open side of the building (Figure 3). Despite hypotheses that it might have been an altar or uncarved stela, there was no direct evidence on the stone itself to support this such as carving or residual pigment on the stone’s face. A substantial concentration of ceramics were found to the south of it, and lithic tools may have been placed at the northwest corner of this stone. Such concentrations were recovered in other areas around Structure I, but those recovered to the south of this stone in Lots 92, 94, 101, and 103 were among the most numerous concentrations associated with this building.

A total of eighty-six lots were defined (Table 2) for the fifteen horizontal subunits delineated for the Suboperation 8 (Structure I) excavations. These horizontal and vertical units of collection will enable spatial comparisons to be made of different areas of Structure I. For temporal analysis, a correlation of these lots among the horizontal subunits is more useful for general analysis.

Figure 2. Photo of Structure I, facing west

Figure 3. Map of Structure I (illustration prepared by Pamela Headrick).

Stratigraphic deposits associated with Structure I are listed below, and the associated lots are summarized in Table 3, and profiles are displayed in Figure 4.

Deposit A) a humic topsoil found within and outside of the wall alignments that defined Structure I, averaging a depth of 20cm. This soil was loosely packed and riddled with fibrous tree roots. Ceramics, lithics, obsidian, and faunal bone were recovered in this layer. The materials in this soil are thought to be contemporaneous with the use of Structure I (Middle Postclassic), and some materials in this soil may postdate the use of Structure I (Late Postclassic). A burial (Burial 6) was found in this topsoil.

Many ceramics deposited around Structure I were recovered in nine dense and distinctly isolated concentrations, ranging from 20-50 cm in diameter (Table 4, Figure 5). In contrast, soil outside of these deposits had very little debris and few ceramics. Faunal bone, freshwater snail shells, fishing weights, and occasional obsidian blades were also present in these concentrations. Isolated sherds and partially reconstructable vessels were present in these features, and some sherds were refitted from different concentrations. These refits imply that the deposits occurred within a short period of time, and that the vessels were broken prior to their deposition in the concentrations. The origins of these concentrations of mixed vessels pose interesting questions and it is uncertain whether the features are primary or secondary deposits. On a general level, however, the composition of vessel forms (including many incense burners) found in these concentrations are used to infer a ritual function for Structure I as described later in this chapter. Other concentrations of materials ("artifact concentrations") were defined, which contained higher proportions of nonceramic debris, such as snail shells, animal bone, and chert or obsidian items, though lithics were always present in low numbers (Table 4). Fishing weights or notched sherds were also occasionally observed.

Ongoing analysis will investigate the explanation of these deposits at Structure I. Smashed vessels are often associated with termination events (Walker 1990), and much of the lithic debris around Structure I appears to have been burned, perhaps with the abandonment of this structure. Breakage of incense burners is also a characteristic of shrines found at sites in northern Belize after their abandonment in Precolumbian times (Sidrys 1983:242-244, Valdez 1987, Walker 1990:399-411, Masson 1997b). Such shrines may be linked to pilgrimage circuits documented in Late Postclassic and Colonial Yucatan (Freidel and Sabloff 1984:74-75, Walker 1990:311,

Masson 1997b), which served significant integrative functions in defining regional polities as they do in ethnographic Zinacantan (Vogt 1969:391, 1976:187, Freidel and Sabloff 1984:74-75).

Table 2. Lot descriptions from Suboperation 8, Structure I.

Table 3. Lots Correlated with Deposits at Structure I (Suboperation 8)

Figure 4. Profile of Suboperation 8, north wall. Northwest corner of Structure I viewed on right side of photo.

Figure 5. Ceramic concentration at Structure I.

Table 4. Concentrations of Artifacts and Ceramics at Structure I.

Burner ceremonies are also frequently linked to calendrical ceremonies in codaical and ethnohistorical accounts. Currently, it is not known whether the broken burner vessels at Structure I are associated with possible calendrical ceremonies contemporary with the use of this building, deposited at the time of its destruction, or placed there after it was abandoned. Regardless of these alternatives, they do signify its importance as a ritual locality. As the ceramics, lithics, and faunal remains largely match those found elsewhere on the island, the possibility that these assemblages are contemporary with the village’s Middle Postclassic occupation seems most probable.

Burials were found around the exterior of Structure I, in what appears to be two distinctive deposits (Figure 6). First, to the west of the building, intruding into the soft limestone substrate, two adult burials were found this year (Figure 6). In 1991, an additional adult burial (Burial 4, male) was found in this location (Figure 6). The two burials from 1996 (Burial 5, male and Burial 9, female) had filed incisors, which may indicate a higher social position for these individuals. This practice appears to have continued in the Postclassic, and supports other information suggesting that the apex of the island was the focus for elite activity. Burials 4 and 5 from this area were seated and flexed, and Burial 9 was flexed on her left side.

A fourth burial, that of a child (Burial 6), was found to the south of Structure I (Figure 6). This burial deposit is distinguished from the other interments to the west by two characteristics. First, it is located in shallow topsoil, just 20 cm below the surface and paralleling the south wall. Second, this individual was in a different position, facing north, lying on the left side and semi-flexed. This child may have been buried at a later point in the island’s history, but shares proximity to Structure I with the other interments.

A flint eccentric appears to have been cached outside the south wall of Suboperation 8 (Figure 7) at a shallow depth of 15 cm below the surface (Masson, Chapter Three, this volume). Flint eccentrics in northern Belize often take the shape of lightning bolts, various geometric patterns, or effigies of animals or humans (Pendergast 1982). The eccentric from Suboperation 8 has two effigy heads of the Maya God K on it, a deity often associated with lightning or rain. The effigies of God K are identified by smoking torches emerging from their foreheads. This artifact was apparently left as a cached offering. God K effigy eccentrics are rare in the Maya lowlands, and are associated with Classic period royal tombs or monumental caches. Other collected pieces are reported by Schele and Miller (1986) and McAnany (1995, from the Dumbarton Oaks collection). The significance of this find as a probably heirloom at the Postclassic village of Laguna de On is discussed in Chapter Three.

Additional caches were found in 1991 from this vicinity. A lenticular biface was found near Suboperation 8 in Suboperation 3 (a courtyard patio surface located immediately west of Structure I (Masson 1993, 1997b)) which also may have been cached in the topsoil. An additional cache was located in 1991 outside the west wall of Structure I, which included a drilled Pomacea shell, a shell bead, an obsidian blade, and a flake (Masson 1993, 1997b). Together, these three sets of possible offerings may identify the significance of caching in this ritually significant locality.

Figure 6. Burials at Structure I.

The lack of large quantities of domestic debris at Structure I (ground stone, staple faunal remains, cooking features, and utilitarian lithic tools) correlates with a high proportion of ritual ceramics, ritual faunas, and cache offering compared to other areas excavated at the site. Ceramics consisted of a high proportion of incense burner types (Figure 8), including Cehac-hunacti composite, Ak applique, Tsabak unslipped, and a single example of Chen Mul Modeled effigy censer (Mock, Chapter Seven, this volume). At Structure I, censers formed at least 20% of ceramics identified by Shirley Mock (Chapter Seven, this volume) during preliminary analysis. This percentage of censers is twice that recovered from Structure II (also a ritual public building) and is three times the amount found in the domestic terrace area of Suboperation 5 (Mock, Chapter Seven, this volume). Other types of vessel fragments are found in low numbers, interspersed with the censer sherds. The inclusion of utilitarian wares in ritual deposits, however, is not uncommon (Smith 1971:Table 20, Levanthal and Baxter 1988:Table 3.1).

Most of the ceramics from Structure I date to the Middle Postclassic. This assignment is based on high frequencies of Payil Red ceramics (31.3%) that are thought to be affiliated with the Tulum Sphere (Mock, Chapter Seven, this volume). Middle Postclassic contexts also contain ceramics related to the Early Postclassic Augustine sphere (Zakpah Orange Red 6.3%), as well as various censer vessels and Rita Red ceramics are similar to those forms common in the Late Postclassic (Chase 1982, Mock, this volume). Late Postclassic Chen Mul modelled sherds scattered around the surface of Structure I have been partially refitted and represent (to date) a single Mayapan style effigy incense burner (Figure 9). This vessel may postdate the earlier materials or signal the initial arrival of this form in northern Belize around beginning of the 14th century. The mixed ceramic lots at Laguna suggests that transitions from the Early, Middle, and Late Postclassic resulted in lingering characteristics (Mock, Chapter Seven, this volume). Securing absolute dating for ceramic chronologies is a primary objective in the ongoing analysis. When ceramic analysis is complete, a seriation analysis is planned that may help to refine the chronology as well.

Cranial elements of large game animals found in association with the ceramic concentrations of Structure I, apparently deposited in the context of the same activities. These cranial bones were disarticulated, and do not represent entire skulls placed in situ. Concentrations typically contain scattered alligator, peccary, deer, or tapir teeth, skull or mandible fragments (Wharton, Chapter Eight, this volume). The deposition of these faunal remains around Structure I at Laguna de On Island may represent behavior associated with calendrical burner ceremonies depicted in the Maya codices (Masson 1997a, 1997b, Taube 1988:244) as well as Colonial and ethnographic accounts (for example Landa in Tozzer 1944:62-63). Landa (in Tozzer 1944:62-63) describes the consumption of animals sacrificed in calendrical rituals. Such ceremonies at Laguna de On Island may have served an important role integrating the community and reinforcing village political leadership.

It is noteworthy that the taxa involved are those that would have represented high status animals, limited to large game (Masson 1995). Smaller, "staple" taxa, such as armadillo, fish and turtle are evenly distributed around the island (Masson 1995). Also equitably distributed are the postcrania of large game around the island, suggesting that these animals may have been redistributed (Masson 1995).

The ritual assemblages of Structure I provide important insight into religious institutions in the Postclassic at Laguna de On Island. The 1996 tests of this structure observe that an occupational zone is represented upon bedrock in this location. Subsequently, a C-shaped structure was built. Associated with this building are concentrations of broken ceramics, artifacts, and large game crania. A possible stelae or altar stone is centrally located at the open face of the building, which was also associated with an artifact concentration and a number of stone tools. Three burials that may be of upper status individual are found intruding into bedrock outside the rear west wall of the building. A fourth burial, that of a child was found in the topsoil outside the south wall. Near this burial, also in the topsoil, a God K effigy eccentric flint was found. Remains of a single Mayapan style censer were also found scattered around the building, but it is not known whether this Late Postclassic form postdates the use of Structure I or not. Further research will focus on the function and chronology of the materials collected in 1996.

Figure 7. Flint God K eccentric in situ at Structure I.

Figures 8. Ceramics from Structure I: top row,

Canton Incised: variety unspecified;

center, Payil Red: Palmul Incised;

bottom row, Payil Red: variety

unspecified.

Figure 9. Ceramics from Structure I: composite censer wares (Cehac-Hunacti).

An eccentric flint was found, apparently cached, outside the south wall of Structure I. On this flint, two God K effigy heads have been chipped bifacially off of a single blade (Figures 10, 11). The eccentric was found at the southwest corner of Structure I, cached 15 cm below the surface. The effigies of the deity God K are identified by smoking torches emerging from their foreheads (Schele and Freidel 1990:414). This artifact was apparently left as a cached offering, although as it is located in the topsoil, stratigraphic support for its identity as a cache (such as an intrusive pit) was not visible.

The flint is manufactured of dark grey, mottled chert with a fine-grained, waxy, and moderately siliceous consistency. The chert is thought to be from Colha, a neighboring stone tool manufacturing center (Thomas Hester, personal communication, January 1997) It exhibits a proximal break in an area of the tool that appeared to have been the handle or base of the artifact. A portion of the surface of this proximal end appears unchipped bifacially, giving the artifact an incomplete appearance. Close examination shows that this planar region is actually covered in clear subcortex, which may have been prohibitive for reduction when the artifact was manufactured. Other areas of the artifact are completely finished, including final retouching of serrations that form the headresses and faces of the effigies. These portions of the tool suggest that the artifact was completed, and that the unchipped surface was intentionally left this way. The break would thus represent a use-related occurrence rather than a manufacturing accident as it is unlikely that bifacial thinning of the shaft would occur after delicate late stage shaping of the effigies in the manufacturing sequence. An additional break is noted on the lateral side of the artifact near the proximal end. A third head may have projected from this narrow constricted projection that appears to have snapped off. This break is shown at the base of the artifact on the left lateral side in Figure 10. A small serrated segment beneath it was probably related to the missing portion of this effigy.

Effigy flints, particularly those elaborate enough to depict a deity, are rare in the Maya lowlands. Contextual recovery of such artifacts is reported from funerary and cache deposits at the Classic period centers of Copan, Honduras (Fash 1991:100, Figures 53-57) and Tikal, Guatemala. Laguna de On provides a third documented location of such an object, although the time period and its location in the surface deposits of an agrarian village are in sharp contrast to the royal burials contexts that have yielded previous examples. Additional collected pieces are reported by Schele and Miller (1986:73, 101, Plates 25, 26, though these specimens are unprovenienced) and McAnany (1995:Figure 2.11), from the Dumbarton Oaks collection. God K effigy flints from Tikal and Copan (Fash 1991) date to the Classic period. Noneffigy eccentrics from Altun Ha (Pendergast 1982) and Colha (Shafer and Hester 1983) date primarily to the Late Preclassic and Classic periods. As large, elaborate eccentrics have not been commonly reported from Postclassic investigations throughout the Maya lowlands (Shafer and Hester 1983, 1988), the Laguna de On Island eccentric is probably in secondary context. Gann (1918:68) reported finding two eccentrics cached just below the surface of Mound 4 at the Postclassic site of Santa Rita, along with Postclassic projectile points, suggesting that the use of these heirloom objects for cached offerings was not confined to Laguna de On. The Laguna eccentric is likely to have been manufactured in the Classic period, and may have been looted in the Middle Postclassic from its original context in a Classic period tomb.

Eccentric flints or flint axes often appear to be associated with lightning, rain rituals, and fertility (Taube 1992). In the Maya codices, Chac is occasionally observed holding eccentric flints which may be representative of lightning bolts or other forces associated with storms. God K ("with the ornamented nose") is a prominent deity that survived in Maya religion from the Classic period through Colonial times (Schellhas 1904:32-34, Taube 1992:69) with numerous referents in the codices. Associations of this deity are significant for interpreting the context of the Laguna de On effigy. Most common themes affiliated with God K are those of lineage (Coggins 1988, Schele 1976, Taube 1992), accession to rulership (Schele and Miller 1986, Grube 1992), and ancestor communication (Schele and Miller 1986, Love 1994). God K effigy axes are depicted in art as "manikin scepters" held by Classic period royals served as emblems of lineage (Coggins 1988). Grube (1992:211) notes that a particular form of Maya dance commemorating accession recorded on Lintels 1 and 42 at Yaxchilan involved the bestowal of the God K scepter.

The God K kawil deity is also closely associated with communication of ancestors through companion spirits or wayob, according to inscriptions read by Grube (1992:209-210). One eccentric flint illustrates a God K head which forms the back of a canoe that appears to be escorting a deceased Maya lord to the underworld (Schele and Miller 1986:286, Plate 114). Posthumous portraits of deceased kings at Palenque and Copan depict them with the God K celt piercing their foreheads, suggesting they transformed into or otherwise portrayed this deity in death (Schele and Miller 1986:73). The fact that the God K effigies that have been documented archaeologically from in funerary and cache contexts provides support for the epigraphic and iconographic association of God K with ancestors. Other noneffigy eccentrics reported from the geographically closer Belizean sites of Altun Ha (Pendergast 1982:121-124) and Colha (Sullivan 1991:34, 70) were also from funerary context.

The procurement of the Classic period effigy expresses the significance of obtaining an heirloom from the past for the purpose of using it as a cached offering. Added to its status as an object of antiquity to the Laguna de On Postclassic village, the artifact’s symbolic representation of a lightning deity with well-documented close ties to the ancestors and other rain deities in the supernatural realm make it pertinent as an offering in a building around which calendrical ceremonies may have taken place and significant ancestors were interred. The association of upper status interments, a public building, incense burning, ceramic smashing, animal sacrifice and game redistribution, and the offering of an eccentric flint provide a window of observation into religious ritualism of the Middle Postclassic. It is clear that rain or calendrical ceremonies with links to ancestor communication that have their roots in the Preclassic and Classic period were reproduced at the village scale at Laguna de On Island.

The orchestrators and prime beneficiaries of such ceremonies at Laguna de On would have been village leaders who may have been descendants of ancestors interred at Structure I through affinal or sanguinal lineage ties. Concurring interpretations have been offered for the archaeological record at Santa Rita (D. Chase 1982, 1986, 1988, D. Chase and A. Chase 1988), where evidence suggests that New Year or Uayeb rituals took place in the residential courtyards of village leaders. The existence and operational mechanisms of a village hierarchy at Laguna de On may thus be viewed at Structure I through mortuary patterns, architecture, features, and artifact composition.

Figure 10. Close-up photo of Flint God K Eccentric from Structure I.

Figure 11. Illustration of Flint God K Eccentric from Structure I (drawn by Anne Deane).

The Suboperation 12 Complex exposed Structure II in a series of sixteen adjacent units including Suboperations 12a through 12p (Figure 1). This area encompassed 77 square meters of excavation and allowed us to document and record the majority of this structure (Figure 12). The surface was visible in the form of outcropping rubble and a vertical alignment of stones along the north edge.

The working hypothesis for the function of this structure is that it represents an offeratory shrine platform. Such a platform was reported by the explorer Thomas Gann when he visited the site in 1927 and collected a number of offerings that lay upon a "pavement of stone" (Gann 1928: 53-54). As the structure is located at the pinnacle of the island, it is in a likely location for such a shrine. It is also the only discernable surface feature (Figures 12, 13) that could be described as a "pavement of stone." The offerings that Gann collected date to the Late Postclassic and Colonial era, which postdate the occupation of Laguna de On Island. Shrines are often built at important ancestral locales in northern Belize (Walker 1990, Hammond and Bobo 1994). Our investigations were undertaken with the goals of determining whether this structure was used as a shrine as well as when it had been built. In particular, we wished to determine the relationship between the Late Postclassic use of this structure and any previous ritual activity dating to the island’s occupation in the Middle Postclassic.

The excavations of Suboperation 12 consisted of three stages: 1) removing and screening the soil matrix covering the structure to collect assemblages associated with the structure’s surface in separate horizontal subunits (Deposits A-C); 2) mapping and photographing the structure after it was maximally exposed in the sixth week of the 1996 season; and 3) excavating beneath the structure ( Deposits D-F) to determine the history of its development. The soil covering this structure was excavated with trowels and the soil beneath the rubble platform was removed with shovels. All soils excavated from this structure were screened through 1/4" mesh and all cultural materials were collected and returned to our field laboratory.

Excavations revealed that Structure II consists of a large rubble platform, oriented with the long axis of the island approximately 20 degrees west of magnetic north. Structure II is a two level, stone structure with an upper platform of 4X3m and a lower patio of 9X2m (Figures 12, 13). Both levels were probably surfaced with marl plaster which has since eroded and is no longer present. Two periods of construction are observed in the upper rubble platform.

Figure 12. Map of Structure II - Suboperation 12 (illustration by Timothy Hare). The upper platform is shaded at the top of the map, and the rubble patio is shown, unshaded to the south.

Figure 13. Photograph of Structure II (Suboperation 12).

The west side of the building is framed in the rear by the upright vertical stone wall alignment that had been visible from the surface. This alignment is parallel to the front edge of the rubble platform. The west side of the structure marked by the vertical alignment appears to represent the first building episode in this location. The east half of the platform is characterized by a "pavement" which partially overlays the original structure on the east side. The pavement thus represents the second construction phase of Structure II. The lower patio that extends to the south of the upper platform was built before or at the same time as the second phase of Structure II, as it runs partially beneath the platform. It may also have been built in two episodes corresponding to the construction phases of the upper platform. Radiocarbon dates will provide a refined understanding of this sequence.

Within the sixteen horizontal excavation units of Suboperation 12, 78 individual lots were designated (Table 5). The comparison of these lots will facilitate spatial analysis among different horizontal areas of the building in ongoing artifact distribution studies. The correlation of these lots across Structure II may also be combined for analytical purposes into basic stratigraphic deposits interpreted to be contemporaneous. The basic stratigraphic deposits at Structure II are summarized below. A list of lots associated with these deposits defined for Structure II is presented in Table 6. These will provide the basis for chronological analysis of artifact assemblages at this location. A profile of an east-west cross section of Structure II (surface rubble) and underlying midden deposits is shown in Figure 14.

Deposit B) a humic topsoil layer overlying the stones comprising the lower rubble patio that extends to the south along the front of Structure II. This patio was not further excavated, and remains intact at the site. These layers contained ceramic sherds, obsidian blades, chipped stone (nonobsidian) tools and flakes, and animal bone. Ceramics date to the Middle Postclassic.

Deposit C) a humic topsoil layer of 20 cm average) found at the surface to the east and north sides of Structure II. These layers contained ceramic sherds, obsidian blades, chipped stone (nonobsidian) tools and flakes, two human phalanges, burned rock concentrations, and animal bone. Ceramics date to the Middle Postclassic.

Deposit D) humic soil within the stones of Structure II in Suboperations 12a-12d, excavated with the removal of these stones to investigate deposits beneath the building. This soil may similarly be divided into west (Deposit D1) and east (Deposit D2) halves.

Deposit E) midden like soil excavated beneath Structure II (average depth 50 cm). This underlying deposit contained lithics, ceramics, animal bone, pits of burned soil. It may represent a zone of occupational debris that predates the construction of Structure II. Ceramics are Early/Middle Postclassic.

Deposit F) a lower zone of midden like soil excavated beneath Structure II that interfaces with marl bedrock. It is of a lighter color due to intermingling of organic soil with limestone bedrock. It also contained lithics, ceramics, and animal bone. Pit-like intrusions are noted into bedrock, but they do not appear to be cultural features and are currently interpreted to represent infilling of natural bedrock anomalies with soil that formed on the island. One of these anomalies was excavated at the end of the 1996 season, and the remainder will be investigated at the beginning of the 1997 season. Ceramics are Middle Postclassic.

Information related to the function of Structure II is evasive. As previously noted, effigy ceramics thought to be from the surface of this building were collected by Gann. However, the structure had few intact artifact concentrations that directly revealed the function of the structure. Ritual areas among living cultures in Mesoamerica are often kept clean from debris, and the low amount of artifacts on the rubble platform may signify the maintenance of ritual cleanliness in this area. However, many small concentrations of highly burned rock were found around the edges of the rubble platform (Figure 12). Further concentrations of these burned rocks were found within the alignment on the west side (construction episode I) and off the edge of the building to the east. Burned rock concentrations found within the alignment of phase I and around the edges of the pavement of phase II may functionally link the two construction episodes. Little was found in these burned rock concentrations that would indicate the purpose for which they were used. Fires were probably kindled in these locations, but the features are not pit shaped and thus do not resemble domestic cooking pits as found elsewhere on the island. The nondomestic nature of these tiny burned rock concentrations also suggests a ritual use of this structure.

Around the periphery of the upper platform, ceramic and artifact concentrations were recovered as shown in Figure 12. Whole stone tools were also placed on the surface of the patio, including a whole, triangular biface with little evidence of wear (Figure 15) which may have been a cached offering on the building’s east side. Roughly manufactured expedient bifaces of a utilitarian nature were also found on the patio and obsidian blades were found around the periphery of Structure II’s upper platform. The blades exhibited very little wear, in contrast to blades found in domestic contexts at the island which were used for a variety of heavy tasks. Little or no visible macrowear on the obsidian blades from Structure II may imply their use for bloodletting or animal sacrifice, though use wear studies planned for these tools will test this hypothesis. To the north east of the structure two human phalanges were recovered as well as two catfish spines.

In comparing artifact counts, Structure II has at least 20% more ceramics than Structure I and all other areas tested (Table 1). This proportion also suggests that activities conducted at Structure II were of a specialized nature. This proportion is due to lower numbers of other types of materials such as lithic debris and faunal remains, as the actually density of ceramics encountered

Table 5. Lot Descriptions for Structure II.

Table 6. Lots correlated with stratigraphic deposits at Structure II.

Figure 14. Profile of soil beneath Structure II, shown in east-west cross section balk of southwest quadrant of Structure II. North wall of Suboperation 12c, with underlying rubble of Structure II visible on surface, brown loam midden or midden/fill below the rubble interfacing with limestone bedrock at base.

throughout excavations was not great. In fact, Structure II had less overall ceramics (5,313) than Structure I (5,828), despite the fact that over 10 square meters more were excavated at Structure II than at Structure I. When results of ceramic analysis are completed, we will evaluate differences in the type of ceramics found at Structure II compared to other areas. The majority of these sherds are from non-utilitarian censers and incense burners. Some sherds were refitted from opposite ends of the Suboperation 12 complex, as well as from the neighbouring Suboperation 8 complex.

Ceramics found from this area (within the platform, in concentrations, and in general scatters) date to the Early/Middle Postclassic. The most common ceramic type (Mock, this volume) identified thus far are Payil Red (38.71%), Zakpah Orange Red (8.93%), and Santa Unslipped (6.82%). Future planned seriation of the complete data set may refine this chronology further. The date of the ceramics found at Structure II suggests that the building was constructed and used at the time of the occupation of the island, in the Middle Postclassic. Radiocarbon dates to be submitted in the Spring of 1997 will provide refinement of this chronology. The Late Postclassic/Colonial materials collected from the surface by Gann thus appear to have been deposited on a previously built and abandoned structure at that time. There is no evidence that during the Late Postclassic or Colonial periods a shrine platform was built on Laguna de On Island. Other deposits of late period effigy ceramics found in northern Belize are similarly found on abandoned "ancestral" structures built in earlier periods (Walker 1990:399-411).

Beneath the rubble platform of Structure II, there is a deposit that may represent domestic occupation (Figure 14). A large pit of burned soil and midden debris was found beneath the rubble in the front of the building. No dedicatory offering was placed in this location, and although two-thirds of the area beneath the building was excavated to bedrock, there is no evidence that this shrine is associated with a cemetery area or even a single burial that might have influenced the location of this building as observed in Structure I. Beneath the south central area of the upper platform, a large pit of burned soil was found. A partially reconstructable vessel and midden debris were found in this pit, and it is thought to be domestic in nature and to be unassociated with the construction of the building. Artifact analysis will determine whether the brown loamy soil deposit beneath the building represents undisturbed domestic trash or whether this soil was transported to this location as fill prior to the construction of the platform. The presence of pits of burned soil and partial vessels broken in place suggests, preliminarily, that beneath the platform an area of primary (undisturbed) domestic refuse is present.

The diverse classes of artifacts and features uncovered at Structure II suggest a non-domestic, nonutilitarian function for this building. These results are consistent with Gann’s (1928) interpretation of this building as a shrine from his collection of Late Postclassic/Colonial materials on the upper platform’s surface. Four distinct temporal episodes are indicated from the 1996 excavations and Gann’s report; 1) a domestic, pre-architectural deposit of probable Middle Postclassic date, 2) an initial construction episode of the west half of the upper platform and patio of probable Middle Postclassic date, 3) a second construction episode when this west half was extended to the east by a rubble pavement and a continuation of the patio of probable Middle Postclassic date, and 4) the use of this structure as a shrine in the Late Postclassic/Colonial period(s) as reported by Gann.

Figure 15. Triangular biface (possibly cached) found at east edge of Structure II.

Suboperation 14 extended units over an area of 6.5m X 4.5m, encompassing terrestrial and aquatic segments of Structure III (Figures 16, 17). Suboperations 14a and 14b (1X2.25m each) tested the terrestrial area where the stones were partially covered by wet humic topsoil (Table 7). Suboperations 14c and 14d (3.25m X 2.25m each) tested stones of the structure that extended out into the water, outcropping above it. Soil was washed out of the stones in 14c and 14d, and tree trunks had created large holes in the structure in several places. A large mangrove tree was removed from the stones in this area prior to excavation. Suboperations 14e and 14f marked the edge of the structure in the water. Stones in this area form a U-shaped projection. Submerged lagoon sand adjacent to Structure III’s edge in Suboperations 14e and 14f was excavated by shovel testing and water screened.

High numbers of artifacts were recovered from this procedure. The artifacts must have fallen or eroded into the water in this area through canoe landing activities or from eroding midden deposits on the island’s shore. The density of artifacts found in Suboperations 14e, 14f, 14a, and 14b suggest that the Structure III vicinity was a high traffic area in which a variety of activities were conducted. Such density might be expected for a dock and major access area, where artifacts were probably broken, lost, or discarded around the landing. Materials found were not just trade items such as ceramics or obsidian, but faunal bone as well, suggesting that domestic middens are located nearby. Cranial and postcranial remains of animals were found. The materials collected from this area represent a broad range of economic activities. This sample provides a valuable contrast to the nondomestic material recovered from Suboperations 8 and 12 and the fill materials found in the lower portion of 5b, and occupational debris from the top of 5b and 5c.

Tests at Suboperation 14 suggest that this stone feature that projects into the water may have served as a canoe dock. The presence of this dock, the only such feature located to date around the island’s periphery, may suggest that access to the island was formally restricted. Considered in conjunction with the idenfication of boulder walls found around the island at Suboperations 5, 13, and 7a (Barnhart and Howard, Chapter Six, this volume), the possibility that the Laguna de On Island village was fortified is emerging. The recovery of high numbers of artifacts that represent a range of economic activities from the dock area also contributes a valuable contrastive sample to the ritual assemblages recovered from Structures I and II that will facilitate further analysis of village production activities and local and long distance trade.

Gann, in his reconnaissance of the island in 1927 (Gann 1928), noted the presence of stone docks around the periphery of the lagoon. These features are observed today at the southeast shore of the lagoon. It is not known whether they date to the historic logging camp (Honey Camp) encountered by Gann in the early decades of this century, or whether they are Precolumbian in origin. The 1997 season will initiate underwater reconnaissance in the vicinity of these features and sampling of associated submerged deposits in an effort to address this question.

The identification of these docks as Precolumbian is significant for assessing the degree to which the lagoon shore was occupied in the Postclassic. Survey on the shore has noted dispersed surface deposits of Postclassic materials that are difficult to locate in areas other than uncleared milpas. It is certain that inhabitants of the island would have cultivated rich sediments surrounding the shore of the lagoon, but the extent of settlement is unknown. Investigations of a Classic period plaza group in 1991 at the southeast shore of the lagoon did exhibit a Postclassic occupation in the surface topsoils (Masson 1993), and it is hoped that future testing will clarify the extent of this shore occupation.

Figure 16. Map of Structure III (Suboperation 14), stone dock projecting into water on east edge of island.

Figure 17. Photo of Structure III (Suboperation 14) vicinity looking east.

Table 7. Lot Descriptions at Suboperation 14.

______________________________________________________________________________

The 1996 Testing Program

Suboperations 5b, 5c, 7a, and 13 formed the testing program of the 1996 season, which is designed to explore the diversity of contexts at Laguna de On Island that will help guide further testing (Figure 18). In 1996, the project sought to identify domestic contexts, and test excavations were placed in areas to the south and north of Structures I and II to collect contrastive information from these ritual buildings. Suboperation 5b located a deep layer of fill that clarified terrace construction efforts at the island (described below and in Masson, Chapter One). At the surface of this terrace deposit, the corner of a rubble platform was uncovered which may represent a domestic structure. Suboperation 5c revealed a shallow midden deposit and a concentration of three burials in a 1X2m unit. Further work in this area will explore the extent of this apparent cemetery and search for domestic structures that may be associated with it. Suboperation 7a was placed over a boulder alignment framing a possible courtyard area in the northwest area of the island’s highest plateau. A high concentration of debris suggests that a number of activities were conducted in this area, and a cache was also found adjacent to the wall. This area will also be targeted for horizontal expansion. Suboperation 13 was placed over an additional concentration of large boulders along the west edge of the upper courtyard. Similar to Suboperation 7a, artifact density was very high. The artifacts collected from Suboperations 5b, 5c, 7a, and 13 provide valuable contrastive assemblages to the materials collected from Structures I and II, and will be analyzed with the goal of reconstructing economic systems of this island community.

Suboperations 5b and 5c were placed in an area away from the ritual precinct with the goal of gathering complementary information about domestic activities, house form, non-apex burial patterns, and status variability (Figure 1). Domestic structures were not encountered this season, although an area of deep artificial terrace fill was revealed in 5b and a possible cemetery area was found in 5c.

In the Suboperation 5 area, terrain slopes gently up from the shore in three wide, ephemeral terraces, the third beginning some thirty meters from the water's edge (Figure 1). The terrace edge closest to the shoreline slopes up almost two meters and exhibits surface indications of having once been a north-south running retaining wall. In 1991, a 2X1m excavation unit was placed parallel to and just above the wall segment (Valdez et al. 1992, Masson 1993). Excavation reached a depth of just over 1.5 meters before encountering bedrock. In 1996, the wall segment was designated Suboperation 5a and briefly investigated. A five meter section of wall was cleaned off and photographed. Ceramic sherds, lithic flakes and small quantities of faunal bone were recovered from the spaces in between the wall stones, supporting the area's identification as a cultural feature. Two other units were opened in the Suboperation 5 area; Suboperations 5b and 5c.

Figure 18. Cross section diagram of island at the vicinity of Suboperation 5 units. Variable depth of surface topsoil, fill, and bedrock on east half of island and relative position of Suboperations 5, 5b, and 5c are shown.

Figure 19. Photo and map of Suboperation 5b, south profile.

Table 8. Lot descriptions at Suboperation 5b.

Table 8. Lot descriptions at Suboperation 5b.

Table 9. Lot Correlations at Suboperation 5b.

Suboperation 5b was a 1X2 meter unit placed upslope from Suboperation 5 that was excavated in 1991. Suboperation 5 was located along the inside of a wall that marks the edge of the island on the southeast side. Artificial fill was observed in Suboperation 5, retained by large boulder walls on the terrace edge. A burial was found beneath the fill intruding into bedrock. This 1991 unit raised important issues of landscape modification. How extensive was this fill spatially, and what was the original contour of the island? Why was the terrace built in this location, perhaps to improve the surface for domestic occupation? To answer these questions, Suboperation 5b was placed between the original Suboperation 5 and the top contour of the island at this location. Suboperation 5c tested a 1X2 meter area just within the top contour of the island. It was hoped that Suboperations 5, 5b, and 5c would provide a good cross section of terrace modification, original island shape, and activities for which the terrace was used. Suboperation 5a was assigned to an area of the terrace wall that was cleaned for photography, although excavations were not conducted in this area.

The deposits of Suboperation 5b match closely those found in Suboperation 5, nine meters to the east along the shore of the island. From this unit, we learned that the original contour of the island, indicated by the depth of bedrock in these two units, is very low and only a few centimeters above water level. We also learned that the one meter of terrace fill is broadly extensive, at least 13 meters in length when the dimensions of the two units are added to the distance between them. This construction documents intensive efforts on the part of the Middle Postclassic occupants of this village in modifying and improving the shape of the island. The social significance of this labor is outlined in the introductory section of this report.

The soil layers within Suboperation 5b slope up to the north and west. Stratigraphic profiles reveal three main soil strata, each containing cultural material (Figures 18, 19, Tables 8, 9). The upper most layer, Deposit A, reaches a depth range from 25 to 75cm below ground surface. This layer includes the naturally formed organic topsoil and contains many roots in a matrix of dark brown loamy soil. The second soil layer, Deposit B, ranging in depth from 25 to 145cm below ground surface, is gray-brown in color and contains tiny chunks of marl and dispersed limestone nodules. This zone is thought to be artificial fill. Deposit C, the third and lowest layer (directly atop bedrock), first encountered at a depth of 95cm below surface, is a mottled combination of two soil types. This soil is thought to be original A horizon soil formed over bedrock. The majority of Deposit C is comprised of a soft yellow limestone, the same material as the harder bedrock below it. Interfingered and pocketed into the soft limestone lies a dark brown organic soil. The organic soil layer is roughly 20 cm in vertical width and covered the majority of the unit floor at a depth of 125-135 cm below surface.

At ground surface of Suboperation 5b, a concentration of rubble was uncovered in the southwest quadrant of the unit. Ceramic sherds, lithic flakes, animal bone fragments and a broken biface were found within the spaces between the stones. Interpreted as a possible corner of a structure close to the surface, the concentration was pedestalled and preserved through the completion of the excavation.

The artifacts in Deposit A of Suboperation 5b were plentiful and found randomly throughout the matrix. No real concentrations were observed. Bioturbation caused by roots has probably moved the artifacts from their original locations. Lithic flakes and eroded slip ceramic sherds comprised the majority of artifacts recovered. Small quantities of animal bone, shell and obsidian were also recovered. A small percentage of the lithic flakes and tiny fragments of limestone showed evidence of heat alteration. Objects of special note found in the bioturbated Deposit A include a small stone adze made of serpentine, a lithic projectile point blank, a molded ceramic spindle whorl top and a single, historic white glazed ceramic sherd, likely evidence of the British logging camp once on the island, occupied briefly around the turn of the century.

The transition between Depost A and Deposit B was marked by an increase in flecks of marl and a change to a more gray-brown color soil (Tables 8, 9). When the top of Deposit B was encountered, vertical collection strategies were changed from natural/cultural lot divisions into 20cm arbitrary levels. The depth of the unit was observably below that generally observed for natural soils on the island, and we suspected that a zone of artificial fill had been encountered. The Suboperation 5c excavation, taking place simultaneously less than 10m away, had hit bedrock at 40cm below ground surface. In contrast, just 9m to the east of Suboperation 5b, the 1991 excavations at Suboperation 5 (Masson 1993) had gone to a depth of 1.5m before encountering bedrock. An arbitrary level division of 20cm per level was thus adapted in anticipation of a deep deposit.

Artifacts in Deposit B were dispersed, consisting primarily of lithic flakes and ceramic sherds. Burned flakes increased in frequency and the first slipped ceramic sherds in the unit were recovered. The matrix of Deposit B contained tiny flecks of charcoal. A small burned rock cluster was encountered in association with a burned biface and charcoal at a depth of 80cm below ground surface. A 14C sample was collected for analysis. At a depth of 110cm below ground surface, sitting on the transition zone between Deposits B and C, a partial vessel was recovered protruding from the south wall of the unit. The fragment was a flaring rim, medial flanged, orangeware bowl. Shirley Mock, project ceramicist, confirmed the vessel's identification as an Early Classic type. Currently, this is the only ceramic artifact recovered from the island that is not of Postclassic date. A large charcoal sample was collected that was in contact with the bowl's interior for carbon dating. Postclassic sherds found at the same depth and below suggest the partial vessel was brought in as part of the fill, rather than representing evidence of Early Classic occupation. The original location of this fill is not known. As Classic period deposits have been located to date only on the shores of the lagoon, it is probable that fill was transported by canoe to the island. Further testing on the island in future years will be used to supplement current conclusions that the island deposits are only of Postclassic date. Carbon dating of soil humates and animal bone in Deposit B are also planned.

The third and deepest strata of Suboperation 5b, Deposit C (Tables 8, 9), was a mottled, dark brown organic soil with soft yellow limestone interfingering horizontally through it. The mottling is due to its mixture with limestone bedrock upon which it rests. Flakes, burned flakes, ceramic sherds and two possible flake scrapers were found within the organic soil patches. A small but significant quantity of flakes and sherds came from the soft limestone layers. The soil in this deepest strata is probably the original horizon formed above bedrock on the island.

The general character of artifacts recovered from Suboperation 5b are quite different from those found at Structures I and II. For example, few incensario ceramics were recovered. Ceramic concentrations were also not observed, and obsidian artifacts are also less frequent. In preliminary observations, current evidence indicates that artifacts from in Suboperation 5b represent domestic debris.

Suboperation 5c added further intriguing information to the construction history of the island. Located just 5.6 meters to the west (uphill) of Suboperation 5b, the depth of deposit was only 40 cm deep (Figure 20, Tables 10, 11). Bedrock was found at this depth! Thus, bedrock is 40 cm deep at the high contour at this part of the island, and is 140 cm deep 5.6 meters downslope at Suboperation 5b. The terrace that extended east to the water, documented in Suboperation 5 and 5b, was essentially raised to the elevation of the high contour found in 5c. Furthermore, a burial area was revealed in the 1X2m area of Suboperation 5c. In this unit, three burials were found; Burial 7, Burial 8, and Burial 10. The skull of Burial 10 was found on the last day of excavation, in the eastern portion of the unit, at a lower elevation than the others. It was not removed due to lack of time. Burial pits were found cut into the bedrock. Burial 7 was located in the western portion of the unit and Burial 8 was found in the center of the unit (Figure 21). During the cleaning of the south wall of the Suboperation 5c for profile drawing, the outline of a fourth pit was revealed, which may represent an additional burial.

The cranium of Burial 7 was found 45cm below ground surface, slightly crushed on the left side (Figure 22). The distal left femur laid at an angle against the back of the cranium. Bone preservation was good to moderate. The body was interred seated, knees up and arms wrapped around the legs. Its head had since slumped down between the knees. At the base of the pit, the individual's hands and feet were found in good condition along side the fragmentary pelvis. Skeletal positioning indicated the body was seated facing west. The mandible was recovered with teeth intact. Initial observation of molar characteristics identified Burial 7 as an adult, 30-40 years of age. The burial pit was clearly visible, roughly circular and marked by the sharp transition between the yellow limestone bedrock and the gray-brown soil within the pit. No artifacts were recovered in direct association with the burial.

Burial 8 was found and excavated simultaneously with Burial 7. The pit outline of Burial 8 touches that of Burial 7 to the west. Above the Burial 8 pit were placed a cluster of five unburned cobbles. Fragmentary pieces of Burial 8's skull were recovered from the perimeter of the rock cluster. Despite the destroyed condition of the skull, the rest of the skeleton, encountered directly beneath the rock cluster, was relatively articulated and in a good state of preservation.

Burial 8 was also seated facing west, with the knees up and arms wrapped around legs. The maxilla was found with teeth still intact, upside down and oriented east (opposite from the orientation of the body). The skull had apparently dropped between the legs postmortem. Molar characteristics identified the individual as an adolescent. The pelvis was fragmentary and gender has not been assigned. No artifacts were found in direct association with the burial pit.

The relationship of this tightly-packed burial area to the rubble platform in Suboperation 5b will be one focus of 1997 investigations. Additional plans are being made for geostratigraphic documentation and radiocarbon dating of the landscape modification in this area. A connecting

trench will be dug which will provide a full cross section view of the island to enable full analysis of these deposits.

Figure 20. Suboperation 5c profile of north wall.

Table 10. Lot Descriptions at Suboperation 5c.

Table 11. Lot correlations at Suboperation 5c.

Figure 21. Photo of Suboperation 5c burials.

Figure 22. Map of Suboperation 5c burials.

Figure 23. Greenstone celt (left) and groundstone bead (center) from Suboperation 5c. Also shown is a ground iron ore bead (right) from Suboperation 8 (Structure I).

The strata above the burials in Suboperation 5c, Deposit A (Tables 10, 11), from surface to an average depth of 40cm, contained a number of interesting artifacts. This dark brown organic topsoil was heavily bioturbated by roots, and overlies bedrock. A single glazed historic sherd was found, and all others resemble Middle Postclassic domestic debris. In the soil above Burial 7, two small stemmed projectile points and a thin, bifacial lanceolate blade were recovered. Between the stones in the cluster above Burial 8, two obsidian blade fragments and a piece of sea turtle shell were found. In the soil from the western half of the Suboperation 5c, a greenstone 2X1cm rectangular bead was recovered (Figure 23). Above the third as yet unexcavated burial, a ceramic tripod foot vessel fragment and a burned chert core were recovered. The association of these artifacts with the burials is unclear as the origination surface of burial pits was not defined due to the similarity of burial fill to Deposit A. Lithic flakes, ceramic sherds, faunal bone (burned and unburned) and tiny fragments of burned rock were found dispersed through Deposit A.

Suboperation 7a tested a 2X3m area adjacent to Suboperation 7, excavated in 1991 (Masson 1993). The 1991 Suboperation 7 (2X3m) tested a small earthen mound in which a scatter of human bone was recovered. This area of the island is located at the west edge of a rectangular plateau at the island’s highest elevation which may have served as a courtyard (Figure 1). Immediately to the west of Suboperation 7 and 7a, the slope of the island’s northwest side begins. Suboperation 7a was placed over some large stones that were visible on the surface in this area, to see whether they formed part of a wall that may have delineated the courtyard edge (Figure 24). Further stones are observed to the north of Suboperation 7a, which probably extend the alignment exposed in this unit. The 1997 season will continue explorations in that direction. These stones appear to mark the edge of a large, flat surface to the northwest of Structure II (Suboperation 12) which may be a courtyard at the island’s apex. Suboperation 7a was placed with the goal of exposing the courtyard-edge wall and collecting further materials for spatial comparison of activities.

The walls uncovered in 7a were comprised of large, flat boulders (Figure 24). The stone alignment within Suboperation 7a runs north-south. The two largest stones, crumbling at their edges, are flat and over 1 meter wide. The 20cm organic topsoil layer surrounding the stones, Deposit A, was excavated in two separate lots, one on the east side, the other on the west (Tables 12, 13). These collection units will facilitate comparisons among artifact assemblages and activities in this location. An eroded plaster surface, Deposit B, was encountered on both sides of the wall (Tables 12, 13). This surface may represent the top of bedrock, but it appears to have been exposed and eroded, perhaps doubling as a living surface. A high concentration of artifacts were found in this area, particularly on the east (or courtyard interior) side of the wall. A pit was found in the northeast corner of the unit, in which a lenticular biface (made of chalcedony) and Pomacea shells were placed as a cached offerings or part of a cooking pit (Figure 25).

Artifact density within the top soil layer was high, and hundreds of flakes, burned flakes and ceramic sherds were recovered from both sides of the stone alignment, as well as small quantities

faunal bone (burned and unburned), obsidian fragments and shell. On the west side of the stone alignment, towards the center of the unit, a concentration of artifacts was discovered directly

Figure 24. Map of Suboperation 7a (right),

showing plan perspective with north to the

top of the page (scale 1:20cm). Note Pomacea

cache pit in northwest corner of unit within

depression indicated. Photo of Suboperation 7a is

shown below, facing south.

Table 12. Lot Descriptions at Suboperation 7a.

Table 13. Lot Correlations at Suboperation 7a.

Figure 25. Suboperation 7a Pomacea cache pit.

besideand below the stones. Fragments of a single, red slipped vessel, four obsidian blade fragments and two Pomacea shells were found in association, mixed with ceramic sherds and lithic flakes. This concentration is similar to that of small ritual offerings found at Structures I and II.

The northeast corner of Suboperation 7a was identified early as a separate feature. A patch of dark brown soil, approximately 50cm in diameter, was identified in the topsoil. Artifacts included a high density of faunal bone (a large mammalian mandible among them), a solid ceramic tripod vessel leg, high numbers of burned flakes, ceramic sherds, obsidian blade fragments and ceramic fishing weights. As this northeast corner was excavated, it became clear that it was a pit that intruded beneath the limestone surface that covers all other areas of the unit. At the base of this pit, two levels of Pomacea shells were recorded to a depth of 60cm below ground surface. The pit walls and base were lined with stones.

The upper Pomacea shell concentration was found at 30cm below ground surface in association with four burned rocks, a fragmentary red slipped plate and a well made lenticular biface, 14cm in length. The three shells were tightly clustered in the center of the pit. The lower Pomacea shell concentration was found 25cm lower than the upper concentration, at 55cm below ground surface. The shells were not tightly clustered, as in the upper layer, but were found in greater quantity. A total of seven Pomacea shells were found in this layer in association with fragments of a single, red slipped vessel. Two large charcoal samples were collected from this level.

Further horizontal exposure of this area planned for 1997 will hopefully reveal associations of this pit of concentrated materials and heaveily littered activity area and permit further interpretation. The 1996 unit accomplished its goals of collecting comparative lithic, ceramic, and faunal material. Documenting this segment of the courtyard wall also provides further testimony to the significant efforts in construction at the island and the organization of public and private space at this community.

Suboperation 13 was excavated to recover comparative information with Suboperation 7a, as it is located in the same general vicinity on the island. Suboperation 13 was placed 10.5 meters to the west of Suboperation 12 and 10.2m south of Suboperation 7a, along the edge of the same courtyard space tested by Suboperation 7 (Figure 1). Suboperation 13a was a 1X2m trench placed in front of an extremely large semi-upright boulder along the courtyard edge (Figure 26). Working hypotheses for the function of these large stones included their use as part of the courtyard wall or perhaps as an altar area. An additional large boulder was visible on the surface of Suboperation 13a as well. Suboperation 13b was a 2X3m trench that encompassed the area around the boulder.

Stratigraphy around the stones was the same found at Suboperation 7a a shallow topsoil layer (Deposit A) lying directly atop a cracked, crumbling limestone surface (Deposit B, Tables 14, 15). In the south portion of Suboperation 13a, the edge of a structure or wall was found in the form of a rubble surface extending into the south and east walls (Figure 26). The north side of the boulders in both units revealed an eroded marl limestone surface (Figure 26) similar to that observed in Suboperation 7a.

Artifact densities were extremely high. Lithic flakes and ceramic sherds were especially abundant compared to other areas tested at the site. Hundreds of large flakes were collected. Other artifact types found in Suboperation 13 topsoil include obsidian blade fragments, fishing weights and small fragments of faunal bone. This material and associated ceramics will be used in analysis to evaluate the use of this area when the island was occupied. It is possible that the large stones are part of the courtyard wall, but they appear to have been dislodged by tree root activity.

Two areas of concentrated artifacts were found in Suboperation 13. On the north side of the large stones, directly at their base, whole Pomacea shells were recovered along with obsidian blade fragments and an anthropomorphic tripod vessel leg. The leg was hollow and molded in the form of a human face. The second concentration, 20cm west of the stones, was first identified as a patch of dark brown organic soil. Unburned fragments of a red slipped plate were recovered from within the patch. The objects were found in loose, root-disturbed soil. Further analysis of these assemblages will examine the nature of activities performed in this area.

The testing program of the 1996 season at Laguna de On succeeded in sampling a four locations at the site. Suboperation 5b and 5c expanded 1991 investigations of the Suboperation 5 vicinity on the southeastern slope of the island, identifying terrace construction activity and a burial area. Further work in this area is planned that will provide context for the burial area and continue the project’s search for domestic structures and midden sampling. The other two locations tested, Suboperations 7a and 13, provide a glimpse of the west edge of a possible courtyard area along the northwestern slope of the island. These tests suggest that this flat plateau may be bordered by a large boulder wall that meets an eroded plaster deposit, currently thought to be bedrock that may have served as an occupational surface. Atop this deposit was recovered evidence of intensive activities that appear both economic and ritual in nature. Further testing will expand these test pits to horizontal units to aid the interpretation of village domestic and ritual activities at the site.

Figure 26. Map of Suboperation 13 (right)

showing plan perspective with north to the

top of the page. A photo of Suboperation 13 is

shown below, facing south.

Table 14. Lot Descriptions at Suboperation 13.

Table 15. Lot Correlations at Suboperation 13.

Ceramics remain a significant portion of the mysterious puzzle of Maya Postclassic settlement economies, social organization, ideology, and religion. This is due, in part, to the fact that the nuances of this time period are not understood. Past attempts to separate ceramics into discrete temporal categories have often been confounded by the admixture of Terminal Classic-Postclassic occupations at many sites. Such methodological difficulties in Belize have been encountered in shallow deposits at Colha (Valdez 1987) and Northern River Lagoon (Mock 1994). The type-variety format itself in many ways, intended to be a valuable tool for comparative analysis, imposes categories that obfuscate our understanding of Postclassic processes. Moreover, sherd surfaces are often eroded, making type assignments difficult.

Previous investigations at the island site of Laguna de On in 1991 began to address some of these questions. Evidence from ceramic data suggested that a small community grew up on the island during the Middle Postclassic (AD 1100-1350). During the Late Postclassic (AD 1350-1500), a shrine was established, becoming the depository of offerings perhaps related to post-abandonment pilgrimages (Masson 1993:46). Ceramics and lithics recovered from initial tests on the island provided the basis for these preliminary interpretations (Masson 1993; Valdez and Masson 1994), although ceramic chronology was not supplemented by absolute dating.

The 1996 field season at Laguna de On continued to address these questions through more extensive excavations and detailed attention to horizontal and vertical variability among the assemblages. Detecting differential representations and the presence or absence of specific ceramics within the excavated portions of the site became crucial not only to illuminate site chronology but to facilitate religious and functional designations of structures on the island.

Approximately 50% of the Laguna de On ceramics were analyzed during my examination in the 1996 field season. Following procedures established at Colha (Valdez 1987) and Northern River Lagoon (Mock 1994), the sherds were sorted into groups according to surface finish or slip color. The groups were then divided into tentative types using the type-variety format and modal form analysis. Established types from Colha (Valdez 1987), Cerros (Walker 1990), Santa Rita (D. Chase and A. Chase 1988), and Lamanai (Graham 1987) were used as indices of classification. Preliminary descriptions of these provisional classifications are provided below.

Originally placed in the Augustine Red Group (Mock 1994), these ceramics date to the Early Postclassic at Colha where they are represented by Zakpah Orange-Red. Forms include bowls, flanged dishes, and large chalices (Mock 1994; Valdez 1987; Valdez et al. 1994). Rims are often direct or slightly padded with rounded or pointed lips. Slip is a reddish orange with frequent fire clouding grading into a golden green (Mock 1994). Paste varies, some unoxidized with coarse grained with large calcite particles. Some sherds displaying surface attributes characteristic of the type have a finer sandy-coral colored paste with fewer black cores. When eroded, Zakpah Orange-Red typically displays more calcite leaching than Payil Red (discussed below)

The small sample of sherds from Laguna de On appear to represent vessel forms such as collared jars or bowls and chalices with direct or slightly padded rims. Based on my own observations of the Northern River Lagoon ceramic collection, I agree with Walker’s assessment (1991) that this group represents a transition from the Terminal Classic Kik Group.

One ceramic type belonging to the Zakpah Group that is rare in the ceramic collection is

identical to Centon Incised: Variety Unspecified at Colha (Valdez 1987:220; Walker 1990). It is typified by grater bowls with interior incised designs. Presumably this rarity is in part due to the non-domestic nature of the structure it was recovered in at Laguna de On.

A tentative type, Zakpah Variety Incised: Variety Unspecified, is being used at this stage of the analysis. The limited number of sherds does not permit placement in either the Zakpah Incised or Zakpah Gouged-Incised types designated at Cerros (Walker 1990). The identifying attributes of both either overlap or are not fully represented in the Laguna de On collection. Another provisional "working" type is Zakpah Orange-Red Composite: Variety Unspecified. It is characterized by applique bands with punctation and incisions on what appears to be Zakpah Orange-red paste.

The slipped ceramics at Laguna de On are dominated by the Payil Red Group and appear identical to the red slipped wares of Tulum and Ichpaatun (Pendergast 1981; Sanders 1960; Smith 1971:30) in Yucatan, and those of Lamanai, in Belize (Graham 1987). The type variety Payil Red is placed in the Middle Postclassic, Early Facet Kanan, complex at Cerros (Walker 1990:86; see also Ball 1978:208). At Colha, Payil Red is placed in the Middle Postclassic Canos Complex, the last functionally complete complex (Valdez 1993:13) at the site. Both plain and incised versions of this ceramic type occur in surface deposits mixed with Terminal Classic ceramics (Gabourel Complex) at the Northern River Lagoon site in Belize (Mock 1994).

Payil Red (Mock 1994) is typically characterized by an opaque red slip and thin walls. The slip has a slightly waxy feel. Paste, ranging from buff to light pink-coral is more oxidized and shows finer calcite intrusions than Zakpah Red. Eroded sherds have a powdery surface.

Forms include basal-break flat-bottom dishes and sag bottom vessels with hollow tripod feet. Vessel forms generally are more diminutive in contrast to Terminal Classic-Early Postclassic vessels. I have observed that there are similarities in manufacturing technology between the Payil Red group and the Late Classic Tinaja Red Group (Mock 1994). Modal continuities also exist in the bowl and jar forms.

Payil Red: Palmul Incised, occurring rarely in the Laguna de On ceramics is characterized by shallow post-slip incising. Redundant elaborated scroll patterns, band panels, and pendant hooks suggesting abbreviated, conflated serpent and bird motifs are a notable characteristic (Mock 1994). Due to the small sherds in the Laguna de On collection designs are difficult to determine.

However, these attributes are present in attenuated form, and appear to be crudely executed. A few segmented flanges present in the collection display the characteristic Payil Red slip.

Payil Red: Punctated Applique is another provisional type assigned on the basis of one sherd. Punctated decoration is infrequent in the Postclassic, in contrast to the Late-Terminal Classic. Again we may have local potters incorporating Classic ceramic traditions into a new, later ceramic type.

Payil Red-Striated, another infrequently occurring provisional type, is characterized by its thin walls and fine calcite paste. Forms are ollas or small jars. Striations below below the juncture of neck and body. This type shares characteristics with Red Neck Mother Striated: Variety Unspecified, Chambel Group at Cerros (Walker 1990:75)

A limited number of sherds from Laguna de On have been tentatively placed in the Rita Red Group (see Chase 1982; Walker 1990:88-90). Represented by unique parenthesis neck tinajas at Laguna de On, this type is dated from the Middle Postclassic to the Late Postclassic (see Sanders 1960:Figure 5c). Since the sherds display the characteristic Payil Red slip, this assignment is pending, however, until the paste is examined petrographically. It is possible that the sherds represent a transitional stage or type between the Middle and Late Postclassic. The jars may have been manufactured at Santa Rita and brought to Laguna de On during pilgrimage visits (e.g., Walker 1990:89).

Initial sorting was easily accomplished due to the unslipped, smoothed, pinkish buff surface with the pock-marked finish peculiar to this group. One noted characteristic at Laguna de On is the exterior thickened and folded rim and round lip. A second sorting divided the sherds tentatively into Cehac-Hunacti Composite:Variety Unspecified and Thul Applique (see also Ak Applique, Valdez 1994) of the Navula Unslipped group Because of the lack of complete vessels and small rim sherds, many of the assignments to either type-variety were arbitrary. Both types were recovered in all deposits at Laguna de On.

Cehac-Hunacti as a type was established at Mayapan by Smith (1971) and specific sherds indicate that it is represented at Laguna de On by a similar globular jar censer form with pedestal base and averted neck. Vessels are decorated with appliqued, punctated bands and circular motifs, and occasionally twisted guilloches (see Smith 1971:95). The paste is heavily tempered with large calcite flecks.

Many of the sherds could also be placed in the Cohokum Group on the basis of distinct paste and surface characteristics (see Cerros [Early Facet Kanan], Walker 1990:100, 102); see also Chase 1982: 541). For example, some appear identical to Santa Unslipped:Taman Variety of the Cohokum Group at Cerros (Walker 1990:100), also characterized by pock-marked surfaces. Walker (1990:102) also notes that Taman Variety shares attributes with the Navula Ceramic Group. Analogous forms occur at Middle Postclassic Mayapan (Smith 1971:94-95).

One new provisional type represented in limited quantities in the collection is Santa Unslipped: Striated. Sample is too small to determine whether they represent jars or bowls. Cerros has Santa Unslipped bowls and ollas in the Early Facet Kanan continuing into the Late Facet Kanan (Walker 1990:100-102). These two groups need to be studied further. It is entirely possible, that the two overlap temporally and what we see are transitions in pottery traditions with residual attributes. Bowl or olla forms used as censers may have evolved to more formal censer forms.

This unslipped group is represented by two provisional type varieties (see also Walker 1990:94-96). Provisional Type I has a dark grey-reddish brown surface and a sandy, flaky paste with dark cores and large inclusions. Similar to Walker’s (1990:91), Tsabak Unslipped: Tsabak Variety, and Valdez’s (1987:220) More Force Unslipped, there are visible layers of silicate arrangements. Forms are short-necked ollas or bowls with averted necks and direct rims. Rim sherds are thick.

Provisional Type II is represented by similar forms but is characterized by a micaceous paste (see Walker 1990:93). At Colha this type is represented by Maskall Unslipped (Adams and Valdez 1979: 37;Valdez 1987). These variety differences suggest the existence of two separate production locales. It is possible that the presence of the two varieties relates to desired functional differences. Some of the vessels may have functioned as a type of censer.

At Colha the Late Postclassic is represented by Chen Mul Modeled ceramic type (Ball 1978: Smith 1971), found only at one elite complex (Operation 2012) presumed to be the focus of ancestor rituals and pilgrimage visits (Valdez 1987:231). Chen Mul Modelled is widespread in the Postclassic, typified by a large, modeled effigy figure with attached bucket. It is similar to the full figure effigy vessels at Mayapan (Smith 1971:102-103. A similar type is also found at Lamanai (Pendergast 1981: Fig. 27, 1985). An analogous form at Santa Rita (Chase 1982:537) and Cerros (Walker 1990:108) is Kol Modeled: Kol Variety of the Cohokum Group.

At Laguna de On this crude censer ware is temporally placed in this group until further analysis and observations of comparative collections. Regardless it is represented by many sherds in mixed Middle Postclassic-Late Postclassic strata (see also Ball 1978). By matching sherds from different lots and levels, one almost complete vessel was reconstructed. Until depositional and contextual factors are examined however, we cannot assume this temporal continuity

I have separated the ceramics into two provisional types on the basis of surface and paste characteristics. Provisional Type I, including the reconstructed vessel, is characterized by a distinct yellowish-red, sandy paste with large calcite and hematite inclusions. Surfaces are reddish brown and cores are typically dark gray. Surfaces are crudely smoothed. Modeled appliques are attached.

Provisional Type II is characterized by a gray paste and gray surface. These differences are presumably due to the clay sources used by different production locales.

Several implications may be derived from this preliminary analysis of the Laguna de On ceramics. First, observation of forms represented suggest that the assemblage is functionally incomplete. There is an obvious rarity of or absence of functional types such as plates and comales in the collection, supporting previous interpretations that Structures I and II were ritualistic settings. Of particular significance are the large numbers of sherds representing special-occasion vessels such as the large red-slipped chalices or effigy censers and ollas or bowls.

A second observation is chronological. In contrast to the sites of Lamanai (Pendergast 1981) and Cerros (Walker 1990), the Postclassic ceramics from Laguna de On include no obvious Terminal Classic-Early Postclassic vessel forms. Terminal Classic indices such as incised and trickle slate wares are notably absent. However, as at other sites (Graham 1987), the Postclassic ceramics from Laguna de On do show continuities in certain Late-Terminal Classic attributes such as red-slipped types, gouged-incised, pedestal bowls, censers, and flaring neck jars.

In general, Postclassic ceramics are unlike those of the Late-Terminal Classic in that they are characterized by a more limited repertoire of designs and forms. This change may relate to a decrease in the importance that individuals and groups ascribed to local social boundaries. The most common slipped ceramic type is Payil Red, referred to elsewhere as Augustine Red or Tulum Red (Pendergast 1981, Graham 1987, Mock 1994, Valdez et al. 1994, Valdez 1994).

A third point observation from my analysis to date concerns distinctions in Postclassic slipped traditions. I agree with Valdez (1994) that the Middle Postclassic forms (Payil Red) continue traditions developed and established in the Early Postclassic. However, based on my preliminary observations of the more diagnostic sherds of the Laguna de On sample I am hesitant to pose clear distinctions in surface treatment as I did at Colha (Mock 1994). Graham (1987) for example, describes the "orange-red slip" of Buk Phase Middle Postclassic ceramics at Lamanai as similar to the "lustrous orange-red slip" of the Augustine Red at Tipu. At Colha, Valdez places the Zakpah Orange-Red in the Early Postclassic (1987), while Walker assigns it to the Middle Postclassic (1990:41).

My observations of the sherds indicate that it can be very difficult to separate out the surface color and incising treatment into two temporal types. Until further comparisons are made I would say that the two overlap in surface treatment. In fact, attributes characteristic of both types may be represented in single vessel.

In addition, neither slip color nor presence of post-slip or preslip incising are a consistent index of classification in the Laguna de On collection. Ceramics with the darker matte-red slip characteristic of Payil Red Group also display deep-gouged designs and the orange-red, fire-clouded surfaces show the shallow post-slip incising characteristic of the Zakpah Orange-red Group (as described in Walker 1990). There are also inconsistencies in pastes and wall thicknesses between the two groups. The latter may simply be related to the portion of the vessel the sherd represents. For instance, a bowl may be thin-walled while the base may be thicker to sustain the weight. Local production by different potters may also be responsible for these differences. If the bowl and base of the large red-slipped chalice vessels were formed separately by different potters it is also reasonable that the vessels may display individual preferences. Admittedly, these observations are preliminary based on the small numbers of incised sherds, and the lack of complete specimens. However, these discrepancies are enough to raise significant issues about distinguishing these two ceramic groups.

Such overlap also occurs in groups and types other than the redwares that I have described in this paper. The main censer type, Chen Mul Modeled, also crosses paste groups and is present in the Middle Postclassic deposits. This observation supports the notion that local production areas were using different clay sources. If Laguna de On is being visited in pilgrimages, one might also expect to find Chen Mul Modeled censers of different pastes from various locations.

These differences may be more significant than the similarities for interpreting behavioral patterns. As Sinopoli (1991:165) points out, the presence of "functionally analogous vessels of distinct materials" is not unique to our modern world. She notes further that "factors of cost, availability, the ‘prestige value’ of goods, taste (of food or in styles), and tradition are all likely to play a role in the choices consumers make among alternate vessel materials."

Until analysis of the entire collection of ceramics is complete I am hesitant to place specific ceramic types in a time period within the Postclassic. Completion of analysis will involve the comparison of these ceramics to other collections. However, for the purposes of promoting discourse on Postclassic chronology as the ceramic analysis and radiocarbon chronologies are established for this site, a tentative assignment can be made according to comparisons to other sites. Based on these preliminary observations of the ceramics, I propose an early facet Middle Postclassic to Late Postclassic occupation of the site. Contextual recovery of ceramics combined with future chemical characterization and mineral petrographic studies of the ceramics from Laguna de On offer some potential in understanding these technological and stylistic overlaps. Radiocarbon dates are needed before the ceramic chronology proceeds much further. From this converging data we can draw significant inferences about the Postclassic period at Laguna and throughout northern Belize.

Figure 27. Decorated ceramics from Laguna de On Island, 1996: Payil Red (top left), Payil Red: Palmul Incised (top right, center, below left, below right).

Figure 28. Ceramic types from Laguna de On Island, 1996 season. Top left is a folded rim vessel from the Navula Group, the top three sherds on the right are identified as Thul Applique, the two effigy vessel supports on the bottom left are Zakpah. All other specimens on the bottom are Payil sag-bottom vessels or vessel supports, with one flange fragment of a Payil chalice form (foreground near scale, bottom right).

Table 16. Ceramic Frequencies from Laguna de On Island 1996 Excavations.

This report represents a preliminary analysis of the faunal remains from the 1996 excavations at the Postclassic Maya site of Laguna de On Island, Belize. Prior faunal studies have been completed at this site by Masson (1993, 1995). This previous study focused on bone collected in 1991 from excavations at the Classic period shore as well as the Postclassic period island at Honey Camp Lagoon (Laguna de On). This analysis provides additional information on Postclassic faunal remains from the 1996 field season which extended previous excavations on the island. A portion of faunal bone samples were analyzed from Suboperations 5, 8, 12, 13, and 14.

This report contains preliminary results of the portion of faunal remains from Laguna analyzed as of December 1996, an estimated 50% of last year’s sample. The remainder is being analyzed for my master’s thesis at SUNY-Albany during the Spring of 1997. This analysis focused primarily on Suboperation 8 in order to verify patterns of faunal distribution detected in the 1991 analysis and observed in the field in 1996.