Figure 4: Bivariate plots of chemical and ANOVA data of the Paxcamán, Trapeche, and Fulano ceramic groups (left) and the Augustine ceramic group (right). Ellipses represent 90% confidence intervals and red squares indi-cate statistical significance.

Chromium (log base-10 ppm)

Van

adiu

m (l

og

bas

e-10

pp

m)

Early Group 1

Early Group 2

Rim Diameters Sherd Thickness

Early Group 1 vs. Early Group 2 .29 Early Group 1 vs. Early Group 2 .32

Level of significance = p < .05

Early 1 Early 2 Augustine 1 .88 .48 Augustine 2 .46 .05 All Petén Augustine .65 .03 Macanché Island Trapeche .005 .33 Macanché Island Fulano .16 .52 All Petén Early Paxcamán .26 .46 All Petén Late Paxcamán .56 .06 All Petén Topoxté .25 .06

Early 1 Early 2 Augustine 1 .07 .02 Augustine 2 .51 .08 All Petén Augustine .30 .02 Macanché Island Trapeche .23 .01 Macanché Island Fulano .17 .04 All Petén Early Paxcamán .94 .15 All Petén Late Paxcamán .10 .0008 All Petén Topoxté .03 .0003

Chromium (log base-10 ppm)

Cer

ium

(lo

g b

ase-

10 p

pm

) Augustine 1

Augustine 2

Augustine 3

Augustine 1

Augustine 2 Augustine 3# All Petén Augustine .80 .37 Early Group 1 .88 .48 Early Group 2 .24 .05 Macanché Island Trapeche .001 .0005 Macanché Island Fulano .16 .07 All Petén Early Paxcamán .09 .07 All Petén Late Paxcamán .19 .05 All Petén Topoxté .10 .005

Level of significance = p < .05

Rim Diameters *

Sherd Thickness

Augustine 2 vs. Augustine 3 .54

Augustine 1 vs. Augustine 2 .13

Augustine 1 vs. Augustine 3 .83

Augustine 2 vs. Augustine 3 .36

*Group 1 did not have rims. #Group 3 only has two samples.

Augustine 2 Augustine 3 All Petén Augustine .11 .72 .35 Early Group 1 .07 .51 .28 Early Group 2 .02 .08 .12 Macanché Island Trapeche .12 .60 .36 Macanché Island Fulano .53 .31 .86 All Petén Early Paxcamán .03 .34 .13 All Petén Late Paxcamán .09 .35 .30 All Petén Topoxté .30 .07 .65

Do Form Measurements Have A Place In Archaeometry?Conjoining Diversity Indices with Chemical Analyses of Pottery from Petén, Guatemala

Leslie G. Cecil and Michael D. Glascock (University of Missouri Research Reactor Center)

Tipuj

Topoxté Island Zacpetén

Ixlú

Nojpetén

Ch’ich’ Lake Petén Itzá

Lake Macanché

Macanché Island

Lake SalpeténLake Yaxhá

↑N

40 km

Tayasal

Mayapán

Chich’en Itzá

Figure 1: Postclassic Archaeological Sites in Central Petén

Methodology

While conducting fieldwork in Guatemala from 1995-1999, Cecil collected rim diameter and sherd thickness measurements for Post-classic slipped tripod plates excavated by Proyecto Maya Colonial from the archaeologi-cal sites of Ch’ich’, Ixlú, and Zacpetén. Cecil also collected the same measurements for Postclassic slipped pottery from Tipuj (curated at Southern Illinois University Carbondale), from Macanché Island (curated at the Florida Museum of Natural History), and from Topoxté Island (curated at the Peabody Museum of Ar-chaeology and Ethnology). Within-group and between-group variability of the Postclassic slipped plate rim diameters and thicknesses were calculated using one-way ANOVA statis-tics. Possible rim diameter and sherd thick-ness variability of the chemical compositional groups was tested and then compared to other ceramic group data. Sample sizes are listed in Table 1. The null hypothesis is that the rim diameters and the sherd thicknesses were the same regardless of chemical or ceramic group.

Trapeche Ceramic Group Sherds

Pink Slips Double Slip Technology Tan to Light Gray Pastes Mostly Calcite Temper Monochrome Slip, Black Line Decoration, Red Line Decoration

Fulano Ceramic Group Sherds

Black Slips Double Slip Technology on most Light to Dark Gray Pastes Mostly Calcite Temper Monochrome Slip, Black Line Decoration

Augustine Ceramic Group Sherds

Red to Red-Orange Slips Orange-Red Colored Pastes Some Volcanic Ash Temper, Some Quartz Temper, Mainly Calcite Tempered Monochrome Slip, Black Line Decoration, Red-and-Black Line Decoration, Incising

Early Paxcamán Ceramic Group Sherds

Red Slips Black Rims (painted and/or fireclouded) Tan to Light Gray Pastes Some Volcanic Ash Temper, Mostly Calcite and Shell Inclusions Monochrome Slip, Black Line Decoration, Incising

Figure 2: Early Postclassic Ceramic Group Characteristics

Paxcamán (later) Ceramic Group Sherds

Red Slips Light to Dark Gray Paste Tempered with calcite, shells, quartz, biotite, feldspar, and hematite Monochrome Slip, Black Line Decoration, Red Line Decoration Red-and-Black Line Decoration, Incising

Topoxté Ceramic Group Sherds

Red Slips Creamy-white to Light Gray Pastes Tempered with calcite, quartz, biotite, chalcedony, and chert Monochrome Slip, Black Line Decoration, Red Line Decoration Red-and-Black Line Decoration, Incising

Figure 3: Late Postclassic Ceramic Group Characteristics

Number of Samples Rim Diameter

Number of Samples Sherd Thickness

Early Group 1 7 7 Early Group 2 6 6 Augustine 1 0 4 Augustine 2 16 17 Augustine 3 2 3 Late Group 1 15 14 Late Group 2 16 16 Late Group 3 5 5 Topoxté Group 1 14 14 Topoxté Group 2 8 8 All Petén Early Paxcamán 46 46 Macanché Island Trapeche 31 31 Macanché Island Fulano 6 6 All Petén Augustine 32 32 All Petén Late Paxcamán 205 205 All Petén Topoxté 58 58 Macanché Island Late Paxcamán 32 32 Topoxté Island Topoxté 12 12 Macanché Island Topoxté 10 10

Table1: Sample Sizes for One-Way Analysis of Variance (ANOVA)

Group 1 Group 2 Group 3 Macanché Island Late Paxcamán .005 .007 .05 All Petén Late Paxcamán .64 .12 .27 Topoxté Group 1 .61 .72 .91 Topoxté Group 2 .71 .75 .91 All Petén Topoxté .71 .32 .55

Chromium (log base-10 ppm)

Cer

ium

(lo

g b

ase-

10 p

pm

)

Late Group 1

Late Group 2 Late Group 3

Rim Diameters

Sherd Thickness

Late Group 1 vs. Late Group 2 .46

Late Group 1 vs. Late Group 2 .30

Late Group 1 vs. Late Group 3 .62

Late Group 2 vs. Late Group 3 .91

Late Group 1 vs. Late Group 3 .83

Late Group 2 vs. Late Group 3 .98

Level of significance = p < .05

Group 1 Group 2 Group 3 Macanché Island Late Paxcamán .10 .95 .17 All Petén Late Paxcamán .46 .28 .41 Topoxté Group 1 .02 .21 .16 Topoxté Group 2 .11 .46 .39 All Petén Topoxté .11 .86 .18

Chromium (log base-10 ppm)

Scan

diu

m (l

og

bas

e-10

pp

m)

Topoxté Group 1

Topoxté Group 2

Level of significance = p < .05

Figure 5: Bivariate plots of chemical and ANOVA data of the Late Postclassic Paxcamán ceramic group (left) and the Topoxté ceramic group (right). Ellipses represent 90% confidence intervals and red squares indicate statistical significance.

Group 1 vs. Group 2 .05

Rim Diameters

Group 1 vs. Group 2 .97

Group 1 Group 2 All Petén Topoxté .56 .69 Topoxté Island Topoxté .16 .30 Macanché Island Topoxté .92 .96 Late Group1 .61 .71 Late Group 2 .72 .75 Late Group 3 .91 .91 All Petén Late Paxcamán .23 .39

Sherd Thickness

Group 1 Group 2 All Petén Topoxté .05 .28 Topoxté Island Topoxté .12 .35 Macanché Island Topoxté .005 .05 Late Group1 .02 .11 Late Group 2 .21 .46 Late Group 3 .16 All Petén Late Paxcamán .001 .05

.39

Conclusions

Archaeometric analyses of pottery manufacture and exchange typically focus on the composi-tion of sherd pastes. While this focus is essential for understanding provenance and ex-change patterns, it cannot always address the more socioeconomic concerns of variation within ceramic traditions and degree of quality control during the manufacturing process. Variation within ceramic traditions also can be interpreted through diversity indices that in-clude rim diameter and wall height measurements. This study demonstrates the value of combining descriptive information (form measurements) with trace chemical data to better un-derstand the changes in Postclassic pottery manufacture (Early vs. Late) and the correspond-ing socioeconomic milieu. Early Postclassic Paxcamán, Trapeche, and Fulano ceramic group pottery separate into two distinct chemical compositional groups. In addition to chemical differences, the majority of the samples in Early Group 1 are from Ch’ich’ and Ixlú while those from Early Group 2 are pri-marily from Zacpetén. This may suggest that manufacture of the earlier Postclassic pottery (especially the Trapeche and Fulano ceramic groups) was near Ch’ich’ and/or at Ixlú. There is also a manufacturing center at Macanché Island as is indicated by the statistical variability of sherds regardless of time period (these samples have yet to be analyzed by INAA). The Augustine ceramic group sherds divide into three chemical composition groups. The majority of sherds in Augustine 2 are from the archaeological site of Tipuj. Because of the dif-ferences in slips (double slipped) and the criterion of abundance, it is possible to suggest that Tipuj is one location of Augustine production. Augustine 3 represents a second possible loca-tion for manufacture at or near Ch’ich’ as clay samples from the lake shore near Ch’ich’ “match” samples from that archaeological site (as well as other sites) (Cecil 2007). In general (although not always statistically proven), pottery of the Augustine ceramic group is larger and thicker than most other central Petén Postclassic pottery. Early Postclassic plates are more variable (larger and thicker) in diameter and thickness than those of the Late Postclassic period. While the variability decreases during the Late Postclassic period, Paxcamán ceramic group pottery could have been manufactured from any number of lake clays in the region. Three paste chemical groups occur that correspond to various decorative modes. Late Group 1 is decorated with red-and-black decoration or black-line decoration and the sherds were excavated from Ixlú and Tipuj. Late Group 2 primarily consists of undecorated sherds from Ch’ich’ and incised sherds from Ixlú, Tipuj, and Ch’ich’. Finally, Late Group 3 contains pottery from Zacpetén that is decorated with red-and-black or red-only motifs as well as minor amounts of undecorated pottery from Ixlú and Ch’ich’. Topoxté ceramic group pottery can also be divided into two chemical groups (the third, ash paste group is not included). Topoxté Group 1 is decorated with red-on-paste decoration and Topoxté Group 2 is decorated with red-and-black or black only decoration. Only Zacpetén and Tipuj had Topoxté ceramic group plates. This pottery was made by the Kowoj only at Topoxté Island and exported to other sites in the Petén lakes region (Cecil 2007; Rice 1979) and in some instances carried to Mayapán (Cecil, personal observation). Therefore, the close simi-larities of the rim diameters of Topoxté Group 1 and Topoxté Group 2 (p=.97) is not unusual. What is interesting is that the sherd thicknesses are statistically different between the two groups and that those Topoxté ceramic group plates excavated from Macanché Island were statistically-significantly thicker than plates being traded to Zacpetén and Tipuj. This may suggest that different thickness of plates were being manufactured depending on the decora-tive program and differential trade patterns among the Kowoj.

Acknowledgements

The INAA research and operating support for the MURR Archaeometry Laboratory was provided by a grant from the National Science Foundation (BCS- 0504015). This research was also supported in part by a grant from the US Department of Energy Office of Nuclear Energy, Science and Technology Award No. DE-FG07-03ID14531 to the Midwest Nuclear Science and Engineering Consortium under the Innovations in Nuclear Infrastructure and Education program. The archaeological research was supported by funding from Proyecto Maya Colonial of Southern Illinois University Carbondale, from the National Science Foundation (Dissertation Improvement Grant SBR-9816325), and from Sigma Xi Grants-In-Aid of Research. I would also like to thank the Florida Museum of Natural History and the Peabody Museum of Archaeology and Eth-nology for allowing me access to their Postclassic pottery collection and the Instituto de Antropología e Historia (IDEAH) of Guatemala for allowing the export of the pottery sample to the United States for study.

References CitedCecil, Leslie G. 2007 Postclassic Maya Ceramic Advances: Conjoining Stylistic, Technological, and Chemical Compositional Data. In Ceramic Research Advances, edited by Frank Columbus. Nova Science Publishers, Hauppauge, NY.

2001 Technological Styles of Late Postclassic Slipped Pottery from the Central Petén Lakes Region, El Petén, Guatemala. Ph.D. dissertation, Southern Illinois University Carbondale, Carbondale. University Micro-films, Ann Arbor, Michigan.

Glascock, Michael D.2007 The Archaeometry Laboratory at the University of Missouri Research Reactor [hyper-text document] http://archaeometry.missouri.edu

Jones, Grant D. 1998 The Conquest of the Last Maya Kingdom. Stanford University Press, Stanford.

Jones, Grant D., Don S. Rice, and Prudence M. Rice 1981 The Location of Tayasal: A Reconsideration in Light of Petén Maya Ethnohistory and Archaeology. American Antiquity 46: 530-547.

Rice, Prudence M. 1979 Ceramic and Nonceramic Artifacts of Lakes Yaxha-Sacnab, El Petén, Guatemala, Part I. The Ceramics. Cerámica de Cultura Maya 11: 1-85.

Introduction

At the time of the Spanish contact in Petén Guatemala (16th century), as many as five socio-political groups occupied territory in the central Petén lakes region (Jones 1998:18). Spanish documents and Postclassic to Colonial (13th-18th century) period ethnohistories of these groups document their presence (especially that of the Itza and the Kowoj), their chang-ing alliances and dominance relations in Petén, and their repeated migrations to and from northern Yucatán to central Petén. The documents suggest that Petén was divided into ad-ministrative provinces headed by and named after a dominant lineage and each lineage con-trolled several subprovinces or territorial regions throughout the area. The best documented Postclassic Maya socio-political group is the Itza. They controlled territory along the southern and western basin of Lake Petén Itzá and had their capital at Noj-peten or Taj Itzaj, on modern Flores Island (Figure 1)(Jones, Rice, and Rice 1981) and may have been an alliance of multiple socio-political groups (Jones 1998). Itza territory may have included the archaeological sites of Ch’ich’, Tayasal, Ixlú, Macanché Island, and Tipuj and most likely changed throughout the Postclassic period as boundaries were contested and changed. Although less well documented, the Kowoj occupied territory from the northern and east-ern shores of Lake Petén Itzá and Lake Salpetén. They claimed to have migrated from May-apán at approximately AD 1530 as a result of Spanish contact and political turmoil (Jones 1998). Kowoj territory may have included the archaeological sites of Zacpetén, Topoxté Island, and possibly Macanché Island, Ixlú and Tipuj. Rulership of Macanché Island, Ixlú and Tipuj may have changed multiple times during the Postclassic period due to changing alli-ances and borders. Throughout the Postclassic period, the Itza and Kowoj manufactured pottery made from la-custrine gray carbonate clays as is evident from the plethora of it at all of the Postclassic ar-chaeological sites in central Petén and western Belize (Figures 2 and 3). They produced pot-tery with three different slips (red, black, and pink) classified by archaeologists as the Pax-camán, Fulano, and Trapeche ceramic groups, respectively (Figure 2). The Early Postclassic period is also characterized by the production of the orange-red paste pottery of the Augus-tine ceramic group. Two probable locations of production have been identitied: Ch’ich’ and Tipuj (Cecil 2007). During the Late Postclassic period, the Kowoj at Topoxté Island produced a creamy-white paste pottery (Topoxté ceramic group) (Figure 3) and traded it primarily to other Kowoj at eastern lake archaeological sites (Cecil 2001; Rice 1979). In addition to chemi-cal differences within the ceramic groups, variability in rim diameters and sherd thicknesses provides insight into Maya pottery manufacture during the Postclassic period.

In order to obtain data concerning chemical grouping, instrumental neutron activation analysis (INAA) was used because it is the best chemical method to address questions about ceramic paste variability (clay, natural inclusions, and temper). Pottery samples were prepared for INAA using procedures standard at MURR (Glascock 2007). Portions of approximately 150 mg of powder were weighed into clean high-density polyethylene vials used for short irradiations at MURR. At the same time, 200 mg of each sample was weighed into clean high-purity quartz vials used for long irradiations. Along with the un-known samples, standards made from National Institute of Standards and Technology (NIST) certified standard reference materials of SRM-1633a (coal fly ash) and SRM-688 (basalt rock) were similarly prepared, as were quality control samples (e.g., standards treated as unknowns) of SRM-278 (obsidian rock) and Ohio Red Clay (a standard devel-oped for in-house applications). The following elements were counted and used for this analysis: Al, As, Ba, Ca, Ce, Co, Cr, Cs, Dy, Eu, Fe, Hf, K, La, Lu, Mn, Na, Nd, Rb, Sb, Sc, Sm, Sr, Ta, Tb, Th, Ti, U, V, Yb, Zn, and Zr. These elements were used to produce bivariate elemental plots reflecting the general compositional differences among the specimens.

ResultsEarly Postclassic Pottery (Figure 4) ● Paxcamán, Trapeche, and Fulano ceramic groups have two chemical compositions. Early Group 1 and Early Group 2 do not exhibit variability in rim diameter and sherd thicknesses. Plates in Early Group 1 are larger than Trapeche plates from Macanché Island. Early Group 2 plates are smaller than Augustine ceramic group plates and thicker than most other plates. ● Augustine 1, Augustine 2, and Augustine 3 rims and thicknesses do not vary. However, they are larger and thicker than most Postclassic samples.

Late Postclassic Pottery (Figure 5) ● Late Postclassic Paxcamán plates have larger rim diameters than those produced at Macanché Island, but are within range of other Late Postclassic plates. ● Because Topoxté pottery was made at only one location, variation is minimal. However, thickness varies between Topoxté Group 1 and Topoxté Group 2.