Graduation Year


Document Type




Degree Granting Department


Major Professor

E. Christian Wells


Ceramics, Geochemical characterization, Plaster, Prospection, Provenance, pXRF


The use of geochemical analysis for characterization studies of archaeological material has been increasing for decades. In recent years, advancements in X-ray fluorescence (XRF) instrumentation have led to hand-held portable XRF (pXRF) instruments capable of on-site, rapid, non-destructive analysis. The addition of pXRF to the archaeologist's toolkit has the potential to revolutionize geochemical characterization studies as research design can be informed by field analysis, once off-limits museum collections may now be analyzable, and data can be gathered on in-situ objects without disturbing their context. This new instrumentation has shown promise in characterization studies on a variety of archaeological materials, including ceramics. Particularly, it has been shown to generate data useful in distinguishing the geochemical provenance of artifacts. While most existing provenance studies that employ pXRF involve obsidian, it is possible that pXRF is a suitable analytical tool for ceramics as well, since instrumental neutron activation analysis (INAA) and laboratory XRF have both been used in provenance studies of obsidian and ceramics. This thesis reports research that utilized a Bruker hand held pXRF analyzer to study ceramics and plaster floors from Classic Period (ca. AD 250-1000) archaeological sites in southeast Mesoamerica. The data gathered from this analysis are used to examine the ability of pXRF to address archaeological questions of archaeological provenance on ceramic material and human activity patterns on archaeological plaster. The data gathered on ceramic material is analyzed to determine if "valley profiles" (or "chemical fingerprints") can be created for samples excavated in the same valley and to investigate inter-valley trade. Data from the plaster samples is examined to determine if pXRF can detect human activity patterns on plaster floors and, if so, if the activity is identifiable. The results suggest that, while pXRF results can be used to create "valley profiles" for provenance research, it is difficult to ascertain if it can be used to identify items of trade. This investigation also shows that, while pXRF data can identify human activity on plaster, it is difficult to determine what those activities were. This leads to the conclusion that pXRF has potential for both provenance and prospection studies on earthen materials. However, methodological development is necessary to fine tune research design. This thesis contributes to a greater understanding of the potentials and limitations of pXRF instrumentation in characterization studies of archaeological materials as well as understanding cultural-historical details in Classic southeast Mesoamerica.