Archaeologists have long labored under the implicit assumption that the archaeological record is a direct reflection of past human behaviors. However, numerous cultural and environmental processes intervene between past behaviors and their reconstruction through archaeological inference. This study examines the interface between household archaeology and formation processes through the study of domestic materials from two contemporaneous sites in the Zapotitan Valley of El Salvador that were occupied by people who spoke the same language and belonged to the same regional political system. Ceren was a small village that was occupied for several decades before it was deeply buried by the eruption of Loma Caldera volcano. San Andres was a much larger center that also was affected by several eruptions, but did not experience long-term catastrophic abandonment or exceptional preservation. The research examines the effects of cultural formation processes, including reuse, discard, abandonment, and post-abandonment disturbance processes, and non-cultural formation processes, such as effects of catastrophic volcanic burial, and the effects of plants and animals. It compares the de facto refuse from Ceren with discarded materials from Ceren, and San Andres using the discard equation and methods developed in accumulations research to build a foundation for more generally applicable models to interpret household remains in western El Salvador and throughout Mesoamerica.

This dissertation examines El Marquesillo, a settlement in an archaeologically unexplored region of the Southern Gulf Lowlands of Veracruz, Mexico. Evidence suggests the site has been consistently occupied from the Early Formative period (c. 1500 BC) to the present. Thus, this investigation presents an opportunity to re-examine the sociopolitical continuum encompassing the Olmec cultural phenomenon (c. 1150-300 BC), the emergence of which has been used repeatedly as an example of incipient social complexity. Theorists have portrayed the development of sociopolitical complexity as a mosaic process in which environmental, social, political, economic, ideological, and demographic variables act independently or in combination to bring about change. In order to examine these variables, a suite of traditional and progressive archaeological techniques -- remote sensing, geophysical survey, GIS, mapping, anthropogenic soil survey -- were employed to prospect, document, and analyze the natural and built environments along with the material record documented at El Marquesillo. I argue that the resulting data do not fit many of the traditional models that have been offered to explain the development of Olmec sociopolitical complexity. The term "traditional Olmec paradigm" is used to describe a collective array of conjectural concepts that have been proposed by theorists to explain how Formative people of the Southern Gulf Lowlands constructed and experienced their reality. Findings from El Marquesillo and other recent Heartland investigations suggest that much of this traditional Olmec paradigm may not be accurate. The Gulf Olmec were not a homogeneous and uniform entity across space and time. At El Marquesillo, idiosyncratic behaviors of the ancients relating to ancestor veneration and their connection to the landscape and worldview have been identified. These noted variations in social expression and the lack of adherence to the traditional Olmec paradigm suggest that some hypotheses regarding the Formative people of the Southern Gulf Lowlands be re-visited and possibly revised in the light of new evidence.

Teotihuacan was the most influential city in the Classic Mesoamerican worldsystem. Like other influential cities in the ancient world, however, Teotihuacan did not homogenously affect the various cultural landscapes that thrived in Mesoamerica during the Classic period (300-900 CE). Even where strong central Mexican influences appear outside the Basin of Mexico, the nature, extent, and strength of these influences are discontinuous over time and space. Every place within the Classic Mesoamerican landscape has a unique Teotihuacan story. In the Tuxtla Mountains of southern Veracruz, Mexico, Matacapan, located in the Catemaco Valley, drew heavily upon ideas and symbols fostered at Teotihuacan, while Totocapan, a peer political capital located in the neighboring Tepango Valley, emphasized social institutions well-entrenched within Gulf Coast cultural traditions. Through a detailed comparison of these two river valleys, I demonstrate that each polity developed along different trajectories. By the Middle Classic (450-650 CE) each polity displayed different political, economic, and ritual institutions. While they shared an underlying material culture style, the data suggest that the regimes of both polities promoted a different ideology. These cultural divergences did not, however, cause hostilities between them. To the contrary, compositional sourcing of Coarse Orange jars indicates that they engaged in material exchanges with each other. Agents at each settlement within the study region made unique decisions with regard to their involvement in local, regional, and macroregional interaction networks, particularly with regard to the adoption or rejection of Teotihuacan cultural elements. As a result, the Classic period Tuxtlas comprised multiple overlapping, but disjoint, landscapes of interaction. Places of human settlement were nodes on the landscape where these disjoint landscapes intersected in space and time. By examining these disjunctures, world-system studies can reveal a trend of increasing cultural diversity that parallels the better-theorized trend of homogenization emphasized by core-periphery models. In this dissertation, I take the initial steps toward developing an archaeology of disjuncture that examines the cultural variability that develops where groups across the landscape employ different strategies of interaction within the world-system.