Xcalumkin was a small, Late Classic Maya center located on the western edge of the Puuc area. Following in the footsteps of Teobert Maler, who was the first to report the site in 1902, and of several later visitors (Harry Pollock in the 1930s, the New World Archaeological Foundation in 1968, and Ian Graham in 1986 and 1988) who produced maps, documented inscriptions, and dug test pits, Pierre Becquelin and Dominique Michelet visited Xcalumkin in 1991. During their visit, they took note of vaulted structures without specialized vault stones. These lay beneath Early Puuc (AD 715–770) and Classic Puuc (AD 770–950) structures and therefore held potential for research on the in situ development and florescence of a Puuc center. Becquelin and Michelet thereafter embarked on five seasons of mapping and excavations (1992, 1993, 1994, 2001, and 2007) with the goals of acquiring fine chronological control and understanding changes in economic, demographic, social, and political organization over time.

The current volume reports on settlement research from these field seasons. A follow-up volume will focus on the results of the excavations. Settlement research consisted of identifying and mapping 37 architectural compounds, 21 in the site center and 16 in the periphery. The mapping covered 35 ha (the book comes with a 40 × 25-inch pullout map at a scale of 1:1,000) but did not locate site boundaries. Thus, the extent of the site is not yet clear. The 37 compounds contain 131 buildings and 280 rooms. Assuming approximately two people per room, Becquelin and Michelet estimate that 550 people lived in these 37 compounds. For each building, Becquelin and Michelet gathered extensive architectural details not visible in the maps. These details make up Appendix 1; at 80 pages it is longer than the book's main text, which consists of an 8-page introduction and a 50-page chapter with summaries of each compound, interpretations, and other findings.

Becquelin and Michelet attribute the pre-Puuc vaulted buildings to the Early Xcalumkin phase, which they date to between AD 650 and 715. They identify 12 buildings pertaining to this phase. All concentrate at what would become the site core in two adjacent plaza compounds: C6 to the north and C9 to the south, separated by a single platform. Most of the Early Xcalumkin buildings contain a single room with an open colonnade on the long side facing the plaza. The lengths of the colonnaded buildings range from 9 to 21 m. Due to their lack of privacy, Becquelin and Michelet do not consider these buildings residential, instead favoring a political or ceremonial function; they likely represented lineage houses. The actual residences of these lineage members have not yet been detected. The fact that Early Xcalumkin buildings cluster into two compounds leads the authors to suggest a dualistic political organization.

Becquelin and Michelet believe that the dualistic organization persevered after later constructions covered all the Early Xcalumkin buildings. During the Early Puuc, the authors see dualism in the pairs of sajals (subordinate lords, in this case subordinate to the ruler at Jaina, located on the Gulf Coast 50 km to the west) attested first in inscriptions at the Hieroglyphic Group, then in Compound 10, and then once again in the Hieroglyphic Group. For the Classic Puuc period, they identify Compound C5, which contains the 10-room “Palace of the Cylinders,” as the principal group for the site as a whole but argue that it mediated a pair of factions, one representing the north side of the site and the other representing the south side. The leaders of the two factions lived at the site's two most costly compounds (C1–2 to the north of the principal group and C18, to the south). The presence of three temples and two adoratorios in Compound C9 leads the authors to suggest that this compound, located directly in the middle of the site and adjacent to C5, served as a religious center for the entire site. They also suggest an alternative in which each faction had its own religious center, with C9 pertaining to the southern faction and four ceremonial structures within C1–2 serving as the spiritual venue for the northern faction.

Becquelin and Michelet's calculation of architectural cost—what they call “relative labor value,” which they developed from their earlier work at Xculoc, Xcochkax, and Chunhuhub—is a schematic scoring system in which vaulted rooms count for 17 points, complex nonvaulted rooms count for 1.8 points, and simple nonvaulted rooms count for 1 point. The volume of basal platforms does not figure in the scores. Using this system, they score each compound and find a relatively continuous distribution of scores. Slight breaks in the distribution allow a tentative proposal of five classes, with 25 of the 37 compounds in the lowest class. All 16 peripheral compounds also fall into the lowest class.

Mapping at Sayil in the 1980s revealed an unexpectedly high proportion of vaulted structures, challenging notions of how to identify “elites.” Of the buildings documented at Xcalumkin, an even higher percentage (77%) are vaulted: 234 of 280 rooms are vaulted. Across the Puuc area, Becquelin and Michelet show that this high percentage of vaulted buildings resembles the percentages at Kiuic (78%) and Hunto Chac (71%) and contrasts with the percentages at Xculoc, Sayil, Xuch, and Xkipche, which each were around 40%. They claim that the most common unit of social organization at Xcalumkin was the extended family. Although larger than nuclear families, these groups would still have needed help from the larger community to build their vaulted homes in a timely fashion. Creative labor-pooling strategies have also recently been proposed at Kiuic. Cooperation and collective action, thought to be more and more common in Mesoamerica these days, thus gains additional purchase in the Puuc.