marked by changes in pottery styles. He identified the preclassic pottery types with those of Copan and Kaminaljuyu in the highlands of Honduras and guatemala and argued that the classic-period inhabitants of western Salvador were cut off from Kaminaljuyu when the latter site was Mexicanized and came under exclusive influence from the Mayan city of Copan. The classic Mayan-affiliated occupation of the area came to an end with the invasion of Mexican Pipil people after a.d. 1000. Longyear explicitly characterized western Salvador as an extension of other regions and portrayed changes in the archaeological record as responses to external influence. In his discussion of eastern El Salvador, he argued that little or no population displacement took place. From the preclassic period through to the classic, the region was comparable to central Honduras, and together they composed a Lenca archaeological culture.

Longyear’s discussion raised one issue that crossed the two regions he defined: the status of the diagnostic preclassic pottery type Usulutan resist. The preclassic period was the focus of a number of tentative new projects in the 1950s, including excavations by Muriel Porter on ash-sealed deposits at Barranco Tovar, by William Coe in the El Trapiche group of Chalchuapa, and by Wolfgang Haberland at Atiquizaya and Acajutla, all in the western part of the country. The latter two projects were curtailed prematurely, and the archaeologists did not have access to their collections for publication. When large-scale archaeological work resumed in the 1960s, one of the major objectives of site-focused projects in both the east and the west was clarification of the preclassic occupation history of the country.

E. Wyllys Andrews V, excavating Quelepa in eastern El Salvador, drew on field notes from unpublished work by Pedro Armillas in 1949. Andrews conducted excavations of monumental architecture and a ball court and constructed a chronological sequence based on construction phases and changes in pottery that spanned the preclassic and classic periods, ending at the very beginning of the postclassic period. His preclassic pottery included abundant examples of Usulutan resist decoration. With the fuller data he commanded on the associations of the Usulutan pottery, he was able to argue for strong connections with contemporary Honduran sites, reaffirming Longyear’s alignment of eastern Salvador with the Lenca. Andrews explicitly identified uncertainty about the sixteenth-century inhabitants of the region as a problem in the interpretation of the archaeological record.

Robert Sharer carried out excavations at Chalchuapa in western El Salvador from 1966 to 1970, following up the unfinished work of his professor at the University of Pennsylvania, William Coe. Sharer’s report expressed explicit concern about the application of scientific method as it was presented by “the new archaeologists.” He discussed the potential bias in project results owing to the impossibility of carrying out probabilistic sampling, both because of financial limitations and because of constraints raised by the local sociopolitical context.

The Chalchuapa project introduced several methodological innovations in Salvadoran archaeology. In addition to traditional excavations in architecture, fine-grained stratigraphic deposits on an old lake shore were carefully excavated in natural levels, and Payson Sheets developed a behavioral model of stone tool production based on the presence of different types of debitage (the detritus that results from making stone tools) and finished products. Through experimental replication, the project produced pottery vessels with the appearance of Usulutan resist in a pioneering attempt to understand the technology of this pottery.

Results from the Chalchuapa and Quelepa projects formalized the separation between east and west that Longyear had begun. Sharer explicitly framed this differentiation as a question of the location and nature of the Mesoamerican frontier. He retrospectively presented the issue of the “frontier” of mesoamerica as one that had concerned earlier archaeologists. Sharer traced the original genesis of Coe’s work at Chalchuapa to a survey by alfred v. kidder of the CIW in 1953 that had been designed to establish a chain of archaeological connectives from the Maya area through its southeastern periphery and into the non-Maya areas of lower Central America. He listed as his first objective