INTRODUCTION

Ccurately dating ancient events or processes is one of the most essential, yet difficult, tasks faced by archaeologists. Mayanists, fortunately, often can rely on calendar dates to establish chronological frameworks for many great Classic centers prior to the eighth-ninth centuries. When we venture into the Terminal Classic and Postclassic periods and attempt to understand the Classic Maya “collapse,” however, we leave behind the comfortable chronological anchorage of Long Count and period-ending dates.

Two decades ago Raymond Sidrys and Rainer Berger (1979) compared Long Count dates on Classic monuments with the then-available radiocarbon dates from both elite and commoner contexts. They concluded that these data sets together suggested the rapid demise of kings, a somewhat later disappearance of associated nobles, and a protracted survival of Maya commoners inconsistent with pervasive conceptions of an extremely catastrophic demographic collapse. But they noted that their radiocarbon sample from commoner contexts—thirty in all—was woefully small, and they urged “Maya archaeologists to obtain more commoner-associated ‘'*C dates, as these dates should eventually be of great value in resolving the depopulation issue and be more representative of the cultural situation as it really was” (Sidrys and Berger 1979; 274).

What must be dated if we are to understand “the cultural situation as it really was” is daunting. A few dates do little to resolve big issues such as how quickly political centralization unraveled, how rapidly populations declined after the

Disappearance of ruling dynasties, and how (or if) ceramic traditions changed. Nor can we necessarily rely any longer on stratigraphic or other information from the great royal centers, which were often largely or wholly abandoned in Terminal Classic times. Archaeologists must instead recover, as Sidrys and Berger advised, many dates from good contexts in lesser places that might relate to what Bey, Hanson, and Ringle (1997) call “post-monumental” times.

The experiments reviewed here demonstrate that at Copan, Honduras, we have amassed a very large, representative, and reliable sample. The sixteen accelerator mass spectrometer (AMS) radiocarbon dates reported below from commoner contexts total more than half the number Sidrys and Berger found in the entire literature in 1979, and they are from clearer contexts and are technically more reliable. But we also have hundreds of obsidian hydration dates from commoner (and elite) households that make excellent sense when checked against these AMS dates and that, not coincidentally, yield basically the same culture-historical patterns that Sidrys and Berger detected.

This data array derives from an obsidian hydration dating project carried out in conjunction with our regional settlement surveys and excavations. Settlement, political, demographic, and land-use reconstructions derived partly from the obsidian dates have already been presented in considerable detail and are not repeated here (Freter 1988, 1992, 1994, 1996, 1997; Paine and Freter 1996; Paine, Freter, and Webster 1996; Gonlin 1993; Webster 1999; Webster and Gonlin 1988; Webster and Freter 1990b; Webster, Freter, and Gonlin 2000; Webster, Sanders, and Van Rossum, 1992; Wingard 1992, 1996). This paper instead reviews a recent set of concordance experiments carried out on materials from eleven sites in the Copan Valley (Figure 11.1) that independently test and strongly support the hydration chronology central to these reconstructions.