Читать книгу Mesoamerican Archaeology - Группа авторов - Страница 15

The amount of time it takes to create the sediments included in a depositional unit can vary from a very rapid interval (the time it took to sweep food scraps off a floor) to a longer period (the building of a temple and its use before abandonment). The amount of time it took to erode or remove deposits, a gap signaled by contacts between depositional units, is equally variable. In depositional histories, archaeologists understand that the layers of the simple model of superposition were created at different speeds, by events of different durations. Part of the archaeologist’s craft is narrowing in on the length of time that it took to create a particular depositional unit while simultaneously identifying the relative segment of time within which the activities that created that depositional unit took place. Both steps are required to interpret large-scale depositional events, like the construction of monuments, in terms of human decision-making and action. Knowing the relative order of things, and even the coordination of different relative orders in which things took place across space, is not enough: the sequence has to be complemented by measurements of real time elapsed.

A number of methods have been used to tie relative chronologies to measured (chronometric) dates. In many areas of Mesoamerica, indigenous historical texts with dates are available. Some of these were created or continued after Spanish colonization. Maya archaeologists often refer to specific years or even days recorded on monuments with inscriptions using an indigenous calendar for which a calibration with European calendars is possible as a result of references to Maya time keeping in texts written using the European alphabet (Aveni 1980: 204–210).

Of course, monuments with inscribed dates aren’t necessarily created at the moments they commemorate. In Mesoamerica, as everywhere in the world, historical texts can and do record earlier events, especially when those texts are political (Chapter 13). Think about the prevalence in the United States today of monuments mentioning dates like July 4, 1776, not one of which was produced before 1790 and most much later in the nineteenth century. Most dates in Maya texts do appear to fall in time spans expected on other grounds, since archaeological phases cover 100 years or more. Within these phases, texts promise the possibility of interpreting sequences of action down to a day so that within a broad phase archaeologists can have a sense of the human scale of historical events. The political histories of the Postclassic Mixtec also allow definition of events and spans of history measured in absolute terms, in lifetimes and actions of individual people.

Determination of equivalents in European calendars for dates recorded using indigenous calendars provided support for beginning and ending dates for the intervals of time proposed as phases using archaeological methods of chronometric dating. The methods involved in chronometric dating take advantage of natural processes of change that occur at known or precisely measurable rates and that affect common materials found at archaeological sites. A prime example is the use of the known rate of decay of radioactive forms of carbon. By measuring the ratio between different isotopes of carbon in organic materials that are the remains of plants and animals, we can produce estimates of the time elapsed since the death of these plants or animals.

While the technique of radiocarbon dating is the most widely employed material analysis supporting Mesoamerican chronologies, other materials can be used. Obsidian hydration is a second example. This method exploits the natural tendency of volcanic glass to absorb water from the atmosphere, creating a “hydrated” rind on fresh glass surfaces exposed when tools were first created. Specific rates of hydration have to be established for obsidian from different sources used in sites with different environments, and fluctuations in climate over time have to be considered; however, the method has the potential to provide dates on a material that is abundant in most Mesoamerican sites.

All chronometric methods produce estimates of the probability that a specific event took place during an interval of time. In the case of obsidian hydration, that event is the exposure by a person of a fresh surface on an obsidian tool, presumably near the time of its use. In the case of radiocarbon dates, the event is the death of the organism from which the carbon sample came. A sample could come from a plant that died long before it was used by human beings or that was reused much later (as when a large piece of timber might be recycled from one building to another). Because trees grow by adding layers of living tissue over dead tissue, even different parts of the same log from a long-lived tree can produce different age estimates.

Systematic concerns arise because what is being dated is the event that began the chemical process that provides the measured passage of time: decay of radioactive forms of carbon into other forms of carbon or absorption of water by the fresh obsidian surface. Samples can be recovered from deposits created long after the events that would be dated. When trash containing obsidian tools and scraps of plant material was swept up and used as part of construction fill for a building, the event of historical interest is the date of construction, but the radiocarbon and obsidian hydration dates will be of earlier events. To counter this, archaeologists produce dates for multiple samples from the same deposits so that samples that do not belong will stand out.

For each sample evaluated, the possible date of the event that began the process on which the method is based is calculated as an estimate of an interval of time, not as a singular year or day. Estimates are reported by the specialist labs that carry out analyses as intervals, with a specific degree of precision. For example, Beta-129129, analysis of a carbon sample from Puerto Escondido, a site in Honduras where I excavated, was reported as 3320 +/− 40 BP (where BP means Before Present and the present is 1950 CE). The reported date brackets an interval of 80 radiocarbon years when the plant might have died.

When radiocarbon analysis was initially introduced, analyses of samples used to measure the beginnings and endings of phases and periods were translated by a simple process of subtraction, counting backward from 1950 CE. Later, researchers realized this simple procedure was based on an incorrect assumption: that the concentration of different forms of carbon in the atmosphere had not changed over time. Specialists in chronometric dating have produced highly detailed graphs showing fluctuations in concentrations and calculated the divergence between the simple radiocarbon age and the real calendrical age. Computer programs allow archaeologists to calibrate radiocarbon years and estimate real calendar years of the interval when the plant or animal producing a carbon sample most likely died.

The effects of calibration can be substantial. The aforementioned sample from Puerto Escondido corresponds to 1690–1510 cal. (calibrated) BCE. If we simply had subtracted the radiocarbon years from 1950 CE, the same sample could have been mistakenly treated as the date 1370 BCE, plus or minus 40 years, or an interval from 1410 to 1330 bce. A plant that actually died sometime between 1690 and 1510 BCE might have been thought to have lived and died centuries later. All the historical events and all the people whose lives that single carbon sample dates could have been falsely thought to have taken place much more recently than they actually did.

Even when a shift in dates seems relatively minor, calibrating radiocarbon dates corrects potential misunderstanding of the length of an interval of time. Sample Beta-129125 from Puerto Escondido was dated 1530 +/− 40 BP in radiocarbon years. If we just counted backward from CE 1950, this would mean the plant producing this carbon died between CE 380 and 460. The calibrated date range was actually cal. 430–625 CE. Not only does calibrating change the date of the depositional event that yielded this carbon sample from Early Classic to Late Classic, but it also greatly increases the span of time within which the event most likely occurred, from 80 to 195 years.

The effects of calibrating samples do not vary in a single predictable way. Puerto Escondido sample Beta-129126, at 2730 +/− 40 BP in radiocarbon years, would correspond to the span from 820 to 740 BCE if we just counted backward from 1950 CE. The calibrated date span of cal. 940–810 BCE shifts the interval earlier (instead of later as in the previous example) and changes the interval of highest probability to 130 calendar years (from 80 radiocarbon years). These effects matter when archaeologists are interested in understanding the rate of change and timing of actions within a society over individual lives and connected generations, the kind of timing seen in indigenous calendar notations, critical to contemporary archaeological interpretation.