Lecture 4 - Chapter 4: Archaeological Methods
The last couple classes, we examined the Maya site of Copan as an example of an archaeological research project. Tonight, I'll explain in more detail how archaeologists dig at site.
Remember from Copan and our discussion of the scientific method is that one of the most important things about beginning an excavation is that it must be problem oriented...you should be digging where you are because it will test a hypothesis that you have formulated. At Copan, the purpose of the overall project was to examine how the prehistoric Maya related to their environment. They then formulated several specific hypotheses that were tested by excavations. But before they dug, they made sure that they knew everything there was to know about Copan up to that moment. Thus, ethnohistoric accounts through the most recent excavation records were carefully studied, as well as preliminary survey work done by methods of remote sensing, such as aerial and satellite photographs. In addition, remember that the first thing they did at Copan was examine the environmental variables...how did the environment change through time? Thus, before they moved any dirt, they knew everything there was to know about the site at that moment and could concentrate on answering new questions.
However, we don't just start digging...Ideally, we would be able to completely excavate every square centimeter of every site so that we could have 100% complete data recovery. Unfortunately, reality does not allow that. It seems that archeologists are always short of money, therefore labor, therefore time. Thus, we have to develop methods of knowing about the whole site, while only excavating a portion it. This is known as sampling.
There are several statistical methods of doing this:
Simple Random Sampling: every unit (artifact, site, square, etc.) has an equal chance of being selected. The investigator determines how large a sample they want (how many units) and uses a random number generator (either a table, or more likely, a computer) to select the individual units to be sampled. Problems: site density is not uniform: most interesting stuff is usually at the center, which is undersampled (minority of squares). also, large gaps between sample points can form resulting in patchy coverage. thus, important things can be missed. (a hypothetical random sample survey of the Teotihuacan valley in Mexico missed the major(!!!) city at Teotihuacan!)
Systematic Sampling: selecting sample using a predetermined system or pattern (i.e. every other transect, checkerboard, etc.) While this does insure that the entire site will receive the same amount of sampling, important features can be missed if they are similarly patterned (i.e. Iroquois longhouses).
Stratified Sampling: this method breaks the sample region (or population) into discrete sections based on obvious variations (topography, geology, arch. features, etc.) How might Copan be divided? For example, within the Copan study area, sites locations ranged from fertile river bottoms, to along minor streams, to hilltops away from water. In addition, remember that they have divided the sites into 5 types...Sanders and Webster wanted to make sure that they excavated several of each type.
Finally, now that we have completed our background research, developed a hypothesis to test, and selected specific units or sites to sample, we can finally start digging.
First, the surface of the site is mapped by having a grid imposed on it. Then, the surface is systematically searched for artifacts...at Copan, this was done over a 110 square-km area! This superficial level of investigation is known as a Stage 1 Survey . Artifact deposits on the surface give us an idea of what is underground.
From this data, area's of potential interest are identified. Those that might be a good test our hypotheses are subjected to test-pitting...a Stage 2 Survey. If the test pits prove to be promising, then they are expanded into large-scale excavations, or a Stage 3 Survey. The purpose of this level of survey is to uncover large areas of the prehistoric landscape. At Copan, this often meant an entire site and its immediate surroundings were excavated.
Both the Stage 2 and 3 surveys involve the kind of excavations that most people associate with archaeology. An excavation unit is mapped out according to the 3-dimentional central grid. This way, the provenience, or spatial context, of any artifacts or features that are found can be recorded to within a centimeter. This permits the site to be virtually reconstructed in the lab after the real site has been destroyed by excavation. Once the limits of a specific excavation unit are mapped, then the soil is carefully removed layer by layer. If the top-soil is disturbed or mixed by plowing, then it is just shoveled off or even scraped off by a machine. Remember, we are trying to capture an artifact in its original context. If this has already been destroyed through plowing, there is no point in carefully trying to recover non-existing context. Once artifacts begin to appear in undisturbed context, then the pace of excavation slows down dramatically. Soil layers are carefully removed using a hand trowel and every handfull of soil is screened. In addition, samples of soil, ash, and any other unusual deposits are saved for floatation analysis. By immersing the soil in water, any organic material will float to the surface. In this way, many samples of charcoal, seeds, fish bones, etc. area collected.
Ok, so the archaeological deposits area found in layers...what's this called again? Right! Stratigraphy. Well, as you dig down, it's often difficult to tell where one layer ends and the next one begins. Thus, a trench will often be dug to expose the layers as a profile. Then, each layer can be excavated by starting at the profile and following it along. Similarly, this is why archaeologists area careful to protect the walls of their units...when a unit is completed, these profiles are mapped and provide a cross-section of the deposit and a useful record of the prehistoric activities in that unit. Usually, excavation is continued until sterile soil (no cultural material) or bedrock is reached.
As artifacts are uncovered, their locations are mapped with as much precision as possible. They are then labeled with this data, and sent to the lab. Here, several things can be down to it and much additional information obtained. Often, the artifact will just be washed, categorized, counted, and stored. However, if the hypotheses require it, much more can be done. Individual artifacts can contain loads of useful information...information that is only useful if its spatial context is recovered.
Dating: many methods of determining when an artifact was manufactured are possible. The most common are:
Dendrochronology: Trees add one ring per year as they grow. A master calendar is constructed beginning with very old tree like sequoias or bristle-cone pine and is extended back by matching the patterns on archaeological wood with those on the existing calendar sequence. Once a calendar is complete, other pieces of wood that are found may be matched to it and dated. Limitations: wood doesn't preserve in most environments, reuse, ring removal.
Radiocarbon: A radioactive isotope of carbon, C-14, occurs naturally in the atmosphere and gets included in carbon-dioxide. CO2 is what plants breath....so C-14 is absorbed by plants when they breath...thus they have to be alive. This C-14 is then absorbed by animals that eat the plants, and humans who eat both the animals and the plants. Thus, plants, animals, and humans maintain a constant intake of C-14 while they are alive, but do not intake any more C-14 when they die. Because C-14 is radioactive, it decays continuously throughout this process. Because its half-life is ~5730 years, after 5730 years, only half the original amount is left. This means that, since the amount of C-14 is fixed in an organism at the time of death and decreases at a fixed rate, by measuring the amount remaining, you can calculate how long ago it was alive. Problem: expensive, somewhat inaccurate (+-40 years), easily contaminated, sample is destroyed.
Obsidian Hydration: Because obsidian is a glass, it absorbs water at a rate determined by its chemistry and environment. This absorption changes the outer layer of the sample. This altered layer gets thicker over time as more water is absorbed. Because this rate of layer growth can be calculated, the thickness of this layer can be measured and the time that has elapsed since manufacture can be determined. Problems: determining the hydration rate is difficult, sample has a chunk cut out of it.
In addition to chronological information, different methods of chemical analyses can match rocks from stone tools or dirt in pottery to the original sources of the raw materials. This can tell us about migrations and trade relationships.
Analysis of use-wear on bones, stone-tools, and pottery can tell us what they were used for.
Chemical and morphological analysis of bones can tell us about what the people are and what diseases they may have had.
Thus, archaeological field work can provide us with tons of tangible physical evidence. How do we transform this evidence into interpretations of the less-tangible realm of human behavior?