Volume vs. Area Density - Baja and California

Results of excavation sampling have been used to estimate the density with which artifacts occur in archaeological sites. Such density estimates may serve several functions, both in evaluating the sites as resources for potential future research and in interpreting them.

Density of artifact occurrence may be used as a measure of the comparative richness of a deposit and therefore as one measure of a site’s potential productivity in information about prehistory, as against the effort that would be required to recover that information.
Density may serve as a basis for estimating the total artifact content of a site, and therefore the total potential yield and the relative magnitude of the site as compared to other sites (e.g., Laylander 1989:138-141).
Density may be used as a measure of the intensity of prehistoric use of a site (e.g., Corum 1991). Intensity of use is a function of a number of variables of archaeological interest, including the length of the site’s occupational history, the duration of the individual occupations, the size of the occupying groups, the compactness of their settlement, and the rate at which they produced artifact residues.

Density of artifact occurrence has been expressed in two ways: as density per volume excavated, and as density per area excavated. For some purposes, volume density appears to enjoy certain advantages:

This method of expressing density seems to be the one that has been used most frequently in the San Diego region. It is therefore the one for which comparative values are most readily available.
Guidelines for archaeological testing mandated by the County of San Diego, which were intended to be used in determining the required test sample size, explicitly used volume density of artifact occurrence as the key variable (cf. Davis 1990).
Volume density links artifact yield to the effort and expense required for artifact recovery more directly than does area density.
If the sedimentary build-up of a deposit can be attributed, to a substantial degree, to cultural activity, then the relative density of artifacts per unit volume within the deposit may be indicative of the character of the cultural activity that produced the deposit (e.g., Hagstrum and Hildebrand 1988:9-10).
If the build-up can be attributed to natural sedimentation occurring at a roughly constant rate, then variation in the density of artifacts per unit volume may be indicative of relative intensity of occupation during particular time periods (e.g., Carrico and Taylor 1983:113).On the other hand, area density also appears to have some advantages:
In excavation reports that do not specifically discuss artifact density, the information needed to calculate area density (i.e., the area excavated) is more consistently reported than the information needed to calculate volume density (i.e., the volume excavated). Therefore the potential data base for the area density statistic may be larger.
Postdepositional downward mixing of artifacts into natural soils appears to be an important factor in many San Diego sites (cf. Borst and Olmo 1983; Gross 1990). In such cases, the depths to which artifacts are distributed, and therefore the deposit volumes from which they are recovered, will be a function of such noncultural factors as the natural soil depth and the intensity of activity by burrowing rodents. Density measurements based on area will be much less distorted by such noncultural factors.
Archaeological deposits in the San Diego region are commonly found to become gradually attenuated in artifactual content with depth. As a result, the field decisions that are made concerning how deep to continue excavating to recover increasingly sparse artifacts may have only a minor influence on the total artifact count for a given area of excavation, but they may have a major effect on the total volume with which that count is associated.

Estimation of total site artifact density and total site artifact content based on area density is somewhat more feasible than estimation based on volume density. Estimation based on area involves only the calculation of the total site area, the excavated sample area, and the artifact sample recovered. Estimation of total density and content based on volume also involves estimating of the mean depth of the deposit throughout the site, and a good estimate for this variable is frequently not available.

In addition to the a priori arguments, some empirical testing of the relative effectiveness of volume and area density statistics is possible. One test of effectiveness would be to compare the coefficients of variation in the sample estimates of volume and area density as calculated for the various excavation units. A relatively small coefficient of variation would indicate that a relatively precise density estimate had been obtained. As an illustration, this calculation has been applied to the densities of lithic wastes for two site complexes from which fairly extensive (although nonrandom) test excavation data are available: SDI-5383, in the Penasquitos area, with 38 test units (Laylander 1989); and SDI-10,996/10,998, in Spring Valley, with 20 test units (Laylander 1992a). In both cases, the coefficients of variation are large, but the coefficients for volume density are somewhat smaller than those for area density:

		SDI-5383	SDI-10,996/10,998
Area	Mean Density / m²	180.7	300.4
Area	Coefficient of Variation	99 %	126 %
Volume	Mean Density / m²	310.3	338.2
Volume	Coefficient of Variation	83 %	97 %

Another empirical test would be to consider the coefficient of variation in the depth of the cultural deposit, in order to see what sort of precision would be available in estimating the total volume of the site deposit. For sites SDI-5383 and SDI-10,996/10,998, the coefficients of variation in this variable are 45% and 49% respectively, indicating that a sizable uncertainly factor in estimating total site density and content would be introduced by using volume density.

A third test would be to consider whether patterns in the vertical variability of volume density, such as a decline in density with depth, can better be explained as products of cultural or of natural processes. If vertical patterning is cultural, the use of volume density may provide additional interpretive possibilities. If the patterning is natural, as noted above, the use of volume density may cause distortions.

PROSPECTS

Future archaeological investigations may be able to test the relative effectiveness of volume and area density statistics for various interpretive objectives and the validity of the assumptions that underlie them.