Neutron Activation Analysis on Olmec Pottery: A View From La Venta

Analysis

Inherent to compositional studies is the provenance postulate (Weigand et al., 1977). In short, variation is greater between two spatially separated regions than the variation within a particular region. Thus, in order to recognize and monitor production and vessel movement from both Olmec sites, variation between San Lorenzo and La Venta must be greater than the variation within each particular site so that the chemical signatures of clays can be distinguished. In this bivariate plot, Figure 4, of chromium and antimony such is the case. Compositions of clays at La Venta ("+") tend to have both smaller concentrations of antimony and higher concentrations of chromium than the San Lorenzo clays ("x"). Having obtained all elemental data from the clays, sands, and pottery, the analysis continued by forming reference groups of the pottery based on elemental concentrations. Within the 174 samples, 2 compositional reference groups were realized (Figure 5). The first was a small, N = 13, group of pottery found only at operation 28 in complex E, while the other was a much larger, N = 147, pottery group found in all three proveniences. The remaining 14 pottery samples were unassigned. The two groups were distinguished by their varying amounts of the elements thorium and antimony. Group 1 has higher concentrations of both of these elements while group 2 has significantly smaller concentrations. Next, the newly formed compositional groups were compared against the backdrop of elemental concentrations of the local clays found at La Venta (Figure 6). This bivariate plot shows a strong correlation between reference group 2 and La Venta clays identifying the pottery samples of group 2 to be of local procurement and manufacture. Group 1 however, shows no compositional similarity to La Venta clays and was interpreted to be of non-local origin.

All pottery samples in this analysis were also typed using the 1955 type collection formulated by Drucker (1952). It was also realized in this research that Drucker’s types are not compositionally sound (Figure 7). The 90% confidence ellipses of Drucker’s wares indicates here that similar clays were used in the production all wares. It is interesting to note also that in an appendix to the Drucker report (1952), Anna Shepard carried out technical analyses on several fine paste wares. She concluded that at least two clays were used in the production of her sample. A finding that is compatible with the results offered here. However, the pottery she analyzed came from contexts of later temporal placement and was centered on fine paste wares.

A test of clay membership probability was then calculated using the GPS data converted to UTM coordinates for locally produced reference group 2 (Figure 8). In this figure, the solid circles represent the sample locations of clays. The darker shaded regions represent areas of clays sampled with higher probabilities of membership that may have been used in the production of the pottery sampled. It can be seen in this figure that the clays most similar to pottery group 2 are found around the levee site Isla Alor, suggesting the locale for procurement of materials in ancient times. Interestingly, the La Venta island itself which has an immense clay resource had the smallest probability (less than .50%) of membership. This indicates that the La Venta island was not exploited for clays used in the production of locally produced reference group 2.

Finally, the non-local pottery group 1 was compared to a clay database from San Lorenzo (Figure 9). All clay samples that were used in this comparison were obtained by R. Sergio Herrera (1998). In this bivariate plot of chromium and thorium, the San Lorenzo clays are designated by an "x" and are plotted against a 90% confidence ellipse for non-local reference group 1.  Concentrations of these elements vary significantly to suggest that San Lorenzo was not the source of clay materials used in their production. Thus, other regions must be considered to find the provenance of these ceramics. Unfortunately, the present compositional database of Mesoamerican clays does not contain adequate analyses to test other regions against non-local reference group 1.

Previous Page  |  Table of Contents  |  Next Page

Return to top of page