Postdepositional Size Sorting

The postdepositional transformation of archaeological deposits in San Diego County by natural processes, particularly bioturbation, is recognized as a commonly occurring phenomenon that is potentially important in interpreting those deposits (e.g. Borst and Olmo 1983). One effect of such transformations may be to blur or completely destroy any original stratigraphic patterning. Another effect may be to create new and potentially misleading vertical patterning arising out of the physics of the transformation process. Bioturbation may preferentially concentrate specific artifact categories upward or downward, based on such physical properties as artifact size, density, and shape. Distinguishing patterns produced by postdepositional sorting from culturally significant depositional patterns is essential if the deposits are to be interpreted correctly.

Size is the physical attribute that is most easily addressed, and perhaps the most significant one. Postdepositional size sorting is probably best studied in the vertical distribution of lithic tools and wastes. Ceramics, bone, and shell are all subject to significant postdepositional breakage, and bone and shell are also subject to chemical attrition; these factors may seriously complicate any size analysis. At a particular site, it may prove useful to establish the extent of postdepositional size sorting by analyzing patterns in the distribution of lithic artifacts, and then to use the results in interpreting size patterning in ceramics, bone, and shell.

Barbara Bocek (1986) discussed postdepositional size sorting by rodent burrowing, illustrating it with results from the Jasper Ridge Site (SMA-204), south of San Francisco. On the basis of rodent behavior, she suggested that “materials between .6 and 2.5 cm in size should be disproportionately numerous near the surface…materials larger than 5 cm, perhaps those larger than 2.5 cm, should cluster slightly below the rodent zones, both ancient and modern” (Bocek 1986:592). At Jasper Ridge, Bocek found some downward concentration of fire-cracked rock and natural rock. She reported that “average sizes of bone, shell, and debitage were significantly smaller in the upper deposit (levels 1-6) than in levels 7-12” (Bocek 1986:600); however, a table (Bocek 1986:597) showed that the mean weight of a piece of debitage in the upper levels was 2.9 grams, while the mean weight in the lower levels was 2.2 grams.

Jon M. Erlandson (1984) documented and discussed the processes of deposit disturbance by rodents at a site in Santa Barbara, SBA-1582. Although Erlandson did not address the question of size sorting, he presented level summaries by count and weight which showed a clear tendency for heavier lithics to be more abundant in the upper levels, contrary to what was predicted by Bocek’s study.

A few San Diego County studies have provided data that may be applied to this question:

• At site SDI-5383 in the Peñasquitos area, investigations by Richard H. Norwood (1982) showed that the average weight of lithic waste items increased with depth (Laylander 1989:143). However, a subsequent investigation at the same site, reporting both average weights of lithic waste items and counts by three size classes (longer than 3.0 centimeters, 1.5-3.0 centimeters, and shorter than 1.5 centimeters), found the opposite pattern, with lithic wastes being smaller in the lower levels (Laylander 1989).
• At Ystagua (SDI-4609) in Sorrento Valley, in data reported for flakes (types 3 through 7, excluding shatter, blade flakes, and wide-platform flakes), items shorter than 3 cm were disproportionately concentrated in the lower half of the deposit (Hector 1985:54).
• At site SDI-10,998 in Spring Valley, a statistically significant concentration of smaller lithic waste items in the upper portions of the deposit was found (Laylander 1992a:325).

PROSPECTS

Future archaeological investigations may be able to clarify the effects of bioturbation on vertical patterning in archaeological deposits and assist in distinguishing those effects from patterning that arose from diachronic changes during the original episodes of deposition. Replicative experiments with simulated cultural deposits may also be able to shed light on this issue.