Shell boring is one of the few quantifiable, well-preserved biotic interactions in the fossil record. While some workers have used boring intensity as a proxy for predation pressure in both Recent and fossil assemblages, others have warned that taphonomic and other effects can alter the boring intensity and lead to paleoecological misinterpretation. When the biocoenosis is known, however, taphonomic effects can be “undone” a posteriori. Because the valve ratio of any bivalved organism must be 1:1 in the biocoenosis, strengths of between-valve taphonomic biases can be calculated directly for any disarticulated assemblage of bivalved fossils. By back-calculating the biocoenotic boring intensities, improved estimates of boring stereotypy may be obtained. These “restored” boring intensities can provide more accurate paleoecological interpretations of boring habit while remaining numerically conservative.

Taphonomic analysis of a bored Ordovician brachiopod assemblage shows that (1) the assemblage had experienced negligible differential transport; and (2) convex (pedicle) valves have been preferentially crushed in place. Comparing the taphonomy of the assemblage to a set of laboratory taphonomic regimes reveals that valve-valve contact may be of great consequence in skeletal taphonomy. In particular, valve-valve contacts appear to promote (1) preferential destruction of convex valves; and (2) subequal destruction of bored vs. unbored valves. Taking into account these taphonomic effects, numerous hypotheses of boring habit—including mixed-motive boring—have been tested using a probabilistic model. The model herein presented indicates a likely contribution of 10–15% predatory boring in the assemblage. The usefulness of probabilistic models for providing simultaneous, realistic tests of multiple hypotheses is emphasized.