Faunal assemblages offer rich data for exploring domestication, subsistence, ritual practice, and political economy. Issues of equifinality, however, frequently complicate interpretations because different agents and processes may create similar archaeological signatures. Analysts are often forced to make interpretations based on qualitative observations, which can be difficult to justify or replicate. I present an alternative method for classifying Andean assemblages by using ethnographic, ethnohistoric, and archaeological data to construct a Bayesian network model. The model is assessed using specifically constructed test datasets and archaeological case studies. Bayesian models can lead to explicit and quantifiable probabilistic interpretations of faunal assemblages.
