Reliability of Species Assessment of Archaeological Cremated Bone Fragments: Confidence in Diagnosis and Interobserver Agreement

ABSTRACT The ancient practice of cremation has been documented across numerous cultures, yet burnt remains can indeed become a huge challenge for anthropologists, both in archaeological and in forensic contexts. The analysis of such skeletal material presents significant challenges associated with the alterations caused by the process of burning. A salient issue in studying cremated remains pertains to the challenge of distinguishing human remains from those of other animals, particularly on small fragments. The objective of this study was to assess the confidence of different operators on species identification (human/nonhuman) in cremated remains using macroscopic observation. To this purpose, 50 unidentified cremated bone fragments dating from the first to third century ce Italy were analyzed by three different observers. In more than 40% of the cases, observers were not able to suggest an identification (human or nonhuman), even as a cautious possibility. These results underline the uncertainty inherent in this type of identification, regardless of the background or the experience of the assessor. Therefore, in this scenario, macroscopic observation alone is insufficient. Indeed, in the field of archaeology, species identification often has to be made based upon the context in which the material was found. Based upon this research, we stress the huge risks in species evaluation on burned and fragmented bones by quantifying the error, and we aim to raise awareness that a more thorough methodology may be required for cases of extremely fragmented cremations.