The Biomedical Ethics Ontology Proposal: Excellent Aims, Questionable Methods

Koepsell et al. (2009) describe an ideal biomedical ethics committee environment with efficiencies such as electronic and universal application forms and consent templates, automated decision-trees, and broad sharing of data. However, it is unclear that a biomedical ethics ontology (BMEO) is necessary or even helpful in establishing such environment. Two features of any applied ontology are particularly problematic in establishing a useful BMEO: (1) an ontology is a description of a domain of reality; and (2) the description is subject to ongoing revision as it is developed through open processes, e.g., the use of a wiki. A BMEO would need to address two main kinds of entities, regulatory definitions and ethical concepts, and is ill-suited to both. Regulatory definitions are fiats and ought to be adopted verbatim to ensure compliance, but in such cases we do not need the assistance of ontologists, and their modes of working (constant revision within open wiki-based communities) might even be counterproductive. Ethical concepts within pluralistic societies are social constructs, not a priori concepts or biological natural kinds, and the prospects of generating intuitive definitions that enjoy broad acceptance across cultures and institutional settings are slim. In making these arguments, I draw from the writings of leading applied ontologists and Koepsell et al.'s own proof of concept.