Classifications and the Law: Doctrinal Classifications vs. Computational Ontologies

Abstract

This paper addresses legal classifications by comparing legal doctrine and computational ontologies. In recent years legal ontologies have attracted a growing interest, not only from knowledge engineers but also from legal scholars. Indeed, several controversial issues arise concerning the elicitation and structuring of domain (legal) knowledge, and legal theory can provide useful insights in this respect. The existing tradition of definition and classification of legal concepts by legal doctrine can provide a source for the extraction and characterisation of concepts to be included in legal ontologies. The question arises as to what extent doctrinal structures can be reused in the construction of legal ontologies, and as to what extent doctrinal analyses can draw inspiration from computational ontologies. Firstly, different theoretical approaches to legal conceptualisation are presented. Special emphasis is put on the tension existing in the law between the possibility of defining legal concepts in abstract and the need to define them in the context of each norm, which is a specially relevant issue for legal ontology building, since it concerns capturing the semantics of legal concepts. Secondly, systems of concepts in legal doctrine are analysed and compared, considering both their structure as well as the kind of semantic relationships they include. Through the examination of several examples taken from a broad temporal span in the history of legal thought, it is observed that the systems of legal concepts in doctrine have different levels of abstraction (from very general legal categories encompassing the whole legal domain, to specialised conceptual systems representing particular domains like private law) and different structures (from is-a hierarchies to more complex functional networks of concepts). Some conclusions are drawn as to the possible mappings between doctrinal systems of concepts and computational ontologies. Further lines of research are suggested to better exploit the possible cross-fertilisation between legal doctrine and ontological research in the framework of knowledge representation for intelligent legal information management.