Legal ontology vs. legal interpretation

Fernández-Barrera and Sartor (2010) argue that legal concepts depend on the legal norms in which they appear. Because of that, establishing legal concepts in an Ontology# is actually a central part of legal interpretation.

[T]he characterization of legal concepts […] concerns establishing what norms hold in a legal system and thus what norms have to be applied by judges and imposed upon the party. It is not an activity dealing with mere descriptive linguistics: it is rather a central aspect of legal interpretation.

It follows that Legal ontology# is constantly evolving, and it needs to be prepared for legal change. Introducing new legal norms, as well as changing and deprecating old ones, makes static conceptual hierarchies also deprecated.

Fernández-Barrera and Sartor suggest that legal interpretation and ontology engineering should evolve together. Furthermore, legal ontologies should reflect up-to-date rulings of court cases.

This requires that lawyers (and ontological engineers working with legal knowledge) have the ability to continuously adjust their onto-terminological constructions as the law evolves (taking into account the need to implement legal values), and at the same time to make conceptual analyses bear on the interpretation of legal norms and on the soluion of legal cases.

According to Boella and Rossi (2010), it is difficult to create ontologies that work for legal rules independent of time and place. Law is a combination of legal texts and their interpretation and implementation in different jurisdictions. Legal concepts may be organized in different ways in different jurisdictions, and natural language terms that correspond to the concepts may be used in different ways in different languages.

For example, the EU laws are written on the EU level, but implemented by the national legislators in the EU countries. Any ontology that represents EU laws needs at least two levels of representation: the European and the national level. There is potential for misalignments:

  • Term and concept don’t match, due to one or both of
    • EU term is already used in national law for a different concept.
    • EU concept is already used in national law with a different term.
  • EU and national levels use the same term for the same concept, but different legal interpretation.
    • Fernández-Barrera and Sartor (2010) give the following example: “personal data” according to the EU definition only covers data concerning a physical person (a human) while the Italian definition also includes data concerning “legal persons, bodies or associations”.

To model the complexity, Boella and Rossi suggest a multi-layer ontology, with 4 different levels. They don’t have a concrete example nor a picture, so this is pretty abstract, and I may have misunderstood something. (For terms like lightweight and core ontology, see Interoperability between ontologies#.)

Level 1: Lightweight domain ontologies

Lightweight means little formalisation; not much more than a list of terms. These terms are obtained from actual documents. There can be multiple Level 1 ontologies of the same domain; for different languages, or for different time.

Level 2: “Service ontologies”

Would be nice to have an actual concrete example in that paper. Just quoting here: “The second layer (L2) is constituted by service ontologies, enabling the definition of roles and behaviours for agents in charge of executing tasks related to the specific domains considered by L1.”

The ontologies L1 and L2 are built independent of each other. Matching and comparison of L1 and L2 is the job of L3.

Level 4: Core ontology

In this model, I didn’t quite grasp what it’s used for, but here’s a quote.

Conversely, the definition of the roles and behaviours for agents in charge of executing tasks related to the scope of MIFID (L2), without considering the need of matching this ontology with the ontology of L1, gives the possibility to work in L3 with a contrastive comparison. This comparison may allow to discover the misalignments between internal company organisation as shown in L2 and the legal statements about company organisation provided by L1. In this way, the ontology of legal potentialities may emerge in the particular company choices, the black letter rule of the law and the legal interpretation about it. The verification of these potentialities can be done by the orthogonal concepts of L4 that may offer solutions adopted in similar cases by the means of broader conceptual frameworks.

logo