Computational Law as an engineering discipline
Русская версия здесь
Currently, there is no widely accepted definition for Computational Law. The term itself also has several related, often almost synonymous forms: AI in Law, Computational Legal, Artificial Legal Intelligence, etc. The literature mostly contains inductive definitions where the notion is recorded as an umbrella term above the list of research and development directions.
Here is the list from International Association for Artificial Intelligence in Law:
- Formal models of legal reasoning
- Computational models of argumentation and decision making
- Computational models of evidential reasoning
- Legal reasoning in multi-agent systems
- Executable models of legislation
- Automatic legal text classification and summarization
- Automated information extraction from legal databases and texts
- Machine learning and data mining for e-discovery and other legal applications
- Conceptual or model-based legal information retrieval
Available data-centric definitions assume Computational Law mainly for digitalization, standartization and legal data logistics: “Computational Law: Expressing Law and Legal Processes as Standard Data Through Interoperable Service Interfaces” (Dazza Greenwood, Lead Scientist at law.MIT.edu). «Actionable law» means availability of necessary legal data in right form, at right time-space and for a specific automated service, which would use it then to execute law as a sort of automated legal process. Many of visible developments of MIT are focused specifically on digitalization of law (a transformation into digital data) of some important legal states: Digital Identity, Digital Assets, Digital Contracts.
Related definitions speak about support for legal tasks using information technologies:
“AI in Law: Applications of advanced information technology to support tasks in the legal domain” (Michael Bommarito, MSE FE University of Michigan). «Computational law is an approach to automated legal reasoning focusing on semantically rich laws, regulations, contract terms, and business rules in the context of electronically-mediated actions.» (Love, Genesereth «Computational Law», 2005)
«Information technology» can mean virtually anything today, starting with printer maintenance, but we will try to identify an important distinction here. The “informational” focus on law implies a deep semantic differentiation of its elements, more than just a transactional and logistic perspective of life cycle of [chained] data blocks. Semantization of law is primarily an identification of regular structures and elements in it, with creation of portable models: logical descriptions (e.g. description logics), statistical projections into a conceptual scheme, ontologies. These models often require rule-based logical reasoning, fact-orientedness.
Climbing up the ladder of the intellect stack, we can propose another important perspective and corresponding definition.
Knowledge is a shared and distributed state of society, which determines its coordinated behavior. In this optics, legal knowledge is a complex and multilayered social agreement regarding legislative regulation of public life. Knowledge supervenes on information and data, thus requiring special theoretical and practical focus.
Epistemization of a law is recognition of actionable knowledge within legal domain, the cognitive determinants of social behavior. And also deployment of such actionable knowledge elements that enable/require desired behavior in target social practices.
Mass digital transactions provide transport of consensual states between loci of social practices, where digital services consume digitized data objects. A variety of logical and ontological models help us understand this diversity and its organization.
A philosophical view on this problem could be useful, but «the philosophers have only interpreted the world, in various ways; the point is to change it.» (Karl Marx, Theses On Feuerbach, 1845) Changing the world in a methodical way is what engineering is all about. Therefore, if we assign Computational Law the task of changing the world, we must define it as an engineering discipline: as a regular activity equipped with a rational method.
Systems engineering has amassed sufficiently strong intellectual tools of a high level of generality to help such a definition take place. At the same time, it is necessary to clearly define the restriction, that most of these conceptual tools are designed to manage systems of interacting 4D physical objects, and for systems of social states that have orders of magnitude greater complexity, a simple transfer and naive use of the former will only complicate matters.
Why did we speak about epistemization of law? Modern system engineering works perfectly with semantics, having accumulated vast experience here, and an engineering definition could solely rest on the semantic level just fine. It appears, however, that a strong engineering definition of this kind of discipline requires a higher level of generalization, than syntactic or semantic approaches could provide.
Ultimately, target system for Computational Law is neither data nor information. The target system is shared and distributed knowledge, since it determines coordinated behavior in using system. For social practices that are the using systems in relation to Computational Law, exactly [regulated, coordinated] behavior of a social agent is what we must consider a key aspect.
Semantic structures, and even more so — data objects, are much weaker determinants of behavior, in a sense that between these and social behavior lies the apparatus of interpretation, axiology and psychology of a social agent. Epistemological focus, based on more general and less volatile states of social consensus, will provide a more stable and manageable picture, because this stays «closer» to behavior.
Computational law: research discipline and a group of knowledge-centric technologies to support law and jurisprudence, with the following objectives:
- representation of legal and other relevant domain knowledge as Turing computable functions;
- analysis, algorithmic inference, and synthesis of legal knowledge;
- presenting output in a form suitable for use by humans or other machines.
automated management of
actionable legal knowledge life cycle
to support social practices
This definition encompasses the following systemic categories:
- the target system: actionable legal knowledge
- the class of using practices: jurisprudence or any other social activity requiring legal support
- the enabling system: domain specialization of the input-processing-output structure
Data-centric practices of Computing Law, as defined above, in the context of this definition should be considered as the level of intellect stack, providing technologically-dependent input-output, including logistics. Semantic technologies of Computational Law should be considered as the level of intellect stack, which provides processing: generalization, specialization, analysis, synthesis. Epistemic technologies is an interface with a human, society and the supersocial; architecture of social effects and governance for legal socio-machine systems.