Melbourne Law School - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemRepresentation and reasoning in law: legal theory in the artificial intelligence and law movement( 1996)Over the past few decades there has emerged a group of researchers who , have used computerised techniques to investigate the nature of legal reasoning. These researchers have formed what is called the 'artificial intelligence and law movement.' Members of the movement have built computer systems which automate legal reasoning, and in so doing have assumed that certain models of legal reasoning are correct. This dissertation argues that in many instances the models relied on by artificial intelligence research are no longer commonly accepted as valid by legal theoreticians. It further argues that until the artificial intelligence and law movement begins to recognise alternative legal theoretical models of reasoning, it is - unlikely to produce accurate, reliable and useful automated legal reasoning systems. The dissertation examines the four main reasoning paradigms in the artificial intelligence and law movement: deductive reasoning, analogical reasoning, inductive reasoning and sub-symbolic (neural network) reasoning. In each of these reasoning paradigms it shows that there is an extensive legal theoretical literature which is largely ignored by artificial intelligence research. It reviews the different models presented by legal theorists in each of these paradigms, in, order to show the limitations of artificial intelligence assumptions about the paradigm. The dissertation reviews a representative sample of artificial intelligence and law implementations in each of the reasoning paradigms, and assesses the type of legal theory implicitly adopted in each It argues that, generally, the models of legal reasoning adopted in each paradigm by artificial intelligence research has been formalistic, static, and mechanical. As a consequence, the implementations have been computationally tractable, but unconvincing in legal theoretical terms. The dissertation shows how alternative legal theoretical models of reasoning may be incorporated into existing artificial intelligence approaches. The dissertation concludes with an indication of how in future artificial intelligence and law research might provide useful models of legal reasoning, and how it might inform legal theory.