The legal fact as a work of art : artificial intelligence and the pragmatics of legal interpretation


University of British Columbia

Date Issued


Document Type



Master of Laws - LLM




Modern legal theory rests on a premise that in revealing a legal concept we are revealing reality. This traditional (Kantian) approach is reflected most acutely in legal positivism. The position of the legal positivist is that of law as constitutive of facts. The fact is our description of relations in the world. We should see immediately that in understanding the fact we should understand how this fact is presented to us. To comprehend the meaning of a fact we must understand the relation of language and reality. The question concerning the legal fact is the question of origin. It comes in response to the question "Why?" For there to be a legal fact there must already be an issue. It is understanding what is an issue and how we express this issue that is the mark of an "expert." An example of the positivist approach to legal reasoning is Ann Gardner [1987]. Gardner is confronted by Hart's positivism that we verify legal norms through legal language. That is to say, that the descriptions found in legal text are what verify our apperceptions of the legal real. She concludes that hard cases cannot be resolved in expert systems. If we accept Hart's premise then Gardner's conclusions are correct. This paper is to reject the positivist premise. This paper sets out legal language as an artifact of the legal institution. If language is an artifact, we must recognize that language is a product of the social world and not a means of representing the social world. From this, it is apparent that we must go beyond the language of the discourse not only to uncover the basis of legal norms but most especially to uncover a basis for building expert systems. This requires that we move from an epistemological basis for knowledge to an ontology of law. The thesis rests on two premises: first that we cannot represent knowledge in computerized legal databases useful for legal reasoning without a practical rationality, and secondly that modern legal jurisprudence, inter alia legal positivism and legal realism, has been unable to explain contradictions within the doctrinal discourse thus it is unable to provide a practical structure of rationality. We conclude that legal positivism cannot be the basis for expert systems in law. The paper is divided into two sections. The first section sets out a philosophical basis for legal understanding. This thinking will draw heavily on Wittgenstein, Heidegger and Merleau-Ponty. It is to put into perspective the conditions necessary for understanding to exist. We examine the social construction of reality. This will lead to a critique of the current mode of representational thinking and the reasoning within institutional practice. In the second section, we will consider the actual computational application of legal knowledge in connection with a "deep structured" approach to legal reasoning. Legal action brings to attention the transparency of human involvement with the world. It is then that we become aware of what we do. Our involvement is an issue. This "shock of recognition" is when meaning is revealed. This enables the Kantian concepts of intentionality and freedom which presuppose our involvement in the world. We will conclude that knowledge representation and expert systems require a sound practical foundation. Positivism proves an inadequate foundation in its attempts to capture the reality of the legal situation. We argue that if an expert system in law is to be of practical value, it must be based on a practical rationality. A theory of actions provides such a coherent basis upon which to develop models of reasoning within computer based systems.

Date Available



For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use




Law, Peter A. Allard School of