PREFACE
This volume (the first of two) is a selection of papers presented at the International
Conference "Logica, Informatica, Diritto" held in Florence, Italy from 6th to
10th April, 1981 and organized by the Istituto per la documentazione giuridica of the
Italian National Research Council ("Consiglio Nazionale delle Ricerche"). The
Conference was arranged in nine working sessions spread over five days with the active
participation of more than two hundred and thirty scholars from twenty two different
countries. This bears witness to the importance of the topic under discussion and the
interest aroused by the papers presented. With this in mind, I have selected and collected
together the papers in this volume under a title which seems to me to be sufficiently
self-explanatory about the ooals of this Conference.
The fundamental theme proposed for the Conference by the organizers was, from a
philosophical point of view, the following: "Is it possible for computing machines to
"think" juridically and to "reason" symbolically?". In other
words, is it possible (or opportune) to hypothesize that a machine helps a lawyer to
interprete and apply the law, when it is now beyond discussion that the computer has
become an irreplaceable auxiliary tool for obtaining legal information and documentation?
From an operational standpoint this question can also be phrased as follows: is it
possible (and desirable) to construct computer programmes simulating behaviour which can
be called "intelligent" and reasoning which can be called "legal" if
they are observed in a "human" lawyer intent on resolving a legal problem? It is
evident that I have taken the title of this volume "Artificial Intelligence and Legal
Information Systems" from here.
In fact, at least twenty years ago, this fascinating problem of whether computing machines
were also "thinking" machines had already arisen and the extensive literature
written on it and on the related problem of cybernetics is well-known. But it is worth
noting that the debate of the 60's was, above all, a general philosophical and
methodological debate between computer scientists, and, for this reason it never spread to
concrete experimentation and "consumer" projects, nor involved, in any
significant way, social scientists, traditionally interested in the application of
computer science. On the other hand, given the state of development of science and
technology of that era, it could not have been otherwise.
Today the situation is certainly more favourable and I hope the debate opened up by this
Conference and evidenced by these Proceedings will be a milestone in the development of
this research area. The most important recent progress in science and technology has been
in their ability to provide us with valid replies to our questions and efficacious
solutions to our problems. Informatics, mathematics, logic, linguistics and the diverse
modern branches of traditional disciplines, such as artificial intelligence (which, not by
chance, makes up part of the title of this volume), knowledge engineering, operational
research and economic analysis (only to mention a few) have evolved methodologies,
techniques and tools of analysis which are already conditioning and revolutionizing not
only entire areas of knowledge but also vast sectors of production and commerce therefore
touching the very foundations of social life. It is sufficient to mention microprocessors,
the newest revolutionary product of the electronic industry, which will become an
essential component in computers of the so-called "fifth generation". It is easy
to forsee that these will stimulate a new experimentation in and broaden the research
areas covered by this Conference.
It is important and urgent for the legal world to take notice of and to examine
introspectively the possible application of the new concepts that science and technology
put at its disposition and thus to furnish politicians with the necessary elements to make
conscious and responsible choices to help overcome the profound crisis in which, without
distinction, modern societies find themselves in this post-industrial epoch.
The reader of this volume before venturing into the technicalities of the papers
presented, will perhaps be curious to know if the participants gave a single and
definitive reply to the question put by the organizers of the Conference. I believe it to
be beyond my task as Editor of this volume to set out in detail the conclusions of a
Conference as complex and productive of ideas as that of the Florence Conference. I will
limit myself to saying that an answer to the question ("Is it possible for a
computing machine to "think" juridically and to "reason"
symbolically?") was given, in a problematic fashion, as follows:
No - if legal thought (and reasoning) is defined as an activity exclusively and peculiarly
"human". Any such behaviour in machines, therefore, would have to be called
"thinking- and reasoning-like" behaviour.
No - if it is claimed that there is something in the essence of legal thought and
reasoning which is "inscrutable", "mysterious", trusted to something
that cannot be better defined than the "hunch" of a lawyer.
Yes - if it is admitted that the problem is to be tacked through experimentation and
observation, comparing the behaviour of the computer with that of human beings to whom the
terms "thought" and "reasoning" are generally applied.
But I shall leave it up to the reader to draw his own conclusions after having read these
papers.
During the Conference not only was the experimentation which has already reached some
promising results talked about but several crucial problems of a theoretical and
methodological nature within this interdisciplinary research area were also discussed. As
some examples of this experimentation, which is generally limited to particular areas of
law, I would like to mention: a) the use of a conceptual knowledge-base in an enhanced
legal retrieval system; b ) the implementation of systems which assist legal operators in
the drafting of routine legal documents; c) the design of a system for computer-aided
legal analysis; d) a computer-based model of legal decision-making procedures and e) a
computerized consultation system in a particular legal problem domain. Among the
methodological and theoretical problems, I would like to mention: 1) the problem of the
natural language understanding by the computer, i.e. the problem of constructing
programmes to make the computer understand communications in natural language; 2) the use
of deductive logic in problem-solving (and, consequently, a computerization of relative
formal calculus); 3) the problem of inductive inference (looked at from another point of
view, this is the problem of the formation of hypotheses by the computer); and finally, 4)
the problem of heuristic learning, i.e. in other words: how can computers (and how do
people ) learn new heuristic special-purpose and general-purpose methods and rules?
In conclusion, the Conference proved, in particular, to be an occasion for the profitable
exchange of ideas between scholars of different disciplines and for critical reflexion
drawing the attention of those participating to the importance of carrying out experiments
and making valuations on these matters without bias.