UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Building rule-based expert systems in case-based law Deedman, Galvin Charles 1987

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1987_A6_4 D43.pdf [ 16.39MB ]
Metadata
JSON: 831-1.0077674.json
JSON-LD: 831-1.0077674-ld.json
RDF/XML (Pretty): 831-1.0077674-rdf.xml
RDF/JSON: 831-1.0077674-rdf.json
Turtle: 831-1.0077674-turtle.txt
N-Triples: 831-1.0077674-rdf-ntriples.txt
Original Record: 831-1.0077674-source.json
Full Text
831-1.0077674-fulltext.txt
Citation
831-1.0077674.ris

Full Text

B U I L D I N G R U L E - B A S E D E X P E R T S Y S T E M S IN C A S E - B A S E D L A W by G. C. D E E D M A N L L . B . , University of British Columbia, 1976 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F LAWS in T H E F A C U L T Y O F G R A D U A T E STUDIES (Faculty of Law) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F BRITISH C O L U M B I A Apr i l , 1987 (?) G. C. Deedman, 1987 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Law  The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 D a t e April 29, 1987 ABSTRACT This thesis demonstrates that it is possible to build rule-based expert systems in case-based law using a deep-structure analysis of the law and commercially available artificial intelligence tools. Nervous shock, an area of the law of negligence, was the domain chosen. The expert whose knowledge was used to build the system was Professor J.C. Smith of the Faculty of Law at the University of British Columbia. i i TABLE OF CONTENTS I. INTRODUCTION 1 II. THE PRESENT STATUS OF ARTIFICIAL INTELLIGENCE AND LEGAL THEORY WITH RESPECT TO CASE-BASED LAW 2 III. EXPERT SYSTEMS 4 IV. KNOWLEDGE ENGINEERING 7 V. THE DEVELOPMENT OF EXPERT SYSTEMS 9 VI. THE EMERGENCE OF SHELL SYSTEMS 10 VII. EXPERT SYSTEMS AND LEGAL REASONING 14 VIII. THE ESSENTIAL IMPORTANCE OF UTILITY IN EXPERT SYSTEMS 19 IX. THE BASIC THEORETICAL PRESUPPOSITION OF NERVOUS SHOCK ADVISOR 21 X. TRADITIONAL LEGAL THEORY AND CASE-BASED LAW 22 A. The Problem of Hard Cases 22 B. Different Schools of Legal Thought 25 1. RULE SKEPTICISM 25 (a) American Legal Realism 28 (b) The Critical Legal Studies Movement 29 2. LEGAL POSITIVISTS 31 3. T H E BALANCING OF INTEREST DOCTRINE OF SOCIOLOGICAL JURISPRUDENCE 32 C. The Distinction Between Private Law and Public Law 34 D. The Limits of Traditional Legal Theories 37 E. The Critical Issues of Legal Reasoning with Respect to Building Case-based Expert Systems 41 i i i XI. THE FLEX METHODOLOGY AND ITS UNDERLYING ASSUMPTIONS 45 XII. THE ADVANTAGES OF THE FLEX APPROACH 45 XIII. THE UNDERLYING STRUCTURE OF THE NERVOUS SHOCK ADVISOR 48 A. Introduction 48 B. The Rule of Precedent 50 C. The Rule of Relevancy 52 D. The Ambiguity of the Concept of "Duty" 55 1. T H E EXTENSION ISSUE 57 2. T H E RISK ISSUE 59 3. T H E STANDARD OF CARE ISSUE 60 4. T H E REMOTENESS ISSUE 61 5. THE CONFUSION OF DIFFERENT KINDS OF ISSUES THROUGH T H E USE OF DUTY LANGUAGE 61 E. A Deep-structure Analysis of an Action in Negligence 63 F. The Ambiguity of Duty in Nervous Shock 65 G. Contrary Doctrines of Remoteness 69 H. The Reformulation of Remoteness Doctrines into a Deep-structure Rule 74 I. A Deep-structure Analysis of Nervous Shock 75 J. Preparing to Build a Knowledge Base Using FLEX Methodology 77 K. The Similarity Between the Way Lawyers Reason and the Back-chaining Control Strategy Used in Rule-based Expert Systems 79 L. How The Nervous Shock Advisor Works 84 XIV. THE FUTURE OF LEGAL EXPERT SYSTEMS 87 iv BIBLIOGRAPHY Books 104 Articles 110 APPENDICES A. Sample Consultations 124 B. Knowledge Base 155 C. Case File Index 201 D. Case File 299 v A C K N O W L E D G M E N T S This research was carried out under the auspices of the U B C - I B M Co-operative Project in Law and Computers which was generously funded by IBM Canada Ltd. Additional funding was provided by the Social Sciences and Humanities Research Council of Canada and the Law Foundation of British Columbia. I owe special thanks to a number of people for their help. My principal advisor, Professor J.C. Smith, was an unflagging source of encouragement, advice and assistance. Dean Peter T. Burns and Professor Robert T. Franson, the Co-operative Project Director, were both consistently supportive of my efforts. Doug Arnold, the Project Co-ordinator, provided invaluable technical help and advice, as did Don Johnson, the Computer Systems Technician. Anna Holeton and Er in Dowd helped me greatly in preparing the manuscript. vi I. I N T R O D U C T I O N The purpose of this thesis is to demonstrate that it is possible to build a useful rule-based expert system in case-based law using readily available artificial intelligence technology. I have not taken a traditional approach to the subject. In other words, I do not simply speculate about the pros and cons of building such a system and then reach some conclusion. Instead I have chosen to build a working system myself because that is the most graphic way of illustrating the validity of my assertion. It also reflects my primary concern which is to produce an end product which is of some use to practitioners The fact that I was able to do this proves a more general point; namely, that people outside the field of traditional Computer Science now have access to technology which will allow them to build expert systems if they are prepared to invest some time in learning the use of a very high level computer language and the skills required to debrief the human expert on whose knowledge any system must be based. Since building an expert system necessarily involves a close collaboration between the domain expert and the person who debriefs him and designs the system (ie. the knowledge engineer), this thesis is also unusual in that it was very much a joint effort. I was the knowledge engineer and Professor J.C. Smith of the Faculty of Law was my domain expert. In the thesis, therefore, I frequently write in the second person and 2 talk about "our" beliefs and what "we" did. To do otherwise would be to deny that the whole enterprise was a partnership from start to finish. The theoretical underpinning of the expert system is our conviction that legal decision making can be reduced to a rule-governed activity. Without this as a conceptual starting point, it would have been futile to even attempt to build a legal expert system because there would be no certainty that logical rules could be formulated to predict all possible outcomes. A corollary of this theory is that the methodology used to build the system in question, Nervous Shock Advisor, can be used with equal effectiveness to build other systems in different areas of case-based law. II. T H E P R E S E N T S T A T U S O F A R T I F I C I A L I N T E L L I G E N C E A N D L E G A L T H E O R Y W I T H R E S P E C T T O C A S E - B A S E D L A W The latest research in artificial intelligence (Al) and legal reasoning in the area of case law makes it evident that we have only just begun to explore the possibilities offered by the marriage of this technology and legal theory. The latest efforts at duplicating case-based legal reasoning within a machine are not at all promising. Anne von der Lieth Gardner recently completed an extensive research project and wrote her dissertation entitled, AN ARTIFICIAL INTELLIGENCE APPROACH TO LEGAL REASONING, which is soon to be published. 1 Gardner used case law reasoning as a domain to do work in A l . She attempted the construction of a program for a mid-size mainframe computer, which would duplicate legal reasoning regarding whether a particular set of facts or events would constitute an offer and an acceptance for the creation of a contract. The program was implemented in a dialect of the LISP (List Processing) language, a computer language which has been a major development tool in A l . Her dissertation combined legal philosophy and computer science, with somewhat greater emphasis on the latter, and required an academic background in both disciplines. It "presents a computational framework for modeling legal reasoning." She assumes that:2 Legal philosophy tell us, among other things, that legal rules do not dictate legal outcomes: there is more going on in the decision of a case than ordinary deduction. One extra element is that there is often room for choice about which, of several possible decisions, is to be preferred. Another is that the choice, once made, sets a precedent, which may change the space of choices available in later cases. Her resulting program is consistent with her above assumption in that:3 On some issues, the program reaches an answer deductively, determines that it knows nothing that might defeat the answer, and jumps to the conclusion that on this issue, the result is certain. On other issues, the program concludes that the human decision-maker has some room for choice. She concludes that, "Reasoning about which would be the better choice remains a major research area for the future."* Her over all conclusion is that legal problems are in general too complex and "open textured" to be modeled within an artificially thinking machine. At the level of hard cases a judge has discretion to choose between various alternatives, and it is this process of choice which cannot be machine duplicated. With this rather pessimistic assessment as our starting point, we began to explore the possibility of overcoming the problem by using different tools and a different methodology. III. EXPERT SYSTEMS There has been a recent major development in the field of A l which raises some interesting possibilities for law. It used to be said that, " / / it works, it's not Al", but that is no longer true. A l has moved out of the lab and into the market place. Research in the field has produced a new technology and methodology called "expert systems" which has already proved effective and efficient in many fields. 5 One of its founders, Professor Edward Feigenbaum of Stanford University, has defined an expert system as:6 ...an intelligent computer program that uses knowledge and inference procedures to solve problems that are difficult enough to require significant human expertise for their solution. Knowledge necessary to perform at such a level, plus the inference procedures used, can be thought of as a model of the expertise of the best practitioners of the field. The knowledge of an expert system consists of facts and heuristics. The "facts" constitute a body of information that is widely shared, publicly available, and generally agreed upon by experts in a field. The "heuristics" are mostly private. The performance level of an expert system is primarily a function of the size and the quality of a knowledge base it possesses. Expert systems do not necessarily attempt to model the mental processes of the brain in the machine. 7 The breaking free from this old A l paradigm and dream has permitted the development of new techniques which have tremendous practical value. The transition was one from general, domain-independent methods of solving problems typified by GPS (General Problem Solver), a program created by Newell, Shaw and Simon in 1957, to specific, domain-dependent methods. The paradigm shift was from power-based techniques to knowledge-based ones. In fact, an essential characteristic of expert systems is that the knowledge which they contain is deep, but extremely narrow in its scope. Experience has shown that it is necessary for these systems to be confined to very narrow domains in order for them to be capable of giving advice at a level which is comparable to that at which a human expert operates.9 This change in approach in A l gave new meaning to the maxim "Knowledge is Power". General analytical skills which are not backed by extensive knowledge of a domain are useless when it comes to finding solutions to real problems. A n interesting illustration of this shortcoming involving a human being's reasoning processes is given by former Liberal M.P., Donald Johnson, in his book UP THE HILL. Speaking of his former boss, Pierre Trudeau, Johnson says: He is determined to rely upon his analytical equipment rather than a knowledge base and if he can destroy or wound an argument which he doesn't like with a quick, logical thrust he would rather do so than have to wrestle with the fundamental merits of his adversary's position.10 In fact, we are now really in the Knowledge Age rather than the Information Age. Raw, undigested, unorganized information has limited usefulness. Knowledge, on the other hand, meaning information which has been processed by the human mind and structured in some usable way is powerful. Expert systems can reason with uncertain data and fuzzy concepts. 1 1 They deliver advice rather than deductive conclusions. They embody the heuristic rules of an expert rather than the fundamental premises of a subject matter hard-coded into deductive procedures. Expert systems are made up of a knowledge base, an inference engine, and may be linked to a data base. The knowledge base contains facts (or means of obtaining them) and rules. The inference engine is the piece of software that reasons logically about the rules and facts in order to solve problems. Unlike conventional computer programs, the path taken through the knowledge base is dynamic rather than predetermined. It depends entirely upon the facts provided by the user during the course of an interactive consultation. The inference engine applies the rules in the knowledge base to the facts elicited from the user and "reasons" about them as the consultation progresses. Different facts cause different rules to succeed or fail at any given stage in the consultation. The conclusions reached by the inference engine are stored in a working memory as they are made. This accumulation of intermediate results determines which further rules are invoked and which other facts are sought as the consultation continues. Once all the necessary facts have been obtained in a particular instance, the system gives its opinion which is usually weighted in terms of certainty. The optional data base is linked to the knowledge base and contains additional information that may be accessed at relevant points during a consultation. Inference engines called "shells", which can apply modens ponens ("if...then...") logic to rules and facts, are now commercially available. There are presently a substantial number of expert systems operating in several different fields, and the expert system building industry is booming. Indeed, a sign of the times is what one author offers as a tongue-7 in-cheek definition of an expert system. An expert system is a piece of software that causes TV producers to lose all sense of proportion.12 IV. KNOWLEDGE ENGINEERING A new group of professionals have evolved, called "knowledge engineers", trained to debrief experts. The knowledge engineer extracts knowledge of the domain from the expert and organizes it into some logical overall structure. Once this has been done, the knowledge engineer translates the knowledge into a series of "if...then..." rules which form part of the knowledge base. There are other ways of representing knowledge in expert systems, but these so-called production rules have proved particularly successful. It is difficult to specify exactly what qualifications a knowledge engineer should have. The bottom line seems to be that it is more of an art than a science and that on-the-job experience is essential. Universities are doing very little to train knowledge engineers. "Universities are more interested in research," says Peter Hirsch, director of the knowledge engineering group at IBM's Science Centre in Palo Alto, Cal. Conventional computer science courses do not provide adequate preparation for the task. Knowledge engineering is a black art, demanding different types of skills and knowledge. Interviewing experts is one such skill. This requires the ability to find an automatic reasoning procedure that best matches the expert's thought processes. Practitioners learn the art by doing.13 It has occurred to me that some of the skills which make a good lawyer are precisely those which knowledge engineering calls for. Lawyers need to be skilled at interviewing people - often experts - about a wide range of different subjects and then to be capable of analyzing what they have been told so that the information can be ordered in some logical manner for incorporation into a document or presentation to a court. This also demands a good command of the language; in other words, an ability to use precise declarative semantics. It is no secret that there are far too many lawyers on the job market at the moment. Perhaps knowledge engineering can offer some of them an alternate career. It may well be that universities will eventually offer formal courses in knowledge engineering. In the meantime, the private sector fills the gap and provides the necessary training in the same way as it did before computer science became a recognized academic discipline. A number of companies offer courses in knowledge engineering methodology. 1 4 Even writers who are extremely skeptical about the potential of expert systems and who speak derisively of the label "knowledge engineering", concede that the field is a growing one. To describe the process of designing, building, and testing expert systems and related Al software, the Al community has coined the term knowledge engineering. This pretentious term does have some advantages: it emphasizes that more than programming is involved; it's consistent with the term "knowledge base"; and it makes knowledge engineering sound more glamorous than usual. Indeed, knowledge engineering is widely touted as the most significant new result of computer technology, and it is the basis for a growing number of commercial ventures}5 Rather than deriding knowledge engineering in a conventional fashion, one writer has acknowledged that it is a growth industry while poking fun at it. Using the psuedonym Rock R. Farmer-Taylor (presumably a take-off on Rick Hayes-Roth, one of the founders of knowledge engineering), he advocates an alternative: ignorance engineering. Ignorance engineering was developed in part to solve the "knowledge engineering bottleneck". This problem arose because each knowledge-based expert system had to be hand-crafted over a long period of time by highly-paid knowledge engineers in consultation with scarce and uncooperative domain experts. The costs associated with this approach are not a problem because expert systems can be sold at any price. The problem is that there are too few knowledge engineers and, like engineers the world over, they much prefer to drink beer, so not enough expert systems can be produced and marketed to support the founders and stockholders of Al start-up companies.16 The point is that, regardless of what people may think about knowledge engineering, they are forced to admit that it is a burgeoning field. V. THE DEVELOPMENT OF EXPERT SYSTEMS The first expert system, D E N D R A L , was built at Stanford in 1964. It was designed to determine the structure of organic molecules on the basis of information fed it regarding the mass-spectrum of the particles after the molecule has been broken up by bombardment in a mass-spectrometer. It now out performs all human experts in substantially less time. The same research group began work on another system in 1972 and produced M Y C I N , 1 8 which diagnoses and prescribes treatment for bacterial infections of the blood, and has also in a number of tests out performed human experts. P R O S P E C T O R , a diagnostic system developed from 1978 onwards for prospecting for minerals, was used to discover a hundred million dollar molybdenum deposit which had been missed by expert prospectors. 1 9 There are now a number of functioning expert systems in a variety for fields ranging from M O L G E N which was designed and constructed for giving expert advice in the field of molecular genetics, 2 0 to M E E S a system for teaching economics. 2 1 VI. T H E EMERGENCE OF SHELL SYSTEMS It was from M Y C I N that the shell systems evolved. The developers of M Y C I N realized that it could be emptied of its specific knowledge and that the inference engine could used to reason about problems in other domains. E M Y C I N (the "E" stands for "empty"), the first shell system, was the result. There are now a number of different shells commercially available for constructing expert systems. The one which we used to build the Nervous Shock Advisor, M.1, marketed by Teknowledge Inc. of Palo Alto, California, is a direct descendant of E M Y C I N . Nowadays it makes no sense to build one's own inference engine from the ground up as some institutions have done in the past and some are still doing. Besides the fact that it would be incredibly time consuming and require considerable assistance from skilled computer scientists, there is little to be gained from it given that high quality shells which use standard methods of inference are readily available. It would be like creating your own custom-made word processor in order to write a book. The user of a shell need only create the knowledge base of rules and facts or meta-facts (ie. ways of determining facts). The knowledge base must be written in the special syntax of a computer language which the inference engine can understand. The language is usually very high level and English-like; therefore, with a certain amount of effort, it can be learned by someone with little or no computing background. Furthermore, the knowledge base can generally be created using a standard text editor. The completed knowledge base is "loaded" into the inference engine which can then reason about its contents when a consultation is initiated. The fundamental importance of shell systems is that they give people without any background in traditional computer science the capability of building useful rule-based systems provided that they are prepared to spend some time learning how to use the shell. To a non-computer scientist, it is intimidating to have to master the complex procedural syntax of the more common high level computer languages in order to be able to do something useful with a computer. However, this is no longer a necessary prerequisite when it comes to A l thanks to the advent of shells. What is now of paramount importance is the ability to think creatively and logically at a higher level of abstraction. This avoids the risk involved in expending a lot of energy on learning a means to an end; namely, that the means becomes an end in itself. Speaking about the student of traditional computer science, Joseph Weizenbaum, an eminent teacher of the subject, sounded the following cautionary note. He may so thoroughly commit himself to what he naively perceives to be computer science, that is, to the mere polishing of his programming skills, that he may effectively preclude studying anything substantive. Unfortunately, many universities have "computer science" programs at the undergraduate level that permit and even 12 encourage students to take this course. When such students have completed their studies, they are rather like people who have somehow become eloquent in some foreign language, but who, when they attempt to write something in that language, find they have literally nothing of their own to say?2 While Weizenbaum's analogy makes a good point very forcefully, it is, however, somewhat misleading to equate what has become known as "computer literacy" with literacy in a natural language. The ability to read and write a natural language carries with it access to a culturally rich written tradition which is inaccessible to the illiterate. Compared to natural languages, computer languages are impoverished. They have virtually nothing to offer besides certain very application-specific advantages when it comes to communicating with the machine. Now that the first flush of enthusiasm about computer literacy has passed, serious questions are being raised about its value as an end in itself. The primary emphasis in computer literacy should be in the historical, economic, legal and philosophical areas. Computers must be seen in their historical context - as part of an ongoing technological process. The economic and legal implications of their use are a rich source of material for exploring many important social issues. It would be a mistake to focus on the computer itself, because treating even such a marvelous machine in isolation can only result in superficial understanding. This is a real danger, if computer literacy courses are taught by programmers with little experience in other areas.23 Shell systems allow substance to take precedence over form. They permit the knowledge engineer to focus on capturing the knowledge of the expert and getting a prototype system up and running as soon as possible. That is the strategy which we adopted. Throughout the development of our own system, we were guided by the following piece of very valuable advice. 13 The single most important piece of advice that one can give a model designer is to build a prototype model as soon as possible. Because expert reasoning problems are frequently poorly specified, one needs to have something concrete to view and "lay hands on." It is particularly important for the expert to see something running early. A running program is worth thousands of words from an unformalized interview with the expert. The initial prototype may be crude, will certainly be incomplete, and may contain inaccuracies, but at least it provides a focused point of departure from which the expert can make his suggestions. One of the sacrifices one has to make in building expert systems, is to be constantly told about the weaknesses and flaws in the system that is being designed. One has to expect that major revisions will have to be made in the system, particularly in the early stages. It is sometimes amazing to watch the pace of useful knowledge acquisition accelerate once a prototype model has been built. Instead of abstract suggestions from the expert and the model designers, one has a much more limited basis for converging to a practical system. And to those who point out that such constraints will fatally bias or limit the ultimate system design, the answer is that one must always keep an open mind to the need for making drastic changes of representation, and must be willing to put in the effort and resources to carry out these changes if they prove necessary.2* In fact, as the knowledge engineer, I had no idea just how difficult the foregoing advice would be to follow in practice. A well-recognized side effect of building an expert system is that the expert will often be forced to re-think her or his own ideas on the subject. This can be a very useful exercise in clarification for the expert, but enormously frustrating for the knowledge engineer who must be prepared to modify the original blueprint drastically to conform to the new reality, as is suggested above. However, it is very difficult for the knowledge engineer to resist the temptation to become attached to a model in which a lot of effort has been invested and which "works" as far as the computer is concerned even though the domain expert tells him that it is conceptually flawed. I was aware of a very strong resistance on my part when it came to making fundamental changes in the system. Perhaps this is an innate psychological trait among humans when it comes to any sort of change ; we prefer to stick with things to which we have become accustomed. However, no matter how traumatic for the knowledge engineer, it is critical that the system should truly represent the expert's knowledge and be free of any distortion dictated by expediency. If the expert gets up one morning and sees his world in a whole different light, the knowledge engineer must be prepared to scrap everything and begin again, no matter how distasteful he finds the prospect. The process of building an expert system seems deceptively simple at the outset. In fact, it is a difficult and time-consuming business, even using a shell. Wrestling with the conceptual structure of the system is every bit as demanding as coping with the relentless, unforgiving logic of the computer at the implementation stage. VII. E X P E R T S Y S T E M S A N D L E G A L R E A S O N I N G There are at present several research teams working on A l projects and expert systems in the field of law. 2 5 A number of expert systems perform a simple form of legal reasoning in the area of corporate tax law. 2 6 T A X M A N I and II combine the facts of tax cases and the relevant concepts of the Internal Revenue Service Code to produce an analysis of the tax consequences of corporate transactions through deductive inference procedures. 2 7 J U D I T H , developed in West Germany, gives simple advice in negligence questions on the basis of the German Civ i l Code. 2 8 LDS uses expert systems techniques to give advice on whether a case should be litigated or settled. 2 9 T A X A D V I S O R is an expert system which gives accountancy advice in the area of estate planning. CCLIPS, a system based on the Louisiana Civ i l Code, is a project which seeks to develop drafting techniques capable of producing statutory texts that can be processed intelligently by computers, and thus, to produce a draft of the code which would allow legal consequences to be deduced by the computer from description of facts and the rules. 3 1 N O R M A L I Z E R is a prototype of a rule-based automatic drafting system for drafting rules, legislation, agreements and related documents, 3 2 and Professor James Sprowl has developed a system that uses regulations to draft legal documents. 3 3 As well as the authors of the systems named above, there are also a number of other pioneers in the field of law and A l . 3 4 The only team to actually attempt to construct a rule based-expert system for case law legal reasoning is the team of Richard Susskind and David Gold at Oxford. Incidentally, unlike us, this group has decided to custom build its own inference engine. They are working on a project in the area of divorce law which involves provisions in the Scottish Civ i l Code and some case law. Like Ann von der Lieth Gardner, Susskind is also very much aware of the significance of legal theory for A l and legal reasoning. He writes in his article, Expert Systems in Law: A Jurists prudential Approach to Artificial Intelligence and Legal Reasoning?5 It is beyond argument, however, that all expert systems must conform to some jurisprudential theory because all expert systems in law necessarily make assumptions about the nature of law and legal reasoning. To be more specific, all expert systems must embody a theory of structure and individuation of laws, a theory of legal norms, a theory of descriptive legal science, a theory of legal reasoning, a theory of logic and the law, and a theory of legal systems, as well as elements of a semantic theory, a sociology and a psychology of law (theories that must all themselves rest on more basic philosophical foundations). If this is so, it would seem prudent that the general theory of law implicit in expert systems should be explicitly articulated... After canvassing the field of A l and law Susskind concludes that, "Despite growing awareness and interest in the application of Al to legal reasoning...there has not yet been developed a fully operational expert system in law that is of utility to the legal profession."36 The reason why little progress has been made in law regarding the construction of expert systems when compared with the success of this technology in other areas of knowledge such as medicine and engineering, is that there is a wide divergence of views about the critical issues of legal theory referred to by Susskind. In law there is no consensus about what law is, nor about the true nature of legal reasoning. In the first place a legal expert in a particular area of law is likely not to have any particular expertise in jurisprudence. Merely introducing an expert in legal theory to a project will furnish no solution as the assumptions about legal decision making of the legal expert will not necessarily be the same as those of the expert in the domain of knowledge to be put into the system. The solution adopted by the Oxford group is to find an area of consensus between the leading legal theorists, and use that as the foundation of the jurisprudential assumptions for the system. 3 7 The problem with the consensus method, however, is that it will furnish no theoretical framework for dealing with hard cases as the issue of how hard cases are to be decided 17 is exactly where little consensus will be found between legal theorists, and it is in the area of hard cases when special expertise is required. On the other hand while expert system building presents many challenges for legal theory, it also offers some interesting opportunities by way of testing propositions of legal theory. If someone constructed an expert system in an area of case law, which upon asking the user a series of questions to ascertain the facts, could then give the user a response in terms of the law, and cite the relevant cases for the legal issue, we could assume that "rule skepticism" is wrong, at least for that area of the law. Equally, if an expert system were successfully constructed upon the bases of certain jurisprudential assumptions about legal reasoning, we would have some evidence that those assumptions were valid. The requirements for constructing expert systems in law raise some problems which are unique to this discipline. In other systems of knowledge from which one would wish to design a knowledge base for an expert system there is generally much more of a consensus about the nature of the theory which informs the discipline. Medical, dental, and engineering schools would generally not offer courses on the nature of medicine, dentistry or engineering because the practice of these professions seldom raises the kind of foundational questions which drives one back to basic theory about the discipline. 3 8 Such issues, however, often arise in the legal practice, consequently, these are exactly the kinds of questions which a course in jurisprudence asks about law. In law there is no consensus about what law is, nor about the true nature of legal reasoning, but it is nevertheless needed if legal rationality is to be self-conscious. Furthermore, lawyers have no body of scientific knowledge which lies at the core of their discipline as do the experts in the fields where expert systems have been successfully created. Rather, law rests upon sets of fundamental values about which there is a great deal of societal dissensus.3 9 In the first place legal concepts and rules have areas of imprecision because the process of legal reasoning constantly entails extending concepts to new situations. Thus the concepts, in their application, are imprecise at their outer edges. Expert systems which must function with this type of imprecision are often referred to as "fuzzy systems".40 Expert systems for law, however, will have a further area of complexity in that the experts will have different, and often contradictory theories about the nature of law and legal reasoning. We might refer to systems such as these will be, having this second layer of uncertainty, as "hairy systems".*1 Hairy systems such as legal expert systems will require two levels of expertise. A n expertise will be required in the area of substantive law which is the subject matter of the system, and an expertise in legal theory will also be necessary. The number of experts in specific areas of law, who also have true expertise in jurisprudence or legal theory, as contrasted with a mere smattering of knowledge, will be rather small. It was fortunate that my own expert, Prof. J.C. Smith, is one of that rare breed. One of the prime purposes of this undertaking is to aid the legal expert who lacks jurisprudential expertise to construct expert systems in case-based law, and as well, to enable those experts with some theoretical background, to adapt and more efficiently use their jurisprudential knowledge. VIII. T H E E S S E N T I A L I M P O R T A N C E O F U T I L I T Y IN E X P E R T S Y S T E M S My strongest motivation in undertaking this work was the desire to achieve a result that would be in some way useful to practitioners. After having worked as a litigation lawyer in the real world for almost 10 years, I had acquired a distinctively practical bent. I was disinclined to embark upon a graduate program that would culminate in a thesis that was an exercise in traditional scholarship and nothing more; hence my attraction to expert systems. Expert systems typically work in real-world problem domains, rather than what Al scientists call toy-domains. In a real-world domain, the problem solver applies actual data to a practical problem and produces solutions that are useful in some effective way. In a toy domain, the problem is usually a gross simplification or unrealistic adaptation of some complex real-world problem. The problem solver handles artificial data that are simplified to make the problem easier and produces solutions that are of theoretical interest only.*2 Believing that the ultimate proof of the pudding is in the eating, I decided that I would try to build a useful legal expert system rather than simply speculate about whether one could be built. Moreover, the system itself would have to be capable of providing advice on a par with that which a practising generalist would expect to receive as a result of consulting an expert in a particular area of the law. In other words, a legal opinion about the strength of the case supported by appropriate authorities from the case law. Anne Gardner chose to focus on the problems posed in the law examinations at the end of a first year contracts course. This is certainly not something that any practising lawyer would need to seek expert advice about. Indeed, Gardner concedes that, in a sense, the problems she tackled were "toy problems". However, she does insist that they are not trivial because their solution called for skills which all lawyers require and which, unlike interviewing, drafting, negotiating and trying cases, are not developed on the job. 4 3 While there is merit in what she says, my interest was in creating a system which would be helpful to lawyers who have already acquired the necessary practical skills, but still need sound expert advice from time to time. After all, there is so much to know in law these days, that nobody can be an expert in every area. Indeed, most lawyers have enough trouble keeping up to date with the developments in their own specialty. J.J. Robinette is one of the elder statesmen of the Canadian bar. He is generally acknowledged to be the best all-round counsel of his day, an accomplishment due in no small measure to his legendary capacity for hard work. In a recent biography, a friend of Robinette's makes it plain that nowadays no amount of dedication will enable any one lawyer to achieve the kind of mastery of so many areas of the law as Robinette did. Times have simply changed. Ed Sexton is similarly amazed at Robinette's range and sees it as something unique and special - and dying - in Canadian courtrooms. "We won't see another John Robinette in this country," Sexton says. "That's partly because of the changing times. Law has become so specialized now that every corner has its experts. And clients are getting more sophisticated. It used to be that they just wanted a good counsel. Now they want a good counsel who's an expert in their particular problem. But the other reason we won't see another Robinette is because nobody else is so able in so many fields. ,.."44 In approaching my task of building an expert system, I was ever mindful of the following piece of advice offered to builders • of expert systems. One must also consider very carefully whether the proposed application system would be of significant value if implemented. While this may seem obvious, many systems have progressed to relatively advanced stages, before it is recognized that there is no market for the application system.45 IX. T H E B A S I C T H E O R E T I C A L P R E S U P P O S I T I O N O F N E R V O U S S H O C K A D V I S O R Having examined the available technology, let us now turn to look at the legal theoretical context. Professor J.C. Smith, my expert, and I chose as the domain of our expert system a narrowly circumscribed area of the law of negligence called nervous shock. We decided to call the system Nervous Shock Advisor. The decision to actually build a working system necessarily implied a belief on our part that it was theoretically possible to do so using a rule-based shell. In other words, the theoretical presupposition of the exercise was that, at some level, legal decision making can be reduced to a rule-governed activity. In this respect, our starting point was at odds with Anne Gardner's conclusion that there is no practical way of using a computer to overcome the problem of open texture. We felt that in any given case there is always a better answer when it comes to choosing between alternatives. Furthermore, we believed that that answer can be determined in a rule-governed way. I shall now go on to discuss at some length how our ideas jibe with other theoretical approaches to case-based law. X. T R A D I T I O N A L L E G A L T H E O R Y A N D C A S E - B A S E D L A W A. The Problem of Hard Cases The distinction between hard and easy cases is common in the parlance of jurisprudence. A n easy case is one where the legal answer to the issue is clear and relatively undisputed. One generally assumes that easy cases are settled whenever the facts are clear. Easy cases which reach the point of litigation do so only in order that questions of fact can be resolved according to the legal rules of proof. Once the court has established the facts, the legal decision should be straightforward and easily reached by the application of the law to the facts according to well-recognized patterns of inference. Since easy cases by definition do not raise difficult legal issues, they are very seldom appealed. Hard cases are an inevitable aspect of law. A hard case is one which does not fall under an existing rule of law, or which appears to fall under two rules, the application of which would lead to differing or opposing solutions one from the other, or a case which falls clearly under a rule of law, the application of which would produce an irrational result. 4 6 Since the possibilities for human interaction are so varied and since we can never fully predict the future it is impossible to anticipate and create a useful rule which would anticipate and resolve every potential aspect of human interaction. When the vagaries and vicissitudes of life throw into the focus of the legal system a unique kind of dispute, arising from a new kind of activity, a new technology, 4 7 or a new perspective which creates a case which fails to fall comfortably under the existing rules, the law must deal with it within its existing conceptual structure, and hence we have a hard case for the law. Sometimes the new situation falls quite well under the criteria of the old rules, but the situation is such that if we apply the old rule we would produce a very undesirable result. A n example of this kind of hard case is Riggs v. Palmer 4 8 where the beneficiary under his grandfather's will murdered his grandfather in order to receive his inheritance more quickly. This is a hard case because, while it falls clearly under the law of wills which prescribes that the person so named by the testator in a valid will is to take under that will, we encourage murder if we give a murderer the benefits of his crime under these conditions. It is impossible to foresee all the situations which can arise where applying a rule of law to a case which clearly falls under the rule will lead to an undesirable result because of one new unforeseen factor now present which had never arisen in the previous cases. It is well known and frequently noted that legal principles often appear to exist in the form of contradictory and/or contrary pairs, or that a legal principle can always be stated in two mutually inconsistent ways, one favourable to the plaintiff's case, and the other favourable to that of the defendant. For example, an agreement for the use of land or premises might be classified either as a lease or a mere license, or a person may have the exclusive right to information to a spectacle because it is their property, or not have an exclusive right because the subject matter is not capable of being a property interest. Legal reasoning often is thought to be circular because there appears to be no clear guide to which of two relevant opposing principles or conceptual categories should be applied to a particular set of facts. While the bifurcation of legal categories is a part of the natural process of categorization involved in systematizing any area of knowledge, the indeterminacy between categories, and the choice offered a judge between inconsistent and alternative ways of categorizing, is an inevitable consequence of the adversarial method of litigation. Lawyers on each side of a legal issue will always state a relevant legal principle in such a way as to be favourable to their client's position. The judge will then generally choose the formulation which will fit the way he has concluded that the case should be decided. These various formulations get embedded in the jurisprudence in a bifurcated form. Consequently, it is not unusual for lawyers to f ind that there are two possible, opposing, and relevant principles which could reasonably be applied to the facts of their cases. Generally only hard cases, that is to say cases which have no clear legal answer even when the facts have been established, reach the appeal stage. Since it is the established practice of case reporting to report only appeal cases, or cases at the trial level which raise interesting legal issues, nearly all the cases which make up the law reports are what would be called "hard cases". It is self evident that a useful rule-based expert system must be able to handle or suggest legal solutions to hard cases, if it is to have much practical value. Any comprehensive legal theory or philosophy of law must entail a theory about how hard cases are resolved within the law. In fact, how hard cases ought to be resolved is one of the central issues of jurisprudential dispute which underlies and is interrelated with many other issues of contention such as the relationship between law and morality, whether law is an autonomous or an open-ended system, and the nature and function of rights. One cannot have a theory of hard cases, however, unless there is first a theory about how easy cases are to be resolved. A theory of law, therefore, must entail a theory of legal reasoning which will explain the resolution of both hard and easy cases. B. Different Schools of Legal Thought 1. RULE SKEPTICISM The degree to which legal reasoning is rule based should determine the degree of difficulty we will be faced with in constructing expert systems in case-based law. If legal reasoning is highly rule determined and if most of those rules can be stated in the form of "if...then..." propositions, then legal expertise ought to be as capable of being captured in an expert system in the same way as any other form of expertise. Many lawyers, judges, and legal theorists, however, are convinced that legal reasoning is not highly rule-determined. While there are a variety of forms and degrees of rule skepticism, the majority of the legal profession would probably consider themselves to be rule skeptics of one sort or another. Rule skeptics hold in common the thesis that legal reasoning is not rule governed. Judges, some claim, start with a conclusion which they "reach" on intuitive grounds and then select the premises which will "justify" it. Radical rule skeptics such as the late professor Fred Rodell of the Yale Law School took the position that the entire legal conceptual structure is a morass of technical abstract concepts which have little to do with reality. "Legal words and concepts and principles," he states, "float in a purgatory of their own, halfway between the heaven of abstract ideals and the hell of plain facts and completely out of touch with both of them. And that is why, in the last analysis, the language of The Law is inherently meaningless. 4 9 The legal reasoning process is therefore fraudulent, according to Rodell in that it affects a non-existent context for its terms. Rodell claimed and demonstrated that, given the facts of any case being appealed to the Supreme Court of the United States, and the judges who would hear the case, he could predict with ninety-percent accuracy the decision or outcome of the case. His predictions would be made on the basis of his study of the beliefs, prejudices and preferences of the judges based on his analysis of their decisions in past cases which he had analyzed according to a set of categories such as criminal, civil liberties, religious issues, or the rights of the accused. 5 0 His viewpoint is similar to that of many lawyers who believe that the outcome of a jury case is virtually a foregone conclusion once the jurors have been selected because of their prejudices. Hence the great importance which is attached to jury selection, especially in the United States where, unlike Canada, prospective jurors can be questioned at length by counsel. One can easily see how an expert system could be constructed using Rodell as the expert, and creating rules of thumb which reflect his analysis. (A Rodell-like expert system could be made capturing the knowledge of an expert using methods of analysis similar to his.) Such a system would produce a prediction based on the make-up of the court. It assumes, as does most forms of skepticism, such a mammoth amount of social and personal self-delusion that it is not usefully entertainable. Rodell's theory will not tell one anything about the law, since it assumes that there is no law, or that whatever the judges decide is what the law is. It would, having such skepticism, be valueless for purposes of legal research. If an expert like Rodell could be found today and an expert system built upon the basis of his expertise in predicting the decision patterns of known judges, it would have little commercial value and be of little theoretical interest. Or it would be a fortuitous but mere epiphenomenon. A useful expert system must be able to give predictions without information as to the personal identity of the judge because the decision to litigate or not must be made before the judge's identity becomes available. Secondly, an expert system, to be useful to lawyers, must produce information about relevant cases and statutes which support its predictions In fact, the main value of experts systems to lawyers will lie in their potential as tools of research. (a) American Legal Realism Rule skepticism has taken a number of forms during its history as a stream of American legal theory. Rule skepticism originally appeared in the United States in the form of a school of legal theory known as American Legal Realism. Legal realism in that form was an application of the philosophical school known as pragmatism, and incorporated the beliefs of the pragmatists about human reasoning in general. Pragmatism is the dominant American version of empiricism in that the basic epistemological and ethical premises are that experience is the source of knowledge, and ethical judgments are subjective in that they cannot be verified as true or false. In this tradition the American legal realists generally avoided systematic or formal analysis of law as such. They denied the applicability of logical deduction for legal reasoning, and in particular denied the assumption that law is certain, questioning certainty both at the level of determining facts, and determining the law. According to the realists legal rules are not the major determining factor in reaching a judicial decision, but are mere rationalizations of a decision reached for other (generally unarticulated) reasons. The judge first reaches a tentative conclusion and then attempts to find a premise in the form of a legal rule or rules which will justify the conclusion, thus moving from conclusion to premises. The father of American legal realism itself was Oliver Wendell Holmes (1841-1935) who played an influential and active role in the development of pragmatism as a philosophical school of thought. The most important ideas which he contributed to the realists movement was his theory of defining legal concepts in terms of the actual or predicted operations of the courts, and his belief that factors such as social policy or prejudices often play a more decisive role in the outcome of a case then do the legal rules. According to Holmes: ...a legal duty so called is nothing but a prediction that if a man does or omits certain things he will be made to suffer in this or that way by judgment of the court; and so of a legal right.51 If you want to know the law and nothing else, you must look at it as a bad man, who cares only for the material consequences which such knowledge enables him to predict, not as a good one, who finds his reasons for conduct, whether inside the law or outside of it, in the vaguer sanctions of conscience 5 2 The prophecies of what the courts will do in fact, and nothing more pretentious, are what I mean by the law.53 (b) The Critical Legal Studies Movement There are very few legal theorists writing in the tradition of the American Legal Realists today. The most active contemporary school of rule skepticism is the Critical Legal Studies movement. The CRITS, as they have become known, are the intellectual heirs of legal realism enriched with a Marxist orientation, along with smatterings of continental European thought such as phenomenology, and sometimes a dash of liberation theology, the particular mixture depending upon the particular writer. 5 4 While they share the realist's views of legal decision making, they articulate more fully the underlying prejudices and beliefs which are the true motivating factors of decision, and explain them as the product of the class consciousness of that class from which most judges, legislators, and lawyers are drawn. Duncan Kennedy, a professor at Harvard Law School, believes that the process of being socialized into accepting the values of the legal class begins at law school and continues throughout a lawyer's professional career. His attitude towards judicial decision making is characteristic of the CRITS. A second hierarchy is that of the judicial system, in which judges play the role of tin gods, exacting an extraordinary servility from their court personnel and the lawyers and litigants who appear before them. Judges are free to treat, and often do treat those who come before them with a degree of personal arrogance, a sense of entitlement to arbitrariness, and an insistence on deference that provide an extreme model of everything that is wrong with legal hierarchy. Lawyers are complicit in this behavior: they expect it, and even enjoy the purity of the experience—the absolute character of the submission demanded, with its suggestion of playing a game which is really and truly for keeps. Beyond that, the judicial system is based on the same extreme specialization of function and differentiation of capacities as the hierarchy of the bar and the internal hierarchy of particular firms. All of this deforms the very idea of justice, rendering it at once impersonal, inaccessible to ordinary human understanding and ordinary human practice, and intensely personal, since everything depends,  most of the time, on the crotchets and whims of petty rf/c£a£or5.[Emphasis added] 5 5 If the rule skeptics are right about legal reasoning, and it is the case that rules are irrelevant to the outcome of cases because judges can always construct out of the available legal materials a set of premises which will justify their conclusion, then there is no future for expert systems in law since Rodell's methodology would be effective only in regard to a highly politicized court such as the Supreme Court of the United States where judges are deliberately selected and appointed by the President because of their particular biases and beliefs which are generally known within legal circles. 2. LEGAL POSITIVISTS The two most commonly accepted theories of how judges decide hard cases are the discretion theory entailed in most versions of legal positivism, and the balancing of interest theory which is the explanation generally accepted by most persons who would not consider themselves to be legal positivists. The statement of the discretionary theory which is most widely accepted by the positivists is that of H . L . A . Hart who drew the distinction between the core of a rule, and its penumbra. According to Hart every rule has a core of clear and settled meaning, and a surrounding penumbra where meaning drifts into ambiguity and uncertainty. 5 6 When one is able to state the facts of a case in a form which corresponds to the core meaning of a rule, one has an easy case and the judge has no choice but to apply the rule. Where, however, the facts fall within the penumbra of the rule, the judge is bound by no standard, but must "legislate" the solution. He has, therefore, a discretion as to how the case is to be decided. Hart gives an example of a law which prohibits bringing vehicles into parks. A motor car would fall within the core meaning of the rule, and so would, prima facie, be an easy case. Toy cars, bicycles, motorized wheel chairs, or even a World War II airplane put in as a showpiece, fall within the penumbra of uncertainty and are arguably vehicles. In those cases the judge must "legislate" or create the solution since one is not given within the core meaning of the rule. This issue and its implications for artificial intelligence in law have been discussed at some length by Gardner . 5 7 Where a case falls within the core meaning of the rule, but an irrational result would be produced if the clear meaning of the rule is applied, the positivists' position would be that nevertheless the judge should consider the case to be a clear one and apply the rule in order that the clarity of the core meaning is not eroded. This is what is meant by the positivistic aphorism that hard cases make bad law. Positivism can be seen to be a mild form of rule skepticism. Rules function in easy cases, but not in the hard cases. If the positivist view of legal reasoning is correct then expert systems can only be built to handle easy cases. If so, then expert systems have a somewhat limited utility in law since it is hard cases which are litigated and require legal research by the practitioner. At best they would be a means of familiarizing some lawyers with the conventional wisdom in areas of the law with which they are unfamiliar or of teaching law students the basics of the law as a form of computer-assisted instruction. 3. T H E BALANCING OF INTEREST DOCTRINE OF SOCIOLOGICAL JURISPRUDENCE Sociological jurisprudence developed in the United States, in part, as a response to the nihilism implicit in the rule skepticism of American Legal Realism. It attempted to furnish a theory of legal reasoning for judges to follow in place of the application of rules. According to this theory it is the duty of the judge to weigh and balance the conflicting interests which are at stake in any particular dispute, and to decide the case accordingly. This view differs from that of the positivist view described above because on this view the judge would not have as full a discretion since his or her decision will be dictated by the relative weights of the interest. Rules may be applied where they represent or incorporate a correct weighing of interests, but not otherwise. Nevertheless the judge decides on which interest weighs more, at least in close cases. One popular contemporary and more sophisticated version of this view is that of Ronald Dworkin. Dworkin uses Riggs v. Palmer as a paradigm example of a hard case decided by the application of a principle (a man shall not profit by his own wrong) rather than by a rule. According to Dworkin, proper legal premises consist of more than rules, and include principles and policies as well. The latter, unlike rules which according to Dworkin function in an all-or-nothing manner, must be weighed and balanced. 5 8 While Dworkin believes that every case has a right answer, it is not always reached by the application of rules. Non-positivist positions such as that of Dworkin entail that in hard cases the court should weigh and balance the conflicting interests which are at stake. According to this view the judge does not have a discretion but should rule for the more important of the conflicting interests. Balancing or weighing interests, however, is an unfortunate and misleading metaphor as it suggests the existence of an objective scale or of criteria where neither are to be found. Since Dworkin's version of the balancing of interest model of judicial decision-making furnishes no order of precedence or criterion for the balancing or weighing process, in the final result, judicial decision-making is almost as discretionary as it is under positivism. 5 9 C. The Distinction Between Private Law and Public Law Traditional legal theories are unsuited for the theoretical foundations for expert systems in case-based law in an additional way. Whatever the particular jurisprudential assumptions which the legal expert may hold about the nature of a legal system, a rule of law, and judicial reasoning, they must be derived from the body of the law itself. Legal analysis ought to be the method by which our jurisprudential assumptions are clarified and critically examined. If we seek to build expert systems in case-based law then it should be case-based law and only case-based law from which we seek the theoretical presuppositions which will underlie the knowledge base of the expert system. Case-based law deals principally with private transactions and interactions between individuals. Private transactions and interactions are governed and ordered by private law. Private law is judge made law or case law. Sometimes it has been necessary to alter or reform private law through legislation, in which case the legislature is carrying out a function which should have been carried out by the courts, and the specific piece of legislation incorporates private law principles which systematically fit in with the now modified case law. On other occasions the legislature intervenes in private transactions by imposing on them a requirement of the public good. In this kind of situation the legislation need not be consistent with or systematically fit with the private law. Rather it sits above it as an imperative to which private transactions must conform. For example, the Wills Variation Act, R.S.B.C. 1979, c. 435 places limits on testamentary freedom by making provision for dependents of the deceased to make claims on the estate where they have been excluded from the will. Sometimes, such as in the Napoleonic or Germanic Civ i l Codes, or codes derived from them, the principles of private law receive their authoritative form through legislative actions. Civilians, however, are well aware of the difference between a code and a piece of legislation. Codes are a product of scholarly formulation, while legislation is a product of a political process. The foundations of private law are to be found in two social practices, that of property, and that of contract. Property principles or law protect the means which we use in accomplishing our ends through our actions from wrongful interference. What constitutes wrongful interference is determined by the principles of tort law or delict. Contract law enables us to increase our potential for action and satisfying our own goals by the process of bargaining. Thus the overriding goal or function of private law is to protect and extend our free agency. 6 0 The principles of private law have their origins in Classical Roman law. Classical Roman law is the foundation of all Western private law systems, whether common law or c i v i l . 6 1 Classical Roman law, the common law, and the civilian codes and cases are the development of these principles within the context of the myriad of possible situations of private interaction. "Rights" and "obligations" are integral parts of the conceptual structure of private law. They are moral concepts, but their moral bases lie within and not outside the law. 6 2 Private law must necessarily, therefore, have a foundation in moral principles, and it is these moral principles which furnish a meaning for "obligation". Law is the natural and inevitable expression of this particular set of moral principles. It is, at least in the area of private law, the morality of liberty. One can disagree with this kind of morality. 6 3 Or one can recognize its validity but argue that it is not sufficient. Its most outspoken critics are the Marxists. Those Marxists, however, who truly understand the nature and structure of private law, such as Pashukanis, recognize the necessary relationship between private law and rights and duties. 6 4 Their critique of rights is based on their denial of the validity of the morality of individualism from which the rights are derived. Private law forms a system within and of itself. Its individual parts are systematically related because they are deduced or inferred from a common set of interrelated principles. Private law is a teleological system because its over-all function, goal, or policy (in the sense of the policy of the law) is the extension of human agency. As argued above it is a moral system because it is the foundation of liberty, the equality of the individual, and of fundamental rights. Since the morality of private law takes the freedom of the individual as basic, it is opposed to systems which take other values as fundamental. Nevertheless this morality of freedom is compatible with these other values such as those centering on community interests so long as they enter the system as the result of voluntary choice and are consistent with and supportive of the autonomy of each individual. As a legal system private law forms what is often called by civilians, "the law of obligations". As such it forms a part, but only a part, of the legal systems of sovereign political bodies. As a juridical system within wider legal-political systems, the law of obligations is very similar in terms of basic principles whether or not it is the law of obligations of a Napoleonic or Germanic code or of the common law. D. The Limits of Traditional Legal Theories Legal theorists generally conceive of the law of particular sovereign political jurisdictions (of which the system known as the law of obligations is only a part) as a legal system. The parts of such a system, however, cannot be systematically teleologically related as there are fundamental teleological conflicts involved between the law of obligations and the statute law of legislative schemes. The theoretical presuppositions or assumptions which underlie a substantive area of law which has its foundations in legislation or a legislative scheme such as tax law, social welfare law, or regulatory bodies, are based on a totally different set of presuppositions than the presuppositions which underlie the law of obligations. This reflects the conflict often referred to as between private rights and the public good, or between the values of individualism and the values of community. Legal theorists traditionally have sought a single overriding unified theory of law which will treat the law of obligations as a part. Since there is no systematic teleological relationship between the law of obligations (whether judge made or codified) and the law of legislative schemes, or between the juridical paradigm of law and the political paradigm of law, any unified theory will either have to describe a non-existent ideal political process which would be consistent with private rights, or adopt a non-teleological basis for the systematic relations. This is why the most successful unified analytical theories of law are positivistic such as those of Bentham, Austin, Kelsen, Hart, or Raz. It also explains why the non-positivistic theories of law such as Pound, Laswell and McDougal, Stone, Fuller, Dworkin, or Finnis tend to be non-systematic and consequently non-analytic. There is still another reason why we believe that traditional legal theories furnish an inadequate theoretical foundation for constructing expert systems in case-based law. No useful expert system can be built in the area of case law unless it can be shown that hard cases can be explained in a rule-governed manner. Traditional theories simply do not furnish adequate explanations for how hard cases are decided. If the rule skeptics are right, and it is correct that hard cases are discretionary for the judge, then the building of an expert system in law will be beyond the present 39 state of the art in A l . On the other hand, rule-based expert systems may be a way of testing legal theory. If a viable expert system can be developed which will predict the outcome of hard cases with a fair degree of accuracy, and will produce the relevant cases and statutes for the legal solution, we may have a method of testing the validity of theories about legal reasoning. Even if traditional legal theories can furnish an adequate theoretical foundation for constructing expert systems in case-based law, the question arises as to which theory or theories should be used. One answer to this problem is the consensus model which Susskind has adopted, and has described as follows: 6 5 ...fljt is submitted that the divergence of views within jurisprudence has been unrealistically accentuated by the typical foci of inquiry, in that legal theorists tend to concentrate on the inherently contentious issues while ignoring "straightforward" matters (which themselves may indeed raise insurmountable difficulties for the less capable). There may very well be consensus over many jurisprudential questions that has remained unarticulated on grounds of it being simplistic or mundane. Indeed, it may be in virtue of this presupposed, unifying substratum of concordance that dialogue between the various schools has been possible. For instance, theorists may all agree on the forms of legal argument that are both possible and desirable in the clearest of cases. This unanimity may not be apparent from the literature because "hard cases" and not crystal "clear cases" have invariably been jurists' object of study. If there is such a concurrence of approach in relation to legal reasoning as well as to legal theory in general, then it is a model culled from that harmony that should be implemented in expert systems in law. If there is not, and if these conflicts affect the expert system enterprise, then a model that clashes as little as possible with the ruling theories should be developed. It is currently being endeavored to determine if a consensus theory of law (albeit of mundane and limited application), can be propounded. The conventional wisdom in knowledge engineering circles is decidedly against the use of multiple experts in any single expert system. Experience has shown that to do so produces inconsistencies within the system. A n expert system is simply a means of capturing the knowledge of an existing human expert. Experts typically differ at least on issues that are by no means straightforward. To expect an expert system to reconcile the views of all experts within any given domain is to attempt the impossible. 6 6 Whether or not attempting to find a consensus among legal theoreticians will raise similar problems to that raised by the issue of a single expert versus multiple experts remains to be seen. While it is highly likely that one can find some theoretical assumptions about which most legal theorists would agree, it seems to us that the principal difficulty with the consensus method is that it will not furnish an adequate theory of how hard cases are to be decided, since, as seen by Susskind, jurisprudential disputes tend to center on this question. Thus the most critical issue for expert systems is the very point upon which one is likely to achieve the lowest degree of consensus between legal theories and legal theorists. Furthermore, this approach seems to entail a departure from standard techniques for building expert systems. It is clear that expert systems for law require two kinds of expertise, expertise in an area of substantive law and expertise in legal theory. The ideal expert for a legal expert system would be one who had expertise in both areas, and who had applied the latter to the former. We see a further difficulty with the consensus method. It is our present opinion that it is the jurisprudential assumptions of the particular expert which are important, and not those of anyone else. It is the theory of law to which the expert committed and uses in her or his analysis of the substantive law, which ought to be built into the expert system containing that person's legal expertise. If some kind of consensus theory of law is imposed on the expert, the system may contain contradictions if her or his own unarticulated jurisprudential assumptions are inconsistent with those of the consensus theory. If the expert is an expert only in the area of substantive law, and is not an expert in legal theory, then the expert will need aid in clarifying and evaluating her or his own jurisprudential assumptions. The expertise of the legal theorist would aid the expert in legal doctrine in this process of clarification. In the first place doctrinal experts should clarify their position on rule skepticism. We should never lose sight of the fact that it is the explicit and implicit heuristic, meta-rules, and rules of the expert in the substantive law which must be built into the expert system. If the doctrinal expert is a rule skeptic she or he must have some theory of what drives decisions, if not rules. A n d if the doctrinal expert takes rules seriously she or he must have some idea about how rules function. E . The Critical Issues of Legal Reasoning with Respect to Building Case-based Expert Systems Issues in legal theory such as the relationship between law and politics, legal theory and political theory, the relationship of law morality, and such questions as how hard cases are to be decided, have implications for work in expert systems. There are at least three critical theoretical issues so far as legal reasoning is concerned. These are: 1. Is law an autonomous system of rules? If so, then law can be distinguished from other sets of norms such as morals or public policy. Law will be recognizable in a rule-governed manner, that is to say through rules of recognition. Anything which is not so recognizable will not be law. The idea of a rule of recognition was formulated by H . L . A . Hart. He drew a distinction between the primary rules which dictate human actions in terms of rights and duties or prohibitions, and secondary rules which function to create, change, or determine methods of adjudication of the primary rules. 6 7 A l l of this assumes a theory of rule-governed social practices. 2. Do judges have a discretion in deciding hard cases? If it is true that judges do have a discretion, then they are free to reach any particular decision on that point. That is to say that they are not bound by any particular standard. They are free to choose their own premises. 3. Is law teleologically neutral? If law is not teleologically neutral then only certain purposes or goals will be appropriate for law, and certain others will not. If "legal obligation" is to be a meaningful concept then law must necessarily have a foundation in some kind of moral principle or principles which can give "obligation" a normative as contrasted with a mere descriptive meaning. 6 8 Each of the above questions may be put in the form of propositions which may be true or false. In order for it to be possible to construct expert systems in the area of case law which will have the capacity to deal with "hard cases", the proposition that law is an autonomous system would need to be true. If we are not able to distinguish between what is law and what is not law, then there would be no limits on what could constitute a proper legal argument. We would not have a discrete body of legal material over which an individual could acquire an expertise if any kind of assertion can function as a legitimate premise in a legal argument. It would also seem necessary in order to construct the rules for an expert system that one be able to recognize what constitutes law in a rule-governed manner. Law, in other words, would need to be an autonomous system, and the expert would need to know the systematic relationship of its parts, at least in so far as her or his own areas of expertise are concerned. If we are to be able to construct expert systems in case-based law, the proposition that judges do not have a discretion in deciding hard cases would need to be true. If judges have a discretion then decisions need not be rule-governed, and if decisions cannot be stated in the form of rules, then it would be most difficult to build a knowledge base for an expert system because the factors which direct a judge to exercise her or his discretion in any particular manner would be too complex and numerous. If expert systems for case-based law are possible it would also need to be true that law is not teleologically neutral. Any knowledge base for an expert system in the area of case law, would have to, in one way or another, entail the doctrine of precedent that relevantly alike cases be decided alike. If this is so then a criterion of relevancy or a principle or principles for deciding relevancy would be necessary. It is difficult to imagine a criterion of relevancy which would not in someway be a teleological one so long as "legal obligation" is to remain a meaningful concept. 6 9 If all of the above remains true, then the goals of the law should be considered to be a part of the law, and related to each other in a systematic way. 7 0 It is the teleological aspects of the law which furnish the meaning for the normative legal concepts of "right", "duty", "obligation", and "justice". Traditional legal theories all take positions in regard to one or more of the above three issues which would seem to be incompatible with the creation of expert systems in case-based law. Theories which entail a significant content of rule skepticism are not helpful. Theories which provide for a significant degree of judicial discretion in decision-making are unlikely to furnish an adequate foundation for the creation of expert systems in case-based law. Theories, such as traditional natural law, or sociological jurisprudence which recognize little distinction between what constitutes law and other matters of morality and social policy will also be inadequate. Theories such as the various forms of legal positivism which, while they recognize law as an autonomous system, hold that law is neutral so far as values are concerned, or to put it in a different way, hold that the goals of the law are not to be considered as being a part of the law, will be unsuitable for expert systems building. XI. T H E F L E X M E T H O D O L O G Y A N D ITS U N D E R L Y I N G A S S U M P T I O N S We have called our methodology F L E X (an acronym for Fast LEgal EXpert). F L E X methodology assumes that law is an autonomous system, that judges do not have a discretion in hard cases, and that the goals of the law are a necessary part of the law. It entails a legal theory which is like legal positivism in that it assumes that law is an autonomous system and can therefore be subject to rigorous analysis, but, unlike positivism, it does not need to assume judicial discretion to explain hard cases; and contrary to positivism, it considers the goals of the law to be a part of the law. Like the non-positivist theories, it assumes that law has a necessary teleology, but unlike the non-positivist theories, it is analytical. The F L E X methodology, assumes, therefore, an analytical-teleological theory of law. To the degree that F L E X methodology proves to be successful, we have evidence of the correctness of the assumptions which are entailed in its construction. It is in this manner that work in expert systems and artificial intelligence should shed light on the nature and structure of legal reasoning. XII. T H E A D V A N T A G E S O F T H E F L E X A P P R O A C H The F L E X methodology offers an alternative to rule skepticism by offering an analysis of a system of laws, individual rules of law, and legal reasoning in terms of an underlying structure which may be assumed by but not necessarily expressed at the doctrinal level. The deep-structure approach to legal reasoning is based on the assumption that judges decide even hard cases in a rule-governed manner even when they are not necessarily aware of the meta-rules which govern their decisions. A n analogy can be drawn with the way children learn to speak a language. Most children learn to speak their native tongue, quite correctly, without any knowledge of the grammar of the language. Speaking a language, therefore, can be a rule-governed activity which people can carry out without being consciously aware of the rules. The rules are a presupposed part of the language. 7 1 The analogy can be carried over into law. The thesis would be that the law contains meta-rules for deciding hard cases which lawyers and judges unconsciously exhibit or otherwise adhere to in applying the law. The decision is then justified in the surface level rules and principles which make up the law, and in various ways reflect the meta-rules. Both the necessity for a deep-structure model of law for building expert systems, and the lack of such models in traditional jurisprudence and legal theory has been pointed out by Waterman, Paul, and Peterson, the creators of LDS (Legal decision making system) and S A L (System for Asbestos Litigation). They write: 7 2 Lack of a deep model. The legal area doesn't have a clear and well understood underlying (deep) model of many of the mechanisms involved in decision making. Some domains do have a fairly clear model that can be used to predict activity and explain reasoning....Deep knowledge of this sort is not easy to find in the legal area. This makes the task of developing competent explanation facilities for legal expert systems somewhat difficult.... One might expect that the large body of legal rulings and regulations that have been accumulated and formalized in the legal domain would make expert system development easier. Unfortunately, this is not the case. Instead, this characteristic of the domain, having rules that already exist, has led to trouble. It is our conclusion that traditional legal theories will not be of all that much use in the construction of expert systems in case-based law. In the first place, they generally present a unified theory for law as a whole, failing to draw any significant distinctions between legislation and case-based law, nor do they focus directly on the law of obligations which is the foundation for most case-based law. Secondly, traditional theories are either analytical or teleological, but not both. Thirdly , apart from Hart's distinction between primary and secondary rules, there has been little search for or recognition of deep structure within traditional theories. While this study presents a particular legal theory, it does so because that theory has been sufficiently rich to use as a theoretical basis for creating expert systems. F L E X methodology, however, does not necessarily presuppose the theoretical foundations of the Nervous Shock Advisor, rather we present it as a starting point, and as a set of assumptions which may be a useful starting point for doctrinal legal experts lacking jurisprudential expertise, or for legal theorists who find that their own jurisprudential assumptions are not helpful in constructing an expert system. We will go on to examine a number of different kinds of indeterminacies which seem to support rule skepticism, which appear in the law of negligence, particularly that part which relates to nervous shock. We will show how each of these kinds of indeterminacies were resolved or dealt with in a rule-governed fashion through the use of deep-structure in constructing the Nervous Shock Advisor. XIII. T H E U N D E R L Y I N G S T R U C T U R E O F T H E N E R V O U S S H O C K A D V I S O R A. Introduction Rule skepticism is both an appealing and a popular view among members of the legal profession. One of the reasons why this is so is that there are so many different sources of indeterminacy in legal reasoning. Ambiguous concepts, conflicting principles, contrary doctrines, circular references, and abuses of logic are to name but a few. The nature of some of the kinds of indeterminacies will probably differ from one field of law to another while other patterns of indeterminacy may prevail throughout all parts of the law. There are several different methods by which these various sources of indeterminacy may be described and categorized. 7 3 The fact that there is no agreed taxonomy makes no difference to our analysis. What we have to say will be applicable whatever their source, or however they are described. From our point of view, what is important is that they all have in common the fact that the greater the degree of indeterminacy, the less rule-governed the area of law will be, and consequently the less will be the degree of predictability of outcome for any set of facts raising a legal issue. It is generally conceded that if an area of law is in the main indeterminate, it would not be amenable to developing expert systems.7 4 If expert systems are to be constructed in case-based law, methods must be found for transforming what appear to be areas of indeterminacy into a rule-governed structure. The methods that work best may differ from one area of the law to the next, but in general, there should be a good deal of carry over. Methods that work in one area of the law should at least suggest solutions for another. Most of our examples in this study are drawn from the law of negligence, with an additional two or three from other areas of law. We have set ourselves, as one of the goals of this project, the task of explaining the F L E X methodology in a way that will allow others to apply it in the areas of law of interest to them. The F L E X methodology has been developed for use in areas of case-based law that contain a high degree of indeterminacy. It is impossible to explain this methodology without furnishing a number of concrete examples. For those examples to be meaningful to the reader, it will be necessary to take the reader into the areas of indeterminacy within the scope of the examples, illustrate the indeterminacy and give the details of the method of solution which allows rules to be formulated. The examples are mainly drawn from the common law of the British Commonwealth of Nations. Civilians, however, should have no difficulty recognizing similar problems, which arise in the law of delict. Readers conversant with the American common law will f ind that all of our examples will have counterparts in their own system. Some of the English legal doctrines and cases will even be familiar to them. In several of the examples American cases and doctrines will be used and referred to by way of comparison. B. The Rule of Precedent The principle of formal justice underlies the F L E X methodology. The underlying presupposition about legal reasoning in the area of case law is that legal judgments are universalizable. 7 5 The doctrine of precedent which prescribes that like cases are to be decided alike can be restated in the form of a modus ponens rule which we will call the principle of formal . 76 justice: If it is the case that in any judgment made in regard to a particular situation, that a particular person is or is not legally obligated to do a particular act, then it logically follows that anyone in a relevantly similar situation is or is not legally obligated to do the same act. Thus each case or judgment which holds that a particular person has a legal obligation to do a particular act, instances a rule of law which applies to all other persons in a relevantly similar situation. The universalizability of legal judgments is reflected in the practice of citing precedents from many different legal jurisdictions to show what the law is in the particular jurisdiction of the court, at least so far as the law of civil obligations is concerned. Thus in any civil jurisdiction of any province in Canada, the law of any other province will be cited, as well as decisions from England, Australia, New Zealand, and the United States. On occasions common law courts will even cite civilian codes, or even THE DIGEST. If, on the other hand, an issue arises that requires a court to determine what the law is in a different jurisdiction, the law cannot be established by merely citing cases or precedents. One must bring in an expert to give testimony as to the law in the other jurisdiction. Thus a lawyer in British Columbia may cite English precedents to establish what the law of British Columbia is, but must call an expert witness to establish what the law of Great Britain is. The universalizability of legal judgments is not only reflected in the doctrine of precedent, but is assumed by, or is entailed in many other legal principles and practices. For example: the courts' duty of impartiality; the rule of law (those who create and administer the law are subject to the law); the right to due process; equality before the law; and the reciprocal nature of rights and duties. 7 7 Use of the principle of formal justice often permits us to ignore the legal doctrine which a judge may cite to justify her or his opinion, while at the same time maintain continuity and consistency within the law. This underlying rule of legal reasoning along with the Boolean connectors "and", "or", and "not", can be used to deduce classes of factual situations from decided cases, and link them with legal conclusions. The principle of formal justice furnishes no criteria of relevancy for deciding when a particular factual situation is relevantly like another factual situation. The criteria of relevancy must, however, be teleological and the teleology must be derived from the law itself. 7 8 The overriding goal or policy of the law of obligations made up of property, tort or delict, and contract, is the protection and extension of the agency of the indiv idual . 7 9 The teleology which lies behind the law of negligence is the reduction of inadvertently caused harm to others as a result of people's actions, with as little interference as possible with the freedom of action. C. The Rule of Relevancy The F L E X methodology is based on the assumption that judges do take the doctrine of precedent seriously - that they do seek and try to follow "the law". It assumes that the criterion of relevancy is teleological and that judges do select between alternative lines of precedents in terms of the potential impact of the possibilities of decision as measured in terms of the goals of the legal practices ( ie. contract, property, etc.). This pattern of judicial decision-making can be stated in the form of a higher-order rule that we call the rule of relevancy: 8 0 If a case CI arises having some facts identical with some facts in Precedent PI and If some other facts are identical with some facts in Precedent P2, and If the application of PI would lead to different results from those that would follow the application of P2, then CI should be decided according to the precedent which, if followed, will, when universalized as a rule of law, bring about, because of the presence of the similar facts, the most desirable consequences in terms of the teleology of the legal system. How such a meta-rule can function can be illustrated by a hypothetical example. Suppose facts (Cl) that the plaintiff has been severely injured by a criminal assault, and that he discovers that the defendant, the lawyer of the person who committed the assault, had prior knowledge of the intent to commit the assault. Let us assume further that the basis of the claim of the plaintiff against the defendant is an alleged duty on the part of the defendant to take some action such as warning the plaintiff, or going to the police, to prevent the harm from happening. And finally let us assume that this is a new case for the law, as it must at some time have been, in that there is no precedent directly in point. The plaintiff's case is based on (PI) Tarasoff v. The Regents of the University  of Ca l i forn ia . 8 1 where the defendant psychiatrists was held liable for failing to warn that his patient intended to physically harm the plaintiff when he had prior knowledge. The defendant's case is based on a clearly recognized legal principle (P2) that a lawyer owes no duty to warn others of an intent on the part of their clients to cause pure economic loss to others through a criminal act. PI, P2 and C l are all alike in that they involve a relationship which entails a duty of confidentiality. In all three cases there is a foreseeable risk of harm if some action is not taken. In PI and C l the risk is of physical harm. In P2 the risk is of economic loss. The legal issue is whether the lawyer-client relationship is to dominate or be dominated by the risk of physical harm to a third party, P2 inclining us toward the former, PI to the latter. In the teleology of the law of torts, however, it is quite clear that the prevention of physical injury is given a much higher priority than is the prevention of economic loss. Unless, therefore, the confidential relationship between a lawyer and his client can be seen to be relevantly different from the relationship between a psychiatrist and his patient, the case, CI , should fall under precedent PI rather than P2. It is our contention that while a judge who decides such a case would not necessarily justify her or his decision in these terms, or might have difficulty explaining exactly why, she or he would probably follow the Tarasoff case, assuming it was accepted as correctly decided. The Tarasoff case is a precedent for the ordering of the goal of the prevention of physical harm in relationship to the goal of maintaining the confidentiality of professional relationships. It entails the proposition that preventing physical injury to persons is more important in terms of the teleology of the law, than the maintenance of professional confidentiality. In the absence of a relevant distinction between the two kinds of professional relationship, it is a precedent for the ordering of certain previously not explicitly ordered goals in relation to lawyer-client confidentiality. The current AIDS (Acquired Immune Deficiency Syndrome) crisis has sparked a lot of debate about whether physicians have a duty to warn the sex partners of people whom they diagnose as carriers of the virus. 8 2 Since people who contract AIDS face virtually certain death, the Tarasoff case would undoubtedly apply. D. The Ambiguity of the Concept of "Duty" The F L E X methodology assumes that judges decide cases in accordance with an underlying goal-set revealed by teleological studies. These goals are systematically related. These teleological relationships are a part of what constitutes the "system" part of the concept of a "legal system". Expert systems in case-based law must take into account this teleology. They should not accept legal doctrine at face value. The law of negligence, particularly that part centering around such concepts as "duty" and "remoteness" are just such areas where legal doctrine may not be trusted. This is why it makes an ideal area of the law for testing some of the assumptions which underlie the F L E X methodology. This is why an alternative analysis is needed. The law of negligence, as a number of the following examples will show, is generally considered to be one of the most indeterminate areas of the law. If it is possible to build expert systems here, it should be possible to construct them in a good number of other areas of case-based law. One of the best examples of the indeterminacies which plague the law of negligence is the concept of duty of care. The source of the indeterminacy is the multiple judicially determined meanings which the concept has been given in the law of negligence. Within this area of the law "duty" has at least four different judicially determined meanings or functions. 8 3 The reason for this ambiguity lies in the historical distinction made within the common law between questions of fact and questions of law. Questions of fact were to be decided by a jury, and the resolution of questions of law remained the prerogative of the judge. Where a judge was reluctant to leave a particular issue to the jury, he could avoid doing so by formulating the issue as a question of law. One way of transforming the issue into a question of law was to state it in terms of whether or not the defendant owed to the plaintiff, or was in breach of, a duty of care. A negligence case can raise a variety of quite different kinds of issues, many of them of a highly factual nature. When judges in the past wanted to keep some of these issues from the jury, and did so by framing them as issues of duty of care, a variety of issues were then stated in the same language framework—Did the defendant owe the plaintiff a duty of care? The law of negligence deals with the question of unintentionally caused harm. If the harm suffered by a plaintiff was intended by the defendant, then the proper cause of action would lie in some nominate tort such as assault or battery, but not in negligence. The law of negligence does not place a prohibition on certain types of activities as do the nominate torts like assault, battery, defamation, or false imprisonment. The law of negligence prescribes that whatever actions we may take, we must carry them out in such a way that we do not cause harm to other persons: it prescribes an adverbial rather than a verbal prohibition. 1. THE EXTENSION ISSUE Since failing to act does not cause harm, we generally need not concern ourselves with the consequences of not acting. While a failure to act may not be moral, it generally does not entail civil liability. A n individual is generally not liable, in the absence of a special duty such as might lie in contract, for a failure to rescue a person in danger, since the person who fails to rescue is not the cause of the harm. If, however, the person who fails to rescue caused the danger in the first place, then that person might well be liable. If the failure to act is in the context of a wider action, and changes the nature of that act so that it has harmful consequences, such as failing to put one's foot on the brake while driving a car, then liability would lie for causing the accident. The distinction in the law of negligence between causing harm and failing to prevent harm from happening is often referred to as the distinction between misfeasance and non-feasance. The legal issue is often stated in the form of the question: Did the defendant owe the plaintiff a duty?—which would mean—Did the defendant owe the plaintiff a duty to perform the act which he or she did not perform, and which, if it had been performed, would have prevented the harm to the plaintiff. The law of negligence does not require us to take care in regard to all possible harmful consequences of our actions, even though they may be quite foreseeable. Nearly every conceivable human action may have potentially harmful consequences for other people. If the law were to attempt to impose liability for all foreseeable harmful consequences of all our actions, action itself would become intolerably restricted. There are several different kinds of harmful consequences which may follow action. The harm might be physical injury to persons or property, emotional suffering, or pure economic loss. Most human actions have economic consequences for other people. To prohibit the causing of pure economic loss to other people as a result of our actions would intolerably restrict action itself. Since, for instance, any business venture may well have negative economic consequences for some competitors, commerce and markets would not be possible if people were to be held liable for all of the economic consequences of their actions. Physical injury to persons or property, however, is not the general and necessary consequence of most action. Actions can generally take place without such consequences following. Therefore the law can require people to take care not to cause physical harm to others without unduly restricting action. One is generally, therefore, liable for the physically harmful consequences caused by one's actions, but not generally liable for the purely economic consequences. The issue of whether the law requires a person to take care in regard to a particular set or kind of consequences is also often stated in the form of the question: Did the defendant owe the plaintiff a duty of care?—which would mean—Did the defendant owe the plaintiff a duty to take care not to cause the particular kind of loss suffered?—for example, a pure economic loss. 2. T H E RISK ISSUE The answer to the above question is determined within the law by the teleological considerations which underlie the structure of rules making up the law of negligence. The basic goal of the law of civil obligations is the facilitation of human action. The law of torts protects freedom of action from "wrongful" interference. The law of negligence imposes a standard of care on agents to require them to take reasonable precautions when acting so as not to interfere with the actions of other persons. That standard must not be so high as to obstruct action. A standard can be imposed regarding the prevention of physical harm without seriously limiting action. Therefore a duty of care is owed regarding the prevention of physical harm when we act. A standard cannot be imposed preventing economic loss to others when we act. Consequently, in general, no duty of care is owed to prevent economic loss to others when we act. Thus teleological considerations which are to be found within the law are the basis for the resolutions which the law produces. The law of negligence imposes liability only for harm caused through or by our actions which we could have reasonably foreseen, and consequently could have prevented. Reasonable foreseeability of harm is, therefore, a necessary condition for responsibility and liability. We cannot be held responsible for harm which we could not reasonably foresee and, therefore, could not have prevented. In every negligence case, therefore, the question will arise as to whether the defendant could have reasonably foreseen the harm which resulted from her or his action. The answer to this question often takes the form of a conclusion that: The defendant owed no duty of care to the plaintiff—because there was no foreseeable risk of harm. Where there is a foreseeable risk of harm as a potential result of a particular act, then a duty of care arises to see that the act is carried out in such a way that the harm does not materialize. The proximity of the plaintiff to the defendant is often a critical element in ascribing a duty in this sense of its usage. A postal clerk would owe no duty to handle a package with special care if he could not foresee that the package contained nitroglycerin. 8 4 3. T H E STANDARD OF CARE ISSUE Negligence is a normative concept in the sense that a person is only judged to be negligent if their conduct has fallen below a certain standard of care: thus the use of the qualifier "reasonable". The standard of care is the norm of conduct which ought to have been followed. Even though an agent may have been non-negligent with respect to foreseeability, the agent may be negligent with respect to the standard of care which ought to have been taken in the light of the foreseen risk Even if a certain action entailed a foreseen risk of harm, the precautions which the defendant took in regard to it, might well have been reasonable, or, in other words, have met the standard of care. The standard of care sets the duty. The defendant owes a duty to meet that standard, and if the standard has not been met, then the defendant would be said to be in breach of a duty of care. A defendant who maintains a cricket field could be found by a court to owe no duty to a plaintiff struck by a cricket ball, to maintain a sufficiently high fence to prevent balls from leaving the field where the probability is that balls would very seldom be struck that high or f a r . 8 5 4. THE REMOTENESS ISSUE The harmful consequences of a negligent act are often themselves conditions for further harm. As a result of an accident caused by the defendant's negligence the plaintiff might require medical treatment. The medical treatment may produce further harmful complications which in turn may cause a further economic loss. Each particular harm must be dealt with separately. If, for example, the defendant, while driving negligently, hits the plaintiff and breaks his leg, and while the plaintiff is lying on the road, a third person steals his wallet, there would be no question of the defendant's liability for the broken leg. However, an issue would arise as to whether the defendant owed a duty of care in regard to the wallet. As a result of wearing a cast, the plaintiff might later fall and break his other leg. The issue of whether the defendant would be liable for the loss resulting from the second break would generally be phrased in terms of duty of care language. 5. THE CONFUSION OF DIFFERENT KINDS OF ISSUES THROUGH THE USE OF DUTY LANGUAGE Where, for example, an experimental laboratory negligently allows a highly contagious virus to escape which results in the death by disease or by governmentally-required slaughter of all the cattle in a particular area, and an auctioneer suffers economic loss as a result of the closing of the cattle market, the legal issue which would arise would be put in terms of whether or not the defendants owed the the auctioneer a duty of care in regard to his loss of business profits. 8 6 This might mean, however, one or more of several different things: 1. It might raise the issue of whether or not people are required to exercise any care at all to prevent others from suffering an economic loss. 2. It might raise the issue of whether or not the economic loss was foreseeable. 3. It might raise the issue of whether, if the economic loss was foreseeable, the defendant did everything which she or he ought to have done to prevent the loss, or in other words whether the defendant met the required standard of care. 4. It might raise the issue of whether the economic loss was too remote from the original negligence relating to the spread of the infectious disease for the courts to award recovery. The essential issue in this case, given that the escape of the virus was due to the negligence of the defendant, is one of remoteness. A finding of negligence is a necessary condition for an issue of remoteness to arise. Physical harm to some can often result in economic loss to others. Where the line for recovery is to be drawn must be determined by teleological considerations. It must be sufficiently wide as to affect people's behaviour by inducing them to take care not to be negligent, but sufficiently narrow so as not to make the price of action so high that action itself is discouraged. Because of the ambiguity of the concept of duty of care, one would not want to use it as a primary term in constructing the knowledge base of an expert system. Rather, one would want an analysis of the essential elements or issues in a negligence action which would be stateable in language independent of the concept of duty of care, but which could be related to it so as to justify outcomes with cases which use the standard duty language. E. A Deep-structure Analysis of an Action in Negligence The following analysis of a cause of action was developed in L I A B I L I T Y I N N E G L I G E N C E as a result of the domain expert's attempt to make some sense of the law of negligence for law students. A n action in negligence is broken into seven steps:87 1. The extension issue: Will the courts extend a standard of care to a particular kind of loss or activity? 2. The risk issue: Did the conduct of the defendant create a reasonably foreseeable risk of harm? 3. The standard of care issue: What is the standard of conduct which the courts will impose in respect to 2? 4. The negligence issue: Was the actual conduct of the defendant below the standard of care? 5. The causal issue: Was their a cause-effect relationship between the actual negligent conduct of the defendant and the loss or damage suffered by the plaintiff, or was the loss due to some other event? 6. The remoteness issue: How far down the chain of causation wi l l the courts impose liability when the claim was occasioned by an act of negligence? 7. The damages issue: How are the damages to be measured? The concept of a duty of care is used in the law of negligence to deal with four different kinds of legal issues, the extension issue, the risk issue, the standard of care issue, and the remoteness issue.88 Consequently these different issues are constantly being confused in the cases and literature. In particular, issues of extension and issues of remoteness are seldom clearly differentiated. In constructing an expert system in an area of law which on the surface appears to have a high degree of indeterminacy, there is one cardinal rule which should be followed, and which we have complied with in constructing the Nervous Shock Advisor. That is: Do not build into it either the doctrinal legal rules of the subject area, nor use the legal concepts found in the doctrine i f they are the source or the subject matter of the indeterminacy. One must use an analysis which is independent of the legal concepts where those concepts are ambiguous or have been the subject of conflicting judicial interpretation. If a concept is clear within the law, and the cases merely refine and develop the concept then there is no reason why that concept cannot be incorporated into the rules which lie at the heart of the expert system, and if the rules of law in a particular area are relatively clear and straightforward, then one should be able to incorporate them and their correlative legal doctrine directly into the knowledge base. If, on the other hand, the concept is the subject of alternative or conflicting judicial interpretation, as is often the case in case-based law, one should seek an alternative conceptual structure for the foundations of the knowledge base. The futility of attempting to reconcile all the cases with one another if they are taken at face value inspired the poet Tennyson to write the following lines. ....The lawless science of our law, That codeless myriad of precedent, That wilderness of single instances.89 Most lawyers have shared Tennyson's sentiments about the common law at one time or another. F . The Ambiguity of Duty in Nervous Shock Where the subject area of the law that will be the object of the expert system contains basic concepts that have several different meanings, the area of law requires an analysis whereby these different meanings can be separated and clarified. The Nervous Shock Advisor is based on the above seven step analysis of a cause of action in negligence. It furnishes the foundation framework upon which the knowledge base was built. The knowledge base is linked to a data base which produces the right cases which will furnish both the precedents and the traditional doctrinal level of discourse which is generally used for arguing the legal issues. In the case of nervous shock, the issues are generally argued in terms of duty and f oreseeability. Consider where, for example, a mother is watching her child on a roller coaster, and the car, as a result of a negligent design, leaves the track and the child is seriously injured in the fall. As a result of witnessing the accident, the mother suffers psychological trauma which prevents her from carrying out her profession for six months. The legal issue that would arise would be put in terms of whether or not the defendant owner and/or operator of the coaster owed the mother a duty of care in regard to her nervous shock. This might mean, however one or more of several different things: 1. It might raise the issue of whether or not the defendant, in the particular circumstances, is required to exercise any care at all to prevent others suffering from nervous shock. 2. The question might raise the issue of whether or not the nervous shock was foreseeable. 3. It might be asking the question of whether the nervous shock was too remote from the original negligence relating to the design of the coaster, for the courts to award recovery. The law of nervous shock furnishes a paradigm example of the confusion of issues of extension with the remoteness issue. As stated by one judge, "no absolutely clear picture emerges and many of the judgments speak with different voices."90 A l l of the judges repeat the litany that a plaintiff cannot recover from a defendant for nervous shock unless the defendant owed the plaintiff a duty of care. Some of the judges, however, treat nervous shock as a cause of action separate from the negligence which produced the harm which caused the shock, by using duty of care language to ask whether the defendant owed a duty to the particular plaintiff who suffered the nervous shock, to take care not to produce nervous shock. 9 1 One judge phrased the question in this way: The first question which arises is whether nervous shock is a substantive tort or whether a particular instance of damage flowing from a particular tort. If the latter, recovery depends upon the question of remoteness. If the former, recovery depends upon a breach of duty. Legal writers and commentators have expressed contrary views, but in my opinion the authoritative view is that nervous shock, other than that flowing from a physical injury suffered by a claimant as a result of a negligent act, is a substantive tort. This then poses the problem of what is the duty the breach of which gives rise to a claim for damages for nervous shock.92 Other judges see the issue as one of remoteness. This perspective of the subject is taken by Lord Denning where he states:93 / cannot see why the duty of a driver should differ according to the nature of the injury. I should have thought that every driver was under a plain duty which he owed to everyone in the vicinity. He ought to drive with reasonable care. If he drives negligently with the result that a bystander is injured, then his breach of duty is the same, no matter whether the injury is a wound or is emotional shock. Only the damage is different...If you view the duty of care in this way, and yet refuse to allow a bystander to recover for shock, it is not because there was no duty owed to him...but simply because it was too remote to be admitted as a head of damage. A different result is reached by viewing the driver's duty differently. Instead of saying simply that his duty is to drive with reasonable care, you say that his duty is to avoid injury which he can reasonably foresee, or, rather, to use reasonably care to avoid it. Then you draw a distinction between physical injury and emotional injury, and impose a different duty on him in regard to each kind of injury, with the inevitable result that you are driven to say there are two different torts—one tort when he can foresee physical injury, and another tort when he can foresee emotional injury. I do not think that that is right. There is one wrong only, the wrong of negligence. The contradiction between the above two views runs throughout the nervous shock cases. Often it is not clear from the words of many of the judgments just which view the judge is taking, or whether the judge is clear about the distinction in her or his own mind. The above analysis of the law of negligence, which underlies the construction of the Nervous Shock Advisor, dictates that nervous shock raises an issue of remoteness, since in the standard case of nervous shock, the shock is the result of a negligent act which causes some other physical injury or risk of such injury. Extension, risk, and a departure from a standard of care are generally assumed before a nervous shock issue can even arise. While the assumption that nervous shock cases give rise to a remoteness issue is an essential one for the organizational structure of the knowledge base and the data files, it would in no way hamper a user who wished to argue for the opposite point of view. The Nervous Shock Advisor does not furnish the user with legal arguments, but only the cases from which the argument should be made. G. Contrary Doctrines of Remoteness To categorize nervous shock as a remoteness issue, however, does not take one very far. The law relating to remoteness of damages arising from negligent acts is a paradigm example of bifurcated doctrines. For many years in both the United States and the Commonwealth there have always been two mutually inconsistent tests for whether or not a particular damage was too remote-- the foreseeability test, and the causation test. The leading American decision which best illustrates the two different approaches is the decision of the New York Court of Appeals in Palsgraf v. Long Island  Railway C o . 9 4 A servant of the defendant railway company, in that case, while attempting to aid a passenger boarding the train, negligently dislodged a package which fell on the rails. The package contained fireworks which exploded when they fell. The shock of the explosion caused a scale to fall over on to the plaintiff, causing her injury. The majority judgment was written by Cardozo C .J. who applied the foreseeability test. He held that the defendant could not be held liable to the plaintiff unless he was in breach of a duty owed to her. Since it was not foreseeable that the package contained fireworks which would explode and cause the scales to fall, no duty to her was owed. The conduct of the defendant's guard, he wrote, if a wrong in its relation to the holder of the package, was not a wrong in its relation to the plaintiff, standing far away. Relatively to her it was not negligence at all.95 The risk reasonably to be perceived defines the duty to be obeyed.96 Andrews J., in his dissenting judgment, held that once a person is found to be negligent, they are liable for all the damage of which their negligence is the approximate cause, whether or not that damage was foreseeable. He wrote: 9 7 As we have said, we cannot trace the effect of an act to the end, if end there is. Again, however, we may trace it part of the way....What we do mean by the word "proximate" is that, because of convenience, of public policy, of a rough sense of justice, the law arbitrarily declines to trace a series of events beyond a certain point. This is not logic. It is practical politics. For many years a parallel to these two approaches could be found in the law of England which was also followed, in this regard, in Canada, Australia, and New Zealand. The analogue of Andrew's test of approximate cause in Palsgraf was the direct cause test of the English Court of Appeal in In Re A n Arbitration Between Polemis and Furness. Withy & Co. L t d . 9 8 In that case the owners of a ship brought an action against the charterers of the ship whose servant had negligently dropped a plank into the hold of the ship causing an unforeseeable chain of events in which a spark ignited benzene fumes which caused an explosion and fire which destroyed the ship. Bankes L.J . described the two conflicting tests, stating: 9 9 According to the one view, the consequences which may reasonably be expected to result from a particular act are material only in reference to the question whether the act is or is not a negligent act; according to the other view, those consequences are the test whether the damages resulting from the act, assuming it to be negligent, are or are not too remote to be recoverable. Warrington L . J . stated: 1 0 0 The presence or absence of reasonable anticipation of damage determines the legal quality of the act as negligent or innocent. If it be thus determined to be negligent, then the question whether particular damages are recoverable depends only on the answer to the question whether they are the direct consequences of the act. The House of Lords, in the later case of Hav (or BourhilO v. Young applied the foreseeability test, and in a judgment in every way consistent with that of Cardozo C.J. in Palsgraf. while not actually over-ruling Polemis gave a judgment quite incompatible with i t . 1 0 1 The "proximate" or "direct" cause test for remoteness is hardly a test at all because there are no criteria for determining what is direct or proximate. It is quite understandable why judges have felt uneasy in applying it. The Privy Council, on an appeal from Australia, in Overseas  Tankship (U.K.) Ltd. v. Morts Dock & Engineering Co. L t d . 1 0 2 ( now commonly referred to as Wagon Mound No. 1) adopted the foreseeability test and expressly rejected the "direct" or "proximate" cause test, declaring Polemis to be wrong in law. The subsequent history of Wagon Mound No. 1 and Palsgraf are very similar. Both have been frequently cited and almost never actually applied or followed. Prosser and Smith stated in 1952 that since Palsgraf was reported in 1928 it had only actually been squarely applied in one case. 1 0 3 Prosser describes how the foreseeability test has been applied in the United States as follows: 1 0 4 Such piecemeal foresight is a rope of sand, and offers neither certainty nor convenience, as the floundering in the cases seems to show. Here is Learned Hand, a great judge blandly assuring us 72 that it is beyond reasonable anticipation that a barge with which the defendant collides will sink, and will be carrying insurance. Here is Pennsylvania twice asserting that no reasonable man could foresee that any object struck by a speeding train or bus would fly off at an angle and hit a person not directly in its path. Here is Wisconsin, affirming that when a child is run down in the street there is no recognizable risk that its mother, in the vicinity, may suffer mental shock. Here is New York solemnly declaring that the foreseeability of the spread of the fire ends at the first ad joining house. I do not believe these things, I think they are rubbish. At the other extreme is another New York case, finding it all foreseeable when a collision forced a taxicab over a sidewalk and into a building and loosened a stone, which fell on a bystander and killed her, while the taxicab was being removed twenty minutes later by a wrecking car. There is also Texas, which had no difficulty at all in foreseeing that a mud hole left by a defendant in a highway would stall a car, that a rescuer attempting to tow it out would get his wooden leg stuck in the mud and that a loop in the tow rope would lasso his good leg and break it. Illustrations might be multiplied, as every negligence lawyer knows, but surely these are enough. Similar examples can be found throughout the common law world. In Canada, for instance, one court found that it was not foreseeable that a mother would suffer nervous shock when told of the death of two of her children and the severe injury of the third in a motor vehicle accident, 1 0 5 while another court found that it was reasonably foreseeable that when a driver negligently losses control of her vehicle when attempting to avoid a cat, she would run into a wire fence; that the impact would pop out the staples holding the wire; that two weeks later dairy cows would come along and eat the staples along with the grass, and contract "reticulitis" or "hardware disease" and would have to be slaughtered. 1 0 6 No sooner was Wagon Mound No. 1 reported than courts began to retreat from i t . 1 0 7 Just five years later, in Overseas Tankship (ILK.) Ltd. v. The Miller Steamship Co. Ptv. Ltd. (now commonly referred to as the Wagon Mound No. 2). a case arising from exactly the same set of facts, only involving a different ship, the Privy Council subtly shifted the foreseeability test from reasonable foreseeability to foreseeable as possible. 1 0 8 At the level of the legal language used by the courts in both cases the courts used a test of foreseeability of the damage or injury. If the injury was foreseeable then it is not too remote, and if it was not foreseeable then it is too remote. There are, however, two versions of the foreseeability test. If the court decides to find the particular damage too remote, it can justify the decision in terms of the Wagon Mound No. 1 test of reasonable foreseeability because the kind of damages which give rise to an issue of remoteness are seldom reasonably foreseeable. If they were, they would be a part of the initial risk of harm on which the judgment of negligence rests, and a question of remoteness would not arise. If the court finds the particular damage not too remote, it can justify the decision in terms of the Wagon Mound No. 2 test of foreseeable as possible because this will cover almost any kind of injury or damage which can follow negligence. Consequently experts in the field of negligence consider the law to be basically indeterminate and, therefore, unpredictable. 1 0 9 Commonwealth courts do not admit that they have available two conflicting tests of remoteness, and in the academic literature concerning this topic most authors will be found to consider remoteness issues to be a matter of policy, or within the discretion of the judges. 1 1 0 Thus remoteness of damages is an area which appears to be one of the most discretionary and the least rule-governed of any part of case-based law. H. The Reformulation of Remoteness Doctrines into a Deep-structure Rule On the basis of the theory of negligence which was derived from the theory of action set out in L A W A N D I T S P R E S U P P O S I T I O N S , a deep-structure rule was clarified which reconciles the two inconsistent Wagon Mound decisions. It takes the form: Physical damage to persons or property resulting from a negligent action are not too remote if they are reasonably foreseeable in the particular, or are one of a reasonably foreseeable class of injuries. In particular it is reasonably foreseeable that: 1. Dangerous activities when carried out negligently create a wide variety of particular kinds of risks of harm; 2. Injury to persons can also result in further damage from particular susceptibilities or medical complication to the injured person or to others: and 3. The creation of a risk invites rescue. It can be concluded from the decided cases that as a general rule: 1. No physical injury or property damage caused by a motor vehicle accident is too remote. 2. No physical injury or property damage caused by a highly dangerous activity such as that involving the handling of explosives, highly inflammable or toxic substances, high voltage electricity, fire, or use of dangerous objects or machinery is too remote. 75 3. No increased physical or emotional injury resulting from an unusual or particular susceptibility of a person suffering damage as a result of a negligent act is too remote. 4. No medical complication resulting from an injury to a person whether or not the complications are due to an act of a third party, whether negligent or not, is too remote. 5. No physical injury or nervous shock suffered by a rescuer is too remote. 6. Nervous shock resulting in some form of incapacity, suffered as a result of an injury or risk of injury to oneself or someone else is not necessarily too remote. Ninety-two percent of the cases which raise issues of remoteness of damages, decided by the courts of England, Canada, Australia, and New Zealand from the latter part of the 19th century up to the present, conform to the above rules whatever legal doctrine was used to justify the outcome. 1 1 1 I. A Deep-structure Analysis of Nervous Shock Nervous shock, as a class of injury, is foreseeable, even though the specific circumstances are often not foreseeable in the particular. However, the cases show that recovery is not always given. When nervous shock is or is not too remote cannot be justified or explained in terms of foreseeability, even though that is the rationale given for the decision in almost every decided case. Recovery is confined, however, to cases where there is a close family relationship between the person suffering the injury and the person suffering the shock, and to cases where the close relative had a certain degree of visual exposure to the event which caused the shock. These limitations create an exception to the more general rule of remoteness stated above. Nervous shock is one of the foreseeable classes of injury which can follow from a negligent act, but it is subject to the qualification that: 1 1 2 Nervous shock resulting in some form of incapacity, is too remote unless it is inflicted on a near family relative as a result of witnessing or coming upon the immediate or near aftermath of a serious accident involving a family member. Almost all the decided nervous shock cases are consistent with this principle. A n analysis of the decided cases dealing with nervous shock indicates that recovery is allowed only in a sub-class of cases within the general category of nervous shock. Nervous shock, at first glance, would thus appear to be an exception to the deep-structure remoteness rule in that, although the class of injury is foreseeable, not every case within the class is recoverable as not being too remote. There are two necessary conditions to recover in an action for nervous shock: a degree of visual exposure to the shock-producing event, and a degree of close family relationship. While at first such limitations may seem arbitrary, the consistency with which they are applied without ever being expressly articulated in the legal doctrine, would indicate that a teleological factor is a work, limiting the range of recovery. This should take us back to the two underlying policies which set the limits of remoteness, (a) the prevention of harm, with (b) the minimum limitation on freedom of action. Life experience, as it is brought before the courts in the form of legal disputes, show that in general nervous shock is suffered only when these two factors, (1) a visual exposure, and (2) a close emotional or familial relationship is present. While nervous shock is a foreseeable consequence of the creation of a physical risk of harm to people, it is not foreseeable over the class of all persons, and under a wide variety of circumstances as is the case with the other foreseeable categories of injury such as dangerous substances or dangerous articles. A careful analysis of the cases leads us, therefore, to a redefinition of the foreseeable class. In the case of nervous shock, while it is never foreseeable in the particular, it is foreseeable that members of a family sometimes do suffer nervous shock if they are visually exposed to an accident or its close aftermath where someone dear and close to them is harmed. Nervous shock, therefore on a closer analysis turns out not to be an exception to the deep-structure remoteness rule after all since, the foreseeable class cannot be described merely in terms of the kind of injury, but must be described as well in terms of who suffers it and under what conditions. J. Preparing to Build a Knowledge Base Using F L E X Methodology The F L E X methodology consists of the following fundamental steps: 1. Define a narrow and deep universe of discourse for the advisor. 2. Build a data base by gathering all the case law which falls within the universe of discourse as far back in time as desired, and within the selected jurisdictions. Make a short brief of each case. 3. Select the goal of the advisor. 4. Analyze the cases in terms of the traditional doctrine as expounded in the cases and in the legal texts. 5. Organize the case data base according to the initial analysis in terms of traditional doctrine. 6. Ascertain to what degree the traditional analysis can account for the decided cases. 7. If the traditional analysis can, in general, account for most of the decided cases, isolate out the "hard" cases and seek underlying patterns which might account for or explain them. 8. If the traditional analysis cannot account for the decided cases, develop a second analysis which is independent of the conceptual terms of the traditional analysis, in terms of which the data base of cases can be explained. 9. Formulate the second analysis into deep structure rules or principles which would appear to explain those patterns of decisions. 10. Divide and analyze the data base of cases in terms of the second analysis. 11. Restructure the analysis and the deep structure rules until you are left with only a few random lower level decisions which may not fit the analysis. These can be assumed at this point to be wrongly decided. 12. Re-examine the cases in the data base and modify the briefs so that they include the relevant factors of the final analysis. 13. When this is done, the expert then conceives of possible hypothetical examples which might arise, and decides which to link into the system, and connects them to the cases which could be argued to be similar by analogy. 79 At this point the expert and the knowledge engineer should be ready to go on to start drafting the specific rules which will form part of the knowledge base. Standard expert system methodology, particularly that which is used for what is termed "fuzzy systems",113 should be quite adequate, and a good commercial shell such as M.1 should be sufficient. 1 1 4 K. The Similarity Between the Way Lawyers Reason and the Back-chaining Control Strategy Used in Rule-based Expert Systems There is a remarkable similarity between the way that lawyers reason and the goal-directed reasoning strategy used in rule-based expert systems like M . l . The pattern of inference entailed in expert systems starts with the identification of a goal that it wishes to achieve. It then moves back from the goal down paths of inference seeking a path to the facts which will sustain the goal (backward chaining), 1 1 5 or seeks a path which will eventually reach the goal (forward chaining). 1 1 6 Backward chaining is the preferred control strategy when it comes to rule-based expert systems. It can be explained in terms of the following set of rules which might represent the knowledge base of a mini expert system which will advise whether an applicant is qualified for admission to a graduate program: 1 1 7 G O A L = Is applicant qualified for the grad program? (i.e. conclusion = applicant is qualified for the grad program or conclusion = applicant is N O T qualified for the grad program) Rule 1 i f applicant has law degree from an accredited law school in Canada, U.S. or Commonwealth country then applicant has undergrad law degree from a reputable institution Rule 2 i f applicant's overall average in each year of law school >70% then applicant's academic performance was above average Rule 3 If applicant has practiced law in Canada with a high degree of competence for > 5 years then applicant has exceptional qualifications Rule 4 i f applicant has post-grad degree in area relevant to thesis topic then applicant has exceptional qualifications Rule 5 i f applicant's thesis topic is in an area of interest to at least 2 faculty members and at least 2 faculty members are prepared to supervise the applicant then applicant's thesis topic meets faculty guidelines Rule 6 if applicant has undergrad law degree from a reputable institution and applicant's academic performance was above average or applicant has exceptional qualifications and applicants' thesis topic meets faculty guidelines then applicant is qualified for the grad program Rule 7 if_ applicant has N O T undergrad law degree from a reputable institution or applicant's academic performance was N O T above average and applicant has N O T exceptional qualifications or applicants' thesis topic does N O T meet faculty guidelines then applicant is N O T qualified for the grad program The inference engine will scan the knowledge base until it finds the first rule with a conclusion which satisfies its goal. In this case, that rule happens to be Rule 6. The inference engine will then attempt to satisfy in turn each of the premises of Rule 6 which precede the conclusion beginning at the first one. It will then wind its way back through the other rules seeking as sub-goals other premises it must satisfy in order to conclude that each of the premises which comprise Rule 6 are true. If any one of the premises of Rule 6 fail (except those connected by or which must both fail) the overall conclusion of the rule must fail and the inference engine passes on to consider Rule 7. Rule 7 will necessarily succeed if Rule 6 fails because failure to satisfy any of the premises of Rule 6 will satisfy the conclusion of Rule 7. The illustration of back-chaining which follows is more interesting from a lawyer's point of view. It is a somewhat simplified but essentially accurate set of rules which could form the knowledge base of a mini-system which determines whether a client will be convicted of the criminal offence of driving with blood alcohol content in excess of .08%. The back-chaining reasoning pattern would be exactly the same in this case as in the previous example. GOAL = will client be convicted of over .08? (i.e. conclusion = client will be convicted of over .08 or client will NOT be convicted of over .08) Rule 1 if each breathalyzer test taken not less than 15 minutes apart and all tests completed as soon as practicable and all tests completed within 2 hours of detention then statutory time limits complied with. Rule 2 if at least two breathalyzer tests taken and statutory time limits complied with and lowest breathalyzer reading over .08 then the certificate shows blood alcohol content over .08. Rule 3 if copy of certificate handed personally to client and significance of certificate explained to client then proper service of true copy on client. Rule 4 i f clear photocopy of original certificate made or carbon copy compared and found identical then a true copy of original certificate made. Rule 5 i f client told of right to counsel and client understood rights then Charter rights were properly given when client detained. Rule 6 i f Charter rights were properly given when client detained and a true copy of original certificate made and proper service of true copy on client then the certificate of analyses is admissible. Rule 7 i f the certificate of analyses is admissible and the certificate shows blood alcohol content over .08 then it is proved client's blood alcohol content exceeded .08 at the time. 83 Rule 8 if the client was actually driving a motor vehicle or the client was sitting behind the wheel with the engine running then the client was in care and control of a motor vehicle. Rule 9 if the client was in care and control of a motor vehicle and it is proved client's blood alcohol content exceeded .08 at the time then client will be convicted of over .08. Rule 10 if the client was N O T in care and control of a motor vehicle or it is N O T proved client's blood alcohol content exceeded .08 at the time then client will N O T be convicted of over .08. I should emphasize that a real knowledge base is much more complex. The rules would be formulated using the syntax of a very high level computer language rather than the abbreviated English I have chosen for the purposes of the example. The knowledge base would also contain means of generating questions to obtain necessary facts, as well as a whole host of other features. I present these two examples simply to illustrate how a back-chaining control strategy works. When lawyers reason they follow a pattern of inference somewhat similar to back-chaining. The lawyer's goal in any particular case is dictated by the needs of the client. The lawyer starts with that as a given and then seeks means of achieving it. Although lawyers are constrained by the rules of the legal system and bound by a code of ethics, they are essentially intellectual mercenaries. They start with a specific goal such as recovery for a plaintiff in a specific cause of action. The particular cause of action such as a tort will have certain necessary conditions. If, for example, the goal is a successful action in malicious prosecution, the lawyer will recall the necessary conditions such as: 1. The initiation of 2. judicial proceedings 3. which result in damage to the plaintiff's reputation, person, freedom, or property, and 4. which are terminated 5. in the plaintiff's favour, and 6. which were brought without reasonable cause, and 7. which were brought with malice. Each of the above necessary conditions can be made the subject of rules or sets of rules, and each necessary condition can be linked with factual phenomena by rules which are of the form (facts a, b, and c) constitute (necessary condition No. x ). The lawyer will reason back through each to discover if a particular set of facts contains all the necessary and sufficient conditions for a successful cause of action. L. How The Nervous Shock Advisor Works The Nervous Shock Advisor (NSA) has all the fundamental characteristics of a rule-based expert system which were discussed earlier, and these will not be reiterated. At this point I shall limit myself to identifying its specific features. It is designed to act as an intelligent assistant to lawyers and others with some legal training. Particular facts are elicited from the user during the course of an interactive consultation which can be run on an IBM PC. The machine requires a minimum of 512K of R A M (Random Access Memory) if the latest version of M.1 is used and the knowledge base is loaded in a compressed form which has been previously saved to a file. If the knowledge base is loaded from source code or an older version of M.1 is used, 640K of R A M is needed. Consultations are initiated by simply typing "go" and then responding to the questions asked. On the basis of the facts provided, the user is advised whether or not a client has a good cause of action in nervous shock, together with the confidence level of the opinion in percentage terms. If there is no cause of action, NSA will inform the user which material element of the plaintiff's case is lacking. In either instance it will supply case citations and summaries of the case reports to justify its opinion if the user wishes to see them. The cases supplied are identified as falling into one or more of three categories: on point, relevant by analogy and contra. When NSA determines that there is a cause of action, it offers the user the option of seeing two additonal groups of cases. First, the cases in which the courts have allowed recovery for the particular symptoms experienced by the client. Secondly, the cases on which the defendant is most likely to rely in order to defend the action. In every consultation, the user is given the opportunity of perusing the full text of the leading case on nervous shock. A l l cases are retrieved automatically from a data base which has been linked to the advisor. 1 1 8 At any point during the consultation, the user can interrogate the system and ask "why" in response to any question. NSA will provide a short text explanation of its reasons and then repeat the question. A hard copy of the consultation can be obtained if the computer being used is linked to a printer. N S A was built at the Faculty of Law as part of the U B C - I B M Co-operative Project in Law and Computers. The initial planning of NSA began in May, 1986. The system was substantially completed in early January, 1987. Minor additions and refinements were finished by March, 1987. Since copyrighted software is required to run NSA, the working system in electronic form cannot be incorporated into this thesis. Arrangements to have the system demonstrated may be made through Professor Robert T. Franson, Project Director. NSA has been demonstrated publicly at various stages of its development. 1 1 9 A number of sample consultations in hard copy are contained in Appendix A. These have been included to give the reader a flavour of how NSA works in the absence of a hands-on demonstration. Hard copies of the Knowledge Base, Case File Index and Case File are to be found in Appendices B, C and D respectively. The Knowledge Base has comments throughout which explain in general terms how it works. The Case File Index aggregates the groups of cases which apply to the conclusions reached by NSA at the end of each of its many paths of inference. The Case File contains summaries of all the relevant cases on 87 nervous shock decided since 1888 in Canada, Great Britain, Australia and New Zealand (74 in all), as well as full text of what is currently the leading case. X I V . T H E F U T U R E O F L E G A L E X P E R T S Y S T E M S Expert systems in law will not be quite the same as the old idea of a slot machine kind of justice where you feed in the facts and the law and the machine cranks out a conclusive decision. In the first place expert systems generally do not contain first principles. Rather they contain merely the heuristic rules of the expert. They, therefore, do not and will not contain "The Law" except as it might implicitly be assimilated in the expert's heuristic rules. Being idiosyncratic to the particular expert whose expertise is built into the advisors, they can have no authoritative standing. They are merely tools of research, but they could turn out to be very powerful and valuable tools, because they can be easily updated and modified with new law, and they can be comprehensive within the particular domain. In fact, expert systems may become indispensable tools of legal research. The writer of an article in the January, 1987, edition of the American Bar Association Journal has suggested that lawyers may be liable for malpractice if an expert system has been created to cover a particular area of the law and they fail to consult i t . 1 2 0 The other side of the coin, as another writer has recently pointed out, is that a lawsuit could arise if expert systems are consulted and fail to perform correctly, or give inaccurate or misleading answers leading to i n j u r y . 1 2 1 I mention this question of potential liablity only in passing because it would make a thesis topic in itself. It does not seem unreasonable to hold lawyers liable for failing to consult an expert system given that they are much more user friendly than existing methods of legal research and that they provide answers which are tailor-made to fit the facts provided by the user. Legal texts, by contrast, merely provide a general discussion of an area illustrated by a limited number of examples. QuicLaw and L E X I S are useful, but they just provide the lawyer with access to a list of cases which contain the key words or phrases specified by her or him. These computerized systems cannot distinguish the relevant cases from those not relevant. Furthermore, they tend to provide either too much or too little in the way of raw material. Legal expert systems are about as close as one can get to the busy lawyer's favourite research tool: the telephone. For practising lawyers, time is of the essence. Their professional lives revolve around recording it in terms of billable hours or lamenting the lack of it. When an unusual problem arises, they do not waste time researching it from the ground up. They pick up the phone, run the facts by a friend with recognized expertise in the area, and get a quick opinion. Once they have some sort of answer which they know is basically correct, they can polish it up for themselves secure in the knowledge that they are on the right track. A legal expert system emulates this process with one important difference; namely, that the system is always available for a consultation whereas the friend might not be. 89 The most exaggerated claim made by a legal writer about expert systems is that made by John Miller, a lecturer in law from New Zealand. He predicts that they will eventually replace lawyers as a result of advanced expert systems which will be produced by the Fifth Generation Project in Japan . 1 2 2 Miller seems to have failed to grasp the fundamental fact that expert systems are based on human expertise, or, if he has, he is unclear about how the human expert will be replaced by a machine. However, as I indicated earlier, given the amount of talent and money being expended on developing Fifth Generation Systems, it is highly likely that some form of advanced expert system will be produced and I cannot see why legal reasoning would be too complex for such systems. ... Nor can it be assumed that legal expert systems would not be created as a matter of priority. A state has an obligation to make details of its law available to all its citizens. This can only be done properly with expert systems. Furthermore, lawyers have never enjoyed great popularity with the public and their suggested replacement, especially by a user friendly expert system would be a popular move. ... The only hope that I can see is for the legal profession to seize the opportunities offered by modern technology to provide inexpensive highly efficient comprehensive legal service to all sectors of society. If this was done, attitudes would change and they would be seen as the appropriate controllers of expert systems rather than as suitable for re placement by them.123 His statements have the ring of political rhetoric rather than dispassionate analysis. It seems reasonable to conclude, therefore, that they are intended for the ears of his government which, having abolished the right to sue for negligently caused personal i n j u r y , 1 2 4 would presumably be receptive to such ideas. In the course of his article, he does admit with respect to his predictions about expert systems and law that: "Most lawyers of course think that this is nonsense. On this issue I am in agreement with the majority of my colleagues. A more realistic assessment of the Fifth Generation Project is given by Terry Winograd in a book he co-authored with Fernando Flores, U N D E R S T A N D I N G C O M P U T E R S A N D C O G N I T I O N . 1 2 6 Winograd's impeccable credentials put him in a much better position to give a balanced viewpoint than Miller. He has been actively engaged for many years in computer science and artificial intelligence research at the Massachusetts Institute of Technology, Stanford University, and Xerox Palo Alto Research Centre. His work has centred around designing systems for the representation and analysis of language and knowledge. His conclusion about the Fi f th Generation Project is as follows. The grandiose goals, then, will not be met, but there will be useful spinoffs. In the long run, the ambitions for truly intelligent computers systems, as reflected in this project and others like it around the world, will not be a major factor in technological development. They are too rooted in the rationalistic tradition and too dependent on its assumptions about intelligence, language, and formalization.127 In essence, we must heed the old adage and know ourselves better before we put too much faith in expert systems. In a sense, these systems are a scam. They skim the cream from the surface of the human mind without being fully aware of how it floated to the top in the first place. Cognitive psychologists working in A l have had some useful insights about the nature of human expertise, but they are a long way from full understanding. By side stepping for the time being the issue of how the mind functions and instead making use of its end products in conjunction with relatively straightforward techniques which were developed in the early days of A l , we have achieved an eminently pragmatic blend of human and machine. In any event, in the long run, the real issue will be not whether human expertise can be removed from legal decision making but whether it ought to be. There are some human functions for which computers should not be substituted regardless of what they can or cannot be made to d o . 1 2 8 As intelligent assistants, legal expert systems can take lawyers very rapidly to a good jumping off point in their search for the solution to a particular problem. From that point onwards they can use their own creativity and ingenuity to reach a uniquely human solution to the problem. In this way, they will be spared the drudgery of sifting through a mass of irrelevant material and the stress of wondering whether they have overlooked some critical detail. Most lawyers I know have experienced the horror of waking up with the cold sweats in the early hours of the morning tormented by the thought that they have missed some vital point in their preparation of a case. They have also known the frustration of being bogged down in a morass of detail without being able to discern any unifying structure. In effect, the expert system reasons by analogy up to a point by fitting the user's facts into a framework constructed from the basic elements that are necessary to support a cause of action based on those particular facts. It is then left to the lawyer's own imagination to refine and extend the analogy in more subtle ways which will best support his client's case. Given the present state of the art in A l , this is the best that 92 can be done to overcome what has been seen to be a conceptual stumbling block in the way of the development of legal expert systems. Most legal problems tend to be poorly defined and broad-ranging. Moreover, they typically involve large amounts of common sense and reasoning by analogy — things that existing expert systems are unable to do.129 The day when computer systems will be able to reason using analogy and common sense in a way that comes close to what humans do is a long way off. The most ambitious attempt to achieve this goal is the C Y C Project undertaken by Microelectronics & Computer Technology Corp. of Austin, Texas . 1 3 0 The project just began and has a 10-year life span. Basically, it involves building a massive knowledge base of real world facts, heuristics and methods for efficiently reasoning over the whole knowledge base which would themselves be contained within it. In effect, it is an attempt to simulate in a machine the rich background of information acquired over a lifetime which humans draw upon when they reason about anything, but which they take completely for granted. In the final analysis, the validity of legal expert systems must be ascertained in terms of their practicality and usefulness. The test will thus be purely pragmatic. Practising lawyers are nothing if not pragmatic. Indeed, they are pragmatic to the point of being considered downright amoral by some people. If practitioners are showing an interest in expert systems - and they are, at least in the U.S. - it is safe to assume that the technology has passed a fairly rigorous kind of acid test. Despite the relatively small number of expert systems that now exist for the law there is much interest in developing commercial legal applications. Many law firms and insurance companies are seriously exploring this technology, some even creating expert system groups or departments to perform in-house research and development. We predict that within a few years most of the ma for insurance companies and many legal firms will be heavily involved in the building of expert systems.131 Expert systems, therefore, present a unique opportunity to test theory in terms of practice. Art i f ic ial Intelligence and legal reasoning is an area where theory meets practice with a vengeance and each stands to gain much from the interaction. Legal problems and patterns of reasoning have furnished an interesting body of material for testing and developing ideas in the field of artificial intelligence. Equally, the technology and tools developed in A l , in particular expert systems, furnish a new kind of empirical test for legal theory. It well may be that in the next few years as more and more A l methods are used in law to develop aids and tools for research, legal drafting, teaching, and prediction, the way we do legal theory may be fundamentally altered. And as jurisprudence may be transformed, the way we practice law may be transformed as well. N O T E S Anne von der Lieth Gardner, AN ARTIFICIAL INTELLIGENCE APPROACH TO LEGAL REASONING, (Ph.D. dissertation, Stanford University, Report No. STAN-CS-85-1045, June 1984, Department of Computer Science). Also see Anne Gardner, Overview of Artificial Intelligence approach to legal reasoning, in Charles Walter (ed.), COMPUTER POWER AND LEGAL REASONING, (West Publishing Co., St. Paul, Minnesota, 1985) at 247-274. Id. at 187. Id. Id. at 188. For a general introduction to the field see, Peter Jackson, INTRODUCTION TO EXPERT SYSTEMS, (Addison-Wesley Publishing Co., Menlo Park, California, 1986). Alex Goodall, THE GUIDE TO EXPERT SYSTEMS, (Learned Information, Medford, New Jersey, 1985). Donald A. Waterman, A GUIDE TO EXPERT SYSTEMS, (Addison-Wesley Publishing Co., Menlo Park, California, 1986). Paul Harmon and David King , EXPERT SYSTEMS: ARTIFICIAL INTELLIGENCE IN BUSINESS (John Wiley & Sons, New York, 1985) at 5. Jackson, supra note 5 at 1-18. R.I. Levine, D .E . Drang, and B. Edelson, A COMPREHENSIVE GUIDE TO AI AND EXPERT SYSTEMS, (McGraw-H i l l , Toronto, 1986) at 3-24. Peter S. Sell, EXPERT SYSTEMS-A PRACTICAL INTRODUCTION, (John Wiley & Sons, New York, 1985) at 2-3. Donald A. Waterman, A GUIDE TO EXPERT SYSTEMS, (Addison-Wesley Publishing Co., Menlo Park, California, 1986) at 26. Donald Johnson, UP THE HILL, (Optimum Publishing International, Montreal, 1986) at 99. Constantin V . Negoita, EXPERT SYSTEMS AND FUZZY SYSTEMS, (Benjamin/Cummings, Menlo Park, California, 1985). Richard Forsyth, Fuzzy Reasoning Systems, in Forsyth (ed.), EXPERT SYSTEMS PRINCIPLES AND CASE STUDIES, ( Chapman and Hal l , New York, 1984) at 51. Richard Forsyth, The architecture of expert systems, in EXPERT SYSTEMS, PRINCIPLES AND CASES STUDIES, Richard Forsyth ed. (Chapman and Hal l , London, 1984) at 10. Paul Kinnucan, Computers That Think Like Experts, (January, 1984), HIGH TECHNOLOGY 30 at 37. I attended a course in December, 1985, at Teknowledge Inc., Palo Alto, California. John Shore, T H E S A C H E R T O R T E A L G O R I T H M A N D O T H E R A N T I D O T E S T O C O M P U T E R A N X I E T Y , (Viking, Markham, Ontario, 1985) at 241. See pp. 241-244 for this writer's negative views about expert systems. It seems to me that his arguments about expert systems being part of the software crisis are refuted by the fact that by their very nature expert systems are small enough to be testable. Furthermore, they are designed to assist human decision making rather than to supplant it. Rock R. Farmer-Taylor, A Brief Review of Ignorance Engineering (June, 1985), C A N A D I A N A R T I F I C I A L I N T E L L I G E N C E 27 at 27. Jackson, supra note 5 at 19. Harman and King , supra note 6 at 134. Harmon and King , Id. at 15. F. Hayes-Roth, D.A. Waterman, D.B. Lenat, B U I L D I N G E X P E R T S Y S T E M S , (Addison-Wesley Publishing Co., Menlo Park, California, 1983) at 9. C. Townsend and D. Feucht, D E S I G N I N G A N D P R O G R A M M I N G P E R S O N A L E X P E R T S Y S T E M S , (Tab Books Inc., Blue Ridge Summit, Pennsylvania, 1986) at 15. Harman and King , supra note 6 at 145. Jackson, supra note 5 at 177. Waterman, supra note 5 at 247. Hayes-Roth, Waterman, and Lenat, supra note 18 at 108. Goodall, supra note 5 at 29. Joseph Weizenbaum, C O M P U T E R P O W E R A N D H U M A N R E A S O N (W.H. Freeman and Co., New York, 1976) at 277-278. Richard S. Rosenberg, C O M P U T E R S A N D T H E I N F O R M A T I O N S O C I E T Y , (John Wiley & Sons Inc., Toronto, 1986) at 362. Sholom M . Weiss and Casimir A. Kulikowski, A P R A C T I C A L G U I D E T O D E S I G N I N G E X P E R T S Y S T E M S , (Rowman & Allanheld, New Jersey, 1984) at 106. For an overview see Charles Walter, supra note 1. Richard E. Susskind, Expert Systems in Law: A Jurisprudential Approach to Artificial Intelligence and Legal Reasoning, (1986), 49 T H E M O D E R N L A W R E V I E W 168. Waterman, supra note 5 at 267. For examples see: D. Shpilberg, L . E . Graham, and H . Schatz, ExperTAXsm: an Expert System for Corporate Tax Planning, (1986) 3 E X P E R T S Y S T E M S 136, and Dean A. Schlobohm, TA—A Prolog Program Which Analyzes Income Tax Issues Under Section 318(a) of the Internal Revenue Code, in Walter, supra note 1 at 765; R. Hellawell, A Computer Program for Legal Planning and Analysis: Taxation of Stock Redemptions (1980), 80 C O L U M B I A L A W R E V I E W 1363. L . Thorne McCarty, Reflections on Taxman: An Experiment in Artificial Intelligence and Legal Reasoning (1977), 90 H A R V A R D L A W R E V I E W 837. The TAXMAN Project: Towards a Cognitive Theory of Legal Argument, in B. Niblet (ed.) C O M P U T E R S C I E N C E A N D L A W , (Cambridge University Press, Cambridge, 1980) at 23. W.G. Popp and B. Schlink, Judith, A Computer Program to Advise Lawyers in Reasoning a Case (1975), 15 J U R I M E T R I C S 303. D.A. Waterman and M.A. Peterson, Models of Legal Decision making, in P. Klahr and D.A. Waterman (eds.) E X P E R T S Y S T E M S , T E C H N I Q U E S , T O O L S , A N D A P P L I C A T I O N S , (Addison-Wesley Publishing Co., Menlo Park, California, 1986) at 135. M . Peterson and D.A. Waterman, An Expert Systems Approach to Evaluating Product Liability Cases, in Walter, supra note 1 at 627. D.A. Waterman, J. Paul, and M . Peterson, Expert Systems for Legal Decision Making (1986), 3 E X P E R T S Y S T E M S 212. Goodall, supra note 5 at 20. Waterman, supra note 5 at 269. Cary G. DeBessonet, A Proposal For Developing the Structural Science of Codification (1980), 8 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 47. An Automated Approach to Scientific Codification (1982), 9 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 27. An automated Intelligent System Based on a Model of a Legal System (1984), 10 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 31. C .G. DeBessonet and G.R. Cross, Representation of Some Aspects of Legal Causality, in Walter, supra note 1 at 205. C .G. DeBessonet, J. Hintze, and W. Waller, Automated Retrieval of Information: Toward the Development of a Formal Language For Expressing Statutes (1979), 6 S O U T H E R N U N I V E R S I T Y L A W R E V I E W 1. Laymen E . Allen, A Language-Normalization Approach to Informational Retrieval in Law (1968), 9 J U R I M E T R I C S J O U R N A L 41. L . E . Allen and C.S. Saxon, Computer Aided Normalizing and Unpacking: Some Interesting Machine-Processable Transformations of Legal Rules, in Walter, supra note 1 at 495. L . E . Allen and C.S. Saxon, One Use of Computerized Instructional Gaming in Legal Education: To Better Understand the Rich Logical Structure of Legal Rules and Improve Legal Writing (1985), 18.2 J O U R N A L O F L A W R E F O R M 383. Grayfred B. Gray, Statutes Enacted in Normalized Form: The Legislative Experience in Tennessee, in Walter, supra note 1 at 467. James A. Sprowl, Automating the Legal Reasoning Process: A Computer That Uses Regulations and Statutes to Draft Legal Documents (1979), 1 A M E R I C A N B A R F O U N D A T I O N R E S E A R C H J O U R N A L at 3. See for example, Carole D. Hafner, A N I N F O R M A T I O N S Y S T E M B A S E D O N A C O M P U T E R M O D E L O F L E G A L K N O W L E D G E , (UMI Research Press, Michigan, 1981) and the authors included in Walter, supra note 1, and in B. Niblet (ed.), supra note 27. Richard E . Susskind, supra note 25 at 183. Id. at 181. Id. at 183-184. The consensus as to what constitutes the proper practice of medicine is not as clear as one might think at first glance. As well as the standard medicine taught in the medical schools there are such methodologies as naturopathic medicine, holistic medicine or homeopathic medicine. The debates about what is the proper or best form of medicine, however, are different than the jurisprudential debates about law. The analogue of the medical debates would be disputes about whether mediation or adjudication is the best way to resolve disputes; or whether government regulation or adjudication is the best method of maintaining safety within an industry or protecting the environment. There is, on the other hand, no analogue in medicine to the issues in legal theory which separates positivists from realists or natural law theorists. For a contrast of medical and legal expertise see, Dietrich Rueschemeyer, Lawyers and Doctors: A Comparison of Two Professions, in Vilhelm Aubert (ed.), S O C I O L O G Y O F L A W , (Penguin, Harmondsworth, Middlesex, England, 1969) 267 at 270-272. Constantin V . Negoita, supra note 11. This term was coined by my expert, Professor J.C. Smith. Donald A. Waterman, supra note 5, at 26-27. Anne von der Lieth Gardner, A N A R T I F I C I A L I N T E L L I G E N C E A P P R O A C H T O L E G A L R E A S O N I N G , supra note 1 at 6. Jack Batten, R O B I N E T T E , T H E D E A N O F C A N A D I A N L A W Y E R S , (Macmillan of Canada, Toronto, 1984) at 210. Sholom M . Weiss and Casimir A . Kulikowski, A P R A C T I C A L G U I D E T O D E S I G N I N G E X P E R T S Y S T E M S , (Rowman & Allanheld, New Jersey, 1984) at 105. J.C. Smith, L E G A L O B L I G A T I O N , (Athlone Press of the University of London, 1976) at 2. A classic example is the courts' attempt to deal with computer technology using a piece of legislation which was enacted decades before computers were even conceived of. The Copyright Act, R.S.C. 1970, c. C-30, was first adopted in 1921 and has remained substantially unchanged ever since. In a recent case, the Federal Court (Trial Division) had to decide whether a computer program embodied in a silicon chip inside a computer is a subject matter in which copyright exists. The court found that it was by concluding that the original source code fell within the definition of "original literary work" and that the object code etched into the chip was a "translation" of that work. See: Apple Computer Inc. and Apple  Canada Inc. v. Mackintosh Computers Ltd. (1986), 3 F .T.R. 118. 115 N. Y . 506, 22 N.E. 188 (1889). Fred Rodell, W O E U N T O Y O U , L A W Y E R S ! , (Pagent-Poseidon Ltd. Brooklyn, New York, 1959) at 136. Rodell taught a seminar at the Yale Law School for many years. In this seminar, he taught and practiced his methodology by obtaining the briefs, factums, and material of cases before they were heard by the Supreme Court. Rodell then made his predictions and had his students make theirs by using his methods, and then verified the results by comparing them with the actual decisions of the Court when they were handed down. Oliver Wendell Holmes, The Path of the Law, in C O L L E C T E D L E G A L P A P E R S , (Harcourt, Brace and Co., New York, 1920) 167 at 169. Id. at 171. Id. at 173. See for example Roberto M . Unger, T H E C R I T I C A L L E G A L S T U D I E S M O V E M E N T , (Harvard University Press, Cambridge, 1986). David Kairys (ed.) T H E P O L I T I C S O F L A W , (Pantheon Books, New York, 1982). Duncan Kennedy, L E G A L E D U C A T I O N A N D T H E R E P R O D U C T I O N O F H I E R A R C H Y , A P O L E M I C A G A I N S T T H E S Y S T E M , (Afar Press, Cambridge, Massachusetts, 1983) at 39-40. H . L . A . Hart, Positivism and the Separation of Law and Morals, in Hart, E S S A Y S I N J U R I S P R U D E N C E A N D P H I L O S O P H Y , (Clarendon Press, Oxford, 1983) 49 at 62-72. H . L . A . Hart, T H E C O N C E P T O F L A W (Clarendon Press, Oxford, 1961) at 120-150. Anne von der Lieth Gardner, A N A R T I F I C I A L I N T E L L I G E N C E A P P R O A C H T O L E G A L R E A S O N I N G , supra note 1 at 38-41. Ronald Dworkin, T A K I N G R I G H T S S E R I O U S L Y (Harvard University Press, Cambridge, Massachusetts, 1977) at 22. Id. at 36. S.C. Coval and J.C. Smith, L A W A N D I T S P R E S U P P O S I T I O N S (Routledge & Kegan Paul, London, 1986) at 115-124. J.C. Smith and D.N. Weisstub, T H E W E S T E R N I D E A O F L A W (Butterworths, London, 1983) at 398-435. S.C. Coval and J.C. Smith, supra note 60. See for example Christian Bay, From Contract to Community, in Fred R. Dallmayr (ed.) F R O M C O N T R A C T T O C O M M U N I T Y (Marcel Dekker Inc., 1978) pp. 31-43, and Bay, Hayek's Liberalism: The Constitution of Perpetual Privilege (1971), 1 T H E P O L I T I C A L S C I E N C E R E V I E W E R 110-123. E.B. Pashukanis, The General Theory of Law and Marxism, in Hugh Babb, trans., S O V I E T L E G A L P H I L O S O P H Y (Harvard University Press, 1951) 193-201. R .E . Susskind, supra note 25 at 183-84. Paul Harmon and David King , supra note 6 at 199-200. Hart, T H E C O N C E P T O F L A W , supra note 56 at 89-96. Smith, supra note 46 at 42-59. Hart, T H E C O N C E P T O F L A W , supra note 56 at 79-88. Smith, supra note 46. S.C. Coval and J.C. Smith, supra note 60 at 115-124. Noam Chomsky, A S P E C T S O F T H E T H E O R Y O F S Y N T A X (M.I.T. Press, Cambridge, Massachusetts, 1965). D.A. Waterman, J. Paul, and M . Peterson, Expert Systems for Legal Decision Making (1986), 3 E X P E R T S Y S T E M S 212 at 214. Julius Stone, L E G A L S Y S T E M A N D L A W Y E R S ' R E A S O N I N G S , (Stevens & Sons, London, 1964) at 235-300. E . H . Levi , A N I N T R O D U C T I O N T O L E G A L R E A S O N I N G , (University of Chicago Press, Chicago, 1949). O.C. Jensen, T H E N A T U R E O F L E G A L A R G U M E N T , (Blackwell, Oxford, 1957). N . Blodgett, Artificial Intelligence Comes of Age, (Jan. 1987) A M E R I C A N B A R A S S O C I A T I O N J O U R N A L 68 at 70. J.C. Smith, L E G A L O B L I G A T I O N , (Athlone Press of the University of London, 1976) at 88-108. The moral philosopher, R . M . Hare, in F R E E D O M A N D R E A S O N (Oxford University Press, Oxford, 1965) at 35 argues that in contrast to moral judgments, legal judgments are not universalizable. He writes: "This is one reason why the word 'ought' cannot be used in making legal judgments; if a person has a certain legal obligation, we cannot express this by saying that he ought to do such and such a thing, for the reason that 'ought'-judgments have to be universalizable, which, in the strict sense, legal judgments are not. The reason why they are not is that a statement of law always contains an implicit reference to a particular jurisdiction..." Hare is wrong in assuming that such judgments contain a specific reference to jurisdiction. One must distinguish between statements of judgments which contain a particular reference and are consequently not universalizable, and statements which, although they may contain a particular reference, are about judgments which do not. A n example of the latter is the statement that "In England the judgment is held that one ought not to kil l bulls for sport". While the statement itself is not universalizable, the judgment which it expresses, "One ought not to kil l bulls for sport," is. Thus, while it is the case that it is the courts of particular legal jurisdictions which make judgments that some action ought or ought not to be done, the judgments themselves are universalizable. See also, S.C. Coval and J.C. Smith, L A W A N D I T S P R E S U P P O S I T I O N S , supra note 60. J.C. Smith, ibid, at 89. Ibid, at 90-97. Ibid, at 174-189. S.C. Coval and J.C. Smith, supra note 60. J.C. Smith, supra note 46 at 185. 529 P. 2d. 553. See for example, Joan Breckenbridge, AIDS carriers put doctors in dilemma, The Globe and Mai l , Saturday, Apr i l 11, 1987, at 1. J.C. Smith, L I A B I L I T Y I N N E G L I G E N C E , (Sweet and Maxwell, London, Carswell, Toronto, 1984) at 7-13. Parrott v. Wells Fargo & Co.. 15 Wall., U.S., 524, 21 L . E D . 206 (1872). Bolton v. Stone. [1951] A . C . 850. Weller & Co. v. Foot and Mouth Disease Research Institute. [1966] 1 Q.B. 569. J.C. Smith, supra note 83 at 3-7. Id. at 7-13. Alfred Lord Tennyson, Aylmer's Field, 1864. Comyn J. in Whitmore v. Eurowavs Express Coaches Ltd. . The Times, 4 May 1984, cited in F .A . Trinade, The Principles Governing the 101 Recovery for Negligently Caused Nervous Shock (1986), 45 C A M B R I D G E L A W J O U R N A L 476 at 476. 9 1 See for example Hav ( or Bourhill) v. Young. [1943] A . C . 92 (H.L.). 9 2 Abramzik v. Brenner (1962), 62 W.W.R. 332 at 335 (Sask. C.A.). 9 3 K ing v. Phillips. [1953] 1 Q.B. 429 at 439 (C.A.). 9 4 1 62 N .E . 99. 9 5 Ibid. 9 6 Id. at 100. 9 7 Id. at 103. 9 8 [1921] 3 K.B . 560. 9 9 Id. at 569. 1 0 0 Id. at 574. 1 0 1 [1943] A . C . 92. 1 0 2 [1961] A . C . 388. 1 0 3 Y.B. Smith and William L . Prosser, C A S E S A N D M A T E R I A L S O N T O R T S (Foundation Press, Brooklyn, 1952) at 231. 1 0 4 William L . Prosser, S E L E C T E D T O P I C S O N T H E L A W O F T O R T S , (University of Michigan, Ann Arbor, 1954) at 219. 1 0 5 Abramzik v. Brenner. (1967) 65 D.L.R. (2d) 651 (Sask. C.A.). 1 0 6 Falkenham v. Zwicker (1978), 93 D.L.R. (3d) 289. 1 0 7 J .C. Smith, L I A B I L I T Y I N N E G L I G E N C E , supra note 83 at 101-114. 1 0 8 [1967] 1 A . C . 617. 1 0 9 J .C. Smith, L I A B I L I T Y I N N E G L I G E N C E , supra note 83 at 112. 1 1 0 Ibid. 1 1 1 Id. at 140. 1 1 2 L A W A N D I T S P R E S U P P O S I T I O N S , supra note 75 at 21-25. L I A B I L I T Y I N N E G L I G E N C E , supra note 83 at 131-161. 1 1 3 Constantin V . Negoita, supra note 11. Marketed by Teknowledge Inc., 1850 Embarcadero Rd. , Palo Alto, C A 94303. Goodall, supra note 5 at 43. Id. at 45. This fairly simple example was put together for a lecture which I gave on Expert Systems in a class on legal reasoning. The linkage between the knowledge base and the data files was written in the C programming language by Doug Arnold, the Project Co-ordinator. Other valuable technical assistance was rendered by Don Johnson, the Project's Computer Systems Technician. The UBC-Queens University Joint Conference of Computers and Law on June 26, 1986. The Conference of Academic and Scientific Executives (CASE), an international conference sponsored by IBM, on October 8, 1986. The Annual Meeting of the Association of American Law Schools at Los Angeles, California, on January 6, 1987. U B C Open House, March 7 & 8, 1987. N. Blodgett, supra note 74 at 70. Janet S. Zeide, Esq. and Jay Liebowitz, Using Expert Systems: The Legal Perspective (Spring 1987) I E E E E X P E R T 19 at 20. John Miller, Expert Systems - The end of the legal profession?, (1984), N E W Z E A L A N D L A W J O U R N A L 85 at 85-88. Id. at 87-88. Accident Compensation Act, 1972, No. 43. This Act erodes the philosophical basis of Western law by permitting individual freedom of action without imposing a corresponding personal liability for negligence. John Miller, supra note 122 at 87. Terry Winograd and Fernando Flores, U N D E R S T A N D I N G C O M P U T E R S A N D C O G N I T I O N (Ablex Publishing Corp., Norwood, New Jersey, 1986) at 133-139. Id. at 138-139. Joseph Weizenbaum, C O M P U T E R P O W E R A N D H U M A N R E A S O N (W.H. Freeman and Co., New York, 1976) at 269-270. A full discussion of human versus machine decision making is beyond the scope of this paper. The subject is a highly controversial one. See for example: Anthony D'Amato, Can/Should Computers Replace Judges?, (1977) n G E O R G I A L A W R E V I E W at 1277; Albert J. Millus, Would You Settle for a 103 Robot on the Bench?, 57 N E W Y O R K S T A T E B A R J O U R N A L at 22; and Joseph J. Spengler, Machine-Made Justice: Some Implications, 2 8 L A W A N D C O N T E M P O R A R Y P R O B L E M S at 37. Paul Harmon and David K i n g , supra note 6 at 7. Doug Lenat, Mayank Prakash, & Mary Shepherd, CYC: Using Common Sense Knowledge to Overcome Brittleness and Knowledge Acquisition Bottlenecks (1986), 6 Al M A G A Z I N E ( N O . 4) 65 at 65-85. 131 D.A. Waterman, J. Paul, and M. Peterson, supra note 72 at 224. B I B L I O G R A P H Y B O O K S Allen, L . and M . Caldwell, eds. C O M M U N I C A T I O N S C I E N C E S A N D L A W : R E F L E C T I O N S F R O M T H E J U R I M E T R I C S . Indianapolis: Bobbs-Merrill, 1965. Anderson, Alan R., ed. M I N D S A N D M A C H I N E S . Englewood Cliffs , N.J.: Prentice-Hall, 1964. Barr, Avron and Edward Feigenbaum, eds. T H E H A N D B O O K O F A R T I F I C I A L I N T E L L I G E N C E . 3 vols. Stanford, California: Heuristech Press, 1 9 8 2 . Batten, Jack, R O B I N E T T E - T H E D E A N O F C A N A D I A N L A W Y E R S . Toronto, Canada: Macmillan Canada, 1984. Bigelow, Robert P., ed. C O M P U T E R S A N D T H E L A W : A N I N T R O D U C T O R Y H A N D B O O K , 3d ed. Chicago: American Bar Association, 1981. Bing, Jon and K.S. Selmer, eds. A D E C A D E O F C O M P U T E R S A N D L A W . Oslo: Universitetsforlegat, 1980. Buchanan, Bruce G. and E . H . Shortliffe. R U L E - B A S E D E X P E R T S Y S T E M S : T H E M Y C I N E X P E R I M E N T S O F T H E H E U R I S T I C P R O G R A M M I N G P R O J E C T . Menlo Park, California: Addison-Wesley Publishing Co., 1985. Campbell, Colin, ed. D A T A P R O C E S S I N G A N D T H E L A W . London: Sweet & Maxwell, 1984. Charniak, Eugene and Drew McDermott. I N T R O D U C T I O N T O A R T I F I C I A L I N T E L L I G E N C E . Reading, Mass.: Addison-Wesley, 1985. Chomsky, Noam, A S P E C T S O F T H E T H E O R Y O F S Y N T A X . Cambridge, Mass.: M.I.T. Press, 1965. Ciampi, C , ed. A R T I F I C I A L I N T E L L I G E N C E A N D L E G A L I N F O R M A T I O N S Y S T E M S . New York: North-Holland Pub. Co., 1982. Coombs, J.J. D E V E L O P M E N T S I N E X P E R T S Y S T E M S . London: Academic Press Inc., 1984. Cordell, Arthur J. T H E U N E A S Y E I G H T I E S : T H E T R A N S I T I O N T O A N I N F O R M A T I O N S O C I E T Y . Ottawa: Science Council of Canada, 1985. Coval, S.C. and J.C. Smith, L A W A N D I T S P R E S U P P O S I T I O N S . London: Routledge & Kegan Paul, 1986. Dennett, D.C. B R A I N S T O R M S : P H I L O S O P H I C A L E S S A Y S O N M I N D A N D P S Y C H O L O G Y . Hassocks: Harvester Press, 1978. Dretske, Fred I. K N O W L E D G E A N D T H E F L O W O F I N F O R M A T I O N . Cambridge, Mass.: M.I.T. Press, 1981. Dworkin, Ronald, T A K I N G R I G H T S S E R I O U S L Y . Cambridge, Mass.: Harvard University Press, 1977. Evans, Christopher R. T H E M I C R O M I L L E N N I U M . New York: Vik ing Press, 1980. Fabien, Claude, C O M P U T E R I Z E D L E G A L R E S E A R C H I N C A N A D A - W O R K I N G P A P E R N O . 3 . Ottawa: Canadian Law Information Council, 1979. Fahlman, S.E. N E T L : A S Y S T E M F O R R E P R E S E N T I N G A N D U S I N G R E A L W O R L D K N O W L E D G E . Mass.: M.I.T. Press, 1979. Feigenbaum, E . A . and P. McCorduck. T H E F I F T H G E N E R A T I O N . New York: Signet, 1983. Fink, Donald G . C O M P U T E R S A N D T H E H U M A N M I N D . Garden City, N.Y.: Anchor Books, 1966. Gardner, Anne von der Lieth. A N A R T I F I C I A L I N T E L L I G E N C E A P P R O A C H T O L E G A L R E A S O N I N G , R E P O R T N O . S T A N - C S - 8 5 - I O 4 5 . Stanford University: Dept. of Computer Science, 1984. Gardner, Howard, T H E M I N D ' S N E W S C I E N C E . New York: Basic Books, 1985. Gevarter, W.B. A I , E X P E R T S Y S T E M S , C O M P U T E R V I S I O N A N D N A T U R A L L A N G U A G E P R O C E S S I N G . New Jersey: Noyes Publication, 1984. Goodall, Alex. T H E G U I D E T O E X P E R T S Y S T E M S . Medford, New Jersey: Learned Information, 1985. Hafner, Carole D. A N I N F O R M A T I O N R E T R I E V A L S Y S T E M B A S E D O N A C O M P U T E R M O D E L O F L E G A L K N O W L E D G E . Mich.: U M I Research Press, 1981. Hammond, P. and M.J. Sergot. T H E L A W A S L O G I C P R O G R A M S . Ontario, Canada: Applied AI Systems, 1985. Hare, R . M . F R E E D O M A N D R E A S O N . Oxford: Oxford University Press, 1965. Harmon, Paul and David King, E X P E R T S Y S T E M S : A R T I F I C I A L I N T E L L I G E N C E I N B U S I N E S S . New York: John Wiley & Sons, 1985. Harris, James W. L A W A N D L E G A L S C I E N C E . Oxford: Clarendon Press, 1979. Hayes, Jean E . and D. Michie, eds. I N T E L L I G E N T S Y S T E M S : T H E U N P R E C E D E N T E D O P P O R T U N I T Y . Chichester, U.K.: Ellis Horwood Ltd. , 1983. Hayes, Jean E . , D. Michie, and Y. Pao, eds. M A C H I N E I N T E L L I G E N C E I O . New York: Halsted Press, 1982. Hayes-Roth, F. , D.A. Waterman and D.B. Lenat, eds. B U I L D I N G E X P E R T S Y S T E M S . Menlo Park, California: Addison-Wesley Publishing Co., 1983. Hofstadter, D.R. M E T A M A G I C A L T H E M A S : Q U E S T I N G F O R T H E E S S E N C E O F M I N D A N D P A T T E R N . New York: Basic Books, 1985. Hustler, Allen, P R O G R A M M I N G L A W I N L O G I C . Waterloo, Ontario: Department of Computer Science, University of Waterloo, 1982. Jackson, Peter, I N T R O D U C T I O N T O E X P E R T S Y S T E M S . Menlo Park, California: Addison-Wesley Publishing Co., 1986. Johnson, Donald, U P T H E H I L L . Montreal: Optimum Publishing International, 1986. Kairys, David, ed. T H E P O L I T I C S O F L A W . New York: Pantheon Books, 1982. Kennedy, Duncan, L E G A L E D U C A T I O N A N D T H E R E P R O D U C T I O N O F H I E R A R C H Y - A P O L E M I C A G A I N S T T H E S Y S T E M . Cambridge, Mass.: Afar , 1983. Kent, Ernest W. T H E B R A I N S O F M E N A N D M A C H I N E S . Peterborough, N.H.: Byte Books, 1981. Klahr, Philip and Donald A. Waterman, eds. E X P E R T S Y S T E M S , T E C H N I Q U E S , T O O L S , A N D A P P L I C A T I O N S . Menlo Park, California: Addison-Wesley Publishing Co., 1986. Kowalski, Robert, L O G I C F O R P R O B L E M S O L V I N G . New York: North-Holland, 1979. Lehnert, W.G. T H E P R O C E S S O F Q U E S T I O N A N S W E R I N G . New Jersey: Erlbaum, 1978. Levi , E . A N I N T R O D U C T I O N T O L E G A L R E A S O N I N G . Chicago: Chicago University Press, 1949. Levine, R.I., D . E . Drang and B. Edelson. A C O M P R E H E N S I V E G U I D E T O A I A N D E X P E R T S Y S T E M S . Toronto: McGraw-Hi l l , 1986. Mason, Mary A. A N I N T R O D U C T I O N T O U S I N G C O M P U T E R S I N T H E L A W . St. Paul, Minnesota: West Publishing Co., 1984. McCorduck, P. M A C H I N E S W H O T H I N K - , A P E R S O N A L I N Q U I R Y I N T O T H E H I S T O R Y A N D P R O S P E C T S O F A R T I F I C I A L I N T E L L I G E N C E . San Francisco: W.H. Freeman, 1979. Michie, Donald, ed. E X P E R T S Y S T E M S I N T H E M I C R O - E L E C T R O N I C A G E . Scotland: Edinburgh University Press, 1979. Michie, Donald, ed. I N T R O D U C T O R Y R E A D I N G S I N E X P E R T S Y S T E M S . London: Gordon and Breach, 1983. 108 Michie, Donald and Rory Johnston. T H E C R E A T I V E C O M P U T E R . London: Viking , 1984. Negotia, Constantin Virgi l , E X P E R T S Y S T E M S A N D F U Z Z Y S Y S T E M S . Menlo Park, California: Benjamin/Cummings, 1985. Niblett, B., ed. C O M P U T E R S C I E N C E A N D L A W . London: Cambridge University Press, 1980. Nilsson, Nils J. P R I N C I P L E S O F A R T I F I C I A L I N T E L L I G E N C E . Palo Alto, California: Tioga Publishing Co., 1980. Oddy, Robert N. , ed. I N F O R M A T I O N R E T R I E V A L R E S E A R C H . London: Butterworths, 1981. Pask, Gordon and Susan Curran. M I C R O M A N : C O M P U T E R S A N D T H E E V O L U T I O N O F C O N S C I O U S N E S S . New York: Macmillan, 1982. Prosser, William L . S E L E C T E D T O P I C S O N T H E L A W O F T O R T S . A n n Arbor, Michigan: University of Michigan, 1954. Rich, Elaine, A R T I F I C I A L I N T E L L I G E N C E . New York: McGraw-Hi l l , 1983. Rodell, Fred, W O E U N T O Y O U , L A W Y E R S ! New York: Pagent-Poseidon Ltd. , 1959. Rose, S. T H E C O N S C I O U S B R A I N . New York: Knopf, 1973. Rosenberg, Richard S. C O M P U T E R S A N D T H E I N F O R M A T I O N S O C I E T Y . Toronto: John Wiley & Sons Inc., 1986. Schank, Roger C. D Y N A M I C M E M O R Y . New York: Cambridge University Press, 1982. Schank, Roger C. and Peter Childers. T H E C O G N I T I V E C O M P U T E R - , O N L A N G U A G E , L E A R N I N G A N D A R T I F I C I A L I N T E L L I G E N C E . Reading, Mass.: Addison-Wesley, 1984. Sell, Peter S. E X P E R T S Y S T E M S - A P R A C T I C A L I N T R O D U C T I O N . New York: John Wiley & Sons, 1985. Sherman, David M . B L U E P R I N T F O R A C O M P U T E R - B A S E D M O D E L O F T H E I N C O M E T A X A C T O F C A N A D A . North York, Ontario: Osgoode Hal l Law School, 1986. Shirai, Yoshiaki and Jun-ichi Tsujii. A R T I F I C I A L I N T E L L I G E N C E - C O N C E P T S , T E C H N I Q U E S A N D A P P L I C A T I O N S . Trans. F .R.D. Apps. New York: John Wiley & Sons, 1984. Shore, John, T H E S A C H E R T O R T E A L G O R I T H M A N D O T H E R A N T I D O T E S T O C O M P U T E R A N X I E T Y . Ontario: Viking, 1985. Smith, J.C. L E G A L O B L I G A T I O N . University of London, London: Athlone Press, 1976. Smith, J.C. L I A B I L I T Y I N N E G L I G E N C E . London: Sweet & Maxwell, 1984. Smith, J.C. and D.N. Weisstub. T H E W E S T E R N I D E A O F L A W . London: Butterworths, 1983. Smith, Y.B. and William L . Prosser. C A S E S A N D M A T E R I A L S O N T O R T S . Brooklyn: Foundation Press, 1952. Stone, Julius, L E G A L S Y S T E M A N D L A W Y E R S ' R E A S O N I N G S . London: Stevens & Sons, 1964. Susskind, Richard Eric, E X P E R T S Y S T E M S I N L A W - A J U R I S P R U D E N T I A L I N Q U I R Y . PhD. thesis. Oxford: Oxford University, 1986. Tapper, Colin, C O M P U T E R S A N D T H E L A W . London: Weidenfeld and Nicolson, 1973. Townsend, Carl and Dennis Feucht. D E S I G N I N G A N D P R O G R A M M I N G P E R S O N A L E X P E R T S Y S T E M S . Blue Ridge Summit, P.A.: T A B Books Inc., 1986. Unger, Roberto M . T H E C R I T I C A L L E G A L S T U D I E S M O V E M E N T . Cambridge, Mass.: Harvard Unvi . Press, 1986. Walter, Charles, ed. C O M P U T I N G P O W E R A N D L E G A L R E A S O N I N G . St. Paul, Minnesota: West Publishing Co., 1985. Waterman, D.A. A G U I D E T O E X P E R T S Y S T E M S . Menlo Park, California: Addison-Wesley Publishing Co., 1986. Waterman, D.A. and F. Hayes-Roth, eds. P A T T E R N - D I R E C T E D I N F E R E N C E S Y S T E M S . New York: Academic Press, 1978. Waterman, D.A. and M.A. Peterson. M O D E L S O F L E G A L D E C I S I O N M A K I N G , R E P O R T R - 2 7 1 7 - I C J . Institute for Civ i l Justice, Rand Corporation, 1981. Weiss, Sholom M . and Casimir A . Kulikowski. A P R A C T I C A L G U I D E T O D E S I G N I N G E X P E R T S Y S T E M S . New Jersey: Rowman & Allanheld, 1984. Weizenbaum, Joseph, C O M P U T E R P O W E R A N D H U M A N R E A S O N . New Tork: W.H. Freeman and Co., 1976. Winograd, Terry and Fernando Flores. U N D E R S T A N D I N G C O M P U T E R S A N D C O G N I T I O N . New Jersey: Ablex Publishing Corp., 1986. Winston, Patrick, A R T I F I C I A L I N T E L L I G E N C E , 2d ed., Reading, Mass.: Addison-Wesley, 1984. Zadeh, Lotf i A . F U Z Z Y S E T S A N D D E C I S I O N A N A L Y S I S . New York: North-Holland, 1978. A R T I C L E S Aikins, J.S. Prototypical Knowledge for Expert Systems. (1983) 20(2) A . I . M A G A Z I N E 163-210. Allen, Layman E . Beyond Document Retrieval Toward Information Retrieval. (1963) 47 M I N N E S O T A L A W R E V I E W 713. Allen, Layman E . A Language-Normalization Approach To Information Retrieval In Law. (1968) 9 J U R I M E T R I C S J O U R N A L 41. Allen, Layman E . Symbolic Logic: A Razor-Edged Tool for Drafting and Interpreting Legal Documents. (1957) 66 Y A L E L A W J O U R N A L 833. Allen, Layman E . and C.S. Saxon. One Use of Computerized Instructional Gaming in Legal Education: To Better Understand the Rich Logical Structure of Legal Rules and Improve Legal Writing. (1985) 18:2 J O U R N A L O F L A W R E F O R M 383. Allen, Layman E . and C.S. Saxon. Computer Aided Normalizing and Unpacking: Some Interesting Machine-Processable Transformations of Legal Rules, C O M P U T I N G P O W E R A N D L E G A L R E A S O N I N G . Ed . Charles Walter. St. Paul, Minnesota: West Publishing Co., 1985. Asimov, I. The Next 70 Years For Law and Lawyers. 71 A M E R I C A N B A R A S S O C I A T I O N J O U R N A L 56. Andrews, David. The Legal Challenge Posed by the New Technologies. (1983) 24 J U R I M E T R I C S J O U R N A L 43. Bauer-Bernet, Helene. Effect of Information Science on the Formation and Drafting of Law. (1974) 14 J U R I M E T R I C S J O U R N A L 235. Bay, Christian. From Contract to Community, F R O M C O N T R A C T T O C O M M U N I T Y . Ed. Fred R. Dallmayr. Marcel Dekker Inc., 1978. Bay, Christian. Hayek's Liberalism: The Constitution of Perpetual Privilege. (1971) 1 T H E P O L I T I C A L S C I E N C E R E V I E W E R 110-123. Bennett, Guy M . and Signa Treat. Selected Bibliographical Material on Computer-Assisted Legal Analysis. (1984) 24 J U R I M E T R I C S J O U R N A L 283. Blodgett, N . Artificial Intelligence Comes of Age. (1987) A M E R I C A N B A R A S S O C I A T I O N J O U R N A L 68. Boose, John H . A Knowledge Acquisiton Program for Expert Systems Based on Personal Construct Psychology. (1985) 23 I N T . J O U R N A L M A N - M A C H I N E S T U D I E S 495-525. Boyd, W.E. and Charles S. Saxon. The A-9: A Program for Drafting Security Agreements Under Article 9 of the Uniform Commercial Code. [1981] 3 A M E R I C A N B A R F O U N D A T I O N R E S E A R C H J O U R N A L 637. Breckenbridge, Joan. AIDS Carriers Put Doctors in Dilemma. (Saturday, Apr i l 11, 1987) T H E G L O B E A N D M A I L 1. Buchanan, Bruce G. New Research on Expert Systems, M A C H I N E I N T E L L I G E N C E 10. Eds. J. Hayes, D. Michie and Y . Pao. New York: Halsted Press, 1982. Buchanan, Bruce G. and F.A. Feigenbaum. Dendral and Meta-Dendral: Their Applications Dimension. (1978) 11(1) A . I . M A G A Z I N E 5-24. Buchanan, Bruce G. and T . E . Headrick. Some Speculation About Artificial Intelligence and Legal Reasoning. (1970) 23 S T A N F O R D L A W R E V I E W 40. Buchanan, Bruce G. and T . M . Mitchell. Model-Directed Learning of Production Rules. P A T T E R N - D I R E C T E D I N F E R E N C E S Y S T E M S . Eds. D. Waterman and F. Hayes-Roth. New York: Academic Press, 1978. Cendrowska, J. and M.A. Bramer. A Rational Reconstruction of the MYCIN Consultation System. (1984) 20(3) I N T . J O U R N A L M A N - M A C H I N E S T U D I E S 229-317. Chandrasekaran, B. and S. Mittal. Deep Versus Compiled Knowledge Approaches to Diagnostic Problem-Solving. (1982) 19(5) I N T . J O U R N A L M A N - M A C H I N E S T U D I E S 425-436. Clancey, W.J. The Epistemology of a Rule-Based Expert System: A Framework for Explanation. (1983) 20(3) A . I . M A G A Z I N E 215-51. Clark, K.I. and F . G . McCabe. PROLOG: A Language For Implementing Expert Systems, M A C H I N E I N T E L L I G E N C E 10. Eds. J. Hayes, D. Michie and Y . Pao. New York: Halsted Press, 1982. Cook, S. The Applications of A.I. to Law: A Survey of 6 Current Projects. (1980) 50 N A T I O N A L C O M P U T E R C O N F E R E N C E P R O C E E D I N G S 689. 113 Coombs, M.J. and J. Alty. Expert Systems: An Alternative Paradigm. (1984) 20(1) I N T . J O U R N A L M A N - M A C H I N E S T U D I E S 21-43. D'Amato, Anthony. Can/Should Computers Replace Judges? (1977) 11 G E O R G I A L A W R E V I E W 1277. Davis, R. Knowledge Acquisition in Rule-Based Systems: Knowledge About Representations as a Basis For System Construction and Maintenance. P A T T E R N - D I R E C T E D I N F E R E N C E S Y S T E M S . Eds. D. Waterman and F. Hayes-Roth. New York: Academic Press, 1978. Davis, R. Interactive Transfer of Expertise: The Acquisition of New Inference Rules. (1978) 11(1) A . l . M A G A Z I N E 5-24. Davis, R., B. Buchanan and E . Shortliffe. Production Rules as a Representation for a Knowledge-Based Consultation Program. (1978) 8 A . I . M A G A Z I N E 15-45. deBessonet, Cary G. An Automated Approach to Scientific Codification. (1982) 9 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 27. deBessonet, Cary G. An Automated Intelligent System Based on a Model of a Legal System. (1984) 10 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 31. deBessonet, Cary G. A Proposal For Developing the Structural Science of Codification. (1980) 8 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 47. deBessonet, Cary G. and G.R. Cross. Representation of Some Aspects of Legal Causality, C O M P U T E R P O W E R A N D L E G A L R E A S O N I N G . Ed. Charles Walter. St. Paul, Minnesota: West Publishing Co., 1985. deBessonet, Cary G. , J. Hintze and W. Waller. Automated Retrieval of Information: Toward the Development of a Formal Language for Expressing Statutes. (1979) 6 S O U T H E R N U N I V E R S I T Y L A W R E V I E W 1. Dickerson, Reed. Some Jurisprudential Implications of Electronic Data Processing. 28 L A W A N D C O N T E M P O R A R Y P R O B L E M S 53. Dickerson, Reed. Electronic Aids to the Drafting of Legal Instruments. (1970) 1 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 75. Doane, Donald G. A Pilot Study For Computerised Legislation. (1974) 48 L A W I N S T . J O U R N A L 180. Drobak, John N. Computer Simulation and Gaming: An Interdisciplinary Survey with a View Toward Legal Applications. (1972) 24 S T A N F O R D L A W R E V I E W 712. Duda, R.O., P.E. Hart, N.J. Nilsson and G.I. Sutherland. Semantic Network Representations in Rule-Based Inference Systems, P A T T E R N - D I R E C T E D I N F E R E N C E S Y S T E M S . Eds. D. Waterman and F. Hayes-Roth. New York: Academic Press, 1978. Dyer, Michael G . and Mar got Flowers. Toward Automating Legal Expertise. C O M P U T I N G P O W E R A N D L E G A L R E A S O N I N G . Ed. Charles Walter. St. Paul, Minnesota: West Publishing Co., 1985. Dym, C.I. and S. Mittal. Knowledge Acquisition From Multiple Experts. (1985) 6(2) A . I . M A G A Z I N E 32-37. Edwards, Thomas H . and James P. Barber. A Computer Method for Legal Drafting Using Propositional Logic. (1975) 53 T E X A S L A W R E V I E W 965. Eldridge, W.B. and Sally F. Dennis. The Computer as a Tool For Legal Research. (1963) 28 L A W A N D C O N T E M P O R A R Y P R O B L E M S 78. Erman, L . D . , F . Hayes-Roth, V .R . Lesser, and D.R. Reddy. The Hearsay-II Speech Understanding System: Integrating Knowledge to Resolve Uncertainty. 12 C O M P U T I N G S U R V E R S 213. Farmer-Tailor, Rock R. A Brief Review of Ignorance Engineering. (June 1985) C A N A D I A N A R T I F I C I A L I N T E L L I G E N C E 27. Finan, John P. Lawgical: Jurisprudential and Logical Considerations. (1982) 15 A K R O N L A W R E V I E W 675. Fleming, Jack W. A Guide to the Use of Computers to Estimate Damages in Complex Litigation. (1980) 2 C O M P U T E R L A W J O U R N A L 863. 115 Forsyth, Richard. Fuzzy Reasoning Systems, E X P E R T S Y S T E M S P R I N C I P L E S A N D C A S E S T U D I E S . Ed. R. Forsyth. New York: Chapman and Hal l , 1984. Forsyth, Richard. The Architecture of Expert Systems, E X P E R T S Y S T E M S P R I N C I P L E S A N D C A S E S T U D I E S . Ed. R. Forsyth. New York: Chapman and Hall , 1984. Fox, John. Knowledge, Decision Making, and Uncertainty. A R T I F I C I A L I N T E L L I G E N C E A N D S T A T I S T I C S . Ed. William A. Gale. Menlo Park, California: Addison-Wesley Publishing Co., 19 . Freiling, M. , J. Alexander, S. Messick, S. Rehfus and S. Shulman. Starting A Knowledge Engineering Project: A Step-by-Step Approach. (1985) 6(3) A . I . M A G A Z I N E 150-164. Friedland, P.E. Acquisition of Procedural Knowledge From Domain Experts. P R O C E E D I N G O F T H E 7TH U C A I - 8 1 . Vancouver, B.C., 1981. Fuller, Lon L . Science and the Judicial Process. (1966) 79 H A R V A R D L A W R E V I E W 1604-1628. Gardner, Anne von der Lieth. Overview of Artificial Intelligence Approach to Legal Reasoning, C O M P U T E R P O W E R A N D L E G A L R E A S O N I N G . Ed . Charles Walter. St. Paul, Minnesota: West Publishing Co., 1985. Gaschnig, J. , P. Klahr, H . Pople, E . Shortliffe, and A. Terry. Evaluation of Expert Systems: Issues and Case Studies, B U I L D I N G E X P E R T S Y S T E M S . Eds. F . Hayes-Roth, D.A. Waterman and D.B. Lenat. Menlo Park, California: Addison-Wesley Publishing Co., 1983. Gold, D.I. and Susskind, R .E . Expert Systems in Law - A Jurisprudential and Formal Specification Approach. 2 N D I N T E R N A T I O N A L C O N G R E S S O N "LOGiCA, I N F O R M A T I C A , D I R I T T O " . Palazzo dei Congressi, Florence, Sept., 1985. Gonser, T .A. , J.T. Soma and E.I. Wilhelm. The Computer as a Tool for Legal Decision Making. (1981) 27 P R A C T I C A L L A W Y E R 11. Grainger, Brian. Hard Times and Automation: Should Computers Assist Judges in Sentencing Decisions. 26 C A N A D I A N J O U R N A L O F C R I M I N O L O G Y 231. Gray, Grayfred B. Statutes Enacted in Normalized Form: The Legislative Experience in Tennessee, C O M P U T I N G P O W E R A N D L E G A L R E A S O N I N G . Ed. Charles Walter. St. Paul, Minnesota: West Publishing Co., 1985. Greenberg, R. The Lawyer's Use of Quantitative Analysis In Settlement Negotiations. 38 B U S I N E S S L A W Y E R 1557. Greenberg, Ronald David. Quantitative Aspects of Legal Analysis. (1976) 645 I N S U R A N C E L A W J O U R N A L 589. Haar, Charles, John P. Sawyer Jr. and Stephen J. Cummings. Computer Power and Legal Reasoning: A Case Study of Judicial Decision Prediction in Zoning Amendment Cases. [1977] 3 A M E R I C A N B A R F O U N D A T I O N R E S E A R C H J O U R N A L 655. Harris, Allen. Judicial Decision Making and Computers. (1967) 12 V I L L A N O V A L A W R E V I E W 272. Harrison, G.A. and J.T. Scott Jr. and C.B. Baker. A Decision Model for Estate Planning—Formula and Computer Aids to Projections of Outcomes. (1968) 107 T R U S T S A N D E S T A T E S 523. Hart, H . L . A . Positivism and the Separation of Law and Moral, E S S A Y S I N J U R I S P R U D E N C E A N D P H I L O S O P H Y . Ed . H . Hart. Oxford: Clarendon Press, 1983. Hayes-Roth, Frederick. Rule-Based Systems. (1985) 28(9) C O M M U N I C A T I O N S O F T H E A C M 921. Hellawell, Robert. SEARCH: A Computer Program For Legal Problem Solving. (1982) 15 A K R O N L A W R E V I E W 635. Hellawell, Robert. CHOOSE: A Computer Program for Legal Planning and Analysis. (1981) 19 C O L U M B I A L A W R E V I E W 339. Hellawell, Robert. A Computer Program for Legal Planning and Analysis: Taxation of Stock Redemptions. (1980) 80 C O L U M B I A L A W R E V I E W 1363. Holmes, Oliver Wendell. The Path of the Law. C O L L E C T E D L E G A L P A P E R S . New York: Harcourt, Brace and Co., 1920. 117 Ishizuka, M . Japanese Work in Expert Systems. (1984) 1(1) E X P E R T S Y S T E M S 51-56. Jackson, P. and P. Lefrere. On the Application of Rule-Based Techniques to the Design of Advice-Giving Systems. (1984) 20(1) I N T . J O U R N A L M A N -M A C H I N E S T U D I E S 63-86. Johnson, P.E. A.S. Duran, F. Ffassebrock, J. Moller, M . Prictula, P.J. Feltovich and D.B. Swanson. Expertise and Error in Diagnostic Reasoning. (1981) 5(3) C O G N I T I V E S C I E N C E 235-284. Kaplan, S.J. The Industrialization of Artificial Intelligence. (1984) 5(2) A . I . M A G A Z I N E 51. Kelso, Louis O. Does The Law Need A Technological Revolution? (1946) 18 R O C K Y M O U N T A I N L A W R E V I E W 378. K i d d , A.I. and M.B. Cooper. Man-Machine Interface Issues in the Construction and Use of an Expert System. (1985) 22(1) I N T . J O U R N A L M A N - M A C H I N E S T U D I E S 246-259. K i m , Y u n Haeng, Seong-in K i m , and Roberta L . Jacobs. Sentencing by Computer—One Model and It's Implications. (1975) 4 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 283. Kinnucan, Paul. Computers That Think Like Experts. (1984) H I G H T E C H N O L O G Y 37. Kolodner, J.L. Towards an Understanding of the Role of Experience in the Evolution from Novice to Expert. (1983) 19 I N T . J O U R N A L M A N - M A C H I N E S T U D I E S 497-518. Kovach, Richard J. Application of Computer-Assisted Analysis Techniques To Taxation. (1982) 15 A K R O N L A W R E V I E W 713. Kowalski, Robert. Software Engineering and Artificial Intelligence in New Generation Computing. (1984) 1(1) F U T U R E G E N E R A T I O N S C O M P U T E R S Y S T E M S 39-49. Kunz, J . C , T.P. Kehler and M.D. Williams. Applications Development Using a Hybrid Al Development System. (1984) 5(3) A . I . M A G A Z I N E 41. Langlotz, C P . and E . H . Shortliffe. Adapting a Consultation System to Critique User Plans. (1983) 19(5) I N T . J O U R N A L M A N - M A C H I N E S T U D I E S 479-496. Lawlor, Reed C. What Computers Can Do: Analysis and Prediction of Judicial Decisions. (1963) 49 A M E R I C A N B A R A S S O C I A T I O N J O U R N A L 337. Leith, Philip. Cautionary Notes on Legal Expert Systems. (1984) 40 C O M P U T E R S A N D L A W 14. Leith, Philip . Logic, Formal Models and Legal Reasoning. (1984) 24 J U R I M E T R I C S J O U R N A L 334. Lenat, Doug, Mayank Prakash and Mary Shepherd. CYC: Using Common Sense Knowledge to Overcome Brittleness and Knowledge Acquisition Bottlenecks. (1986) 6(4) A . I . M A G A Z I N E 65-85. Liebowitz, Jay. Useful Approach for Evaluating Expert Systems. (1986) 3(2) E X P E R T S Y S T E M S 86. Loevinger, Lee. Jurimetrics: Science and Prediction in the Field of Law. (1961) 46 M I N N E S O T A L A W R E V I E W 255. Loevinger, Lee. Jurimetrics: The Next Step Forward. (1949) 33 M I N N E S O T A L A W R E V I E W 455. Lorinczi , George G . When Does the Computer Engage in Unauthorized Practice? (1968) 34 U N A U T H O R I Z E D P R A C T I C E N E W S 14. Lowry, Houston Putnam. Normalization: A Revolutionary Approach. (1970) 20 J U R I M E T R I C S J O U R N A L 140. Mace, J.R. and P.F. Pope. Tax Planning and Computer Simulation. (1980) B R I T I S H T A X R . 45. Maechling, Charles Jr. Systems Analysis and The Law. (1976) 62 V I R G I N I A L A W R E V I E W 721. Maggs, Peter B. and Cary G. deBessonet. Automated Logical Analysis of Systems of Legal Rules. (1972) 12 J U R I M E T R I C S J O U R N A L 138. Mamdani, F . H . and J. Efstathiou. An Analysis of Formal Logics as Inference Mechanisms in Expert Systems. (1984) 21(3) I N T . J O U R N A L M A N - M A C H I N E S T U D I E S 213-227. Marx, Stephen M . Citation Networks in the Law. (1970) 10 J U R I M E T R I C S J O U R N A L 121. Manuel, Tom. What's Holding Back Expert Systems? (1986) 59(28) E L E C T R O N I C S 59. McCarty, L . Thorne. The Taxman Project: Towards A Cognitive Theory of Legal Argument, C O M P U T E R S C I E N C E A N D L A W . Ed. B. Niblett. London: Cambridge University Press, 1980. McCarty, L . Thorne. Intelligent Legal Information Systems: Problems and Prospects. (1983) 9 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 265. McCarty, L . Thorne. Reflections on TAXMAN: An Experiment in Artificial Intelligence and Legal Reasoning. (1977) 90 H A R V A R D L A W R E V I E W 837. McCarty, L . Thorne and N.S. Sridharan. The Representation of an Evolving System of Legal Concepts II: Prototypes and Deformations. 7TH I N T E R N A T I O N A L J O I N T C O N F E R E N C E O N A R T I F I C I A L I N T E L L I G E N C E . Vancouver, 1981. McCoy, R.W. and W.A. Chatterton. Computer-Assisted Legal Services. (1968) 1 L A W A N D C O M P U T E R T E C H N O L O G Y 2. Meldman, Jeffrey A. A Structural Model For Computer-Aided Legal Analysis. (1977) 6 R U T G E R S C O M P U T E R & T E C H N O L O G Y L A W J O U R N A L 27. Meyer, Perry. Jurimetrics: The Scientific Method In Legal Research. (1966) 44 C A N A D I A N B A R R E V I E W 1. Meyer, Perry. The Computer—A Replacement for Lawyers? (1966) C A N A D I A N B . P A P E R S 111. 120 Michaelsen, Robert H . An Expert System For Federal Tax Planning. (1984) 1 E X P E R T S Y S T E M S 149. Miller, John. Expert Systems—The End of the Legal Profession? [1984] N E W Z E A L A N D L A W J O U R N A L 85. Millus, Albert J. Would You Settle for a Robot on the Bench? (1985) 57 N E W Y O R K S T A T E B A R J O U R N A L 22. Morrise, Mark. Emerging Computer-Assisted Legal Analysis. (1980) B R I G H A M Y O U N G U N I V E R S I T Y L A W R E V I E W 116. Mulder, Richard V . and Helen M . Gubby. Legal Decision Making By Computer: An Experiment With Sentencing. 4 C O M P U T E R L A W J O U R N A L 243. Myers, Martha A. Predicting the Behavior of Law: A Test of Two Models. (1980) 14 L A W A N D S O C I E T Y R E V I E W 835. Nagel, Stuart S. Bibliography on Decision Science Applied to Law Practice. (1985) 25 J U R I M E T R I C S J O U R N A L 197. Nau, D. Expert Computer Systems. (1983) 16 C O M P U T E R 63-85. Niblett, Bryan. Expert Systems For Lawyers. (1981) 29 C O M P U T E R S A N D L A W 2. O'Neil, D.P. and G.D. Wood. Artificial Intelligence and Law. (1985) iv(4) I N F O R M A T I O N T E C H N O L O G Y Q U A R T E R L Y 3. Pashukanis, E.B. The General Theory of Law and Marxism, S O V I E T L E G A L P H I L O S O P H Y . Trans. Hugh Babb. Cambridge, Mass.: Harvard University Press, 1951. Peterson, M . and D.A. Waterman. An Expert Systems Approach to Evaluting Product Liability Cases, C O M P U T I N G P O W E R A N D L E G A L R E A S O N I N G . Ed. Charles Walter. St. Paul, Minnesota: West Publishing Co., 1985. 121 Popp, Walter G. and Bernhard Schlink. JUDITH, A Computer Program To Advise Lawyers in Reasoning a Case. (1975) 15 J U R I M E T R I C S J O U R N A L 303. Richer, Mark H . An Evaluation of Expert System Development Tools. (1986) 3(3) E X P E R T S Y S T E M S 166. Rissland, Edwina L . Artificial Intelligence and Legal Reasoning. P R O C E E D I N G S O F T H E 9 T H I J C A I - 8 5 . University of Calif., Los Angeles, 1985. Rissland, Edwina L . Examples in Legal Reasoning: Legal Hypotheticals. P R O C E E D I N G S O F T H E 8 T H I J C A I - 8 3 . Karlsruhe, W. Germany, 1983. Rueschemeyer, Dietrich. Lawyers and Doctors: A Comparison of Two Professions. S O C I O L O G Y O F L A W . Ed . Vilhelm Aubert. England: Penguin, 1969. Ruoff, Theodore. The Solicitor and the Silicon Chip—Drafting Without Tears. (1980) 124 S O L I C I T O R ' S J O U R N A L 285. Sagalowicz, D. Development of an Expert System. (1984) 1(2) E X P E R T S Y S T E M S 137-147. Samek, R .A . Language, Communication, Computers and the Law. (1985) 9 D A L H O U S I E L A W J O U R N A L 196. Saxon, Charles S. Computer-aided Drafting of Legal Documents. [1982] 3 A M E R I C A N B A R F O U N D A T I O N R E S E A R C H J O U R N A L 685. Schank, R .C . The Current State of Al: One Man's Opinion. (1983) 4(4) A . I . M A G A Z I N E 3. Schloobohm, Dean A. TA—A Prolog Program Which Analyzes Income Tax Issues Under Section 318(a) of the Internal Revenue Code, C O M P U T I N G P O W E R A N D L E G A L R E A S O N I N G . Ed. Charles Walter. St. Paul, Minnesota: West Publishing Co., 1985. Sherman, D .M. Interactive Computers and Legal Problem Solving. 15 L A W S O C I E T Y G A Z E T T E 368. Shpilberg, D., L . E . Graham and H . Schatz. ExperTAXsm: An Expert System for Corporate Tax Planning. (1986) 3 E X P E R T S Y S T E M S 136. Smith, R . G . On the Development of Commercial Expert Systems. (1984) 5(3) A . I . M A G A Z I N E 61-73. Soloman, Lewis D. and Garry S. Grossman. Computers and Legal Reasoning. (1982) 121 T R U S T S A N D E S T A T E S 43. Spengler, Joseph J. Machine-Made Justice: Some Implications. 28 L A W A N D C O N T E M P O R A R Y P R O B L E M S 37. Sprowl, James A. Automating the Legal Reasoning Process: A Computer That Uses Regulations and Statutes to Draft Legal Documents. [1979] 1 A M E R I C A N B A R F O U N D A T I O N R E S E A R C H J O U R N A L 3. Sprowl, James A. and Ronald W. Staudt. Computerizing Client Services in the Law School Teaching Clinic: An Experiment in Law Office Automation. [1981] 3 A M E R I C A N B A R F O U N D A T I O N R E S E A R C H J O U R N A L 699. Staudt, R.W. and J.A. Sprowl. Automating Administrative Systems in Law School Teaching Clinic: Designing a Computer System to Process Case and Time Data for Management and Research. [1981] 3 A M E R I C A N B A R F O U N D A T I O N R E S E A R C H J O U R N A L 1111. Stefik, M . , J. Aikins, R. Balzer, J. Benoit, L . Birnbaum, F. Hayes-Roth and E . Sacerdoti. The Organization of Expert Systems, A Tutorial. (1982) 18(2) A . I . M A G A Z I N E 135-173. Susskind, Richard Eric. Expert Systems in Law: A Jurisprudential Approach To Artificial Intelligence and Legal Reasoning. (1986) 49 M O D E R N L A W R E V I E W 168. Thomas, Bart T. Unauthorized Practice and Computer Aided Legal Analysis Systems. (1979) 20 J U R I M E T R I C S J O U R N A L 41. Thompson, B. A. and W.A. Thompson. Inside An Expert System. (April 1985) B Y T E 315-330. Trinade, F .A . The Principles Governing the Recovery for Negligently Caused Nervous Shock. (1986) 45 C A M B R I D G E L A W J O U R N A L 476. Tyree, Alan L . Can A "Deterministic" Computer Judge Overrule Himself? (1980) 7 J O U R N A L O F C O M P U T E R S , T E C H N O L O G Y A N D L A W 381. Waterman, D.A. User-Oriented Systems for Capturing Expertise: A Rule-Based Approach, E X P E R T S Y S T E M S I N T H E M I C R O - E L E C T R O N I C A G E . Ed. D. Michie. Scotland: Edinburgh University Press, 1979. Waterman, D.A. , J. Paul and M.A. Peterson. Expert Systems for Legal Decision Making. (1986) 3 E X P E R T S Y S T E M S 212. Waterman, D.A. and M.A. Peterson. Evaluating Civil Claims: An Expert Systems Approach. (1984) 1 (1) E X P E R T S Y S T E M S 65-76. Waterman, D.A. and M.A. Peterson. Models of Legal Decision Making. E X P E R T S Y S T E M S , T E C H N I Q U E S , T O O L S , A N D A P P L I C A T I O N S . Eds. P.Klahr and D.A. Waterman. Menlo Park, California: Addison-Wesley Publishing Co., 1986. Welch, John T. LAWGICAL: An Approach to Computer-Aided Legal Analysis. (1982) 15 A K R O N L A W R E V I E W 655. Wiener, Frederick Bernays. Decision Prediction by Computers: Nonsense Cubed—and Worse. (1962) 48 A M E R I C A N B A R A S S O C I A T I O N J O U R N A L 1023. Wilkins, Leslie T. and Hans Toch. Machine Inferences and Clinical Judgments - A Debate. (1984) 11 C R I M I N A L J U S T I C E A N D B E H A V I O R 387. Willick, Marshal S. ARTIFICIAL INTELLIGENCE: Some Legal Approaches and Implications. (1983) IV(2) A . I . M A G A Z I N E 5. Zeide, Janet S. and Jay Liebowitz. Using Expert Systems: The Legal Perspective. (Spring 1987) I E E E E X P E R T 19. A P P E N D I X A S A M P L E C O N S U L T A T I O N S S A M P L E C O N S U L T A T I O N N O . 1 H Y P O T H E T I C A L A man is standing on a downtown street talking to a friend. His ten-year-old son is part way down the same block on the sidewalk watching workmen on a construction site where a new office tower is being built. The skeleton of the building is in place and work is going on at several different levels which are as yet unenclosed. On one of the upper floors, a workman has negligently failed to secure a piece of machinery properly. The machine begins to roll towards the edge of the building where it borders the street. One of the workmen yells a warning just before the machine rolls over the edge. The father turns and sees the machine fall and crush his son to death. He suffers nervous shock. He is often depressed about his son's death and sometimes contemplates suicide. Can he recover? C O N S U L T A T I O N F L E X >go * .* .* .* Welcome to Nervous Shock Advisor *-*-*-* I will tell you whether or not your client has a cause of action in nervous shock. Simply type your answers to my questions on the keyboard located below the screen. If you wish to know why a particular question is being asked, feel free to type "why" in response to the question. You may also type "unknown" if you are unable to answer a particular question. If so, I shall canvass whatever other options are available to me. [The banner appears whenever a consultation is started, but will not be duplicated in the subsequent examples.] Was the incident which caused the nervous shock the result of an act on the part of the defendant which you, as a lawyer, would characterize as negligent or deliberate with respect to any of its harmful results? 1. negligent 2. deliberate » 1 Was the nervous shock sustained as the result of an incident in which any person or persons, including the plaintiff, suffered physical harm or the risk of such harm? ( N O T E : Option No.3 necessarily implies that there was more than one victim). 1. harm 2. risk 3. both 4. neither » 1 Who was the victim of the incident which caused the nervous shock: the plaintiff, a third-party or both? ( N O T E : If there was more than one third-party victim, your responses to any questions about a third-party should relate to the most serious case only). 1. plaintiff 2. third-party 3. both » 2 Was the other person killed? » y e s Is your client closely related to the other person? » y e s Is the other person a member of your client's immediate family; namely, the child, spouse, parent or sibling of your client? ( N O T E : Regard a step-relationship as equivalent to a natural one if the person was treated like a member of the natural family and a grandparent as equivalent to a parent if the grandparent had been filling the role of a parent). 1. child 2. spouse 3. parent 4. sibling 5. neither » 1 Did your client see the incident happen? » y e s Did the nervous shock cause an expense, a loss of income or an incapacity which was sufficiently serious to support a claim for damages? » y e s Into which of the following general categories did the symptoms of nervous shock fall: physical illness; mental disorder; sleep disturbance? Separate each option by a comma if more than one is applicable 1. physical-illness 2. mental-disorder 3. sleep-disturbance » 2 J U D G M E N T : an action lies 100 percent. The following is a breakdown of the various elements involved in this decision. If any element has a certainty factor of less than 100% you should regard it as a weak link in the chain and focus on bolstering your case in that area. Negligence established 100 percent. Appropriate triggering event 100 percent. Third-party principal victim 100 percent. Certainty not increased because only one victim involved. Sufficiently serious incident 100 percent. Plaintiff connected to the victim 100 percent. Plaintiff exposed to incident 100 percent. Damage proved 100 percent. I am pleased to inform you that your client has a cause of action for nervous shock. Would you like to peruse the cases which support my decision? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O Y O U R F A C T S I T U A T I O N : * D E A T H — C H I L D — W I T N E S S * C A S E S O N POINT: 1 Howes v. Crosby (1984), 45 O.R. (2d) 449, 6 D.L.R. (4th) 698, 29 C . C . L . T . 60 (C.A.). 2 Richards v. Baker, [1943] S.A.S.R. 245 (S.C). * Other Family Relationships 3 Storm v. Geeves, [1965] Tas. S.R. 252 (S.C). 4 Walker v. Broadfoot, [1958] O.W.N. 173 (H.C). * R E L E V A N T B Y A N A L O G Y : * Injury—Related—Witness 5 Pollard v. Makarchuk (1958), 26 W.W.R. 22 (Alta. S.C). Pg Dn to continue ... Enter the number of the case you wish to see: 1 [When the selection is made and the "Enter" key pressed, the text of the case summary appears on a new screen.] Howes v. Crosby (1984), 45 O.R. (2d) 449, 6 D.L.R. (4th) 698, 29 C . C . L . T . 60 (C.A.). F A C T S The plaintiff mother was walking along the shoulder of a road holding her young daughter's hand when the girl was struck from behind by a 128 car driven negligently by the defendant. The brother and sister of the victim were also walking with their mother and witnessed the accident. The father came to the scene immediately and attempted to give mouth to mouth resuscitation to his child but the young victim died. The entire family sued and recovered for nervous shock. S Y M P T O M S P. 452 ... "As noted earlier, the attack is made on the awards to Susan and David. Janet was five at the time of her death, Susan was seven and David was sixteen. (a) Susan [In cases which are relevant for their F A C T S , the full S Y M P T O M S portion of the case will not be shown.] [Exit from data file.] You said that your client had mental problems. Would you like to peruse the cases which relate to these symptoms? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O M E N T A L S Y M P T O M S : 1 Walker v. Broadfoot, [1958] O.W.N. 173 (H.C.) (depression, suicide). 2 Hinz v. Berry, [1970] 2 Q.B. 40, [1970] 2 W.L.R. 684, [1970] 1 A l l E.R. 1074 (CA. ) (depression, suicidal ruminations). 3 Brice v. Brown, [1984] 1 A l l E.R. 997 (Q.B.) (insanity). 4 Tsanaktsidis v. Oulianoff (1980), 24 S.A.S.R. 500 (S.C.) (severe depression). 5 Montgomery v. Murphy (1982), 37 O.R. (2d) 631, 136 D.L.R. (3d) 525 (H.C.) (depression). 6 Kohn v. State Government Insurance Commission (1976), 15 S.A.S.R. 255 (S.C.) (intensified depression). 7 McLoughlin v. O'Brian, [1982] 2 W.L.R. 982, [1982] 2 A l l E.R. 298 (H.L.) (personality change). 8 Fenn v. City of Peterborough (1976), 14 O.R. (2d) 137, 1 C . C . L . T . 90, 73 D.L .R. (3d) 177, varied 25 O.R. (2d) 399, 9 C . C . L . T . 1 ( C A . ) (loss of pride and self-respect). 9 Hogan v. City of Regina (sub. nom. McNally v. City of Regina), [1924] 2 W.W.R. 307, 18 Sask L .R. 423, [1924] 2 D.L.R. 1211 (CA. ) (amnesia). Pg Dn to continue ... Enter the number of the case you wish to see: 2 Hinz v. Berry, [1970] 2 Q.B. 40, [1970] 2 W.L.R. 684, [1970] 1 All E.R. 1074 (C.A.). FACTS The plaintiff, her husband, and their children were on an outing. The plaintiff was picking flowers on the one side of the street while her husband was on the other near the parked van with the children. A car driven by the defendant went out of control and crashed into the husband and children. Upon hearing the crash, the plaintiff turned around and saw the aftermath of the accident. There was blood everywhere. She ran across the road to attend to her family. The husband died and the children were severely injured. As a result of witnessing this tragedy the plaintiff suffered a morbid depression and had frequent suicidal ruminations. Her claim for recovery in nervous shock was permitted. SYMPTOMS P. 43 ..."she is suffering from a morbid depression; she is now officially ill. ... irritable with the children ... exhausted, has frequent suicidal ruminations and at the same time is covered with guilt at being like this. At the trial, five years after the accident, she frequently broke down when giving evidence." P. 44 ... "period of not far short of six years, in a sad and depressed state. ... she has been and still is in a positively morbid state. There is a recognisable psychiatric illness. ... the lady is not only sad and depressed but is in a morbid state of depression..." [Exit from data file.] The most effective defence to an action in nervous shock is often that of insufficient proof of damages. Would you like to see the cases where this argument succeeded? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O Y O U R F A C T S I T U A T I O N : * N O P R O O F O F L E G A L L Y R E C O G N I Z E D D A M A G E S * Nervous shock is a legal, not a medical term. It includes a variety * of symptoms resulting from the mental trauma produced by an accident * or its aftermath. In contrast, damages are not recoverable for pure * grief and emotional suffering. Damages for nervous shock are often * difficult to prove so often the most effective defence is the argument * that there is insufficient proof of the kind of damages which are * recoverable in an action for nervous shock. * Plaintiff Harmed A n d Third Party Harmed 1 Griffiths v. C.P.R. (1978), 6 B.C.L.R. 115 (C.A.). 2 Kernested v. Desorcy, [1978] 3 W.W.R. 623, affirmed [1979] 1 W.W.R. 512 (Man. C.A.). 3 Rowe v. McCartney, [1976] 2 N.S.W.L.R. 72 (C.A.). * Plaintiff Risk A n d Third Party Harmed Pg Dn to continue ... Enter the number of the case you wish to see: 3 Rowe v. McCartney, [1976] 2 N.S.W.L.R. 72 (C.A.). F A C T S The plaintiff was the owner of a car and reluctantly permitted her friend to drive it. The plaintiff and her friend were involved in a motor vehicle accident and, as a result, the plaintiff suffered some physical injuries and her friend was rendered a quadriplegic. The accident was due solely to the negligence of the plaintiff's friend but the plaintiff suffered depressive neurosis as a result of her feelings of guilt for having permitted her friend to drive the car. Her claim for recovery on the basis of nervous shock was dismissed. [Exit from data file.] Would you like to read what is currently the leading case on the law of nervous shock in order to get a good overview of the area? » y e s Press Enter to continue... L E A D I N G C A S E 1 Ful l Text: McLoughlin v. O'Brian, [1982] 2 A l l E.R. 298 (H.L.). * The following are also important and leading cases for the law * of nervous shock. You may wish to look at their summaries. 2 Jaensch v. Coffey (1984), 54 A . L . R . 417, 58 A.L.J .R. 426 (H.C.), affirming 33 S.A.S.R. 254 (Sup. Ct. F.C.). 3 Hambrook v. Stokes Brothers, [1925] 1 K.B . 141 (C.A.). 4 Hinz v. Berry, [1970] 2 Q.B. 40, [1970] 2 W.L.R. 684, [1970] 1 A l l E.R. 1074 (C.A.). 5 Boardman v. Sanderson, [1964] 1 W.L.R. 1317 (C.A.). 6 K i n g v. Phillips, [1953] 1 Q.B. 429, [1953] 2 W.L.R. 526, [1953] 1 A l l E.R. 617 (C.A.). 7 Hay (or Bourhill) v. Young, [1943] A . C . 92, [1942] 2 A l l E.R. 396 (H.L.). Enter the number of the case you wish to see: 1 Ful l Text: McLoughlin v. O'Brian, [1982] 2 A l l E.R. 298 (H.L.). F A C T S Lord Wilberforce, Lord Edmund-Davies, Lord Russell of Killowen, Lord Scarman and Lord Bridge of Harwich 15, 16 February, 6 May 1982 A P P E A L The plaintiff, Rosina McLoughlin, appealed against the judgment of the Court of Appeal (Stephenson, Cumming-Bruce and Griffiths LJJ) ([1981] 1 A l l E R 809, [1981] QB 599) given on 16 December 1980 dismissing her appeal against the judgment of Boreham J on 11 December 1978 whereby the judge dismissed her claim against the defendants, Thomas Alan O'Brian, A E Docker & Sons Ltd , Raymond Sygrove and Ernest Doe & Sons Ltd , the respondents to the appeal, for damages for shock, distress and injury to her health. The facts are set out in the opinion of Lord Wilberforce. Michael Ogden Q C and Jonathan Haworth for the appellant. Michael Turner Q C and John Leighton Williams for the respondents. Their Lordships took time for consideration. 132 6 May. The following opinions were delivered. L O R D W I L B E R F O R C E . My Lords, this appeal arises from a very serious and tragic road accident which occurred on 19 October 1973 near Withersfield, Suffolk. The appellant's husband, Thomas McLoughlin, and three of her children, George, aged 17, Kathleen, aged 7, and Gil l ian, nearly 3, were in a Ford motor car; George was driving. A fourth child, Michael, then aged 11, was a passenger in a following motor car driven by Mr. Pilgrim; this car did not become involved in the accident. The Ford car was in collision with a lorry driven by the first respondent and owned by the second respondent. That lorry had been in collision with another lorry driven by the third respondent and owned by the fourth respondent. It is admitted that the accident to the Ford car was caused by the respondents' negligence. It is necessary to state what followed in full detail. As a result of the accident, the appellant's husband suffered bruising and shock; George suffered injuries to his head and face, cerebral concussion, fractures of both scapulae and bruising and abrasions; Kathleen suffered concussion, fracture of the right clavicle, bruising, abrasions and shock; Gil l ian was so seriously injured that she died almost immediately. At the time, the appellant was at her home about two miles away; an [Note: The judgment continues for many more pages.] [Exit from data file.] Thank you for consulting Nervous Shock Advisor. F L E X > 133 S A M P L E C O N S U L T A T I O N N O . 2 H Y P O T H E T I C A L Jane is in her early twenties and lives with her boyfriend Dick who is approximately the same age. They both commute by bus to their jobs in different areas of the downtown core. They usually arrive home at their trendy townhouse at about the same time. One day Jane is walking home from the bus stop, when she sees a group of people standing around a person lying in the roadway who has been struck by a negligent driver some minutes earlier. Jane discovers that the victim is Dick and that he has been seriously injured. She suffers nervous shock. Dick ultimately makes a ful l recovery, but in the meantime Jane has recurring nightmares about this awful experience. Can she recover? C O N S U L T A T I O N Was the incident which caused the nervous shock the result of an act on the part of the defendant which you, as a lawyer, would characterize as negligent or deliberate with respect to any of its harmful results? 1. negligent 2. deliberate » 1 Was the nervous shock sustained as the result of an incident in which any person or persons, including the plaintiff, suffered physical harm or the risk of such harm? ( N O T E : Option No.3 necessarily implies that there was more than one victim). 1. harm 2. risk 3. both 4. neither » 1 Who was the victim of the incident which caused the nervous shock: the plaintiff, a third-party or both? ( N O T E : If there was more than one third-party victim, your responses to any questions about a third-party should relate to the most serious case only). 1. plaintiff 2. third-party 3. both » 2 Was the other person killed? » n o Did the other person suffer visibly serious injuries? » y e s Is your client closely related to the other person? » n o Did your client and the other person live together in an intimate personal relationship analogous to a marriage? » y e s 134 Did your client see the incident happen? » n o Did your client arrive at the scene right after the accident and see the condition of the other person? » y e s Did the nervous shock cause an expense, a loss of income or an incapacity which was sufficiently serious to support a claim for damages? » y e s Into which of the following general categories did the symptoms of nervous shock fall: physical illness; mental disorder; sleep disturbance? Separate each option by a comma if more than one is applicable 1. physical-illness 2. mental-disorder 3. sleep-disturbance » 3 JUDGMENT: an action lies 70 percent. The following is a breakdown of the various elements involved in this decision. If any element has a certainty factor of less than 100% you should regard it as a weak link in the chain and focus on bolstering your case in that area. Negligence established 100 percent. Appropriate triggering event 100 percent. Third-party principal victim 100 percent. Certainty not increased because only one victim involved. Sufficiently serious incident 100 percent. Plaintiff connected to the victim 70 percent. Plaintiff exposed to incident 100 percent. Damage proved 100 percent. I am pleased to inform you that your client has a cause of action for nervous shock. Would you like to peruse the cases which support my decision? » y e s Press Enter to continue... 135 T H E F O L L O W I N G CASES A R E R E L E V A N T T O Y O U R F A C T S I T U A T I O N : T H I R D P A R T Y I N J U R Y - N E A R B Y A N D A T T E N D E D - L I V I N G T O G E T H E R * R E L E V A N T B Y A N A L O G Y : | 1 Marshall v. Lionel Enterprises Inc., [1972] 2 O.R. 177, 25 D.L.R. (3d) | 141 (H.C.). | * Minor Injury—Lucky Escape—Related—Witness « 2 Boardman v. Sanderson, [1964] 1 W.L.R. 1317 (C.A.). 3 Pollard v. Makarchuk (1958), 26 W.W.R. 22 (Alta. S.C). * Death-Related-Plaint i f f Nearby and Attended 4 Hinz v. Berry, [1970] 2 Q.B. 40, [1970] 2 W.L.R. 684, [1970] 1 A l l E.R. 1074 (C.A.). 5 Howes v. Crosby (1984), 45 O.R. (2d) 449, 6 D.L.R. (4th) 698, 29 C . C . L . T . 60 (C.A.). Pg Dn to continue ... Enter the number of the case you wish to see: 1 Marshall v. Lionel Enterprises Inc., [1972] 2 O.R. 177, 25 D.L .R. (3d) 141 (H.C). F A C T S The victim, husband of the plaintiff, purchased a snowmobile with a defective clutch from the defendant. The husband suffered grievous injury while riding the vehicle. The plaintiff did not see the actual accident but came upon the aftermath of it, saw her injured husband, and suffered nervous shock as a result. The defendants attempted to have the plaintiff's statement of claim struck but the court dismissed this motion. [Exit from data file.] You said that your client had sleep problems. Would you like to peruse the cases which relate to these symptoms? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O S L E E P S Y M P T O M S : 1 Howes v. Crosby (1984), 45 O.R. (2d) 449, 6 D.L.R. (4th) 698, 29 C . C . L . T . 60 ( C A . ) (insomnia, nightmares). 2 Brice v. Brown, [1984] 1 A l l E.R. 997 (Q.B.) (insomnia). 3 Storm v. Geeves, [1965] Tas. S.R. 252 (S.C.) (sleeplessness, nightmares). 4 Fenn v. City of Peterborough (1976), 14 O.R. (2d) 137, 1 C . C . L . T . 90, 73 D.L .R. (3d) 177, varied 25 O.R. (2d) 399, 9 C . C . L . T . 1 (CA. ) (nightmares). 5 Hogan v. City of Regina (sub. nom. McNally v. City of Regina), [1924] 2 W.W.R. 307, 18 Sask L.R. 423, [1924] 2 D.L.R. 1211 (CA. ) (insomnia). Enter the number of the case you wish to see: 3 Storm v. Geeves, [1965] Tas. S.R. 252 (S.C). F A C T S The plaintiff mother, who was inside her home, heard the noise of an accident from outside but did not know what it was. The plaintiff sibling, who had witnessed the accident, immediately informed her that her daughter had been struck by a truck outside the house. The plaintiff mother ran out to see her child squashed beneath a truck. Due to difficulties in releasing the trapped child, she was pinned beneath the truck for a good portion of an hour. The child victim died. As a result, the plaintiff mother fell in a state of reactive depression for a 15 month period following the death and suffered recurring nightmares. The plaintiff sibling showed a marked behavioural change and became silent, moody, depressive and apathetic. Both plaintiffs recovered for nervous shock. S Y M P T O M S P. 267 ... "She suffers from a recurring nightmare... various sedatives have been prescribed. ... she is still in a state of reactive depression... pathological mourning. ...depressed, obsessed by the loss; she remains sleepless, agitated, tense. ...cried and displayed deep emotion." P. 269 ... "The case of the boy Hendrikus is much more serious. ... a marked change in his personality. He became moody, silent, depressed and apathetic to work and play. His school work deteriorated. ... boy had suffered grave psychic injury,..." [Exit from data file.] The most effective defence to an action in nervous shock is often that of insufficient proof of damages. Would you like to see the cases where this argument succeeded? » n o Would you like to read what is currently the leading case on the law of nervous shock in order to get a good overview of the area? » n o Thank you for consulting Nervous Shock Advisor. F L E X > 138 S A M P L E C O N S U L T A T I O N N O . 3 H Y P O T H E T I C A L Elmer is a bit of a wimp but his jock friends manage to persuade him to take up windsurfing by assuring him that the sport is perfectly safe. One day about a month later, Elmer is practising his rudimentary skills in an area just off the beach reserved for novice windsurfers. The area has been cordoned off by floats. Tony, a daredevil sailor, is out in the bay on his sail boat with a girlfriend. There is a stiff breeze blowing and Tony has one hand on the tiller and the other arm around the lady's waist. Turning to glance at his girlfriend, Tony becomes mesmerized briefly by her beauty. In the meantime, his speedy boat sails onward. He negligently fails to keep a proper lookout and sails into the supposedly safe area where Elmer is barely managing to balance on his board. When Elmer sees Tony's boat bearing down on him, he is paralyzed by fear. At the last moment Tony snaps out of his trance and takes evasive action. The hull of his boat strikes Elmer's windsurfer a glancing blow and Elmer falls into the water. As he flounders around in a panic, the mast of the windsurfer lands on his head and he suffers a small cut above the eye. The injury is not serious and no stitches are required. However because of his delicate disposition, Elmer suffers nervous shock. Every time thereafter that he sees a sailboard he feels nauseated. Eventually the strain gives him an ulcer and he requires surgery. Can he recover? C O N S U L T A T I O N Was the incident which caused the nervous shock the result of an act on the part of the defendant which you, as a lawyer, would characterize as negligent or deliberate with respect to any of its harmful results? 1. negligent 2. deliberate » 1 Was the nervous shock sustained as the result of an incident in which any person or persons, including the plaintiff, suffered physical harm or the risk of such harm? ( N O T E : Option No.3 necessarily implies that there was more than one victim). 1. harm 2. risk 3. both 4. neither » 1 Who was the victim of the incident which caused the nervous shock: the plaintiff, a third-party or both? ( N O T E : If there was more than one third-party victim, your responses to any questions about a third-party should relate to the most serious case only). 1. plaintiff 2. third-party 3. both » 1 Was your client injured, even slightly, in circumstances that could well have resulted in far more serious injury? » y e s Did the nervous shock cause an expense, a loss of income or an incapacity which was sufficiently serious to support a claim for damages? » y e s Into which of the following general categories did the symptoms of nervous shock fall: physical illness; mental disorder; sleep disturbance? Separate each option by a comma if more than one is applicable 1. physical-illness 2. mental-disorder 3. sleep-disturbance » 1 J U D G M E N T : an action lies 100 percent. The following is a breakdown of the various elements involved in this decision. If any element has a certainty factor of less than 100% you should regard it as a weak link in the chain and focus on bolstering your case in that area. Negligence established 100 percent. Appropriate triggering event 100 percent. Plaintiff principal victim 100 percent. Certainty not increased because only one victim involved. Sufficiently serious incident 100 percent. Connection to a third-party victim not essential to recovery on these facts. Plaintiff exposed to incident 100 percent. Damage proved 100 percent. I am pleased to inform you that your client has a cause of action for nervous shock. Would you like to peruse the cases which support my decision? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O Y O U R F A C T S I T U A T I O N : * P L A I N T I F F O N L Y - H A R M * C A S E S O N POINT: 1 Negro v. Pietro's Bread Co., [1933] O.R. 112, [1933] 1 D.L.R. 490 (C.A.). * R E L E V A N T B Y A N A L O G Y : * Plaintiff O n l y - R i s k 2 Dulieu v. White & Sons, [1901] 2 K.B . 669. 3 Duwyn v. Kaprielian (Infant) (1978), 22 O.R. (2d) 736, 7 C . C . L . T . 121, 94 D.L .R. (3d) 424 (C.A.). 4 Taylor v. Weston Bakeries Ltd. (1976), 1 C . C . L . T . 158 (Sask. Dist. Ct.). * Plaintiff Harmed A n d Third Party Harmed Pg Dn to continue ... Enter the number of the case you wish to see: 1 Negro v. Pietro's Bread Co., [1933] O.R. 112, [1933] 1 D.L.R. 490 (C.A.). F A C T S The plaintiff purchased bread from the defendant company and, while eating the bread, discovered broken pieces of glass. The plaintiff was able to remove the glass from within his mouth with no damage save a slight scratch of the throat. However, the plaintiff suffered extreme shock and excitement over the possible serious consequences and, from then on, he became nauseous whenever he attempted to eat bread. The plaintiff's claim for nervous shock was permitted and Coultas was not followed. S Y M P T O M S P. 115 ... "The result of this excitement and fear of serious consequences and the shock suffered by finding himself in what he thought to be serious danger produced a deplorable condition of nausea which was renewed whenever he attempted to eat bread. [Exit from data file.] You said that your client had physical problems. Would you like to peruse the cases which relate to these symptoms? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O P H Y S I C A L S Y M P T O M S : 1 Pollard v. Makarchuk (1958), 26 W.W.R. 22 (Alta. S.C.) (headaches, vomiting, weightloss). 2 Negro v. Pietro's Bread Co., [1933] O.R. 112, [1933] 1 D.L .R. 490 ( C A . ) (nausea). 3 Schneider v. Eisovitch, [1960] 2 Q.B. 430, [1960] 2 W.L.R. 169, [1960] 1 A l l E.R. 169 (dermatitis). 4 Richards v. Baker, [1943] S.A.S.R. 245 (S.C) (exacerbation of thyroid condition). 5 McLaughlin v. Toronto Railway Co. (1916), 9 O.W.N. 407, varied as to the amount of damages 10 O.W.N. 135 (CA. ) (miscarriage). 6 Jaensch v. Coffey (1984), 54 A . L . R . 417, 58 A.L.J .R. 426 (H.C) , affirming 33 S.A.S.R. 254 (Sup. Ct. F . C ) (gynaecological problems resulting in hysterectomy). 7 Dulieu v. White & Sons, [1901] 2 K.B . 669 (premature delivery). 8 Hambrook v. Stokes Brothers, [1925] 1 K.B . 141 (CA. ) (hemorraging, stillbirth). Pg Dn to continue ... Enter the number of the case you wish to see: 5 McLaughlin v. Toronto Railway Co. (1916), 9 O.W.N. 407, varied as to the amount of damages 10 O.W.N. 135 (C.A.). F A C T S The plaintiff passenger in a night train belonging to the defendant was hit by a metallic sign board which became detached from where it was hanging in the car. The plaintiff's husband, sitting beside her, received a serious cut on his head which had to be stitched. The plaintiff wife was also struck but she sustained only a slight swelling on the top of her head. The plaintiff alleged that she suffered nervous shock and claimed that she miscarried as a result. Plaintiff's claim was permitted. S Y M P T O M S P. 407 ... "The wife was also struck, but the extent of her outward injury was a slight swelling on the top of her head, according to the evidence of a medical man to whom they repaired, before they went home, for the purpose of having the husband's head stitched up. The wife was then in a nervous and excited condition. She swore that she was pregnant at the time. The main contest was as to whether she was ever pregnant, and whether certain appearances which the medical man who attended her (the one who stitched her husband's head) described were consistent with a miscarriage produced by the accident or were attributable to some diseased condition of the uterus." The learned Chief Justice finds that the injury which she suffered, of whatever nature, it was, was the result of (1) the physical injury and (2) of shock resulting therefrom and (3) of the nervous excitement and shock caused by her being present while her husband's injuries were being attended to; but is unable to determine the respective proportions in which these three elements were contributing causes -he attaches more importance to the first and second than to the third." * *N .B . The case is not clear as to whether the nervous shock suffered was a result of the plaintiff's own injuries or as a result of witnessing the more serious injuries sustained by her husband. I would assert that it is more likely the former because of the date of the case; Coultas was still being followed and it mentions that the defendant was attempting to argue that the nervous shock was a result of seeing a doctor attend to her husband's injuries. [Exit from data file.] The most effective defence to an action in nervous shock is often that of insufficient proof of damages. Would you like to see the cases where this argument succeeded? » n o Would you like to read what is currently the leading case on the law of nervous shock in order to get a good overview of the area? » n o Thank you for consulting Nervous Shock Advisor. F L E X > S A M P L E C O N S U L T A T I O N NO. 4 H Y P O T H E T I C A L Mother is in the kitchen preparing lunch. Her four-year-old daughter is playing outside on the unfenced front lawn a few feet from the sidewalk. A car being driven at excessive speed rounds a nearby corner and the negligent driver loses control. The vehicle mounts the sidewalk and its wheels cross the lawn about a foot from where the child is playing. The driver regains control of the car and brings it safely to a stop back on the pavement. He is shaken up but the child is unhurt and not particularly alarmed. While the driver is making sure that no damage has been done, a neighbour who witnessed the incident runs into the house and tells the mother exactly what happened. The mother rushes outside, becomes distraught at the prospect of what might have happened and suffers nervous shock. She suffers from periodic attacks of acute anxiety as a result of the incident. Can she recover? C O N S U L T A T I O N Was the incident which caused the nervous shock the result of an act on the part of the defendant which you, as a lawyer, would characterize as negligent or deliberate with respect to any of its harmful results? 1. negligent 2. deliberate » 1 Was the nervous shock sustained as the result of an incident in which any person or persons, including the plaintiff, suffered physical harm or the risk of such harm? ( N O T E : Option No.3 necessarily implies that there was more than one victim). 1. harm 2. risk 3. both 4. neither » 2 Who was the victim of the incident which caused the nervous shock: the plaintiff, a third-party or both? ( N O T E : If there was more than one third-party victim, your responses to any questions about a third-party should relate to the most serious case only). 1. plaintiff 2. third-party 3. both » 2 Is your client closely related to the other person? » y e s Is the other person a member of your client's immediate family; namely, the child, spouse, parent or sibling of your client? ( N O T E : Regard a step-relationship as equivalent to a natural one i f the person was treated like a member of the natural family and a grandparent as equivalent to a parent if the grandparent had been filling the role of a parent). 1. child 2. spouse 3. parent 4. sibling 5. neither » 1 Did your client see the incident happen? » n o J U D G M E N T : no cause of action. R E A S O N : plaintiff not exposed incident. I regret to inform you that the damage is too remote to permit recovery for nervous shock because your client was not sufficiently involved in the incident. Would you like to peruse the cases which support my decision? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O Y O U R F A C T S I T U A T I O N : * P L A I N T I F F H A D I N S U F F I C I E N T E X P O S U R E T O I N C I D E N T 1 Abramzik v. Brenner (1967), 62 W.W.R. 332, 65 D.L.R. (2d) 651 (Sask. C. A.). 2 Babineau v. MacDonald (No. 2) (1975), 59 D.L .R. (3d) 671, 10 N.B.R. (2d) 715 (C.A.). 3 Duwyn v. Kaprielian (Mother) (1978), 22 O.R. (2d) 736, 7 C . C . L . T . 121, 94 D. L.R. (3d) 424 (C.A.). 4 Bourque v. Surette (1978), 23 N.B.R. (2d) 357, 44 A.P.R. 357 (Q.B.). 5 Dietalbach v. Public Trustee, [1973] 5 W.W.R. 93, 37 D.L .R. (3d) 621 (B.C.S.C.). 6 Bunyan v. Jordan (1937), 57 C.L.R. 1, 54 W.N. 61, 37 S.R. (N.S.W.) 119 (H.C.). 7 Pratt and Goldsmith v. Pratt, [1975] V.R. 378 (Sup. Ct. F .C. Enter the number of the case you wish to see: 3 Duwyn v. Kaprielian (Mother) (1978), 22 O.R. (2d) 736, 7 C . C . L . T . 121, 94 D .L .R . (3d) 424 (C.A.). F A C T S The defendant negligently backed his car into a parked car in which the plaintiff's infant son and mother were seated. The defendant dented the parked car and shattered glass into the interior. No physical injuries were sustained as a result of the accident. The plaintiff mother of the infant was in a nearby building at the time of the accident and upon hearing the commotion, came out to see people gathered around her vehicle and hear her baby screaming. As a result of the accident, the plaintiff mother alleged nervous shock evidenced by her need to take tranquilizers. The mother's claim for nervous shock as a result of the incident was dismissed. S Y M P T O M S P. 755 ... "From the foregoing it seems reasonably clear that the condition of emotional upset of which she justly complains was caused by the constant obligation of having to cope with Brent and not by the initial upset which she underwent immediately following the accident and, as such, is not recoverable (Hinz v. Berry, supra) but further, as indicated, I doubt whether, on the evidence, it qualifies as legally compensable nervous shock." [Exit from data file.] Would you like to read what is currently the leading case on the law of nervous shock in order to get a good overview of the area? » w h y There are usually one or two cases from the highest courts in jurisdictions governed by comparable law which review and restate the law on a particular subject. The persuasive force of these leading cases makes it essential that a lawyer dealing with a similar sort of case be aware of them. Judges assume that lawyers have read these cases. To be unfamiliar with them can be a source of considerable embarrassment. Would you like to read what is currently the leading case on the law of nervous shock in order to get a good overview of the area? » n o Thank you for consulting Nervous Shock Advisor. F L E X > S A M P L E C O N S U L T A T I O N N O . 5 H Y P O T H E T I C A L Max prides himself on being a practical joker. On Apr i l 1st he is walking past a restaurant in Whistler Village carrying his skis over his shoulder after a beautiful day of skiing in the spring sunshine. Crowds of people are sitting at tables outside enjoying drinks in the sun. As he passes by, Judy, the wife of Max's good friend John, calls him over. She tells him that she is waiting for John who is late and wonders whether Max has seen him. Being a quick thinker, Max immediately assumes a tragic air and says to Judy, "Didn't you hear about the avalanche?" Judy's face turns ashen. Max goes on to say that he last saw John heading towards the West Bowl with a group of other skiers and that a ski patroller subsequently told him that there was a massive slab avalanche over there just afterwards. At this point Judy becomes hysterical and collapses. Max tries to explain that it is an Apr i l Fool's joke, but by then Judy requires medical attention. John arrives shortly afterwards and his friendship with Max ends on the spot when he finds out what has happened. Judy needs bed rest for several days because of the nervous shock she has suffered. Can she recover? C O N S U L T A T I O N Was the incident which caused the nervous shock the result of an act on the part of the defendant which you, as a lawyer, would characterize as negligent or deliberate with respect to any of its harmful results? 1. negligent 2. deliberate » 2 J U D G M E N T : an action may well lie for causing nervous shock. R E A S O N : in general, nervous shock resulting from an intentional wrong is not considered to be too remote. However, I am afraid that Nervous Shock Advisor limits itself to providing advice about cases where the nervous shock was suffered as a result of the defendant's negligence. The intentional infliction of nervous shock falls within a different realm of the law of torts. 147 Would you like to peruse the cases which support my decision? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O Y O U R F A C T S I T U A T I O N : * I N T E N T I O N A L L Y I N D U C E D N E R V O U S S H O C K 1 Wilkinson v. Downton, [1897] 2 Q.B. 57. 2 Janvier v. Sweeney, [1919] 2 K.B . 316, affirmed [1919] 2 K.B . 326 (C.A.). 3 Bielitzki v. Obadisk, [1922] 2 W.W.R. 238 (Sask. C.A.). 4 Timmermans v. Buelow (1984), 38 C . C . L . T . 136 (Ont. H . C ) . * N E R V O U S S H O C K R E S U L T I N G F R O M A N I N T E N T I O N A L W R O N G F U L * A C T 5 Purdy v. Woznesensky, [1937] 2 W.W.R. 116 (Sask. C.A.). 6 Stevenson v. Basham, [1922] N.Z.L .R. 225 (Auck. S.C). 7 Edmonds v. Armstrong Funeral Home Ltd., [1931] 1 D.L.R. 676 (Alta. S.C.-A.D.). * H E L D F O R D E F E N D A N T : 8 Radovskis v. Tomn (1957), 65 Man. R. 61, 9 D.L.R. (2d) 751 (Q.B.). Enter the number of the case you wish to see: 1 Wilkinson v. Downton, [1897] 2 Q.B. 57. F A C T S The defendant had the intention of playing a practical joke on the plaintiff and he told her that he had been sent to inform her that her husband had just been hit in a accident and suffered two broken legs. The defendant instructed the plaintiff to fetch pillows and go the accident site to retrieve her husband. The statements were false; however the plaintiff acted on the words spoken by the defendant. As a result of the news, the plaintiff suffered nervous shock to her system, displaying symptoms of vomiting and other effects on her reason. The court held that there is liability for the intentional infliction of nervous shock where there is an element of wilfulness on the part of the defendant and the plaintiff's illness is a direct consequence of the defendant's conduct. The court found that the defendant's intent to play a practical joke in making the statements sufficient to constitute the element of wilful infliction of nervous shock. S Y M P T O M S The plaintiff suffered nervous shock to her system, displaying symptoms of vomiting and other effects on her reason. [Exit from data file.] Would you like to read what is currently the leading case on the law of nervous shock in order to get a good overview of the area? » n o Thank you for consulting Nervous Shock Advisor. F L E X > S A M P L E C O N S U L T A T I O N N O . 6 H Y P O T H E T I C A L Joe, a mild-mannered bank clerk on his summer vacation, arrives at the scene of a serious accident. A crowded tour bus has been struck by a negligently driven train at a crossing in the rural part of beautiful British Columbia. A number of people have been killed and there are severely injured survivors trapped in the mangled wreckage. Joe offers his assistance to the lone R C M P officer who is doing his best to cope with the situation. Needless to say, the offer is gratefully accepted. His exposure to this gruesome scene causes Joe to suffer nervous shock. His mind attempts to suppress the awful memory, resulting in periods of amnesia. The victims were all strangers to him. Can he recover? C O N S U L T A T I O N Was the incident which caused the nervous shock the result of an act on the part of the defendant which you, as a lawyer, would characterize as negligent or deliberate with respect to any of its harmful results? 1. negligent 2. deliberate » 1 Was the nervous shock sustained as the result of an incident in which any person or persons, including the plaintiff, suffered physical harm or the risk of such harm? ( N O T E : Option No.3 necessarily implies that there was more than one victim). 1. harm 2. risk 3. both 4. neither » 1 Who was the victim of the incident which caused the nervous shock: the plaintiff, a third-party or both? ( N O T E : If there was more than one third-party victim, your responses to any questions about a third-party should relate to the most serious case only). 1. plaintiff 2. third-party 3. both » 2 Was the other person killed? » y e s Is your client closely related to the other person? » n o Did your client and the other person live together in an intimate personal relationship analogous to a marriage? » n o Were your client and the other person engaged to be married? » n o Did your client attempt to assist the other person after the incident? » y e s Did the nervous shock cause an expense, a loss of income or an incapacity which was sufficiently serious to support a claim for damages? » y e s Into which of the following general categories did the symptoms of nervous shock fall: physical illness; mental disorder; sleep disturbance? Separate each option by a comma if more than one is applicable 1. physical-illness 2. mental-disorder 3. sleep-disturbance » 2 JUDGMENT: an action lies 70 percent. The following is a breakdown of the various elements involved in this decision. If any element has a certainty factor of less than 100% you should regard it as a weak link in the chain and focus on bolstering your case in that area. Negligence established 100 percent. Appropriate triggering event 100 percent. Third-party principal victim 100 percent. Certainty not increased because only one victim involved. Sufficiently serious incident 100 percent. Plaintiff connected to the victim 70 percent. Plaintiff exposed to incident 100 percent. Damage proved 100 percent. I am pleased to inform you that your client has a cause of action for nervous shock. Would you like to peruse the cases which support my decision? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O Y O U R F A C T S I T U A T I O N : * T H I R D P A R T Y - R E S C U E R * C A S E S O N POINT: 1 Chadwick v. British Transport Commission, [1967] 2 A l l E.R. 945 (Q.B.). 2 Mount Isa Mines Ltd. v. Pusey (1970) 45 A L J R 88. Enter the number of the case you wish to see: 1 Chadwick v. British Transport Commission, [1967] 2 A l l E.R. 945 (Q.B.). F A C T S Following a train collision near his home, the plaintiff voluntarily took an active part in rescue operations at the scene of the accident. The accident claimed ninety lives and many others were trapped and injured. As a result of the horror of the experience, the plaintiff suffered a prolonged and disabling anxiety neurosis which required hospital treatment. The plaintiff's claim for nervous shock was allowed. S Y M P T O M S P. 948-9 ..."After the accident, he started sleeping badly, waking up in the night and talking of the little boy whom he had seen. Mrs. Chadwick found that he was not sleeping and about Christmas time, that is to say four or five weeks after the accident, he stopped working. He was shaking. Mrs. Mills, another witness, whom Mr. Chadwick had helped with whist drives, also described how she saw him change. She first noticed him shaking about five weeks after the accident, and there is no doubt that a marked change was noticed within a few weeks [Exit from data file.] You said that your client had mental problems. Would you like to peruse the cases which relate to these symptoms? » y e s Press Enter to continue... T H E F O L L O W I N G CASES A R E R E L E V A N T T O M E N T A L S Y M P T O M S : | 1 Walker v. Broadfoot, [1958] O.W.N. 173 (H.C.) (depression, suicide). | 2 Hinz v. Berry, [1970] 2 Q.B. 40, [1970] 2 W.L.R. 684, [1970] 1 A l l | E.R. 1074 ( C A . ) (depression, suicidal ruminations). | 3 Brice v. Brown, [1984] 1 A l l E.R. 997 (Q.B.) (insanity). | 4 Tsanaktsidis v. Oulianoff (1980), 24 S.A.S.R. 500 (S.C.) | (severe depression). | 5 Montgomery v. Murphy (1982), 37 O.R. (2d) 631, 136 D.L.R. (3d) 525 (H.C.) | (depression). | 6 Kohn v. State Government Insurance Commission (1976), 15 S.A.S.R. | 255 (S.C.) (intensified depression). | 7 McLoughlin v. O'Brian, [1982] 2 W.L.R. 982, [1982] 2 A l l E.R. 298 | (H.L.) (personality change). | 8 Fenn v. City of Peterborough (1976), 14 O.R. (2d) 137, 1 C . C . L . T . 90, | 73 D.L.R. (3d) 177, varied 25 O.R. (2d) 399, 9 C . C . L . T . 1 (CA. ) | (loss of pride and self-respect). | 9 Hogan v. City of Regina (sub. nom. McNally v. City of Regina), [1924] | 2 W.W.R. 307, 18 Sask L.R. 423, [1924] 2 D.L.R. 1211 (CA. ) | (amnesia). | Pg Dn to continue ... Enter the number of the case you wish to see: 9 Hogan v. City of Regina (sub. nom. McNally v. City of Regina), [1924] 2 W.W.R. 307, 18 Sask. L.R. 423, [1924] 2 D.L.R. 1211 (C.A.). F A C T S The plaintiff mother was strolling with her family when a streetcar belonging to the defendant was negligently backed into them. The streetcar ran over and killed two of the plaintiff's children and knocked the plaintiff down so that she was brought to her hands and knees. However, the plaintiff did not sustain any physical injuries as a result of the fall. The plaintiff sued on the basis that she suffered nervous shock at seeing her two children killed. Her nervous shock manifested itself in symptoms of headaches, dizziness, insomnia, loss of memory and bodily weakness. The plaintiff recovered on the basis that the jury found that physical injury was sustained. SYMPTOMS P. 311 ..."The evidence, however, showed that prior to the accident Mrs. McNally's health had been good, that she had been strong and well, but that ever since the accident she had been suffering from nervous shock, from headaches and dizziness, and from insomnia and loss of memory, as well as bodily weakness. Dr. Rothwell, in his evidence, stated that, while there had been no organic injury, that is, no destruction of any part of the body, Mrs. McNally appeared to him to be suffering from a form of neurosis called neurasthenia, and he gave it as his opinion that this condition was due to the "experience of the shock of the accident, and also the mental worry that continued on after the accident", and he pointed out that from the shock there might be a real physical reaction. He further said that a woman in pregnancy, as Mrs. McNally was at the time, was more susceptible to nervous shock and permanent nervous disabilities than she otherwise would be, and he intimated that Mrs. McNally's condition, if allowed to become permanent, might lead to insanity." P. 317 ..."According to her own evidence, she has been nervous since the accident; somewhat sleepless at night; cannot do her work as well as usual, but in reality has taken charge of her home and family; suffers from headaches and is dizzy at times. Dr. Rothwell testified that she was suffering from functional neurosis, which means that the nervous element in the body is not acting as it ordinarily would, and that such condition generally follows a great muscular and nervous exhaustion, or exhaustion of the body from sickness, and also from emotional shock; that there was no organic destruction or symptoms of any organic charge and that he believed the plaintiff was suffering from emotional shock." P. 319 ... "It has been pointed out in recent judicial decisions that the nervous system is just as much a part of man's physical system as an arm or a limb, perhaps very much more important. The relation between fright or shock and the nervous structure of the body is a matter which depends upon scientific and medical testimony, and I do not see how a Court can lay down as a matter of law that, if negligence causes fright, and such fright in turn affects the nervous structure of the body in such a way as to cause injury to health, such injury cannot be a consequence which, in the ordinary course of things, would flow from the negligence of the defendant. On this point I agree with the reasoning and the conclusion of my brother Lamont." **N.B. In this case, the court followed Coultas and stated that the law required physical injury to the plaintiff before recovery for nervous shock would be permitted. Here the jury found that the plaintiff did sustain physical injuries though the evidence clearly states that she did not suffer any physical injuries as a result of being knocked down by the streetcar. There was impact however. There is no evidence as to whether the plaintiff feared for her own safety. No damages were awarded on the basis of near-miss of serious injury. [Exit from data file.] The most effective defence to an action in nervous shock is often that of insufficient proof of damages. Would you like to see the cases where this argument succeeded? » n o Would you like to read what is currently the leading case on the law of nervous shock in order to get a good overview of the area? » n o Thank you for consulting Nervous Shock Advisor. FLEX > APPENDIX B KNOWLEDGE BASE /* N E R V O U S S H O C K A D V I S O R C A L D E E D M A N , M A R C H 18, 1987 U B C L A W S C H O O L C O M P U T E R S A N D T H E L A W P R O J E C T Copyright (c) 1987 by G.C.Deedman. A l l rights reserved. "Flesh and blood is weak and frail susceptible to nervous shock ... " T.S.Eliot, "The Hippopotamus". This system is designed to provide advice to a legally trained user about whether or not, on a given set of facts, a plaintiff has a cause of action for negligently inflicted nervous shock. The domain expert on whose knowledge the system was based, is Professor J.C.Smith of the Faculty of Law at the University of British Columbia. The knowledge engineer who designed and built the system was G.C.Deedman. The knowledge base is largely self-documenting because of the declarative semantics used to formulate the rules and the text explanations linked to all the questions which the advisor can ask through the rules which trigger them. Additional comments have been added where necessary to clarify the way in which the system operates. */ initialdata = [advice-given]. / * C O N F I G U R A T I O N * / configuration(banner) = [nl, nl, U N I V E R S I T Y O F BRITISH C O L U M B I A ' , n l , n l , F A C U L T Y O F LAW',nl ,nl , IBM-UBC',nl , JOINT P R O J E C T O N C O M P U T E R S IN LAW' , nl.nl.nl]. configuration(prompt) = ' F L E X > '. / * N E G L I G E N C E * / automaticmenu(negligence). enumeratedanswers(negligence). question(negligence) = 'Was the incident which caused the nervous shock the result of an act on the part of the defendant which you, as a lawyer, would characterize as negligent or deliberate with respect to any of its harmful results?'. legalvals(negligence) = [negligent, deliberate]. rule-001: if negligence = negligent then right-tort. explanation(rule-OOl) = [nl, 'There is a fundamental distinction to be drawn in law between cases of negligently inflicted nervous shock and cases where the nervous shock suffered by the victim was attributable to a deliberate unlawful act on the part of the defendant. This Advisor limits itself to giving advice on the law as it relates to negligence.',nl]. rule-002: if negligence is unknown and display([nl, 'Let us proceed with the consultation anyway, but please understand that the results you will get are predicated on the assumption that there WAS indeed negligence. I hope that is clear!',nl,nl]) then right-tort. rule-003: if negligence = deliberate then wrong-tort. / * H A R M - T Y P E • automaticmenu(harm-type). enumeratedanswers(harm-type). presupposition(harm-type) = right-tort. */ question(harm-type) = 'Was the nervous shock sustained as the result of an incident in which any person or persons, including the plaintiff, suffered physical harm or the risk of such harm? ( N O T E : Option No.3 necessarily implies that there was more than one victim).'. legalvals(harm-type) = [harm, risk, both, neither]. /* F A L S E - R E P O R T */ presupposition(false-report) = (harm-type = neither) or (harm-type is unknown). presupposition(false-report) = not(cached(victim-type = X)). presupposition(false-report) = not(cached(victim-type is unknown)). question(false-report) = 'Was the nervous shock caused by the reception of a false report of some matter seriously affecting the plaintiff? (e.g. That a close relative had been killed).'. legalvals(false-report) = [yes, no]. rule-004: if harm-type = harm or harm-type = risk or harm-type = both then valid-harm-claim. explanation(rule-004) = [nl, 'The basic dichotomy in nervous shock cases is between those where there was actual harm and those where there was simply a risk of harm. Where there was harm, the extent of the harm must be determined.',nl]. rule-005: if false-report then valid-news-claim. explanation(rule-005) = [nl, 'In a limited number of cases, the courts have allowed recovery for nervous shock where it was caused by the reception of a particularly distressing piece of false news.',nl]. rule-006: if harm-type = neither or harm-type is unknown then invalid-harm-claim. rule-007: if false-report = no or false-report is unknown then invalid-news-claim. / * P U B L I C A T I O N * / presupposition(publication) = false-report. question(publication) = 'Did the defendant disseminate or cause the dissemination of the false report?'. legalvals(publication) = [yes, no]. rule-008: if publication then false-news. explanation(rule-008) = [nl, Tn cases of nervous shock induced by the reception of a false report, it must be established that the defendant was responsible for the report being "published" in the legal sense of the word.',nl]. rule-009: if false-report and publication = no or publication is unknown then false-news = no. rule-010: if false-news then serious-incident cf 55. rule-Oil: if false-news then connection-to-victim. rule-012: if false-news then exposure-to-incident. / * B E S T - D E F E N C E - * / presupposition(best-defence) = cached(judgment = an-action-lies cf X). question(best-defence) = 'The most effective defence to an action in nervous shock is often that of insufficient proof of damages. Would you like to see the cases where this argument succeeded?'. legalvals(best-defence) = [yes, no]. rule-013: if best-defence = no or best-defence is unknown then defences-given. explanation(rule-013) = [nl, 'It is important for lawyers to understand the arguments on both sides of a case if they are to be fully prepared. Even if you are acting for the plaintiff, you should be aware of the standard defence to this kind of action. If you are acting for the defendant, it goes without saying that you should be familiar with these cases.',nl]. / * The following rule accesses a linkage to an external data file from which cases are retrieved if the user wishes to examine them. Subsequent rules with similar syntax perform the same function. * / 160 rule-014: if best-defence and text(defence-cases) = [ D E F E N C E ] and str ingof(DEFENCE) = C A S E L I S T and external(selectcases, [CASELIST]) = Y S T R I N G then defences-given. / * T E X T ( D E F E N C E - C A S E S ) * / nocache(text(def ence-cases)). text(defence-cases) = [7]. / * L A T E S T - C A S E * / question(latest-case) = 'Would you like to read what is currently the leading case on the law of nervous shock in order to get a good overview of the area?'. legalvals(latest-case) = [yes, no]. nocache(overview). rule-015: if latest-case = no or latest-case is unknown and display([nl, 'Thank you for consulting Nervous Shock Advisor.', nl]) then overview. explanation(rule-015) = [nl, 'There are usually one or two cases from the highest courts in jurisdictions governed by comparable law which review and restate the law on a particular subject. The persuasive force of these leading cases makes it essential that a lawyer dealing with a similar sort of case be aware of them. Judges assume that lawyers have read these cases. To be unfamiliar with them can be a source of considerable embarrassment.',nl]. rule-016: if latest-case and full-text(leading-cases) = [ T E X T ] and stringof(TEXT) = C A S E L I S T and external(selectcases, [CASELIST]) = Y S T R I N G and display([nl, 'Thank you for consulting Nervous Shock Advisor.', nl]) then overview. / * F U L L - T E X T ( L E A D I N G - C A S E S ) * / nocache(full-text(leading-cases)). f ull-text(leading-cases) = [1]. / * - V I C T I M - T Y P E * / presupposition(victim-type) = right-tort. presupposition(victim-type) = valid-harm-claim. presupposition(victim-type) = not(cached(harm-type = both)). automaticmenu(victim-type). enumeratedanswers(victim-type). question(victim-type) = 'Who was the victim of the incident which caused the nervous shock: the plaintiff, a third-party or both? ( N O T E : If there was more than one third-party victim, your responses to any questions about a third-party should relate to the most serious case only).'. legalvals(victim-type) = [plaintiff, third-party, both]. rule-017: if victim-type = plaintiff then primary-victim. explanation(rule-017) = [nl, 'The elements which need to be established to support a case of this kind differ depending on the identity of the victim of the primary event which caused the nervous shock.',nl]. rule-018: if victim-type = third-party then secondary-victim. rule-019: if harm-type = both then multiple-victim. rule-020: if victim-type = both then multiple-victim. rule-021: if primary-victim or secondary-victim or multiple-victim or cached(false-news) then victim = ascertained. rule-022: if not(primary-victim) and not(secondary-victim) and not(multiple-victim) then victim = unascertained. / * O T H E R - H A R M presupposition(other-harm) = (not(cached(harm-type = risk))). presupposition(other-harm) = multiple-victim. presupposition(other-harm) = poor-plaintiff. presupposition(other-harm) = (not(harm-type = both and plaintiff-harmed)). question(other-harm) = 'Was the other person injured?'. legalvals(other-harm) = [yes, no]. / * O T H E R - R I S K presupposition(other-risk) = multiple-victim. presupposition(other-risk) = (not(cached(harm-type = harm))). presupposition(other-risk) = not(cached(other-harm)). presupposition(other-risk) = poor-plaintiff. presupposition(other-risk) = (not(harm-type = both and plaintiff-at-risk)). question(other-risk) = 'Did the other person narrowly escape being seriously injured?'. legalvals(other-risk) = [yes, no]. rule-023: if other-harm or other-risk then serious-incident. rule-024: if other-harm = no and other-risk = no and not(cached(poor-plaintiff)) then serious-incident = no. rule-025: if multiple-victim and plaintiff-harmed or plaintiff-at-risk and not(cached(other-harm)) and not(cached(other-risk)) then connection-to-victim. / * F A T A L I T Y presupposition(fatality) = valid-harm-claim. presupposition(fatality) = (victim = ascertained). presupposition(fatality) = secondary-victim or poor-plaintiff = no. presupposition(fatality) = not(primary-victim). presupposition(fatality) = not(cached(harm-type = risk)). presupposition(fatality) = not(cached(harm-type = neither)). presupposition(fatality) = not(cached(other-X = no)). presupposition(fatality) = not(cached(serious-incident = yes cf X)). question(fatality) = 'Was the other person killed?'. legalvals(fatality) = [yes, no]. / * SERIOUS-INJURIES * / presupposition(serious-injuries) = not(cached(harm-type = risk)). presupposition(serious-injuries) = not(cached(harm-type = neither)). presupposition(serious-injuries) = poor-plaintiff = no or secondary-victim. presupposition(serious-injuries) = not(primary-victim). presupposition(serious-injuries) = (victim = ascertained). presupposition(serious-injuries) = not(cached(other-X = no)). presupposition(serious-injuries) = not(cached(serious-incident = yes cf X)). question(serious-injuries) = 'Did the other person suffer visibly serious injuries?'. legalvals(serious-injuries) = [yes, no]. / * L U C K Y - E S C A P E * / presupposition(lucky-escape) = not(cached(harm-type = risk)). presupposition(lucky-escape) = not(cached(harm-type = neither)). presupposition(lucky-escape) = poor-plaintiff = no or secondary-victim. presupposition(lucky-escape) = not(primary-victim). presupposition(lucky-escape) = (victim = ascertained). presupposition(lucky-escape) = not(cached(other-X = no)). question(lucky-escape) = 'Was the other person slightly injured in a narrow escape from death or serious injury?'. legalvals(lucky-escape) = [yes, no]. / * N E A R M I S S - * / presupposition(nearmiss) = valid-harm-claim. presupposition(nearmiss) = (harm-type = risk) or (harm-type = both). presupposition(nearmiss) = (poor-plaintiff = no) or secondary-victim. presupposition(nearmiss) = not(cached(primary-victim)). presupposition(nearmiss) = (victim = ascertained). presupposition(nearmiss) = not(cached(lucky-escape)). presupposition(nearmiss) = not(cached(other-X = no)). question(nearmiss) = 'Did the other person come very close to being killed or seriously injured?'. legalvals(nearmiss) = [yes, no]. rule-026: if primary-victim and plaintiff-harmed then serious-incident. explanation(rule-026) = [nl, 'The basic dichotomy in nervous shock cases is between those where there was actual harm and those where there was simply a risk of harm. Where there was harm, the extent of the harm must be determined.',nl]. rule-027: if primary-victim and plaintiff-at-risk then serious-incident cf 80. explanation(rule-027) = [nl, 'The basic dichotomy in nervous shock cases is between those where there was actual harm and those where there was simply a risk of harm. In the case of a risk, the victim must have come very close to being injured.\nl]. rule-028: if multiple-victim and plaintiff-harmed or plaintiff-at-risk and not(cached(other-harm)) and not(cached(other-risk)) then serious-incident. rule-029: if harm-type = harm or harm-type = both and fatality or serious-injuries then has-incident. explanation(rule-029) = [nl, 'The basic dichotomy in nervous shock cases is between those where there was actual harm and those where there was simply a risk of harm. Where there was harm, the extent of the harm must be determined.\nl]. rule-030: if harm-type = harm or harm-type = both and lucky-escape then has-incident cf 80. 165 explanation(rule-030) = [nl, 'The basic dichotomy in nervous shock cases is between those where there was actual harm and those where there was simply a risk of harm. Where there was harm, the extent of the harm must be determined.',nl]. rule-031: if harm-type = risk or harm-type = both and not(primary-victim) and nearmiss then has-incident cf 50. explanation(rule-031) = [nl, 'The basic dichotomy in nervous shock cases is between those where there was actual harm and those where there was simply a risk of harm. In the case of a risk, the victim must have come very close to being seriously injured.',nl]. rule-032: i f has-incident then serious-incident. rule-033: if not(cached(primary-victim)) and not(cached(poor-plaintiff)) and not(cached(serious-incident = yes cf X)) and fatality is unknown and serious-injuries is unknown and lucky-escape is unknown and nearmiss is unknown then serious-incident = no. rule-034: i f not(cached(primary-victim)) and not(cached(serious-incident = yes cf X)) and not(has-incident) then serious-incident = no. rule-035: if primary-victim then connection-to-victim. rule-036: if primary-victim then exposure-to-incident. / * P L A I N T I F F - I N V O L V E D * / rule-037: if multiple-victim then plaintiff-involved. rule-038: if primary-victim then plaintiff-involved. 166 rule-039: i f plaintiff-involved then exposure-to-incident. / * P L A I N T I F F - H A R M E D * / presupposition(plaintiff-harmed) = plaintiff-involved. presupposition(plaintiff-harmed) = not(cached(harm-type = risk)). question(plaintiff-harmed) = 'Was your client injured, even slightly, in circumstances that could well have resulted in far more serious injury?', legal vals(plaintiff-harmed) = [yes, no]. / * P L A I N T I F F - A T - R I S K * / presupposition(plaintiff -at-risk) = plaintiff-involved. presupposition(plaintiff-at-risk)=not(cached(plaintiff-harmed)). presupposition(plaintiff-at-risk)= not(cached(harm-type = harm)). question(plaintiff-at-risk) = 'Did your client narrowly escape being seriously injured?'. legalvals(plaintiff-at-risk) = [yes, no]. rule-040: if plaintiff-harmed = no and plaintiff-at-risk = no and not(cached(multiple-victim)) then serious-incident = no. rule-041: i f plaintiff-harmed or plaintiff-at-risk and not(cached(primary-victim)) then poor-plaintiff. rule-042: if plaintiff-harmed = no and plaintiff-at-risk = no and not(cached(primary-victim)) then poor-plaintiff = no. / * A G G R A V A T E D - I N C I D E N T — - * / rule-043: if multiple-victim and plaintiff-harmed and other-harm then aggravated-event. explanation(rule-043) = [nl, Tn cases where multiple victims were involved in the incident which caused the nervous shock, there need only have been relatively minor injury.',nl]. rule-044: i f multiple-victim and plaintiff-harmed and other-risk then aggravated-event. explanation(rule-044) = [nl, 'In cases where multiple victims were involved in the incident which caused the nervous shock, there need only have been relatively minor injury or a risk of injury.',nl]. rule-045: if multiple-victim and plaintiff-at-risk and other-harm then aggravated-event. explanation(rule-045) = [nl, 'In cases where multiple victims were involved in the incident which caused the nervous shock, there need only have been relatively minor injury or a risk of injury.',nl]. rule-046: if multiple-victim and plaintiff-at-risk and other-risk then aggravated-event. explanation(rule-046) = [nl, 'In cases where multiple victims were at risk in the incident which caused the nervous shock, there need only have been a risk of injury.',nl]. rule-047: if aggravated-event then aggravated-incident. rule-048: if not(aggravated-event) then aggravated-incident = no. / * C E R T A I N T Y - A U G M E N T E D The following rules increase the certainty of recovery in aggravated incidents once it has been established that all the requisite elements of a cause of action are present and that the increase would raise the pre-existing factor.*/ 168 rule-049: if aggravated-incident and close-relationship and not(cached(associated)) and cached(judgment = an-action-lies cf X) and X < 100 and do(reset judgment = an-action-lies cf X) and do(set judgment = an-action-lies cf 100 because aggravated) then certainty-raised. rule-050: if aggravated-incident and not(cached(close-relationship)) and associated and cached(judgment = an-action-lies cf X) and X < 100 and do(reset judgment = an-action-lies cf X) and do(set judgment = an-action-lies cf 100 because aggravated) then certainty-raised. rule-051: i f aggravated-incident and not(cached(close-relationship)) and not(cached(associated)) and harm-last-resort and cached(judgment = an-action-lies cf X) and X < 100 and do(reset judgment = an-action-lies cf X) and do(set judgment = an-action-lies cf 100) then certainty-raised. rule-052: if aggravated-incident and not(cached(close-relationship)) and not(cached(associated)) and not(cached(harm-last-resort)) and risk-last-resort and cached(judgment = an-action-lies cf X) and X < 100 and do(reset judgment = an-action-lies cf X) and do(set judgment = an-action-lies cf 100) then certainty-raised. rule-053: if aggravated-incident and close-relationship and not(cached(associated)) and cached(judgment = an-action-lies cf X) and X = 100 then certainty-raised = no. rule-054: if aggravated-incident and not(cached(close-relationship)) and associated and cached(judgment = an-action-lies cf X) and X = 100 then certainty-raised = no. rule-055: if aggravated-incident and not(cached(close-relationship)) and not(cached(associated)) and harm-last-resort and cached(judgment = an-action-lies cf X) and X = 100 then certainty-raised = no. rule-056: if aggravated-incident and not(cached(close-relationship)) and not(cached(associated)) and not(cached(harm-last-resort)) and risk-last-resort and cached(judgment = an-action-lies cf X) and X = 100 then certainty-raised = no. rule-057: if certainty-raised then certainty-increase. rule-058: if certainty-raised = no then certainty-increase = no. rule-059: i f certainty-increase or certainty-increase = no then certainty-augmented. / * C O N N E C T I O N - T O - V I C T I M presupposition(connection-to-victim) = valid-harm-claim. presupposition(connection-to-victim) = secondary-victim or multiple-victim. presupposition(connection-to-victim) = serious-incident. / * - R E L A T E D presupposition(related) = valid-harm-claim. question(related) = 'Is your client closely related to the other person?'. legalvals(related) = [yes, no]. /* T Y P E - R E L A T I O N S H I P •*/ presupposition(type-relationship) = valid-harm-claim. automaticmenu(type-relationship). enumeratedanswers(type-relationship). question(type-relationship) = 'Is the other person a member of your client's immediate family; namely, the child, spouse, parent or sibling of your client? ( N O T E : Regard a step-relationship as equivalent to a natural one if the person was treated like a member of the natural family and a grandparent as equivalent to a parent if the grandparent had been filling the role of a parent).'. legalvals(type-relationship) = [child, spouse, parent, sibling, neither]. / * C O H A B I T A T I O N - * / presupposition(cohabitation) = valid-harm-claim. question(cohabitation) = 'Did your client and the other person live together in an intimate personal relationship analogous to a marriage?'. legalvals(cohabitation) = presupposition(engaged) = valid-harm-claim. question(engaged) = 'Were your client and the other person engaged to be married?'. legalvals(engaged) = [yes, no]. / * R E S C U E R * / presupposition(rescuer) = valid-harm-claim. presupposition(rescuer) = bodily-harm. question(rescuer) = 'Did your client attempt to assist the other person after the incident?'. legalvals(rescuer) = [yes, no]. /* E N G A G E D */ [yes, no]. /* ' H A R M - L A S T - R E S O R T presupposition(harm-last-resort) = multiple-victim. presupposition(harm-last-resort) = plaintiff-harmed. presupposition(harm-last-resort) = not(cached(close-relationship)). presupposition(harm-last-resort) = not(cached(associated)). question(harm-last-resort) = 'Did the plaintiff suffer the nervous shock all or in part because of injury to himself?'. legalvals(harm-last-resort) = [yes, no]. / * R I S K - L A S T - R E S O R T * presupposition(risk-last-resort) = multiple-victim. presupposition(risk-last-resort) = plaintiff-at-risk. presupposition(risk-last-resort) = not(cached(close-relationship)). presupposition(risk-last-resort) = not(cached(associated)). question(risk-last-resort) = 'Did the plaintiff suffer the nervous shock all or in part because of a fear of injury to himself?'. legalvals(risk-last-resort) = [yes, no]. rule-060: if harm-type = harm and fatality or serious-injuries then bodily-harm. rule-061: if related and type-relationship = child or type-relationship = spouse or type-relationship = parent or type-relationship = sibling then close-relationship. explanation(rule-061) = [nl, 'The closest relationship recognized by the law is that between members of the so-called nuclear family.',nl]. rule-062: if related and type-relationship = neither then close-relationship = no. rule-063: i f related and close-relationship then connection-to-victim. explanation(rule-063) = [nl, 'Some sort of proximate relationship must be established between the person who suffered the nervous shock and the victim of the event which triggered it. Generally speaking, a plaintiff cannot recover for nervous shock arising out of a traumatic incident which involved a stranger. A close family tie is the strongest legal relationship.',nl]. rule-064: if cohabitation or engaged then associated. explanation(rule-064) = [nl, 'Some sort of proximate relationship must be established between the person who suffered the nervous shock and the victim of the event which triggered it. Generally speaking, a plaintiff cannot recover for nervous shock arising out of a traumatic event which involved a stranger.',nl]. rule-065: if cohabitation = no and engaged = no then associated = no. rule-066: if related and close-relationship = no or close-relationship is unknown and associated and cached(cohabitation) then connection-to-victim cf 70. rule-067: if related and close-relationship = no or close-relationship is unknown and associated and cached(engaged) then connection-to-victim cf 60. rule-068: if related = no or related is unknown and associated and cached(cohabitation) then connection-to-victim cf 70. rule-069: if related = no or related is unknown and associated and cached(engaged) then connection-to-victim cf 60. rule-070: if related = no or close-relationship = no and associated = no and fatality and rescuer then connection-to-victim cf 70. explanation(rule-070) = [nl, 'While there must generally be a legally proximate relationship between the person who suffers the nervous shock and the victim of the event which triggers it, the law makes a policy exception in the case of rescuers. This policy is obviously designed to encourage good Samaritans.\nl]. rule-071: if related = no or close-relationship = no and associated = no and fatality = no and serious-injuries and rescuer then connection-to-victim cf 50. explanation(rule-071) = [nl, 'While there must generally be a legally proximate relationship between the person who suffers the nervous shock and the victim of the event which triggers it, the law makes a policy exception in the case of rescuers. This policy is obviously designed to encourage good Samaritans.',nl]. rule-072: i f related is unknown and associated is unknown and fatality and rescuer then connection-to-victim cf 70. rule-073: if related is unknown and associated is unknown and fatality = no and serious-injuries and rescuer then connection-to-victim cf 50. 174 rule-074: i f harm-last-resort or risk-last-resort then connection-to-victim. explanation(rule-074) = [nl, 'Even in the absence of some sort of proximate relationship between a third-party victim of nervous shock and the plaintiff, recovery is usually allowed where the plaintiff was injured or nearly injured.',nl]. rule-075: i f harm-last-resort = no and risk-last-resort = no then connection-to-victim = no. rule-076: if related = no or close-relationship = no and associated = no and not(rescuer) then connection-to-victim = no. rule-077: if related is unknown and associated is unknown and not(rescuer) or rescuer is unknown then connection-to-victim = no. / * E X P O S U R E - T O - I N C I D E N T * / presupposition(exposure-to-incident) = valid-harm-claim. presupposition(exposure-to-incident) = secondary-victim. presupposition(exposure-to-incident) = connection-to-victim. / * E Y E W I T N E S S * / question(eyewitness) = 'Did your client see the incident happen?'. legalvals(eyewitness) = [yes, no]. / * - I M M I N E N T - A C C I D E N T * / question(imminent-accident) = 'Did your client see the accident about to happen, but not the accident itself or-its aftermath?'. legalvals(imminent-accident) = [yes, no]. 175 / * N E A R B Y S T A N D E R - * / question(nearbystander) = 'Did your client arrive at the scene right after the accident and see the condition of the other person?'. legalvals(nearbystander) = [yes, no]. / * T R A C E S * / presupposition(traces) = imminent-accident = no. question(traces) = 'Did your client arrive at the accident scene not long after the other person had been removed and, knowing who that person must have been, see obvious signs of a serious accident having taken place?'. legalvals(traces) = [yes, no]. / * H O S P I T A L - V I S I T * / presupposition(hospital-visit) = traces = no. question(hospital-visit) = 'Did your client see the other person at hospital before the injuries were treated or the dead body cleaned up?'. legalvals(hospital-visit) = [yes, no]. / * D E F O R M I T Y * / presupposition(deformity) = (hospital-visit = no). question(deformity) = 'Was the other person so permanently disfigured, deformed or otherwise altered by the harm they suffered that their appearance would be shocking to the average person?'. legalvals(deformity) = [yes, no]. rule-078: i f related = no or close-relationship = no and associated = no and rescuer then saw-victim. rule-079: if related is unknown and associated is unknown and rescuer then saw-victim. rule-080: if eyewitness then saw-victim. explanation(rule-080) = [nl, 'In order to make out a successful case of nervous shock, the plaintiff must have had some visual exposure to the incident itself or its after-effects.',nl]. rule-081: if bodily-harm and nearbystander then saw-victim. explanation(rule-081) = [nl, 'In order to make out a successful case of nervous shock, the plaintiff must have had some visual exposure to the incident itself or its after-math.',nl]. rule-082: i f bodily-harm and imminent-accident then saw-victim cf 50. explanation(rule-082) = [nl, 'In order to make out a successful case of nervous shock, the plaintiff must have had some visual exposure to the incident itself or its after-effects. At the very least, this would involve seeing the accident about to happen, although not the actual accident, provided a serious accident did in fact occur.',nl]. rule-083: i f bodily-harm and traces then saw-victim cf 70. explanation(rule-083) = [nl, 'In order to make out a successful case of nervous shock, the plaintiff must have had some visual exposure to incident itself. Seeing the gruesome after effects of a serious accident meets this criterion.',nl]. rule-084: if bodily-harm and hospital-visit then saw-victim. explanation(rule-084) = [nl, 'In order to make out a successful case of nervous shock, the plaintiff must have had some visual exposure to the incident itself. The shock of seeing an untreated victim satisfies this requirement.',nl]. rule-085: if bodily-harm and deformity then saw-victim cf 60. explanation(rule-085) = [nl, Tn order to make out a successful case of nervous shock, the plaintiff must have had some visual exposure to the incident itself. In these unusual circumstances, although there was a delay between the incident and the exposure, the horror of the visual impact satisfies this requirement.\nl]. rule-086: if eyewitness is unknown and nearbystander is unknown and imminent-accident is unknown and traces is unknown and hospital-visit is unknown and deformity is unknown then saw-victim = no. rule-087: if saw-victim then exposure-to-incident. rule-088: if not(saw-victim) then exposure-to-incident = no. / * I N C A P A C I T Y * / presupposition(incapacity) = right-tort. presupposition(incapacity) = valid-harm-claim or valid-news-claim. presupposition(incapacity) = serious-incident. presupposition(incapacity) = connection-to-victim. presupposition(incapacity) = exposure-to-incident. question(incapacity) = 'Did the nervous shock cause an expense, a loss of income or an incapacity which was sufficiently serious to support a claim for damages?'. legalvals(incapacity) = [yes, no]. / * A F T E R - E F F E C T S * / presupposition(after-effects) = incapacity. automaticmenu(after-effects). enumeratedanswers(after-effects). multivalued(af ter-ef f ects). question(after-effects) = 'Into which of the following general categories did the symptoms of nervous shock fall: physical illness; mental disorder; sleep disturbance? Separate each option by a comma if more than one is applicable'. legalvals(after-effects) = [physical-illness, mental-disorder, sleep-disturbance]. rule-089: if incapacity then affected-by-incident. explanation(rule-089) = [nl, 'Every successful lawsuit requires proof of legally claimable damages. In this respect, nervous shock is no different from any other case.',nl]. rule-090: if incapacity = no then affected-by-incident = no. rule-091: if after-effects is unknown then affected-by-incident = no. rule-092: if affected-by-incident = no then disability = no. rule-093: if incapacity and after-effects is known then disability. explanation(rule-093) = [nl, 'In order to recover damages in a case of nervous shock, the plaintiff must have displayed symptoms which fall into one of the categories which the courts will recognize as legitimate manifestations of nervous shock.',nl]. rule-094: if incapacity is unknown or after-effects is unknown then disability = no. / * S Y M P T O M * / multivalued(symptom). rule-095: if disability and after-effects = physical-illness then symptom = physical. rule-096: if disability and after-effects = mental-disorder then symptom = mental. rule-097: i f disability and after-effects = sleep-disturbance then symptom = sleep. / * S Y M P T O M S - M E S S A G E The following set of rules ensures that the exact symptoms or combinations thereof entered by the user are referred to when the user is given the option of perusing the cases. * / presupposition(symptoms-message) = disability. rule-098: if disability and listof(X,(symptom) = X) = [ONE] and display([nl, 'You said that your client had ' , O N E , ' problems.',nl,nl]) then symptoms-message. rule-099: if disability and listof(X,(symptom) = X) = [ONE,TWO] and display([nl, 'You said that your client had \ O N E , ' and ' ,TWO,' problems.',nl,nl]) then symptoms-message. rule-100: if disability and listof(X,(symptom) = X) = [ O N E , T W O , T H R E E ] and display([nl, 'You said that your client had ' ,ONE, ' , \ T W O , ' and \ T H R E E , ' problems.', nl,nl]) then symptoms-message. / * VIEW-CASES * / presupposition(view-cases) = symptoms-message. question(view-cases) = 'Would you like to peruse the cases which relate to these symptoms?'. legalvals(view-cases) = [yes, no]. / * G E T - X - C A S E S * / rule-101: if symptoms-message and view-cases and casesfor(physical) = [PHYSICAL] and stringof(PHYSICAL) = C A S E L I S T and external(selectcases, [CASELIST]) = Y S T R I N G then get-physical-cases. explanation(rule-lOl) = [nl, 'I can provide you with cases in which the courts allowed recovery for the general type of symptoms which your client suffered. Furthermore, within the overall category, specific symptoms are clearly identified. If you choose to read the case digests, you will f ind verbatim quotes from the judgments. In this way, you can be sure about how closely the symptoms of the successful litigant resembled those of your client.',nl]. rule-102: if symptoms-message and view-cases and casesfor(mental) = [ M E N T A L ] and str ingof(MENTAL) = C A S E L I S T and external(selectcases, [CASELIST]) = Y S T R I N G then get-mental-cases. explanation(rule-102) = [nl, 'I can provide you with cases in which the courts allowed recovery for the general type of symptoms which your client suffered. Furthermore, within the overall category, specific symptoms are clearly identified. If you choose to read the case digests, you will f ind verbatim quotes from the judgments. In this way, you can be sure about how closely the symptoms of the successful litigant resembled those of your client.',nl]. rule-103: if symptoms-message and view-cases and casesfor(sleep) = [SLEEP] and stringof(SLEEP) = C A S E L I S T and external(selectcases, [CASELIST]) = Y S T R I N G then get-sleep-cases. explanation(rule-103) = [nl, 'I can provide you with cases in which the courts allowed recovery for the general type of symptoms which your client suffered. Furthermore, within the overall category, specific symptoms are clearly identified. If you choose to read the case digests, you will f ind verbatim quotes from the judgments. In this way, you can be sure about how closely the symptoms of the successful litigant resembled those of your client.\nl]. rule-104: if not(cached(get-physical-cases)) and not(cached(get-mental-cases)) and not(cached(get-sleep-cases)) or view-cases = no then get-no-cases. /* S Y M P T O M S - F O U N D multivalued(symptoms-found). rule-105: if disability and symptom = physical and casesfor(physical) = P H Y S I C A L and get-physical-cases and do(reset casesfor(physical)) then physical-search. rule-106: i f disability and symptom = mental and casesfor(mental) = M E N T A L and get-mental-cases and do(reset casesfor(mental)) then mental-search. rule-107: if disability and symptom = sleep and casesfor(sleep) = S L E E P and get-sleep-cases and do(reset casesfor(sleep)) then sleep-search. rule-108: if disability and get-no-cases then no-search. rule-109: if physical-search is sought and mental-search is sought and sleep-search is sought and no-search is sought then symptoms-found. rule-110: if disability = no then symptoms-found. / * C A S E S F O R ( S Y M P T O M ) multivalued(casesfor(X)). nocache(casesfor(X)). casesfor(physical) = [2]. casesfor(mental) = [3]. casesfor(sleep) = [4]. /* J U D G M E N T rule - I l l : if right-tort and valid-harm-claim and victim = ascertained and aggravated-event is sought and serious-incident and connection-to-victim and exposure-to-incident and disability then judgment = an-action-lies. rule-112: if right-tort and valid-news-claim and false-news and serious-incident and disability then judgment = an-action-lies. rule-113: i f wrong-tort or (invalid-harm-claim and not(valid-news-claim)) or victim = unascertained or serious-incident = no or connection-to-victim = no or exposure-to-incident = no or disability = no then judgment = no-cause-of-action. rule-114: i f wrong-tort or (not(valid-harm-claim) and invalid-news-claim) or false-news = no or disability = no then judgment = no-cause-of-action. / * G R E E T I N G - D O N E nocache(greeting-done). rule-115: if display([ ' * .* .* .* Welcome to Nervous Shock Advisor * .* .* .* ' s nl, nl, T will tell you whether or not your client has a cause of action in nervous shock. Simply type your answers to my questions on the keyboard located below the screen. If you wish to know why a particular question is being asked, feel free to type "why" in response to the question. You may also type "unknown" if you are unable to answer a particular question. If so, I shall canvass whatever other options are available to me.', nl, nl]) then greeting-done. / * R E S U L T S - S H O W N The following variable rule calls up appropriate text messages which describe the outcome of a particular consultation. As the consultation progresses, the user is informed when his case fails for lack of a material element. If his case succeeds, he is advised accordingly. * / nocache(results-shown). rule-116: i f event = E V E N T and event-description(EVENT) = D E S C R I P T I O N and display(DESCRIPTION) then results-shown. / * E V E N T * / rule-117: if wrong-tort then event = tort-wrong. rule-118: if invalid-harm-claim and invalid-news-claim then event = no-claim. rule-119: if false-news = no then event = not-spread. rule-120: i f primary-victim and plaintiff-harmed = no and not(cached(plaintiff-at-risk)) and not(cached(f alse-news)) then event = no-incident. rule-121: if primary-victim and plaintiff-at-risk = no and not(cached(plaintiff-harmed)) and not(cached(f alse-news)) then event = no-incident. rule-122: i f victim = unascertained and not(cached(f alse-news)) then event = no-victim. rule-123: if serious-incident = no and not(cached(f alse-news)) then event = no-incident. rule-124: if connection-to-victim = no and not(cached(f alse-news)) then event = no-connection. 184 rule-125: i f exposure-to-incident = no and not(cached(false-news)) then event = no-exposure. rule-126: if disability = no then event = no-disability. rule-127: if right-tort and valid-harm-claim and victim = ascertained and serious-incident and connection-to-victim and exposure-to-incident and disability then event = recovery. rule-128: i f right-tort and valid-news-claim and false-news and disability then event = recovery. / * D E C I S I O N - G I V E N The following set of rules display messages which tell the user either why a case has failed or that there is good cause of action. * / nocache(decision-given). rule-129: if wrong-tort and display([nl, ' J U D G M E N T : an action may well lie for causing nervous shock.',nl, ' R E A S O N : in general, nervous shock resulting from an intentional wrong is not considered to be too remote.',nl]) then decision-given. rule-130: if invalid-harm-claim and not(cached(valid-news-claim)) and display([nl, ' J U D G M E N T : no cause of action.',nl, ' R E A S O N : event not a legally recognized cause of nervous shock.',nl]) then decision-given. rule-131: if invalid-news-claim and not(cached(valid-harm-claim)) and display([nl, ' J U D G M E N T : no cause of action.',nl, ' R E A S O N : event not a legally recognized cause of nervous shock.',nl]) then decision-given. 185 rule-132: if false-news = no and display([nl, ' J U D G M E N T : no cause of action.',nl, ' R E A S O N : defendant not responsible for publishing the false report.\nl]) then decision-given. rule-133: if victim = unascertained and not(cached(false-report)) and display([nl, ' J U D G M E N T : no cause of action.',nl, ' R E A S O N : victim"s identity unknown.',nl]) then decision-given. rule-134: if primary-victim and not(cached(false-report)) and plaintiff-harmed = no and not(cached(plaintiff-at-risk)) and display([nl, ' J U D G M E N T : no cause of action.',nl, ' R E A S O N : incident not serious enough.\nl]) then decision-given. rule-135: if primary-victim and not(cached(false-report)) and plaintiff-at-risk = no and not(cached(plaintiff-harmed)) and display([nl, ' J U D G M E N T : no cause of action.',nl, ' R E A S O N : incident not serious enough.',nl]) then decision-given. rule-136: i f serious-incident = no and not(cached(false-report)) and display([nl, ' J U D G M E N T : no cause of action.\nl, ' R E A S O N : incident not serious enough.\nl]) then decision-given. rule-137: if connection-to-victim = no and not(cached(false-report)) and display([nl, ' J U D G M E N T : no cause of action.',nl, ' R E A S O N : no connection between plaintiff and victim.',nl]) then decision-given. rule-138: if exposure-to-incident = no and not(cached(false-report)) and display([nl, ' J U D G M E N T : no cause of action.',nl, ' R E A S O N : plaintiff not exposed incident.\nl]) then decision-given. rule-139: if disability = no and display([nl, ' J U D G M E N T : no cause of action.',nl, ' R E A S O N : damage not proved.',nl]) then decision-given. / * X - F A C T O R The following set of rules builds a profile of the constituent elements of the successful cause of action which is tailored to fit the particular fact pattern supplied by the user. Certainty factors are attached to each element so that the user can see the weak points in his case at a glance. Generally speaking, the overall chance of success will be no stronger than the weakest link in the chain. However, in cases in which the plaintiff was also a victim of the primary event which caused the nervous shock, the total chance of recovery i