Computer-Mediated Legal Argument: Towards new Opportunities
in Education

Arno R. Lodder
Computer/Law Institute
Vrije Universiteit Amsterdam
lodder@rechten.vu.nl

and

Bart Verheij
  Section Law and Informatics
Department of Metajuridica
Universiteit Maastricht
bart.verheij@metajur.unimaas.nl

(The second author gladly acknowledges the financial support by the Dutch National Programme Information Technology and Law  ( ITeR ) for the research reported in this paper (project number 01437112 )).

Versions of this paper have been presented at the 13th Lodder and Verheij 1998 ) and at the 2nd French-American Conference on Artificial Intelligence and Law ( Verheij and Lodder 1998 ).
 

Abstract

Argumentation is a key activity of lawyers. Therefore in law school teaching argumentation is essential. Information technology can provide useful support in argumentation courses. New opportunities come from a recent topic of research in the field AI & Law, viz. computer-mediated legal argument (e.g., Gordon , Lodder , Loui ). In this paper, we suggest how computer-mediated legal argument can provide opportunities for teaching legal argument. The main part of the paper is devoted to a description of two experimental systems for computer-mediated legal argument, viz. Lodder's DiaLaw and Verheij's Argue!.

The starting point for the research on computer-mediated legal argument is that a computer system can support lawyers by mediating the process in which they draft and generate arguments: the system can administer and supervise the argument process by keeping track of the reasons adduced and the conclusions drawn, and by checking whether the users of the system obey the pertaining rules of argument, e.g., those related to the division of burden of proof.

Computer-mediated legal argument poses a new problem: how should an argument be presented to the users of the mediating system? Especially with regards to recently developed logical tools (e.g., Hage , Prakken , Sartor , Verheij ), there is little experience with argument presentation. There is a natural division of approaches to argument presentation in two classes: the verbal and the graphical approaches. In the verbal approach, the argument is mainly presented in a verbal style, for instance in the form of a text or a written-out dialog. In the graphical approach, the argument is mainly presented in a graphical style, for instance in the form of a tree of sentences.

In the paper, we reconstruct elements of a Dutch Supreme Court case on tort law (March 20, 1992) and its sequel at the Court of Justice of The Hague (September 15, 1994) in two prototypical systems for the mediation of legal argument, viz. DiaLaw by Lodder and Argue! by Verheij . The first system takes the verbal approach; the second system takes the graphical approach.

The two approaches to argument presentation are discussed against the background of teaching legal argument. We speculate on the possible uses of computer-mediated legal argument in teaching argumentation skills. We expect that neither of the two approaches can be fully satisfactory for teaching legal argument if it is taken to its extreme, and recommend that a hybrid combination of verbal and graphical elements is striven for.

This is a Refereed Article published on 13 April 1999.

Citation : Lodder A and Verheij B, 'Computer-Mediated Legal Argument: Towards new Opportunities in Education', 1999 (2) The Journal of Information, Law and Technology (JILT). <http://elj.warwick.ac.uk/jilt/99-2/lodder.html>. New citation as at 1/1/04: < http://www2.warwick.ac.uk/fac/soc/law/elj/jilt/1999_2/lodder/>

1. Introduction

Argumentation is a key activity of lawyers. Therefore in law school teaching argumentation is essential. Information technology can provide useful support in argumentation courses. For instance, recently Aleven and Ashley ( 1997 ) have shown that instructing case-based argumentation skills by their CATO system led to improvement in students' basic argumentation skills that was comparable to the improvement realised by an experienced legal writing instructor.

One could also think of systems teaching classical logic, such as first-order predicate logic. Several systems are available. An example is the acclaimed Tarski's World by Barwise and Etchemendy, that teaches the meaning of the language of first-order predicate logic by depicting a block's world on screen, in which sentences can be interpreted.

Notwithstanding their availability, systems meant for teaching classical logic are hardly used in law school, since only the basic elements of classical logic are considered essential for lawyers. Certainly, classical logic can be used to analyse the argument structure of legal texts and discourse. However, the resulting analysis is (by the very nature of classical logic) rather limited. Probably, the main reason for teaching elements of classical logic in law school (if taught at all) is that it serves the purpose of training general analytical skills, which is useful for anyone.

Recently, much research has been devoted to legal logic . In this research program, attempts are made to include legal argument forms into logic. Some of the topics addressed are exceptions, the weighing of conflicting reasons, rule applicability and the division of burden of proof. Arguments using these have long been adopted by lawyers, but in an intuitive way and without a systematic basis. Legal logic attempts to provide this basis. Verheij, Hage and Lodder ( 1997 ) provide an overview of the recent research on legal logic with examples from Dutch tort law.

Supposedly, legal logic has more to offer that is readily useful for law school than classical logic, since the argument forms that can be analysed by legal logic, occur regularly in actual legal argument. The question arises how the findings of legal logic could be taught in an effective way.

New opportunities come from a recent topic of research in the field AI & Law, viz. computer-mediated legal argument (e.g., Loui ( 1992 ), Gordon ( 1995 ), Lodder ( 1998 ), Verheij ( 1998a , 1998b )). The starting point for the research on computer-mediated legal argument is that a computer system can support lawyers by mediating the process in which they draft and generate arguments: the system can administer and supervise the argument process by keeping track of the reasons adduced and the conclusions drawn, and by checking whether the users of the system obey the pertaining rules of argument, e.g., those related to the division of burden of proof.

A difficulty for such systems is the presentation of arguments. There is a natural distinction between two approaches to argument presentation, viz. the verbal and the graphical approach. The first presents arguments in a verbal way. The chosen language can be natural language (or a fragment of natural language) or a formal language, for instance, that of predicate logic. Gordon's Pleadings Game ( 1995 ) is an example of the verbal approach, using a formal language. The second approach presents arguments in a graphical way. The structure of arguments is somehow reflected in its graphical layout. An example is Loui et al.'s 1997 ), where sentences are shown in boxes. If one box is inside another box, the sentence in the first box expresses a reason for the conclusion in the second box.

As far as legal systems for teaching argumentation-like skills are concerned, besides the already mentioned CATO-system only few of them are really used in practice or have left the stage of a prototype. Examples of legal systems for training students are OBLIGATIO by Fernhout et al. ( 1987 ) (that was used in the early nineties to train law students in Maastricht), the LITES system by Span ( 1994 ), the STATUTOR system ( Routen ( 1991 ), Centinia et al. ( 1995 )), the PROSA-system by Muntjewerff & Groothuismink ( 1998 ), the system for international tax law by Lodder ( 1996 ), and, recently, Bench-Capon et al. ( 1998 ).

In this paper, we describe two mediation systems that have been developed at the Universiteit Maastricht that can serve as the starting point for future systems for teaching students legal argumentation. We will be brief about the argumentation theory behind the systems. The interested reader is referred to the work of Hage ( 1996 , 1997 ), Lodder ( 1998 ), and Verheij ( 1996 , 1998a ).

The first system, called DiaLaw, takes the verbal approach. It is a Prolog implementation of the dialog game for legal justification DiaLaw ( Lodder & Herczog,  1995 ); Lodder, 1998 ). The second system, called Argue!, takes the graphical approach. It is a Delphi implementation of the procedural model of argumentation with arguments and counterarguments CumulA ( Verheij, 1996 ). Both systems are prototypes for the mediation of legal argument. In our discussion of the two systems, we suggest opportunities for teaching legal argument.

In the following, we discuss an interesting case of Dutch tort law (section 2). The case is used to illustrate the two prototypical systems for the mediation of legal argument. First the system that takes the verbal approach is described (section 3), then the system that takes the graphical approach (section 4).

2. A Dutch Case of Tort

In this section, we give a brief overview of the regulation of torts in the Netherlands, and discuss an interesting case: the 'bussluis' case.

2.1 Dutch Tort Law

In civil law systems, the liability for damages is amongst others related to the notion of a tort, or wrongful act. For instance, if someone clumsily parks his car, thereby damaging another already parked car, he commits a tort against the owner of that car and has to compensate for the damages. In Dutch civil law, the essence of the relation between the liability for damages and a tort is regulated in the articles 6:162 and 6:101 of the civil code.

Art. 6:162.1 of the civil code states that a person who commits an unlawful act toward another which can be imputed to him, must repair the damage which the other person suffers as a consequence thereof. Art. 6:162.2 contains the three forms of unlawful acts in Dutch law: a violation of a right, an act or omission violating a statutory duty, and an act or omission violating a rule of unwritten law pertaining to proper social conduct.

Art. 6:101.1 of the civil code states (in synopsis) that, if the damages are partially caused by a circumstance that can be imputed to the party that suffered the damages, the duty to repair the damages is diminished, relative to the amount of imputability of the party that suffered damages and the party that has the duty to repair the damages.

As a result of the Dutch regulations, there are two steps that determine the duty to compensate damages. First, the general duty to compensate damages is established on the basis of art. 6:162 of the civil code. Second, the relative amount of imputability determines the portion of the damages that have to be compensated for.

2.2 The 'Bussluis' Case

In the eighties a new phenomenon to regulate traffic in inner cities was introduced. The so-called 'bussluis', which is nowadays well-known, was shortly after its introduction rather unknown. The 'bussluis' is an obstacle in the road that can only be passed without problems by buses. Ordinary cars get stuck in it, or if they nevertheless pass through the 'bussluis' the car is damaged. The latter happened to a cab-driver in the mid eighties who ruined his car by driving into a 'bussluis'. The cab-driver could be blamed, because he was driving a one way street in the wrong direction, and had ignored warning signals.

The Dutch Supreme Court (March 20, 1992) decided that the local authorities were guilty of tort and therefore had to compensate for the damages. Since the Supreme Court in the Netherlands only decides legal matters and not factual matters, the case was referred to a Court of Justice in order to determine the amount of damages that had to be compensated. Rather surprisingly, the Court of Justice of the Hague (September 15, 1994) determined that the amount damages that had to be paid was nil. In their opinion, the damages were caused by the cab-driver's own fault for the full 100%. So although the local authorities were guilty of tort, they did not have to pay for the damages after all.

The arguments the Court used to justify that the amount of damages that had to be compensated was nil, were the following:

In summary, the unexpected character of the 'bussluis' case was that, although the Municipality had committed a tort against the cab-driver and therefore had the general duty to repair the damages (on the basis of art. 6:162 of the civil code), the thereafter determined portion of the damages to be compensated for (on the basis of art. 6:101) was nil because the damages were fully imputed to the cab-driver.

In the following, the case of the cab-driver is used to illustrate both the verbal and the graphical approach to mediating legal argument.

3. DiaLaw: A Verbal Approach to Legal Argument Mediation

The verbal approach is demonstrated by an example in the dialog game for legal justification DiaLaw ( Lodder & Herczog, 1995 ; Lodder, 1998 ). After introducing DiaLaw, a sample dialog based on the 'bussluis' case illustrates the opportunities of the verbal approach for education.

3.1 Introducing DiaLaw

In the AI & Law-community a wide range of dialogical models have been developed (e.g., Gordon, 1993 ; Loui et al., 1993 ; Hage, Leenes & Lodder, 1994 ; Lodder & Herczog, 1995 ; Farley & Freeman, 1995 ; Bench-Capon, 1995 ; Prakken & Sartor, 1996 ; Kowalski & Toni, 1996 ; Nitta & Shibasaki, 1997 ). These are all formal or software models. The tradition of informal models is longer. Perelman's rhetorical theory ( Perelman & Olbrechts-Tyteca 1958 ), and Habermas' ( 1973 ) consensus theory of truth have influenced most dialogical research. Today's best-known defenders of dialogical models in law are Aarnio, Alexy and Peczenik, who in the beginning of the eighties ( Aarnio, Alexy & Peczenik, 1981 ) integrated their theories that were developed independently in the seventies. DiaLaw has been developed over the last five years.

DiaLaw is a two-person dialog game, in which both players make moves. The goal of the game is that the proponent convinces the opponent of the correctness of his own assertions, or the incorrectness of the opponent's assertions. In their moves the players express an illocutionary act with a propositional content ( Searle, 1969 ). The illocutionary act is one of the following four:

If a player claims a proposition (act 1) the content of that proposition is in principle free. The only restriction is that the dialog rules allow it. For instance, a player is not allowed to claim a proposition if he is committed to the opposite. The acts 2-4 are referring to a previous claim, so the proposition is the one of this previous claim.

The first move of the dialog is always a claim by one of the players. Suppose the player Bert claims that the cab-driver is liable. At that moment Bert starts a dialog.

Bert: The cab-driver is liable

Ernie can do several things in the second move. He can accept that the cab-driver is liable, ask why the cab-driver is liable (question), or even claim that the cab-driver is not liable at all. In the last case the roles change: now Ernie has the burden of proving that the cab-driver is not liable, so he becomes the proponent and Bert the opponent.

Commitment plays an important role in the dialog. Commitment originates after a proposition is claimed, or after a proposition is accepted. Commitment terminates when a proposition is withdrawn. To avoid that the dialog remains an informal talk, a player needs means to force his opponent to accept a proposition. This is what is called forced commitment, and plays a crucial role in the dialog. It occurs when a player is forced to accept a proposition, due to his commitment to other propositions.

The dialog rules describe formally when a move is allowed, and what the consequences of a move for the commitment of the players are. In the Prolog implementation of the game, the levels are initiated, the commitment is settled, and it is indicated who's turn it is. If a player enters a move (speech act and proposition), the program checks whether the move is allowed. If it is, the dialog continues; if it is not the player gets another change.

3.2 An Example of the Verbal Approach in DiaLaw

As in most mediating systems, in DiaLaw both roles in the dialog game are performed by humans. Although it is already instructive for law students if their mutual argumentation is monitored by a system, a goal to strive for is a computer in its role as opponent in the debate. The examples show interaction between humans, but probably in the future one of the human-players will be replaced by a computer-player. For the sake of readability the sample dialogs are informal.

Bert and Ernie, well-aware of damages that can be caused by driving into a 'bussluis', are discussing who should be held responsible for damages.

For the computer to be helpful as a coach and teacher, a specific domain has to be chosen, for instance that of tort. Another way to restrict the subject of the dialog is to offer students a case description, and a list of statements from which they can select their favourite.

In the above dialog, several argumentation skills are trained. While explaining and elaborating on the subsequent moves of the dialog, it is indicated which specific skill is trained.

Although the Court of Justice decided differently, Bert chooses a reasonable first statement. He claims that the municipality has to compensate for only half of the damages. In the second move Ernie questions this statement. Questions are a powerful tool to use in education, since Bert now has the duty to defend his statement. Therefore, he is forced to make explicit the reasons why his statement holds. The first reason he claims is that the Municipality committed a tort. Ernie accepts this reason, not surprisingly, because it is a good reason. If someone questions the existence of a duty to compensate, a first step is to indicate the ground on which the duty is based. In the current case the fact that the Municipality committed a tort is the correct ground. If Ernie had been a computer-player, he could have asked for the reasons why the Municipality committed tort.

The next statement claimed by Bert is also a good one. He indicates explicitly that there exists an obligation to compensate. Ernie also accepts this statement, but still is not convinced about the correctness of the first claimed statement. After claiming the necessary preliminary statements, in the seventh move Bert claims a statement essential for the position he defends. Although Ernie could have questioned this statement in order to hear Bert's reasons, he follows another strategy.

The eighth move is an interesting point in the dialog. In that move Ernie takes over the initiative. He denies that the cab-driver was guilty for 50%. As long as a player just questions and agrees, the opponent holds the initiative. A way to grab the initiative and become the proponent of a statement is by denying. Ernie has become the proponent of the statement that the cab-driver is not guilty for 50%. Bert questions the denial, because he wants to know why Ernie thinks that his 50% is not appropriate. Ernie mentions the driving in a one way street in the wrong direction, and that without this violation there would not have been damage. To this point Bert agrees. He denies, however, that the violation was the only cause of the damage. As a reason backing his denial, he claims that the municipality is responsible for the damages too, because they acted negligent. This gives Ernie the insight that both the cab-driver and the municipality are to blame for the damages, and he accepts the first claimed statement of the game. Obviously, Bert and Ernie decided from a legal point of view better than the Court did. Only because the procedure already lasted so long the cab-driver did not appeal again. Experts in this field (e.g., the Dutch tort specialist professor Van Maanen) are sure the Supreme Court would have set aside the decision of the Court. In the working out of the example in the system CumulA, the result of the argumentation is different. In that discussion, the decision by the court is used as an authoritative argument. Since there are no points left open for discussion, the dialog ends.

In sum, the students were confronted with the following argumentation skills:

• claiming (defensible) statements;

• claiming reasons that justify questioned statements;

• taking over the initiative in the dialog by denying.

The mentioned skills are important for any lawyer. For instance, these basic skills are used in legal practice frequently. First, positions are taken. In case the claims of the opponent are reasonable they can be accepted. However, probably not all claims of the other party are accepted, but some will be questioned, others will be denied. In the former case the opponent has to provide reasons for his claims, in the latter case initiative is taken over. Finally, convincing reasons justifying ones position can be based on law books or court decisions.

DiaLaw has been implemented in Prolog as a prototype. While consulting the following screen-dump, one should be well-aware that for use in an educational setting the user-friendliness is for now far too low. One of the necessary improvements is that the players should be offered a list of natural language sentences to choose from.

4. Argue!: A Graphical Approach to Legal Argument Mediation

The graphical approach is demonstrated by the Argue!-system. Its underlying argumentation theory is based on CumulA, a procedural model of argumentation with arguments and counterarguments ( Verheij, 1996 ). After introducing CumulA, again a sample session based on the 'bussluis' case is used to illustrate the opportunities of the graphical approach for education. Verheij ( 1998a ) gives a more technically oriented description of the Argue!-system.

4.1 Introducing CumulA

CumulA ( Verheij, 1996 ) is a procedural model of argumentation with arguments and counterarguments. It is based on two main assumptions. The first assumption is that argumentation is a process during which arguments are constructed and counterarguments are adduced. The second assumption is that the arguments used in argumentation are defeasible, in the sense that whether they justify their conclusion depends on the counterarguments available at a stage of the argumentation process.

The goal of argumentation is to (rationally) justify conclusions. In CumulA, the focus is on the process of argumentation, and on the defeasibility of the arguments used in argumentation. Argumentation is a process, in the sense that during argumentation arguments are constructed and counterarguments are brought up. Arguments are assumed to be defeasible, in the sense that if an argument at some stage of the argumentation process justifies its conclusion, it not necessarily justifies its conclusion at all later stages. The defeat of an argument is caused by a counterargument that is itself undefeated.

For instance, if the Municipality has committed a tort against the cab-driver, a conclusion would be that the Municipality has the duty to repair 100% of the damages. The conclusion can be rationally justified, by giving support for it. E.g., the following argument could be given:

Recall that in Dutch tort law, the general duty to repair damages and the portion of the damages to be repaired are established consecutively (see section 2.1).

An argument as above is a reconstruction of how a conclusion can be supported. The argument given here consists of two steps.

An argument that supports its conclusion does not always justify it. For instance, if in our example it turns out that the damages are fully imputed to the cab-driver (as in the 'bussluis' case), the conclusion that the Municipality has the duty to repair 100% of the damages would no longer be justified. The argument has become defeated . In the example, the argument

does not justify its conclusion because of the counterargument

CumulA is a procedural model of argumentation with arguments and counterarguments, in which the defeat status of an argument, either undefeated or defeated, depends on:

• the structure of the argument;

• the counterarguments;

• the argumentation stage.

We briefly discuss each below. The model builds on the work of Pollock ( 1987 , 1995 ), Loui ( 1991 , 1992 ), Vreeswijk ( 1993 , 1997 ) and Dung ( 1995 ) in philosophy, artificial intelligence, and was developed to complement the work on Reason-Based Logic (see, e.g., Hage ( 1993 , 1996 , 1997 ) and Verheij ( 1996 )).

In the model, the structure of an argument is represented as in the argumentation theory of Van Eemeren, Grootendorst and Kruiger ( 1981 , 1987 ). Both the subordination and the co-ordination of arguments are possible. It is explored how the structure of arguments can lead to their defeat. For instance, the intuitions that it is easier to defeat an argument if it contains a longer chain of defeasible steps ('sequential weakening'), and that it is harder to defeat an argument if it contains more reasons to support its conclusion ('parallel strengthening'), are investigated.

In the model, which arguments are counterarguments for other arguments is taken as a primitive notion (cf. Dung, 1995 ). So-called defeaters indicate when arguments can defeat other arguments. It turns out that defeaters can be used to represent a wide range of types of defeat, as proposed in the literature, e.g., Pollock's ( 1987 ) undercutting and rebutting defeat. Moreover some new types of defeat can be distinguished, namely defeat by sequential weakening (related to the well-known sorites paradox) and defeat by parallel strengthening (related to the accrual of reasons).

In the model, argumentation stages represent the arguments and the counterarguments currently taken into account, and the status of these arguments, either defeated or undefeated. The model's lines of argumentation, i.e., sequences of stages, give insight in the influence that the process of taking arguments into account has on the status of arguments. For instance, by means of argumentation diagrams, which give an overview of possible lines of argumentation, phenomena that are characteristic for argumentation with defeasible arguments, such as the reinstatement of arguments, are explicitly depicted. In contrast with Vreeswijk's ( 1993 , 1997 ) model, we show how in a line of argumentation not only new conclusions are inferred, but also new reasons are adduced.

To summarise, CumulA shows

4.2 An Example of the Graphical Approach in Argue!

The implementation of CumulA is another example of a system for the mediation of legal argument. It takes the graphical approach. Students can use the system to construct a line of argumentation. We give an example session, based on the 'bussluis' case.

As a start, a statement is typed, 'The Municipality has committed a tort against the cab-driver':

Statements can be justified by adding reasons (in the figure: 'The Municipality has acted against proper social conduct'), and can be used to draw conclusions ('The municipality has the duty to repair the damages'). This is graphically depicted in a straightforward way, by arrows connecting the statement-boxes.

The reader may have noticed that the statement 'The Municipality has committed a tort against the cab-driver' was first in a grey box, and now is in a white box. This is due to the different statuses that statements can have: if a statement is unevaluated it is in a grey box, if it is undefeated (i.e., justified), it is in a white box. In the example, the statement 'The Municipality has acted against proper social conduct' is undefeated, since it has been added as an assumption. The other two statements become undefeated since there is an undefeated reason for them.

The line of argument continues in order to determine the amount of damages that the Municipality has to pay. At first, the conclusion is drawn that the municipality has the duty to repair 100% of the damages. However, the student recalls something about the importance of imputability:

The statement that the damages are fully imputed to the cab driver is a counterargument to the argument that the municipality has the duty to repair 100% of the damages because the Municipality has committed a tort against the cab-driver. In order to indicate that one argument is a counterargument to another, a special graphical structure is used:

Since the statement that the damages are fully imputed to the cab driver is as yet unevaluated, the statement that the Municipality has the duty to repair 100% of the damages is still justified.

In order to justify the statement that the damages are fully imputed to the cab driver, the relevant case is cited. Since the corresponding statement that the Court  decided on the imputability, is added as a assumption, the conclusion that the damages are fully imputed to the cab driver, becomes justified:

As a side effect, the statement that the Municipality has the duty to repair 100% of the damages, has become defeated (graphically indicated by the cross in the corresponding box), since the argument that the damages are fully imputed to the cab-driver, now is a counterargument.

Now it is concluded that the Municipality has the duty to repair 0% of the damages, on the basis of the reason that the damages are fully imputed to the cab-driver. If desired, the rule that warranted the connection between the reason and the conclusion, can be made explicit by the user of the system:

When the user has stated that the rule of art. 6:102 of the civil code determines the portion of the damages, the session ends:

This example differs from the example in DiaLaw. Here the content of a Court decision was taken over, whereas in DiaLaw the arguments of the court were adduced but defeated by a counter-argument (the municipality was negligent). We do not say that one outcome is better than the other. Both outcomes are defensible, and just show another aspect of argumentation. It would have led to far astray to show an example in which the decision of the Court was criticised by arguments of expert lawyers (like the already mentioned tort specialist professor Van Maanen), but these discussions must be possible if teaching systems are built, e.g. based on either DiaLaw or CumulA or on a combination.

4.3 Opportunities of the Graphical Approach

The implementation of CumulA, illustrated above, can be used in legal education for training several types of argumentation skills:

• Students are forced to make arguments in an explicit reason-conclusion structure. For instance, they become aware of the fact that the same statement can be used as the reason in an argument and as the conclusion of an argument.
   

• Students see that assumptions are needed to justify conclusions and how justified reasons make their conclusion justified.
   

• Students can practice the different ways of using legal rules (as warrants behind reason/conclusion-connections) and legal cases (as support for conclusions) in arguments.
   

• Students experience how counterarguments can be used to defeat arguments, that were previously undefeated. They undergo the successive changes of status. The relation of counterarguments and argument-structure is also clarified.
   

• Students get a feeling for the role of process in argumentation by the gradual construction of arguments and counterarguments, and by the occurrence of status changes.

In training these skills the graphical approach to the mediation of legal argument can be beneficial, e.g.:

• The graphical lay-out gives direct insight in the reason-conclusion structure in a line of argumentation.
   

• The graphical lay-out directly shows which arguments are counterarguments to other arguments.
   

• The changes of statuses (e.g., from undefeated to defeated) are directly noticeable by graphical changes.

The graphical approach is not appropriate in all respects. For instance, it partly requires an unusual attitude towards argumentation. People do not necessarily think in graphical terms of explicit reason/conclusion-structures. The use and presentation of counteraguments is even less familiar. A more fundamental issue is the limitation of the graphical interface. A long line of argument can easily go 'off-screen' and result in a complex and hard to understand structure of statements, reason-arrows, and counterargument-structures. The ArguMed-system ( Verheij 1998b ) has been developed as a successor to the Argue!-system in an attempt to enhance the familiarity of the interface and the transparency of the underlying argumentation theory.

5. Towards a Hybrid Approach

Verbally presenting arguments has a long tradition. The old Greeks already paid attention to argumentation, understandably in verbal style only, while focussing on the core element of current mediating systems: dialectic. One of the first graphical approaches, and yet still widely used in AI & Law, are the argument schemata by Toulmin ( 1958 ).

We have suggested how both the verbal and the graphical approach to the mediation of legal argument can be useful as tools in teaching legal argument. The verbal approach fits in nicely with legal practice. As a result, students can practice skills they need in their professional career. They are trained in choosing arguments with the right meaning and rhetorical power.

The graphical approach has the advantage that it can provide a clear overview of a line of argumentation at a glance. One easily 'gets the picture'. The graphical approach also forces to think of arguments in a new way, namely in terms of pictures.

An additional tool of DiaLaw that has been formally defined,  but is not implemented yet is that the dialog history can be depicted by means of trees ( Lodder, 1998 ). By way of the dialog trees a good insight into the layered structure of the dialogs is provided. Implementing these trees could be a first step towards combining the verbal and graphical approaches.

We recommend that both the verbal and the graphical approach are combined in one system, in order to profit from the best of each. E.g., the system could provide means to switch between different presentations. In order to encourage students to use the system, a game-like element is essential. In a dialog game they can try to win by beating their opponent, e.g., by drawing a convincing argument.

We think that integrating systems such as Argue! and DiaLaw could lead to a good result. Obviously, a hybrid system does not have to be based on an integration of the systems described in this paper, and we hope that others take up the thread by realising a genuine hybrid system for the mediation of legal argument. The next step would then be to use and test the system in an educational environment.

6. Conclusion

In this paper opportunities of computer-mediated legal argument in education are suggested. Two approaches to the presentation of arguments, the verbal and the graphical, have been discussed. The DiaLaw system and the Argue! system have been described as examples. Although both approaches have their specific merits, we think that a combination of the two would be most promising in an educational environment. We recommend that future research on automated tools for teaching legal arguments focuses on developing systems in which arguments are presented both graphically and verbally. By the development of attractive systems, e.g., with a game element, students can be encouraged to train their argumentation skills. Since training in these skills does often not receive much attention in the overloaded curricula of legal education, computer-mediated legal argument could become a valuable addition to argumentation courses in legal education.

References

Aarnio, A., Alexy, R., and Peczenik, A. (1981). The foundation of legal reasoning. Rechtstheorie 21, pp. 133158, 257-278, 423-448.

Aleven, V., and Ashley, K.D. (1997). Evaluating a Learning Environment for Case-Based Argumentation Skills. The Sixth International Conference on Artificial Intelligence and Law. Proceedings of the Conference , pp. 170-179. ACM, New York (New York).

Bench-Capon, T. (1995). Argument in Artificial Intelligence and Law, in: J.C. Hage et al. (eds.), Legal Knowledge Based Systems: Telecommunication and AI and Law, JURIX '95, Kloninklijke Vermande, Lelystad.

Bench-Capon, T.J.M., Leng, P.H., and Staniford, G. (1998). A Computer Supported Environment for the Teaching of Legal Argument. The Journal of Information, Law and Technology (JILT), 1998 (3), < http://www2.warwick.ac.uk/fac/soc/law/elj/jilt/1998_3/bench/>.

Centinia, F., Routen, T., Hartmann, A., and Hegarty, C. (1995). STATUTOR: Too intelligent by half? Legal knowledge based systems JURIX '95: Telecommunication and AI and Law (eds. J.C. Hage, T.J.M. Bench-Capon, M.J. Cohen, and H.J. van den Herik), pp. 121-132. Koninklijke Vermande, Lelystad.

Dung, P.M. (1995). On the acceptability of arguments and its fundamental role in nonmonotonic reasoning, logic programming and n -person games. Artificial Intelligence , Vol. 77, pp. 321-357.

Eemeren, F.H. van, Grootendorst, R., and Kruiger, T. (1981). Argumentatietheorie. Uitgeverij Het Spectrum, Utrecht.

Eemeren, F.H. van, Grootendorst, R. and Kruiger, T. (1987). Handbook of Argumentation Theory. A Critical Survey of Classical Backgrounds and Modern Studies . Foris Publications, Dordrecht. Translation of van Eemeren et al. (1981).

Farley, A.M. and Freeman, K. (1995). Burden of Proof in Legal Argumentation, Proceedings of the Fifth International Conference on Artificial Intelligence and Law , ACM, New York, pp.156-164.

Fernhout, F.J., Cohen, M.J., Crombag, H.F.M., Pinckaers , R., and Temme, W. (1987). OBLIGATIO: computer simulation of legal cases. Leren studeren in het hoger onderwijs, pp. 204-209. Almere.

Gordon, T.F. (1993). The Pleadings Game - Formalizing Procedual Justice, Proceedings of the Fourth International Conference on Artificial Intelligence and Law , ACM, New York.

Gordon, T.F. (1995). The Pleadings Game - An Artificial Intelligence Model of Procedural Justice. Kluwer, Dordrecht.

Habermas, J. (1973). Wahrheitstheorien. Wirklichkeit und Reflexion, Festschrift f. W. Schulz , pp. 211-265. H. Fahrenbach, Pfullingen.

Hage, J. (1993). Monological reason based logic. A low level integration of rule-based reasoning and case-based reasoning. The Fourth International Conference on Artificial Intelligence and Law. Proceedings of the Conference , pp. 30-39 . ACM, New York (New York). Also published as report SKBS/B3.A/93-08.

Hage, J. (1996). A Theory of Legal Reasoning and a Logic to Match. Artificial Intelligence and Law , Vol. 4, pp. 199-273.

Hage, J. (1997). Reasoning with Rules. An Essay on Legal Reasoning and Its Underlying Logic. Kluwer Academic Publishers, Dordrecht.

Hage, J, Leenes, R. and Lodder, A.R. (1994). Hard cases: a procedural approach. Artificial Intelligence and Law 2: 113-167.

Kowalski, R. and Toni, F. (1996). Abstract Argumentation. Artificial Intelligence and Law , Vol. 4, pp.275296.

Lodder, A.R. (1996). Legal debate as an educational tool. Verso un sistema esperto giuridico integrale/Towards a global expert system in law, Tomo II (eds. C. Ciampi, F. Socci Natali & G. Taddei Elmi), pp. 129-138. CEDAM, Padova.

Lodder, A.R. (1998). DiaLaw - on legal justification and dialog games . Dissertation, Universiteit Maastricht.

Lodder, A.R., and Herczog, A. (1995). DiaLaw - A dialogical framework for modeling legal reasoning. The Fifth International Conference on Artificial Intelligence and Law. Proceedings of the Conference , pp. 146-155. ACM, New York (New York).

Lodder, A.R., and Verheij, B. (1998). Opportunities of computer-mediated legal argument in education. Proceedings of the BILETA-conference - March 27-28 . Dublin, Ireland.

Loui, R.P. (1991). Ampliative Inference, Computation, and Dialectic. Philosophy and AI. Essays at the Interface (eds. R. Cummins and J. Pollock), pp. 141-155. The MIT Press, Cambridge (Massachusetts).

Loui, R.P. (1992). Process and Policy: Resource-Bounded Non-Demonstrative Reasoning. Report WUCS-92-43. Washington University, Department of Computer Science, Saint Louis (Missouri).

Loui, R.P., Norman, J., Altepeter, J., Pinkard, D., Craven, D., Lindsay, J., and Foltz, M. (1997). Progress on Room 5. A Testbed for Public Interactive Semi-Formal Legal Argumentation. The Sixth International Conference on Artificial Intelligence and Law. Proceedings of the Conference , pp. 207-214. ACM, New York (New York).

Loui, R.P., Norman, J., Olson, J., and Merill, A. (1993). A Design for Reasoning with Policies, Precedents and Rationales, Proceedings of the Fourth International Conference on Artificial Intelligence and Law, ACM, New York, pp. 202-211.

Muntjewerff, A., and Groothuismink, J. (1998). PROSA. A Computer Program as Instructional Environment for Supporting the Learning of Legal Case Solving. Legal Knowledge-based Systems.  JURIX: The Eleventh Conference (eds. J.C. Hage, T.J.M. Bench-Capon, A.W. Koers, C.N.J. de Vey Mestdagh, and C.A.F.M. Grutters), pp. 85-100. GNI, Nijmegen.

Nitta, K., and Shibasaki, M. (1997). Defeasible reasoning in Japanese criminal jurisprudence. Artificial Intelligence and Law , Vol 5, pp. 139-159.

Perelman, Ch., and Olbrechts-Tyteca, L. (1971). (Original French version in 1958). The New Rhetoric. A Treatise on Argumentation . University of Notre Dame Press, London.

Pollock, J.L. (1987). Defeasible reasoning. Cognitive Science , Vol. 11, pp. 481-518.

Pollock, J.L. (1995). Cognitive Carpentry: A Blueprint for How to Build a Person. The MIT Press, Cambridge (Massachusetts).

Prakken, H. and Sartor, G. (1996). A dialectical model of assessing in conflicting arguments in legal reasoning. Artificial Intelligence and Law , Vol. 4, pp. 331-368.

Routen, T. (1991). Complex Input: A Practical Way of Increasing the Bandwidth for Feedback and Student Modelling in a Statute-Based Tutoring System. The Third International Conference on Artificial Intelligence and Law. Proceedings of the Conference , pp. 77-80. ACM, New York (New York).

Searle, J.R . (1969). Speech acts: an essay in the philosophy of language . Cambridge university press.

Span, G. (1994). LITES: an Intelligent Tutoring System Shell for Legal Education. International Yearbook of Law Computers and Technology , Vol. 8, pp. 103-113.

Toulmin, S.E. (1958). The uses of argument. University Press, Cambridge.

Verheij, B. (1996). Rules, Reasons, Arguments. Formal studies of argumentation and defeat . Dissertation Universiteit Maastricht. A summary and table of contents are available on the World-Wide Web at < http://www.metajur.unimaas.nl/~bart/proefschrift/ >.

Verheij, B. (1998a). Argue! - an implemented system for computer-mediated defeasible argumentation. NAIC '98. Proceedings of the Tenth Netherlands/Belgium Conference on Artificial Intelligence (eds. Han La Poutre and Jaap van den Herik), pp. 57-66. CWI, Amsterdam.

Verheij, B. (1998b). ArguMed - A Template-Based Argument Mediation System for Lawyers. Legal Knowledge Based Systems. JURIX: The Eleventh Conference (eds. J.C.Hage, T.J.M. Bench-Capon, A.W. Koers, C.N.J. de Vey Mestdagh and C.A.F.M. Grutters), pp. 113-130. Gerard Noodt Instituut, Nijmegen.

Verheij, B., Hage, J., and Lodder, A.R. (1997). Logical tools for legal argument: a practical assessment in the domain of tort. The Sixth International Conference on Artificial Intelligence and Law. Proceedings of the Conference , pp. 243-249. ACM, New York (New York). An abstract is available on the World-Wide Web at < http://www.metajur.unimaas.nl/~bart/papers/icail97.htm >.

Verheij, B., and Lodder, A.R. (1998). Computer-mediated legal argument: the verbal vs. the visual approach. Proceedings of the 2 ^nd French-American Conference on Artificial Intelligence and Law - June 11-12 . Nice, France.

Vreeswijk, G.A.W. (1993). Studies in defeasible argumentation. Doctoral thesis, Vrije Universiteit, Amsterdam.

Vreeswijk, G.A.W. (1997). Abstract argumentation systems. Artificial Intelligence , Vol. 90, pp. 225-279.

Bibliography

Freeman, K., and Farley, A.M. (1996). A Model of Argumentation and Its Application to Legal Reasoning. Artificial Intelligence and Law , Vol. 4, pp. 163-197.

Gordon, T.F., and Karacapilidis, N. (1997). The Zeno Argumentation Framework. The Sixth International Conference on Artificial Intelligence and Law. Proceedings of the Conference , pp. 10-18. ACM, New York (New York).

Loui, R.P., and Norman, J. (1995). Rationales and argument moves. Artificial Intelligence and Law , Vol. 3, pp. 159-189.

Moore, D.J. (1993). Dialogue games and computer aided learning . Doctoral dissertation, Leeds Metropolitan University.

Prakken, H. (1997). Logical Tools for Modelling Legal Argument. A Study of Defeasible Reasoning in Law. Kluwer Academic Publishers, Dordrecht.

Yoshino, H. (1995). The Systematization of Legal Meta-inference. The Fifth International Conference on Artificial Intelligence and Law. Proceedings of the Conference , pp. 266-275. ACM, New York (New York).

Links

Thomas Gordon - < http://nathan.gmd.de/persons/thomas.gordon.html >
Arno R. Lodder - < http://www.rechten.vu.nl/~lodder/ >
Ronald P. Loui - < http://www.cs.wustl.edu/~loui/ >
John L. Pollock - < http://www.u.arizona.edu/~pollock/ >
Henry Prakken - < http://www.cs.vu.nl/~henry/ >
Bart Verheij - < http://www.metajur.unimaas.nl/~bart/ >

Tarski's World - < http://csli-www.stanford.edu/hp/ >
The OSCAR project - < http://www.u.arizona.edu/~pollock/ >
Room 5 - < http://www.cs.wustl.edu/~room5/ >
The ZENO project - < http://nathan.gmd.de/projects/zeno.html >
Argue! and ArguMed - < http://www.metajur.unimaas.nl/~bart/aaa/ >

Dutch National Programme Information Technology and Law (ITeR) - < http://www.nwo.nl/iter/ >
ITeR project 01437112 - < http://www.nwo.nl/iter/thme4l3.html >