Jusletter IT

This paper discusses the notions of situation and case. They are characterised by different methods of legal informatics. We start with a note that a situation stands for a type of behaviour and a case stands for an instance.

Fig. 1: A sample road situation which is presented to an examinee to take a driving theory test. The examinee has to answer a question. In this situation the question is «Which order does automobile 2 drive?» (Blue-red beacon and sound signal are on 4). The right answer: the third, along with 1.

In this paper we attempt to develop a theoretical framework (a conceptualisation). Two theses are defended. First, situations are governed primarily by the principle «Roles, not rules.» Second, in legal cases two kinds of subsumption – textual and legal – are in the foreground.

An example of a situation is a crossroad description. The roles comprise pedestrians and the drivers of different types (car, bus, tram, ambulance, motorcyclist, motorbike, bicyclist, horseman, animal driver, etc.). A sample situation – a crossroad with priorities – is shown in Fig. 1. This is familiar to examinees who take a driving theory test. The test has situation descriptions and questions. Road elements, automobiles and road signs are involved. Drivers’ intentions are described with text and symbols such as arrows. This description makes part of situation’s teleology. The alternatives are essential. For example, examinees answer multiple choice questions.

Another example shows schematic representations of situations in air traffic (Fig. 2).

Fig. 2: Sample representations of situation types in air traffic.

Situations and cases are characterised differently:

 

Situations and cases are characterised differently:

1. Type. A situation constitutes a generic behaviour pattern whereas a case – a concrete one.
2. Ex-ante/ex-post. A situation is related with ex-ante analysis whereas a case – with ex-post.
3. Time. A situation concerns the future whereas a case – the past.
4. Alternatives.

   Situation: alternatives are possible. This is the essence of a situation.

   Case: Concrete behaviour. Events are passed.

   However, alternatives can appear in hypothetical evaluations such as «Should the actors perform another manoeuvre, the accident would not happen.»

5. Language

Situation. Situations have no language at all.

1. A situation is mentally – visually, acoustically, sensibly – interpreted. Suppose a driver in a crossroad (see e.g. Fig. 1). A mental language is non-textual and non-professional. Sensual (visual, aural, etc.) comprehension dominates and textual descriptions appear on the periphery. Hence, a situational language is non-professional. A communication language needs not to be textual; cf. gestures. Therefore a situational language is loosened and differs from case languages.
2. Roles are inherent in situations, e.g. «pedestrian», «driver», «pilot». Actors’ legal status may be implicit because rights and obligations are comprised by their roles.
3. Artificial agents can use formal languages. As an example suppose multi-agent systems. Agent’s beliefs, desires and intentions are represented in computers in a knowledge representation language.
6. Placing onto the Is and Ought stages
   Situation
   1. i. Situations are assigned to Is. A situation is always real, factual. As an example suppose a crossroad with red light on. You would like to cross but do not want to show children a bad example while they learn the custom law from your behaviour.
   2. In contrast, the type of a situation is assigned to Ought. A situation type allows visual representations such as a schema. This appears in technical devices.
   Case
   
   1. Cases are also assigned to Is. Every case has passed. The reference range is not important. A case is fixed in the text.
   2. A case is on the Is stage but can be viewed from two perspectives. Firstly, the case is assigned to the subjective law. Secondly, the case is assigned to legal proceeding, hence, to the objective law. Here argumentation arises. The players are assigned legal proceeding roles such as plaintiff, defendant, witness, expert, etc.
7. Web applications. e-Government application examples in Austria:
   Situations see in www.help.gv.at. Cases see in www.ris.bka.gv.at.
8. Legal instruments. Distinct legal instruments are concerned:

Situation: (i) the roles of actors; (ii) assumptions (hypothetical facts); (iii) rules which govern the situation; (iv) additional regulations which govern the situation.
Case: (i) claim; (ii) evidence; (iii) attacks; (iv) litigation can consist of several cases (e.g. criminal and civil).

9. Formalisms. Distinct legal instruments are concerned:

Situation: deontic logic, abstract normative systems, etc.
Case: case modelling approaches, factors, etc.

10. Representation in computer science. Knowledge representation: frame-based.
    Situation: slots are partially filled (slotName := value). Alternatives result when empty slots are filled in.
    Case: slots are completely filled (probably with the value «unknown»).
11. Customary law, machine law and statutory law

Situation: customary law and machine law are in the foreground. As an example suppose a zebra crossing. Pedestrians aim to cross it. Statutory law (the road rules) regulates. However, ordinary people are governed primarily by customary law which superimposes. And finally the situation is governed by traffic lights – machine law steps in.
Cases: traditional hierarchies of legal sources prevail.

Cases appear on Is whereas legal rules appear on Ought (Fig. 3). Here we use a visualization pattern that is composed of two stages, the vertical Ought stage and the horizontal Is (Fig. 3). The two stages depict Kelsen’s categorical distinction between Is and Ought; see [Kelsen 1967, § 3 ff.].

Fig. 3: A horizontal and a vertical stage correspond to Is and Ought.

 

Attribute	Situation	Case
1. Examples	A crossroad Airways separation pattern	Traffic accident Mid-air collision
2. Type	Generic, i.e. a type, class.	Concrete, individual, an instance of class.
3. Analysis	Ex-ante	Ex-post
4. Time	Future	Past
5. Alternatives	Possible	Not possible
6. Language	Mental, non-textual, non-professional.	Textual, professional. Witnesses speak in non-professional language and jurists in professional. Visual representations, e.g. schemas, are supplementary
7. Language elements	Roles. Legal status implicit in the roles	Roles: plaintiff, defendant, witness, expert. Argumentation
8. Placing on Is and Ought	Is. In contrast, the type of a situation appears on Ought	Is. Subjective law. Legal proceedings are related to the objective law.
9. Web example	www.help.gv.at	www.ris.bka.gv.at
10. Legal instruments	Roles, assumptions, rules which govern the situation, additional regulations	Claim, evidence, attack
11. Representation formalisms	Deontic logic worlds which can be accessed in the future	Facts in the past. Different narratives of the players
12. Frame-based representation	Slots are partially filled	All slots are filled
13. Customary vs. statutory law	Ordering: 1) machine law, 2) customary law, 3) statutory law	Traditional hierarchies of legal sources
14. Bridging	Story-telling

Table 1: A comparison of structural elements of situation and case

Situations and cases are described by different concepts. Thus distinct conceptualisations result. A situation is viewed as a state of the world. A situation is not finished. Different scenarios can evolve from the sole state. In terms of logic, central are the different worlds that can be accessed from the given situation. In terms of normative systems, the legal status of the players (normative positions) is a key. Therefore formalisms to represent possible worlds and legal statuses are in the background. Examples are deontic logic (see e.g. [Prakken & Sergot 1996]), abstract normative systems [Tosatto et al. 2012], input-output logic of [Makinson & van der Torre 2012].

Cases are finished and the outcome is known, e.g. «Mr. Lammers is dead shot with a gun». Thus, the scenario is finished. However, the players may have different stories and this is central. For example, the plaintiff and the defendant may render contrary arguments, e.g. «Rijkbloem shot» and «Mrs. Lammers shot». There is a variety of case modelling approaches; see e.g. a special issue of Artificial Intelligence and Law 16(2) (2008), esp. Prakken. Formalisms for arguments, stories and evidence comprise a hybrid formal theory [Bex et al. 2010] that combines argument-based and story-based approaches.

Which elements are important in the constellations of situations and cases? Figure 4 shows core elements which make the differences.

Fig. 4: Situation and Case. Attention and Subsumption.

Situation. Here attention is the most important element. Attention can be compared with a cursor that can change to different positions. The players reside similarly as on a stage and make scenes like people around a table. A script assigns the players their roles.

Case. Here legal subsumption, i.e. bringing the text under norms, is in the front. It is important that the case elements are relevant to norms. The elements of the issue have to be named in a professional legal language. Cases are marked by verbalisations. Here relationships – references – of the text to the relevant norms are dealt with.

A situation appears on the horizontal Is stage and is described by the following entities (Fig. 5):

• Situative elements. These are constituents of the situation. They are denoted by small letters, e.g. a, b, c, driver, pedestrian, etc. They exist in time and space.
• Relations between the situative elements. There are many kinds of relations: causal (^c→), teleological (^te→), instrumental (^instr→), context (^contx→), etc. These relations are comprised by both legal relations such as debt, but also of empirical non legal relations. The relations represent different perspectives.

Fig. 5: A notation (conceptualisation) for situation.

The alternatives of behaviour are important. Suppose you do not have money and therefore you cannot take a tram, but if you have you can take a transport. For example, the notation for a causality relation «a causes b» is a ^c→ b.

A predicate language can serve to represent situations; cf. block worlds in early artificial intelligence. Situative elements are represented by constants (0-arity predicates). A relation of arity n is represented by a predicate R(a₁, a₂,…, a_n).

Distinguishing representations for humans and computers. Human beings and computers use different formats of representations and reasoning to act in a situation. Humans employ essentially more formats than computers.

Human beings comprehend a situation with their senses. Human’s brain makes a decision which action to take. In a situation, a human reasons primarily in terms of roles not legal rules. Predicate logic is not in the forefront in human decision making unless a computer decision support system is employed (e.g. air traffic controllers). However, predicate logic can be employed to model and justify an action that was taken.

Computers in a multi-agent systems use computer knowledge representation formats. Predicate logic and other representations of n-ary relations can be used. Computer agents use machine inference to take a decision on an action. However, computers can be more effective than humans in specific tasks. Suppose making an emergency action based on instrumental sensors. An example is an airborne collision avoidance system (ACAS)¹.

The picture in Fig. 1 should be used by an individual to pass the driving test. However, a similar situation can happen to three drivers on a real road, too. An unmanned vehicle could also occur there. Each actor would use different representation of the situation and different decision making.

The facts of a case are transformed into legal terms. The facts are obtained from story-telling, witnesses, etc. For example, an action, a, is treated as a theft, A, not a burglary. This makes a first kind of legal subsumption, called terminological subsumption. A notation is a == A. The instance-of notation of computer science can also be used, a instance-of A. A prefix form instance-of(a,A). A pool of legal terms is used for the terminological subsumption. Legal ontologies can serve here.

Fig. 6: A notation (conceptualisation) for case.

The second step is a normative subsumption. Here the norm Norm( obligatory A(x) otherwise sanction B(x) ) is applied to subsume a sanction B. The first step, terminological subsumption, corresponds to the unification. It is linked with the minor premise. The second step, normative subsumption, corresponds to the major premise xD Norm( A(x) / B(x) ).

The conceptualisation above is similar to the following form of inference (cf. the syllogism and also the modus ponens rule):

Minor premise: Socrates is a human

Major premise: Humans are mortal.

Therefore, Socrates is mortal.

The notation (actually a conceptualisation) for the case concept is shown in Fig. 6.

Story-telling serves as a bridge between situations and cases. As an example suppose a judge in a court announcing the verdict of a case. The meaning of the judge’s speech act is a legal act. The verdict contains a statement of grounds. It is presented as a kind of story-telling. It summarises the established facts and their interpretation and is the official version of the story.

To model the subsumption procedure, conceptual modelling formalisms which are used in computer science can be applied. The general relationships is-a, instance-of and part-of are used in object-oriented analysis and systems development. Suppose the fact that my-door is open and the norm N «The doors ought to be closed». The norm can be formalised with the following rule: if x is an instance of Door, then x ought to be closed. Formally, x xDoor  O closed(x), where O is the deontic operator and closed a predicate. The situation (fact) with the instance my-door is from the Is world. The fact is interpreted according to a norm from the Ought world which contains the door concept Door. Then my-door is matched with Door, formally match(my-door, Door). This can be simplified and expressed with a truth statement instance-of(my-door, Door) or my-doorDoor. This truth statement is from the Is world. A graphical notation is shown in Fig 5. A duty which is conferred on me, to close my-door, is from Ought. In the Is world I can decide to leave my-door open thus violating the norm.

Fig. 7: Graphical notation of the instance-of relationship, instance-of(my-door, Door) or my-doorDoor. This visualizes that my-door (from the Is world) is matched with the Door concept within the norm N which is from Ought.

This paper aims to contribute to formalising informational processes in the domain of legal informatics. The first step is to identify the notions which distinguish legal situations from legal cases. They are summarised in the table. Two kinds of subsumption – terminological and normative – are central in the case concept. Providing a more rigorous notation is a work in progress.

V. Čyras has been supported by the project «Theoretical and engineering aspects of e-service technology development and application in high-performance computing platforms» (No. VP1-3.1-ŠMM-08-K-01-010) funded by the European Social Fund.

Bex, Floris J.; van Koppen, Peter J.; Prakken, Henry; Verheij, Bart, A hybrid formal theory of arguments, stories and criminal evidence. Artificial Intelligence and Law 18(2):123–152 (2010).

Bex, Floris; Bench-Capon, Trevor; Atkinson, Katie, Did he jump or was he pushed? Abductive practical reasoning. Artificial Intelligence and Law 17(2):79–99 (2009).

Kelsen, Hans, Pure Theory of Law. 2nd ed., Max Knight, trans. University of California Press, Berkeley, 1967.

Prakken, Henry; Sergot, Marek, Contrary-to-duty obligations. Studia Logica 57(1):91–115 (1996).

Sartor, Giovanni, Legal Reasoning: A Cognitive Approach to the Law. A Treatise of Legal Philosophy and General Jurisprudence, vol. 5. Springer, Heidelberg (2007).

Tosatto, Silvano Colombo; Boella, Guido; van der Torre, Leendert; Villata, Serena, Visualising normative systems: an abstract approach. In: T. Ågotnes, J. Broersen, D. Elgesem (Eds.): Deontic Logic in Computer Science – 11th International Conference, DEON 2012, Bergen, Norway, July 16-18, 2012, Proceedings. Lecture Notes in Computer Science 7393, pp. 16–30. Springer (2012).

 

---

 

Vytautas Čyras, Docent, Vilnius University.

 

Friedrich Lachmayer, Professor, University of Innsbruck.

 

---