APPLYING HOHFELDIAN ANALYSIS TO PROCESS MODELING

(1)

APPLYING HOHFELDIAN

ANALYSIS TO PROCESS

MODELING

MASTER THESIS

Yahia Tharwat Mahmoud El-Sherbini

Master Information Studies: Business Information Systems

Student Number: 10629076

Supervisor: Alexander Boer

(2)

2

Abstract: This thesis aims to conceptualize a framework that implements Hohfeld's jural

relations to process modeling with the ability to use Petri Nets as a computational tool for story acquisition. The framework consists of five layers of abstraction that combine different modeling tools with the fifth layer transforming the models to Petri Nets. The research was conducted by exploring the use of Hohfeld's analysis as a tool for expressing normative social behavior in various contexts. The framework provides a tool to acquire scenarios of non-compliance that can be used to detect and handle it for business and legislative processes.

(3)

3

ACKNOWLEDGMENTS

This research is the final requirement for the Msc degree in Information Studies at the University of Amsterdam.

I would like to thank Giovanni for the great amount of inspiration he provided me and the valuable feedback which helped me get on track, many times over.

Lastly, my thanks and appreciations go to Alexander. I highly enjoyed and learned from the mind-opening discussions we had over the course of this research that ranged various topics: Culture, history, politics, and most importantly, all things Hohfeld.

Yahia El-Sherbini Amsterdam, 2014

(4)

4

Introduction

Non-compliance is an inevitably occurring issue in any story where it involves a transaction or legal service. The notion of preventing non-compliance is considered unrealistic due to the underlying nature of human behavior. However, expecting and detecting non-compliance allows for it to be handled when it occurs. Therefore, stories of non-compliance are crucial for the enhancement of the design space in any field. When designing a process, scenarios of non-compliance are considered by collecting stories. For example, it has been used in the legal domain as a mean for interpreting social behavior to improve case law (Sileno, Boer, & van Engers, 2013). With this in mind, utilizing scenario based modeling as a tool, for acquiring different aspects of conclusions of stories, would improve the overall structure. For this reason, it is essential to pave the way for transforming simple narrative models to intricate computational tools. To achieve this task, there has to be an understanding of modeling tools and legislative rights.

Hohfeld's analysis of legal interests provides an understanding of human behavior in legal transactions by defining the relationship between two actors involved. Therefore, allowing a better assessment for legal conclusions (Hohfeld, 1913). Since transactions usually occur between at least two persons, applying Hohfeld's rights to process modeling by defining non-normative actions would allow the handling of non-compliance.

Process modeling tools like BPMN and UML diagrams are popular among businesses for visualizing their processes. Processes are, in their simplest form, derivatives of narratives constrained by business rules. So, experimenting on these tools for story acquisition is not off the grid with regards to their original use. Applying Hohfeld's rights to the models would shed light on the legal positions of each actor in the different phases of a process. However, this thesis also aims to allow the seamless transformation from simple process modeling tools like BPMN and UML to an elaborate, and more importantly, computational tool like Petri Nets. This thesis aims to find a way to achieve the aforementioned endeavor. It will serve as a contribution to a methodology proposed by Sileno, Boer, & van Engers (2013). It will focus on creating a framework to implement Hohfeld's analysis as an institutional element to process modeling. The paper will discuss the interpretations of Hohfeld's analysis and the methodologies presented in literature for applying the analysis to process modeling. The conception of the framework will be discussed while referencing the relevant literature. The framework will be explained with a meta-model and finally tested with two cases. The first case will entail a sale transaction while the second will entail traffic rules.

(7)

7

Research Motivation

The motivation of the research is to enhance the performance of scenario-modeling. In any business, creating a business process usually follows the happy-flow of the process. The happy-flow is the expected behavior of the actors in the process that is controlled by the business rules. It is usually the case where that happy-flow is disrupted in practice because not everyone tends to follow the rules. Therefore, in legislative processes a heavy analysis of stories and different outcomes are studied before applying them. In my studies, I had the opportunity to model processes about foreigners applying for residence permits in the Netherlands. The goal of the assignment was to consider non-compliance from different aspects so that they can be handled when they occur. These kinds of processes require an extensive study of previous scenarios and speculating other scenarios to create the processes. Another aspect in legislative processes is the reliance on law for both applying it and handling non-compliance. Looking at business processes, business rules are the laws which they derive from. It is a matter of adding institutional elements in scenario modeling to provide expectancy of non-compliance.

A Hohfeldian Approach

As mentioned in the introduction, Hohfeld's analysis will be the institutional element which will help in modeling social behavior in scenarios. An extensive literature research has thus been conducted to first understand the elements and consider the different views in the field of law. The focus of the research was on how these elements are used.

Hohfeld's analysis

In 1913, Wesley N. Hohfeld formalized a theory of defining rights called the fundamental legal conceptions (Hohfeld, 1913). The theory explained eight jural relations to the concept of rights as a mean to simplify legal relationships between two persons. The eight jural relations are divided into pairs, opposites and correlatives: the former cannot exist together while the latter must:

Opposites: Right –No-right, Privilege – Duty, Power – Disability, Immunity – Liability. Correlatives: Right – Duty, Privilege – No-right, Power – Liability, Immunity – Disability. The eight relations are the "Lowest common denominators of law" as Hohfeld defined them (Hohfeld, 1913, p.58). For the purpose of this thesis, a clear understanding of each relation is needed to develop a concrete concept.

Right-Duty:

It is when a person holds a right or claim against another. For example, A has a right(claim) that B pays him the money he owes. Therefore B has a duty to pay the money he owes to A. A right cannot exist without it's correlative duty.

Privilege-No-right:

A privilege is the absence of a duty against another or it can be considered at liberty of doing an action. A has a privilege of smoking in designated areas while B, who does not want someone smoking around him, has no- right to prevent A's act of smoking.

(8)

8 Power-Liability:

A power is the ability to change a legal relation with another. A has the legal power to go into a selling contract with B, thus B is liable to A changing their legal relation with each other. Immunity-Disability:

Immunity is B having no power over A, thus A is immune to B changing their legal relationship. For example, B has no power to go into legal contract with A since A is a minor. This means that B has a disability which is the opposite of power.

Different interpretations of Hohfeld's analysis

Although Hohfeld defined the eight jural relations to help with assessing legal cases, he did not explicitly state what counts as rights, duties, privileges etc. This leaves room for interpretation especially when trying to label the two person's positions before a case is made. Since this thesis aims to define the positions in a transaction, justifying the choice of relations in different contexts is highly important.

One of the hurdles faced with the jural terms are the meaning of the terms. The word "Privilege" means having an advantage, which is sometimes the case when using its jural relation. But in certain circumstances the word "Liberty" defines it better (Williams, 1956). Commons (1932) argued the substitution of the word "limits" instead of "opposites" to define the jural relations opposites. Since liberty means the absence of a duty, he claims that it is a quantitative matter of a limit to duties. Following this line of thought, Commons (1932) suggests the use of the term "Exposure" instead of "no-right" for its quantitative nature. Using alternative terms for the analysis, however, may help understand and assign it better but it does not refute it as Anderson & Moore (1957) discussed.

Commons' work attempted to relate the jural relations to transactional economics (Fiorito, 2010). He criticized Hohfeld's bilateral nature of the jural relations because of the lack of connection to state officials with regard to the relation at hand. Since transactions usually follow law to allow for enforcement, he introduced state officials as an evident actor. Therefore, he considers his interpretation of Hohfeld's analysis to be multilateral. Although there are many alterations and criticism to Hohfeld, the analysis still stands as a powerful analytical and rhetorical tool of legal relations.

Ownership

Ownership is a complicated matter when applying Hohfeld's rights since it is not merely one right of ownership but a complex collection of rights, privileges, etc.

There are two theories of rights; rights in rem and rights in personam. A right in rem is a relationship between a person and an object which makes it a right held against no specific person, while a right in personam is a relationship between a person and another which makes it held against a specific person. Since Hohfeld's analysis does not equal a legal right but a bundle of rights, privileges, etc, he replaced this theory with multital rights and paucital rights (Hohfeld, 1917). Where rights in personam are paucital rights held against a definite number of persons and rights in rem are multital rights held against an indefinite number of persons.

(9)

9

Ownership and property fall under multital rights since it is an object and an owner can have rights, privileges, etc against specific persons and the whole world.

There is another aspect to consider when mentioning ownership. Cook (1918) mentioned Hohfeld's analysis with regard to "exclusive" and "concurrent" rights. "Exclusive" rights are exclusive to the laws of equity, where they have the upper hand. "Concurrent" rights are concurrent between common law and equity. This theory corrects the perception of rights with regard to ownership. Someone who owns an object, may it be a car, has the perception that he may due as he pleases with it (I.e destroy it) but certain laws prevent him to do so. For example, blowing up the car in the middle of the street would invoke certain duties.

(10)

10

Research Design

Goals

One of the problems faced with using Hohfeld's analysis for computational tools is the uncertainty that befalls their use. The analysis can interpret any scenario in different ways that, with enough discussion, can be proven valid. This lack of one strict interpretation makes it hard for the theory to be applied with computational tools. The goal of this thesis is to create a framework for Hohfeld's analysis that can help apply them to process modeling. The framework provides a template for their use with regards to general transactions. It is expected to be improved the more it is used. The second goal for the thesis is to transform the models created to Petri Nets for story-acquisition.

Research Question

The research question posed is:

How can process modeling depict Hohfeldian analysis and be applied to Petri Nets?

As mentioned in the introduction, the diagrams that are widely used for process modeling are UML diagrams and BPMN due to their simple structure. The question aims to find a way to implement Hohfeld's analysis for these diagrams. The research focuses on assigning the activities in the processes with jural relations to provide a layer of institutional actions for the models. The aim, however, is not to complicate the graphical notation of the models to allow it to be easy to understand and transform to Petri Net.

Proposition and methodology

Hohfeld's jural relations are the common denominators of law as Hohfeld defined them (Hohfeld, 1913). They exist with regard to people's actions towards one another. A transaction is considered an act of exchange between people by performing tasks to conclude. The connection between the two is inevitable since handling non-compliance in transactions is often done by law. I propose that the tasks portrayed in processes can be associated with the jural relations. Once this association is added to the modeling of the processes, translating them to Petri Nets for computational tools is a matter of assigning the jural relations to the graphical notation of the models. Designing a business process often requires several diagrams to be modeled for different depictions. When creating a database for instance, designers make use of UML Use case, sequence diagram, activity diagrams and state diagrams for implementation. Creating a process to portray Hohfeld's jural relations can be done with the combination of different diagrams.

The method of doing such a task is to find related literature that have approached this subject and combine different theories to create a coherent framework. The topics that are concerned with the literature review are:

1. Logical frameworks for Hohfeld's analysis. 2. Methodologies for scenario-based modeling. 3. Transformation of diagrams to Petri nets.

Validity

The validity of the research will be provided with the use of literature, cases and sufficient argumentation. The framework constructed in this research relies on theories found in

(11)

11

literature that will be discussed to prove their relevancy. These theories are related to the use of Hohfeld's analysis and transformation between process models. Moreover, the use of each theory will be discussed sufficiently for validity. The framework uses diagrams that will be chosen after considering relevant factors for their use. Finally, two cases will be applied to the framework to test its validity.

Academic and Scientific Relevance

As mentioned in the introduction, this research is a contribution to a methodology proposed by Sileno et al. (2013). The underutilized field of story acquisition by using computational tools is where the methodology by Sileno et al. (2013) comes into place. Their focus on the intentional layer of the agents in the stories provides a different approach to scenario-based modeling. As there is an intention behind every action taken, these intentions are not necessarily revealed before the action is taken. The ability to do an action depends on the agent's perception of his own affordance. Another aspect they argue is the institutional layer which expresses the obligations and powers for the agents that generate intentions. The obligations and powers are expressed by Hohfeld's analysis to identify the required elements for an agent to promise a certain action. A buyer needs to have the power to accept to create the obligation to deliver. They further elaborate on the methodology in (Sileno et al., N.D) where they use different visualization tools. They suggest the use of "Models of concurrent computation" similar to the actor model provided by Hewitt (As cited in Sileno et al., N.D, p. 10) but with the implementation of the intentional and institutional layers.

Singh, Chopra, Desai, & Mallya (2004) provided an approach to process modeling that proposes the use of protocols. Protocols are the specifications related to a certain interaction, the combination of protocols constitute a business process. Singh et al. (2004) proposes that the use of protocols provide flexible and easy to reuse interactions to create a business process. The paper, however, only offers a tool to create protocols for business processes by defining key aspects. One of the key aspects mentioned is commitment. Singh et al. (2004) explain commitment as a representation of contractual relationships; it provides a way to detect the violation of obligations and the consequences of the participant's actions. Singh et al. (2004) also refers to Hohfeld's analysis as the notion which captures the obligations in commitment.

Although this research does not aim to create a computational protocol, Singh et al. (2004) proposition of commitment bolsters the relevancy of creating a model which includes institutional elements for various benefits to process modeling. In fact, many literature suggest the implementation of contractual and legal relationships when creating business processes (Bons, Lee, Wagenaar, & Wrigley, 1995, Mallya & Singh, 2007, Goedertier & Vanthienen, 2006).

(12)

12

Hohfeld in scenario modeling

During the extensive review of literature, little research was found that implements Hohfeld's analysis in scenario modeling. Some of the literature has different goals for implementing the analysis: Siena, Perini, Susi, & Mylopoulos (2009) provide a model for law compliant requirements. With the use of Hohfeld's analysis, the model allows the modeling of requirements for law-regulated domains. Daskalopulu (2000) models legal contracts as processes. With the use of state diagrams, instances of the contract are depicted as states for the actors involved. However, the method focuses on transforming written legal contracts to processes for the portrayal of the status of the agreement.

In the field of artificial intelligence, literature regarding the implementation of deontic logic for intelligent agents by using Hohfeld's analysis was found (Krogh & Herrestad 1999, Singh 1999, McCarty, 1983, Lindahl, 2006, Allen & Saxon, 1995, Jones & Sergot, 1992). Giving the agents normative positions would improve their interaction by recovering and handling mistakes. These methods use programming languages for implementation and the research found aim to create multi-agent models for simulation.

What is different about this model?

Sileno et al. (N.D) aim to use alternative visualizations for their methodology to target non-IT experts. The model presented in this research follows this note by using process modeling tools like UML and BPMN because of their simplicity to both understand and use. The research focuses on transactions to help businesses and legislation create processes that can detect and handle non-compliance by using legal rights. The model allows acquiring possible unconventional scenarios, like sale transactions in a black market or social scenarios without any contracts, with legal positions applied. These unconventional scenarios can be used for creating laws and requirements as well. Although these scenarios are discussed in the artificial intelligence field, the model in this research aims to streamline the ability to acquire scenarios.

(13)

13

Findings

This section discusses the findings in the research and the development of the model.

Theoretical Findings

Hohfeld's Cube

Hohfeld's jural relations define two persons legal relationship at a given moment in time. However, a transaction goes through different phases before it concludes, which results in different jural relations changing in the process. The change in the relation has to be logical and follow a realized formula that explains a connection between all of Hohfeld's relations. Hohfeld's cube places the eight jural relations on a cube which gives them a unified logical structure (Andrews, 1982). It will serve as an important tool for developing the concept of this thesis.

Figure (1): Hohfeld Cube (Andrews, 1982, P. 478)

The cube positions the jural relations carefully to portray the connection between them. The correlatives are drawn at each end of a horizontal line, while the jural opposites are connected diagonally. As figure (1) shows, Right and duty are on the top left corner while Privilege and No-right are on the bottom left corner. Andrews (1982) then mentions the second set of jural relations and how to portray them on the cube. He analyzed Hohfeld's observations to the connection between right and power. Hohfeld mentioned that right implies the existence of power. For this reason it is positioned under power vertically. Liability, on the other hand, is corollary of duty. Hohfeld said "It is a liability to have a duty created" (As cited in Andrews, 1992, P. 476). This line of reasoning claims that power creates a future right which creates a duty, for the other person, from the liability. It also claims that having a right and duty implies that power and liability exist as well. As seen on the cube this is portrayed with Power and liability placed under Right and Duty. Immunity and Disability would logically be placed diagonally from power and liability to portray their opposite nature. Andrews (1982) also further backs this theory up by Hohfeld's analogous relations: right is to privilege as power is to immunity. As previously mentioned, this cube will serve as a tool for the concept to be created in this thesis. It will follow the same line of reasoning to justify the change in jural relations and the connection between them.

(14)

14

Commons interpretation of Hohfeld

In the background section, commons' interpretation of Hohfeld was discussed as it provided a relation between transactional economics and legal relations. Commons' alterations to Hohfeld's schema will not be fully implemented in the model as it tries to stay close to Hohfeld's original interpretation. However, certain elements will be noted as they help in conceptualizing the model. Commons viewed the jural relations as multilateral relations with five actors involved: The buyer, the seller, the alternative buyer, the alternative seller and the state official (Fiorito, 2010). His depiction of an authorized transaction involved Duty, Right, Liberty and Exposure assigned to the relationship between the buyer and the seller. The buyer has a duty to pay while the seller has the correlative right. If the duty is breached then the buyer has the power to make the state official enforce because of the state's correlative liability. This depicted the relationships between the buyer or seller to the state official as Power, Liability, Disability and Immunity. The actor with the right can enforce due to his power over the state official. On the other hand, when the actor has an exposure to the correlative liberty, that actor has a disability to enforce with regard to the state official's immunity. This interpretation is similar to Andrews' (1982) attempt to unify the jural relations by assuming that right draws upon the existence of a power and vice versa. What will be taken from this interpretation is the handling of enforcement. The model will retain Hohfeld's original view of bilateral jural relations but the transaction will assume that duty confirms that there is a power to enforce while privilege or liberty assumes that there is immunity to enforcement. The nature of enforcement will not translate to the inclusion of state officials for enforcement as this will require another transaction between the enforcer and the state official to maintain the bilateral nature of the relation.

Process Modeling and transformations

Information systems consist of process and system models for development as they follow user requirements to develop a working system. There are three levels of abstraction that process modeling follows: Process meta-level (Generic concepts), Process model (Way of working) and Development Runs (What happens) (Roland, 1993). These abstractions are used to generate process models and execute them. There are different tools that are used for process modeling; UML is one of the most famous modeling languages used. UML uses a wide variety of diagrams to model system development that are divided into object-oriented and activity oriented diagrams. However, UML can also be used for business process modeling; the diagrams often used are Activity diagrams and Use case. The combination of both object oriented diagrams and activity-oriented diagrams help in successfully implementing a business information system ( Loos & Allweyer 1998).

Petri Nets is a modeling tool that incorporates graphical and mathematical elements. It offers the ability to simulate processes by tracking the flow of tokens in the diagram. Many literature focused on transforming process modeling tools like UML to Petri nets to evaluate the performance of the models created (Lopez, Merseguer, & Campos, 2004, Kuske, Gogolla, Kollman, & Kreowski, 2002, Ameedeen & Bordbar 2008). Although these transformations make use of programming languages, others have simply defined a direct translation of all the elements available in a modeling tool to Petri nets (Raedts et al., 2007, Trickovié, 2000). Murata (1989) discusses simple rules to enable transitioning to Petri Nets which can vary from computer-aided transitions or informal transitions. The way transitioning works is by

(15)

15

identifying the nodes used in Petri Nets; places represent conditions while transitions represent events. The input and output to and from transitions may vary in numbers which is represented by the tokens. Murata (1989) further explains the possible adaptations of inputs, transitions and outputs. This guide provides an understanding of how the process can be implemented rather than a line to line transition which is crucial for this research because it strives to transition the different models used concurrently to have a coherent story.

The model's conception

A session of trial and error

After concluding the literature research, several attempts of depicting Hohfeld's jural relations came to mind. It seemed like a logical choice to use the jural relations as merely labels to the process in the diagram to define the actors' legal positions. However, this attempt proved to be useless due to the lack of effect on the model. Having the jural relations as merely labels to the process would provide no effect in Petri nets when computing the outcomes of the process.

Discarding this method opened up more attempts on making the jural relations integral to the process and the diagrams. Consequently, an attempt on modeling the jural relations alongside a process in the diagrams was made. This was thought up as having two processes alongside each other. For example, the BPMN or activity diagram would have a process of two actors interacting to reach a goal and a second process following it with the jural relations changing according to certain actions in the first process. This attempt also failed due to the random transitions from jural relation to another without a coherent structure. Moreover, it proved to be unpleasant when transforming them to Petri nets.

Modeling tool Pros Cons Verdict

Use Case - Portrays more than one actor.

- Static in structure - Can't portray positions

- Not valid

- The static structure cannot be fixed

Sequence diagram - Dynamic in structure. - Portrays more than one actor

- Hard to translate - Can't portray positions

- Not valid

- The hard translation can be fixed but the portrayal of positions cannot. Activity diagram - Dynamic in structure

- Portrays more than one actor.

- Easy to translate.

- Can't portray positions - Not valid

- The portrayal of positions cannot be fixed. State diagram - Dynamic in structure

- Able to portray positions - Easy to translate.

- Only one actor per diagram

- Valid

- The portrayal of more than one actor can be fixed.

Table (1) Choosing a diagram for Hohfeld's analysis

The last attempt for the depiction was inspired by the use of different diagramming tools for different purposes. As seen in table (1), different diagrams were considered to portray Hohfeld's jural relations. State diagram was chosen due to certain factors that were in mind initially. Mainly, the diagram that was to be chosen had to be dynamic in structure since it would translate the process' activities from BPMN or Activity diagram. The second factor is

(16)

16

the ability to portray the position of the actor at a given moment in time to show the jural relation at hand. The third factor is the inclusion of two actors to show the correlatives of the jural relations. State diagrams met the first two factors, but the third factor was not met. However, using association lines between two state diagrams provided a solution.

When applying Hohfeld's analysis to the diagram, the process always had to consider the possibility of non-compliance. For example, If there is right(claim) and duty then the model has to portray the handling of not fulfilling this duty. Also, Sileno et al. (N.D (B)) discuss the three areas of action in a transaction: direct action, reception and evaluation. These areas are important to implement in the diagram to consider the interaction between the actors from a different perspective. The actors interact by noticing each other's powers, then they comply to certain duties and then they evaluate whether an enforcement is needed. This layer was added to the methodology as well.

Another aspect that needed to be considered was the tracking of the actor's rights before and after the transaction. Consequently, another layer for the initial rights of the actors was added. With these four layers and the fifth layer being the Petri nets, the conceptualization of the framework was concluded.

The layers of abstraction

For the purpose of applying Hohfeld's rights to process modeling; five layers of abstractions are drawn up as seen in table (2).

Layers Modeling tool Description

The first layer State diagram Portrays the exchange of rights between actors regarding the object, The second layer State diagram Portrays the three phases of the

transaction

The third layer BPMN or Activity diagram Portrays the transaction process. The fourth layer State diagram Portrays Hohfeld's analysis in the

transaction process.

The fifth layer Petri Nets Portrays the whole process with all the layers.

Table (2) The layers of abstraction.

The first four layers of abstraction are created in tandem as they are connected in theory but this connection is only shown visually in the fifth layer. Different diagrams will be used for the layers: State diagrams and BPMN. State diagrams will be used for its portrayal of states as the jural relations and phases of the transaction, while BPMN will showcase the flow of the process. BPMN is not the only diagram that can be used; activity diagrams or sequence diagram can be used in that same layer. The reason different diagrams will be used in the layers is the different aspects each diagram provides with respect to the layer. The fifth layer is the combination of all the above layers in Petri nets to showcase the connection and be able to compute the possible outcomes of the process/story.

The methodology is a tool to help acquire the possible outcomes for one process or story. The five layers can be considered a pyramid scheme where it elaborates on more detail the higher it goes. To explain the methodology, meta-models will be presented for each layer. However, certain elements will be discussed, before the meta-model is presented, in the following chapters that are used in the layers.

(17)

17

Exclusive and Concurrent Rights1

Generally, transactions show an exchange between people. This exchange can be an object, action, service etc. To avoid confusion it shall be called an "entity". For example, if the entity is something one of the actors owns, then we have to consider the rights of owning something. According to Hohfeld's description of ownership, it is multital rights therefore it is considered a bundle of exclusive rights. It is labeled as exclusive because one person owns that entity. The act of exchanging the entity transfers the exclusive rights of the owner to the other actor in the transaction.

However, the transaction may not simply be about exchanging the entity as a whole, but certain rights towards it (I.e renting or selling a service, doing an action). In this case, the owner wouldn’t exchange all his exclusive rights but some of them, which would create concurrent rights between both actors. This theory is derived from Cook's (1918) mention of common law and equity. "Exclusive" rights are exclusive to the laws of equity, where they have the upper hand. "Concurrent" rights are concurrent between common law and equity. The owner of the entity may be considered equity and the other actor common law. The owner would still have exclusive rights that can overcome the rights given to the other actor. For example, A rents a house to B, this gives B certain rights like living in the house but he does not have the right to sell it, that is still exclusive to A. B may rent it to someone else without A knowing but if A finds out then he can invoke his rights. Renting or selling a service would change the exclusive rights to concurrent rights, not concurrent in the sense that they are the same rights but different rights regarding the object.

Three phases of action2

A transaction process has different phases of actions. Sileno et al. (N.D (B)) focused on a sale transaction when defining the areas of action it undergoes. These areas of action will be used in the second layer but with a few changes to apply to the methodology at hand. The three phases in the transaction are Recognition, Execution and diagnosis. Recognition is when the two actors come into contact but it may not be under both actors consensus opinion. For this reason it is considered as both actors recognizing each other's existence and position in the transaction. Execution is the consequent conclusion of the first phase; it is the execution of each actor's expected performance. Lastly, diagnosis is where non-compliance is detected depending on the previous phase's conclusion. According to the diagnosis the transaction either successfully ends or returns back to the recognition phase for enforcement. This focus on the conclusions of each phase to proceed with the process is inspired by Breuker's (1994) problem solving technique.

Institutional vs. brute acts

The fourth layer of abstraction is the layer where Hohfeld's rights are applied to the transaction. It is considered a transformation of the diagram created in the third layer. In the third layer each actor performs activities for the transaction to proceed. These activities are considered brute acts, as in offering or paying, that create institutional acts, which are Hohfeld's jural relations.

For example, if the transaction was about buying from a supermarket: The customer would gather the items to buy, go to the cashier to calculate the sum, then pay and the cashier would

1_{This theory will help implement the first layer.} 2_{This theory will help implement the second layer.}

(18)

18

give him the items. These activities are considered brute acts that can be modeled in the third layer and it will be translated as institutional acts for the fourth layer as follows: The customer has the power to gather the items he needs which puts the cashier at a liability to calculate everything for him then the customer has the duty to pay to receive the items which is correlative to the cashier's claim to withhold the items until he pays.

Unifying Hohfeld's jural relations

Hohfeld's analysis is used to assess legal cases under a discussion in court. The privilege of discussing how they are used in each story is not possible for this framework. For this reason, a theory of what the jural legal relations represent, as a general outline to all the stories to be modeled, is possible to create. The theory's ambition, however, is not to set in stone the fundamental use of Hohfeld's analysis for this methodology. It is merely to suggest a logical representation of them that can be updated the more it is used. This diagram will be presented in the fourth layer.

The diagram is loosely based on Hohfeld's cube (Andrews, 1982), which positions all eight legal relations on the cube to resemble a unified progression. With that in mind, the diagram here aims to portray the unified structure of the jural relations with regard to transactions. In Hohfeld's cube, the relation between power and right(claim), liability and duty, are explained as derivatives of each other. A right implies the existence of a power and a liability creates a future duty. The start of a transaction is when someone has a certain power he would practice, thus creating a liability to the other person in the transaction. That power would result in imposing a task on the other person for the transaction to proceed which would create a

right(claim) for the one who imposed the task and a duty for the other.

This line of reasoning is derived from the nature of transactions. A transaction means an exchange between people. This exchange can happen if someone has the power to do so and another one who is at a liability to this exchange. Consequently, actions need to be done for the exchange to happen either imposed by the actors or the expected behavior in society for something like a sale (Paying/delivering). These actions are considered the rights and duties of Hohfeld's jural relations. However, Non-compliance is sometimes expected in a transaction, mostly in the duties given to each actor. A person has a duty while the other has the right(claim) over this duty. If it hasn't been fulfilled, it is assumed that the person with the

claim would gain the power to enforce a punishment. This can be derived from Hohfeld's

reasoning; a right implies the existence of a power (Hohfeld, 1913). Moreover, according to commons privilege results in immunity to enforcement (Fiorito, 2010). Since the other party has no-right which results in disability to enforce3.

The actors in this meta-diagram do not represent strictly both actors in the transaction but it follows Andrews' (1982) representation of the cube as the A side and the B side. The A side represents the person with the power, right, no-right, or disability. While the B side represents the person with the liability, duty, privilege or immunity. (Andrews, 1982) follows Corbin's definition of A and B in Hohfeld's jural relations (As cited in Andrews, 1982, P. 474): where A is the person that holds the jural position in question and B the correlative. Using A and B in this diagram denotes that either actors can be considered A and B in the same diagram according to the activities.

(19)

19

The Meta-model

In this section the meta-model will be presented of each layer of abstraction with a concluding note.

The First layer

Figure (2) Meta-model first layer.

The diagram used in figure (2) is a state diagram that portrays the first layer. State diagrams can only portray one process at a time and this depiction provides two diagrams that are associated with each other through the association lines (dotted lines). As a meta-model, the states are not defined because they are different with every case presented. The actors involved in this layer are the two actors who represent the transaction (A and B) and a third actor representing the rest of the world (C). The possible states for this layer are: Exclusive

rights, Duties and Concurrent rights. The initial state of the actors can be any of the three

states but the states will be correlative to each other. Exclusive rights must exist with duties while concurrent rights must exist with another concurrent rights. The correlatives are to be represented with the association lines portrayed in the above diagram. The transition between the states will be caused by events. The event will transition between the states for different reasons; following the transaction or repairing failure in the transaction to return the actors to the initial states. These events are the transaction to be modeled in the proceeding layers. The last event will be the event ending the process; it can be an event which the transaction was successful or disability to enforce punishment for non-compliance. After modeling the first layer, the position of each actor and the world with regard to the entity is portrayed before and after the transaction. The proceeding step is to model the transaction.

(20)

20

The second layer

The second layer's role in the methodology is to represent a sense of progress and goal in the transaction process. Each phase has different conclusions and goals to continue. Portioning out the transaction will help analyze and solve problems that occur in any phase according to its goal. The goal of the recognition phase is for both actors to recognize each other's position and interact in the process, it also portrays the progress of the transaction which is still in the first phase where there is still no action taken. The goal of the execution phase is that both actors are supposed to take the expected action that derived from the previous phase. It is considered the middle of the transaction. The goal for the last phase, diagnosis, is to monitor the actions taken and conclude it

Figure (3) Meta-model second layer.

The diagram in figure (3) shows the second layer using state diagram. The states are the phases and the events are the conclusions. The initial phase is recognition, after both actors

recognize the state transitions to execution. The execution phase's transition is execute where

both actors execute their duties. Execute can also mean that they failed to fulfill their duty in the time given, therefore it proceeds to diagnosis. There are two events that would transition the Diagnosis phase: Success or Failure. Failure would transition it back to the recognition phase but this time for the enforcement while success would end the process.

As mentioned earlier, the second layer is to portray progress and goals in the transaction process. The third layer in the methodology is where the transaction is modeled with BPMN. The second layer serves as an extension to the third layer to label the phases on the diagram.

The third layer

In the third layer, the transaction is portrayed using activity-oriented models like BPMN and activity diagram. The three phases of the second layer are labeled to define the goals of the activities. The meta-model presented here showcases how the phases will be added as an extension. Similar to the first layer, the activities will not be defined in the meta-model.

(21)

21

Figure (4) Meta-model third layer.

As seen in figure (4), the phases are applied with dotted squares to group the activities done. They follow the phases' goals. The activities in the first phase will relate to the actors initial interaction with each other and they either ask or notify what should be done for the following phase. The activities in the second phase are the actions done by each actor, and the last phase portrays the diagnosis of the action and the activities that will portray the handling of non-compliance.

The fourth layer

Figure (5) Meta-model fourth layer

The meta-model presented in figure (5) is modeled with state diagrams, where the states are Hohfeld's rights and the events are the activities from the third layer (or Brute acts). The

(22)

22

model has two systems shown, one for each actor, which are correlated through association lines (dotted lines) to portray the correlative jural relations.

As previously mentioned, the events portray the activities (or brute acts). The event of

imposing a task would create a right for actor A and the event perform task would create a duty for Actor B. If the transaction is successful, then the transaction along with the system

ends with the event successful, otherwise the event failed to fulfill occurs which would transition the states back to power and liability but in this case to enforce. The other states that can occur with the event perform task are No-right and Privilege because in certain cases the actors may not be susceptible to duties. Let's assume that the transaction takes place in a black market where paying or delivering is not backed up by a claim because going to court is not an option. In this context power to offer and accept transition the actors to

privilege(liberty) to pay and deliver while they also have no right(claim) to the

aforementioned action. If the transaction is successful then it wouldn't be of importance, but in case of non-compliance the following transition would be immunity and disability. Since the act of paying, for example, is a privilege then that act is immune to enforcement if it is not fulfilled. Consequently, the person with a no-right to the payment has a disability to enforce.

The fifth layer

The fifth layer is the last layer in the methodology where the first four layers of abstraction are transformed to Petri Nets. The goal for this transformation is to utilize Petri nets for calculating the possible outcomes to the story. This is considered a preliminary implementation and it requires testing and investigations to improve it.

Using Murata's (1989) theory, state diagrams will be transformed by having states as places and transitions as Petri Nets transitions. When a state has two outputs it is considered as a

choice in Petri Nets and will be translated as two possible transitions.

The transformation of the first and second layer's state diagrams into Petri Nets is shown in figure (6)4. In the first layer, the places that belong to a certain actor are defined in the name

4_{The transformation of the first and second layer here is only for the purpose of explaining how the} diagrams are transformed. They do not need to be transformed standalone for the model.

(23)

23

of the place A: State 1. The transitions after the places A: State 2 and B: State 2 portray the two transitions in the state diagram as Second Event, which revert the actors to their initial states, and Last Event, which ends the process. The second layer is portrayed simply with the phases as places.

When combining and transforming the four layers in Petri Nets, there has to be a consideration of how the story unfolds. As mentioned before, the layers are like a pyramid scheme where they are modeled sequentially. The model will start and end with the first layer as it portrays the holistic view of the transaction. Moreover, the third and fourth layers are similar in structure but divert in focus; the third stresses on the activities while the fourth on Hohfeld's analysis. Consequently, transforming them to Petri Nets will require one depiction. After combining the layers, there has to be a synchronization of time between the layers. For example, the recognition phase has to transition to execution before the actors receive their duties and claims. This will be implemented by using inhibitor arcs to put conditions for certain transitions. There has to be no token in recognition for the transitions after A:Claim and B: Duty. This implementation will be shown in the appendix for the cases.

Figure (7) Meta-model fifth layer

The Petri Nets in figure (7) showcases all the layers together as one story. It starts with the first layer by defining the places of each actor and the Event that transitions the places opens up the other layers. The first layer remains on the outer edges of the diagram, parallel to the process ongoing inside5. The second layer starts with the third and fourth layers, proceeding according to their progress and ends when they conclude. The third and fourth layers are the core of the diagram as they hold the inner transaction which affects the first layer according to the outcome. If the actors end the transaction successfully then the process ends with the first layer's last event, if they fail then the diagram focuses on the first layer again to portray the changes with the transition Second event and then concludes.

The method of transforming the layers to Petri Nets is as follow: The first layer should be transformed with the states as places and the state transitions as Petri Net transitions. The second, third and fourth layer should be transformed together. The third and fourth layer should have one depiction of the process with the places as Hohfeld's jural relations (Institutional acts) portrayed in the state diagram and the transitions as the activities (Brute

(24)

24

acts) portrayed in the BPMN. The second layer should be depicted with states as places and events as transitions. Now with all the layers portrayed the main focus should be on the connection between them. The first layer connects to all the other layers with the event that transitions the initial state. As mentioned before, the event that transitions the states in the first layer is in reality the transaction which occurs in the following layers. Therefore, this transition should connect to the other layers in Petri Nets. The tokens should start with the first layer then moves along to the other layers while maintaining an idle state until a certain place is reached in the other layers. The tokens should be in the recognition phase and the start of the transaction process in the third and fourth layer. When the actors recognize the powers, the second layer's token moves to the execution phase. Consequently the first layer's tokens move to the second states. When the actors finish the execution phase, the second layer's token to the diagnosis phase while the first layer remains idle. According to the third and fourth layer's activities, the diagnosis is either successful transaction or failure. If it is successful then the first layer ends the diagram. If it failed, then non-compliance is portrayed in the third and fourth layer and then the first layer portrays the punishment then ends the process.

The fifth layer then portrays a coherent story for the model that connects all the layers together. This connection allows the choices and progression of the story in any layer to cause a ripple effect that translates to the other layers.

Meta-model conclusions

With the use of the meta-models, the concept of the methodology was explained. The layers of abstractions put into account the important aspects regarding scenarios of transactions that are influenced by them. The four layers are considered the building blocks for the fifth layer to create a coherent process for Petri Nets to be executed for the different outcomes. Moreover, the literature discussed backs up the theories speculated in this research for validation. The following chapter will present two cases that will be applied to the model to show it in practice and prove validity.

(25)

25

Testing the model

To test the model created, two cases will be applied. The first case will entail a simple sale transaction and the second will portray traffic rules. The reason for choosing those specific cases is because they display different variations of transactions, the first is a real transaction while the other will have an interaction between two drivers as a metaphor for a transaction. The cases will also portray different elements from the model that wouldn't be possible with one case.

How the model is used

The model is divided into four layers of abstractions, with each layer more elaborate than the one before it. The layers should be modeled from one to five sequentially to coherently develop the story. The first layer should model the big picture of the transaction, showing the before and after effect of it. However, the second layer is consistent in every story with the only adaptation it should receive would be in the terms used. The third layer should be modeled with the second layer's phases in mind while the fourth should adapt the third layer faithfully. The creation of the fifth layer should aim to link all the layers together in its adaptation to have a coherent structure.

Case 1: A sale transaction

As mentioned above, the first case is a simple process of a sale transaction akin to the case used by Sileno et al. (2013). It is important to choose a generic case where certain information is not defined to show the effect of the methodology on it. The transaction process will be between two persons: A seller and a buyer. It starts when the seller offers a good then the buyer evaluates it then accepts. This binds them to the next step of paying and delivering. Moreover, a diagnosis of failure in payment or delivery that is backed up by enforcement is done. Without adding more information, the process can have many outcomes: failure in either duties, failure to enforce etc.

The first layer

When developing the first layer for the sale transaction, an important detail was added. There are four actors in this layer: The seller, the buyer, and alternative seller and buyer. Usually, in any sale there is always an alternative option for both the buyer and the seller which puts pressure on the sale by expecting that there is a better option available6. The alternative actors will not play a role in the transaction but their existence is still considered important to portray Hohfeld's multital rights towards the whole world.

6_{Commons talked about the five actors of a transaction being the buyer, seller, alternative buyer,} alternative seller and the state (Fiorito, 2010).

(26)

26

Figure (8) Case 1 first layer

According to Figure (8), the first actor in the diagram is the seller, who is the one giving up his rights to the buyer. The initial state of the seller is Exclusive rights to the object/service, the final state is Duties. The Exchange is considered the event that will transition the states which is the transaction portrayed in the third layer. The second actor is the buyer whose initial state is Duties, which is correlative to the seller's Exclusive rights as portrayed by the association line in the diagram. In the final state, there are two events that may occur, which handle non-compliance7. The diagram shows the ability to Repair failure if it is possible, but in case of punishment or remedy it is out of scope so the process ends. It is considered out of scope because the first layer observes the rights with regard to the object so only repair would be relevant to it while remedy and punish are considered another process with another narrative. Destroy is the chosen term for this event, not necessarily to point out that the item is destroyed but to represent an irredeemable enforcement with regards to the object.

The second layer

The second layer in the case remains unchanged to the depiction provided in the meta-model. As mentioned before, the second layer is a general model of the phases which can be adapted as is with only the terms used altered if needed.

7_{Generally, there are 3 ways of handling non-compliance with regards to breaching a duty: Punish,} remedy or repair. Punishment is considered to be unbeneficial to the enforcer while only punishing the guilty (I.e going to prison). Remedy allows the enforcer to receive some sort of compensation either from the guilty or equity. Repair, on the other hand, allows the enforcer to regain what he lost before the occurrence.

(27)

27 The third layer

Proceeding with the third layer, the flow of the sale transaction will be portrayed using BPMN and it will incorporate the second layer's phases.

Figure (9) Case 1 third layer

As shown in figure (9), the diagram follows basic actions to proceed with the process. The important notice is the recognition of actions as seen in the Recognition phase. An offer needs to be recognized for it to be valid, same goes for the acceptance. However, the parallel flow between pay and recognize acceptance provides a generic outlook to the transition between the phases. In some cases, the buyer can proceed with the payment before or after the seller recognizes the acceptance, this can happen in an online shop for example. The recognition in the process is what makes the actors proceed with the process. In the execution phase, the recognition of payment or delivery will result in the action in the diagnosis phase. If delivery is recognized then it is successful, otherwise the actor enforces.

The fourth layer

The fourth layer portrays Hohfeld's analysis of the sale transaction for each actor. The association lines referring to the correlatives are missing in this depiction to make the diagram clearer and easier on the eye but all the states still have their respective correlative.

(28)

28

Figure (10) Case 1 fourth layer

As shown in figure (10), the initial state shows the seller's power and the buyer's liability. The seller offers the object thus transitioning into a liability which gives the buyer the power to accept. Both of those actions are recognized by each actor. After the buyer accepts the two diagrams have parallel flows with the buyer paying and the seller delivering which transitions them into duty and claim. If each actor fulfills his duty and recognizes the other's then the transaction is successful. Otherwise, the claims transition to power with the event Payment

failure or Delivery failure, which puts the one with the duty to liability and a scenario of claim and duty for enforcement occurs.

The fifth layer

Figure (11) Case 1 fifth layer. S stands for Seller and B stands for Buyer.

With all the layers combined, the fifth layer portrays the sale transaction in Petri Nets. The transformation follows the method explained in the meta-model. The first layer is transformed first and then it breaks up into the other layers with the transitioning event. In this case it is

(29)

29

called Exchange. Moreover, the other layers are portrayed concurrently to in the Petri Net with the transaction process connected to the second layer to portray the phase the transaction is at. According to the activities of the transaction, the phases change then the first layer's places change.

As mentioned in the meta-model, the conditions for certain transitions are implemented by using inhibitor arcs. The diagram in figure (11) hides the inhibitor arcs because it over complicates the model and renders it almost unreadable. A figure of the model with the inhibitor arcs will be in the appendix. Moreover, in the failure and enforcement scenarios some tokens remain when the process ends. These tokens are flushed out by using biflow transitions before repair and destroyed. This is also left out to keep the diagram simple and it will be shown in the Appendix as well.

Case 2: Traffic rules

The second case applied to the model is a traffic rules scenario. The scenario portrays the driver's rights as he is driving forward in a street. The scenario assumes that there are no traffic lights and the driver has to abide by the rules of the road to yield and make way for any car approaching from the right. This scenario puts the driver and any car approaching from the right in a transaction where they recognize each other's existence and the driver makes way. The scenario does not include the driver turning and only focuses on him moving forward.

The first layer

As seen in figure (12), the two actors in this layer are the driver and the traffic, which entails all the cars on the road. The initial states of both actors are Concurrent rights since both have rights of way as long as they do not meet anyone coming from the right. The event, Car

approaching from the right, transitions the driver's state to Duties to make way for the

approaching car, while the traffic transitions with the correlated event Car approaching from

the left to Exclusive rights of way. There are two possible transitions from the second state,

either the car from the right passes and they return to their initial states or something goes wrong and they crash.

(30)

30 The second layer

Similar to the previous case, the second layer remains unchanged.

The third layer

As mentioned before, the transaction in the case will be considered the interaction between the driver and every car approaching from the right. The third layer portrays this transaction using BPMN with the two actors as the Driver and Car on the right.

Figure (13) case 2 third layer

In figure (13), the start of the process is the activity Drive done by the driver which is then recognized by a car on the right. Proceeding with the activity approaching, the driver recognizes it and two parallel activities are created thus ending the recognition phase. Pass and Yield, activities performed by the car on the right and the driver respectively, entail the execution phase. The Yield activity is a choice where the driver yields then ends the process or does not and then crashes as shown with the activity Crash. The activity Pass for the car concludes the actor's activities since it assumes that it has the right of way.

(31)

31

Figure (14) Case 2 fourth layer

Portraying Hohfeld's analysis in figure (14) starts with the driver's event Drive and the car

Recognize which initiates the states Liability and Power respectively. The transaction entails

the driver's interaction with cars on the right, therefore when he drives he is liable to all cars on the right and the cars on the right recognize. The transition from Power, for the car, is

Approaching for the state Claim and correlative to that power is the state Duty for the driver

which is transitioned with the event recognize. As translated from the third layer, the outcome for the Duty is either Yield or Crash for the driver and Pass or Crash for the car on the right. The fifth layer

Figure (14) Case 2 fifth layer. COR stands for Car on the right.

Concluding the case, the fifth layer shows the Petri Nets created for the case with each layer depicted concurrently to achieve a coherent structure of the flow of the story. The inhibitor arcs will be presented in the appendix.

(32)

32

Conclusions

The research's focus is to add Hohfeld's analysis to process modeling to enhance the possibilities of scenarios that do not include the happy-flow. The research question posed "How can process modeling depict Hohfeldian analysis and be transformed to Petri Nets?" was met with a framework. The layers of abstraction provided building blocks to create the process with the inclusion of important aspects that coherently ties them together. The use of different diagramming tools and theories derived from the literature helped define and distinguish the importance of each layer. The final layer of abstraction transforms the diagrams into one story using Petri Nets that realizes the model's structure.

To test the model, two cases were applied to showcase the model in practice. The cases were different in structure to validate the use of the model in different contexts. The sale transaction provides a generic sale where the two actors interact for the intention of the exchange while the traffic rules case provides a scenario where the actors interact unwillingly. Although future work is needed to further validate the model and enhance its features, the model as it stands provides a core theoretical concept that implements Hohfeld's analysis with the use of process modeling and can provide acquisition of stories through Petri Nets.

Discussion

One of the model's limitations is the lack of automatic transformation of the first four layers to the fifth layer. Transforming the layers to Petri Nets automatically is not an impossible task but it requires the use of programming languages to be done which is out of scope of this research. Another limitation is the perspective of how Hohfeld's analysis is unified, it can be argued whether this is a faithful adaptation of it due to the different interpretations of the analysis in general. Moreover, the complicated nature of Hohfeld's analysis may pose as an obstacle for using such a model in many fields as it requires an understanding of case law. With these limitations in mind, the recommended future work for this research is to further enhance and add different elements that will expand the model to more than just transaction processes. Moreover, the transformation of Petri Nets should be investigated to provide alternative ways to portray the time and conditions in the diagram.

Another domain to consider is to implement automatic transformations for the models. It can be implemented to a user-friendly software that helps create the model while explaining the methods behind it. For example, the software can help automatically create the correlative jural relations for the user to understand how it works.

The model can be used as a tool to create new processes, but it can also be implemented using existing processes. The depiction of Hohfeld's analysis provides a great incentive for legislative processes to be created with the model by implementing the legislative rules behind them. Business processes can also be depicted in the diagram with the use of the third layer as it is a basic flowchart diagram that will build the other layers around it. Finally, with the use of famous and simple diagramming tools, the framework is easy to grasp by non-IT experts.

APPLYING HOHFELDIAN ANALYSIS TO PROCESS MODELING