Towards Automation of the Swedish Property Formation Process: A Structural and Logical Analysis of Property Subdivision

The ongoing digitalization of public administration and increased automation of legal decision-making bears promise to benefit citizens, businesses and other stakeholders through simpler and more efficient civil processes, and thus has great impact on the urban planning and building process. However, automation of decision-making that is directed or constrained by normative systems such as laws, regulations and policies, requires a detailed and accurate representation of these concepts and their constituent parts, and the domain to which they are applied. This paper combines two perspectives on formalisation and classification of legal relations within the urban planning and building domain. In a crossdisciplinary fashion, the paper analyses and describes a small part of this domain at a higher level of abstraction and formalization using two different analysis instruments. Using these tools, we perform structural and conceptual as well as logical analyses of two specific snapshots of a fictitious property subdivision case in Sweden, focusing on the legal relations between different entities and parties involved in the specific situations. The structural analysis uses the Land Administration Domain Model ISO 19152:2012 standard formalism, and the logical analysis is based on the notion of atomic types of legal relations. The paper discusses some of the strengths and weaknesses of the two tools regarding the formal representation of rights, restrictions and responsibilities of different parties in the land administration domain, as well as how the tools relate to each other and how they can be aligned. This paper takes one step towards a deeper understanding of the domain, and identify areas for future research that may provide better conditions for efficient and transparent use of geospatial information, and automation of the property subdivision process and other related civil processes.

of different parties in different situations that follow from them) closer to each other. A suitable method for this is to perform a number of interrelated case studies of selected subprocesses, including the property formation process. Within such a 'process case study' it is possible to single out a number of process 'snapshots' (i.e. specific situations in the studied process) to be structurally and logically analysed. For such a process snapshot, a 'situation case study' may be performed, in which the types of entities and parties involved in the specific situation, as well as their relationships and their legal positions, are analysed and formally described. The aim of this paper is to take a first step towards developing this theoretical and conceptual framework, by performing a situation case study within the property subdivision process. The point of departure is a specific property subdivision case, in which the types of entities and parties involved in the specific situation, as well as their relationships and their legal positions, are analysed and formally described. The formalism offered by the Land Administration Domain Model (LADM) ISO 19152:2012 standard (see Section 2.3 in this paper), as well as the logic-based theory of normative positions (see Section 2.4 and 2.5) are utilized for this. In a cross-disciplinary fashion, the paper thereby combines two different perspectives on the formalisation and classification of such legal positions.
Joining and aligning analysis tools and formalisms from two different research areas, the paper aims to gain insight on the strengths and weaknesses of the two tools regarding the formal representation of rights, restrictions and responsibilities of different parties in the case at hand. The attempt is to outline the further work needed to gain a better understanding of the land administration domain (including how to analyse and describe it at a higher level of abstraction and formalization using partly new analysis tools). The long-term goal is to contribute to efficient and transparent use of geospatial information in, and increased automation of, currently manual civil processes that will benefit many actors within the urban planning and building process.
Cadastral (and other) authorities can make decisions manually by one or more decision-making officers or by automated procedures (FL, 2017, 28 §). It is outside the scope of this paper to describe how the motives for the decisions made in the investigated cadastral processes are documented or archived by the authorities. Furthermore, research concerning the emergence and importance of real property rights due to social, political and economic factors has been conducted through several decades (see e.g. Ekbäck, 2009;Libecap, 1989 andUmbeck, 1981) and will not be investigated in this paper. This paper is structured as follows. Section 2.1 discusses the notions of (land use) rights and real property ownership, and Section 2.2 discusses property ownership as an intermediate concept. Sections 2.3-2.5 present the theoretical frameworks and formalisms that will be applied for describing and analysing rights, restrictions and responsibilities in a property subdivision case. A specific property subdivision case is presented and structurally and logically analysed in Section 3. Section 4 discusses the analysis and its implications for automation of the property subdivision process and identify topics for future work, and Section 5 concludes the results.

Theoretical framework
Land is a limited resource and has to be administered in order to regulate the various private and public interests of individuals, companies or the State. The way land is administered includes decisions on access to land, land rights, land use and land development. Human activities (housing, farming, husbandry, forestry, recreation etc.) presuppose certain rights. These rights are defined according to law or custom. This section discusses the notion of (land use) rights and presents some theoretical frameworks and formalisms for describing and analysing rights.

Land use rights and real property ownership
Real property legislation, and in a broader context land use legislation, is concerned with regulating what may be done with land. "Real" in real property usually is associated with something solid, fixed and permanent, which has to do with land. Land use rights are links between the legal owner(s), the right, restriction or responsibility and the area(s) of land in question, and thus have at least three dimensions: what is included in a right, who is the holder of this right and the physical extension this right has (Larsson, 2010). A recent discussion of the real property concept and the relation owner, right and property, as well as its relation to digital processes, can be found in Ekbäck (2019).
There is no universally accepted definition of the term right, but a right has been given a number of rather similar definitions, such as an "action, activity or class of actions that a system participant may perform on or using an associated resource" (ISO, 2007, Section 4.38), "a claim or title to or an interest in anything that is enforceable by law" (Gifis, 1984, p. 416), and "[a]right to a specific property, whether tangible or intangible" (Garner, 1891(Garner, , p. 1096. A restriction has been defined as a formal or informal obligation to refrain from doing something (ISO, 2012, Section 4.1.19), "[...] a limitation [...] placed on the use or enjoyment of property" (Garner, 1891(Garner, , p. 1089. 2 A responsibility has e.g. been defined as a "formal or informal obligation to do something" (ISO, 2012, Section 4.1.18), "a liability" (Garner, 1891(Garner, , p. 1087, and "an obligation" (Gifis, 1984, p. 408). 3 As previously mentioned, land use rights often follow from (or is interconnected to) real property ownership. Thus, real property ownership is a central notion concerning the relation between person and land. The authors are not aware of a commonly accepted definition of 'ownership', but it has been argued that ownership can be described as the greatest possible interest in a thing which a mature system of law recognizes (Honoré, 1987). A common approach is to regard real property ownership as a right of its own, 'ownership right', which in turn is a combination of several rights. Together in a 'bundle' these rights form the concept of real property ownership: The right of unlimited possession of the property; https://doi.org/10.30672/njsr.78170 the right to use the property; the right to manage the property and exclude others from the property; the right to added value of the property; the right to transfer the property to somebody else according the the owner's choice (see e.g. Honoré (1987), Snare (1972), Bergström (1956), andHohfeld (1917;1913)). The concept is illustrated in Figure 1.

Real property ownership as an intermediate concept
Another way to treat the notion of real property ownership is to regard it as a socalled intermediate (legal) concept, also known in the literature under names such as 'intermediaries', 'ground-consequence-terms', 'middle terms' or 'coupling terms'. 4 According to Lindahl (1985), it is an old observation that a number of legal concepts, e.g. contract, ownership and tenure, are linked both to certain legal facts and to certain legal consequences, and thus can be regarded as syntactic tools for formulating legal rules and 'vehicles of inference' for legal reasoning. In fact, ownership is a classic example of an intermediate concept, whose function (and, thus, meaning) is tied to its role as a vehicle of inference linking factual grounds for ownership with legal consequences of ownership. In this view, the term being the owner of functions as a bridge or transition between different conceptual systems, one containing facts 5 (e.g. events, actions, or circumstances) and one containing normative positions 6 like obligations, claims, legal powers, etc. According to this view, ownership is attached to certain facts, and different normative positions are attached to ownership. In other words, the term ownership (like other intermediate concepts) connects legal information of two different sorts, factual (descriptive) and normative, and is in itself neither a purely descriptive nor a purely normative concept. This idea is illustrated by the scheme shown in Figure 2 where O denotes ownership, G 1 , ... G p denote factual (legal) grounds for ownership and C 1 , ..., C n denote legal consequences of ownership. Each G iS may represent circumstances 4 The term intermediate concept, for concepts that lie conceptually in between purely descriptive and purely normative concepts, derives from the discussion between Scandinavian legal philosophers Ekelöf, Ross, Wedberg and others. An overview of this discourse is given in, for example, Lindahl and Odelstad (2013, Section 1.7).

Ownership
Right to use, manage and exclude Right to added value Right to transfer Land (Object) executes in that hold, or events or actions that take place, in a particular situation. G i often takes the form of a condition on a number of agents 7 , such as a binary condition g i (x,y) involving two agents x and y. For example, if g i represents the condition having inherited from, then g i u (x,y) may be read as x has inherited property unit u from y.
In Figure 2, G 1 , G 2 , … , G p express different legal grounds for x being the owner of (e.g.) a property unit u, and C 1 , C 2 , … , C n express different consequences of x being owner of u (cf. Odelstad, 2017, p. 34). Examples of legal grounds for ownership of u may be having lawfully purchased u or having inherited u. Some possible legal consequences of ownership of u are mentioned above; e.g. having the right of unlimited possession of u and having the right to transfer (ownership of) u to somebody else. It thus seems that the view of ownership as a 'bundle of rights' mainly focuses on the normative side of the concept.
According to the scheme, it holds for all i, 1≤ i ≤ p, and all j, 1 ≤ j ≤ n, that G i implies O O implies C j Thus the communicative function of O is to link the grounds G 1 , ... G p to the consequences C 1 , ..., C n . This syntactical tool offers economy of expression since it only requires p+n implications instead of p*n compared to formulating the rules by attaching each G i to each C j (Lindahl & Odelstad, 2013, p. 231). It is not uncommon that intermediate concepts form chains, so that what constitutes a consequence of a certain concept in turn constitutes a ground for another concept. For example, as illustrated in Figure 3, the condition having inherited u has certain legal grounds, such as being the sole heir to u and being the heir to u according to will, and certain consequences, such as being the owner of u. Thus, having inherited u is a ground for being the owner of u, which in turn has various legal consequences. Inheritance of property u and ownership of property u thus form part of a chain or network of intermediate concepts. 7 Here, agent is used as a generic term for various legal parties such as persons, groups, organisations, or other entities capable of action. The term actor will mainly be used to indicate an agent that is in some sense 'active' in a particular scenario.

Figure 2.
A schematic view of ownership (O) as an intermediate concept, linking its grounds G 1 , G 2 , … , G p to its consequences C 1 , C 2 , … , C n (Lindahl & Odelstad, 2013, p. 553 Lindahl and Odelstad (2013, Section 1.7.5). Regarding real property ownership, it seems reasonable to believe that in most legal systems and traditions, both its grounds and its consequences are fairly well specified, but the exact extent to which real property ownership is a ground-open and/or consequence-open intermediate concept might vary from one legal system to another. A more interesting example is the condition being enduringly suited to its purpose (see Section 1.1) which Section 4 will discuss further. In a decision process, a ground-open intermediate concept is of special significance, since it functions as a 'point of decision' where it must be decided if the grounds of the concept are fulfilled, and thus its normative consequence applies.

The Land Administration Domain Model
The Land Administration Domain Model, LADM, ISO 19152:2012(ISO, 2012, is an international standard and a tool for structuring land administration. It is not limited to any legal system or tradition, thus possible to use as a reference model regardless of a nation's legal system. The LADM is developed by the International Organization for Standardization, ISO, and even accepted as a national standard by a number of countries and as a European standard. The purposes of the model are several; to be used as a conceptual basis for development and/or maintenance of effective and efficient land administration systems and to enable communication and transfer of real property and land administration terms based on a shared  Figure 23 in Lindahl & Odelstad, 2013). As in Figure 2, the intermediate concept is placed to the right of its grounds and to the left of its consequences.
vocabulary. In LADM, RRRs are seen as general relations between a (legal) person and land. The LADM does not focus on technical implementation of IT systems for land management, but describes the legal and spatial relations between e.g. a right-holder (e.g. an owner of a real property) and the RRRs that affect a specific piece of land. The basic LADM classes are -Party (a person or organization playing a role in a rights transaction (ISO, 2012, p. 4)), -RRR (a right, restriction or responsibility (ISO, 2012, p. 5)), -BAUnit (a basic administrative unit subject to registration or recordation or customary or social entity with RRRs associated to it (ISO, 2012, pp. 2-3)), and -SpatialUnit (one or multiple areas of land and/or water, or one or multiple volumes of space (ISO, 2012, p. 6)). In Figure 4, the classes have the prefix "LA_" attached to them to make them unique in the ISO series of geographic information standards.
LADM is an important part of the foundation for automation within the cadastral field, since it offers a standardized terminology for describing entities in the domain and their relationships, including different types of rights (or absence of rights) of different parties in particular situations. To build further on this foundation, a logical analysis of the notions of RRRs is close at hand.

Fundamental jural relations
A natural point of departure (see for example Paasch, 2012) for a logical analysis of RRRs is the work by Hohfeld (1913;1917) on the "fundamental jural relations" (often also referred to as "fundamental legal conceptions") right 8 (claim), privilege 8 A note on terminology: The term 'right' is somewhat ambiguous, since it is sometimes used in the generic sense of what in Section 2.3 was referred to as 'right, restriction or responsibility' (RRR). In the following, the more specific term 'claim' ('claim-right') is used instead of 'right' when referring to the fundamental jural relation, while 'legal position' will be used for the generic term that also includes restrictions and responsibilities, i.e. the absence of claim-rights. In other words, a (claim-) right is a kind of legal position, but not  LADM's LA_Party, LA_RRR, LA_BAUnit and LA_SpatialUnit (ISO, 2012, p. 9). https://doi.org/10.30672/njsr.78170 (liberty, freedom), power, and immunity, and their 'correlatives ' duty, no-right (no-claim), liability, and disability. See Figure 5.
A common view is that the so-called first-order 'Hohfeldian incidents' Privilege and Claim (shown in the left part of Figure 5) directly regulate what actions people may perform, while the second-order incidents Power and Immunity (right part) regulate the introduction and changing of other incidents. That some agent x has a power versus some other agent y means that x has the 'legal capacity' (according to some legal system) to alter y's Hohfeldian incidents, and that x has an immunity versus y means that y lacks the legal capacity to alter x's Hohfeldian incident. Privilege and Power are sometimes referred to as "active" rights, i.e. rights that concern the actions of the bearer of the right, while Claim and Immunity are referred to as "passive" rights that regulate the actions of others.
The fundamental legal conceptions are correlated to each other 9 in the following way (cf. Figure 5): -If x has a Claim versus y regarding some state of affairs F, then y has a Duty versus x regarding F. -If x has a Privilege versus y regarding F, then y has a No-claim versus x regarding not:F, i.e. the negation of F.
-If x has a Power versus y regarding F, then y has a Liability versus x regarding F. -If x has an Immunity versus y regarding F, then y has a Disability versus x regarding not:F. Also note that some fundamental legal conceptions are the 'opposites' of others: -If x has a Claim versus y regarding F, then x does not have a No-claim versus y regarding F.
-If x has a Privilege versus y regarding F, then x does not have a Duty versus y regarding not:F.
all legal positions are (claim-) rights. The acronym RRR will be used when referring to the notion of legal position in the LADM context.

9
A comprehensive review of the relationships between Hohfeld's fundamental legal conceptions is given by Lindahl (2006, pp. 327-331). -If x has a Power versus y regarding F, then x does not have a Disability versus y regarding F. -If x has an Immunity versus y regarding F, then x does not have a Liability versus y regarding not:F. As an example, let us assume that subdividing the property unit u O into two property units, the residual property (here denoted u R ) and the new subdivided property (u D ), is in line with the current municipal development plan. Let us further assume that both u R and u D are deemed enduringly suited to their purposes. Then if a person x owns u O , x has (versus any person y that does not own u O ) a legal position of type Immunity with respect to subdividing u O into u R and u D . In other words, y does not have a Power versus x regarding x's subdividing u O , i.e. y does not have the legal capacity to make x subdivide u O . Furthermore, x also has a legal position of type Immunity with respect to not subdividing u O into u R and u D . That is, y does not have a Power versus x regarding not subdividing u O ; y does not have the legal capacity to prevent x from subdividing u O . Hohfeld regarded the fundamental legal conceptions as "the lowest common denominators of the law" that could be used to express jural relations exhaustively and with high precision. It appears that what in LADM is referred to as restrictions and responsibilities may be expressed as no-claims, duties, disabilities or liabilities in Hohfeld's terminology. The fundamental legal conceptions and their correlatives thus seem to have the potential to capture the notions of RRRs (see Section 2.3) with higher precision. The observation, that what is expressed as a right for one party may also be expressed as a restriction or responsibility for some other party, is also in accordance with the LADM (ISO, 2012).

The theory of normative positions
A logical reconstruction of Hohfeld's theory was suggested by Kanger (1957) who combined the standard operator 'Shall' from deontic logic, i.e. the logic that deals with concepts such as obligation and permission, and the action operator 'Do'. This combination together with the negation operation 'not' gives us a powerful language for expressing normative sentences. For example, Shall Do(x, not:F) can be interpreted as 'it shall be (the case that) x sees to it that not F' or 'it shall be that x brings it about that not F'. Similarly, not:Shall Do(x, F) can be read as 'it is not the case that it shall be that x sees to it that F' or 'it is not the case that it shall be that x brings it about that F'. Hohfeld's fundamental jural relations may be formally stated in this language, and when combined with standard logical connectives, they can be used to express complex conditional normative sentences. Despite its compactness, the logical formalism has great expressive power that makes it possible to formulate and analyse normative systems with high precision. This, in turn is a prerequisite for automated application of norms, i.e. what Olsson et al. (2018, Sect. 2.3) refer to as "rule checking". https://doi.org/10.30672/njsr.78170 Kanger distinguished between four simple types of legal positions (claim, freedom 10 , power, immunity) and four simple 'counter-types' (counter-claim, counter-freedom 11 , counter-power, counter-immunity), where "counter" refers to the negation of some event or state of affairs. Thus, the expression 'x has versus y a privilege regarding not F' is synonymous to 'x has versus y a counter-privilege regarding F' (Lindahl, 1977, p. 43), where F may denote (cf. the example in Section 2.4) that the property u is subdivided into u 1 and u 2 . The relationship between Hohfeld's and Kanger's primitive concepts is shown in Table 1.
In the logical language described above, Kanger's explication of Claim(x, y, F) is Shall Do(y, F) 12 and the explication of Privilege(x, y, F) is not:Shall Do(x, not:F) 13 . The explication of each of the eight simple types is shown in Table 2, where May P is used as an abbreviation of 'not:Shall not:P', and the generic symbol F is replaced by a condition f(x,y) representing some binary relation f that may hold between x and y.
Note that for Privilege, Counter-privilege, Power and Counter-power it is the rights-bearer x that is 'active' in the sense of 'seeing to it that'. These types express x's freedom and capacity. Claim, Counter-claim, Immunity and Counterimmunity express obligations of the counterparty y, who is the active agent. Also, note that, for example, the explication Shall Do(y, F) of Claim(x, y, F) only explicitly references one of the agents involved, in this case the counterpart y. However, it is common that F represents some condition f on a number of agents, such as the binary condition having received the down payment for property u. In this case, Shall Do(y, f(x,y)) can be interpreted as follows: 'y shall see to it that x receives the down payment for u from y'. This implicitly represents that x has a claim on y regarding receiving the down payment for u.
A list of legal positions regarding f(x,y) may be constructed by forming the conjunction of each simple type, either negated or unnegated, and removing those conjunctions that are logically inconsistent given the underlying logic of Shall, May and Do. This list contains 26 'atomic' legal positions (see Section 3.4. for 10 Here, 'privilege' will be used instead of 'freedom'.

11
'counter-privilege'. 12 'y shall see to it that F'. 13 'it is not the case that x shall see to it that not F'. Table 1. Hohfeld and Kanger (based on Lindahl 1977, p. 49 Lindahl (1977) into three systems of types of normative positions. That x versus some counterpart y has a certain type of normative position with respect to, e.g., the state of affairs f(x,y) means that x has, or does not have, certain (types of) rights versus y as regards f(x,y). The simplest of these systems of types of normative positions, the so-called one-agent types of normative positions, is shown in Table A1 in the appendix.
Although not without its problems as a theory of rights (see, e.g., Makinson, 1986), the so-called Kanger-Lindahl theory of normative positions is generally regarded (Sergot, 2013, p. 355) as the most comprehensive and best developed attempt to formalize Hohfeld's fundamental jural relations, which may be expressed as logical combinations of normative positions.

Subdivision of a real property unit: A 'Situation Case Study'
Since property formation is central to the urban planning and building process, property subdivision (which, by far, is the most common case of property formation in Sweden) is selected as the object of study. A Swedish property formation process is generally divided into the four phases initiation, preparation, decision and registration (see for example Figure 31 in Vaskovich, 2012, for an overview). The analysis focuses on the decision phase of the property subdivision process, described in Section 3.1. Two 'Situation Case Studies' analyse snapshots of the property subdivision process, along with the types of entities and parties involved in the specific situations. Furthermore, the case studies formally describe the relationships and legal positions of said entities and parties. The paper studies (i) the situation where the cadastral surveyor is about to take the cadastral decision, and (ii) the situation just after the decision has entered into force. The goal of the analysis is to be more acquainted with the selected decision situation, in order to get a better understanding (and formal  Lindahl, 1977, p. 43.) Simple type Explication Active agent May not:Do(x, not:f(x,y)) x (rights-bearer) Shall not:Do(y, not:f(x,y)), i.e. not:May Do(y, not:f(x,y)) y

Counter-type Explication
Shall not:Do(y, f(x,y)), i.e. not:May Do(y, f(x,y)) y description) of the domain of the property subdivision process. A challenge here is to delimit the scope of the analysis. Explicitly mapping out all the legal positions that hold between all different parties and objects, is way beyond the scope of this paper. The aim is not to perform an exhaustive analysis of the decision situation, but to compare and discuss different tools for analysis and formal representation of legal positions between different parties. Therefore, Sections 3.3-3.5 focus on a subset of the legal positions that is manageable, yet rich enough to illustrate the 'analytic capacity' and expressiveness of the different analysis tools. Section 3.6. further discusses this approach.

The property subdivision process
A subdivision process is exclusively performed through decision of a cadastral surveyor. A completed subdivision may only be changed by another decision of a surveyor or through a court decision. The surveyor is in general rather free to arrange the subdivision process in the way seen as most suitable, not being bound by statutory regulations in this aspect. During the subdivision process, the cadastral surveyor makes several legal decisions, viz. the cadastral decision, the cost-distribution decision and the completion decision. The central decision is the actual cadastral decision to form a new property unit for e.g. housing purposes, which legalises (the existence of) the newly formed property (FBL, 1970, 4 chap. 25 a §). The decision must be made by the surveyor after all details of subdivision are investigated, i.e. after, among others, preparation, rearrangement of related property rights and the necessary surveying have been done. The decision includes the assessment of the requirement in the Swedish Property Formation Act that a newly formed real property unit must be enduringly suited to its purpose (FBL, 1970, 3 chap. 1 §), which includes being suitably designed and having (potential) access to both a road and acceptable sewerage arrangements and water supplies. Therefore, the surveyor takes the size of the new land plot and the future access to road, water and sewerage systems under particular consideration. Moreover, the surveyor assesses whether a requested subdivision is consistent with the existing municipal detailed development plan as well as with public land policy. If needed, the surveyor may (choose to, or be required to) consult with the municipality (typically regarding sewerage, water supply and/or development plan issues), the Road Authority (typically regarding road access), and the County Board (regarding environmental protection).
As soon as the cadastral decision is taken, a bundle of respective property rights is legally attached to the new property unit. When the appeal period expires, the cadastral surveyor finalises the registration of the new property unit. Subdivision completes by the respective entry regarding newly formed property units into the real property register.
The following Section 3.2 describes a fairly normal and uncomplicated property subdivision case, based on certain simplifying assumptions. The aim is to highlight important parts of the process and show how it may be carried out in this specific case. In general, the process can be much more complex.

Case description
The owner of a property (the 'original' property, in the following referred to as u O ) would like to subdivide the land parcel into two smaller property units, with the aim of keeping ownership of one part (the 'residual' property, u R ) and selling the other (the 'subdivided' property, u D ) to the buyer. The buyer plans to build a house on u D , and a preliminary building permit has already been applied for and granted by the municipality. There is no mortgage attached to u O , and there will be no mortgage attached to u D . A servitude with u O as the dominant property and the neighbouring property u N as the servient property grants the owner of u O the right of access to and use of the well situated on u N . (See Figure 6 for an overview of the intended situation after subdivision.) In the case presented here, the seller and the buyer make purchase arrangements and sign a purchase contract before subdivision, and then submit an application for subdivision. As soon as the application is registered at Lantmäteriet, a new cadastral dossier is created and the case is assigned to a cadastral surveyor. The cadastral surveyor examines the general conditions of subdivision, including a bundle of attached property rights, and the requirement that the new property unit is enduringly suited to its purpose. As already mentioned, this means that a newly formed property unit needs (potential) access to water supply and sewerage arrangements as well as free passage from the property (i.e. right of way). The requested subdivision must also be consistent with the existing municipal detailed development plan as well as with public land policy.
To ensure right of way for the owner of the subdivided property u D over the residual property u R to the public road, a new servitude has to be created. 14 14 Instead of creating servitudes, it would be possible to create a so-called joint facility, i.e. a construction (facility) beneficial for two or more real property units (AL 1973). Since it is more common to use servitudes to secure rights of access when very few properties are involved, and to avoid unnecessary complexity in the example, the study abstains from creating a joint facility here. Likewise, to ensure the owner of u D the right of use to the well on the neighbouring property u N , a servitude (with u N as the servient property and u D as the dominant property) can be created if an agreement is reached with the owner of u N . The existing servitude ensuring right of use of the well on u N to the owner of u O is not affected by the property subdivision process and will thus remain unchanged, since it still belongs to u O (now called u R ). The new servitudes are given unique designations and database id's in the real property register.
Subdividing the original property u O into u D and u R results in a new cadastral boundary being created to separate the properties. The area of u O is thus changed (i.e. reduced) in the property formation process, but the residual property u R keeps u O 's property unit designation and database identification number in the real property register. Some of u R 's former cadastral boundaries now mark the subdivided property together with the new created boundary dividing the properties 15 , thus creating a closed geographic area. u D receives a new database identification number and a new real property designation within the series used for the cadastral district in which the property is located.
As part of the property formation process, the cadastral surveyor has to ensure that the subdivided property u D will be connected to a sewage network or otherwise be able to discharge of its waste. In this example, the buyer has chosen to construct a small plant for sewage discharge treatment on the property. A permission for the construction of such a facility has been obtained from the municipality´s environmental department prior to the subdivision.
To summarise, the buyer (in the following referred to as a B ) is assumed to sign an agreement with a N , the owner of u N , regarding right of access to the well also for the owner of u D , i.e. that a N agrees to the creation of a new servitude for this purpose. The creation of a servitude that grants the owner of u D access to the road on u R , is likewise assumed, as well as that preliminary building permits for a new building and a sewage discharge facility have been obtained. Furthermore, this property formation is assumed uncomplicated in the sense that it is entirely in line with the current municipal detailed development plan and public land policy, and no consultation with other authorities is necessary.

Preamble to analysis: Actors and entities
The 'Situation Case Study' is prepared for by first identifying in the case description the parties (actors and stakeholders) that are directly involved in or affected by the Other parties that can be extracted from the case description, but do not occur explicitly in the analysis, include other neighbours, a technical surveyor, the Municipality, the County Board, the Road Authority, and the State.
The Use Case diagram in Figure 7 shows the main actors of the real property subdivision process. Zooming in on the particular decision situation, the cadastral decision, it can be seen that some actors are more directly involved than others are. The cadastral surveyor a C is the main actor in the sense of being the decision maker, while the seller a S and the buyer a B are also central actors that are directly affected by the decision. To some extent, other stakeholders such as an owner of a neighbouring property or some other private citizen (who, for example, for some reason opposes the formation of the new property) are also affected, whereas the previously mentioned public authority parties are not directly involved in or affected by this step of the process.
Next, central concepts are identified that occur in the case description. The following entities occur explicitly in the analysis: Examples of other entities that can be extracted from the case description, but do not occur explicitly in the analysis, are the subdivision application, the road on the residual property, the preliminary building permits for a building and a sewerage arrangement on u D , and 'technical' entities such as cadastral boundaries and database id's.

Structural analysis using LADM terminology
This section structures the case of property subdivision from Section 3.2 with the help of LADM object diagrams, i.e. UML diagrams following the LADM standard (ISO, 2012). The notations from Section 3.3 are utilized, slightly adapted. The original owner, and seller of the subdivided part, will be denoted aS in the diagrams (not a S as in the text). This and similar adaptations are made to make the diagrams more readable. Further, the class names of the LADM (e.g.LA_BAUnit) are utilised, but package names (e.g. Administrative) are not written out. Thus, LA_BAUnit stands for the class whose full name according to the LADM standard is Administrative::LA_ BAUnit.
A starting point for modelling the case according with LADM is the class diagram in Figure 8, which is adapted from Figures 9, 10, and 11 in the LADM standard (ISO, 2012). The classes are drawn from three of LADM's main packages (Party, Administrative, and Spatial Unit). As the case study focuses on RRRs, spatial relationships are not included in the models. E.g., the fact that a passageway over some property must spatially be 'within' this property, will not be represented in the models. (See, e.g., ISO (2012, Figure C.21) for an example.) Although UML class diagrams are static structure diagrams, it is worth noting that the LADM can be used for state-based modelling as well as eventbased modelling, see Appendix N of the LADM standard (ISO, 2012). Section 3.4.1 represents the situation before subdivision, and Section 3.4.2. the situation after subdivision.

Situation before subdivision
Before subdivision, the seller a S owns the original property u O . A servitude attached to the original property u O and the neighbouring property u N grants a S the right to use the well situated on u N . LADM can model the servitude as a restriction on the neighbouring property, in the sense that the ownership rights of the neighbour a N are restricted. Further, in the (non-normative) code list LA_RestrictionType there is a type servitude. The choice to represent a servitude as an instance of LA_Restriction is discussed in Section 3.6. Figure 9 follows the approach of Figure C.21 of the LADM standard (ISO, 2012), although in a different context. The well on the neighbour's property is represented by the LA_BAUnit instance w N , which is spatially located within the neighbour's property. According to the class diagram and other specifications in the LADM, each instance of LA_Restriction is associated to (linked to) exactly one LA_BAUnit, and to one (or none) LA_Party. 16 The LA_Restriction object is associated to an instance of LA_ AdministrativeSource, where the latter represent documents (in this case files in the cadastral dossier d) internal or external to the land administration organization. Not shown in the figure is the association between the administrative source for the original servitude and the cadastral surveyor (LA_Party) who once established it.

Situation after subdivision
When the original property u O has been subdivided into a remaining part u R and a subdivided part u D , both properties will have the right to access the neighbour's 16 A difference from Figure C.21 of the LADM standard is that we do not link the owner a S of the original property u O directly to the Restriction object s w,O . Instead, we follow Swedish law, which says that it is u O that has the servitude, not the owner of u O . (In order to follow the LADM specifications, we had to insert an LA_Party object representing u O acting as a party, viz. the nameless object between u O and s w,O .) well. In Figure 10, this has been represented as the two properties each having a unique servitude on u N . 17 For simplicity, Figure 10 excludes the instances of classes SpatialUnit and RequiredRelationshipSpatialUnit from the object diagram. Instead, a link is drawn directly from w N (the LA_BAUnit representing the well) to u N (the LA_BAUnit representing the neighbour's parcel). This might be seen as a link derived from a spatial relationship that is not shown. Section 3.6. discusses the spatial relationship between the two BA_Unit objects u R and u N further.

Logical analysis of atomic types of rights
To perform a logical analysis of the legal relations that hold between different parties (see section 3.3) regarding different conditions is not a simple task. The list of possible conditions to examine is practically endless, and each condition may be instantiated with different combinations of parties, and in different situations (e.g. before or after subdivision of u O ). However, some conditions that might be relevant to consider can be inferred from the process and case descriptions in Sections 3.1 and 3.2: 1. using the well on u N 2. using the road on u R 3. subdividing u O into u R and u D 4. receiving agreed payment for u D 5. retracting an application to subdivide u O into u R and u D 6. appealing a decision to subdivide u O into u R and u D 7. appealing a decision to deny subdivision of u O into u R and u D .
The following analysis, based on the fundamental jural relations and their logical explications (or, as will be discussed in Section 3.6, a generalisation of these notions) that were presented in Section 2.4, focuses on items 1 and 3. As in 17 An object diagram showing the situation where the two properties u R and u D have a common servitude on u N would be slightly different, e.g. using the LADM LA_GroupParty class to group the two real properties as a single party (not shown). Section 3.4, this section starts with the situation before subdivision of u O into u R and u D , and then look at the situation after.

Situation before subdivision
First, this section looks at the condition using the well on u N and the two parties a S and a N . If x is an agent, let w N (x) represent the event x uses the well on u N . 18 Which simple types of legal relations hold in this situation between the seller a S and the neighbour a N regarding w N (a S ), i.e. a S 's using the well on u N ? Given the assumptions in 3.3, which includes the existence of a servitude for the owner of u O (i.e. a S ) to use the well on u N , it seems reasonable to say that a S does not have a duty (versus a N ) to not use the well. It is natural to express this as the fundamental jural relation not:Duty(a S , a N , not:w N (a S )), (1a) which (since Privilege and Duty are 'opposites'; cf. Figure 5) is the same as Privilege(a S , a N , w N (a S )). (1b) By substituting the binary condition f in Table 2 for the unary condition w N , a possible logical explication is obtained of (1ab): May not:Do(a S , not:w N (a S )) (2) This can be read as 'it may be that a S does not see to it that a S does not use the well on u N '. It also seems reasonable to claim that a N lacks capacity to prevent a S from using the well. Below it is suggested how to express this as a fundamental jural relation, together with a possible logical explication: not:Power(a N , a S , not:w N (a S )) (3a) Alternatively put: Immunity(a S , a N , w N (a S )) (3b) not:May Do(a N , not:w N (a S )) (4) A possible interpretation of (4) is 'it is not the case that a N may see to it that a S does not use the well on u N '. By similar reasoning applied to the remaining simple types of legal relations, leaving out the details for brevity, Table 3 shows the suggested analysis.

Simple type
Possible logical explication Privilege(a S , a N , w N (a S )) not:Shall Do(a S , not:w N (a S )) Counter-privilege(a S , a N , w N (a S )) not:Shall Do(a S , w N (a S )) not:Claim(a S , a N , w N (a S )) not:Shall Do(a N , w N (a S )) not:Counter-claim(a S , a N , w N (a S )) not:Shall Do(a N , not:w N (a S )) Power(a S , a N , w N (a S )) May Do(a S , w N (a S )) Counter-power(a S , a N , w N (a S )) May Do(a S , not:w N (a S )) Immunity(a S ,a N , w N (a S )) not:May Do(a N , not:w N (a S )) Counter-immunity(a S ,a N , w N (a S )) not:May Do(a N , w N (a S )) 18 Despite some risk of confusion; in previous sections w N is also used to denote a domain entity, viz. the well on u N . From this follows that the atomic type of legal relation between a S and a N (before subdivision) with regard to a S 's using the well on u N is the following: The difference between (5), (6) and (7) is further discussed in Section 3.6. Similar analyses (not shown here due to lack of space) may be performed for other combinations of parties, as well as for the listed conditions 2, 4, 5, 6, and 7. An important observation is the following: the atomic legal relations regarding conditions 1-2 and 4-7 are significant for decisions on whether some party at some point does or does not act in compliance with applicable regulations and agreements 19 , but they do not affect (i.e. put restrictions on) the actual cadastral decision that our example focuses on. In contrast, condition 3, subdividing u O into u R and u D , is directly related to the cadastral decision. The paper will therefore examine examine this condition instantiated with the cadastral surveyor a C , and consider the two parties a S and a C . Let s O,R,D (x) represent the event x subdivides u O into u R and u D . 19 For example, a court decision on whether a N at some point was entitled to prevent a B from using the well. (Before subdivision, a N has this right, but no longer after.)

Simple type of legal relation Possible logical explication
Privilege ( Given the assumptions in 3.2 (that u R and u D are considered enduringly suited to their purposes, etc.), which simple types of legal positions regarding s O,R,D (a C ) hold between the seller a S and the cadastral surveyor a C ? Leaving out the details, Table  4 shows the suggested analysis. As Section 3.6 will discuss, the analysis in Table 4 can be questioned, but if it is accepted the atomic type of legal relation between a S and a C regarding the event s O,R,D (a C ) is the following: Given that this atomic legal relation holds in the cadastral decision situation, the cadastral surveyor has a duty to perform the requested subdivision, i.e. a C 's room for manoeuvre is restricted in this situation.

Situation after subdivision
First, looking at the condition w N , how do the atomic types of legal relations discussed in the previous section change as a result of performing the subdivision? After subdivision, the servitude for the owner of u O (now called u R ) to use the well on u N remains attached to u R . Therefore, the atomic type of legal relation between a S and a N regarding w N (a S ) that holds after subdivision is the same as before, viz.
(5). The same type of legal relation now also holds between a B and a N regarding w N (a B ), due to the creation of a similar servitude for (the owner of) u D , i.e. the type of atomic relation changes from (6) to (5). The legal relation between the parties a S and a B regarding w N (a B ) is not changed during the subdivision process, i.e. the type of atomic legal relation (7) still holds. As for the condition s O,R,D , subdividing u O into u R and u D , it no longer represents a meaningful condition: the event s O,R,D (x) is not possible for any agent x, since the property u O no longer exists in its original form. Therefore, after subdivision it is not meaningful to talk about the atomic legal relation between any two agents as regards s O,R,D (x).

Experiences and reflections
Sections 3.4 and 3.5 apply two different analysis tools to analyse and describe two 'snapshots' of the subdivision process. Although being simplistic in many ways, the case is still both realistic and rich enough to put the formalisms to the test, and no claims are made that the analyses are exhaustive. Here, the paper will reflect upon perceived strengths and weaknesses of the two formalisms, as well as their potential relation.
The analysis in Section 3.4 gives several insights regarding LADM as a tool for structural analysis. The LADM standard describes many useful 'off the shelf' tools for modelling the example situation(s). The paper has included notions of class diagrams (Figure 8) and object diagrams (Figure 9 and Figure 10), the latter to describe two specific situations related to the case, viz. the situation immediately before resp. immediately after the cadastral decision regarding subdivision. As Section 3.4 already mentions, object diagrams have not been utilized in the same https://doi.org/10.30672/njsr.78170 way as in the property subdivision example illustrated in Figure C.38 in the LADM (ISO, 2012).
For structurally analysing two situations of the example process, the LADM offers the possibility to represent formally (and with high precision) various RRRs of different parties in relation to each other and to land. The extension to the LADM classification proposed by  offers even higher level of precision, for example making it possible to model the right-of-use to the well on the neighbouring property (i.e. s w,O in Figure 9) as an LA_PartyToPropertyRight or perhaps an LA_PropertyToPropertyRight (see Figure 11 in Section 4). Since the focus of the LADM is conceptual rather than technical, and the standard was designed to give room for national implementations that adhere to different legal traditions, there are several degrees of freedom to its application, and the examples discussed in the standard sometimes give limited guidance. For example, Section 3.4.1 utilized the RRR subclass Restriction to model the servitudes in Figure 9 and Figure 10. However, since the paper refers to the owner of u O (resp. u D ) having a right to access the well for water, it might be perceived as more natural to model this with an instance of the Right class (attached to the dominant property) than an instance of the Restriction class. The paper refrained from this for two reasons. First, in order to follow the example C.21 of the LADM standard, and secondly, since the RestrictionType code list has an entry servitude, while the RightType code list has not. It is not clear from the description of the standard if any of the two options should be preferred, so it appears that the servitudes could have been modelled both as (subclasses of) LA_Right and LA_Restriction, perhaps depending on perspective. This freedom may be a benefit for a conceptual analysis, but at the same time a challenge when approaching technical implementation. Due to this and to the focus on concepts and static situation descriptions, the LADM support for further automation of land administration processes is rather limited.
Note that the two BA_Unit objects u R and u N in Section 3.4.2 originate from u O , and that the two LA_SpatialUnit objects in Figure 10 together constitute the original LA_SpatialUnit object in Figure 9. These relationships could be represented in the diagram by means of a LA_RequiredRelationshipBAUnit and a LA_RequiredRelationshipSpatialUnit, respectively. These relationships could be very important in various context, but were omitted from Figure 10 to reduce the complexity.
As demonstrated, the LADM with extensions can be used to structurally with high level of granularity model different categories of (e.g. public and private) RRRs, and how they are attached to different parties and administrative units. As Section 3.5 shows, the structural analysis can be complemented with a logical analysis that adds better support for legal reasoning. Through the analysis a deeper insight is gained into how to (with very high precision) state basic legal positions of different parties regarding some state of affairs or event in logical language. By doing so, it is possible to exhaustively formalise the leeway of the agents involved in the decision situation, and thus take another step towards increased automation of legal decision processes. Consider again, for example, the simple types of legal relations (5), (6) and (7) regarding the condition using the well on u N (abbr. w N ) in Section 3.5: Before subdivision, (5) is the type of legal relation between a S and a N with regard to a S 's using the well on u N , (6) is the type of legal relation between a B and a N with regard to the same event, and (7) is the type of legal relation between a B and a N with regard to a B using the well on u N . The difference between (5) and (6) is due to the existing servitude that gives a S a right-of-use (manifested as a Privilege, a Power and an Immunity) that a B does not have. The reason for the difference between (5) and (7) is that a S does have legal capacity regarding his/her own use of the well on u N , thanks to the well servitude, but not regarding a B 's use of the well.
(The servitude does not give a S the right to permit another person to use the well.) Regarding the condition subdividing u O into u R and u D (abbr. s O,R,D ), some of the suggested simple legal relations in Table 4 are rather straightforward, while others are not. It is not evident, for example, how to understand the notions of Power and Counter-power in this case. In the current situation, does a S versus a C have 'legal capacity' (power) regarding that a C subdivides u O ? As for Counterpower, does a S versus a C have capacity regarding that a C does not subdivide u O ? In Table 4 it is suggested that both Power(a S , a C , s O,R,D (a C )) and Counter-power(a S , a C , s O,R,D (a C )) hold.
The analysis in terms of simple legal relations may be translated to logical language through Table 2. Note, however, that the conditions using the well on u N (w N ) and subdividing u O into u R and u D (s O,R,D ) are unary conditions, i.e. conditions on one agent x. Kanger's logical explication of the fundamental jural relations (Section 2.5) originally presupposes binary conditions, i.e. conditions on two agents x and y. Thus, by generalising to unary instead of binary conditions, as in Table 3 and Table 4, the formalism is stretched a bit. This is not a limitation, since the main purpose of the logical analysis was nothing more than demonstrating how to put the logical language to work, but should be kept in mind when interpreting the result of the translation. For example, in the situation before subdivision, it seems reasonable to say that Counter-claim(a N , a B , w N (a B )) (9a) or, in other words, Claim(a N , a B , not:w N (a B )) (9b) holds, i.e. that a N (versus a B ) has a claim regarding that a B does not use the well on u N . On the other hand, equally reasonably, it holds that not:Claim(a S , a B , not:w N (a B )), i.e. it is not the case that a S (versus a B ) has a claim regarding that a B does not use the well. However, the generalised explication of (9ab) is Shall Do(a B , not:w N (a B )), (11) while the generalised explication of (10) is not:Shall Do(a B , not:w N (a B )), (12) i.e., there is a logical contradiction. Thus, when generalising the logical explication of simple types of legal relations to unary conditions, either the rights-bearer or counterparty may 'disappear' from the logical analysis, which may be potentially problematic. Regarding the interpretation of Power and Counter-power in Table  4, the generalised explication of Power(a S , a C , s O,R,D (a C )) is May Do(a S , s O,R,D (a C )), which may be interpreted as 'it may be the case that a S sees to it that a C subdivides u O '. The fact that a S has the right to have u O subdivided (provided that all necessary prerequisites are fulfilled) seems to already be adequately modelled by the simple type Claim, and what it would mean in the present situation that 'a S sees to it that a C subdivides u O ' is not wholly clear. The explication of Counter-power(a S , a C , s O,R,D (a C )) is May Do(a S , not:s O,R,D (a C )); 'it may be the case that a S sees to it that a C does not subdivide u O '. In this case, one might perhaps say that a S has the right to withdraw the subdivision application, and thus may see to it that a C does not subdivide u O . (See also the example in Section 2.4.) It could be argued that by performing a logical analysis of this simple decision situation, one takes a sledgehammer to crack a nut. In fact, the complexity of mapping out all the legal positions of different parties with respect to different conditions in the decision situation seems to be way out of proportion to the complexity of the actual decision. Besides, sorting out this complexity by hand is likely a very difficult and time-consuming task. These are valid objections, but it should be kept in mind that the aim here is to take a first step towards developing the theoretical and conceptual framework for digitalization and automation of the urban planning and cadastral process. By one step at a time applying the analysis tools to a number of situation case studies, this domain becomes more and more familiar, and at the same time the strengths and weaknesses of the applied tools become more evident. The rapid development of modern machine-learning techniques potentially offers a new kind of tools that for example makes it possible to derive formal descriptions of the normative systems that regulate the property formation decision-making from unstructured or semi-structured legal text and/or descriptions of real subdivision cases. A prerequisite for employing such approaches, however, is a thorough understanding of the domain, regarding both structure and logic.
An interesting question is how the two analysis tools relate to each other, i.e. how they 'fit together'? Since they have different purposes, i.e. conceptual and structural vs. logical analysis, they can be regarded as complements to each other rather than competitors. It is natural to ask whether it would be possible (and, if so, useful) to add support in the LADM for expressing RRR's as atomic types of legal relations, in order to allow for even more fine-grained analyses and smoother co-existence and integration of the two formalisms.

Discussion
The analyses in Section 3 focus on a subset of the domain and a subset of the legal positions that hold in the selected decision situation. (A more comprehensive analysis of the selected decision situation and/or decision process is beyond the scope of this paper, and thus left for future work.) This approach might seem limiting, but it should be noted that the legal positions of different parties in specific situations are rarely explicitly written down in legal text like laws or other normative systems. In the process of formulating a normative system that is internally consistent and precise, it may certainly be helpful to be able to formally express (some or all of) the legal positions that hold in a particular situation. However, Lindahl and Odelstad (2013, p. 547) argue that a set of sentences that contain individual names is not an appropriate representation of a normative system, since normative systems express general rules where no individual names occur. (Lindahl & Odelstad, 2013, p. 547) The application of a particular normative system to a particular situation usually does not require being able to map out all legal positions 'by hand'. Instead, the specific legal positions of the parties with respect to different states of affairs or events should follow from application of general regulations to the specific situation. What is needed for automated application of normative systems is therefore (i) a theory of representation of normative systems consisting of individual general norms, together with (ii) a mechanism for deducing specific legal positions of different agents by applying general norms to specific situations, (iii) a mechanism for checking the compliance of the agents' actions with these legal positions, and (iv) a computational framework for instrumentalising these components into executable code. One example of how to deal with items (i) through (iv) is briefly discussed below.
Individual items of a normative system, i.e. what is here referred to as norms, are often formally expressed as conditional sentences of the simple form P implies N (Q), where P and Q are descriptive sentences, often conditions on a number of agents, and N is a norm-creating operator. If the left part P of the implication holds, then the right part N (Q) is in effect, and potentially regulates the behaviour of some agent(s). The norm-creating operator N may be a deontic operator such as Shall and May, or an operator based on atomic types of legal relations (see Sections 2.5 and 3.5) or one-agent types of normative positions (see Table A1). Odelstad and Boman (2004) employ an algebraic version of the theory of normative positions, based on the notion of a condition implication structure (cis). In the cis approach to the formal representation of normative systems (item i), a conditional norm is represented as an ordered pair ⟨p, Nq⟩ where p (the 'ground' of the norm) and q are descriptive conditions and Nq (the 'consequence') is a normative condition on a number of agents. Applying an individual norm ⟨p, Nq⟩ is done through instantiating the ground p with the parties x i involved in the particular situation, and checking if p(x 1 , x 2 , ... , x n ) holds in this situation. If so, the following derivation scheme (see https://doi.org/10.30672/njsr.78170 Odelstad & Boman, 2004, p. 146) is used to infer a consequence in the form of a normative n-ary condition instantiated with n agents: The last step is to check whether the agents' actions comply with the normative consequence Nq(x 1 , x 2 , … , x n ). Hjelmblom (2015) demonstrates possible mechanisms for norm instantiation and rule checking, i.e. items (ii) and (iii), instrumentalised into a computational framework (item iv).
It can be noted that norms of the more complex form P 1 implies: P 2 implies N(Q), occur frequently in law; not least in normative systems containing intermediate concepts (see Section 2.2). For example, P 1 could denote that there is a servitude attached to the property unit u O regarding right-of-use of the well on u N , P 2 could denote that a S is the owner of u O , and N(Q) could denote the atomic legal relation (5) in Section 3.5.1. The consequence of P 1 is itself a conditional norm, since it is conditional on P 2 , the intermediate concept being the owner of u O . Jurists often call such consequences hypothetical legal consequences. For a discussion of the formal treatment of norms with hypothetical consequences, see for example (Lindahl and Odelstad, 2000) and Odelstad (in press).
During its development as well as after its publication in 2012, the LADM has been the subject of numerous research activities, with topics ranging from technical implementation issues and the registration of real property to legal and organisational aspects. Liedholm Johnson et al. (2015) showed that it is relevant to use a standardized approach for obtaining an overview of, and thus comparing the multifaceted nature of, private and public interests in land. A survey  showed that there has been limited focus on research on legal and organisational matters, such as how to organise and manage interests in land. The rather coarse classification in the LADM today could benefit from a higher level of specialisation by adding an extended classification, as mentioned in  and . Both discuss a proposed development of the LADM, an extension focusing on expanding the standard's terminology for providing a more detailed classification of land use than possible in the original standard. Figure 11 shows a possible extension of the LADM's legal right class, showing an extended classification for privately and publicly imposed rights.
An interesting line of work would be to examine the possibility (and usefulness) to extend the LADM even further with support for expressing RRR's as simple types of legal relations with regard to some state of affairs or event, for example by adding more subclasses or more developed property code-lists.
The logical analysis presented in Section 3.5 was based on Kanger's typology of atomic types of legal relations. As mentioned in Section 2.5, this typology was developed by Lindahl (1977) into three systems of types of normative positions. Recently, this theory has attracted attention within computer science and has been been put to work by, e.g., Jones and Sergot, Krogh and Herrestad, and Odelstad and Boman (see for example Sergot, 2013, Krogh & Herrestad, 1999, and Odelstad & Boman, 2004. As already mentioned, the latter work is based on an algebraic version of Lindahl's system of one-agent types of normative positions (see Table A1). In the algebraic approach, a normative system in its simplest form is represented as a so-called Boolean joining system where conditional norms are represented by ordered pairs ('joinings') which correlate descriptive conditions with normative conditions. It is straightforward to perform the analysis in Section 3.5 in terms of one-agent types of normative positions instead of atomic legal relations, and formulate corresponding normative systems as algebraic entities. Together with the work by Hjelmblom (2015), which demonstrates how to instrumentalise this approach into executable code, this prepares the ground for automated decision-making.
When developing automated decision-making systems within such complex areas as the urban planning and building process, and in particular property formation, it is important to consider and analyse in detail both technical, legal and organizational aspects. Ongoing projects seem to mainly focus on technical, and to some extent organizational, issues (Ekbäck, 2019), which makes it even more important to consider the legal aspects. The urban planning and building process consists of many different decision processes and includes different kinds of decisions. Some of them (such as the surveyor's cadastral decision as an example, or a building permit approval by a municipality) are related to specific cases and are taken by individual officials, while others (for example establishing municipal detailed development plans or comprehensive plans) are made by local assemblies. What these decisions processes all have in common is that they are directed and constrained by a legal framework which in the Swedish setting Figure 11. Specialization of the LADM's legal right profile (white) with an extended classification of privately and publicly imposed rights (yellow) (Paasch et al., 2015, p. 684). includes laws such as the Real Property Formation Act, the Planning and Building Act, the Joint Facilities Act, the Administrative Procedure Act, the Environmental Code, the Land Code, etc. This legal framework constitutes a normative system with which all decisions must comply, i.e. the law must always be obeyed, but the latitude given by the legislation can vary considerably. In the case of the cadastral decision discussed in this paper, the scope of action for the cadastral surveyor is very narrow: if the formal requirements for property formation are fulfilled, then the surveyor must approve the application. In other cases, for example regarding establishing local regulations or municipal development plans, the scope of action given by the legislation can be much wider, with room for making political choices (based on different preferences or ideological stances) between several options. Ekbäck (2019) discusses how digital processes could handle the many qualitative assessments that are required by law, where the variables that must be evaluated are neither quantifiable nor well defined, or may be based on normative political positions. He claims that change of property ownership and property transactions would be somewhat easier to make automated since no particular qualitative assessments are needed, but raises the question whether or not it would be possible to design the technology to handle the balance between different public and individual interests. This paper hopes to further contribute to this discourse, by suggesting that these neither quantifiable nor well-defined variables are to be understood as open intermediate concepts (see Section 2.2), and by discussing the theory of such concepts and their role in the decision-making process. Groundopen intermediate concepts are of special significance, and require special attention, since they function as 'decision points' in a decision process (Odelstad, 2019, pp. 106f). In the property subdivision process, for example, the cadastral surveyor must aggregate information of different sorts in order to decide on whether the factual grounds of, e.g., the previously mentioned condition being enduringly suited to its purpose apply in the specific case 20 , and thus its legal consequences are in effect. This potentially includes weighing together different legal facts and balancing sometimes conflicting interests. (This approach to open intermediate concepts, analogous to that of weighing together different aspects in a multi-criteria decision problem, is outlined in Odelstad, 2002, ch. 12-3.) Thus far, the paper has discussed one potential approach to the logical analysis and automated application of normative systems within the land administration domain, based on the algebraic approach to norms by Lindahl and Odelstad (2013) and its instrumentalisation by Odelstad and Boman (2004) and Hjelmblom (2015). Naturally, there are other interesting approaches to the formal representation and instrumentalisation of normative systems, such as Input/output logic (see the overview by Parent & van der Torre, 2013). Two recent examples within the land administration domain are the work by Lee et al. (2016) and Malsane et al. (2015) on formalising and digitalising building requirements and regulations. However, a particularly interesting feature of the work by Lindahl and Odelstad 20 I.e., that there are no legal impediments to forming the new property, as regards its suitability.
is that it is an application of their so-called Theory of Joining Systems (TJS). Since one of the aims of the development of this theory was "to provide tools for a rational reconstruction of a legal system with intermediaries" (Lindahl & Odelstad, 2013, p. 625), TJS and its application to normative systems containing intermediate concepts has the potential to be a useful part of the framework for digitalization and automation of the urban planning and building process. 21 Further development of the theory includes developing an algebraic version of the system of n-agent types of normative positions (Lindahl, 1977), to potentially address some of the limitations (briefly discussed in Section 3.6) of the simple one-agent system, and investigating the formal treatment of norms with hypothetical legal consequences. Further work (including computational logic considerations) on the instrumentalisation of the theory into executable logic programs is also of interest.
This paper has presented a simple case on property formation and a starting point for how an automated decision process could be achieved. However, when adding more complexity, as often is the case in real life situations, and including the assessments made by various authorities in several steps, additional considerations would have to be made. The required information and documentation as a basis for the assessment and decision-making has to be more standardized. In many cases, a combination of automation and manual assessments of more qualitative aspects might be necessary, at least during the initial phases before further development of the automated system. For example, formal analyses as performed here, together with analyses of relevant normative systems regarding the occurrence of open intermediate concepts, may lay part of the groundwork for semi-automated decision-making where a computer identifies decision points and presents a complex decision situation (and possibly suggest or recommend a particular decision) to a human decision maker, who then makes the necessary judgments and trade-offs. This, in turn, is an important step towards further automation of complex decisions.

Conclusion
This paper uses two different analysis instruments to perform structural and logical analyses of two specific snapshots of a fictitious property subdivision case in Sweden, focusing on the legal relations between different entities and parties involved in the specific situations. The structural analysis used the LADM ISO standard formalism, and the logical analysis was based on Kanger's atomic types of legal relations. By (i) combining two perspectives on formalisation and classification of legal relations within the urban planning and building domain, (ii) discussing some of the strengths and weaknesses of the two tools regarding the formal representation of RRRs of different parties in this domain, and (iii) discussing how the tools can be aligned, the paper has presented one way to analyse 21 Lindahl and Odelstad (2013, p. 546) argue that "[a] theory of representation for normative systems will be incomplete unless attention is paid to the role of intermediate concepts within the system (for example, the role of legal concepts such as ownership)". https://doi.org/10.30672/njsr.78170 and describe the land administration domain at a higher level of abstraction and formalization using different analysis tools.
Furthermore, the paper has provided suggestions of future research in several directions, including to model the general subdivision process by mapping out the (kinds of) different parties involved in or affected by the process and analysing what kind of decisions emerge where in the process. Another direction is to analyse further the normative systems (such as Swedish laws, regulations and municipal development plans) that regulate the process and how they form networks or strata of intermediate concepts.
Similar analyses of a wider range of subdivision process snapshots as well as analyses of other property formation processes are one suggested future research path. The paper has highlighted the need for more basic research on the theoretical tools themselves, such as to explore possible extensions of the LADM standard, and to further develop the Theory of Joining Systems (TJS) and put it to work within the land administration domain. Another suggestion for further work is the instrumentalisation of TJS into executable prototypes, and investigation of the possibility to use machine-learning approaches within the theoretical and conceptual framework developed here. This paper has taken one step towards a deeper understanding of the domain, and outlined some of the work needed to proceed even further, in the hope of providing better conditions for more efficient and transparent use of geospatial information, and increased automation of the property subdivision process and other related civil processes.