Huntrex 277 (Pty) Ltd v Margo Berzak and Others

CONSTITUTIONAL COURT OF SOUTH AFRICA Case CCT 69/23 In the matter between: HUNTREX 277 (PTY) LIMITEDApplicant and MARGOT BERZACKFirst Respondent REGISTRAR OF DEEDSSecond Respondent CITY OF CAPE TOWN MUNICIPALITYThird Respondent Neutral citation:Huntrex 277 (Pty) Ltd v Berzack and Others [2025] ZACC 6 Coram:Zondo CJ, Madlanga ADCJ, Bilchitz AJ, Chaskalson AJ, Dodson AJ, Majiedt J, Mathopo J, Mhlantla J and Tshiqi J Judgments:Dodson AJ (majority): [1] to [146] Chaskalson AJ (dissenting): [147] to [189] Bilchitz AJ (dissenting): [190] to [235] Heard on:12 March 2024 Decided on:30 April 2025 Summary:Servitude — whether praedial or personal — approach to interpretation of servitude — whether servitude compliant with principles of property law Praedial servitude — requirements of perpetuity and utility — whether utility satisfied by recreational use Personal servitude — whether servitude constituting personal servitude of usus Section 66 of Deed Registries Act 47 of 1937 — prohibiting registration of certain personal servitudes extending beyond the lifetime of the holder Common law of servitudes — whether same should be developed or modernised by way of new statutory regime Servitude — exercise required to be civiliter modo or reasonable ORDER On application for leave to appeal from the Supreme Court of Appeal (hearing an appeal from the High Court of South Africa, Western Cape Division, Cape Town). The following order is made: Leave to appeal is granted. The appeal is upheld. The order of the Supreme Court of Appeal is set aside and replaced with the following order: “(1)Leave to appeal is granted. (2)The appeal is upheld. (3)The order of the High Court is amended to read as follows: ‘1.The second respondent is ordered to rectify clause P of deed of transfer No. T10518/2017 executed at the office of the Registrar of Deeds, Cape Town on 28 February 2017 so as to read as follows: ‘P.SUBJECT FURTHER to the following conditions imposed by the Transferor in favour of herself personally, which conditions are as follows: (a)The property hereby transferred is subject to a servitude area 20 (twenty) metres wide, the Western boundary of which shall be parallel to the boundary marked DE on Diagram No 5253/1981 of the Remainder of Erf 380 CONSTANTIA, held by the said Transferor, Margot Berzack (born Illman) married out of community of property to Jeffrey Cyril Berzack, under Deed of Transfer No. 38631 dated 31st December 1970. (b)The said servitude shall be subject to the following terms and conditions namely: (i)No wall or fence of any description shall be erected on the servitude boundary except extension of existing type of fencing (wooden pole fencing). (ii)The Transferor shall have the right to plant, control, care for and renew the existing garden situated within the servitude area more fully described above. (iii)The Transferor shall have full rights of access to such servitude area in fulfilment of the rights hereby granted.’ 2.It is declared that clauses P(b)(i), (ii) and (iii) of the servitude— (a)were in their original form not lawfully capable of being registered in the title deed, or of operating, as a praedial servitude over Erf 8478 Constantia, (a) in favour of the remainder of Erf 380 Constantia, or (b) beyond the lifetime of the first respondent in favour of her successors-in-title, and the second respondent erred in permitting those clauses to be so registered; (b)were in their original form and are in their rectified form, lawfully capable of operating only as a personal servitude of usus over Erf 8478 Constantia in favour of the first respondent; (c)in their rectified form shall cease to have force or effect upon the death of the first respondent, or the alienation of Erf 380 by the first respondent, or if otherwise lawfully terminated, whichever is the earlier. 3.The balance of the relief sought by the applicant is dismissed. 4.The first respondent is ordered to pay two thirds of the costs of the applicant, including the costs of two counsel.’ (4)The applicant is ordered to pay two thirds of the costs of the first respondent in the applications for leave to appeal and the appeal to the Supreme Court of Appeal, including the costs of two counsel.” The first respondent is ordered to pay the applicant’s costs of the application for leave to appeal and the appeal in this Court, including the costs of two counsel. JUDGMENT DODSON AJ (Madlanga ADCJ, Majiedt J, Mathopo J, Mhlantla J and Tshiqi J concurring): Introduction A dispute between neighbours in the leafy Cape Town suburb of Constantia has found its way to this Court. The neighbours disagree about a servitude registered over their respective properties. The servitude allows the one neighbour to access and tend a garden on part of the property of the other. To resolve the dispute, it must be decided whether the servitude is praedial or personal. If praedial, it will bind the respective properties and their current and future owners in perpetuity. If personal, it is for the sole benefit of the current owner of the dominant property and cannot extend beyond her lifetime, at the latest. Parties The applicant is Huntrex 277 (Pty) Limited (Huntrex). It has as its directors and shareholders Ms Samantha Bain and her husband, Mr Rodney Bain. The property owned by Huntrex is their home. I will refer to the applicant as “the Bains” rather than using the company name. The first respondent is Ms Margot Berzack. The property owned by her is similarly her home. The second respondent is the Registrar of Deeds, who filed a report for purposes of the proceedings before this Court, presumably in terms of section 97(1) of the Deeds Registries Act. The Registrar of Deeds does not oppose the application. The third respondent is the City of Cape Town Municipality. It has not participated in any of the proceedings in the High Court, Supreme Court of Appeal or this Court. Background Ms Berzack was originally the owner of the entirety of Erf 380, Constantia (the property). She purchased it and took transfer in 1970. The property was some 8 310 m2 in extent. Her house was situated on the eastern side of the property. The garden extended westward from the house. The end of the garden was demarcated by a wooden fence. The land west of the wooden fence, although still forming part of the property, was left unused and untended. In 1982 Ms Berzack decided to subdivide and sell off the unused area. However, she ran into a problem. A local land use ordinance imposed a minimum erf size of 4 000m2 and the unused area was less than this. A subdivision of the property into two 4 000m2 erven would have divided the part of the property where Ms Berzack had created her garden. The new boundary would have crossed the paving alongside her swimming pool. In order to avoid the impact of the ordinance, Ms Berzack decided to create a servitude. The servitude would allow her and her successors-in-title to retain “the exclusive right to use, access and tend to that portion of the garden” that fell on the newly-subdivided area (the Bains’ property). The terms of the servitude were agreed with a Mr Arnold Gerhard Wellens who wished to purchase the 4 000m2 newlysubdivided erf. He has provided an affidavit in which he confirms that it was their mutual intention in creating the servitude to give effect to Ms Berzack’s decision. Mr Wellens purchased the newly-subdivided erf and the servitude was duly registered on 21 September 1983 when the erf was transferred to Mr Wellens. The servitude conditions were retained in each successive deed of transfer, the most recent being the deed in terms of which the Bains’ property was transferred to them on 28 February 2017. The servitude is set out in the relevant deeds of transfer as follows: “P.SUBJECT FURTHER to the following conditions imposed by the Transferor in favour of [herself] and [her] successors in title as owner of the REMAINDER of ERF 380 CONSTANTIA, which conditions are as follows: (a)The property hereby transferred is subject to a servitude area 20 (twenty) meters wide, the Western boundary of which shall be parallel to the boundary marked DE on Diagram No 5253/1981 in favour of the Remainder of Erf 380 CONSTANTIA, held by the said Transferor, Margot Berzack . . . under Deed of Transfer No. 38631 dated 31st December 1970. (b)The said servitude shall be subject to the following terms and conditions namely: (i)No wall or fence of any description shall be erected on the servitude boundary except extension of existing type of fencing (wooden pole fencing). (ii)The Seller shall have the right to plant, control, care for and renew the existing garden situated within the servitude area more fully described above. (iii)The Seller shall have full rights of access to such servitude area in fulfilment of the rights hereby granted. (the term Seller shall include her Successors-in-Title)” Diagrammatically, the servitude may be represented as follows:Diagram 1Diagram 1 Diagram 1 Diagram 1 In the above diagram, the Bains’ property is “Erf 8478” and Ms Berzack’s property is “Remainder Erf 380”. The servitude area is represented by the polygon “HGAKJH”. The solid line AG represents the boundary between Ms Berzack’s property and the Bains’ property. The dotted line “HJ” represents the eastern boundary of the servitude area. The dotted lines within the servitude area, with the word “Fence” appearing alongside them, represent the wooden pole fencing enclosing the garden which forms the subject matter of the servitude. It is common cause that the wooden pole fence completely precludes the Bains from accessing the part of the servitude area that is gardened by Ms Berzack. What precipitated the dispute was that the Bains’ small and adventurous dogs were able to climb through the wooden pole fence. The Bains could not. So, retrieving their dogs was difficult. They asked an architect to design an impermeable, transparent fence to replace the wooden pole fence. This would keep the dogs on the Bains’ side of the fence and give them sight of Ms Berzack’s garden area. The Bains sought Ms Berzack’s consent to the replacement of the fence, but she refused. She relied on her rights in terms of the servitude, particularly clause P(b)(i), requiring that “no wall or fence of any description shall be erected on the servitude boundary except extension of existing type of fencing (wooden pole fencing)”. It was at this point that the Bains investigated the matter further and for the first time came to appreciate more fully the implications of the servitude. Ms Berzack’s unwillingness to agree to replacing the fence is not the only complaint the Bains have. Ms Berzack, they say, treats the servitude area as being de facto (as a matter of fact) in her ownership. This is not seriously disputed by Ms Berzack. She asserts that the terms of the servitude give her exclusive access to and control over the entire servitude area. Photographs taken from Ms Berzack’s property of the servitude area would give the impression that at least that part of the servitude area that lies on her side of the wooden fence, forms part and parcel of her property. Notwithstanding her assertion of exclusive access to and control over the servitude area, she declined a request by a previous owner of the Huntrex property to pay a proportionate share of their municipal rates. She persists in this stance. Litigation history In the High Court The battle lines were drawn. The Bains applied to the Western Cape Division of the High Court, Cape Town (High Court). They took the attitude in their founding affidavit that clause P(b)(i) did not preclude them from erecting a wooden pole fence along the common boundary and giving Ms Berzack access to the servitude area by way of a gate in the envisaged boundary fence. Neither did it prevent them from demolishing the existing wooden pole fencing in the servitude area. The Bains asserted that the garden servitude was a personal servitude of usus (use); that the Registrar of Deeds had overlooked this; and that it had been impermissibly registered as a praedial servitude. Accordingly, the Bains sought rectification of the servitude and declaratory relief consistent with their stance. In answering the case sought to be made by the Bains, Ms Berzack insisted that a praedial servitude had been properly and lawfully registered over the Bains property as the servient tenement, in favour of her property as the dominant tenement. The servitude was not one of usus. It afforded her “a right of exclusive use” over the servitude area. The Bains’ envisaged new fence on the boundary between them would run almost along the edge of her swimming pool, over the surrounding paving, and would “sequester the servitude garden from [her] property”. On Ms Berzack’s interpretation, the servitude granted the owner of her property “sole and absolute discretion” to decide whether the existing wooden fence should be demolished or extended into the servitude area and the exclusive right to control access to the servitude area. Ms Berzack counter-applied for relief declaring that the servitude was to be interpreted in accordance with her understanding. Alternatively, she sought rectification to this effect. In the further alternative, she asserted rights equivalent to her interpretation of the servitude acquired by way of acquisitive prescription. The High Court identified the requirements for a praedial servitude. It focused on the requirement of utility. This requires that, for the servitude to be a praedial one, it “must offer some permanent advantage or benefit to the owner of the dominant land qua [(in their capacity as)] owner and must not merely serve his or her pleasure or caprice”. The High Court went on to hold as follows: “An inspection in loco revealed that the portion of the servitude area located on Ms Berzack’s property is but a part of the very large and beautiful garden and an additional area, as it were, for the private exclusive enjoyment and pleasure of the owners of Ms Berzack’s property. It is not an area that can be used separately from the rest of the garden and cannot be said to increase its economic, industrial or professional potential. It is at best a pleasant add-on to an already large and pleasant garden, providing pleasure to the owners of Ms Berzack’s property. The fact that they have exclusive use and access thereto for other pleasurable activities does not detract from this. I am inclined to agree that these rights do not increase the utility of Ms Berzack’s property. It does no more than enable Ms Berzack to engage in a personal pleasure or caprice, and is therefore one which may only be conferred by a personal servitude of usus. The servitude is thus in substance a personal servitude of usus and as such can only be in favour of Ms Berzack personally, and not her successors in title or in favour of Ms Berzack’s property. This being so, the servitude was prohibited from being registered as it is contrary to section 66 of the Deeds Registries Act . . . which prohibits the registration of a personal servitude of usus, purporting to extend same beyond the lifetime of the person in whose favour it was created.” Insofar as the counter-application based on acquisitive prescription is concerned, the High Court rejected this claim. It reasoned that— “if the servitude was originally legally flawed, because it amounted to the registration as a praedial servitude of what was actually a personal servitude of usus, that legal flaw was not cured by the operation of [acquisitive] prescription. . . . Put differently, [Ms Berzack] could not have, by acquisitive prescription, acquired a praedial servitude by subverting the provisions of Section 66 of the Deeds Registries Act”. Insofar as the Bains sought orders permitting demolition of the existing wooden pole fence enclosing the garden and authorisation to erect a wooden pole fence along the boundary between the erven, the High Court found that there was nothing in clause P(b)(i) of the servitude that precluded such relief. Accordingly, the High Court granted the relief sought by the Bains, including rectification of the terms of the servitude to reflect it as being a personal servitude of usus, not a praedial servitude; deletion of clause P(b)(i); the substitution for it of a clause stating “that no wall or fence of any description shall be erected on the aforesaid servitude area except extension of existing type of fencing (wooden pole fencing)”; and granting a series of declarators including those permitting demolition of the existing wooden pole fence and authorising erection of a wooden pole fence along the boundary between the erven. Ms Berzack’s counter-application was dismissed and she was ordered to pay the Bains’ costs of both the application and the counter-application. Ms Berzack’s application to the High Court for leave to appeal failed. Supreme Court of Appeal Ms Berzack then applied to the Supreme Court of Appeal for leave to appeal. The Supreme Court of Appeal set the matter down for oral argument on both the application for leave to appeal and the merits of the appeal. Ms Berzack also applied to the Supreme Court of Appeal in terms of section 19(b) of the Superior Courts Act for the admission on appeal of new evidence. This took the form of affidavits by a Mr Richard Moffat and a Ms Margaret Boag, both attorneys and conveyancers with over 35 years’ experience. Each claims to have registered what they consider to be praedial servitudes identical to the one in question, along with other servitudes which they contend are similar. These include servitudes to protect the view from a dominant property by restricting building on a servient property, and “grant[ing] the dominant property the use of the servient property for certain defined purposes which may be recreational in nature, for example, walking, hiking and jogging”. They go on to express the view that the High Court judgment creates uncertainty in the law and conveyancing practice. The Supreme Court of Appeal disregarded these affidavits on the basis that they had no bearing on the merits of the appeal but were, rather, filed in order to bolster Ms Berzack’s application for leave to appeal. I will, similarly, treat them as being relevant only to whether this Court has jurisdiction and whether it is in the interests of justice to grant leave to appeal. The Supreme Court of Appeal was divided. By a majority of 3:2, it upheld the appeal, set aside the decision of the High Court, and replaced it with a decision dismissing the Bains’ application. The minority agreed with the majority insofar as it overturned the relief allowing the Bains to demolish the existing wooden pole fence, to erect a new wooden pole fence on the boundary between the erven and to have clause P(b)(i) deleted. Save for that, their judgments diverged. The majority held that the High Court had misinterpreted the servitude and misconstrued the meaning of utilitas (utility) in finding that this requirement for a praedial servitude was not satisfied. In its view, the element of utility was indeed present in that— “[t]he right to the garden is reserved on the servient land and it inures in favour of the Berzack property, serving the pursuit of Ms Berzack’s personal pleasure or caprice. . . . The fact that the servitudal rights are enjoyed by the owner of the dominant tenement is a natural feature of the praedial right.” The utility requirement was also satisfied insofar as the servitude increased the economic potential of Ms Berzack’s property. The majority held further that the garden servitude in question was akin to the praedial servitude of view; and that the intention of Ms Berzack and Mr Wellens in 1983, was that the garden servitude should be reserved for the benefit of both Ms Berzack and her successors-in-title in perpetuity. Its registration as a praedial servitude by the Registrar of Deeds could therefore not be faulted. The minority pointed out that the determination of the rights under a servitude is not simply a matter of interpreting its terms. Sometimes the peremptory principles of property law will override the intention of the parties. It placed considerable emphasis on the presumption against a servitude, including that component of it that operates in favour of a personal over a praedial servitude. According to the minority, the operative provisions of the servitude conferred on Ms Berzack the right to use the servitude to tend her garden and to access it for that purpose. The only indications that the servitude was praedial were the reference to successors-in-title and the identification of the two properties, one being subject to a servitude area and one in whose favour the servitude area operated. The minority held further that the rights imposed were, by their nature, personal to Ms Berzack. On this basis it concluded that the intention of the parties was to create a personal servitude allowing Ms Berzack to “enjoy gardening in the servitude area”. The servitude fell squarely within the definition of the personal servitude of usus. To the extent that Ms Berzack sought to impose the servitude not only in favour of herself but also on her successors-in-title, section 66 of the Deeds Registries Act precluded its registration. The minority also considered the matter from the perspective of the requirements for a praedial servitude. It considered that the requirement of utility was not satisfied by a mere view of the servient property. Whilst an increase in market value might afford utility, this they held must flow from the enhanced usefulness of the dominant property deriving from the servitude. For perpetuity, the servient property must have a particular attribute affording some permanent advantage to the dominant property. Neither requirement was satisfied. In this Court Applicant’s submissions In their application for leave to appeal the Bains do not seek to reinstate the relief granted in their favour by the High Court, allowing them to demolish the fence in the servitude area and erect a fence on the boundary. They seek only to secure the relief that would have been granted by the minority of the Supreme Court of Appeal, namely to rectify the servitude to reflect it as a personal servitude of usus in favour of Ms Berzack. Consistent with this stance, the Bains in essence base their case on appeal on the reasoning of the minority in the Supreme Court of Appeal. The Bains make the point that “a significant difference between a praedial servitude and a personal servitude is that the intensity of the burden imposed by the latter is much greater than the intensity of the burden imposed by the former”. Personal servitudes allow the servitude holder to exercise certain central entitlements of ownership. If personal servitudes were allowed to operate as praedial servitudes, thus permitting the servitude holder to exercise the central entitlements of ownership in perpetuity, he or she would, the Bains submit, become the de facto owner of the subject property, with the registered owner left as the nominal or token owner of the property. On the basis of this analysis the Bains make the submission that— “the underpinning policy of section 66 of the Deeds Registries Act is to avoid this outcome by specifically providing that the personal servitudes of usufruct, usus and habitatio [(habitation)] shall not extend beyond the lifetime of the person in whose favour it is created”. Yet, they contend, this is the effect of the servitude in question. The Bains assert jurisdiction solely under this Court’s general jurisdiction in terms of section 167(3)(b)(ii) of the Constitution on the basis that the matter raises an arguable point of law of general public importance. First respondent’s submissions Ms Berzack places considerable reliance on the assertion that this Court’s general jurisdiction is not engaged. She submits that the matter raises questions of fact, not law. To the extent that it might be held that the Bains raise as a question of law that praedial servitudes only impose benign restrictions on the main entitlements of ownership, whereas personal servitudes divest the owner of the central entitlements of ownership, she submits that the point is not arguable. This is because there is no authority for it and the argument is at variance with the central tenets of the common law on servitudes. What in truth distinguishes personal servitudes from praedial servitudes, she says, is that the former are inseparably attached to the identity of a particular beneficiary, not to the dominant property. Ms Berzack contends that the garden servitude in question imposes no greater burden than other recognised praedial servitudes, including those which afford the dominant landowner the exclusive right to trade on the property of another; the praedial servitude of jutting beams, which provides for overhangs on balconies, and praedial servitudes that afford the owner of the dominant property the right to establish and use parking bays on the servient property. Ms Berzack contends that the lack of prospects of success preclude a finding that it is in the interests of justice to grant leave to appeal. While acknowledging the principle of restrictive interpretation of servitudes, she argues that where the servitude is framed in unambiguous terms, a court is not entitled to ignore that meaning in order to achieve a lesser curtailment. Moreover, she says the servitude in question satisfies the five requirements for a praedial servitude including those of utility and perpetuity. On this basis, Ms Berzack asks that the application for leave to appeal be dismissed. Analysis Jurisdiction The Bains do not contend that the matter raises any constitutional issue. The enquiry is therefore confined to whether the matter raises arguable points of law of general public importance which ought to be considered by this Court. Since this is a major focus of Ms Berzack’s resistance to the application, it deserves more attention than would ordinarily be the case. Relying on the second judgment in Media 24, Ms Berzack contends that this Court will only assume general jurisdiction on “purely legal issues”, because it was said that “mixed questions of fact and law require evaluative assessments and it is precisely those assessments that it is not the function of this Court to tread into”. Apart from the fact that the second judgment in Media 24 was the minority judgment on the issue of jurisdiction, the quote is taken out of context. On the view taken in that minority judgment, resolving the point there in issue unavoidably required assessing the relative merits of conflicting expert evidence. The minority considered this a factual enquiry which it was not the function of this Court to perform, particularly where the matter was on appeal from a specialist court, the Competition Appeal Court. In this case, we are not concerned with that scenario. As this Court pointed out in Paulsen, it will not entertain purely factual matters. In truth it will seldom be that a question comes before this Court as a pure question of law, divorced from the underlying facts of the case. Of course, the point of law must be arguable, requiring a degree of plausibility that provides some prospects of success. Criteria to support making that assessment include whether “[t]he Supreme Court of Appeal . . . expressed itself on the matter by a narrow minority”, whether a minority view in the Supreme Court of Appeal is “quite forceful”, whether the matter raises a “new and difficult question of law” or where “the answer to the question in issue may not be readily discernible”. Taking into account the majority and minority judgments of the Supreme Court of Appeal, these criteria are satisfied in relation to at least the following points of law: The question whether the garden servitude in question constitutes a praedial servitude or a personal servitude. Sections 3(1)(o), 66, 67 and 76 of the Deeds Registries Act all refer variously to personal and praedial servitudes. This judgment will give meaning to these terms in the statute and explicate the test for identifying praedial servitudes, an issue that is by no means settled. Section 66 of the Deeds Registries Act expressly incorporates by reference three personal servitudes that derive from Roman and Roman‑Dutch law, namely usufruct, usus and habitatio (habitation). The meaning of the term usus in section 66 is a question of law that must be determined and then applied to the facts of this matter in order to render a decision. The content of the utility requirement for a praedial servitude is not settled. According to the late Professor van der Walt, there are different approaches to the utility requirement discernible amongst the Roman‑Dutch authorities and in the case law. It is necessary to explore these in order to decide this case. The content of the perpetuity requirement for a praedial servitude has received relatively scant attention in the case law and is central to the decision of this case. Contrary to the submission made by Ms Berzack, the Bains’ argument that a distinguishing feature of personal servitudes is the intensity of the burden imposed by them on the servient property is indeed arguable and enjoys the support of well-respected academic writers on the law of property. Are the above questions of law ones of general public importance? This raises the question whether a dispute between two neighbours over a gardening servitude “transcends the narrow interests of the litigants and implicates the interest of a significant part of the general public.” Ms Berzack contends that it does not. In my view it does for the following reasons: If the servitude is held to be praedial it binds all future owners of both the dominant and servient properties in perpetuity. The future owners in perpetuity are, potentially, an infinite number. It is apparent from the affidavits of both Ms Boag and Mr Moffat, both experienced conveyancers, that the Registrar of Deeds has registered similar garden servitudes involving other properties as praedial servitudes without demur. It is important for the Registrars of Deeds that their obligations in this regard are clarified. The Registrar of Deeds has, according to them, also registered praedial servitudes over servient properties for other recreational or pleasurable purposes such as walking, hiking and jogging. In their view the High Court judgment will preclude the future registration of all these kinds of servitudes and lead to the expungement of “countless numbers of praedial servitudes that are already registered” with “serious consequences for the rights of the owners of the properties that benefit from these praedial servitudes and rely upon them”. This, too, points to the dispute transcending the narrow interests of the litigants. Leave to appeal In Paulsen, the enquiry as to whether the matter was one that “ought to be considered” by this Court was treated as co-extensive with the enquiry whether it is in the interests of justice to grant leave to appeal. Considerations that are important in this enquiry include the importance of the issues raised, the prospects of success and whether it is in the public interest that the matter be entertained. The split decision in the Supreme Court of Appeal, with a carefully reasoned dissent by two of the five judges, who aligned themselves largely with the High Court judgment, points to there being reasonable prospects of success. The Deeds Office will benefit from clarification of its duties with reference to section 66 of the Deeds Registries Act. The case presents an opportunity to provide some clarity on areas of the law that have remained unclear for close-on two millennia. I am accordingly satisfied that the Court has jurisdiction in terms of section 167(3)(b)(ii) and that it is in the interests of justice to grant leave to appeal. The legal nature of a servitude In Link Africa, this Court provided an overview of the origins and nature of servitudes. Our law in relation to servitudes derives from Roman law. A servitude is “a real right carved out of the full dominium [(ownership)] of the owner and transferred to another”. LAWSA defines servitudes as follows: “A servitude is a limited real right that imposes a burden on movable or immovable property by restricting the rights, powers or liberties of its owner in favour of either another person (in the case of a personal servitude) or the owner of another immovable property (in the case of a praedial servitude).” A praedial servitude involves immovable properties. It confers rights on one property, the dominant property, over another, the servient property, upon which it imposes corresponding burdens. The owner of the dominant property enjoys the benefits of, and rights conferred by, the servitude over the servient property, but only for as long as she is owner. Since the rights attach to the dominant property and not its owner, when the owner transfers the land she loses the rights and they pass to the new owner for as long as she remains owner. This continues in perpetuity. The successive owners of the servient property are similarly bound in perpetuity. Nowadays it is recognised that there is no limit to the kinds of praedial servitudes that may be registered, as long as they satisfy the requirements for a valid praedial servitude. Roman Law distinguished between rural and urban praedial servitudes. Examples of rural praedial servitudes inherited from Roman law include a right of way and a range of water servitudes. Examples of urban servitudes inherited from Roman law include a servitude of light and a servitude of view. A personal servitude is a real right of use, vesting in a person, over the movable or immovable property of another. The right attaches to the person holding the servitude, not to a dominant property. It burdens the servient property for the duration of the life of the servitude holder, unless a shorter period is specified in the agreement or will creating the servitude. For as long as the servitude-holder is alive, the servient property remains bound, notwithstanding a transfer of ownership. As with praedial servitudes, subject to these qualifications, there is no limit on the range of personal servitudes that may be registered. In contrast, Roman Law only recognised four personal servitudes namely usufruct, use, habitation and services, the latter in the form of the use of a slave or animal. Save for services, not only have these servitudes been inherited as part of our common law, but they are also recognised and regulated by statute in section 66 of the Deeds Registries Act. Praedial and personal servitudes are created either by conditions included in a deed of transfer of immovable property or by registration of a notarial deed. In this case the former method was employed. Against this backdrop I now turn to consider how the provisions in a deed of transfer or notarial deed creating a servitude are interpreted. Interpretation The starting point in interpreting a servitude was described in Willoughby’s as follows: “Whether a contractual right amounts in any given case to servitude – whether it is real or only personal – depends upon the intention of the parties to be gathered from the terms of the contract construed in the light of the relevant circumstances. In case of doubt the presumption will always be against a servitude, the onus is upon the person affirming the existence of one to prove it.” That presumption operates at three levels. First, it is presumed that the property is free of any servitudes at all. Second, if the servitude is proven, it is presumed to impose the least onerous burden. Third, if it is unclear whether the servitude is praedial or personal, the presumption favours a personal servitude. However, where the terms of the servitude are clear and unambiguous, the presumption does not operate and effect must be given to the servitude on its terms. I pause here to point out that the language used in the authoritative judgments on the interpretation of servitudes is outdated. This is no doubt because they generally predate the judgments of this Court and the Supreme Court of Appeal that modernise the approach to the interpretation of legal instruments. I will base the interpretation of the terms of the servitude in question on the modern judgments, which require that the exercise be approached “holistically: simultaneously considering the text, context and purpose”. Applying these principles of interpretation to the present matter, there are two components to the interpretive exercise. The first is to establish the content of the rights capable of being exercised by Ms Berzack, and the forbearance imposed on the Bains, under the servitude. The second is to determine whether the servitude in question is praedial or personal. The operative part of the servitude for purposes of identifying the primary rights afforded and forbearance imposed, is in clause P(b)(ii). That affords Ms Berzack the right to “plant, control, care for and renew the existing garden situated within the servitude area”. It is common cause that the gardening does not take place over the entire servitude area. The word “existing” would mean that the garden itself would be confined to the limits that existed when the servitude was first established in 1983. It is not disputed that this is precisely the area in which Ms Berzack gardens today. Clause P(b)(iii), conferring upon “the [s]eller . . . full rights of access to [the] servitude area”, determines access for purposes of exercising the gardening rights. In Ms Berzack’s own words in her answering affidavit, she says: “Since 1983, I have continued openly and as of right, to exercise (as owner of my property) exclusive control over the portion of my garden that now falls on the Huntrex property in the same manner I did before the subdivision.” She is supported in her stance that this is what was intended by the servitude, by the evidence of Mr Wellens, who was the first owner to receive transfer of the subdivided portion and consented to the terms of the reservation of the servitude in the deed of transfer. This provides important context. Taking into account this evidence of the factual matrix at the time of the conclusion of the original servitude document, the words “full rights of access” in clause P(b)(iii) must mean exclusive access, at least in that part of the servitude area that is bounded by the wooden pole fence in existence at the time of the creation of the servitude and still in place today. It is common cause that the effect of the wooden pole fence is to preclude entirely the rights of access and possession that the Bains would, as owner of the land, otherwise be able to exercise in the servitude area. There was some suggestion that clause P(b)(i) (which precludes erection of any wall or fence on the servitude boundary other than extension of the existing wooden pole fence), exists as a distinct, negative praedial servitude separate from clause P(b)(ii) and (iii). I disagree. In my view it is ancillary to, and part and parcel of, the servitude created by clause P(b)(ii) and (iii). Although the wording is a little obscure, in my view it serves to protect Ms Berzack’s exclusive use of the area that she gardens. This is because it is implicit in the wording that the existing wooden pole fence can only be extended, it cannot be removed. The corollary of Ms Berzack’s exclusive access to the servitude area for gardening is that the Bains are entirely deprived of any access to, possession of, or use for any purpose of the servitude area, or at least that part of it behind the wooden pole fence. None of the ordinary incidents of ownership accrue to them in respect of that portion, although counsel for Ms Berzack pointed out that the Bains would, as the owners of their property, be entitled to object to any use by her that fell outside the terms of the servitude. Is the servitude personal or praedial? The terms of the servitude are in my view ambiguous in this regard. As the minority pointed out, “the wording . . . pulls, at times, in different directions”. In one breath the servitudinal rights are conferred on the person of Ms Berzack. In the next, they are conferred on Ms Berzack’s property. That ambiguity brings the presumption favouring a personal servitude into play. But the reference twice in the conditions to the conferral of the rights also on successors-in-title, favours a praedial servitude. The evidence of both Ms Berzack and Mr Wellens as to what they sought to achieve at the time also tends to favour a praedial servitude. On balance, and notwithstanding the presumption, I accept that based purely on the wording of the deed of transfer creating the servitude, they intended to create a praedial servitude. That however is not the end of the enquiry, because, as is pointed out by Van der Walt— “the nature and content of a servitude depends to a degree on the intention of the parties creating it, but the law will not give effect to the intention of the parties if they intended to do something that is not possible according to the principles of property law.” (Emphasis added.) This raises the questions whether the servitude complies with the principles of property law pertaining to the requirements for a valid praedial servitude and whether, if it does not, the servitude is a personal servitude registered in conflict with the requirements of section 66 of the Deeds Registries Act. The requirements for a praedial servitude are essentially duality, vicinity perpetuity, utility and passivity. In this case, we have two separately-owned, adjacent properties, so the duality and vicinity requirements are satisfied. In relation to the passivity requirement, the Bains have no positive duties arising from the servitude. This means that the passivity requirement is satisfied. What then of the remaining two requirements, perpetuity and utility? Perpetuity LAWSA has the following explanation of the perpetuity requirement: “The use made of the servient land must be based on some permanent feature or attribute of the servient land. The servient tenement must be capable of continuously fulfilling the needs of the dominant tenement. This is known as the requirement of perpetual purpose.” (Emphasis added.) The requirement traces its origins to a passage in the Digest based on views attributed to the jurist Paulus. It refers to the requirement of perpetuity in the context of water servitudes and states that “all real [that is, in this context, praedial] servitudes are required to depend on the causes that are perpetual”. Paulus’ view is that water servitudes will thus only be recognised where there is a natural supply of water or rainwater. The view is also expressed in the same part of the Digest that “what is effected by human hands has not perpetual cause”. On the basis of this statement, the late Professor de Waal questions the accuracy of this part of the Digest as a basis for a general, self-standing perpetuity requirement. He does so because by the time that Paulus wrote, Roman law had developed to accept that a servitude may be based on something that was constructed by human hands, such as a building. He refers in this regard to the various urban praedial servitudes that are based upon a construction on the servient property. Voet, however, accepts the perpetuity requirement as part of Roman-Dutch law and bases his view on the above extract from the Digest. Professor de Waal’s view is that the perpetuity and utility requirements are closely related and the former may be subsumed into the latter. He does, however, accept perpetuity as a continuing and important requirement. He summarises the combined requirements of utility and perpetuity as follows: “A praedial servitude can only be exercised to the use and advantage of successive owners of the dominant tenement if the character, quality or a feature (kenmerk) of the servient tenement that provides the use or advantage is constant (standhoudend), not accidental or transient.” The requirement has received some attention in the case law. Watermeyer J in Dreyer, like Paulus, regarded a servitude of aqueduct as requiring a natural source (aqua viva) of water. By contrast, Professor de Waal refers to various cases where water servitudes were recognised in respect of man-made structures like dams and where no regard is had to the source of the water. The perpetuity requirement was seemingly first referred to expressly in Venter where Hoexter JP said “[a] praedial servitude comes into existence only if the right to be acquired by the praedium dominans [(dominant property)] is for its perpetual benefit”. In Willoughby’s, the fact that the buildings erected to give effect to the contract sought to be registered as a servitude were temporary in nature, not involving any durable construction, was an indicator for the court that there was no intention to establish servitudinal rights. Based on this survey of the academic writings and the case law, I am satisfied that perpetuity remains a distinctive requirement for the existence of a praedial servitude, albeit one that is closely related to the utility requirement. The servitude must therefore “be based on some permanent feature or attribute of the servient land”. Turning to the facts of this case, whilst the garden has been maintained by Ms Berzack for more than 50 years, it cannot be considered a “permanent feature or attribute” or a “characteristic” that inheres in the Bains’ property that will be constantly present in perpetuity. A garden is subject to the vagaries of drought, flood, hail, pests, climate change, neglect by a disinterested or absent owner or tenant and so on. Without more, it involves no durable natural feature, nor any permanently constructed feature. It requires constant upkeep, including watering, fertilising, replanting, pruning, mowing and trimming. The wooden pole fence is neither a distinctive feature nor a durable installation in the way that a brick and mortar structure might be. The underlying land on which the garden is kept is like any other land, including vacant land on Ms Berzack’s property. It lacks any permanent feature or attribute to serve Ms Berzack’s property in perpetuity. In the circumstances, I am of the view that the garden servitude does not satisfy the perpetuity requirement of a praedial servitude. Utility Utility is the next requirement to consider. LAWSA describes this requirement as follows: “A praedial servitude must offer some permanent advantage or benefit to the owner of the dominant land qua [(in their capacity as)] owner and must not merely serve his or her personal pleasure or caprice. This is known as the requirement of utilitas. The locus classicus [(classic example)] of this requirement is a Digest text which states that a praedial servitude cannot be constituted to allow the owner of the dominant tenement to pick fruit or to promenade or dine on another’s land. Such a right advances the pleasurable pursuits of a person and does not increase the utility of landed property.” Significantly for the present matter, Voet explains the rationale behind the rule precluding a servitude based on pleasurable pursuits, as follows: “To promenade, to dine and to pick fruit is a right not of such an estate but of a person, and is related rather to the delectation of human beings than to the benefiting of landed estates. . . . Apart from that it is not to be doubted that picking fruit, promenading and dining on another’s ground can without question be brought into a servitude of use, restricted in a definite manner.” Voet is supported in this strict approach to praedial servitudes by Huber, who adds singing, dancing and playing to the list of examples of activities that do not establish a servitude. A number of respected South African academic writers express the firm view that a praedial servitude cannot be established for the personal pleasure of the owner of the dominant property, although it may form the basis for a personal servitude. This was stated as being part of our law in the judgment of Davis AJ in De Kock. Indeed Ms Berzack in her written submissions described the fifth requirement as “utility provided to the dominant tenement that is not merely aimed at the pleasure or caprice of the person who happens to be the owner thereof”. In applying this requirement to the facts of this case, we are constrained by the factual finding of the High Court, the majority and the minority of the Supreme Court of Appeal, on a conspectus of the evidence put up, that the garden servitude in this case, to quote the majority of the Supreme Court of Appeal, served “the pursuit of Ms Berzack’s personal pleasure or caprice”. The High Court arrived at its conclusion also having had the benefit of a site inspection. As an appellate court, we may not lightly interfere with that factual finding. It is so that Voet acknowledges that the utility requirement is satisfied where a servitude affords both pleasure and utility. This is discussed under the heading “[p]raedial servitudes may be granted for both pleasure and benefit.” Relevant to this authority, Ms Berzack argues that the gardening servitude also provided a view over the servitude area. I do not accept this argument. Any pleasurable right given over the property of another that involves access to it, will always afford a view over that servient property. Indeed, in many urban homes, there is a view over the neighbour’s garden simply by reason of its proximity, not by reason of any servitude. In any event, the view contemplated by a servitude of view is surely the ability to see into the distance beyond one’s immediate confines, not 20 metres into the neighbour’s garden. The view cannot therefore provide the requisite additional element to constitute the form of utility envisaged for a praedial servitude, and the majority of the Supreme Court of Appeal was in my view incorrect in this regard. Ms Berzack attempts to deal with this difficulty by saying— “Voet confirms [that] a servitude which is directed at increasing the pleasure of the dominant landowner is likely also to increase the enjoyment of the property as a whole, and that is likely to increase the market value of the property, which is sufficient to satisfy the requirement of utility”. Reliance is placed on Hollmann in this regard. There are the following difficulties with this argument: The Appellate Division in Hollmann only went as far as saying that Voet and Brunneman “appear to be of the view that a servitude would qualify as a praedial servitude if it would raise the price of the dominant [property]”. Moreover, the comment is obiter (made in passing) because the Appellate Division went on to decide that the utility in that case lay in the exclusive right of the owner of the dominant property to trade on the servient and dominant properties and the protection from competition that came with that restraint. De Waal’s comment on the extract from Voet relied on for this view is that it is inconsistent with Huber and with Voet’s approach to the utility requirement in every other respect. To the extent that Hollmann is to be considered as authority for an increase in market value providing the requisite utility, this has been criticised by most South African authors “because the decision effectively does not require any direct land-use link between the servitude and the dominant land and therefore abandons the utility requirement altogether”. I am accordingly not satisfied that value enhancement is a sufficient basis for satisfying the utility requirement. Yes, it may be a consequence of the utility otherwise afforded by a servitude. But value enhancement is not sufficient in and of itself to satisfy the utility requirement. Judgments suggesting otherwise were not correctly decided. Even if this were not the case, the question whether or not a servitude enhances the market value of a property is something which cannot simply be speculated on. Qualified valuers must provide evidence on market value and there is authority as to the methodology to be used in assessing the enhancement or reduction in value brought about by the registration of a servitude. Although both the minority and majority in the Supreme Court of Appeal (and the Bains) accepted that there may be an enhancement in market value brought about by the servitude in question, this was not based on any reliable evidence of market value. Accordingly, even if an increase in market value was sufficient to establish utility, there is no evidentiary basis for a finding in this regard. Accordingly, the utility requirement for a praedial servitude is also not satisfied. Consequences of absence of utility and perpetuity Before dealing with the consequences of the failure to satisfy the perpetuity and utility requirements, regard must be had to Ms Berzack’s reliance on the judgment of the England and Wales Court of Appeal in Ellenborough Park. That case concerned a large rectangular garden, 350 yards by 100 yards, around which houses had been built on three sides, with the fourth side open to the sea. The houses were designed so that they would have one communal garden rather than individual gardens. The original deeds of transfer conferred on the owners of the houses and their successors-in-title, jointly, full rights of access to and enjoyment of the garden, subject to a fair share of the costs of maintaining it. The issue in dispute was whether this amounted to a binding easement, the equivalent for present purposes of a praedial servitude. The English law of easements was strongly influenced by Roman Law. There was however debate about whether that extended to the rejection of easements for purely recreational or pleasurable purposes. The Court of Appeal said— “we do not think that the right to use a garden of the character with which we are concerned in this case can be called one of mere recreation and amusement . . . No doubt a garden is a pleasure – on high authority, it is the purest of pleasures; but, in our judgment, it is not a right having no quality either of utility or benefit as those words should be understood. Its use for the purposes, not only of exercise and rest but also for such normal domestic purposes as were suggested in argument – for example, for taking out small children in prams or otherwise – is not fairly to be described as one of mere recreation or amusement, and is clearly beneficial to the premises to which it is attached.” A binding easement was therefore found to exist, benefitting not only the original owners, but also the successors-in-title to the houses surrounding the garden. It is noteworthy that the Court of Appeal found that the garden in question was distinctive and that the utility went beyond “mere recreation and enjoyment”. This was, to a degree, an acknowledgement of the continuing impact of the principle deriving from Roman Law. In Regency Villas, the United Kingdom Supreme Court nevertheless treated Ellenborough Park as a development of the common law so as to include rights in a communal garden that were “essentially recreational” as affording the requisite “service, utility and benefit” to constitute an easement. On this basis, the Supreme Court considered whether or not to endorse Ellenborough Park. That case concerned a timeshare development in which the owners of the timeshare apartments as the dominant properties had an easement purporting to confer rights to a park and manor house, as servient property, with a range of sporting facilities, including a golf course, tennis courts, squash courts, a gym, a billiards room and a heated pool. In deciding on its validity, the Supreme Court took into account judgments in Canada and Australia that held that an easement could be established in relation to recreational rights. The Supreme Court held that it— “should affirm the lead given by the principled analysis . . . in In re Ellenborough Park by a clear statement that the grant of purely recreational (including sporting) rights over land . . . may be the subject matter of an easement.” However, this was subject to compliance with the other requirements for a valid easement, including that the rights genuinely benefit the dominant property. The Supreme Court took into account that sport and recreation are “a beneficial part of modern life.” The present matter is distinguishable from these two judgments in important respects. The ouster principle in English law precludes an easement that has the effect of depriving the owner of the servient property of lawful possession and control. In both judgments, the courts enquired into the question whether the servient property owners were so ousted and held that they were not. By contrast, the Bains and their successors-in-title are entirely deprived of possession in the servitude area. There is no sharing of possession of the servitude area with the owners of the servient property, or, as was the case in Regency Villas, with members of the public. In both Ellenborough Park and Regency Villas, the properties were developed in ways that the garden and the recreational facilities were permanent features serving to benefit the dominant properties. Under South African law, unlike Ms Berzack’s garden, this would have satisfied the perpetuity requirement. In Ellenborough Park, the costs of maintaining the garden were shared amongst the owners of the dominant and servient properties. No provision equivalent to section 66 of the Deeds Registries Act came into play in the English cases. Nevertheless, the development of the common law in the United Kingdom and other countries to accommodate recreational servitudes is something that cannot be ignored in this country. Arguably it would be a dystopian society that failed to recognise the utility of recreational praedial servitudes that are otherwise compliant with property law principles, particularly if they are created for the benefit of the public or communities. Our courts have shown a willingness, when called upon to do so, to develop the common law of servitudes in terms of section 173 of the Constitution so as to adapt to modern conditions. It may also be appropriate in particular circumstances to relax the strict application of one or more of the five requirements for a praedial servitude in order to develop, on a case by case basis, new praedial servitudes that are manifestly of benefit to society and are not unduly oppressive to the owner of the servient property or precluded by statute. For a number of reasons, this is not the case for a development of the law along the lines of English law or otherwise. Ms Berzack did not seek the development of the common law. Both parties pleaded and argued the case based on the common law as it stands. Before developing the common law, consideration would have to be given to whether or not a legislative amendment is a better manner of proceeding. Even if we were to develop the law in relation to the utility requirement, Ms Berzack does not satisfy the perpetuity requirement. The fact that the servitude in this case was conceived in the context of an attempt to circumvent a land use ordinance does not provide fertile ground for a significant development of the common law. And the complete ouster of the Bains’ possession of the servitude area presents a significant difficulty. I will return to this later. The upshot is that the garden servitude in question does not qualify as a praedial servitude. Section 3(1)(o) of the Deeds Registries Act provides that the Registrar of Deeds “shall, subject to the provisions of the Act . . . register any servitude, whether personal or praedial”. Section 76(1) provides that “[a] praedial servitude in perpetuity . . . may be created in a transfer of land only if the servitude is imposed on the land transferred in favour of other land registered in the name of the transferor”. In terms of these provisions the Registrar of Deeds purported to register the servitude in question as a praedial servitude upon transfer of what is now the Bains’ property from Ms Berzack to Mr Wellens on 21 September 1983. However, because the servitude did not qualify as a praedial servitude, there was no lawful basis to do so. Was the servitude lawfully registered as a personal servitude? Having found that the servitude in question was not capable of registration as a praedial servitude, the question that follows is whether the servitude was registerable as a personal servitude. Section 67 of the Deeds Registries Act provides that “[a] personal servitude may be reserved by condition in a deed of transfer of land if the reservation is in favour of the transferor”. As discussed above, Voet explains why pleasurable servitudes may not be registered as praedial servitudes but goes on to say that “it is not to be doubted that picking fruit, promenading and dining on another’s ground can without question be brought into a servitude of use, restricted in a definite manner”. From this it is clear that Voet considers that a servitude over immovable property that is for the pleasure of a person may be recognised as a personal servitude of use. However, he qualifies it by saying that it must be “restricted in a definite manner”. This requires that the servitude must be framed in accordance with its character as personal, not praedial, and in accordance with what is recognised as a servitude of use. Section 66 of the Deeds Registries Act also comes into play. It is headed “Restriction on registration of personal servitudes”, and reads as follows: “No personal servitude of usufruct, usus or habitatio purporting to extend beyond the lifetime of the person in whose favour it is created shall be registered, nor may a transfer or cession of such personal servitude to any person other than the owner of the land encumbered thereby, be registered.” Applying this provision to the servitude in this case, it is clear from the reference in clause P, the opening paragraph of the servitude, to “and [her] successors-in-title as owners of the Remainder of Erf 380 Constantia” and from the words in parentheses at the end of the servitude, “the term Seller shall include her Successors-in-Title”, that it purports to extend beyond the lifetime of the person in whose favour it has been registered. Accordingly, if the underlying servitude is one of usufruct, usus or habitatio, it has been registered in breach of section 66. Similarly, the servitude will not have been “restricted in a definite manner”, as required by the common law. There was no suggestion that the servitude was one of usufruct or habitatio. Is the servitude one of usus? In making this assessment, it seems to me that one must, in applying section 66 and having established that it cannot be a praedial servitude, ignore the components of the servitude that purport to extend beyond the life of the person in whose favour it is created and look at the essence of the right created. The essence of the right created is to be found in paragraphs (b)(i) to (iii) of the servitude, interpreted in the factual context provided by the evidence. Van der Walt describes usus as follows: “The personal servitude of use . . . is similar to but narrower than usufruct. The beneficiary of a servitude of use . . . can, like the usufructuary, use the property of another for her lifetime or for the specified term of the servitude, for her own benefit or for the benefit of her family, provided that the substance of the property is preserved and returned to the owner when the servitude of use is terminated.” In my view, the nature of the access to and conduct within the servitude area provided for in paragraphs (b)(i) to (iii) of the servitude falls squarely within the above description of usus. It simply affords Ms Berzack exclusive occupation and use of the servitude area for the purpose of gardening. She herself described it in her answering affidavit as “a right of exclusive use” (emphasis added). The fact that, as found by the High Court and all of the judges of the Supreme Court of Appeal, she uses it for the pursuit of her personal pleasure or caprice, is dispositive of it being a personal servitude, and, on the authority of Voet, is entirely compatible with the servitude of usus. There is another perspective from which the matter may be assessed. As I have found above, the terms of the servitude, interpreted in the context provided by the affidavits of Ms Berzack and Mr Wellens, confer exclusive access to and use of at least that part of the servitude area that falls behind the wooden pole fence. The Bains are entirely deprived of virtually all of the main components of the rights that are constitutive of ownership. That represents a significant burden on the property from their perspective. Silberberg and Schoeman identify as one of the criteria on the basis of which praedial and personal servitudes may be distinguished, the nature of the burden imposed by the servitude. In this regard, they state the following: “If one considers the duration of a praedial servitude, [it] imposes a significant burden on the servient tenement because the burden is in principle perpetual. However, if one considers the use and enjoyment entitlements of the servient owner, a personal servitude (like usufruct) places a more intense burden on the servient owner over a relatively shorter period of time than a praedial servitude (like the right of way or a grazing right) which places a burden of comparabl[y] lower intensity on the servient owner perpetually.” This is precisely the complaint raised by the Bains about the nature of the servitude that has been registered against their property. Their property must bear the “more intense burden” associated with a personal servitude over the period of infinite duration associated with a praedial servitude. The effect is to sterilise some 25% of their property in perpetuity (34% if one includes the full servitude area including that outside of the area gardened). With justification, they complain that Ms Berzack cannot have it both ways. Section 66, they say, is there to prevent precisely this situation. Ms Berzack’s answer to this is to point to recognised servitudes that she contends have a similar impact. In this regard, she identifies the servitudes associated with overhanging balconies and beams jutting into a neighbour’s property. However, these servitudes manifestly have a lesser intensity than the servitude she asserts here. The praedial servitude affording an exclusive right to trade in Hollman could be exercised anywhere on a servient property spanning 836 morgen in extent. The fact that the right was not exercised for over a third of a century suggests that if there had been any market for the envisaged trade, the stores were not going to take up much of the 836 morgen, and certainly not 25 to 34%. The example of a servitude allowing the owner of the dominant property to establish parking bays on the servient property is offered as another example of a praedial servitude of equal or greater intensity. For this submission, Ms Berzack relies on the judgments of the High Court and the Supreme Court of Appeal in Olive Marketing. These do not avail Ms Berzack. That matter involved a praedial servitude obliging the owner of the servient property to make available at least 250 parking bays to the dominant property (where an ice rink and cinema were located) in perpetuity. The case is distinguishable from the present matter in several respects. The servitude did not permanently sterilise the property in the hands of the owner of the servient property. It was entitled to charge for the use of the parking bays at the rate prevailing in the area. The property was in any event statutorily bound by a longstanding town planning scheme that compelled the servient owner to provide the parking bays. On the factual finding of the High Court, which was not disturbed on appeal, there was sufficient parking within the three levels of parking at the bottom of the retirement complex for both its residents and the users of the 250 bays. The utility and perpetuity requirements for a praedial servitude were plainly satisfied. The parking bays were a permanent installation, affording the necessary permanent feature and the losses sustained by the plaintiff when the first defendant breached its obligation to provide the parking bays, demonstrated the utility of the servitude. The grounds of the challenge to the validity of the servitude were different from those in this case. It did not involve a servitude for the pleasure of a particular person. In the circumstances, it is quite clear that the servitude is indeed a personal servitude of usus, which purports to have been registered beyond the lifetime of Ms Berzack. This was done in breach of section 66 and should never have been allowed by the Registrar of Deeds. The second judgment The second judgment suggests, with reference to Malan, that this judgment overlooks the fact that the decisive factor in distinguishing servitudes is whether there is a dominant property. If one considers this observation in Malan in its full context, the only point the court was making is that in a praedial servitude the servitudinal rights attach to a property, being the dominant property, whereas in a personal servitude they attach to a person, so there is no dominant property. That begs the question whether a praedial servitude has lawfully been registered in favour of the dominant property. If a servitude is registered as a praedial servitude either in conflict with the principles of property law, or in breach of section 66 of the Deeds Registries Act, the mere fact that a dominant property is referred to in the deed purportedly establishing the servitude takes the matter no further. It is precisely this enquiry as to the lawfulness of the registration in favour of Ms Berzack’s property that we are concerned with in this case. The second judgment emphasises the “clear wording” of the servitude. Whilst I do not agree that the wording is clear, it is so that the wording of the servitude reflected an intention to register and reflect a praedial servitude in the records of the deeds registry. The primary indicator of that intention is that it purports to extend beyond the lifetime of Ms Berzack in favour of her successors-in-title. But this leaves unanswered the question as to whether the servitude is lawful. That requires an enquiry in this case into whether the servitude in fact satisfies the requirements for a praedial servitude or those of a personal servitude. This must include the evidence of the parties as to how the servitude operates in practice. What I have set out earlier in this judgment demonstrates why the finding is unavoidable that the servitude in question is personal. Central to this is a factual finding of the High Court and all of the judges of the Supreme Court of Appeal that the servitude, to quote the majority, “serv[es] the pursuit of Ms Berzack’s personal pleasure or caprice”. The finding leaves no room but to hold that the servitude is personal, particularly when taken in conjunction with Ms Berzack’s own description of it as “a right of exclusive use”. And, on good authority, the more exclusive the use, the more probable that the servitude is personal, not praedial. The second judgment asserts that if the servitude was to be treated as personal, it would extend beyond Ms Berzack’s alienation and vacation of the dominant property. However, that assumes that a personal servitude always endures for the life of the servitude holder. As I have pointed out, it does not invariably do so. Given the factual circumstances here, particularly the integration of the garden on the servitude area with the garden on the dominant property, with the purpose being the pursuit of Ms Berzack’s personal pleasure or caprice, the personal servitude was bound to end when she alienated and vacated the property. Even in its form as currently registered, the servitude does not purport to give her use of the property beyond alienation. The second judgment prefers the line of academic thought that treats perpetuity as being closely related with utility and subsumed by it as a requirement. I agree that perpetuity and utility are closely related. The permanent feature of the servient property is what affords the utility to the dominant property. But to collapse them into one overlooks an essential feature of a servitude. It risks the registration of servitudes that unfairly burden the servient property owner’s land, as is the case here if the Bains are refused relief. Even if perpetuity is subsumed by utility, the servitude in this case does not withstand scrutiny as a praedial servitude for the reasons I have given. The second judgment characterises the perpetuity requirement as being whether the condition establishing the servitude is “designed to have permanent operation”. I respectfully disagree. The perpetuity requirement contemplates “some permanent feature or attribute of the servient land” that is capable of permanently fulfilling the needs of the dominant property. Its focus is on a feature or attribute of the land, not on the deed creating the servitude. The reference in this judgment to the impermanence of a garden is a reference to the absence of a permanent feature on the land constituting the servient property to satisfy the perpetuity requirement. A garden is, for the reasons given, not a permanent feature or attribute of the land on the servient tenement for the purposes of satisfying this requirement. Nor can the land itself constitute the permanent feature of the land to satisfy the perpetuity requirement. Logically that must be so. To suggest that land itself can be the permanent feature, as the second judgment also does, is to do away with the perpetuity requirement. Land is required for the existence of any servitude and it will always be there. The danger with recognising land in and of itself as providing the requisite permanent feature of the land, is that the distinction between ownership and servitudinal rights is undermined. One may then end up with precisely the unjust situation that confronts us here. The owners of the servient property in perpetuity have, to all intents and purposes, forfeited ownership under the exclusive terms of the servitude. The second judgment suggests that this judgment is at odds with my own judgment in Pickard and with that of the High Court in De Kock. This is not so. Neither judgment turned on the perpetuity requirement. The question in those cases was whether a permanent loss of the utility of a servitude resulted in the permanent loss of the servitude itself. Pickard found that it did. De Kock assumed that it did, without deciding the question, and held on the facts, that the utility of the servitude was not permanently lost. If anything, Pickard implicitly supports a perpetuity requirement, particularly if it is seen as a component of utility (as the second judgment does). On that basis a permanent loss of perpetuity means the permanent loss of the servitude. Van der Walt makes the same point. The second judgment refers to Van der Walt’s conception of the perpetuity requirement as being one that is not absolute but relative. However, in saying this, Van der Walt still requires that there be a “feature or characteristic of the servient land” that is relatively durable. Even if this were to be accepted, a planted garden is in my view not a “feature or characteristic” of the land for the purposes of satisfying the perpetuity requirement; nor is it “relatively durable” for the reasons I have given. The second judgment cites several examples of servitudes recognised by the old authorities that survive periods of interruption. This, it seems to be suggested, casts doubt upon the existence of a perpetuity requirement as understood in this judgment. It is important to consider the context in which the examples are cited by the old authorities. At least one of the examples derives from a discussion of the loss of praedial servitudes through non-use. The section of Van Leeuwen relied upon in the second judgment also pertains to the loss of servitudinal rights due to non-use for the period of prescription. This has no bearing on the perpetuity requirement. The fact that a servitude will survive non-use by the owner of the dominant property for up to thirty years, and sometimes longer in the case of a negative praedial servitude, is not an indication that there is no perpetuity requirement or that it is somehow attenuated, as I will explain. The example in question comes from the Digest. It is based on the negative praedial servitude that prevents building above a certain height on the servient property so that light may flow to the dominant property. It assumes that the servitude is not exercised by the dominant property owner, who covers their own windows for an extended period. The servitude is not lost through the conduct of the owner of the dominant property in covering their windows. This is unless the owner of the servient property has built a structure on their property blocking the light and has kept it in place for thirty years without the dominant owner acting to stop them. If one analyses this example from the perspective of the perpetuity requirement, the permanent feature of the servient property is that it provides an uninterrupted flow of light into the dominant property. That permanent feature of the servient property is not interrupted or removed if the dominant property owner chooses to block up their own windows. On the contrary, it remains in place, whether or not the dominant owner decides to unblock the windows and enjoy the servitude that is in place. The owner’s non-use does not give rise to non-compliance with the perpetuity requirement. If, however, the owner of the servient property erects a high building on it that blocks the light, the effect is to eliminate the permanent feature of the servient property. If the obstructing building remains in place for a period exceeding the extinctive prescription period, without the dominant owner acting to have it removed, the servitude is lost. It is lost because the feature allowing the uninterrupted flow of light is permanently lost. That confirms rather than undermines the perpetuity requirement as explicated in this judgment. The permanent feature must be in place for the servitude to continue to exist. Analysis of the example of an urban servitude allowing a beam to be extended into a wall of a neighbouring servient property, yields a similar result. Voet says that such a servitude is “renewed” when the house on either the servient or the dominant property falls down and is then rebuilt. The permanent feature is the supporting wall on the servient property. If only the dominant owner’s house falls down, the permanent feature of the supporting wall remains in place. If the house with the supporting wall on the servient property falls down and is rebuilt, the permanent feature has been restored and the servitude remains in place. As the second judgment points out, a river in an arid area might support a water servitude, even if the river runs dry periodically. But here the riverbed remains throughout a permanent feature or characteristic of the land on the servient property, capable of supplying the dominant property with water when the rain fills it. The access afforded by a right of way to and from a particular point over the servient property remains a permanent feature or characteristic, even if it becomes overgrown through non-use. A servitude of grazing or pasturage affords a permanent supply of natural grasses suited to feeding cattle, reviving naturally after they were depleted during a drought or destruction for some other reason. The various examples do not therefore undermine the perpetuity requirement. They confirm it. The second judgment expresses the view that the perpetuity requirement could not have survived the enactment of section 76(1) of the Deeds Registries Act because it provides for “a praedial servitude in perpetuity or for a limited period”. I respectfully disagree. All that it does is introduce the statutory option of a praedial servitude for “limited periods”. It expressly retains the option of a servitude in perpetuity and, if anything, adds statutory force to the distinct common law requirement of perpetuity. Insofar as utility is concerned, the second judgment relies on Voet’s assertion that “praedial servitudes may be granted for both pleasure and benefit”. As appears from what I have said earlier in the judgment, I agree with this statement of the law. The second judgment identifies an increase in market value deriving from both the aesthetic appeal and exclusive access to and control over the servitude area, as providing the requisite additional benefit. Here I respectfully differ. As authority for the increase in market value constituting benefit, the second judgment cites Voet’s example of a servitude that allows for water to be led from a servient property, not for irrigation but for the dominant property’s “leaping fountains” and “gently murmuring waterfalls”. Of this example, Voet says that such rights “will not be praedial servitudes for any other reason than that the price of the tenement . . . is raised because of them.” As regards Voet’s suggestion that the additional, non-aesthetic utility can be provided by an increase in the market value of the dominant property, De Waal says: “Myns insiens is dit egter te geïsoleerd en ongemotiveerd om enigsins as oortuigend aangemerk to word. ‘n Uitgangspunt só vreemd aan die oorwegende benadering tot die utilitas-vereiste sou immers meer gesag vereis as hierdie enkele stelling van Voet.” Van der Walt and other authors are also critical of this view. Referring to it as the widest approach to utility, he goes on to say— “CG van der Merwe . . . argues that the most important objection against the widest approach is that it effectively abolishes the utility requirement, which could result in the imposition of unbearable burdens on servient land and inhibit a healthy land market.” The outcome on the reasoning of the second judgment would indeed give rise to the unbearable burden to which Van der Merwe refers. There is also, with respect, a contradiction inherent in the second judgment insofar as it on the one hand finds as a source of increased value (and thereby utility) the exclusively controlled “encroach[ment]” twenty metres into the servient property, yet on the other hand hints at the possibility of further proceedings by the Bains on the basis that Ms Berzack does not exercise the servitude civiliter modo (in a civil way). If the exercise of exclusive control by Ms Berzack is unlawful on this ground, it cannot sustain the utility requirement. Nor does the potential for legal proceedings restricting exclusive control leave space for assuming that the servitude enhances the value of the dominant property. I do not join in the second judgment’s assertion that there is no scope for the development of the common law of servitudes in terms of section 39(2) of the Constitution. Indeed, it would be in conflict with section 39(2) and section 173 of the Constitution for this Court to declare in advance that it would not perform the constitutional obligations resting on it in this regard. The common law of servitudes cannot simply be cast aside because it is based on an absolutist theory of property ownership. Yes, whilst aspects such as the presumption against servitudes might be open to criticism on that basis, that is a matter that could be addressed by way of a development of the common law. The outcome in this judgment demonstrates that the common law principles retain a logic that is relevant in today’s society. In Link Africa this Court demonstrated how the law of servitudes could be harnessed in support of a socially desirable outcome. In terms of section 22(4)(c) of the Land Reform (Labour Tenants) Act, the Land Court may award a servitude to a labour tenant to protect rights of access to water and the like. Other statutes also provide for the expropriation of servitudes for the public weal. The application of the common law of servitudes together with the relevant provisions of the Deeds Registries Act in this judgment gives rise to a fair result which is not absolutist by any means. The common law of servitudes is intertwined with the common law of property, so a reconfiguration and codification would have to cover both. It would be a massive and complex undertaking, likely spawning more litigation on the interpretation of the new statute or statutes. My own view, for what it is worth, is that the time of the Executive, the Legislature and the Judiciary would be better spent ensuring, as best they are able, that the land reform provisions in section 25 of the Constitution bear the fruit that they promise. The third judgment I respectfully differ from the reasoning and proposed order in the third judgment. Amongst other things, no case was pleaded on the basis that the servitude was not exercised civiliter modo. This is not remedied by questions having been put to counsel from the bench at the hearing in this Court. The interpretation of the servitude in the third judgment does not, with respect, take into account the factual and legal context in which the servitude was established, as set out earlier in this judgment. In relation to the modification of the fencing provided for in the third judgment’s proposed order, the relief granted to the Bains by the High Court in this regard was overturned by both the majority and the minority in the Supreme Court of Appeal. In the founding affidavit in the application for leave to appeal to this Court, the Bains adopted without reservation the reasoning and order that would have been granted by the minority. That renders the question of the modification of the fencing res judicata (a matter already decided) against the Bains, notwithstanding the broad wording of the notice of application for leave to appeal. This Court is therefore precluded from entertaining that issue, as the third judgment seeks to do. Remedy What is the appropriate remedy where a court finds that a servitude has been registered without statutory authorisation to do so in terms of section 76(1) and in contravention of section 66 of the Deeds Registries Act? The Act makes no express provision for the appropriate form of relief. In Pickard, it was held that the superior courts have a broad jurisdiction under the Deeds Registries Act in appropriate cases to order changes to the title deeds. It referred to various provisions of the Act pointing to such a jurisdiction and went on to say: “Such powers derive also from the courts' inherent jurisdiction at common law and under section 173 of the Constitution. Thus in Ex parte Millsite Investment Co (Pty) Ltd Vieyra J examined the historical development of land registration and the role of the courts in that regard in Holland and the Cape from the sixteenth century onwards, and went on to hold as follows: ‘In its origin then the acts of transfer, hypothecation and similar burdening of land and all matters ancillary thereto were judicial acts. In view of the inherent jurisdiction above referred to and the historical circumstances regarding land registration it would be surprising if the Courts did not claim to have the right of surveillance over the whole system and to interpose their authority in the interests of justice. A consideration of various decided cases shows that this is what happened. Thus it has intervened to authorise rectification of deeds of transfer: see Nhlapo v Nhlapo, where a review of various cases will be found at [page] 499. . . . It seems to me, accordingly, that there is ample support for the view that the Supreme Court does possess the inherent power of interposition in matters pertaining to land registration and matters ancillary thereto.’” In those circumstances, the High Court plainly had the power to order the rectification of the deed of transfer in question by changing it to reflect the personal servitude that ought originally to have been registered, based on the evidence before the Court. This should have rendered the deed compliant with sections 66 and 76(1) of the Deeds Registries Act. However, the rectification ordered by the High Court needs to be corrected in the following respects to reflect the outcomes in the Supreme Court of Appeal and in this Court. Paragraphs 1 to 4 of the order of the High Court allowed the Bains to demolish Ms Berzack’s wooden pole fence and erect a new one along the boundary between their respective erven. Both the majority and the minority in the Supreme Court of Appeal agreed that those paragraphs should be set aside. That component of the order of the Supreme Court of Appeal was not challenged before this Court and it must stand. Paragraph 6 of the order of the High Court ordered that “clauses P(b)(ii) and (iii) shall cease to have force or effect upon the death of [Ms Berzack].” It fails to provide for the scenario where Ms Berzack alienates the property before she dies. On the order of the High Court, she would remain entitled to exercise her servitudinal rights over the servitude area, despite no longer being the owner of the property from which she accesses the garden. As has already been pointed out, this would make no sense. Paragraph 6 should also have provided for the possibility of lawful termination of the servitude by any other means earlier than her death or alienation of the property. Paragraph 8 of the High Court’s order fell short in the following respects: The High Court deleted clause P(b)(i) of the servitude when no case was made out for this relief. The High Court inserted a new clause Pbis to replace clause P(b)(i) and amend it so as to align with the relief granted in relation to the fences. Because that relief was wrongly granted, clause Pbis stands to be deleted and P(b)(i) reinstated, including the cross-references to that clause. Paragraph 10 of the order of the High Court directed that Ms Berzack pay the Bains’ costs. This needs to be revisited because the relief granted in paragraphs 1 to 4 ought to have been refused by the High Court. In my view, an order granting the Bains two thirds of their costs in the High Court would be a fair reflection of what the outcome should have been in that Court. Ms Berzack successfully appealed against paragraphs 1 to 4 in the Supreme Court of Appeal. That relief was not challenged before us. In this Court, the Bains, broadly speaking, successfully defended the balance of the relief granted by the High Court. An order that Ms Berzack pay two thirds of the Bains’ costs in the Supreme Court of Appeal would fairly reflect what the outcome ought to have been there. The Bains have been wholly successful in this Court. They are accordingly entitled to their costs in this Court. The Registrar confirms in his report that “[f]rom a registration point of view there are no objections to the order being granted as prayed”, subject to compliance with the usual requirements for any act of registration consequent upon a court order. Order The following order is made: Leave to appeal is granted. The appeal is upheld. The order of the Supreme Court of Appeal is set aside and replaced with the following order: “(1)Leave to appeal is granted. (2)The appeal is upheld. (3)The order of the High Court is amended to read as follows: ‘1.The second respondent is ordered to rectify clause P of deed of transfer No. T10518/2017 executed at the office of the Registrar of Deeds, Cape Town on 28 February 2017 so as to read as follows: ‘P.SUBJECT FURTHER to the following conditions imposed by the Transferor in favour of herself personally, which conditions are as follows: (a)The property hereby transferred is subject to a servitude area 20 (twenty) metres wide, the Western boundary of which shall be parallel to the boundary marked DE on Diagram No 5253/1981 of the Remainder of Erf 380 CONSTANTIA, held by the said Transferor, Margot Berzack (born Illman) married out of community of property to Jeffrey Cyril Berzack, under Deed of Transfer No. 38631 dated 31st December 1970. (b)The said servitude shall be subject to the following terms and conditions namely: (i)No wall or fence of any description shall be erected on the servitude boundary except extension of existing type of fencing (wooden pole fencing). (ii)The Transferor shall have the right to plant, control, care for and renew the existing garden situated within the servitude area more fully described above. (iii)The Transferor shall have full rights of access to such servitude area in fulfilment of the rights hereby granted.’ 2.It is declared that clauses P(b)(i), (ii) and (iii) of the servitude— (a)were in their original form not lawfully capable of being registered in the title deed, or of operating, as a praedial servitude over Erf 8478 Constantia, (a) in favour of the remainder of Erf 380 Constantia, or (b) beyond the lifetime of the first respondent in favour of her successors-in-title, and the second respondent erred in permitting those clauses to be so registered; (b)were in their original form and are in their rectified form, lawfully capable of operating only as a personal servitude of usus over Erf 8478 Constantia in favour of the first respondent; (c)in their rectified form shall cease to have force or effect upon the death of the first respondent, or the alienation of Erf 380 by the first respondent, or if otherwise lawfully terminated, whichever is the earlier. 3.The balance of the relief sought by the applicant is dismissed. 4.The first respondent is ordered to pay two thirds of the costs of the applicant, including the costs of two counsel.’ (4)The applicant is ordered pay two thirds of the costs of the first respondent in the applications for leave to appeal and the appeal to the Supreme Court of Appeal, including the costs of two counsel.” The first respondent is ordered to pay the applicant’s costs of the application for leave to appeal and the appeal in this Court, including the costs of two counsel. CHASKALSON AJ (Zondo CJ concurring): I have read the judgments of my colleagues Dodson AJ (first judgment) and Bilchitz AJ (third judgment). I find myself disagreeing with the reasoning of the first judgment and its conclusions on the merits of the appeal. In my view, the servitude in this case is a praedial servitude and I would therefore grant leave to appeal but, dismiss the appeal. The first judgment’s focus on the elements of permanence and utility overlooks the fact that: “It is the existence or non-existence of a dominant tenement which is the decisive factor in differentiating between personal and praedial servitudes.” In the present case, there is a dominant tenement, namely the Berzack property. The primary clause of the servitude is clearly designed to create a praedial servitude which operates in favour of the Berzack property as opposed to a personal servitude vesting rights only in Ms Berzack. It states: “The property hereby transferred is subject to a servitude area 20 (twenty) metres wide, the Western boundary of which shall be parallel to the boundary marked DE on Diagram No. 5253/1981 in favour of the Remainder of ERF 380 CONSTANTIA, held by the said Transferor, MARGOT BERZACK (born ILLMAN) married out of community of property to Jeffrey Cyril Berzack under Deed of Transfer No. 38631 dated 31st December 1970.” (Emphasis added.) While clauses (b)(ii) and (b)(iii) of the servitude might be thought to contemplate a personal servitude by vesting rights in “[t]he Seller”, the servitude makes clear that these are not rights personal to Ms Berzack because it states that “the term SELLER shall include her Successors in Title”. Quite aside from its clear wording, the garden servitude has no rationale unless it is seen as being in support of the Berzack property, not least because the garden which Ms Berzack and her successors in title are entitled to plant, control, care for and renew in terms of the servitude, is a garden which straddles both the Berzack property and the Bains property. Ms Berzack is not given the rights in respect of a self-contained garden on the Bains property. She is given rights in respect of a single integrated garden, which has always served the house on the Berzack property and a substantial part of which remains on the Berzack property. Moreover, the servitude cannot be seen as a personal servitude vesting rights in Ms Berzack because it does not vest any rights in her, independent of her ownership and occupation of the dominant tenement, namely the Berzack property. If the servitude was to be treated as a personal servitude, it would entitle Ms Berzack to enter the Bains property to plant and care for that part of the garden on the Bains property even after she sold the Berzack property and ceased to occupy it. That would clearly be an absurd result. The absurdity would be magnified by the fact that after Ms Berzack sold the Berzack property, she would no longer have the right to enter onto the Berzack property at all, still less to plant and care for any part of the garden on the Berzack property. In order to address the absurdity of this outcome, the first judgment is forced to treat the servitude as a personal servitude in favour of Ms Berzack but one which is limited to the period in which she remains owner of the Berzack property. Thus the logic of the first judgment depends on the creation of a hitherto unrecognised hybrid of personal and praedial servitudes – a personal servitude that is defined with reference to the ownership of a dominant tenement by the person in whose favour it is created. Apart from these complications which flow from treating the servitude as a personal servitude in favour of Ms Berzack, that conclusion can only be achieved by simply ignoring the express wording of the servitude that makes clear that the rights that the servitude vests in “the Seller” are rights which vest in Ms Berzack’s successors in title to the Berzack property when they take ownership of the Berzack property. What then of the permanence and utility requirements for praedial servitudes? I shall assume for the purposes of argument that permanence is an independent requirement for a praedial servitude despite the fact that there is a strong argument to the contrary. In this regard, as Van der Walt points out: “[T]he majority of South African authors currently merely emphasise that there is a close correlation between perpetuity and utility, which suggests that perpetuity is regarded as an aspect of utility rather than an independent requirement.” It is also difficult to see how a common law requirement of perpetuity for praedial servitudes could have survived the enactment of section 76(1) of the Deeds Registries Act, which states: “A praedial servitude in perpetuity or for a limited period may be created in a transfer of land only if the servitude is imposed on the land transferred in favour of other land registered in the name of the transferor, or is imposed in favour of the land transferred on other land registered in the name of the transferor.” (Emphasis added.) In any event, the notional requirement of permanence would be satisfied in this case. The wording of the servitude described makes clear that the servitude was designed to have permanent operation. That interpretation of the servitude is confirmed by the evidence of both Ms Berzack and the first owner of the Bains property, Mr Wellens, who concluded the servitude agreement with Ms Berzack. Notwithstanding the wording of the servitude and the evidence of the parties who created it, the first judgment finds that the servitude fails the “permanence” requirement for praedial servitudes because— “[a] garden is subject to the vagaries of drought, flood, hail, pests, climate change, neglect by a disinterested or absent owner or tenant and so on. Without more, it involves no durable natural feature, nor any permanently constructed feature. It requires constant upkeep, including watering, fertilising, replanting, pruning, mowing and trimming.” This reasoning in the first judgment conflates two separate issues: the permanence in relation to the servitude of the land and soil on which the garden grows, and the permanence of the garden itself. The feature of the Berzack property which is required to satisfy the requirement of “permanence” is the former, not the latter. Thus, as was pointed out in De Kock: “Indeed in Digest 8.3.13 it is stated that a servitude can be acquired for a particular type of estate, for example a vineyard, because in this case the right attaches to the soil rather than to the surface. For this reason the servitude remains in force even if the vineyard is removed.” (Emphasis added.) Moreover, logic dictates that any “permanence” requirement must be approached in this fashion. Even before climate change, the interior of South Africa has always been an arid region with periodic droughts in which perennial rivers run dry. This does not mean that no praedial servitudes allowing the flowing of water across a servient tenement into a dominant tenement could ever exist in the interior of our country. Similarly, a right of way, if not maintained, will become overgrown and unusable. That does not mean that no right of way can ever satisfy the “permanence” requirement of a praedial servitude. A garden that is neglected or harmed by drought can be revived when the drought is broken or the owner stops neglecting it. So a praedial servitude providing for a garden would not fall foul of the “permanence” requirement because of the mere possibility that it may be damaged by drought or fall into disrepair if the owner of the dominant tenement neglects it. That is generally true. It is clearly the case in the present matter where the servitude with which we are concerned is “the right to plant, control, care for and renew the existing garden”. The first judgment seeks to distinguish the permanence of the garden servitude in the present case from the permanence of servitudes of unblocked light, supporting beam extensions, rights to draw water and rights of way. In my view, there is no distinction. Just as “the riverbed remains throughout a permanent feature or characteristic of the land on the servient property, capable of supplying the dominant property with water when the rain fills it”, the land and soil remain throughout a permanent feature of the servient property capable of supporting a garden when a drought breaks or when overgrowth is cleared. It is also difficult to see how, despite its vulnerability to drought and weather, grass on land satisfies a notional “permanence” requirement of praedial servitudes when it is necessary to support the well-recognised praedial servitude of grazing, but not when it is necessary to support a garden servitude. Van der Walt treats “permanence” not as absolute but merely as requiring durability. On his review of the authorities, he sees permanence/durability as an element of the utility requirement rather than a free-standing requirement of its own. He points out: “The possibility that the feature or characteristic of the servient land that renders the servitude useful might in the future terminate does therefore not mean that the servitude is not durable and thus not useful; if that characteristic or feature appears relatively durable and therefore useful at the time when the servitude is created it satisfies the utility requirement, even though there is a possibility that the situation could change in the future. Even if it is clear from the outset, at the time when the servitude is created, that the beneficial characteristics or features that render the servitude useful to the dominant land will not endure and that they will cease to exist at a certain point in the future, the utility requirement is still met, provided that the benefit is sufficiently durable to render the servitude beneficial for the use of the dominant land (and its successive owners, rather than just one individual) for the interim.” Confirmation of the correctness of this approach to the “permanence” requirement can be found in analogous cases dealing with the related question of when loss of utility causes a servitude to lapse. In this regard, Dodson AJ’s own judgment in Pickard is instructive. In Pickard reference is made to a passage in Voet that describes how a praedial servitude that has lapsed by losing all utility for a period of time, is revived if the intervening feature that caused it to lose utility is subsequently removed “within the time prescribed for the loss of servitudes”. Consistent with this position, the conclusion reached in Pickard was that it was only when the utility of the praedial servitude had permanently ceased that the servitude would be extinguished. Voet shows with several of his other examples that servitudes do not die when the capacity of the dominant or servient property to accommodate or benefit from the servitude is interrupted for a sustained period. The Digest is similarly clear on this issue, as is Van Leeuwen. Thus, a praedial servitude lapses only when there is no prospect of any future utility being provided by the servitude, not when there is a sustained period in which the utility of the servitude is interrupted. If interruption of the utility of a praedial servitude does not lead to its termination, it would be anomalous if the mere prospect of such interruption would preclude a praedial servitude from coming into existence. I therefore disagree with the first judgment’s conclusion in relation to the permanence requirement. What then of the utility requirement? In my view, it is clear that the owner of a dominant property derives utility qua (in the capacity of) owner from the right effectively to ensure that the garden on the property encroaches 20 metres onto the servient property and is exclusively controlled and accessed by the owner of the dominant property. In this regard, the garden servitude clearly enhances the aesthetic appeal (and price) of the dominant property. There is an obvious similarity between the servitude in the present case and the praedial servitude in the example cited by Voet of a right to draw water from the servient tenement to supply fountains on the dominant tenement to enhance the aesthetic appeal and price of the dominant tenement. Indeed, in the very passage that deals with this example, Voet states that any rights granted to a tenement for the purpose of bringing enjoyment to the owner of the dominant tenement, will be sufficient to found a praedial servitude if they increase the price of the dominant tenement. The first judgment places great store on commentaries on a Digest passage stating that a praedial servitude cannot be constituted to allow the owner of the dominant tenement to pluck fruit or to promenade or dine on another’s land because such a right merely serves the caprice of the owner of the dominant property and does not increase the utility of the property. It proceeds to argue: “A number of respected South African academic writers express the firm view that a praedial servitude cannot be established for the personal pleasure of the owner of the dominant property, although it may form the basis for a personal servitude.” There is an important difference between the proposition that a praedial servitude cannot be established merely to serve the caprice of the owner of the dominant property without increasing the utility of the property and the proposition that a praedial servitude cannot be established for the personal pleasure of the owner of the dominant property. The authorities cited by the first judgment support the first proposition, but not the second proposition. Voet makes clear that the second proposition is wrong. He states that “praedial servitudes may be granted for both pleasure and benefit” and that “in a praedial servitude the benefit and enjoyment of persons can also go hand in hand with the benefit attaching to the dominant tenement”. So the utility question must be viewed objectively and without regard to the personal preferences of Ms Berzack. If the garden servitude increases the utility of the Berzack property, on the authorities, it matters not that it may have been created to ensure that Ms Berzack could continue to enjoy and to tend to her existing garden. It seems to me self-evident that in contemporary urban Cape Town the right to ensure that an existing garden is not destroyed, but is allowed to extend over an area of 1 362 m2 into the property of a servient neighbouring property, enhances the utility of the dominant property. Apart from the aesthetic considerations I have described above, the right to enjoy a substantially larger garden on a property is self-evidently a benefit to any owner of that property and cannot be characterised as an element or as a feature which would be irrelevant beyond the mere caprice of Ms Berzack. It is common cause that the servitude in the present case enhances the value of the Berzack property. The utility of a garden to a property is also supported by the reasoning of the Court of Appeal in Ellenborough Park: “No doubt a garden is a pleasure – on high authority, it is the purest of pleasures – but, in our judgment, it is not a right having no quality either of utility or benefit as those words should be understood. The right here in suit is, for reasons already given, one appurtenant to the surrounding houses as such, and constitutes a beneficial attribute of residence in a house as ordinarily understood. Its use for the purposes, not only of exercise and rest but also for such normal domestic purposes as were suggested in argument – for example, for taking out small children in perambulators or otherwise – is not fairly to be described as one of mere recreation or amusement, and is clearly beneficial to the premises to which it is attached.” (Emphasis added.) In the circumstances, I conclude that there is no basis to override the clear language of the servitude and the clear intention of the parties in the present case to create a praedial servitude. The servitude cannot be characterised as a personal servitude by contending that it does not meet the permanence or utility requirements for praedial servitudes. The Bains argued, in the alternative, that the registration of the servitude should be set aside because the servitude was designed to subvert a local land use ordinance which prescribed a minimum stand size of 4 000 m2. The common cause facts support this argument. Ms Berzack makes clear that she wanted to subdivide her original property on the line of the wooden fence, but that would have left the Bains property substantially below the 4 000 m2 requirement. The servitude was accordingly designed to circumvent the requirement in the ordinance and to give her effective control of the servitude area, depriving the owners of the servient tenement of almost all ownership rights of the servitude area and leaving them de facto (in fact) able to exercise ownership rights only over an area that was substantially less than 4 000 m2. She candidly admits that she would not have sold the Bains property without being sure that she would have “full authority and rights over the servitude area”. Counsel for the Bains argued that, on these facts, the registration of the praedial servitude was unlawful because it was part of a scheme that had the effect of violating the ordinance. The principle on which they relied was trenchantly described by Schreiner JA in his famous dissent in Collins, where he would have struck down the apartheid government’s scheme to disenfranchise “coloured” South Africans: “In general the parts of a scheme take their character from the whole. A scheme to defraud is an obvious example. Another is a scheme to get round a legislative obstacle. So in the case of the Colonial Banking and Trust Company, supra, each part of the scheme taken by itself was lawful but the combination constituted an unlawful evasion of section 48(1), and the cheque, which by itself was certainly a cheque payable on demand, was in the circumstances not a cheque at all, for the purposes of the sub-section. The same principle should apply where the obstacle is a constitutional protection against legislation and the attempted means of avoiding it is legislative. There was of course no intention in the case of the Senate Act to do anything known to be illegal, but neither was there any such intention in the Colonial Banking and Trust Company case. There the parties obviously hoped that it would turn out to be legal, yet the existence of the scheme was the decisive factor in the case.” (Emphasis added.) While there is much to commend this line of argument, I am unable to uphold it in the present case, because the parties have not been able to identify the local government ordinance which the servitude sought to circumvent. Without access to the local government ordinance that the subdivision and servitude scheme was apparently designed to circumvent, it is not possible for this Court to hold that the registration of the servitude was part of a scheme that unlawfully conflicted with that ordinance. It follows that I would dismiss the appeal. This is a dispute between two private parties, both of whom appeared to accept that costs would follow the result. Huntrex would accordingly pay the costs of Ms Berzack. I am mindful of the fact that my disposition of the appeal has the consequence that a substantial part of the Bains’ property will permanently be controlled by the owner of their neighbouring property and that they maintain that their current neighbour, Ms Berzack, exercises her servitudinal rights in a manner that is unfair to them. I express no views on whether Ms Berzack has behaved unfairly. Even if this were the case, however, we cannot find that the servitude is a personal servitude because the de facto existing state of affairs may be unfair to the Bains. The Bains may or may not be able to obtain some relief from their alleged predicament by invoking the principle that servitudes have to be exercised civiliter modo (respectfully and with due caution). In relation to the Bains’ complaints about municipal rates, objections to municipal valuations of both properties that fail properly to take account of the effect of the servitude on the respective values of the two properties may be able to provide them with adequate redress. These, however, are not issues before us in the present appeal. I accordingly express no views on them, and differ with the third judgment in so far as it seeks to provide a solution to the dispute between the Bains and Ms Berzack on the basis of issues that were not canvassed in the application for leave to appeal that was brought to this Court. The only issue before us is whether the servitude was validly registered as a praedial servitude. I would find that it was. There remains an important issue which I feel obliged to address. In the course of preparing for the hearing of this case and the writing of this judgment, I found myself spending a great deal of time reading the translated texts of 17th and early 18th century writers in Holland who were expressing their views on what they thought were the rules by which the ancient Romans organised social relations on the land. In this exercise, the 17th and 18th century writers in Holland were limited by the access they had to the surviving classical texts which had been written in a period well over a thousand years earlier. In turn, I was limited by being confined to those texts which had been translated into English, rather than being available only in the original Latin or 17th and 18th century Dutch. This process of legal research was a fascinating intellectual exercise, but it became increasingly clear to me that it was a wholly inappropriate process to use in deciding the legal rules that will govern social relations on the land in 21st century South Africa. There are three principled reasons that underlie my convictions in this regard. The first is the obvious point that our world and the worlds of ancient Rome and 17th century Holland are completely different. We need to regulate social relations on the land with reference to a set of rules that are designed for our world, and not by trying to squeeze results which we see as just for our world, from a set of rules which, to the extent that they are ascertainable, were designed for very different contexts. The law starts to look faintly ridiculous when in regulating relations between neighbours in contemporary suburban Cape Town, it has to seek guidance from 17th century Dutch or Latin commentaries on a 6th century Latin passage describing the nature of rights to pluck fruits from trees on neighbouring properties in ancient Rome. The second reason is that our entire body of common law rules applicable to servitudes are rules that derive from a legal environment with an absolutist notion of property. The principle of restrictive interpretation of servitudes is premised on the notion that absolute rights of ownership should be constrained as little as possible. The very distinction between non-transferrable personal servitudes and enduring praedial servitudes is designed to limit the types of servitudes that can “interfere” with absolute property rights beyond a single lifetime: personal servitudes terminate with the death of the person in whose favour they are created; if a servitude is to last more than a single lifetime, it has to satisfy the requirements of a praedial servitude. We are now in a post-constitutional era where South African law has rejected an absolutist notion of property. This calls for a fundamental reconsideration of the existing common law of servitudes which was constructed around a central principle that no longer represents the approach taken by South African law. Finally, there is a pressing rule of law concern. The existing law of servitudes requires legal answers to be found in an analysis of the available surviving ancient texts which may or may not have been translated into English from the original Latin or Dutch and which frequently contradict one another, and from a body of case law which lacks consistency precisely because the source and meaning of the underlying legal principles is so obscure. Almost any thorough investigation into a rule of the Roman Law and/or the 17th century Law of Holland applicable to servitudes will find authorities that are fundamentally irreconcilable with each other. By way of illustration, I refer to the valiant attempts of Kotze JP in Salmon to divine the Roman-Dutch Law applicable to the extinction of servitudes by merger (confusio) when ownership of both dominant and servient properties is acquired by the same individual who subsequently sells one of the properties: “An important question of law arises in this case, namely, whether, where a merger of a servitude has occurred, through the owner of the res serviens becoming by purchase also the owner of the res dominans, the servitude on a subsequent sale of either or both the praedia will revive? . . . in Steffens v Bam (12 CTR 1002) the point was touched upon by counsel and mentioned by BUCHANAN, J., who spoke of the rule as laid down by Voet. (8, 6, 3) as well established; but no decision in support of it was mentioned, and several of our most approved Roman-Dutch authorities, some of whom have been relied on by Mr Hutton, have expressed a contrary view. . . . There are several leges, some of which I shall have occasion to consider later on, in the Corpus Juris, which relate to instances of temporary merger, and will be found mentioned by Cujacius, ad Dig. 8, 4, lex 9; Huber, Praelect vol. 2, lib. 8, tit. 6, sec. 1; Voet, 8, 6, 3; and Glück, vol. 10, sec. 688, et in notis. Nowhere is the rule of the civil law so clearly and concisely put as by Glück, who says that servitudes are lost ‘by merger (confusio) when both praedia come to belong to one owner. Here the rule applies nemini res sua servit. If the merger again ceases, the question arises whether the former servitude revives? Simon van Leeuwen (Cens. For. 2, 14, 7) has no difficulty in answering this in the affirmative. Wolfg. Ad Schöpff, on the other hand, point blank denies this, but it is more correct to draw a distinction, as Voet (8, 6, 3) does, whether the merger has arisen through the acquisition of a purely temporary or of a permanent property (praedium).” The lack of clarity in the existing South African law of servitudes is illustrated by the first two judgments in this case. The first judgment and this judgment reach diametrically opposed conclusions. Yet both judgments are able to marshal detailed references to the Roman-Dutch authorities on servitudes to support their conclusions. There are very few trained lawyers who would be able to explain the common law of servitudes with any confidence. For lay persons who are bound by the law of servitudes, the law must seem utterly impenetrable. The rule of law requires law to be clear and accessible. Our current law of servitudes falls a long way short in this regard. It is a branch of the law which serves to mystify law and to preserve it as a domain that is intelligible only to a chosen few. For anyone else to understand their rights and obligations in relation to servitudes, they must be able to access the wisdom and pronouncements of those chosen few. This is fundamentally offensive to the rule of law. There was no request from the parties in this appeal to develop our law of servitudes in accordance with section 39(2) of the Constitution. In any event, the problems with our law of servitudes that I have described above are fundamental. They cannot be cured under section 39(2). They require the thoroughgoing reconfiguration of the existing law of servitudes, not its incremental development. The existing common law that was apparently developed to deal with social relations on the land in Ancient Rome and 17th century Holland under an absolutist theory of property ownership must be replaced with a publicly accessible set of rules that are designed appropriately for 21st century South Africa and that are at least written in an official South African language. This is a task that requires speedy attention, but it is a task for the South African Law Reform Commission (SALRC) and/or Parliament, not for the courts. The first judgment suggests that Parliament and the SALRC should not waste their time investigating and codifying a new law of servitudes because land reform is a more pressing priority. I would hope that our organs of state could find time for both tasks. Conclusion If I had commanded the majority, I would have granted the application for leave to appeal, and dismissed the appeal with costs, including the costs of two counsel. BILCHITZ AJ: Introduction The dispute in this case had its origins in a request by the applicant (Huntrex, which is wholly owned by Samantha and Rodney Bain) to the respondent (Ms Berzack) to replace a wooden fence to prevent the dogs they care for from escaping through it. The fence was erected broadly to demarcate the division between that area of the Huntrex property over which Ms Berzack held a servitude and the rest of the Huntrex property. The refusal of Ms Berzack to agree to this request led to litigation in which the Bains sought to have the existing fence on the servitude boundary demolished and replaced with a fence on the boundary between the two erven with a gate providing Ms Berzack with access to the area over which she held a servitude. The case itself involved a contestation about the nature of the servitude – as to whether it was praedial or personal – as well as the interpretation of the content of the servitude. I have had the pleasure of reading the judgments of my Colleagues Dodson AJ and Chaskalson AJ. I am persuaded by Chaskalson AJ’s reasoning that the servitude in this case was intended to be praedial in nature and is able to meet the objective legal principles that apply in that regard relating to perpetuity and utility. I also agree with the reasons he provides why the law of servitudes is ripe for an engagement by Parliament and that it needs to be transformed in light of the conditions of modern-day South Africa as well as the Constitution. Despite my agreement with Chaskalson AJ on these matters, I depart from both him and my Colleague Dodson AJ in their conclusion that a finding that the servitude is either praedial or personal is sufficient to resolve this matter. I am particularly concerned that neither judgment addresses the problem that lay, in my view, at the heart of this dispute, namely, whether it was permissible for Ms Berzack to exercise the servitude in an absolutist manner without regard to the interests of the Bains and the companion animals (or indeed other vulnerable parties such as children) they care for. Without addressing that issue, after years of working its way through the court system, the dispute will not be resolved and could result in further litigation. The consequences of the first judgment’s finding that the servitude is personal are that it will end upon Ms Berzack’s passing or, as stipulated by the first judgment, upon the sale of her erf (whichever is the earlier). If, as the second judgment holds, the servitude is praedial in nature, it can continue to burden the servient property in perpetuity, effectively preventing the Huntrex owners (or their successors-in-title) from enjoying a large portion of their property. To avoid the unfairness involved in allowing a burdensome servitude to continue indefinitely, it is tempting to find that the servitude is personal in nature. Yet, in my view, that is to go against the express intention of the parties as captured in the servitude, to adopt an unduly restrictive approach to the objective characteristics of a praedial servitude and to cause serious disruption to the existing registration of servitudes. Indeed, evidence was presented in the record – from the affidavits of Mr Moffat and Ms Boag – of many praedial servitudes having been registered that are similar in nature to that of Ms Berzack. The first judgment is likely, as a result, to cause serious disruption to existing servitudes – if there is to be a significant reform of the existing law of servitudes, that would be better accomplished by a parliamentary intervention as indicated by my Colleague Chaskalson AJ which can also create transitional provisions to address existing servitudes. The first question I consider is the content of the entitlements that are granted by this servitude. I depart from my Colleague Dodson AJ in that I consider certain clauses of the servitude itself to be ambiguous which thus requires the application of the principle that a servitude must be interpreted in the least restrictive manner. In my view, this principle, in the context of the current servitude, entails that it does not confer on the servitude holder (currently Ms Berzack) an exclusive right of access and control over the servitude area. Such a recognition, however, would still not solve the problem that prompted this litigation – which related to the permeability of the fence on the Bains’ property that enabled the dogs they cared for to escape into the servitude area and onto Ms Berzack’s property. To address this aspect, it is important to have regard to the exercise of the servitude itself. The central common law principle here is the duty on a servitude holder to exercise a servitude civiliter – that is, in a manner that does not impose an unreasonable burden on the servient owner (in Latin – civiliter modo – often translated literally as “in a civil manner”). This Court should, on the usual casuistic basis, develop the common law in specific instances whether or not there is a major overhaul by Parliament. In my view, the Constitution requires this Court to do so. The civiliter modo principle should be suitably developed in light of our Constitution to require the servitude holder to consider the interests not only of the servient owner but of all affected by the exercise of their entitlements including unlawful occupiers, vulnerable children or animals. That, in turn, provides a constitutionally informed approach both to temper any unfairness inherent in a praedial or personal servitude as well as to resolve the dispute in this case. Characterisation of the legal issues One of the differences between this judgment and the other two judgments is in the characterisation of the legal issues in the dispute in this Court. The other two judgments focus mainly on whether Ms Berzack’s servitude is praedial or personal in nature. This judgment considers that the dispute concerned not only that question but also the interpretation and exercise of the servitude itself. My view is supported by the original pleaded case in the High Court. The case concerned not simply whether the servitude was praedial or personal but also whether the owners of the servient property, the Bains, were permitted in terms of the existing servitude to demolish the existing fence and erect another one on the boundary between Ms Berzack’s property and their own. Ms Berzack also lodged a counter-application to confirm her interpretation of the servitude as granting her exclusive access to the servitude area. The Huntrex owners were successful in the High Court and the servitude was interpreted to allow them the relief they had requested. The appeal to the Supreme Court of Appeal concerned both the question of whether the servitude was praedial or personal as well as the interpretation of the servitude. Whilst diverging on the first issue, both the majority and minority judgments rejected the relief relating to the fencing, effectively interpreting the servitude in favour of Ms Berzack. The notice of application for leave to appeal to this Court clearly is against the entire judgment and orders of the majority of the Supreme Court of Appeal – that includes its finding on the interpretation of the servitude. There is also a more general request for further and/or alternative relief. The question of the interpretation of the servitude thus remained an issue in this Court. It was raised in the oral hearing but also, importantly, was at the heart of this case since its commencement in the High Court. Indeed, the first judgment, prior to ascertaining whether the servitude is praedial or personal, provides an interpretation of the content of the servitude. This is unavoidable as it is necessary in order to determine whether the servitude was personal or praedial. Thus, in my view, the interpretation of the servitude cannot be divorced from a consideration of whether it is praedial or personal. Therefore, the matter could not be res judicata (the matter is already decided) as it is integrally connected to the question that the first and second judgments regard as central. Interpreting a servitude restrictively My Colleague Dodson AJ recognises that there are several interpretive presumptions that generally apply in favour of not burdening a property with servitudes or limiting their extent. This approach was given expression to in the Willoughby’s case as follows: “Whether a contractual right amounts in any given case to servitude – whether it is real or only personal – depends upon the intention of the parties to be gathered from the terms of the contract construed in the light of the relevant circumstances. In case of doubt the presumption will always be against a servitude, the onus is upon the person affirming the existence of one to prove it.” As my Colleague correctly writes, this approach leads to three presumptions. The relevant presumption here is that referred to by the Supreme Court of Appeal in Joles Eiendomme as the “well-established rule of construction that because a servitude is a limitation on ownership, it must be accorded an interpretation which least encumbers the servient tenement”. The presumption applies where the terms of the servitude are not clear and unambiguous. I agree with Van der Walt that the presumption should also apply in circumstances of constructing a servitude that imposes a significant burden on the servient property where the terms thereof are wide and general rather than precise and specific: in such circumstances too, a servitude should be interpreted in such a way so as to create the least onerous burden on the servient property. In my view, this is a case of a servitude that does not unambiguously create a right of exclusive access for the servitude holder to the servitude area. It is also a servitude that, if interpreted to grant exclusive access, would almost completely deprive the servient property owners of their enjoyment of that part of their property. The presumption thus operates in favour of a restrictive reading of the servitude. The relevant part of the current servitude reads as follows: “The said servitude shall be subject to the following terms and conditions namely: (i)No wall or fence of any description shall be erected on the servitude boundary except extension of existing type of fencing (wooden pole fencing). (ii)The Seller shall have the right to plant, control, care for and renew the existing garden situated within the servitude area more fully described above. (iii)The Seller shall have full rights of access to such servitude area in fulfilment of the rights hereby granted. (The term of Seller shall include her successors-in-title).” A plain reading of the wording of the servitude is that it grants the servitude holder a right of continuing access to the servitude area and a right to plant, control, care for and renew the garden in that area. There is no reason why the enjoyment of these rights necessarily requires excluding the owners of the servient property from the servitude area. In fact, guaranteeing a right of access to the servitude holder specifically suggests that the owners of the servient property may seek to prevent such access due to their desire to enjoy that part of their property. Moreover, the right to garden on a property in no way excludes the owners of the servient property from enjoying that garden. The word “control” is not related to access but to the “existing garden” – which indicates that the servitude holder has the right to determine, for instance, what is planted in the garden and its aesthetic properties. The first part of the servitude prohibits erecting a wall or fence on the servitude boundary except in relation to the extension of the existing type of fencing. This provision is consistent with the removal of any fencing between the servitude area and the rest of the servient property – there is indeed no prohibition on removing existing fencing. Whilst there is an existing fence between the servitude area and the rest of the servient property, that appears to be for purposes of demarcating the servitude area rather than preventing the owners of that property from accessing the area – otherwise, logically, there would be a prohibition on removing that fence. The type of fencing (wooden pole fencing) is specified in a consistent manner with the rest of the servitude – namely, to preserve an aesthetic sensibility (consistent with the essence of the servitude which surrounds gardening). It is conceivable that much stronger and permanent fencing could have been utilised if the goal was to keep the owners of the servient property out of the area. Whilst evidence was presented that in the past the servitude was exercised by Ms Berzack in a manner that involved her having exclusive access to that area, such an entitlement in no way emerges clearly from the terms of the servitude itself. It is precisely because the servitude is unclear on this point that Ms Berzack counterapplied in the High Court to have a ruling specifying that the servitude conferred on her – as the servitude holder – exclusive rights over the servitude area. Yet, the law clearly provides that the opposite approach must be adopted: where there is ambiguity, servitudes must be interpreted restrictively so as to burden the servient property no more than is necessary. The interpretation of this servitude not to confer on the servitude holder a right of exclusive access to the servitude area would enable the servient property owner to gain access to that area and to enjoy that part of their property too. That interpretation preserves the essence of the servitude but also does not result in the obliteration of all rights the servient property owners (currently the Bains) have over that portion of their property. The alternative would be to exclude them entirely from enjoying part of their property which they own. That consequence would arise irrespective of whether the servitude is personal or praedial. My approach is similar to that adopted by Griesel J in Roeloffze. That case related to a right of way and the respective rights of the servitude holder as well as the owner of the servient property. Griesel J laid out some of the applicable principles in the following quote: “In the present case the phrase in question must be interpreted in the light of the servitude as a whole, seen against the background of well-established common-law principles. Applying those principles, the rights of the owner of the servient property must not be unduly burdened by the servitude. A strict interpretation of a servitude is therefore called for. Moreover, on my understanding of the law, the creation of a servitude does not preclude the owner of the servient tenement from using the land which is the subject matter of the servitude; he or she ordinarily retains all the rights flowing from his or her ownership of such property, provided that the exercise of such rights may not interfere with the rights of the servitude holder.” As I have mentioned, it is unclear to me why in order to exercise the central elements of her servitude, Ms Berzack must exclude the Bains from retaining any rights associated with the ownership of their property. It is important, when interpreting servitudes, to do so in a restrictive manner such that, where possible and consistent with the terms of the servitude, a court does not extend the servitude holder’s entitlements in such a way so as to deprive the servient property owners of all the entitlements flowing from the ownership of their property. The interpretation I have adopted, however, does not address the problem that brought about this litigation – whether the fencing can be modified to enable the Huntrex owners to prevent the dogs they care for from escaping into the area over which Ms Berzack holds a servitude and, without any barrier, onto her own property too. The existing servitude provides that no wall or fence may be erected on the servitude boundary “except extension of existing type of fencing (wooden pole fencing)”. This provision clearly provides that the existing fence on the servitude boundary may only be modified through “extension” – that could include the augmenting of additional fencing or the modification thereof in other ways. The existing type of fencing could involve wooden pole fencing but also could include fencing of a similar nature that would have a similar aesthetic effect. The latter understanding seems to be more consonant with the purpose of the servitude which is to enable the servitude holder to create a beautiful extension of their garden onto the adjoining property. Given these various possibilities, exactly which modifications would be permissible can properly be understood to lie within the discretion of the servitude holder – in this case, Ms Berzack. When the dispute first arose, the Bains asked Ms Berzack for permission to make additional modifications to the fencing to protect the dogs they care for. Ms Berzack refused to permit any modification to the existing fencing – that refusal was an exercise of the powers conferred upon her by the servitude. Yet, in the exercise of her powers in terms of the servitude, she was not permitted to adopt an absolutist and intransigent approach to the interests of the servient property owner or those on their property. Indeed, the exercise of Ms Berzack’s rights, implicates the common law principle that a servitude must be exercised in a reasonable manner, a matter which I now elaborate on. The duty to exercise a servitude reasonably The common law of servitudes recognised that the exercise of a servitude by a servitude holder without regard to the interests of the owner of the servient property could result in excesses that unreasonably interfered with the enjoyment of the servient owner’s property. It therefore developed the principle that the servitude holder must exercise their servitude entitlements reasonably – and in a way that has regard to the interests of the servient owner (civiliter modo). In Anglo Operations, the Supreme Court of Appeal formulated the principle as follows: “The holder of the servitude is in turn bound to exercise his rights civiliter modo, that is, reasonably viewed, with as much possible consideration and with the least possible inconvenience to the servient property and its owner.” Van der Walt captures the purpose of the principle as follows: “The principle protects the owner of the servient property against unnecessary and unwarranted burdens, which in this context means burdens that are neither required for the proper exercise of the servitude nor clearly specified or included by implication in the servitude grant.” The principle includes a key element of balancing the interests of the servitude holder and the servient property owner. That balancing, in our constitutional era, includes facets of a proportionality test: the servitude holder must be able to exercise their servitude effectively but do so in a way that is not overly burdensome on the servient property. A right of way, for instance, must follow a route that causes the least inconvenience to the servient owner. These principles embody an acknowledgment that a servitude itself is essentially relational in nature – it involves recognition of a limited real right that can be exercised over property that someone else owns. That requires exercising such servitude entitlements with due concern for the entitlements of others. The civiliter modo principle is consistent with and warrants development in our constitutional era. It was referred to by Yacoob J in Motswagae which concerned the right of a municipality in terms of a servitude to conduct public works on a property in relation to the provision of public services. Given the significant intrusion that bulldozing caused near the property, the Court emphasised the common law principle that a servitude be exercised civiliter modo which it translated to mean “respectfully and with due caution”. In Link Africa, Cameron J and Froneman J were also concerned with a public servitude. They too recognised the principle that a servitude must be exercised respectfully and with due caution. In their judgment, they reject what they term an “outdated, over-rigid and absolute notion of ownership. That conception is alien to the holding of property under the common law, the Constitution and other applicable law”. Instead they state: “this Court has recognised that property as an individual right is not absolute. It is subject to societal imperatives”. The approach of the Court in these cases recognises that property rights – and servitudes in particular – are essentially relational in nature. Since the law plays an essential role in the recognition of property rights, it must do so in such a way that achieves a balance between the various interests involved. In his concurring judgment in Daniels, Froneman J recognises “[t]he social boundedness of property in our current law has also been recognised and emphasised in many other areas of our law”. The idea of property rights being exercised in isolation from society, other individuals and the wider environment is also at odds with the transformative shift the Constitution has brought about that requires a consideration of African and feminist approaches to property. For instance, Okoth-Ogendo writes, that the key shift required by an African approach to property is to consider “how individuals on their own, or in community with others, relate not simply to the physical solum, but to each other in respect of that solum and its associated resources”. Such approaches are relational in nature and appreciate our interdependence with other humans, animals and the environment. Recognising the need for a shift in the relationship between humans and the environment induced by the climate crisis, Nedelsky writes: “To embrace a conception of a genuinely shared Earth, humans would see themselves as embedded in an interdependent community of life forms as well as water, air, soil, rocks, and minerals (Davies, Godden and Graham 2021). The goal of a property regime would then be to organise relationships of care and respect between humans and all other members of the Earth community.” A shift brought about by our Constitution in the approach to property thus requires a rejection of absolutist property rights which are exercised without due concern for those affected by the exercise of such rights. Arguably, the common law – through principles such as the duty to exercise a servitude civiliter modo – already recognises the relational dimension of property. In developing the common law, it is necessary to expand upon that underlying foundation. Developing the common law Section 173 of the Constitution provides that “[t]he Constitutional Court, the Supreme Court of Appeal and the High Court of South Africa each has the inherent power to protect and regulate their own process, and to develop the common law, taking into account the interests of justice”. Whilst that section is permissive and recognises the power of the superior courts to develop the common law, section 39(2) provides in more mandatory terms that “[w]hen interpreting any legislation, and when developing the common law or customary law, every court, tribunal or forum must promote the spirit, purport and objects of the Bill of Rights”. In Carmichele, this Court held that “there can be no question that the obligation to develop the common law with due regard to the spirit, purport and objects of the Bill of Rights is an obligation which falls on all of our courts including this Court”. It further held “that this duty upon Judges arises in respect both of the civil and criminal law, whether or not the parties in any particular case request the Court to develop the common law under section 39(2)”. These are duties on the Court irrespective of whether the case has been framed as requiring a development in terms of section 39(2) – indeed, to do that would tether common law development to whether it is advantageous for parties to argue for such a development. The Court must also apply the law holistically and not allow parties to deflect from the true nature of a dispute. It must also consider whether the interests of those who are vulnerable – and who lack a voice – have adequately been addressed. As I have already discussed, the issue surrounding the nature of the fencing was clearly at issue since the commencement of this litigation in the High Court. Moreover, at the hearing in this Court, the question of the approach adopted by Ms Berzack to the servitude in light of the civiliter modo principle was raised in oral questions, providing the parties with an opportunity to engage this issue. The common law approach to exercising a servitude civiliter modo applies with respect to the interests of the servient property owner – in this case, there would be a duty on Ms Berzack to consider the interests of the Bains and both their right and duty to ensure the safety of the companion animals they care for. This Court has recognised in Dawood that the right to dignity protects the right of individuals to enter into meaningful personal relationships and thus “to achieve personal fulfilment in an aspect of life that is of central significance”. For many human individuals, their relationships with other sentient creatures and the caring, compassionate bonds that such relationships can entail are amongst the most centrally important dimensions of their lives. The Bains thus have a constitutionally protected interest in choosing to form and sustain relationships with companion animals and to ensure that they are not harmed and that they are adequately provided for. The Bains also have statutory responsibilities in terms of the Animals Protection Act to avoid unreasonably committing any act or omitting to do any act that could result in an animal being subject to unnecessary suffering. Failing to act reasonably to prevent the companion animals they care for from escaping their property and being subject to the dangers of an adjoining property (where there is a swimming pool, for instance) could contravene this statutory duty, and result in a criminal conviction. However, given the relational nature of property, the exercise of entitlements in relation to a servitude can affect the interests of parties other than the servient property owner. The question thus arises as to whether a servitude holder must also, in exercising their servitude reasonably, consider the interests of those who can be affected but who are not the servient owner. Consider, for instance, if the Huntrex property owners had young children who could easily crawl through the fence and fall into the swimming pool – would Ms Berzack have a duty to consider the danger to the children in her decision concerning whether to permit the modification of the existing fencing? It could also be that there are unlawful occupiers on a property who temporarily have a right to remain on the servient property pending the provision of alternative accommodation – must a servitude holder consider their interests and those of their children (even if they diverge from that of the servient tenement’s owner)? In this case, we are presented with a novel scenario. The Bains initially became aware of the praedial nature of the servitude and challenged it pursuant to a concern that the current fence was inadequate to prevent the escape of the dogs they care for. In the event the dogs did escape, the Bains were concerned that the fence inhibited their ability to retrieve those animals. I have thus far addressed the interests of the Bains in relation to the companion animals they care for. However, in our constitutional era, the dogs themselves have important interests that must be considered. In Openshaw, Cameron J wrote the following about animal protection laws: “The statutes recognise that animals are sentient beings that are capable of suffering and of experiencing pain.” In Lemthongthai, the Supreme Court of Appeal further held that “[c]onstitutional values dictate a more caring attitude towards fellow humans, animals and the environment in general”. This Court in NSPCA built on these dicta to find that, in the constitutional era, “the rationale behind protecting animal welfare has shifted from merely safeguarding the moral status of humans to placing intrinsic value on animals as individuals”. If dogs are beings with intrinsic value, their interests must be considered too in any balancing process relating to how a servitude holder may exercise their servitude. Dogs are not simply inanimate objects: if the fence is left as it is, the dogs cared for by the Bains can escape onto Ms Berzack’s property. Outside of their own property, the Bains are not able to control the dangers those animals are exposed to – such as in relation to the swimming pool – that can result in injury or the loss of life. It may also be possible for the animals to escape onto other properties and public spaces which, in urban settings, may result in injury or death. The inability to recover the animals easily in the servitude area may also prevent the Bains from realising their obligations to ensure the animals they care for can access adequate food and water. There are thus multiple reasons why it is in the interests of both the Bains and the dogs they care for to ensure that there is a barrier through which those dogs cannot escape. Applying these principles In this case, when approached, Ms Berzack simply refused to modify in any way the wooden fence that was erected between the servitude area and the rest of the Bains’ property. A central question this case poses is whether that refusal was an exercise of the servitude civiliter modo. In my view, it was not. Through consenting to a modification of the fence, Ms Berzack would not have been disturbed in her enjoyment of her garden. There are multiple variations of what could have been done: this could have included modifying the existing fence to become sturdier and less permeable to a replacement of the wooden fence with material of a similar nature that would have been more effective in keeping the dogs in the Bains’ property. It would not be reasonable to require the Bains to erect additional fencing on their own property given that the servitude already significantly burdens and divides their property. Any modification of existing fencing would be entirely consistent with what the servitude sought to accomplish. It would lead to virtually no reduction in the enjoyment by Ms Berzack of her entitlements whilst enabling the Bains to enjoy their property to the maximum and take account of the interests of the dogs they care for. A blanket refusal to consider any such changes was thus unreasonable and inconsistent with an ethos requiring the exercise of property rights in a socially responsible manner. Consequently, in my view, the recognition that Ms Berzack has a praedial servitude must go hand in hand with a restrictive interpretation of the servitude as well as a recognition of her duty to exercise the servitude reasonably and with due regard to the interests of the Bains and the dogs they care for. Consent to the modification of the fence would have avoided this litigation and would have conformed both to the wording of the servitude and Ms Berzack’s common law obligations – now developed in the constitutional era – to exercise the servitude civiliter modo with due regard to all the interests involved. In relation to the order that can be issued by this Court, I wish to address the suggestion by the first judgment that the question of the modification of fencing is res judicata. I have already expressed my view that the interpretation of the servitude was before this Court, both in terms of the pleadings – such an order would clearly fall within the terms of the notice of application for leave to appeal – and in relation to the fact that there is an integral connection between the interpretation thereof and a finding concerning whether the servitude is praedial or personal. Given that essential connection, it is also entirely permissible for this Court to grant relief that relates to the interpretation of the servitude. The first judgment also cannot, consistently, hold this view: even though there is no express provision that the servitude would terminate on the alienation by Ms Berzack of her property, the first judgment justifies effectively rectifying the servitude so as not only to terminate upon her passing, but also upon the alienation of the property. An order is made to that effect. This demonstrates that an order concerning the interpretation of the servitude is entirely appropriate. I have already provided a justification for providing an order relating to the exercise of the servitude civiliter modo. The question of relief in this case raises a wider matter of this Court’s powers when exercising its general jurisdiction. In constitutional matters, it is clear that the Constitution provides in section 172(1)(b) that a court “may make any order that is just and equitable”. In relation to its general jurisdiction, courts are governed by section 173 which confers “the inherent power to protect and regulate their own process, and to develop the common law, taking into account the interests of justice”. This provision must be understood to confer a similar power on this Court to provide just and equitable relief where its general jurisdiction is engaged. Its common law powers in this regard must clearly be exercised “in the interests of justice”. This is particularly important when the interests of a vulnerable party – such as a child or nonhuman animal – are at stake. It seems to me that a court should not be shoehorned by the parties to address a narrow question with the result that its finding leads to relief that entrenches an unjust or inequitable state of affairs which can imperil the interests of vulnerable individuals. That is inconsistent with a court’s central duty to do substantive justice between the parties. In relation to this case, even if the focus of argument was on whether the servitude was praedial or personal, that fact does not preclude this Court from ensuring that a finding either way does not result in unjust and inequitable consequences or addressing the interests of vulnerable individuals affected by its order. Had I commanded the majority, I would have concluded that the servitude is a praedial one but ordered, in addition, the following: Clauses P(b)(ii) and (iii) of the servitude are to be interpreted such that the servitude holder has all the rights contained therein but does not have exclusive rights of access to the servitude area. Clause P(b)(i) is to be interpreted to enable the servient property owners to have access to the servitude area up to the boundary of their property. Any use the servient property owners make of that area must not interfere with the full enjoyment by the servitude holder of their rights under the servitude. The servitude holder must exercise the servitude civiliter modo which, in this instance, requires the granting of permission to modify the existing fence to create a barrier that does not allow the companion animals cared for by the Bains to escape into the servitude area. The modification must also be consistent with allowing the Bains to gain access to the servitude area. Seeing that both parties have been partially successful, each party would have been responsible for their own costs in this Court. For the Applicant: For the First Respondent: S P Rosenberg SC and T R Tyler instructed by Lamprecht and Associates Incorporated. J G Dickerson SC and S G Fuller instructed by Dorrington Jessop Incorporated.