Municipal Employees’ Pension Fund v Chrisal Investments (Pty) Ltd
In considering the actio communi dividundo the SCA decided there is no closed list of instances of bound co-ownership and if the ‘relationship gives rise to bound co-ownership the co-ownership will endure for so long as the primary extrinsic relationship endures. Once it is terminated then . . . . it will become free co-ownership and be capable of being terminated under the actio. This judgment is particularly important and interesting because numerous relationships concerning common ownership are discussed, such as partnerships and joint ventures as well as sectional title owners who share in the common property.
‘ In summary therefore, I conclude, in accordance with the authorities discussed above, that the distinction between free and bound co-ownership is that in the former the co-ownership is the sole legal relationship between the co-owners, while in the latter there is a separate and distinct legal relationship between them of which the co-ownership is but one consequence. Co-ownership is not the primary or sole purpose of their relationship, which is governed by rules imposed by law, including statute, or determined by the parties’ themselves by way of binding agreements. The relationship is extrinsic to the co-ownership, but is not required to be exceptional. In other words it requires no special feature for the co-ownership consequential upon the relationship to qualify as bound co-ownership. Whether it is depends upon the terms upon which the relationship is constituted. The mere fact that co-owners decide to exploit their co-ownership commercially will not of itself constitute the co-ownership as bound co-ownership. That will depend upon the nature and terms of the commercial agreement between the parties and matters such as the provision made for its termination.”
Quotations from judgment
Note: Footnotes omitted and emphasis added
 Three companies (the Adamax co-owners), which owned a business operating three shopping centres, disposed of a 55 percent share in that business to the first appellant, the Municipal Employees’ Pension Fund (the MEPF), for a total price of R550 million. Simultaneously with that disposal their holding company, the fourth respondent, Adamax Property Projects Menlyn (Pty) Ltd (Adamax), concluded a detailed co ownership agreement (the COA) with the MEPF.
This agreement encompassed
- the distribution of income from the business;
- the incurrence of costs and other obligations;
- the constitution of an executive committee;
- property and financial management of the business; and
- the costs of managing the shopping centres.
The COA was firmly tied to the sale agreement by way of a condition precedent that made the sale agreement dependent on the conclusion of the COA. It contained detailed provisions in regard to its duration and the manner in which either party might dispose of their interest in the business.
 Was it open to the sellers, the day after these agreements had been concluded and implemented and at all times thereafter, to ignore all those detailed contractual arrangements and bring proceedings under the actio communi dividundo to terminate the joint ownership of the business and cause the properties on which the shopping centres stand to be sold?
The high court’s judgment answers that question in the affirmative. I do not.
 The first, second and third respondents are the Adamax co-owners. They owned the shopping centre business and the immovable properties on which the Parkview, Glen Village South and Glen Village North shopping centres are constructed. The precise relationship between them and even whether each one owns the properties on which one shopping centre stands does not emerge from the papers. What is clear is that they are merely corporate vehicles through which Adamax conducted the shopping centre business. The other appellants are the administrator of the MEPF and an associated company. They play no role in the legal issues in this case.
 The two agreements that are relevant to this case are the sale agreement and the COA. There was also a Property Management Agreement governing the management of the shopping centres, but that was itself a by-product of the other two agreements. The subject of the sale agreement was ‘the Enterprise’, which was defined in clause 2.6.4 as:
‘the letting enterprise carried on by the Sellers as a going concern on the Properties consisting of the Properties and all right, title and interest in and to the Leases.’
Clause 2.6.12 defined the ‘Properties’ as a ‘55% undivided share in the properties owned by the Sellers’, that is, the first to third respondents. Clause 2.6.6 defined the ‘Leases’ as the written agreements of lease in respect of the letting and hiring of the premises forming part of the Properties.
 In terms of clause 4.1 the Sellers sold the Enterprise to the MEPF for a price of R550 million. Clause 4.3.8 said that the Properties and the Leases were an ‘enterprise’ as defined in section 1 of the Value-Added Tax Act 89 of 1991 and clause 4.3.9 provided that:
‘such enterprise is being disposed of as a “going concern” within the meaning of Section 11(1)(e)(i) of the VAT Act.’
The significance of an enterprise being disposed of as a going concern is that it then attracts a zero rating for VAT purposes. In order to qualify for this in terms of the VAT Act the enterprise must be an income earning activity. A business operating three shopping centres qualifies as such.
 Furthermore, clause 3 recorded that the sale constituted a merger between the MEPF and the first to third respondents for the purposes of the Competition Act 89 of 1998 and required the approval of the Competition Commission. In terms of the definition in s 12 of the Competition Act a merger arises where one firm directly or indirectly acquires or establishes direct or indirect control over the whole or part of the business of another firm. This was a merger because the MEPF acquired control, as defined in the Competition Act, over the business of the Adamax co-owners.
 Importantly, these provisions demonstrate that under the sale agreement the MEPF did not merely purchase certain immovable properties, but acquired an interest in the business of operating the three shopping centres (the Letting Enterprise). Its principal tangible assets were the shopping centres themselves and the leases with tenants that provided the revenue stream of the business.
The importance of the leases is apparent from clause 5 under which the sellers provided the MEPF with a guarantee that the net income of the MEPF’s 55 percent undivided share in the Letting Enterprise would amount to at least R49.5 million per annum for the first four years. Without the leases the shopping centres would have been white elephants with little commercial value. The leases and their income earning potential were what transformed the buildings from empty shells into a business valued for the purposes of the sale agreement at R1 billion.
Leases to suitable tenants, whose wares and services were aimed at the target market of shoppers and visitors to the business premises in the centres, would generate footfall through the centres and create the goodwill attaching to them. Their importance was demonstrated by the extensive warranties concerning the existence and commercial viability of the leases given by the Adamax co-owners to the MEPF in clauses 9.3, 9.4 and 9.5 of the sale agreement.
In turn, without the right to occupy the centres and turn them to commercial account, the shopping centre business and the letting enterprise that was the subject of the sale would not have existed. Counsel for both parties agreed that, in the words of an old song, ‘You can’t have one without the other.’
 I stress that the subject of the sale agreement was the business described as the Letting Enterprise, although the immovable properties together with the leases constituted the Letting Enterprise. However, the application was brought on the premise that it concerned only the joint ownership of the properties on which the shopping centres are situated.
The deponent to the founding affidavit, Dr de Muelenaere, said:
‘The purpose of this application is to obtain an order under the common law in terms of the actio communi dividundo to dissolve the current co-ownership between the Adamax Co-owners and the MEPF and to provide for the structured unbundling of the common property.
As will appear more fully herein below, the Adamax co-owners and the MEPF are co-owners of immovable property in the form of three shopping centres in Pretoria. The MEPF is the owner of a 55% undivided share in the common properties and the Adamax co-owners are the owners of the remaining 45% share in the common property.’
 In referring to co-ownership Dr de Muelenaere was concerned only with co-ownership of the immovable properties on which the shopping centres are situated. He made this clear a little later in the affidavit, where he said that the Adamax co-owners ‘owned a number of properties essentially comprising of, or adjacent to, three shopping centres’ and that on 9 November 2011 they ‘sold an undivided share in certain properties (including the properties comprising the three aforementioned shopping centres) to the MEPF’.
He described these as the co-owned properties and in dealing with the relief sought explained that the Adamax co owners were not in a position to buy the MEPF’s ‘share in the properties’. The suggested relief was that ‘the property portfolio’ should be placed on the market by a liquidator to be sold.
 It is true that in delivering the Letting Enterprise to the MEPF a 55 percent share in each of the immovable properties owned by the Adamax co-owners was transferred to the MEPF. But there was also a cession and assignment of their rights and obligations under the leases to the extent of a 55 percent interest. Together this resulted in the transfer to the MEPF of a 55 percent interest in the Letting Enterprise purchased by it.
Had the undivided share in the immovable properties not been transferred to the MEPF, the sale of the 55 percent stake in the Letting Enterprise would not have been affected, although its terms would have needed to be adjusted, as the Adamax co-owners would still have been obliged to allow the shopping centres to be used by the Letting Enterprise for the purposes of conducting its business. No doubt part of the reason for effecting a transfer of the 55% interest in the immovable properties was to provide the MEPF with greater security and because in terms of clause 9.2 of the COA each party was entitled to procure and register a mortgage bond over their pro rata share in order to secure financing from a financial institution without the consent of the other party.
 Turning to the COA it was concluded by Adamax, both on its own behalf and on behalf of an entity, described in the COA as ‘Propco’ and consisting of Adamax and the Adamax co-owners, thereby demonstrating the complete control that Adamax exercised over the Adamax co-owners. The other party was the MEPF.
It recorded that the MEPF had a ‘55% undivided share in the Letting Enterprise’ which it defined as meaning ‘the retail and commercial concern namely the letting enterprise’ comprised by the immovable properties and the right, title and interest in and to the leases and the revenues relating to the properties. In other words, it was a 55 percent share in the overall shopping centre business.
Propco, that is Adamax and its subsidiaries, had a 45 percent share. The subject of the co-ownership was therefore the Letting Enterprise, not the immovable properties separated from it. The Letting Enterprise and the immovable properties could not be separated.
 It was the intention of Adamax and the MEPF that they would conduct the business of the Letting Enterprise together for an indefinite period. This emerges from clauses 4 and 5 of the COA, which read as follows:
‘4 Constitution of Co-Ownership
4.1 The Parties agree to carry on the Letting Enterprise under such name and style as they may choose, currently, Parkview, Glen Village North and Glen Villlage South.
4.2 The Parties agree that should this Co-Ownership be dissolved for any reason whatsoever, that the use of any name applicable to the business or any part or derivative thereof shall remain the absolute right of that Party which may continue to operate such business and that no proprietary right in and to any name so used will vest individually in the Co-Owners.
5.1 The Co-Ownership will commence with effect from the Effective Date and shall terminate:
(2) on the disposal of the Letting Enterprise by the Co-Owners; or
(3) if either of the Co-Owners disposes of its share in the Letting Enterprise, subject however to the provisions of clause 22 hereof.’
 Clause 22 is headed ‘Prohibition on Sale’ and reads as follows:
‘It is expressly agreed amongst the Parties that no Party shall be entitled to sell its undivided share in the Property without having complied in full with the terms and conditions of this Agreement, more particularly the pre-emptive rights referred to in Clause 23 below.’
It is unnecessary to set out the terms of clause 23 in full. It is a relatively standard clause affording the party which does not wish to sell their share in the Letting Enterprise a right of pre-emption if the other party or parties wish to do so.
Two provisions are important.
- The one is that if the remaining co-owner acquires the interest of the departing co-owner they become obliged to seek the release of the latter from any suretyship or guarantee provided by the seller to any creditor of the co-ownership and, if unable to do so, to indemnify the seller against any claim under such suretyship or guarantee.
- The other is that it was to be a condition of a sale to a third party that they agree to be bound by the terms and conditions of the COA. In other words, the business relationship involved in conducting the Letting Enterprise of the shopping centre business was to continue.
Whilst the co-ownership between the MEPF and the Adamax co-owners would have terminated in terms of clause 5.1(3), a fresh relationship of co-ownership would come into existence between the remaining co-owner and the new purchaser.
 Notwithstanding the creation of this complex web of ongoing commercial relationships, Adamax contended that it could be brought to an end in a manner other than that provided by the COA. Dr de Muelenaere set out its position in the founding affidavit:
‘I am advised that co-owners have a right under the common law, more specifically under the actio communi dividundo to, at any time, demand that co-owned assets be divided. I am further advised that no co-owner is obliged to remain a co-owner against his will. In the absence of an agreement to the contrary, any co-owner may consequently demand partitioning of the common property at any time.’
 The breadth of this contention is apparent from the repeated statement that the actio may be invoked ‘at any time’. In other words, Adamax claimed that it was open to it to invoke the actio immediately after the conclusion and implementation of the sale agreement and the COA or at any time thereafter. One could add ‘for any reason’ because the claim is unqualified by any restraint on the exercise of the right to demand partition of the co-owned property.
Notionally, it could be because the one party had fallen out with the other over an issue unrelated to the running of the business, or because they thought that selling the entire enterprise would generate a greater return for them than would a sale of their undivided share.
A disagreement over the prescribed matters in Schedule 1 to the COA, on which clause 11.9 required unanimity, could be met with the nuclear option of a demand for division of the co ownership, even though clause 21 contained detailed provisions to resolve a deadlock over any operational issue.
 Adamax’s contentions meant that it mattered not that this would destroy the very basis upon which the MEPF entered into the arrangement, namely in order to secure a stable stream of income to meet its obligations to pensioners. The blunt proposition was that resort could be had to the actio communi dividundo because it afforded Adamax a right to demand that the co-owned assets be divided. Resistance to such a demand entitled it to bring the present application.
 The MEPF resisted this contention, pointing to the terms of the sale agreement and the COA as the basis for opposing the application of the actio communi dividundo. They contended that the application was an endeavour to subvert the provisions of the COA, which they said regulated all incidents of its operation and constituted the complete record of the parties’ agreement.
Their primary case was that the relationship between the parties was governed by the contracts and nothing else. The actio was accordingly unavailable to Adamax. It is unfortunate that they did not refer in their argument to the distinction between bound and free co-ownership, their attention to it having been prompted by questions posed by this court prior to the hearing.
Greater clarity might then have emerged at an earlier stage of the proceedings, but these are merely jurisprudential labels and the essential case advanced by the MEPF was always that the agreements created a situation of bound co-ownership.
 In the high court the judge characterised this argument as being based on a tacit term that would exclude the actio communi dividundo. He reached that conclusion notwithstanding the absence of any suggestion in the answering affidavit that reliance was being placed upon any such tacit term.
He thought, erroneously, that the availability of the actio was one of the naturalia (inevitable legal consequences) of any agreement giving rise to co-ownership. In the result his entire approach to the case was flawed.
In refusing leave to appeal he compounded these errors by burdening the MEPF with an onus to prove as a defence the exclusion of the actio, when the true question was whether it was available at all given the terms of the agreements. He then refused leave to appeal.
Had he paid regard to the consequences of holding that the actio was available in the circumstances of this case he should have granted leave to appeal to this court, notwithstanding his view that his judgment was correct. All too frequently this court bemoans the grant of leave to appeal in matters of no great complexity raising no significant legal issue. This was not such a case and the leave to appeal that should have been granted was granted by this court.
The elements of free and bound co-ownership
 There are countless situations that may create co-ownership of property. The old authorities and the older cases are almost all concerned with immovable property or inheritances. A common situation was where under a will, or in accordance with the law governing intestacy, property was inherited by two or more people in undivided shares. A deed of donation of property in undivided shares to more than two people was another. The purchase of immovable property by two or more individuals in undivided shares was a third.
Marriage in community of property and partnership are common examples. The establishment of a trust, which is not itself a separate juristic entity, renders the trustees the co-owners in undivided shares of the trust property. The members of a common law universitas that owns property, such as a club, a political party, a church and many non-governmental organisations, own the property of the universitas in undivided shares. The owners of units in a sectional title development are co-owners of the common property in undivided shares.
 Co-ownership may be either free or bound co-ownership. The distinction is explained by Professor C G van der Merwe in LAWSA:
‘Common ownership may either constitute the only legal relationship between the co owners or it may result from some other legal relationship between the parties. In the former case (designated free co-ownership) the relationship between the co owners is more individualistic in that the community of property can be dissolved by any co-owner and in that each co-owner is allowed to use and enjoy the common thing in accordance with his or her undivided share in it. … In the latter instance (designated bound co-ownership), which is, for example, the result of a marriage in community of property or a partnership, the relationship is more permanent in that no division of property can be requested during the course of the community of property and in that the undivided shares allotted to each owner do not have real significance. The legal relationship between the parties further largely determines the rights and duties of the co-owners.’ (My emphasis)
 Professor van der Merwe writes more extensively, but to similar effect, in his book Sakereg. In regard to bound co-ownership he says the following:
‘Gebonde mede-eiendom ontstaan wanneer ’n besondere regsverhouding tussen gemeenskaplike eienaars geskep word, soos by mede-erfgenaamskap, huweliksgoederegemeenskap, ń vennootskap en verenigings sonder regs persoonlikheid. Die besondere regsverhouding wat tussen die deelgenote bestaan, bepaal die manier waarop eiendomsbevoegdhede ten opsigte van die gemeenskaplike saak uitgeoefen word. Die feit van mede-eiendom is slegs één van die gevolge van die regs betrekking wat tussen die partye bestaan. By vrye mede-eiendom daarinteen is die feit dat die mede-eienaars tesame eienaars van dieselfde saak is, die enigste regsverhouding wat tussen die partye bestaan.’ (My emphasis.)
‘Bound co-ownership arises whenever a particular legal relationship is established between co owners, as in co-inheritance, marriage in community of property, a partnership and associations without legal personality. The particular legal relationship that exists between the co-owners determines the manner in which their ownership rights in relation to the joint property are exercised. The fact of co-ownership is only one of the consequences of the legal relationship that exists between the parties. By contrast, with free co-ownership the fact that the co-owners are owners of the same thing is the only legal relationship between the parties.’ (Tranlated by MJD Wallis JA)
 The fundamental point to be distilled from these passages is that in bound co-ownership the existence of the co-ownership arises from a legal relationship between the parties other than the co-ownership itself. In other words, there is a legal relationship between them going above and beyond the fact that they happen to be the co-owners of property. The co ownership arises from and is constituted as a consequence of that relationship. It is not the source of the relationship between the parties.
 This point is emphasised by Kleyn and Wortley, writing jointly in comparative terms about bound co-ownership in South Africa and its equivalent, joint ownership, in Scotland:
‘As far as bound co-ownership is concerned, South African authorities mention the following. There exists some other (special) relationship between the parties which is of a more permanent nature; this relationship determines the rights of the parties; co ownership is just one of the consequences of that relationship; no division can be called for unilaterally during the existence of the relationship; and a co-owner cannot deal independently with the undivided share during the relationship. Scots law provides a similar picture. As Lord Cooper pointed out in Magistrates of Banff v Ruthin Castle, an independent relationship is ‘the indispensable basis of every joint right’ and the ‘attributes [of joint property] are … the consequences flowing from the relationship … there is no entitlement to division and sale.’ …
The importance, in bound co-ownership, of an extrinsic relationship is emphasised in both systems. All the consequences of bound co-ownership flow from that relationship. Although the two jurisdictions identify a limited number of different kinds of relationships that constitute bound co-ownership, without such a special relationship the co-ownership must always be free.’
Counsel for Adamax relied on this last sentence, and two later ones to similar effect, as restricting the potential for a relationship to be one of bound co-ownership. For my part I do not understand the reference in the last sentence to ‘a limited number of different kinds of relationships’ or a ‘special relationship’ to mean that there is a closed list of situations of a peculiar nature in which bound co ownership can arise.
That would be inconsistent with the recognition earlier in the same passage that a relationship between the parties that determines their rights and obligations, and of which the co-ownership is only one consequence, gives rise to bound co-ownership. It is simply a statement that in the two jurisdictions in question only a few instances of bound co-ownership have been identified. Insofar as the authors intend it to go further I do not agree with them.
 South Africa recognises various sources of extrinsic legal relationships giving rise to bound co-ownership. It may arise as a matter of law from the fact that the parties have entered into a particular relationship. An example of this is a marriage in community of property, where the common law, as varied by the Matrimonial Property Act 88 of 1984, imposes co-ownership upon the parties to the marriage.
Another is the co-ownership of the common property in a sectional title development, by virtue of the provisions of s 16(1) of the Sectional Titles Act 95 of 1986.
It may arise from an act such as the execution of a trust deed by the founder of a trust and the acceptance by the trustees of office under that deed. Another possibility is an agreement between the co owners, as in a partnership or the constitution of a universitas. In the case of trust deeds, partnership agreements and constitutions the parties are usually free to vary their terms and the terms of the relationship between the co-owners.
 The leading Scottish judgment on joint ownership is that of Lord Justice Clerk Cooper in Magistrates of Banff v Ruthin Castle.
Lord Gill, a recent successor of his as Lord President, said that this judgment
‘eruditely expounded the distinction in Scots law between joint ownership, being the class of right typified by the ownership of co trustees, and ownership in common, being the right typified by the ownership of two or more persons in whom the right of a single subject has come to be vested’.
The key passage reads as follows:
‘Joint property, on the other hand, has received little doctrinal exposition as a mode of holding property, probably because its attributes are not so much the incidents of the joint right as the consequences flowing from the relationship existing between the persons who alone can have a joint right. So far as has been traced, there is no instance of a joint right in the strict sense having been held to exist except in persons who were inter-related by virtue of some trust, contractual or quasi-contractual bond — partnership or membership of an unincorporated association being common examples — and it seems to me that such an independent relationship is the indispensable basis of every joint right. The distinctive feature of the right of such joint proprietors is the jus accrescendi, which excludes the possibility of separate shares in the several joint owners, and still more emphatically excludes the possibility of severance of the tie, except, of course, by dissolution of the relationship on which the joint ownership rests. Finally, the considerations of public policy that nemo in communis invitus detineri potest, have no application to the entirely different situation created by joint ownership.’ (My emphasis.)
 This statement of Scots law is undoubtedly authoritative in that jurisdiction and wholly consistent with the views of Professor van der Merwe.
It stresses the fact that the incidents of the relationship between the parties are derived not from the co-ownership itself, but from the extrinsic relationship between the co-owners, separate and distinct from their co-ownership. It accepts that the source of a joint ownership (bound co-ownership in South African terminology) can be a trust or a contractual or quasi-contractual bond.
Lastly, it makes the point that in cases of joint co-ownership the public policy rule that a person cannot be compelled to remain a co-owner against their will has no application.
 Neither Professor van der Merwe, nor Lord Justice Clerk Cooper, suggested that bound co-ownership was restricted to any defined class of agreement. The former referred to legal relationships extrinsic to the co ownership and mentioned various examples of such relationships.
The latter said that it arose in relation to persons ‘who were inter-related by virtue of some trust, contractual or quasi-contractual bond’, without in any way limiting the type of contractual or quasi-contractual relationship that might give rise to bound co-ownership. The only other Scots case I have found in the course of my research is Munro v Munro.
There the court held that joint co-ownership was created where a property was passed to family members on terms that it would descend to the last surviving of their number and the family members contractually agreed to this. The court held that this created a bound co-ownership that none of the family members could defeat by way of a claim for division of the property.
On its face that appears to be consistent with the view of Lord Justice Clerk Cooper. I cite the case with a measure of hesitation because it has been criticised in its country of origin, both because it creates problems for a good faith purchaser of a share in the property and because it creates some difficulties in regard to the complex Scots law provisions governing special destination clauses, such as this one.
However, neither criticism has any purchase in South African law, at least insofar as the issue of creating bound co-ownership by contractual agreement is concerned.
 Adamax relied upon certain passages from Kleyn and Wortley’s article to contend that bound co-ownership only arises from an extrinsic exceptional relationship in which the rights and obligations of the parties are defined by common law or statute. The contention was that there is only a limited number of relationships that constitute bound co-ownership and that these are special relationships of a more permanent nature that cannot be altered by agreement between the parties.
 Insofar as Scots law is concerned, neither Magistrates of Banff v Ruthin Castle nor Munro v Munro lends any support to these conclusions. Nor is it apparent that the relevant passages from Kleyn and Wortley’s article are dealing with the issue of whether the relationship between the parties is one of bound or free co-ownership. They mainly occur in a section of the article that commences as follows:
‘However clear the distinction between free and bound co-ownership may be in theory, it can be undermined by the freedom of contract that the parties have in cases of free co-ownership. The question then arises as to whether co-owners can arrange matters in such a way that free co-ownership is converted into bound co-ownership. In other words is it for the parties to decide whether the ownership is bound or free, or does property law dictate the issue?’ (Emphasis added.)
This is a puzzling passage. It is dealing with free co-ownership, not with whether in any particular case the co-ownership is bound or free. If the premise is that save in specific limited cases, such as partnership, all co ownership is free, it is unhelpful, as questions that assume the answer always are. Furthermore, the two propositions are not the same. The former deals with the conversion of free co-ownership to bound co ownership and the latter with the parties’ freedom to determine contractually the nature of their co-ownership.
Approaching the problem from an assumption of free co-ownership and asking whether this can be altered by contract, is the wrong question, or at least the wrong question in the present context. We are not dealing with a relationship commencing as free co-ownership and asking whether the parties can alter it by contract. Our concern is whether from the outset the relationship created by the sale agreement and the COA was one of bound or free co-ownership.
 Other passages from this section of the article also deal expressly with free co-ownership. For example, the authors say:
‘It is clear that the parties can, to some extent, change the ordinary rules of free co-ownership by means of agreement. … However, the matter is different with respect to the right to call for division, which personifies the individualistic nature of free co-ownership. Both jurisdictions seem to accept the principle that a free co-owner can agree not to call for partition within a limited time, but that an agreement never to ask for partition is void, a result which is in accordance with Roman Law’. (Emphasis added)
In the result these passages are unhelpful in resolving the issue with which we are confronted. That is not a criticism of the authors who said at the outset that their focus was free co-ownership.
 The one South African judgment relied on by Kleyn and Wortley is that of Ex parte Geldenhuys. That involved a will bequeathing a farm to the children of the testatrix in undivided shares on the basis that division would take place when the eldest child reached majority and then by a drawing of lots. The court held that this was permissible and the condition could be registered against the title deeds showing the children’s undivided shares in the property.
De Villiers JP (as he then was) said:
‘By the common law, each owner of an undivided share has the right to claim a partition at any time, and can claim that such partition shall be effected either by agreement or by the Court. The will, therefore, modifies the common law right, or dominium, which an owner of an undivided share possesses. That this can validly be done by a will (and presumably also by agreement inter vivos) seems to me to be clear on principle, for the rights of an owner of an undivided share are not sacrosanct or unalterable any more than the rights of an owner of a defined share are. Portions of the dominium of an owner of an undivided share can be parted with as undoubtedly as portions of the dominium of an owner of a defined share can be parted with. There is, in fact, the express authority of Grotius, if authority were needed, that an owner of an undivided share can by will be deprived for a specified time of his right to claim a partition (Grotius 3.28.6, Maasdorp’s Institutes of Cape Law, bk. 2, ch. 14).
The rights of a joint owner in regard to partition can therefore be validly limited by last will at any rate, and the limitations now under discussion (i.e., as to the time of partition and as to the drawing of lots) are therefore valid.’ (Emphasis added.)
 In fact, the provision of the will in Geldenhuys effectively removed the right to demand division entirely, by specifying that it would take place in a specific way at a defined future date. At no stage before the arrival of that date would any of the children be entitled to claim division of the property in accordance with their co-ownership.
And, as De Villiers JP said, if that could be done by will it could also be done by agreement inter vivos. Therefore the judgment does not support the view that bound co-ownership is confined to ‘a limited number of relationships that constitute co-ownership’ and that without such a ‘special relationship’ the co-ownership will always be free.
That view is also not supported by any South African or Roman Dutch authority to which they refer and is incompatible with Professor van der Merwe’s descriptions of bound co ownership.
 I have already mentioned that the old authorities address situations far removed from the present one. They also do not use the terminology of free and bound co-ownership, which we owe to an article by Professor van Warmelo.
As such, their discussion of the division of common property is apt to situations of free co-ownership, but does not address bound co-ownership and does not consider a situation comparable to that in this case.
However, there are indications that there are circumstances in which the right to division may be restricted.
Grotius 3.28.6 says that co-owners are entitled to division unless it has by last will been forbidden to dissolve within a certain time, although he accepts that a perpetual prohibition would be invalid.
Voet 10.2.32 goes further and says that there is no reason why a testator may not forbid the division of an inheritance for a long period or even the lifetime of a person. He draws an analogy with partnership, pointing out that a partnership can be constituted for the lifetime of the partners, which he describes as ‘for ever’, and says:
‘… it can be arranged that within a definite period no division shall take place nor any departure from the partnership.’
 It is clear that these writers accepted that there were situations of community of property, as they describe situations of co-ownership, that were not susceptible to the action for division until the broader relationship between the parties giving rise to the community had ended.
Van Leeuwen says this in relation to partnerships and marriages in community of property. Recognition of the fact that partnership gives rise to bound co-ownership of partnership assets, both tangible and intangible, involves an acceptance of the proposition that parties can by agreement so order their affairs as to create a situation of bound co ownership.
Van der Linden gives the example of joint shipowners being obliged to remain such, at least until the completion of the voyage, in terms of the deed of ship ownership. It is important to note that this would not necessarily be a partnership, but the parties could contract for co-ownership and restrict any right to claim division of the jointly-owned property, namely, the ship.
 It is necessary to consider in greater detail the article in which Professor van Warmelo first applied the expressions free and bound co ownership in our jurisprudence. The learned author drew a distinction between two forms of co-ownership.
The first was where the relationship between the co-owners and their rights to the property were determined by the law and were unalterable by the exercise of free will on the part of the parties. He said this could be called ‘bound co-ownership’. The instances he gave of this form of co-ownership were joint heirs in relation to a joint estate; marriage in community of property; and ‘boedelhouerskap’, where the surviving spouse from a marriage in community of property is appointed as executor of the predeceased spouse, administrator of their estate and guardian of the minor children born of the marriage, and continues the community pending the children’s majority.
 The second possibility was the relationship he called free co ownership. This was where a thing was owned by more than one party, but the law did not impose unalterable consequences upon the relationship of the parties to the property or among each other. Here there were two possibilities.
- The first was where the only relationship between the parties flowed from their co-ownership of the property.
- The second was where there was a further arrangement emanating from the parties themselves, that either confirmed the conventional legal relationship as determined by the law, or changed it in whole or in part.
The typical example of this was a partnership, under which the partners could regulate their relationship to the property and one another. Having said that he went on to explain that the further discussion in the article would be of free co-ownership other than partnership, because partners were free to determine their own rules and in so doing alter the rules that would otherwise apply to the partnership relationship.
 It will be immediately apparent that this description of bound and free co-ownership is not the same as the distinction drawn by Professor van der Merwe, who places partnership in the category of bound co ownership.
Professor van Warmelo derived his terminology from a Dutch work, but acknowledged that its authors used it in a different way, one that accords with Professor van der Merwe. The Dutch usage was that free co-ownership applied to situations where the only relationship between the parties was the co-ownership, while bound co-ownership was where there was another legal tie between the co-owners governed by its own rules.
 The argument by counsel for Adamax was that bound co ownership is limited in accordance with the approach adopted by Professor van Warmelo. That approach was inconsistent with
- the approach of Professor van der Merwe;
- this court’s approval of Professor van der Merwe’s approach;
- the views of Kleyn and Wortley, and
- the Scottish cases.
Professor van Warmelo’s view is cited in Dr Kuyler’s thesis but without any detailed analysis or consideration, which is hardly surprising as the subject of the thesis was free co-ownership, not bound co-ownership. The force of Dr Kuyler’s endorsement of Professor van Warmelo’s approach is considerably diluted by the fact that on the following page he says that partnership is a case of bound co-ownership, whereas Professor van Warmelo said that it is a case of free co-ownership.
 The exposition of the nature of the actio communi dividundo in Robson v Theron does not assist in resolving the issue of free and bound co-ownership.
The case arose from the dissolution of a partnership where the one partner had appropriated and retained the goodwill of the partnership practice. It concerned a claim by the other partner to be compensated for his share of the goodwill. It was not concerned with the distinction between free and bound co-ownership as the dissolution of the partnership had brought the co-ownership to an end.
The parties had agreed on what was to be paid for the retiring partner’s share and the real dispute between the parties was the factual one of whether the other partner had retained the goodwill for his own benefit and was obliged to compensate the retiring partner therefor.
 The judgment commences with a discussion of the remedies available to the departing partner on dissolution of the property, namely the actio pro socio and the actio communi dividundo. The discussion of the former contains the following quotation from Pothier:
‘In order to dissolve the community which subsists, after the dissolution of the partnership between the former partners, and to discharge the respective debts for which they may be liable to each other, each of the former partners, or his heir, has a right to demand of his partners or their heirs to proceed to an account and distribution of the partnership effects. To effect this they can each maintain the actio pro socio or the actio communi dividundo, at their option.’ (Emphasis added.)
 Joubert JA endorsed the view of Pothier, concluding his discussion of the actio pro socio in the following terms:
‘Pothier, on the other hand, makes it abundantly clear that, in the liquidation of a partnership, distribution or division of the partnership assets may be effected among the partners in such a manner as not to involve the physical division or partition of tangible assets of the partnership. According to him the actio pro socio or the actio communi dividundo may be used for the distribution or division of partnership assets. His approach is both logical and practical. In following it Van der Linden introduced it into Roman-Dutch law. It is also basically in conformity with our present day practice of liquidating partnerships.’
 In discussing the actio communi dividundo Joubert JA did not revisit these principles. The discussion is brief and concentrates on the modes of division and the powers of the court in dividing the co-owned property. The only substantive proposition was the statement that the actio has been extended to intangible things held in common ownership of which the goodwill in dispute in that action was an example.
 A brief summary of the provisions of the two actiones followed. In dealing with the actio pro socio it started by making the point that the actio was available during the existence of the partnership to claim specific performance and after dissolution to wind up the partnership.
Turning to the actio communi dividundo the summary commenced by saying that:
‘No co-owner is normally obliged to remain a co-owner against his will.’ (My emphasis)
Significantly, there was no suggestion that the actio was available during the subsistence of the partnership. Given that this was a brief summary of the actio, it is no surprise that Joubert JA did not go on to set out in which circumstances a co-owner would be obliged to remain a co-owner against their will. The obvious case would be that of bound co-ownership (of which partnership is a quotidian example) until the relationship giving rise to the tie had itself been terminated. It would be surprising, given his academic background, were Joubert JA not aware of this. He continued by saying that the actio was available to any co-owner, whether or not the co-owners were partners.
 The third point may be misconstrued. It contained the following passage:
‘Hence this action may be brought by a co-owner for the division of joint property where the co-owners cannot agree to the method of division. Since a partnership asset is joint property which is held by the partners in co-ownership, it follows that a partner may as a co-owner bring this action for the division of a partnership asset where the co-partners cannot agree to the method of its division. This would obviously cover the position where, after dissolution of a partnership, a continuing partner as a co-owner retains possession of an undivided partnership asset.’
While the second sentence is not in terms limited to a division after dissolution of the partnership, it should not be understood as postulating that the actio is available during the subsistence of the partnership, although it may possibly be invoked in conjunction with a claim that the partnership has been dissolved or for its dissolution. It is unnecessary to express a firm view on this.
Given the entire discussion in the judgment it does not support the proposition that the actio is available during the subsistence of the partnership, as opposed to at the stage of its dissolution. The relevant principles are discussed solely in the context of the dissolution of the relationship leading to a claim for division of the co-ownership.
 Accordingly, the judgment casts no light on how to determine whether co-ownership is free or bound. The case is not authority for the general proposition that no co-owner may be compelled to remain a co owner against their will. That ignores the context and the careful qualification that this is ‘normally’ the position. Bound co-ownership is precisely the case where a co-owner is obliged to remain such against their will, unless and until the tie that creates the bound co-ownership has been severed.
 In summary therefore, I conclude, in accordance with the authorities discussed above, that the distinction between free and bound co-ownership is that in the former the co-ownership is the sole legal relationship between the co-owners, while in the latter there is a separate and distinct legal relationship between them of which the co-ownership is but one consequence. Co-ownership is not the primary or sole purpose of their relationship, which is governed by rules imposed by law, including statute, or determined by the parties’ themselves by way of binding agreements. The relationship is extrinsic to the co-ownership, but is not required to be exceptional.
In other words it requires no special feature for the co-ownership consequential upon the relationship to qualify as bound co-ownership. Whether it is depends upon the terms upon which the relationship is constituted. The mere fact that co-owners decide to exploit their co-ownership commercially will not of itself constitute the co-ownership as bound co-ownership. That will depend upon the nature and terms of the commercial agreement between the parties and matters such as the provision made for its termination.
 There is no closed list of instances of bound co-ownership. If the relationship gives rise to bound co-ownership the co-ownership will endure for so long as the primary extrinsic relationship endures. Once it is terminated then, as in Menzies and Robson v Theron, it will become free co-ownership and be capable of being terminated under the actio. I consider the facts of this case in accordance with those principles.
Is this a case of bound co-ownership?
 It is necessary to start by identifying both the subject matter of the co-ownership and how it arose. That is complicated by the fact that there are different things that are the subject of co-ownership. As explained earlier in analysing the sale agreement and the COA the primary subject of the co-ownership is the Letting Enterprise. The COA regulates the operation of the business relationship between the MEPF and Adamax and the Adamax co-owners. Pursuant to the obligation of the Adamax co owners to deliver the 55 percent share in the Letting Enterprise to the MEPF in terms of the sale agreement, a 55 percent interest in each of the immovable properties was transferred to the MEPF in undivided shares.
 The primary relationship between the parties is therefore their relationship in respect of the Letting Enterprise governed by the COA. Their co-ownership of the immovable properties is a consequence of that relationship, albeit not a necessary consequence, in that transfer of an undivided share in the properties was not essential in order to give effect to the sale of the Letting Enterprise or its operation in terms of the COA.
Conceivably the Letting Enterprise could have come into existence without transferring an undivided share in those properties to the MEPF. Some other mechanism for securing the right to conduct the Letting Enterprise in the shopping centres could have been devised. It is not, as I understand it, suggested that in that case the Adamax co-owners would have been entitled to dispose of the properties to a third party and thereby deprive the Letting Enterprise of access to and control over the shopping centres and the MEPF of its interest in the Letting Enterprise.
 The premise upon which the Adamax co-owners brought the application was therefore incorrect, because it undermined the entire relationship between the parties in terms of the COA. It took as its starting point the subsidiary and consequential co-ownership of the immovable properties and treated it not simply as if it were the primary legal relationship, but as if it were the only legal relationship.
That was plainly incorrect. Any attempt to invoke the actio had necessarily to start with the co-ownership of the Letting Enterprise and the question whether that constituted bound co-ownership. It is to that question that I now turn.
 At the outset I reject the proposition in the respondents’ heads of argument that the starting point is that in co-ownership the availability of the actio is implied by law, so that it must be excluded unambiguously. That is incorrect. It puts the cart of a conclusion – ‘This is free co ownership’ – before the horse of the question – ‘Is this free or bound co-ownership?’.
The common law is that the actio is always available in the case of free co-ownership and never available in bound co-ownership. In any particular case the question of the proper characterisation of the co-ownership arises at the outset. Only once it has been answered can one decide what the common law attributes of the co-ownership are. One cannot therefore start with a predisposition in favour of free co ownership.
That also renders irrelevant the reliance placed upon the statement by Marais JA in First National Bank of SA Ltd v Rosenblum, that:
‘In matters of contract the parties are taken to have intended their legal rights and obligations to be governed by the common law unless they have plainly and unambiguously indicated the contrary.’
Where, as here, we are dealing with one of two different forms of co ownership, one of which affords the right to invoke the actio and the other does not, it is necessary first to identify which form of co ownership is applicable in order to determine what the common law is. As it happens, Marais JA was dealing with an exception clause and was articulating the approach our law adopts to such clauses, which is that they are strictly construed and must clearly and unambiguously exclude liability for the loss suffered by the claimant. The quoted dictum must be read and understood in that light.
 The relevant features of the co-ownership of the Letting Enterprise have already been identified in the analysis of the sale agreement and the COA. It was a business relationship of indefinite duration with careful and detailed provisions for the conduct of the relationship.
In clause 5 it contained an express provision in regard to its duration, providing that
- it would terminate as a result of an election by the parties if transfer had not been achieved by 30 June 2012;
- if the parties agreed to dispose of the Letting Enterprise; or,
- if either co-owner disposed of its share in the Letting Enterprise after affording the other a right of pre-emption in terms of clauses 22 and 23 of the COA.
Although not mentioned in clause 5 it could also be terminated as a result of cancellation for breach, or as a result of a forced sale upon insolvency or winding up.
 During the course of argument there was some debate with counsel for the MEPF over whether, in the light of the decision in Putco Ltd v TV & Radio Guarantee Ltd and Other Related Cases, these were the only grounds upon which the COA could be terminated.
I would be hesitant without full argument to reach any conclusion on this point, as neither side’s counsel had addressed their minds to the issue and the implications of an ability to terminate the COA by notice were not explored in the affidavits. However, it is unnecessary to do so because the PUTCO case is clearly distinguishable. It concerned a contract to secure advertising to be placed on the appellant’s buses.
The contract was embodied in a letter containing the following paragraphs:
‘This letter, although binding upon both of us, is intended to be a temporary interim arrangement, and we confirm that, in due course, a detailed agreement will be concluded between us as a result of the negotiations we have been conducting.
… [W]e reserve to ourselves the right to withdraw from this arrangement should damage result to our image flowing from your activities, or should our earnings at any time be insufficient from this scheme, or should the arrangement become administratively impracticable in regard to your ability to signwrite and maintain the advertising copy.’
 In that context it was said:
‘[W]hen parties bind themselves to an agreement which requires them to work closely together and to have mutual trust and confidence in each other, of which the agreement under consideration is an example, it is reasonable to infer that they did not intend to bind themselves indefinitely, but rather contemplated termination by either party on reasonable notice.
Where an agreement is silent as to its duration, it is terminable on reasonable notice in the absence of a conclusion that it was intended to continue indefinitely. The inclusion in the agreement of three specific grounds for termination does not exclude termination by reasonable notice. The logical consequence of an argument that only three specific grounds for cancellation of the agreement exist would be that, provided those grounds for cancellation do not arise, the agreement would continue indefinitely. This would not be a proper construction to place on the agreement as it ignores the intention of the parties when entering into the agreement …’
 That situation was by no means comparable to the present case. The parties clearly expressed the intention that their agreement was an interim and a temporary arrangement. A detailed agreement was to be negotiated which would no doubt have dealt with matters such as the ability of either party to terminate it on notice. The nature of the relationship did not involve the joint and co-ordinated operation of a business, but the provision of services to PUTCO.
All of this is absent from the present case. We are dealing with a suite of agreements of some considerable commercial complexity, in a transaction involving more than half a billion Rand. It was clearly intended to continue for more than four years, as that was the period in respect of which the income guarantee in terms of clause 5 of the sale agreement would operate.
Construing the COA as permitting a termination on notice within that period would run counter to an express provision of the sale agreement. Similarly, permitting the Adamax co-owners to bring the actio during that period would also be utterly inconsistent with the sale agreement that created the co-ownership of the Letting Enterprise and the immovable properties. Yet, as pointed out at the outset of this judgment, that is the approach of the Adamax parties.
 It may be, as debated in argument, that the COA is terminable on reasonable notice duly given. I express no firm view on that question. The fact that it may terminate in other circumstances does not necessarily exclude that possibility, but it does not affect the question whether the relationship gives rise to bound co-ownership.
A partnership undoubtedly does so, but most partnership agreements provide for their termination on reasonable notice and in the ordinary course that is one of the naturalia of a partnership. Nothing was drawn to our attention to suggest that this would affect the partners’ co-ownership of either the partnership business or any movable or immovable property co-owned by them.
 The mention of partnership brings me to a consideration of the nature of the legal relationship constituted under the COA. The requirements of a partnership are
- that each party bring something into the business, be it assets, finance, skill or labour;
- that the business is carried on for the joint benefit of the partners; and
- that the business is conducted for the purpose of making a profit.
Any joint venture that exhibits the characteristics of a partnership is itself a partnership. The COA exhibits all of these features. Of course, if it is a partnership then cadit quaestio, as it is accepted that co-ownership of a partnership business or property is bound co-ownership.
 As with termination on notice, the possibility that the COA gave rise to a partnership was not explored in argument. No doubt this was because clause 7 of the COA, said that:
‘Nothing in the agreement shall be deemed to constitute a partnership or a joint venture of whatsoever nature and/or description and none of the Parties shall be entitled to bind the other in any manner.’
This clause cannot detract from the other provisions of the agreement or alter its proper legal characterisation. Its principal purpose is to make it clear that there is no agency relationship under which the parties would be able to bind one another contractually, as would conventionally be the case with a partnership. Whether it has effect in accordance with its terms to exclude the COA from being characterised as a partnership or joint venture is another matter.
 I have reservations whether the mere fact that the parties say that it is not a partnership can affect the legal position where the relationship has all the hallmarks of a partnership.
In Rhodesia Railways Stratford AJA said:
‘I asked Mr. Tindall, who appeared on behalf of the appellant, if he had found any authority or any decided case in which these four essentials have been present where the relationship has been held not to constitute a partnership, and he was unable to cite any. Where all these four essentials are present, in the absence of something showing that the contract between the parties is not an agreement of partnership, the Court must come to the conclusion that it is a partnership. It makes no difference what the parties have chosen to call it; whether they call it a joint venture, or letting and hiring. The Court has to decide what is the real agreement between them.’
The nature of ‘something showing that the contract between the parties was not an agreement of partnership’ was explained by Ogilvie-Thompson JA in Purdon v Muller in the following terms:
‘The meaning of this qualification is, I think, that, although the presence in an agreement of the four essentials will prima facie establish the existence of a partnership, such presence is not necessarily conclusive but must yield to contrary intention as revealed in the agreement itself read in the light of the other admissible evidence. (Cf. Estate Davison v Auret, 22 S.C. 10 at p. 16). In the ultimate analysis the question is always one of construction.’
I doubt that a boilerplate clause in an agreement suffices to avoid the relationship being characterised as a partnership.
 Be that as it may, it is unnecessary to reach any final conclusion on this question. The parties have described the business venture on which they embarked as a Letting Enterprise, an expression that has no specific legal meaning.
Its basic character is that the business lets premises in shopping centres and seeks to make a profit from them. Its primary assets are the shopping centres and the tenant leases that generate the revenue of the business.
Such a business has employees, letting agents, cleaning and maintenance staff or contractors, and a need to engage in marketing both to potential tenants and to potential customers, because ‘footfall’ is vital.
We were not told how the accounting for this enterprise worked, but there seems to be no reason to believe that the revenue was not used to pay the bills and the surplus, subject to provisions or building up a contingency fund, was to be distributed between the parties in proportion to their interest in the enterprise.
 The relationship constituted by the COA was plainly an ongoing joint business venture, regulated in terms of the three agreements concluded by the parties. It had many characteristics of a partnership and counsel accepted that the central document, the COA, was based on a conventional shareholders’ agreement in a private company.
A convenient term is to describe it as a joint business venture, without the need to go further and place it in some jurisprudential pigeonhole. Its importance is that it established the terms of the contractual bond between the parties. In my view those terms make it sufficiently similar to a partnership or joint venture in the conventional sense that the co ownership of the letting enterprise (the primary asset) and that of the properties (the subsidiary assets) should like partnership be a case of bound co-ownership.
 I am not concerned at the prospect of bound co-ownership being created by way of a commercial agreement between contracting parties. Partnership is a product of a commercial agreement between private contracting parties yet it has always been recognised as bound co-ownership. Going back to the old authorities they always accepted that a partnership would terminate in accordance with the provisions of the partnership agreement.
The co-ownership arising from a marriage in community of property is likewise the result of a private agreement between the parties to the marriage. Both a will and a deed of trust are capable of creating a situation of bound co-ownership where the actio is excluded for a significant period of time, yet they are also the product of private arrangements.
If the co-ownership is exercised through the medium of a company the actio is excluded. The similarities between a conventional shareholders’ agreement and the COA in this case have already been noted.
 In summary, my view is that the correct analysis must start from the primary fact that what was bought and sold was a business, the Letting Enterprise, and that the terms of the co-ownership of the Letting Enterprise, rather than the consequential co-ownership of the immovable properties, should determine whether this is bound or free co-ownership. The legal relationship under the COA is governed by conventional principles of the common law in regard to contracts. It is separate from and extrinsic to the co-ownership of the immovable properties. It is impermissible to separate the co-ownership of the immovable properties from the co-ownership of the Letting Enterprise and the agreement to operate it for the joint benefit of the MEPF and Adamax.
To quote Professor van der Merwe, co-ownership of the immovable properties is not the sole relationship between the parties. It results from another legal relationship, namely the co-ownership of the Letting Enterprise and the COA and it can only be dissolved when the latter relationship is terminated in any manner that may be permissible.
 In the result I make the following order:
1 The appeal is upheld with costs, such costs to include those consequent upon the employment of two counsel.
2 The order of the high court is set aside and replaced by the following order:
‘The application is dismissed with costs, such costs to include those consequent upon the employment of two counsel.’
“Actio Communi Dividundo – Contract providing for co ¬ownership of property in undivided shares and letting enterprise – whether co-ownership an instance of bound or free co-ownership – whether resort to actio competent.”