---

SAFLII Disclaimer: This is an unofficial translation of the judgment in Brisley v Drotsky (432/2000) [2002] ZASCA 35 (28 March 2002) produced for SAFLII. The original Afrikaans text takes precedence and should be cited at all times. The paragraph numbers in this translation have been altered to run sequentially. SAFLII has received this unofficial translation from a third party who prefers to remain anonymise.

 

THE SUPREME COURT OF APPEAL OF SOUTH AFRICA

Reportable

Case no: 432/2000

In the appeal of:

MADELEIN BRISLEY                                                                              Appellant

and

ANTOINETTE DROTSKY                                                                        Respondent

CORAM:                   HARMS, OLIVIER, STREICHER, CAMERON and

                                  BRAND JJA

Date of hearing:         7 March 2002

Date of judgment:     28 March 2002

Summary:                  Shifren case affirmed; bona fides in contract discussed; section 26(3) of the Constitution interpreted.

JUDGMENT

HARMS, STREICHER and BRAND JJA (translated):

[1]     Three questions arise in this appeal, namely (a) the scope of the so-called Shifren principle,[1] according to which a term in a written contract stipulating that any amendment to the contract must comply with certain formalities (a non-variation clause) is binding, (b) in connection with this, the scope of the so-called principles of bona fides in contract law, and (c) the reach of section 26(3) of the Constitution, which states, inter alia, that no one may be evicted from his or her home without a court order and after considering all relevant circumstances. We deal with these questions in turn.

[2]     The appellant (the tenant) agreed a lease with the respondent (the lessor) in terms of which the tenant rented a townhouse in exchange for payment of a rental fee of R3500.00, payable on the first of each month. The tenant failed from the outset to pay timeously, and when she had not yet paid the January rent in full by 31 January 2000, the lessor terminated the tenancy and allowed the tenant fourteen days to vacate. Motion proceedings followed in which the lessor sought to evict the tenant. Every conceivable delaying tactic was used, assisted by courts which repeatedly refused to adjudicate the matter on the merits. As the case dragged on, the tenant embellished her version of events, to the extent that one seriously doubts her credibility. Be that as it may, the case is and will be judged on the tenant’s version, because the lessor has opted for motion proceedings.

[3]     Vorster AJ granted an eviction order. Without explanation, but apparently in reliance on court rule 45A, he gave her ten days to vacate the premises. Although the contract provided for a costs order on the attorney-client scale, he did not grant one, merely because he was not ‘inclined’ to do so. Evidently he did not take into account that in this regard he had only a limited discretion.[2] But the lessor has not filed a cross-appeal against either the ten-day interval or the limited costs award, and accordingly these aspects were not argued during the appeal and merit no further consideration.

[4]     The contract between the parties is a so-called “CNA contract”, in other words a standard form contract which one can buy at a well-known stationery store. As mentioned, it provides for the advance payment of rent on or before the first of each month. It provides further that, in the event that rent is not paid timeously, the lessor may cancel the contract immediately. An acceptance of late payment does not prevent the lessor from cancelling the contract by reason of any other failure to pay on time. Postponements, concessions, failures to enforce the agreement’s terms, etc, do not affect the lessor’s rights. Reliance upon estoppel and waiver are limited. And then there is the offending provision:

‘No alteration, variation or cancellation of any of the terms or conditions of this LEASE shall be of any force or effect unless it is recorded in writing and signed by the parties thereto.’

[5]     The tenant’s defence to the eviction application was initially threefold: she relied on a later oral agreement that allowed her, for the first six months, to pay the rent ‘as it suits me in the course of the month in question’; she relied on estoppel because the lessor suggested to her that she could pay late; and she made only partial payment in January because she had had to bear the expense of repairing the sewage system. After she was given the opportunity, in supplementary pleadings, to invoke section 26(3) of the Constitution, she exploited this opportunity to raise a rectification defence: the monthly contract was, in fact, for three years, and the parties’ payment arrangement was now made before they concluded the contract. Her version about the sewage system also acquired a new dimension.

[6]     THE ATTACK ON SHIFREN: The Shifren principle must, we believe, be seen in its historical and jurisprudential context. Shortly beforehand, there had been a constitutional crisis when the parliament of the time sought to bypass an entrenched statutory provision with a simple majority vote. This court held in the first Harris case[3] that this could not be done. In Shifren, the argument made was analogous to that in Harris: self-imposed restrictions can be undone in just the same way they were brought into being, namely by simple agreement; further, limiting the parties’ ability to amend the restricting clause orally would be a limitation on their power to exercise their contractual freedom (parliamentary sovereignty).

[7]     Aware of the then-existing objections (which are still raised today – in different words, but in substance the same) from academic and other sources against the enforcement of the non-variation clause, this court attached greater value to the parties’ initial exercise of their contractual freedom than to their power to undo those initial choices without restraint. The legislature often does this by prescribing that certain types of contract must be in writing, as must all amendments to them. The parties do this by agreeing in advance that a contract comes into being only when certain formalities are complied with. The purpose is to limit or prevent disputes. Naturally the parties remain free to ignore the formalities and to behave as if a particular law does not exist. But if a dispute arises, anyone is entitled – and the court is obliged – to apply the strict law. And why should it be otherwise in an autonomous contractual relationship? There is also a common myth that this type of provision exists only for the benefit of the economically powerful and that it produces inequality in contractual relationships. This is probably why the constitutional principles of equality was relied upon. But this provision serves to protect both parties. One can only wonder how the tenant would have reacted if the lessor had claimed that an increased rent had been agreed orally.

[8]     The Shifren principle is ‘trite’[4] and the question that arises is why, after almost forty years, it should be overturned. One can hardly imagine the commercial implications, legal uncertainty, and evidentiary problems that would arise.[5] The tenant argues, however, that the principle can be shown to be absurd, and refers in this connection to ‘pithy’ remarks by Hiemstra J that ‘made a fool’ of this court.[6] Of course one can think of cases where the principle may lead to less acceptable results, but this court was well aware of that. A court often has to make a policy choice between two conflicting views, with weighty arguments on both sides. The one answer is not necessarily ‘right’ and the other ‘wrong’ in absolute terms. but when a choice has been made, we should stand by it, unless there are good reasons not to. Non-variation clauses are not contra bonos mores in principle; after all, our system of constitutionalism is grounded upon an analogous idea, which is often embodied in legislation.

[9]     In addition, the Shifren principle does not involve unreasonable ‘coercion’. The general principles of contract law still apply. The courts have often come to the assistance of a party in the past, albeit sometimes on dubious grounds.[7] The tenant’s counsel advanced each and every one of these solutions during argument, such as waiver, pactum de non petendo, alteration by conduct, election, rectification, collateral agreements, estoppel and, last but not least, bona fides (more on this later). He repeatedly ran aground on either the other terms of the contract or the facts. The agreement provides, for example, as already mentioned, that the conduct of the lessor cannot amount to an amendment, waiver or estoppel. And the tenant’s reliance upon rectification is not only in conflict with her founding affidavit, which alleged an amendment, but is otherwise untenable. She accepts that she read the contract before signing it, and that she discussed it with her advisor and her mother. There is also no evidence that the lessor would have been willing to contact on terms other than those contained in the written document; indeed she insisted on the written contract. And in terms of that contract, money could not be withheld by the tenant for the so-called sewage costs.

[10]   The consequences of the attack on Shifren are even more far-reaching: would the other clauses that provide, for example, for a ban on estoppel, novation or waiver also now be open to question? If so, on what grounds? Though contract law promotes the value of legal certainty, that would now seem to be built on sand. Because counsel could provide no answers to these questions, he was compelled to argue that the written contract had been terminated by the parties and that an oral contract of lease was concluded. But this was never the tenant’s case, and there are no facts upon which this argument could be sustained.

[11]   BONA FIDES: We deal next with the bona fides question. The argument, according to the tenant’s counsel, was that the lessor’s reliance upon the non-variation clause should not be enforced because in the circumstances it would be unreasonable, unfair and in conflict with the principles of bona fides (good faith). Here he relied upon the decision of Ntsebeza AJ in Miller and Another NNO v Dannecker 2002 (1) SA 928 (C) as well as the aforementioned article by Prof Dale Hutchison which appeared after the decision had been given.

[12]   Prof Hutchison starts from the premise (at 721) that the court has a general discretion either to enforce or not to enforce a non-variation clause. The premise is, in our view, mistaken. In principle, a court has no discretion to refuse to enforce a valid contract term (which non-variation clauses plainly are).[8]

[13]   The Miller case does support the tenant’s argument that the court may refuse to enforce a non-variation clause if it would amount to a breach of the principle of good faith. In our view, however, this decision cannot be accepted as correct.

[14]   As authority for his view that he was permitted, based upon considerations of good faith, to depart from the decision of this court in Shifren, Ntsebeza AJ relied especially upon the minority judgment of Olivier JA in Eerste Nasionale Bank van Suidelike Afrika Bpk v Saayman NO [1997] ZASCA 62; 1997 (4) SA 302 (SCA) 318 et seq. It seems the minority judgment was regarded as authority for two propositions: first, that decisions of this court may be departed from if their application would, in the judge’s view, be contrary to the principles of good faith; and second, that considerations of good faith provide a self-standing, independent basis for the invalidation or non-application of contractual terms and the principles of contract law.[9]

[15]   That the first proposition cannot be correct is self-evident. According to the doctrine of precedent, lower courts are bound by the judgments of this court and cannot prefer the opinions of single judges of this court, however persuasive, over them.

[16]   As far as the second point is concerned, it is indeed so that Olivier JA made a plea that good faith be given a more prominent place in our contract law. The conclusion he reached for the purposes of his judgment, namely that a commercially competent person can be relieved of liability on the basis of considerations of fairness, is also far-reaching. In our view, the statements in his judgment should be approached with caution. In the first place, it is a minority judgment representing the views of a single judge based upon an account of the facts with which the other four judges did not agree. Second, the issue was not argued in the case. Third, the majority judgment did not give any indication that this interpretation of the law is correct. Admittedly in NBS Boland Bank v One Berg River Drive CC 1999 (4) SA 928 (SCA) 937G the remark was made that—

‘An analogous conclusion may well be reached if one applies the modern concept of public policy, bona fides and contractual equity to the question in issue (see, for example, Eerste Nasionale Bank van Suidelike Afrika Bpk v Saayman NO [1997] ZASCA 62; 1997 (4) SA 302 (SCA) 318–331, per Olivier JA).’

Although these remarks appear to indicate approval of the minority judgment in Saayman, they were made in passing and without analysis. Moreover, the remarks did not form part of the ratio decidendi. Hence the views expressed in Olivier JA’s judgment remain those of a single judge only.

[17]   In addition, the conclusion of Olivier JA’s judgment is based in large part on dubious grounds. We content ourselves with a few examples. The judgment, so it seems to us, seeks to revive the exceptio doli generalis indirectly and to introduce substantially the same principles by another name.[10] The judgment’s starting point is a quotation from Neugebauer & Co Ltd v Hermann 1923 AD 564 at 573 in which Innes CJ made a general statement about bona fides in contractual relationships. The problem is that the quotation does not convey the essence of the judgment. What was decided was that, when a bidder commits fraud at an auction, he cannot invoke the auction’s terms. That this was about fraud was something Innes CJ repeatedly made clear.

[18]   The second judgment Olivier JA relied upon was MacDuff & Co Ltd (in liquidation) v Johannesburg Consolidated Investment Co Ltd 1924 AD 573, where there were also general remarks made about bona fides in contractual relationships. It was a case in which the court decided that, if a party intentionally (translated as ‘dolo’) prevents the fulfilment of a condition, the condition must be deemed to be fulfilled. It is noteworthy that the court did not use a lack of bona fides as the criterion.

[19]   In relation to foreign law, Olivier JA (at 329–331) found support for his position in the ‘special equity theory’ of Scott LJ in Barclays Bank plc v O’Brien and Another [1992] 4 All ER 983 (CA), a case that dealt specifically with suretyships. Besides the fact that the principles of our law differ substantially from English law, the ‘special equity theory’ was rejected by the House of Lords in Barclays Bank plc v O’Brien and Another [1993] UKHL 6; [1993] 4 All ER 417 (HL).

[20]   It is now said that the idea that Shifren should not be applied if that would be contrary to norms of bona fides was already present in a passage in Resisto Dairies (Pty) Ltd v Auto Protection Insurance Co Ltd 1962 (3) SA 565 (C) 571E–F (per Rosenow J). First, this decision was given before Shifren. Moreover, it was overturned on appeal and it appears from the judgment of this court that it was exclusively about estoppel.[11] Still more important is the authority to which Rosenow J referred. The one judgment dealt at the points identified with the exceptio doli generalis,[12] and the other, Wells v South African Aluminite Co 1927 AD 69 73 (incidentally by Innes CJ again), is instructive:

‘No doubt the condition [a provision that the purchaser may not rely on a misrepresentation] is hard and onerous; but if people sign such conditions they must, in the absence of fraud, be held to them. Public policy so demands.’

[21]   It appears from reported decisions that Olivier JA’s judgment has already given rise to much debate. Van Zyl J[13] interpreted the judgment as authority for the proposition that the age-old requirements of the actio quanti minoris may be departed from when the judge considers these requirements to be unfair. And Davis J[14] found in the judgment a basis for the idea that the boni mores, or the convictions of the community, which apply in the law of delict, are valid also in the law of contract. If it is meant that the enforceability of contract terms depends on the views of the community, we cannot agree with it. There are significant policy differences in the approach to contracts and that which applies to delicts. In the former, the parties arrange their legal relationship voluntarily and consider themselves bound by their expressions of will. They determine the nature and extent of their legal relationship. In the case of delicts, the parties have no say in the creation of their legal relationship, and the community’s convictions determine whether a legal relationship should exist and what its content should be. If these decisions are accepted unqualifiedly, it would produce a situation of unacceptable disorder and uncertainty in our contract law.

[22]   As far as the role of good faith is concerned, we substantially agree with the opinion of Prof Hutchison (at 743–744) that good faith does not provide an independent, or ‘free-floating’, basis for the invalidation or non-application of contract terms. Good faith is a fundamental principle that underlies contract law and finds expression in its particular rules and principles. Or, as he puts it:

‘What emerges quite clearly from recent academic writing and from some of the leading cases, is that good faith may be regarded as an ethical value or controlling principle based on community standards of decency and fairness that underlies and informs the substantive law of contract. It finds expression in various technical rules and doctrines, defines their form, content and field of application and provides them with a moral and theoretical foundation. Good faith thus has a creative, a controlling and a legitimating or explanatory function. It is not, however, the only value or principle that underlies the law of contract; nor, perhaps, even the most important one.’

[23]   Another value underlying contract law is emphasised by Rabie CJ in Magna Alloys and Research (SA)(Pty) Ltd v Ellis [1984] ZASCA 116; 1984 (4) SA 874 (A) 893I–894A when he pointed out that—

‘it is in the public interest that persons abide by agreements they have entered into. In this connection, Steyn CJ in SA Sentrale Ko-op Graanmaatskappy Bpk v Shifren en Andere 1964 (4) SA 760 (A) at 767A referred to—

“the elementary and fundamental general principle that contracts freely and seriously concluded by competent parties will, in the public interest, be enforced.”’

[24]   The task of courts in general, and of this court in particular, is to weigh up these fundamental values, which sometimes come into conflict with one another, and, when it seems necessary, to make adjustments, gradually and with caution. Or, as Lord Simon of Glaisdale put it in Miliangos v George Frank (Textiles) Ltd:[15]

‘[J]udicial advance should be gradual ... [O]ne step is enough ... It is, I concede, a less spectacular method of progression than somersaults and cartwheels; but it is the one best suited to the capacity and resources of a Judge.’

To all of a sudden give judges a discretion to disregard contractual principles when they have been deemed unreasonable or unfair is in conflict with this method. The result, after all, would be to disregard the principle of pacta sunt servanda, since the enforceability of contract terms will depend on what a particular judge considers to be reasonable and fair in the circumstances. The criterion is then no longer the law but the judge.[16] From the contracting parties’ perspective, they will not be able to act on the general expectation that, when there is a dispute between them, their contract will be enforced according to its terms. They would have to wait and see if the individual judge regards the terms as reasonable and fair. That such a general approach will not meet the needs of commercial practice speaks for itself.[17] A court cannot seek refuge in the shadow of the Constitution to attack and overthrow principles from there; albeit that the law, in light of the Constitution, can and must be developed. In this context, constitutional values were vaguely referred to, without being specific. Wide judicial discretion is not such a value and we are not able ‘to discren any societal value which is imperilled’[18] by the application of Shifren or by the refusal to introduce a ‘special equity theory’ into contract law.

[25]   With regard to non-variation clauses in particular, it is clear, as was already in part mentioned, that the ultimate decision in the Shifren case was the result of a balancing of various valid considerations. The court therefore asked, for example, whether there was anything wrong or unacceptable in the considerations that the non-variation clause was based on. To this Steyn CJ gave the following answer (at 766H):

‘Their [the contracting parties’] apparent purpose with such a clause is to guard against the disputes and evidentiary difficulties that can arise from oral agreements. To protect both parties against these, they expressly agree that oral amendments ... even when entered into animo contrahendi, will be null and void as between them.’

[26]   It thus remains true that a non-variation clause is not in itself invalid, even though relying upon it necessarily amounts to a refusal to give effect to an oral agreement entered into animo contrahendi. What needs emphasis in the quotation is that non-variation clauses protect both parties – and this was their own choice. The potential inequality of the parties’ bargaining power or financial means, and the protection of weaker contracting parties, therefore do not arise.

[27]   The viewpoint in the Miller case on which the tenant relies essentially amounts to the claim that, although no objection can be made to the non-variation clause itself, an attempt to apply the clause is contrary to the principle of good faith. Pure logic shows that such a view is untenable. The alternative argument was that, even if the invocation of the non-variation clause would not in general be contrary to the principle of good faith, its enforcement in the present case would be so unfair that the court should refuse, in the public interest, to enforce it. As authority for this argument, reliance is placed on Magna Alloys and on Sasfin (Pty) Ltd v Beukes 1989 (1) SA 1 (A).

[28]   Magna Alloys deals with the enforceability of an agreement in restraint of trade, which differs in two respects from a non-variation clause. First, as is clear from the judgment (at 893–894), in the enforcement of a restraint of trade there are two fundamental public policy considerations that come into conflict and must be weighed against one another. On the one hand, there is the consideration expressed in the maxim pacta sunt servanda. On the other is the consideration that everyone should be allowed to freely exercise his trade or practice his profession. This last-mentioned consideration does not apply to non-variation clauses.

[29]   A further aspect of restraints of trade appears from the following remarks of Botha J in National Chemsearch SA (Pty) Ltd v Borrowman and Another 1979 (3) SA 1092 (T) 1107E–H:

‘At first sight it might be thought strange that a contract can be void at one point of time and valid at another. But that, I believe, is not the right way of looking at a situation where the decisive question is whether the Court should decline to enforce a contract in restraint of trade because of considerations of public policy. In an article in 1941 SALJ 335 at 346, Aquilius gives the following definition of a contract against public policy:

“A contract against public policy is one stipulating a performance which is not per se illegal or immoral, but which the Courts, on grounds of expedience, will not enforce, because performance will detrimentally affect the interests of the community.” (My italics.)

With regard to agreements in restraint of trade, it seems to me that the Court’s concern is to assess the effect of an order enforcing the agreement in the light of the dictates of public policy, and that the proper time for making that assessment is the time when the Court is asked to make the order, taking into account the relevant circumstances existing at that time. Public policy does not require the Court to penalise the party seeking to enforce the agreement, by declining to do so, because at the time when it was entered into it was so worded that it could not be accurately forecast whether it would be reasonable or not to enforce it when the occasion for its enforcement should arise.’ (Underlining added.)

[30]   Because of these considerations, an agreement in restraint of trade already carries with it, upon its conclusion, the potential that it would later prove to be unenforceable, and the parties accept this risk. In respect of non-variation clauses, however, the effect of Shifren is precisely that it does not carry this risk. Accordingly, the considerations that apply to agreements in restraint of trade do not apply to non-variation clauses.

[31]   Sasfin applied the principle that contract terms that are so unfair that they are contrary to public policy are, for that reason, invalid. Seeing as the non-variation clause is not in itself invalid, the Sasfin principle has no direct application. On the assumption that the Sasfin principle can be broadened to prevent the enforcement of contract terms (which are not per se in conflict with public policy), such application must necessarily be limited to cases analogous to Sasfin, in other words cases where the enforcement of the non-variation clause would be so unfair that it can be described as ‘inimical to the interests of the community’ (at 8C–D). Moreover, the considerations appearing from the following dicta of Smalberger JA (at 9B–E) will also apply:

‘The power to declare contracts contrary to public policy should, however, be exercised sparingly and only in the clearest of cases, lest uncertainty as to the validity of contracts result from an arbitrary and indiscriminate use of the power. One must be careful not to conclude that a contract is contrary to public policy merely because its terms (or some of them) offend one’s individual sense of propriety and fairness. In the words of Lord Atkin in Fender v St John-Mildmay 1938 AC 1 (HL) at 12:

“... the doctrine should only be invoked in clear cases in which the harm to the public is substantially incontestable, and does not depend upon the idiosyncratic inferences of a few judicial minds”

...

In grappling with this often difficult problem it must be borne in mind that public policy generally favours the utmost freedom of contract, and requires that commercial transactions should not be unduly trammelled by the restrictions on that freedom.’[19]

[32]   Even if the Sasfin principle were to apply, the tenant’s case falls far short, in our view, of this strict test of extraordinary unfairness. Even on the tenant’s version, the lessor is merely invoking the terms of the non-variation clause. The purpose of the non-variation clause is, inter alia, to avoid a dispute about the existence of a later oral agreement. That is precisely the purpose for which the lessor now wishes to use it. If the tenant were entitled to make this argument, it would mean that the unfairness of relying upon the non-variation clause would be bound up with the existence of an oral variation; in effect this would, through a circular argument, allow in by the back door the very dispute that the non-variation clause seeks to prevent.

[33]   The tenant’s further objection is that the lessor cannot, without notice, rely upon the strict terms of the lease after having in the past allowed the tenant to deviate from those terms. The question is: why not? First, there are provisions in the agreement that expressly exclude reliance upon estoppel or waiver merely on the ground that late payments were allowed in the past.[20] Second, it is not clear why the tenant is now in a worse position than she would have been in if the lessor had cancelled the lease after the first late payment.

[34]   For these reasons, the tenant’s reliance upon the principle of good faith also fails.

[35]   SECTION 26(3) OF THE CONSTITUTION: This brings us to the last part of the argument, namely that, even if the lease was validly cancelled, the court a quo should not have granted an eviction order because of section 26(3) of the Constitution. The section provides as follows:

‘No one may be evicted from their home, or have their home demolished, without an order of court made after considering all the relevant circumstances. No legislation may permit arbitrary evictions.’

According to the tenant, the mere fact that the lessor was the owner and that the tenant was unlawfully occupying the property does not justify an eviction order. She argues that the circumstances in which the lease was cancelled, as well as her and her mother and child’s socioeconomic circumstances, were relevant circumstances that the court a quo should have taken into consideration.

[36]   The tenant relies upon the decision in Ross v South Peninsula Municipality 2000 (1) SA 589 (C). In that case, the lessor applied for the eviction of the tenant on the ground that he was the owner of the property at issue and that the tenant occupied the property without any legal right to do so. On appeal, the court, per Josman AJ, decided that section 26(3) had altered the common law as laid down in Graham v Ridley 1931 TPD 476. According to the common law, as it was before the Constitution came into effect, a plaintiff who alleged and proved that he was the owner of property and that the defendant was in possession of it was entitled to an eviction order, unless the defendant could show that he had a right to occupy the property.[21] According to Josman AJ, section 26(3) had the result that a plaintiff wishing to evict an occupier in unlawful occupation of his property may no longer succeed on the basis of those facts alone, not even if the case is unopposed by the occupier. The plaintiff must, over and above that, provide additional relevant circumstances that justify the court in making such an order (at 596H).

[37]   What more the plaintiff must allege and prove, Josman AJ did not say. According to him, it fell outside the scope of the appeal before him to consider precisely which circumstances would be relevant (at 596I). He believed, however, that the Prevention of Illegal Eviction from and Unlawful Occupation of Land Act 19 of 1998 could be helpful in this regard, more specifically section 4(6) and (7), which grants a court a discretion to make an eviction order in certain circumstances if the court considers it just and fair, having taken into account the needs of the elderly, children, disabled persons, and households headed by a woman (at 599B–C).

[38]   Obviously the factors that are relevant in the case where a court has a discretion do not indicate what factors are relevant in the case where the court has no direction. The judge therefore erred in taking the Act to indicate the relevant circumstances for the purposes of section 26(3) before he had established that section 26(3) grants the court a discretion and what the nature of that discretion is.

[39]   The court a quo decided, following Betta Eiendomme (Pty) Ltd v Ekple-Epoh 2000 (4) SA 468 (W) 473A–B, that Ross was wrongly decided, and that section 26 has vertical effect only and thus has no application to natural persons and legal persons other than organs of state.

[40]   Although section 26(1) and (2) probably has vertical effect only and the last sentence of section 26(3) places a duty upon the state not to pass certain legislation, the court a quo decided, wrongly in our view, that this applies to the pertinent part of section 26(3). Section 8(2) of the Constitution provides:

‘A provision of the Bill of Rights binds natural and juristic persons if, and to the extent that, it is applicable, taking into account the nature of the right and of any duty imposed by the right.’

There is no reason why section 26(3) cannot be applied (or should not be applied) to all natural and legal persons. In Betta the court decided that the whole of section 26 has vertical effect only because it could not be expected of persons such as the applicant in that case to provide housing to the general public (‘all comers’) or to the tenant in that case (at 473A). But those considerations are no reason for the finding that section 26(3) does not have horizontal effect.

[41]   It follows that, even in a case such as the present, section 26(3) requires a court, before making an eviction order, to consider all relevant circumstances. Davis, Cheadle and Haysom[22] write:

‘The relevant circumstances which the court must consider before it makes such an order should include at least the personal circumstances of those being deprived of accommodation, and the availability of alternative accommodation. These were among the factors which were considered relevant to the exercise of a court’s discretion in issuing an ejectment order in terms of s 46(2) of the repealed Group Areas Act 36 of 1966 (S v Govender 1986 (3) 969 (T) at 971H–J). These factors should be weighed against the reasons for the party seeking the order (for example, an illegal occupation, a default on bond or rent payments, the resettlement of a community in another area).’

[42]   We do not agree that the abovementioned circumstances will be relevant without more. Section 26(3) requires all relevant circumstances to be taken into account, but does not itself stipulate that any circumstances will be relevant. For that, one must refer to the generally applicable law. Circumstances will be relevant only if they are legally relevant. If the section has granted a court a discretion to refuse an eviction order in certain circumstances, for example where the court considers it just and fair, all the circumstances that are relevant to the question of whether it is just and fair in a particular case will of course be relevant to the exercise of that discretion. But the section does not confer any discretion on the court to refuse to grant an eviction order to an owner who would otherwise have been entitled to one.

[43]   Legally, an owner is entitled to possession of his property and to an eviction order against a person who unlawfully occupies his property, unless that right is limited by the Constitution, another law, a contract, or on some other legal basis. An example of such a limitation is found in the Prevention of Illegal Eviction from and Unlawful Occupation of Land Act, which, as indicated above, subjects an eviction order, in the circumstances mentioned in that Act, to the exercise of a discretion by the court.[23] And section 26(2), which imposes certain housing obligations on the state, might in particular cases place such a limitation upon the state’s ownership rights. For purposes of this case, however, it is not necessary to decide if that is indeed so.

[44]   In S v Govender, to which Davis, Cheadle and Haysom refer in the passage quoted above, the court did not deal with the right of an owner to an eviction order against an occupier. In that case, the court decided that a court that finds a person guilty of violating the now repealed Group Areas Act 36 of 1966 has, in terms of that law, a wide discretion to evict that person from the land occupied by him. Section 46(2) of the Act provides, in other words, that the court which finds a person guilty of infringing certain sections in the Act, in addition to any punishment issued, can (‘may’) grant an eviction order. Goldstone J said in this regard:

‘The power to make such an ejectment order is a wide one. It is one which may, and in most cases will, seriously affect the lives of the person or persons concerned. It may, and frequently will, interfere with the normal contractual relationship which exists between landlord and tenant. Such an order should not therefore be made without the fullest enquiry. Whether such enquiry takes place before or after sentence does not appear to me to matter. However, a court should not make such an order unless requested to do so and there appears to me to be no onus upon the convicted person to dissuade the court from granting the order.’

The circumstances to which Goldstone J referred were seen as relevant by him because the statute grants the court a discretion.

[45]   In the present case the lessor is the owner of the property in question. The tenant’s lease was cancelled and she therefore has no contractual right to occupy the property. The tenant also did not argue that she had any statutory right, other than that which may be found in section 26(3), to occupy the property, and the court has no discretion to refuse an eviction order. The circumstances which the tenant claims the court must take into account before granting an eviction order are therefore not relevant circumstances that must be considered in terms of section 26(3). As owner, the lessor is entitled to possession. If the tenant does not have a right to possession and the court does not have a discretion to nevertheless refuse an eviction order, then the only relevant circumstances that the court may take into account are the fact that the plaintiff is the owner and the fact that the defendant is in possession. Ross was therefore wrongly decided.

[46]   It follows that the court a quo did indeed consider all relevant circumstances and rightly found that the tenant [sic] is entitled to an eviction order.

[47]   The appeal must therefore be dismissed with costs. As mentioned, the contract provides that the tenant is liable for costs on an attorney-client scale. No reasons were given why the agreement should not be upheld in this regard too.

The appeal is dismissed with costs assessed on an attorney-client scale.

________________

LTC HARMS

JUDGE OF APPEAL

________________

PE STREICHER

JUDGE OF APPEAL

________________

FDJ BRAND

JUDGE OF APPEAL

OLIVIER JA (translated):

[48]   Simply stated, the main question in this appeal is whether the principle laid down in SA Sentrale Ko-op Graanmaatskappy Bpk v Shifren en Andere 1964 (4) SA 760 (A) (“the Shifren principle”) should be departed from, or whether it should be maintained. The answer to this question by and large determines the outcome of the appeal.

[49]   In terms of a written contract of lease dated 15 September 1999, the respondent (“Drotsky”) leased a townhouse she owns in Pretoria to the appellant (“Brisley”). The rent amounted to R3 500,00 per month, payable in advance on the first day of each month starting on 1 October 1999. Brisley failed to fulfil this obligation and paid as follows:

(i)      For October 1999:

On 18 October                        R 3 000,00

On 19 October                         R 500,00

 

(ii)     For November 1999:

On 12 November                     R 3 500,00

 

(iii)    For December 1999:

On 10 December                      R 3 500,00

 

(iv)    For January 2000:

On 15 January                          R 3 000,00

 

(v)     For February 2000:

On 9 February                          R 3 500,00

[50]   The written contract of lease provided that, in the event of late or non-payment of rent, Drotsky had the right to terminate the lease summarily. This she did on 31 January 2000 by delivering a written notice of termination to Brisley in which the latter was told to vacate the premises within 14 days.

[51]   Brisley refused to comply with this notice, with the consequence that Drotsky filed an application in the Transvaal Provincial Division on 31 March 2000 requesting a declaration that the lease had been cancelled and an order of eviction.

[52]   Brisley opposed the application. She does not deny the dates of the rental payments as set out above. She also does not deny that, according to the terms of the written contract, the payments were late. What she claims is that, after the conclusion of the contract of lease—

‘... I approached the applicant [Drotsky] for relief from the requirement that rent be paid promptly on the first day of the month and asked that I be allowed to pay it belatedly for the first six months, and that after that I should pay it on or before the 7^th of each month.

The applicant said that she didn’t mind at all; that the written contract is a “CNA contract” which she used only to have something in writing; and that I can make payment as suggested, and as it suits me over the course of the month in question, and that she waived the prompt payment on the first of the month; furthermore, that I must pay directly into her bank account.’

[53]   In regard to the reduced payment of R500,00 in January 2000, Brisley argued that this was money she withheld for the expense of repairing the sewer. She claimed that Drotsky was aware of this and accepted that the part-payment was in order.

[54]   In light of the factual dispute between the parties, Drotsky’s application could not succeed unless it was adjudicated after a trial. But this referral to a trial was not done or requested by either of the parties. Drotsky’s response was different: the defence Brisley relied upon in respect of both the late and incomplete payment (i.e. a permission granted in an oral agreement between the parties after the conclusion of the written lease) is legally inadmissible in virtue of clause 19 of the written contract, which reads as follows:

“19.0    ENTIRE AGREEMENT, VARIATION AND WAIVER:

19.1     This LEASE constitutes the entire agreement between the parties as to the subject matter hereof. No warranties, representations, stipulations or conditions, other than as recorded herein, shall be binding on the LESSOR.

19.2     No alterations, variation or cancellation of any of the terms or conditions of this LEASE shall be of any force or effect unless it is recorded in writing and signed by the parties hereto.

19.3     No latitude, indulgence, consent or forbearance or any other similar act by the LESSOR in enforcing any provision of this LEASE shall constitute a variation or novation of this agreement or a waiver of or estoppel in respect of its rights in terms of this agreement.

19.4     Without derogating from the generality of the aforegoing:

19.4.1              no acceptance by the LESSOR of any rent, or any other payments after due date (whether on one or more occasions), shall preclude the LESSOR from exercising any of its rights under this LEASE by reason of any subsequent payment not being made strictly on due date;

19.4.2              the receipt by the LESSOR of any rent or other payment shall in no way prejudice, or operate as waiver, recission or abandonment of any cancellation effected or right of cancellation acquired prior to such receipt.”

[55]   Brisley’s reliance on an oral modification, and the counter-reliance on clause 19, places the applicability of the Shifren principle inexorably before us. This principle holds that, where contracting parties have agreed that no later modifications of the contract terms are valid unless they comply with certain formalities, a modification (including a modification of the entrenched clause itself) that fails to fulfil those formalities will be unenforceable.

[56]   The court below rightly regarded clause 19 of the lease as a term of this kind, applied the Shifren principle, and granted Drotsky’s application. It is against this decision that Brisley, with leave of the court below, has appealed.

[57]   The Shifren decision has often been questioned in the legal literature and even described as ‘surprisingly naïve’, a denial of the estoppel principle, contrary to the requirements of good faith and the public interest, and out of date, given the new approach in our contract law to contractual justice. (See JS McLennan, The Demise of the 'Non-variation' clause in Contract, 2001 (118) SA Law Journal 574–580). It is also argued that, unless Shifren is appropriately qualified by recognized common-law doctrines, it should be abandoned as inconsistent with current constitutional principles. (Dale Hutchison, Non-variation clauses in contract: Any escape from the Shifren straitjacket? 2001 (118) SA Law Journal 720–746.)

[58]   It was argued for the respondent that, despite the academic criticism, this court will not deviate from its own decision in Shifren because of the principle of stare decisis, which this court formulated as follows in Bloemfontein Town Council v Richter 1938 AD 195 at 232:

‘The ordinary rule is that this court is bound by its own decisions and unless a decision has been arrived at on some manifest oversight or misunderstanding that is there has been something in the nature of a palpable mistake a subsequently constituted court has no right to prefer its own reasoning to that of its predecessors – such preference, if allowed, would produce endless uncertainty and confusion. The maxim "stare decisis" should, therefore, be more rigidly applied in this the highest court in the land, than in all others.’

This court can also depart from a previous decision if

‘... its attention was not drawn in the previous decisions to relevant authorities’.

See John Bell and Co Ltd v Esselen 1954 (1) SA 147 (A) at 153 in fine.

[59]   It was argued further on behalf of the respondent that Shifren cannot be said to have rested on a manifest oversight or misunderstanding or mistake. Before Shifren, there were several conflicting provincial decisions about the binding nature of a non-variation clause of the kind currently under discussion. For a useful overview, see PMA Hunt in 1963 Annual Survey of South African Law 137–141. Hunt’s overview – written before the Shifren decision in this court – strongly argued against the acceptance of non-variation clauses’ absolutely binding force. The debate was therefore nothing new. Shifren allowed a particular point of view to prevail due to policy considerations, for reasons of convenience, and in line with the court’s point of view. It cannot rightly be said, therefore, that Shifren rested upon a clear oversight or misunderstanding or that any relevant authorities were overlooked by the court.

It was thus argued on behalf of Drotsky that, even if doubts remain about Shifren’s correctness, this court should not lightly depart from it, particularly since it was a considered and unanimous decision (see too Tuckers Land and Development Corporation (Pty) Ltd v Strydom 1984 (1) SA 1 (A) at 16G–17D; Government of Lebowa v Government of the Republic of South Africa and Another 1988 (1) SA 344 (A) at 361B–D; Catholic Bishops Publishing Co v State President and Another 1990 (1) SA 849 (A) at 866 G–J).

[60]   Speaking generally, I cannot find fault with the said approach. Mr Pieters, who appeared on behalf of Brisley in this court, could point to no oversight or misunderstanding or disregard of binding authority in the Shifren decision. He did say, however, that the decision rested upon an error of logic. He argued that this court committed a ‘category mistake’ in Shifren, conflating the material clauses of the contract, on the one hand, and the formal clauses or procedure for concluding the contract, on the other. He said these are of different natures, and that the non-variation clause can apply only to the first one. Therefore the parties could have validly concluded an oral agreement to modify or depart from the procedures on which they had previously agreed.

[61]   This criticism – and the logic behind it – are unpersuasive. The difference between material and procedural clauses is one created by the advocate and has no substance. Formal requirements for validly concluding a contract are just as binding as the requirement of consensus as to its contents. They are no less essential. Indeed, insisting on compliance with the formal requirements protects the material content of a contract against the kind of disputes that often arise from alleged oral agreements.

[62]   It was argued further, on behalf of Drotsky, that the Shifren principle has now been accepted as valid law for almost 40 years; that a Shifren clause appears in almost every written contract; that it is an important part of commercial life; and that any watering-down of it at this point would cause considerable legal and commercial uncertainty, not to mention a run on the courts. We were referred to a number of decisions of this court in which the abovementioned principle of continuity and legal certainty has been repeatedly emphasized and applied, among others Harris and Others v Minister of the Interior 1952 (2) SA 428 (A) at 454A–B; John Bell and Co Ltd v Esselen, supra, at 154A–B; Attorney General, Northern Cape v Brühns 1985 (3) SA 688 (A) at 701C–D; Cullinan v Noordkaaplandse Aartappelkernmoerkwekers Koöperasie Bpk 1972 (1) SA 761 (A) at 767F–768A; Tuckers Land and Development Corporation (Pty) Ltd v Strydom 1984 (1) SA 1 (A) at 16G–17D. The principle applies all the more, so it was argued, where this court has in recent times referred to Shifren and applied the relevant principle – compare Barclays Western Bank Ltd v Ernst 1988 (1) SA 243 (A) at 253J; Randcoal Services Ltd and Others v Randgold and Exploration Co Ltd 1998 (4) SA 815 (A) at 841E–G; Golden Fried Chicken (Pty) Ltd v Sirad Fast Foods CC and Others 2002 (1) SA 822 (HHA) at 826F.

[63]   It is hard to criticize, as a general guideline, the argument just stated on Drotsky’s behalf. To bury the Shifren principle in its entirety could create commercial chaos, place a question mark behind every existing contract, and, worst of all, be counter-productive in light of the investments and undertakings that in the long run serves everyone’s interests, such as the construction and letting of rental properties. If the risk of disputes and lawsuits arising from alleged oral agreements cannot be effectively eliminated, prudent businesspersons and even private traders will either refrain from their investments and undertakings or resort to other, less desirable protections. There is thus much to be said for preserving the Shifren principle.

[64]   Nevertheless it must be conceded that Hutchison is correct when he says the following about Shifren (see (2001) 118 SA Law Journal 720 at 721).

‘The principle in Shifren’s case has consistently been reaffirmed, albeit with the rider in a recent case that non-variation clauses are to be restrictively interpreted since they curtail freedom of contract. It is therefore still good law, despite the fact that the courts have frequently felt uncomfortable about applying the principle, and have resorted to all sorts of ingenious stratagems to avoid doing so. The reason is quite simply that, no matter how logical its theoretical justification, in practice the principle would be productive of injustice if applied without a good deal of discretion and qualification. For, on the face of it, Shifren appears to allow a party to go back on his or her word, even when another has in good faith relied thereon. Take the all too common situation represented by Shifren itself: a contract of lease containing a non-variation clause requires the written consent of the landlord for any cession by the tenant of its rights under the contract; the landlord orally consents to such a cession but later, after the cession has taken place, purports to cancel for breach, averring that the oral agreement is of no force or effect in view of the non-variation clause. To permit the landlord to cancel the contract in such circumstances seems not merely unjust but a violation of the principle that parties to a contract are expected to behave in accordance with the dictates of good faith.’

[65]   Firstly, then, there is the possibility of restrictive interpretation in the present case. Clause 19 has already been quoted above and is unambiguous. It covers all ‘alterations’ and ‘variations’ and therefore cannot be restrictively interpreted per se.

[66]   But what about waiver? In cases of the present kind, waiver can take one of two forms: the entrenched provision itself can be waived, or the right to cancel for breach of contract can be waived (as stated in Shifren at 765B–C). But whichever form of oral waiver is argued, the fact remains that waiver is, in the contexts in which it arises here, a bilateral legal act, in casu an agreement. (Administrator, Orange Free State v Mokopanele [1990] ZASCA 69; 1990 (3) SA 780 (A) at 787F; Resisto Dairy (Pty) Ltd v Auto Protection Insurance Co Ltd 1962 (3) SA 565 (C) at 571A–B; Hutchison at 723 for useful discussion and citations to the literature.) It follows from this that oral waiver amounts in this case to an oral agreement which undoubtedly ‘alters’ the terms of the written contract – and thus falls squarely within the prohibition in clause 19.2. The provisions of clauses 19.2 and 19.3 are in each case so wide that they rule out any reliance upon waiver.

[67]   The same must be said in respect of estoppel. Clause 19.3 forbids any reliance upon estoppel to depart from the terms of the written agreement (see the discussion in Hutchison at 731–739).

[68]   Mr Pieters also argued that in the present case the Shifren principle should not be applied because of public interest considerations, as crystallized in the bona fide norm-complex. This is not a new idea. It was said already in Resisto Dairy at 571E–F:

‘The plaintiff Company agreed in advance to a condition which is hard and onerous, and it seems to me that unless it can be shown that it would, indeed, in the circumstances of this case, be fraudulent or unconscionable, or a manifestation of bad faith, to rely on this condition, effect should be given to it. (Wells v. South African Alumenite Co. 1927 A.D. 69 at p. 73; Zuurbekom Ltd. v. Union Corporation Ltd. 1947 (1) S.A. 514 (A.D.) at p. 537)’

In recent times, the principles of the public interest and good faith have rightly been emphasized again (see among others Sasfin (Pty) Ltd v Beukes 1989 (1) SA 1 (A) at 7I–9G; Magna Alloys and Research (SA) (Pty) Ltd v Ellis 1984 (4) 874 (A) at 891G et seq on the public interest and, in respect of good faith, Eerste Nasionale Bank van Suidelike Afrika Bpk v Saayman NO [1997] ZASCA 62; 1997 (4) SA 302 (A) at 318H et seq; NBS Boland Bank Ltd v One Berg River Drive CC; Deeb v Absa Bank Ltd; Friedman v Standard Bank of SA Ltd 1999 (4) SA 928 (SCA) at 937F–G; Janse van Rensburg v Grieve Trust CC 2000 (1) SA 315 (C) at 326I–327D; Miller and Another NNO v Dannecker 2001 (1) SA 928 (C) at 938 para 19).

[69]   Support for the application of good faith has also been sought recently in the Constitution and its penumbra. Davis J stated in an obiter dictum in Mort NO v Henry Shields-Chiat 2001 (1) SA 464 (C) at 474J–475F:

‘Like the concept of boni mores in our law of delict, the concept of good faith is shaped by the legal convictions of the community. While Roman-Dutch law may well supply the conceptual apparatus for our law, the content with which concepts are filled depends on an examination of the legal conviction of the community – a far more difficult task. This task requires that careful account be taken of the existence of our constitutional community, based as it is upon principles of freedom, equality and dignity. The principle of freedom does, to an extent, support the view that the contractual autonomy of the parties should be respected and that failure to recognise such autonomy could cause contractual litigation to mushroom and the expectations of contractual parties to be frustrated. See G B Glover (1998) 61 THRHR 328 at 334.

          But the principles of equality and dignity direct attention in another direction. Parties to a contract must adhere to a minimum threshold of mutual respect in which the “unreasonable and one-sided promotion of one’s own interest at the expense of the other infringes the principle of good faith to such a degree as to outweigh the public interest in the sanctity of contracts”: Zimmermann (supra at 259–60). The task is not to disguise equity or principle but to develop contractual principles in the image of the Constitution. For an instructive insight into this approach see Derek van der Merwe 1998 Tydskrif vir Suid Afrikaanse Reg 1.

            In short, the constitutional State which was introduced in 1994 mandates that all law should be congruent with the fundamental values of the Constitution. Oppressive, unreasonable or unconscionable contracts can fall foul of the values of the Constitution. In accordance with its constitutional mandate the courts of our constitutional community can employ the concept of boni mores to infuse our law of contract with this concept of bona fides. See in this regard Janse van Rensburg v Grieve Trust CC 2000 (1) SA 315 (C) at 325-6.’

[70]   Hutchison, too, makes a strong plea that bona fides, supported by the Constitution, receive greater recognition in our contract law. In a chapter titled ‘Good faith in the South African law of contract’ in Roger Brownsword, Norma J Hird and Geraint Howells Good Faith in Contract: Concept and Context (1999) 213 at 230-1 he writes:

‘What emerges quite clearly from recent academic writings, and from some of the leading cases, is that good faith may be regarded as an ethical value or controlling principle, based on community standards of decency and fairness, that underlies and informs the substantive law of contract. It finds expression in various technical rules and doctrines, defines their form, content and field of application and provides them with a moral and theoretical foundation. Good faith thus has a creative, a controlling and a legitimating or explanatory function. It is not, however, the only value or principle that underlies the law of contract nor perhaps, even the most important one. In the words of Lubbe and Murray:

“It does not dominate contract law but operates in conjunction (and competition) with notions of individual autonomy and responsibility, the protection of reasonable reliance in commerce, and views of economic efficiency in determining the contours of contract doctrine. However, it will ensure just results only if judges are alert to their task of testing existing doctrines and the operation of particular transactions against the constantly changing mix of values and policies of which bona fides is an expression.”

On this view of things, which seems to be correct, the influence of good faith in the law of contract is merely of an indirect nature, in that the concept is usually if not always mediated by some other, more technical doctrinal device. Thus, for example, while good faith does not empower a court directly to supplement the terms of a contract, or to limit their operation, it might in appropriate cases enable the court to achieve these same results indirectly, through the use of devices such as implied terms and the public policy rule.’

[71]   Though the above perspectives must be endorsed, the difficult question is how bona fides should be applied to contractual disputes. The operation of bona fides in our contract law has not yet been fully explored and given content. This will have to take place over many years and on the basis of many judgments. Eventually, one hopes, a new framework and pattern of thought will emerge.

[72]   In 1988 the majority in this court, per Joubert JA, did not hesitate to deliver a eulogy for the exceptio doli generalis, as appears from Bank of Lisbon and South Africa Ltd v De Ornelas 1988 (3) SA 580 (A). This exceptio was an important legal remedy which gave reasonableness and fairness powerful effect, i.e. it was used in certain cases to soften the strict law. But only a year later this court did not flinch before weighing up contractual freedom and legal certainty against ‘the doing of simple justice between man and man’ (Sasfin (Pty) Ltd v Beukes 1989 (1) SA 1 (A) at 9A–C). Since then, as mentioned, reasonableness and fairness in the form of bona fides have come more and more to the fore. It is clear that our law is in a phase of development in which contractual justice is emerging, more than ever before, as a moral and juridical norm of great importance. This trend will in all likelihood, as academics like Neels have pointed out, be reinforced by constitutional values. (See Jan Neels, ‘Regsekerheid en die korrigerende werking van redelikheid en billikheid (deel 3)’ (1999) 3 TSAR 477 at 489; see too A van Aswegen ‘The Future of South African Contract Law’ in Die Toekoms van die Suid-Afrikaanse Privaatreg (1994) 44, 46–51; A van Aswegen ‘The implications of a bill of rights for the law of contract and delict’ 1995 SAJHR 50 et seq). One can point, in particular, to Chapter 2 of the Constitution, which protects values like freedom, equality, and dignity; section 39(2), which requires that the courts promote the spirit, purport, and objects of the Bill of Rights when developing the common law; and to section 173, which provides that the superior courts have the inherent power to develop the common law taking into account the interests of justice.

          It is also increasingly clear that contemporary community needs, including the protection of weaker contracting parties, require the courts to be more active. I therefore associate myself with Prof CFC van der Walt’s plea in ‘Beheer oor onbillike kontraksbedinge – quo vadis vanaf 15 Mei 1999’ (2000 TSAR 33 at 41) (translated):

‘Almost 80 years after the judgment in Neugebauer & Co Ltd v Hermann and despite the “line of decisions” that followed it, it must be concluded that the courts themselves will probably not reach the point where they will apply relevant values to contract terms. As long as the courts are left to themselves and the system of precedent to distinguish between clauses that will be enforced or not, and between clauses that are valid or not, they will not get there. In the meantime, the courts will probably continue to apply the “fundamental value” of good faith, under the guise of various other legal figures, remedies and discretions. This latter way of working should not be underestimated, of course, from the point of view of judicial activism against unfair contractual situations. Yet unless there is rapid progress with the direct approach, an acceptable balance of rights and obligations in the field of contract law (that which is termed fair and just) will not be achieved by the courts.’

(See too Neels, above, at 489 et seq.)

[73]   The question before us is summarized as follows by Hutchison ((2002) 118 SALJ op 746):

‘To what extent the principle of good faith may be used to overcome a non-variation clause is uncertain. One possibility is that the principle may be employed directly, on the grounds that it affords a judge an equitable, discretionary power, based on public policy, to refuse to enforce a provision in a contract whenever a party’s attempt to rely on the provision is unconscionable or in bad faith. The more widely accepted view is that good faith operates indirectly, in that it is always mediated by other, more concrete rules or doctrines. In terms of the latter view, the courts would be justified, even obliged, to develop the technical rules of the common law to ensure that the Shifren principle is applied in a way that is consistent with the dictates of good faith.’

[74]   Whichever of the routes suggested by Hutchison is followed, the ultimate question is whether the application of the Shifren principle in each particular case is so unreasonable, so socially and ethically unacceptable, that it cannot, or cannot fully, be applied. This approach is no different from the analogous case of agreements in restraint of trade. In Magna Alloys and Research (SA) (Pty) Ltd v Ellis 1984 (4) 863 (A) this court instated the test for the enforceability of such an agreement as the public interest. In turn, the public interest is tested by, among other things, the reasonableness question (see 891C–I; 893H–894E; 898A–E).

[75]   Whether a Shifren clause should be given effect in any particular case, or whether the later oral amendment should be upheld, must therefore, in light of the abovementioned considerations, be based upon the community’s convictions about the public interest and reasonableness – in short, the boni mores.

[76]   How should bona fides, or rather reasonableness and fairness, be applied in contract law? Because of the excellent formulation of the answer to this question by Neels ((1999) 4 TSAR at 700) I quote him verbatim (translated):

‘The courts, as has already been repeatedly argued, should exercise the power of correction in a principled manner and with restraint. The principles of legal certainty and autonomy require that the consensus that the parties have reached, or the reasonable trust that has been created, must serve as the starting point (preliminary judgement). Only in cases where the unreasonableness or unfairness of the preliminary judgement is clear, manifest, obvious or unmistakable (marginal test) should that be corrected in the final judgement, on the basis of newly refined rules and principles.’

[77]   Next, the question is whether this clause should, given the facts of this case, be enforced: in light of the circumstances and the interests of all parties, is it objectively reasonable, in the sense just discussed, to enforce it? By analogy to agreements in restraint of trade, should it be about (inter alia) the interest that the party seeking to rely on the Shifren clause has in its enforcement? To what extent did the other party rely on the oral agreement or change his position as a result? Did the one who relied on the Shifren clause give the other reasonable notice that the oral agreement would no longer be valid, but only the written contract? What are the mutual consequences of upholding, or not upholding, the clause? These and other questions will have to be carefully scrutinized.

[78]   It may be that the proposed approach would introduce a degree of legal and commercial uncertainty, but that is the price one must pay for a virile system of law which views fairness as just as important as legal certainty: a balance must be found between the continuity of the legal system and the reality of social conditions (Neels, 1999 (2) TSAR at 266 et seq; (1998) TSAR at 702, 716–717; (1999) 4 TSAR 685–698).

[79]   In the present case, it cannot be said that the reliance upon the written contract is so unreasonable that it should not be allowed. On the one hand, it is true that, in light of the oral agreement (which for the purposes of motion proceedings must be accepted as proven), Drotsky’s about-face, in relying upon the written document, seems unjustified. Apart from what is said below, there was no reason to take an unapproachable attitude towards Brisley. No notice was given to Brisley of the turnaround and the imminent cancellation. No reasonable opportunity was given to Brisley to comply with the provisions of the written lease in the future. The duration of the oral agreement was very short and from the evidence presented it appeared that comparable alternative accommodation was readily available. Although this is a borderline case, I think that Drotsky’s reliance on the entrenched clauses in the written contract should be allowed.

[80]   But Mr Pieters, on behalf of Brisley, was not content with his attack on Shifren. He argued in the alternative that his client cannot be evicted from the lease premises, by reason of the provisions of section 26(3) of the Constitution of our country, Act 108 of 1996. The English text, until now the only official one, reads as follows:

‘26.      (1)        Everyone has the right to have access to adequate

housing.

(2)        The state must take reasonable legislative and

other measures, within its available resources, to achieve the progressive realisation of this right.

(3)        No one may be evicted from their home, or have

their home demolished, without an order of court made after considering all the relevant circumstances. No legislation may permit arbitrary evictions.’

In the course of the proceedings in the court below, Patel AJ ordered the parties to submit affidavits relating to their ‘relevant circumstances’ in terms of section 26(3). This has been done in abundance.

[81]   The first question is whether s 26(3) of the Constitution applies to an ordinary lease between private individuals or entities like companies. I think so.

[82]   The meaning and socioeconomic scope of s 26 of the Constitution is fully dealt with in Government of the Republic of South Africa and Others v Grootboom and Others 2000 (11) BCLR 1169 (CC) in the judgment of Yacoob J. Although the judgment deals mostly with the state’s duties to provide housing, it follows from what was said that s 26(3) also has horizontal effect, i.e. applies to private relationships. This is expressly recognized in para 35 of the said judgment:

‘A right of access to adequate housing also suggests that it is not only the State who is responsible for the provision of houses, but that other agents within our society, including individuals themselves, must be enabled by legislative and other measures to provide housing. The State must create the conditions for access to adequate housing for people at all economic levels of our society. State policy dealing with housing must therefore take account of different economic levels in our society.’

Indeed it would be futile if the state were to fulfil its obligations in respect of housing, but private persons or entities could frustrate it.

[83]   Section 26(3) should be interpreted against this background and close attention paid to the provision itself. The English text provides:

‘No one may be evicted ...’

‘No one’ is a term of wide and unqualified meaning. While some legislation applies only to specific persons – like squatters or hired workers – this constitutional provision applies to all persons. This includes private tenants. The statement to the opposite effect by Flemming DJP in Betta Eiendomme (Pty) Ltd v Ekple-Epoh 2000 (4) SA 468 (W) at 472 J, para 7.2, is therefore mistaken; similarly the statements in para 7.3.

[84]   The words ‘their home’ in s 26(3) of the Constitution carry a similarly broad and indefinite meaning. Therefore, in light of the Grootboom decision and the conclusions I have come to above, it includes a house, dwelling, townhouse, flat, room, etc, in an urban area which a private person or entity rents from another.

[85]   Read in this way, the first prohibition of s 26(3) of the Constitution does not involve anything beyond what the common law already provides: no one, not even an owner or landlord, may at his or her discretion evict a tenant from a residence or premises, even after lawful cancellation (see Nino Bonino v De Lange 1906 TS 120 at 122, 125).

[86]   The only new thing in s 26(3) of the Constitution, in relation to evictions in the case of private leases, is the requirement that the court take into account all relevant circumstances before making an eviction order. In the common law, no such requirement applies: if the lawful termination of the lease, whether by lapse of time or cancellation due to a breach of contract by the lessee, is proven, the court has no power to refuse an eviction order (see Human v Rieseberg 1922 TPD 157 especially at 163–166, approved in Oatarian Properties (Pty) Ltd v Maroun 1973 (3) SA 779 (A) at 785). Whether a court has the power to suspend the execution of an eviction order for a period of time is open to doubt (see Potgieter and Another v Van der Merwe 1949 (1) SA 361 (A) at 373–374; Evans v Schoeman, NO 1949 (1) SA 571 (A) at 580; Bhyat’s Departmental Store (Pty) Ltd v Dorklerk Investments (Pty) Ltd 1975 (4) SA 881 (A) at 886F–887A).

[87]   What do the words ‘all the relevant circumstances’ mean in relation to eviction orders after the termination of a lease? The values of good faith, reasonableness and fairness, and contractual justice would be surrendered if it were held that summary eviction orders, without any exception and without consideration of their humaneness, must follow the lawful cancellation or expiration of a lease. In my view, reasonableness and fairness require, in appropriate cases, that the court can at least suspend the execution of the eviction order for a reasonable period of time. In the court a quo it was ordered that Brisley must vacate the property in question within 10 days of the service of the court order upon her. It was not argued that this period of time was unreasonable; in light of the evidence regarding the ready availability of alternative housing and of moving contractors, this also seems to me to be a fair order.

I would dismiss the appeal with costs.

CAMERON JA (in its original English):

[88]   I have had the benefit of reading the judgment of Olivier JA, as well as that of Harms, Streicher and Brand JJA (‘the joint judgment’). I concur in the joint judgment, and wish to add some observations. All law now enforced in South Africa and applied by the courts derives its force from the Constitution. All law is therefore subject to constitutional control, and all law inconsistent with the Constitution is invalid. That includes the common law of contract, which is subject to the supreme law of the Constitution. The Bill of Rights applies to all law, and binds the judiciary no less than the legislature, the executive and all organs of state. In addition, the Constitution requires the courts, when developing the common law of contract, to promote the spirit, purport and objects of the Bill of Rights.

[89]   These propositions, if they ever were controversial, are no longer so. They derive from the provisions of the Constitution itself,[24] as the Constitutional Court has interpreted and applied them.[25] They bear on this case. In it, the appellant asks this court to reverse the doctrine that contracting parties may validly agree in writing to an enumeration of their rights, duties and powers in relation to the subject matter of a contract, which they may alter only by again resorting to writing. This court nearly four decades ago upheld the validity of such clauses.[26] It did so after some years of academic and judicial controversy, and after full argument, which canvassed the opposing contentions. Its decision expressly considered the paradox at the core of such provisions: that they limit contractual freedom, but do so by the prior design and agreement of the parties themselves,[27] in the exercise of their contractual freedom, and in order to enhance certainty in their future dealings and to minimise disputes between them.

[90]   The appellant’s attack invites us to reconsider that decision. We are obliged to do so in the light of the Constitution and of our ‘general obligation’, which is not purely discretionary,[28] to develop the common law in the light of fundamental constitutional values. For the reasons the joint judgment gives, I do not consider that the attack can or should succeed. The Shifren decision represented a doctrinal and policy choice which, on balance, was sound. Apart from the fact of precedent and weighty considerations of commercial reliance and social certainty, that choice in itself remains sound four decades later. Constitutional considerations of equality do not detract from it. On the contrary, they seem to me to enhance it. As the joint judgment observes (para 7), it is fallacious to suggest that insistence on only written alterations to a contractual regimen necessarily protects the strong at the expense of the weak. In many situations the reverse is likely to be true. And where a contracting party, strong or weak, seeks to invoke the writing-only requirement in deceit or to attain fraud, the courts will not permit it to do so.[29]

[91]   The jurisprudence of this court has already established that, in addition to the fraud exception, there may be circumstances in which an agreement, unobjectionable in itself, will not be enforced because the object it seeks to achieve is contrary to public policy.[30] Public policy in any event nullifies agreements offensive in themselves – a doctrine of very considerable antiquity.[31] In its modern guise, ‘public policy’ is now rooted in our Constitution and the fundamental values it enshrines. These include human dignity, the achievement of equality and the advancement of human rights and freedoms, non-racialism and non-sexism.[32]

[92]   It is not difficult to envisage situations in which contracts that offend these fundamentals of our new social compact will be struck down as offensive to public policy. They will be struck down because the Constitution requires it, and the values it enshrines will guide the courts in doing so. The decisions of this court that proclaim that the limits of contractual sanctity lie at the borders of public policy will therefore receive enhanced force and clarity in the light of the Constitution and the values embodied in the Bill of Rights.

[93]   I share the misgivings the joint judgment expresses about over-hasty or unreflective importation into the field of contract law of the concept of ‘boni mores’. The ‘legal convictions of the community’ – a concept open to misinterpretation and misapplication – is better replaced, as the Constitutional Court itself has suggested, by the ‘appropriate norms of the objective value system embodied in the Constitution’.[33] What is evident is that neither the Constitution nor the value system it embodies give the courts a general jurisdiction to invalidate contracts on the basis of judicially perceived notions of unjustness or to determine their enforceability on the basis of imprecise notions of good faith.[34]

[94]   On the contrary, the Constitution’s values of dignity and equality and freedom require that the courts approach their task of striking down contracts or declining to enforce them with perceptive restraint.[35] One of the reasons, as Davis J has pointed out,[36] is that contractual autonomy is part of freedom. Shorn of its obscene excesses,[37] contractual autonomy informs also the constitutional value of dignity:

‘If we look at the law simply from the point of view of the persons on whom its duties are imposed, and reduce all other aspects of it to the status of more or less elaborate conditions in which duties fall on them, we treat as something merely subordinate, elements which are at least as characteristic of law and as valuable to society as duty. Rules conferring private powers must, if they are to be understood, be looked at from the point of view of those who exercise them. They appear then as an additional element introduced by the law into social life over and above that of coercive control. This is so because possession of these legal powers makes of the private citizen, who, if there were no such rules, would be a mere duty-bearer, a private legislator. He is made competent to determine the course of the law within the sphere of his contracts, trusts, wills, and other structures of rights and duties which he is enabled to build.’[38]

[95]   The Constitution requires that its values be employed to achieve a careful balance between the unacceptable excesses of contractual ‘freedom’, and securing a framework within which the ability to contract enhances rather than diminishes our self-respect and dignity. The issues in the present appeal do not imperil that balance.

E CAMERON

JUDGE OF APPEAL

---